Reflections on the outsourcing decision-making

process and the use of outsourcing from a

management accounting perspective

Lars Br̊ad Nielsen

PhD dissertation

Accounting Research Centre

Department of Business Studies

Aarhus School of Business, Aarhus University

Thesis advisor: Hanne Nørreklit

October, 2010

Table of contents

Acknowledgements ii

Dansk resumé (Danish summary) iv

Introduction and summary vii

Chapter 1 1A retrospective on management accounting�s role in strategic outsourcing

decision-making

Chapter 2 75The outsourcing decision-making process: Two extensions to the

management accounting and strategy toolkit

Chapter 3 128The impact of outsourcing on investments in �rm-speci�c human capital

under di¤erent contract regimes

i

Acknowledgements

This dissertation marks the end of my PhD studies at the Aarhus School of Busi-

ness, Aarhus University, and was written in the period from August 2007 to August

2010. Admittedly, I have been looking forward to writing these acknowledgements

for a long time since the process of writing a dissertation would not have been as

rewarding without the direct and indirect support of several people and institutions.

I would like to thank the Aarhus School of Business and the Accounting Research

Centre for providing me with the administrative and �nancial support to complete

my work. Among the support sta¤, Berit Jensen and Didde Trolle deserve special

thanks for proofreading my work �often with very short notice. Also, I appreciate

the help of Ingrid Lautrup who time and again has assisted me with the adminis-

trative paperwork, thus making my life as a PhD student somewhat easier.

In particular, I am grateful to my thesis advisor Hanne Nørreklit who has shown

great interest in my project from the very beginning. Without your never-ending

support, knowledge, and inspiration, this thesis would have been much less than it

is today. I feel that the process of writing this dissertation has been quite a journey

for both of us and has developed into a really good friendship along the way. We

have shared thoughts on almost every aspect of my work, and although you have

had many other projects to attend to, you have always taken the time needed to

help me overcome the next obstacle. I feel very privileged to have been your PhD

student and to call you my friend.

I also owe a debt of gratitude to Hans Frimor for introducing me to the �won-

ders�of agency theory and for providing me with the possibility of attending two

interesting conferences on analytical accounting in Copenhagen. Indeed, your skill-

ful insights and guidance helped shape my article on �rm-speci�c human capital.

My stay at Fuqua School of Business, Duke University, was a great experience for

me from both an academic and a personal perspective, and I would like to thank

the many people with whom I interacted. First, I am thankful to Børge Obel and

Richard Burton for helping me get in contact with the Accounting Group, and to

Richard Burton in particular for making me feel very welcome. Second, I owe a

debt of gratitude to Qi Chen for letting me attend an excellent course on analytical

accounting and for fruitful discussions on my research project. I appreciate the fun

times we had on several occasions in both North Carolina and Copenhagen and I

sincerely hope to visit you again in the near future.

ii

On a personal level, I made a number of really good friends at Duke Univer-

sity. Isabel Verniers, without you things there would have been much more trivial.

Although we were both constantly a bit stressed, we had a lot of fun running and

memorable talks on how to improve our happiness. Bart Boersma and Sara Morey,

needless to say my social life improved considerably after I met the two of you. I

really enjoyed your company and our many social events �not the least the Duke

basketball games. And to my Italian friend, Gianfranco Siciliano: besides the amaz-

ing dinners that you prepared, I appreciate the many interesting talks we had on

accounting theory along with anything else worth discussing. I hope to see all of

you again soon.

At Aarhus School of Business, I thank my colleagues and fellow PhD students for

providing a friendly work environment. Particularly, I thank my two good friends

and marathon companions, Abhijit B. Bendre and Stig V. Møller, for many enjoy-

able moments at the school as well as on the forest trails around Aarhus. Also, I

am grateful to Sudarshan P. Pillalamarri for o¤ering his ear on several occasions

through the ups and downs of being a PhD student, and to Morten Jacobsen for

helping me plan my teaching. Moreover, I would like to thank my o¢ ce mate during

the �rst year, Pernille Jessen, for good company and for many fun discussions back

in the early days. Someday, I am sure you will come to understand that I was right

about most things! And �nally, to a very special girl that continues to mean a lot to

me, Dominyka Sakalauskaité (Dominga): although things did not work out, I would

not have missed the last year with you for anything. I truly hope we can remain the

best of friends.

But of all, I would like to thank my mom and dad for all your love and support. All

through my education, you have encouraged me to do my best while also trying to

make me not take things too seriously when they did not work out the way I had

hoped for. I have not always succeeded in the latter �sometimes causing quite a bit

of frustration and stress �yet you have always been there to gently help me move

on when I most needed it. Thus, it goes without saying that I dedicate this thesis

to the two of you.

Lars Bråd Nielsen

Aarhus, October, 2010

iii

Dansk resumé (Danish summary)

Denne afhandling består af tre selvstændige papirer, der hver for sig beskæftiger

sig med outsourcing beslutningstagning samt brugen af outsourcing som strate-

gisk værktøj. Første papir bærer titlen �A retrospective on management account-

ing�s role in strategic outsourcing decision-making� og er en historisk analyse af

den måde, hvorpå strategisk outsourcing beslutningstagning er blevet behandlet i

økonomistyrings- henholdsvis strategilitteraturen. I særdeleshed søges det afdækket,

i hvilket omfang beslutningsmodellerne fra den praktisk orienterede del af økono-

mistyringslitteraturen over tid har været i stand til at bibringe information, som

har kunnet understøtte outsourcing beslutningstagning inden for rammerne af de

stadigt mere so�stikerede virksomhedsteorier udviklet i strategilitteraturen. Under-

søgelsen er funderet i videnskabsteori og bygger på en analyse af den underliggende

ontologi og epistemologi indenfor hvert litteraturområde for herigennem at klarlægge

de respektive strømninger. I så henseende påpeger analysen, at fundamentale æn-

dringer i strategilitteraturen har skabt et gab mellem den information, der betragtes

som vigtig for beslutningstagning i strategilitteraturen, og den information som nu-

værende økonomistyringsmodeller er i stand til at levere. Hvor økonomistyring til

stadighed har bevaret fokus på at afdække og indsamle information, der kan valid-

eres på baggrund af kvantitative empiriske observationer, har strategilitteraturen

�yttet fokus i retning af mere subtile kvalitative faktorer, som vanskeligt lader sig

måle. Følgelig konkluderer artiklen, at økonomistyring i sin nuværende forfatning

har mistet betydelig relevans i forhold til strategisk outsourcing beslutningstagn-

ing, hvorfor det diskuteres, hvorledes fremtidig forskning inden for de respektive

akademiske felter kan hjælpe med til at reducere gabet. Således foreslås det, at

et passende første skridt kunne være at adressere de underliggende ontologiske og

epistemologiske forskelle. Dette vil hjælpe med til at klarlægge, hvilke tiltag, der

bør iværksættes, for at de to felter kan komme hinanden i møde og i fællesskab ud-

vikle værktøjer, der kan forbedre strategisk outsourcing beslutningstagning i praksis.

Andet papir bærer titlen �The outsourcing decision-making process: Two exten-

sions to the management accounting and strategy toolkit� og undersøger, hvorledes

økonomistyringsinformation i praksis interagerer med strategi i forbindelse med out-

sourcing beslutningstagning. Med henblik på at muliggøre en sådan undersøgelse

udvikles indledningsvist en konceptuel ramme, der klarlægger hvilke opgaver, som

hidrører under strategi henholdsvis regnskab, samt hvordan disse opgaver på det

generelle plan er relateret til hinanden undervejs i outsourcing beslutningsprocessen.

Formålet er at bane vejen for en kvali�ceret diskussion af, hvorledes strategi og

iv

økonomistyring i fællesskab kan understøtte outsourcing beslutninger, der både kan

realiseres og er pro�table. Med udgangspunkt i den konceptuelle ramme analy-

seres efterfølgende outsourcing beslutningsprocessen i to case-virksomheder, der

begge er beskæftiget inden for fremstilling. I så henseende konkretiserer analy-

serne, hvordan strategi framer outsourcing beslutningsprocessen i overensstemmelse

med virksomhedernes respektive situationer, samt hvordan denne framing former

den økonomistyringsinformation, der konstrueres og anvendes i forbindelse med

beslutningstagning. Følgelig illustrerer de to casestudier, at både strategi og økono-

mistyring spiller en afgørende rolle i forbindelse med outsourcing beslutningstagning,

samt at de er tæt forbundne. I forhold til (komplekse) outsourcing beslutninger, er

det således i bedste fald simpli�cerende og i værste fald vildledende at studere kon-

struktion og anvendelse af økonomistyringsinformation uafhængigt af den strategiske

ramme. I relation til praksis kan analyserne derfor betragtes som en udvidelse af

den fælles strategi- og økonomistyringsværktøjskasse, idet de tjener som inspiration

til, hvordan outsourcing beslutningsprocessen med fordel kan struktureres under

givne omstændigheder. Ydermere bidrager artiklen til litteraturen om beslutning-

steori, hvor det til stadighed diskuteres, hvorvidt beslutningstagning bør studeres

som et mekanisk system frem for en organisk proces baseret på menneskelig intu-

ition. Særligt peger artiklens analyser på, at begge tilgange �ndes i praksis.

Tredje papir bærer titlen �The impact of outsourcing on investments in �rm-speci�c

human capital under di¤erent contract regimes�og belyser, hvordan outsourcing kan

bruges som redskab til at inducere investeringer i virksomhedsspeci�k humankapi-

tal. Følgelig udvikles en to-periode principal-agent model med det formål at vise,

hvorledes en trussel om afskedigelse (dvs. outsourcing af arbejdsopgaver til tred-

jepart) kan anvendes af en virksomhed som instrument til at få en medarbejder til

at foretage investeringer i humankapital (læring), idet der skabes eksplicitte karri-

erebekymringer (career concens). Analysen foretages under såvel kortsigtede som

langsigtede kontraktforhold.

I det tilfælde hvor langsigtede kontrakter kan etableres (fuld binding), vises det,

at outsourcing �eksibilitet ingenlunde er nødvendig for virksomheden for at induc-

ere en investering i virksomhedsspeci�k humankapital. Årsagen er, at medarbe-

jderens incitament kan reguleres ved, at virksomheden forpligtiger sig til at ud-

betale en tilstrækkelig høj medarbejderkompensation, efter investeringen er fore-

taget. Dog kan det alligevel være formålstjenligt for virksomheden at introducere

et outsourcing alternativ grundet to potentielt værdiskabende e¤ekter: en realop-

tionse¤ekt samt en adfærdse¤ekt. Realoptionse¤ekten forøger virksomhedens for-

v

ventede pro�t, idet adgang til tredjepart reducerer afhængigheden af intern læring,

såfremt virksomheden kan bestemme medarbejderens fratrædelsesgodtgørelse uden

større begrænsninger. Adfærdse¤ekten reducerer de incitamentsomkostninger, der

er forbundet med at få medarbejderen til at foretage en bestemt investering i virk-

somhedsspeci�k humankapital. Bevæggrunden er, at introduktion af et outsourcing

alternativ automatisk bibringer medarbejderen incitament til at investere for heri-

gennem at gøre sig selv attraktiv og undgå afskedigelse.

I det tilfælde hvor kun kortsigtede kontrakter kan etableres, vises det, hvordan

fraværet af outsourcing �eksibilitet betyder, at medarbejderen ikke har noget in-

citament til at investere i virksomhedsspeci�k humankapital. Baggrunden er, at

virksomheden ikke troværdigt kan binde sig til fremtidig kontrakter, hvorfor denne

altid vil udnytte, at medarbejderens investering er virksomhedsspeci�k. Følgelig vil

virksomheden altid gennemtvinge en lønnedgang, efter investeringen er foretaget,

hvorved investeringen mister sin værdi for medarbejderen. I det tilfælde hvor læring

er produktivt, kan dette være skadeligt for virksomheden. Således demonstreres det,

hvordan virksomheden kan pro�tere ved introduktion af et outsourcing alternativ, da

der skabes karrierebekymringer hos medarbejderen. Som resultat heraf investerer

denne i virksomhedsspeci�k humankapital for at signalere attraktivitet og undgå

afskedigelse. Årsagen er, at det stadigt er indtægtsgivende for medarbejderen at

undgå afskedigelse til trods for den lønnedgang, som virksomheden gennemtvinger.

I modsætning til tilfældet med langsigtede kontrakter er det derfor helt nødvendigt

for virksomheden at introducere outsourcing �eksibilitet under kortsigtede kontrak-

ter, hvis virksomheden ønsker, at medarbejderen skal investere i virksomhedsspeci�k

humankapital.

vi

Introduction and summary

Over the last two decades, globalization has intensi�ed the competition in many in-

dustries with competitors existing in both home and international markets (McIvor

2005). This increase in competitive rivalry has reduced the pro�ts of many com-

panies and compelled them to search for new ways to reduce costs and improve

customer responsiveness and quality. For instance, these impacts have been very

pronounced in the European airline industry, where competition from low cost air-

lines such as Easyjet and Ryanair, partly driven by deregulation, has forced many

established European airlines including British Airways, Swiss-Air and Air France

to embark upon radical restructuring programmes in order to eliminate ine¢ ciencies

and reduce costs (Franke 2004; Hanlon 1999). Likewise, these trends can be observed

in several consumer goods industries. With consumers becoming more sophisticated

and powerful, they will no longer settle for whatever companies are o¤ering. In turn,

this forces companies to be more responsive to changing customer needs and special

preferences. As an example of this, it is now possible on Nike�s website �nike.com

�to virtually customize shoes for the individual buyer before manufacturing.

The drive for responsiveness and cost reductions has led many companies to focus

increasingly on a limited number of key areas while looking for ways to outsource

many, less important, activities currently carried out in-house. Speci�cally, out-

sourcing involves re-drawing the boundaries between the company and its supply

base. Although, the term has become in vogue in later years, companies have al-

ways made decisions on determining the boundaries of the organization. Yet, as

the concept continues to develop it remains a fascination for both academics and

practitioners.

Traditionally, outsourcing has been related to the transfer of business support

functions such as cleaning or security to external suppliers in order to obtain higher

levels of performance at a lower cost with relatively little upheaval for the company.

Yet, over time, it has progressed from involving only peripheral business activities to-

wards more critical business activities that contribute to the company�s competitive

advantage. Hence, today outsourcing also embraces major organizational changes

that involve dismantling the company�s traditional structure, transferring sta¤to ex-

ternal suppliers, re-de�ning sta¤ terms and conditions and altering the expectations

of the employees that remain within the outsourcing company. For example, many

automotive manufacturers, such as Volkswagen, have been adopting modular-type

arrangements with their suppliers in order to achieve �exibility and responsiveness

(McIvor 2005). A modular product architecture is a special form of product design

vii

that uses standardized interfaces between components to create a �exible product

architecture (Sanchez and Mahoney 1996; Schilling and Steensma 2001). Flexibility

is obtained because product variations can be leveraged by substituting di¤erent

modular components into the product architecture without having to redesign other

components. Hence, at Volkwagen�s truck and bus plant in Brazil, parts are not

just bought from outside (Simionian 1996). Instead, the company has asked the

component makers to design and develop entire sub-assemblies, such as chassis or

engines, that can be �tted directly on the production line.

Since the trend in outsourcing �very much driven by evolving strategic concepts

� is towards ever more complex activities being contracted out, it is essential to

understand how management accounting can continue to function as an attention-

directing decision-support tool when it increasingly requires dealing with strategic

issues that are di¢ cult to quantify and measure. We should thus expect to �nd

extensive academic research on the outsourcing decision-making process within the

practically oriented part of the management accounting literature. However, so far,

the literature has mainly focused on studying outsourcing relations while only minor

attention has been given to the actual decision-making process (see e.g. Widener and

Selto 1999; Seal et al. 1999; van der Meer-Kooistra and Vosselman 2000; Tomkins

2001; Lang�eld-Smith and Smith 2003; Nicholson et al. 2006; Thrane and Hald

2006). Furthermore, even within the sporadic empirical work (see e.g. Sartorius

and Kirsten 2005; Lamminmaki 2008) that currently exists on outsourcing decision-

making, the role of strategy in structuring the decision-making process as well as its

a¤ect on the production and use of accounting information has been neglected. In

fact, this lack of scholarly work seems to be rooted in a much more general prob-

lem within the �eld of management accounting where research tends to shy away

from dealing with the complexity of strategic issues (Lang�eld-Smith 2008). Most

noticeably, this is observed in relation to the strategic cost management/accounting

literature:

By 2000, there was a �fteen-year history of great "beginnings", "pilot"

projects, and "cameo" appearances for SCM [Strategic Cost Manage-

ment], but not much more. It has been a great topic on the lecture circuit

and in cost management symposia. In military parlance, "it briefs well"!

But the topics had not been gaining traction in mainstream academe or

in the corporate world... (Shank 2007, p. 359).

In consequence of this weak understanding of the interaction between strategy and

accounting, the outsourcing decision-making process is still often portrayed in the

viii

literature as a simpli�ed comparison between internal costs and quotations prices,

while the more subtle strategic aspects are either highly reduced or left for a side

note (see e.g. Horngren et al. 2005, pp. 255-258).

Intrigued by the growing importance of outsourcing and puzzled by the lack of

research on the interaction between strategy and management accounting, this PhD

thesis sets out to study both the outsourcing decision-making process and the use

of outsourcing from a management accounting perspective. Speci�cally, drawing

on philosophy (the theory of science), case studies, and agency theory, the thesis

investigates from a multitude of angles how management accounting interacts with

strategy in the outsourcing decision-making process and how outsourcing can be

used as a tool for cleaning up the organization as well as an incentive mechanism for

regulating employee behavior. The linkages among the main elements in the thesis

are illustrated in �gure 1 below.

Management accounting

Strategy

The outsourcing decision-making process

The use of outsourcing…

… as a strategic tool for clean-up

… as a strategic tool for regulating behavior (incentives)

1.1 Interaction

1.2

1.3

2

3

4

Figure 1: Outline of the elements in the thesis.

The thesis consists of three chapters. The �rst chapter entitled "A retrospective on

management accounting�s role in strategic outsourcing decision-making" is a histor-

ical investigation of strategic outsourcing decision-making in both the management

accounting and strategy literature. Speci�cally, the chapter examines the extent to

which decision models in the practically oriented part of the management accounting

literature have, over time, been able to provide vital and supportive information for

outsourcing decision-making within the increasingly sophisticated frameworks de-

veloped in the strategy literature. The study is rooted in the theory of science and

relies on an analysis at the level of ontology and epistemology to critically review

the trends within both literature streams. In particular, the analysis reveals that

changes in the underlying fundamentals of the strategy literature have created a gap

ix

between the information deemed important for outsourcing decision-making within

the strategic frameworks, and the information which accounting models are able

to supply. While management accounting has remained focused on uncovering and

collecting information that can be credibly gauged through empirical observation,

strategy has moved on to consider the importance of more subtle and qualitative

factors. Accordingly, the chapter concludes that accounting models have lost signif-

icant relevance for strategic outsourcing decision-making. It thus re�ects upon how

future research can help reduce the gap between the two academic �elds. To this

end, it is suggested that an appropriate starting point is to address the di¤erences

in the underlying ontology and epistemology. Only by doing this will it be possible

to identify which compromises are required by both �elds in order to create a mutu-

ally inspiring and bene�cial relationship that can help advance strategic outsourcing

decision-making in practice. Referring to �gure 1, the concern of the �rst chapter is

thus mainly with the linkages described by 1, although the historical nature of the

analysis touches upon the changing use of outsourcing in practice throughout the

history (linkage 2).

Building on the �ndings in the �rst chapter, the second chapter "The outsourc-

ing decision-making process: Two extensions to the management accounting and

strategy toolkit" studies how accounting information interacts with strategy in prac-

tice when deciding on outsourcing. To make this possible, a conceptual framework

is initially developed to distinguish between the area of strategy and management

accounting, and to investigate the interaction between the two areas in outsourc-

ing decision-making. The purpose is to pave the way for a quali�ed discussion

of how strategy and management accounting in combination can help ensure that

the decisions made are both feasible and pro�table. Subsequently, the outsourcing

decision-making process is analyzed in two case companies operating in manufactur-

ing. The cases reveal how strategy frames the outsourcing decision-making process

in accordance with the company�s contextual circumstances and how this framing

shapes the accounting information produced and used. Consequently, the cases show

that strategy as well as management accounting plays a crucial role in outsourcing

decision-making and that they are tightly integrated. Separating the study of how

accounting information is produced and used, on the one hand, from the strategic

context, on the other hand, is thus at best simplistic and at worst deceptive in

relation to (complex) outsourcing decision-making . For practice, the results are

extensions to the management accounting and strategy toolkit since the cases can

serve as inspiration for how to structure the outsourcing decision-making process.

x

Also, the chapter adds to the part of the academic literature on decision theory de-

bating the appropriateness of studying the decision-making process as a mechanical

system instead of an organic process based on human intuition. Speci�cally, the

results indicate that both standpoints are legitimate and can be found in practice.

In terms of �gure 1, the focus of the second chapter is thus again on the interaction

between strategy and accounting in the outsourcing decision-making process (the

linkages described by 1) �however, this time in a practical context. Furthermore,

the chapter re�ects upon how di¤erent structures of the decision-making process

might prove more capable for certain types of organizational clean-up (linkage 2

and 3).

The third chapter entitled "The impact of outsourcing on investments in �rm-speci�c

human capital under di¤erent contract regimes" studies how outsourcing can be used

as a tool to induce the employees in a company to invest in �rm-speci�c human cap-

ital. Speci�cally, the chapter develops a two-period agency model to show how

the threat of layo¤ (outsourcing of job tasks to a third-party supplier) can help a

company to trigger or ease employee investments in �rm-speci�c human capital by

creating explicit career concerns for the individual employee. Results are provided

under long-term as well as short-term contracting regimes.

Under long-term contracting (full commitment), it is shown that introducing

outsourcing �exibility is not necessary to induce a certain investment in �rm-speci�c

human capital since the employee�s incentive can be regulated (strengthened or

weakened) through commitment to a given compensation ex post the investment.

Still, it is demonstrated that it might be alluring for the company to bring about the

outsourcing �exibility as it comprises two potentially value creating e¤ects; a real

option e¤ect and a behavioral e¤ect. The real option e¤ect increases the expected

pro�t since the access to a third-party supplier reduces the dependence on in-house

learning if the company can choose the severance fee paid to the worker at dismissal

without constraints. The behavioral e¤ect reduces the incentive cost of making the

employee undertake a certain investment in �rm-speci�c human capital as the mere

introduction of an outsourcing alternative naturally incentivizes him to invest in

order to signal attractiveness and avoid layo¤.

Under short-term contracting, it is shown that the absence of outsourcing �exi-

bility leaves the employee with no incentive to invest in �rm-speci�c human capital.

Speci�cally, since the company cannot credibly commit to any future contracts,

it will always exploit the �rm-speci�city of the worker�s investment and enforce a

salary cut ex post the competence improvement that reduces the value of the �rm-

xi

speci�c investment to zero. As such, this can be harmful to the company if learning

is productive. It is thus demonstrated how the company can bene�t from intro-

ducing outsourcing �exibility into the setting to create explicit career concerns for

the employee and make him undertake long-term actions to signal attractiveness

and maintain production in-house. That is, although future compensation will be

adjusted by the company to o¤set the e¤ect of �rm-speci�c investments under short-

term contracting, there are still rents to be collected by the worker if not dismissed.

Consequently, compared to full commitment, the introduction of outsourcing �ex-

ibility is absolutely necessary to encourage a positive investment in �rm-speci�c

human capital under short-term contracting.Referring to �gure 1, the third chapter thus revolves around the outsourcing

decision-making process as a constrained pro�t maximization problem (linkage 2)and the use of outsourcing as a strategic tool for regulating employee behavior(linkage 4).

References

Franke, M. (2004). Competition between network carriers and low-cost carriers-retreat battle or breakthrough to a new level of e¢ ciency? Journal of AirTransport Management 10 (1), 15�21.

Hanlon, P. (1999). Global Airlines. Oxford: Elsevier Science.

Horngren, C. T., G. L. Sundem, and W. O. Stratton (2005). Introduction toManagement Accounting (13th ed.). New Jersey: Pearson Prentice Hall.

Lamminmaki, D. (2008). Accounting and the management of outsourcing: Anempirical study in the hotel industry.Management Accounting Research 19 (2),163�181.

Lang�eld-Smith, K. (2008). Strategic management accounting: how far have wecome in 25 years? Accounting, Auditing & Accountability Journal 21, 204�228.

Lang�eld-Smith, K. and D. Smith (2003). Management control systems and trustin outsourcing relationships. Management Accounting Research 14 (3), 281�307.

McIvor, R. (2005). The Outsourcing Process: Strategies for Evaluation and Man-agement. Cambridge: Cambridge University Press.

Nicholson, B., J. Jones, and S. Espenlaub (2006). Transaction costs and controlof outsourced accounting: Case evidence from India. Management AccountingResearch 17 (3), 238�258.

Sanchez, R. and J. T. Mahoney (1996). Modularity, �exibility, and knowledgemanagement in product and organization design. Strategic Management Jour-nal 17 (Special Issue: Winter), 63�76.

xii

Sartorius, K. and J. Kirsten (2005). The boundaries of the �rm: why dosugar producers outsource sugarcane production.Management Accounting Re-search 16 (1), 81�99.

Schilling, M. A. and H. K. Steensma (2001). The use of modular organizationalforms: An industry-level analysis. Academy of Management Journal 44 (6),1149�1168.

Seal, W., J. Cullen, A. Dunlop, T. Berry, and M. Ahmed (1999). Enacting aEuropean supply chain: a case study on the role of management accounting.Management Accounting Research 10 (3), 303�322.

Shank, J. K. (2007). Contemporary Issues in Management Accounting, Chapter16: Strategic cost management: upsizing, downsizing, and right(?) sizing, pp.355�379. Oxford: Oxford University Press.

Simionian, H. (1996, March). Prophet of the production line. Financial Times,19. London.

Thrane, S. and K. Hald (2006). The emergence of boundaries and accounting insupply �elds: The dynamics of integration and fragmentation. ManagementAccounting Research 17 (3), 288�314.

Tomkins, C. (2001). Interdependencies, trust and information in relationships,alliances and networks. Accounting, Organizations and Society 26 (2), 161�191.

van der Meer-Kooistra, J. and E. G. J. Vosselman (2000). Management con-trol of inter�rm transactional relationships: the case of industrial renovationand maintenance - design and methods. Accounting, Organizations and Soci-ety 25 (1), 51�77.

Widener, S. K. and F. H. Selto (1999). Management control systems and bound-aries of the �rm: Why do �rms outsource internal auditing activities? Journalof Management Accounting Research 11, 45�73.

xiii

Chapter 1

A retrospective on management accounting’s role

in strategic outsourcing decision-making

A retrospective on management accounting�s role

in strategic outsourcing decision-making

Lars Bråd Nielsen�

July 27th, 2010

Abstract

The question of what to outsource is one of the most important decisions acompany can face as it relates to the fundamental issue of why �rms cometo exist. Motivated by this signi�cance, the interest of the present paperis with strategic outsourcing decision-making and its historical treatment inboth the management accounting and strategy literature. Speci�cally, thepaper investigates the extent to which decision models in the management ac-counting literature have, over time, been able to provide vital and supportiveinformation for strategic outsourcing decision-making within the increasinglysophisticated frameworks developed in the strategy literature. The study isrooted in the theory of science and relies on an analysis at the level of on-tology and epistemology to critically review the trends within both literaturestreams. Particularly, the analysis reveals that changes in the underlying fun-damentals of the strategy literature have created a gap between the informa-tion deemed important for outsourcing decision-making within the strategicframeworks, and the information which accounting models are able to sup-ply. While management accounting has remained focused on uncovering andcollecting information that can be credibly gauged through empirical obser-vation, strategy has moved on to consider the importance of more subtle andqualitative factors. Accordingly, the paper concludes that accounting modelshave lost signi�cant relevance for strategic outsourcing decision-making. Itthus re�ects upon how future research can help reduce the gap between thetwo academic �elds. To this end, it is suggested that an appropriate startingpoint is to address the di¤erences in the underlying ontology and epistemol-ogy. Only by doing this will it be possible to identify which compromises arerequired by both �elds in order to create a mutually inspiring and bene�cialrelationship that can help advance strategic outsourcing decision-making inpractice.

�Aarhus School of Business, Aarhus University, Fuglesangs Allé 4, DK-8210 Aarhus V., Den-mark. E-mail: labn@asb.dk.

1 Introduction

In recent years, much attention has been called to the changing role of management

accounting in decision-making. Traditionally, many organizations had an Account-

ing or Finance Department where accountants were sitting isolated from the rest of

the business, surfacing only to discuss the monthly accounting �gures with business

managers, if at all (Burns and Baldvinsdottir 2005). Management accountants were

regarded as experts in the preparation and interpretation of business information

for decision-making and control. They were viewed as independent and objective

assessors of the �nancial performance of di¤erent business functions (Hopper 1980),

expected to take on a passive role since the numbers were assumed to �speak for

themselves�.

Today, this picture has changed as the role of management accounting has be-

come more �consulting-based� (Siegel and Sorensen 1999; CIMA 2009). Rather

than being isolated from the rest of business, management accountants are today

expected to engage proactively in a variety of tasks including assessment of the �nan-

cial implication of operational decisions, risk assessment, and strategy formulation

(Scapens et al. 2003). Thus, they are transforming into what has become known in

the management accounting literature as hybrid accountants (Burns and Baldvins-

dottir 2005) �or business partners (CIMA 2009) �who need to combine accounting

knowledge with a detailed understanding of the business process in which they work

in order to assist strategic decision-making as well as day-to-day operations (Burns

and Baldvinsdottir 2005; Scapens et al. 2003; CIMA 2009).

However, while it is evident that management accountants of today have to adapt

their behavior to business, it remains unclear how they are supposed to do so. In

particular, the prevailing conception of business has changed over time as new theo-

ries of the �rm have seen the light of day in the strategy literature (e.g. Williamson

1975; Porter 1991; Barney 1991), thus adding to the complexity of the information

deemed necessary for optimal (strategic) decision-making. For this reason, it seems

surprising that so much attention in the management accounting literature has been

given to the changing role of the accountant while only little research has been done

on the extent to which current practically oriented accounting models can actually

satisfy the information needs raised by business.

The purpose of this article is thus to open a debate on whether accounting models

can actually satisfy managerial information demands. This is done by investigating

the extent to which decision models in the management accounting literature have,

historically, been able to provide supportive information for the strategic outsourc-

3

ing decision-making within the increasingly sophisticated frameworks developed in

the strategy literature. To assure an in-depth analysis, I have chosen to focus the in-

vestigation on outsourcing, where speci�c attention will be given to how outsourcing

decision-making in both the management accounting and strategy literature have

evolved over time. The choice of outsourcing is motivated by the fact that this topic

has been a central issue in both accounting and strategy for many years, just as it

continues to be one of the most fundamental questions that a company can face in

practice (Hätonën and Eriksson 2009). Furthermore, the outsourcing concept has

developed considerably over time as new approaches have been introduced within

strategy. This allows for a historical comparison of the shifting importance of ac-

counting in strategic decision-making.

In particular, this article �nds that the combination of changes in the business

environment and emerging ideas in the strategy literature have turned today�s deci-

sion to outsource into an often very complex strategic matter. Yet, the models in the

management accounting literature have not kept pace with this development. Where

early considerations, in the strategy literature, on outsourcing were very focused on

cost issues there was a natural coherence with cost accounting and thus an important

role for management accounting to play in relation to outsourcing decision-making.

Over time, however, outsourcing has become increasingly in�uenced by strategic

factors for which reason this coherence is no longer obvious. Even today the main-

stream treatment of make-or-buy in the management accounting literature is often

reduced to comparing alternatives on the basis of simple cost allocation, where only

a small part of the total indirect cost is re�ected in the cost of each alternative (see

e.g. Horngren et al. 2005, pp. 255-258). The reason is that accounting researchers

in general are afraid of compromising credibility in the calculations �partly due

to attitude and partly due to an inadequate accounting toolkit �for which reason

they shy away from estimating the impact of more profound strategic factors. As a

result, these aspects are not re�ected in the actual cost calculations but merely left

for a side note. The question is thus whether the management accounting literature

�being an ambitious discipline �can be satis�ed with this simplistic treatment as

ever more advanced strategic concepts di¤use from the strategy literature into prac-

tice. In this respect, the literature analysis in this article suggests that the simplistic

treatment of outsourcing only helps to cover up a severely underdeveloped area of

management accounting research that continues to reduce the value of accounting

information as decision support for outsourcing decision-making, since management

accounting models simply cannot supply the information deemed relevant for the

4

strategic frameworks being adopted.

Consequently, the contribution of this article is the revelation of a paradox within the

management accounting literature. Indeed, accounting researchers have acknowl-

edged the appearance of the hybrid accountant and the importance of combining

traditional accounting information with strategic issues when designing accounting

models for practical use. Yet, a literature study of one of the most fundamental

questions that a company can possibly face � the question of what to outsource

�reveals that management accounting research has surprisingly little of relevance

to o¤er for practice in general and for the hybrid accountant in particular. In a

broader perspective, the results of this paper thus serve as evidence to a point al-

ready made in Baldvinsdottir et al. (2010) that an unfortunate tendency has come

to mark management accounting research. A tendency where academic creativity

and e¤orts are only, and to a limited extent, concentrated around developing models

for the improvement of real world decision-making.

The remainder of the paper initially seeks to clarify outsourcing as a concept and the

interdependency between strategy and accounting in outsourcing decision-making.

Then, the methodology used for literature analysis is outlined. The intention is to

build a platform sophisticated enough to e¤ectively allow a paradigmatic study of

the historical developments in accounting and strategy research, yet simple enough

to require the reader to hold no philosophical background let alone in-depth under-

standing of philosophical technicalities. Based on this methodology, the historicaldevelopments that have marked the outsourcing literature in both areas are inves-

tigated. The investigation is divided into descriptions of historical eras and the

actual analysis is conducted in relation to each of these. Following this, the results

are discussed and suggestions are provided to how the practical oriented part of man-

agement accounting research on outsourcing decision-making might be restarted and

taken towards new frontiers to catch up with strategy research. The paper closes

with some ideas and re�ections on further research.

2 Conceptualizing outsourcing and outsourcing

decision-making

In this section, a short introduction to outsourcing as a concept follows �rst. Subse-

quently, the terms of strategy and management accounting are de�ned as they are

to be understood in this paper, and the interdependency between them in relation

5

to outsourcing decision-making is conceptualized.

2.1 Outsourcing: Meaning and scope

When studying outsourcing decision-making, it is appropriate to clarify the meaning

and scope of outsourcing as a concept. So what is outsourcing? Literally, the term

outsourcing is a made-up word meaning �outside resource using� (Arnold 2000;

Koppelmann 1996, p. 2). Outside insinuates value creation on the basis of something

that is not within the company while resource points to the need for a strategic focus

on what external resources to exploit. Finally, using emphasizes that it is not enough

to know about valuable external resources. Indeed, they must be used by and for

the company in order to reinforce the competitive position.

Often, and perhaps mostly in a manufacturing context, outsourcing is used in-

terchangeably with the term �make-or-buy�, which is normally associated with the

decision to manufacture a component internally or buy it from an external supplier.

McIvor (2005, p. 7) argues that this term is perhaps more appropriate as it implies

that there should be an evaluation of the suitability of either internal or external

supply whereas the term outsourcing suggests that the decision to use an external

supplier has already been made without any consideration of whether it is appro-

priate for the organization. Nevertheless, in this paper the two terms will be used

interchangeably.

Having clari�ed the basic meaning of outsourcing, it is equally appropriate to under-

stand the possible scope of make-or-buy. Indeed, and as I will return to in section

4 below, outsourcing is a discussion about the boundaries of the �rm, yet not all

boundary discussions are related to outsourcing. In general, the boundaries can

be moved in a number of ways that include conglomeration, horizontal integration,

vertical integration, and internal integration of supporting activities (Lonsdale and

Cox 2000). Conglomeration occurs when a �rm acquires or merges with other �rms

that are neither at di¤erent stages of the supply chain nor are competitors at the

same supply chain activity. Related to this but still di¤erent, horizontal integration

refers to the merging of two or more �rms that are operating in the same supply

chain (competitors). Essentially, none of these boundary expansion forms can be

regarded as outsourcing.

<Insert �gure 1 here>

Thus, in a strict sense, only vertical integration and internal integration of support

activities are to be classi�ed as outsourcing. Looking at �gure 1, vertical integration

6

occurs when the boundary of the �rm shifts along its primary supply chain �either

upstream or downstream. Moreover, internal integration of support activities cap-

tures the boundary change from outsourcing or internalizing supporting activities.

These activities belong to the secondary supply chains and are necessary for the

�rm to function, yet they are not a part of the supply chain that delivers the market

o¤erings. For the purpose of the present paper, no strict distinction between the

di¤erent boundary expansion forms are as such needed. Still, the distinction guides

and delimits the choice of papers analyzed so that focus will be kept on papers

mainly concerned with the strict form of outsourcing.

In relation to this, it is also important to make clear that several types of potential

contract relationships with varying sourcing control exist. These range from full

ownership to short-term contracting, and with intermediate contract forms such as

partnerships or long-term contracts in between (Quinn and Hilmer 1994). However,

in this paper, the interest is not with the details of the various types of outsourcing

arrangements, and so only two contract regimes given by make (absolute internal

control) and buy (absolute external control) will be considered.

2.2 The interdependency between strategy and accounting

in outsourcing decision-making

In order to examine outsourcing decision-making and help structure the selection of

literature, a conceptual framework that outlines the interdependency between strat-

egy and management accounting is needed. In general, Grant (2005, p. 21) notes

that several de�nitions of strategy exist, yet all of them revolve around the aspects of

where to compete and how to compete. Mintzberg, distinguishes between intended

strategies (the company leadership plans and intentions) and realized strategies (the

company actions actually taken) (Mintzberg 1978; Mintzberg and Waters 1985). As

such, all strategies can be compared on the basis of the �t between intentions and

realization. When the intended strategy turns out to be the strategy actually real-

ized, it is referred to as a deliberate strategy, although the match between intention

and realization is not always perfect. Mintzberg thus introduces the concept of an

emergent strategy to describe the situation when the actual strategic actions deviate

from the original intentions and plans or when neither initial intentions nor plans ex-

ist at all. Since the interest in this article is with planned decision-making, I will not

focus on this latter concept but instead concentrate on the idea of a deliberate strat-

egy where in the initial phase the company plans the strategic actions to be carried

out subsequently. In particular, the de�nition of a strategic plan found in Horngren

7

et al. (2005) is thus adopted, as this fully captures Mintzberg�s conception1 and the

two fundamental aspects pointed out in Grant (2005):

A strategic plan refers to setting the overall objectives for an organization

and outlining how they will be attained (Horngren et al. 2005, p. 299).

Keeping this de�nition in mind, Horngren et al. (2005, p. 11) point out that

decisions in general can be divided into two groups concerned with either (strategic)

planning or control, where the latter refers to the implementation of the (strategic)

plan and the use of feedback to attain objectives. Therefore, albeit both types

of decisions are potentially crucial, strategic decisions tend to involve issues more

important and fundamental to a company than is the case for decisions related to

control.

In continuation of this, Horngren et al. (2005, p. 11) de�ne decision-making as

"the purposeful choice from among a set of alternative courses of action designed to

achieve some objective." Anderson (1983) develops this even further by decomposing

the task of making a decision into �ve subtasks: (1) identifying the relevant goals;

(2) searching for alternative courses of action; (3) predicting the consequences of fol-

lowing each alternative; (4) evaluating each alternative in terms of its consequences

for goal achievement; and (5) selecting the best alternative for achieving the goal.2

Hence, focusing on strategic decisions, the use of Anderson�s (1983) decomposi-

tion of subtasks for decision-making shows that the de�nition of strategy essentially

covers subtasks 1, 2, and 5: identi�cation of goals, determination of decision alterna-

tives, and selection of the best alternative. However, to facilitate this �nal selection,

an evaluation of the alternatives and an assessment of the consequences are needed

�and this is where accounting comes into play. As such, Horngren et al. (2005)

de�ne management accounting as:

the branch of accounting that produces information for managers within

an organization. It is the process of identifying, measuring, accumu-

lating, analyzing, preparing, interpreting, and communicating informa-

tion that helps managers ful�ll organizational objectives (Horngren et al.

2005, p. 5).

1Notice that the most extreme version of a deliberate strategy is referred to as a planned strategy(Mintzberg and Waters 1985).

2Notice that I use the model of rational action to de�ne decision-making (Eisenhardt andZbaracki 1992). However, other frameworks of decision-making do exist. See for instance Nutt(1976).

8

Consequently, in the context of Anderson (1983), management accounting covers

subtasks 3 and 4: information provision and evaluation of alternatives.

With these general classi�cations in mind, outsourcing in this article is seen as

a sequence of decisions that can be divided into two overall phases; a fundamental

strategic decision (the outsourcing decision) and a number of decisions related to

control (implementation).

The �rst phase, the outsourcing decision, is conceptualized as an interdependent

decision-making process of steps that shifts back and forth between the domains of

strategy and management accounting. A process, where strategy initially has the ini-

tiative to "de�ne the rules of the game" and outline the overall decision alternatives

available. Based on these delimitations, management accounting subsequently takes

control to specify and evaluate the consequences of each decision-alternative relying

on proper information collection and accounting decision-models. The �ndings and

related advises are �nally reported back to strategy which weights the information

provided with other (strategic) factors to determine the optimal decision alternative.

Given the chosen course of action, outsourcing enters the second phase, imple-

mentation. During this phase, management accounting continues to play a sig-

ni�cant role in seeing through that the outsourcing arrangement is implemented

appropriately and that necessary adjustments are undertaken when needed.

Figure 2 below depicts all of these dynamics where focus, for the purpose of this

paper, will be kept on the initial phase encompassing the outsourcing decision (steps

1, 2, and 3), while the implementation phase (step 4) will be left aside. Hence, bare

in mind that strategic outsourcing decision-making is conceptualized as always be-

ing initiated by a framework in strategy, which subsequently triggers management

accounting to take action by applying relevant accounting models (i.e. techniques

and methods) to assist in evaluating the alternatives.3 In this respect, the choice

of literature for the analysis will follow the same principle. That is, the trends on

outsourcing in the strategy literature will determine and constrain the selection of

accounting papers to be studied. I will return more thoroughly to the choice of

literature in section 4 after the methodology for the analysis has been established.

<Insert �gure 2 here>

3Notice that in relation to the study of the outsourcing literature in this paper, a framework istaken to be an overall structure that encapsulates how to think about the outsourcing situation. Inaccord with this, the term model will be used to encompass the accounting techniques and methods(techniques tailored to a specifc problem cf. Arbnor and Bjerke (1997, pp. 16-17) ) available forevaluating the decision alternatives.

9

3 Methodology

In this section, the methodology used for the literature analysis is outlined. Since

the interest is with investigating the extent to which accounting models can support

decision-making within strategic frameworks, a set of concepts is needed, which

permits a study of how the discernment and creation of knowledge coheres across

literature streams.

As such, this paper looks to the theory of science, and applies the collective term

of a paradigm to conceptualize the understanding and creation of knowledge within

both strategy and accounting.4 Henceforth, a paradigm is thus taken to consist of

two aspects; epistemology and ontology.5 On the basis of this, and in accord with

the conceptualization and understanding of outsourcing in the previous section, a

framework has thus been developed in preparation for analyzing a literature that

takes the perspective of a decision-maker who wants to make an informed decision

of whether to outsource a given part of the company or not. Figure 3 illustrates this

framework and the connection between the separate parts.

<Insert �gure 3 here>

The theoretical underpinning and exact scope of the concepts in the framework pre-

sented in �gure 3, will be thoroughly outlined in the subsections to come. However,

to motivate this outline and provide a general understanding of the framework, a

brief explanation highlighting some of the main points in relation to outsourcing

decision-making will follow �rst.

Consider thus a make-or-buy situation, where a decision-maker has to determine

on what grounds his decision should be based. This implies establishing the models

and arguments to be considered as valid sources for knowledge creation (epistemol-

ogy). Speci�cally, he has to decide whether knowledge generated from pure thought

(rationalism) alone is acceptable or whether it has to be grounded in empirical evi-

dence (empiricism).

In determining this, the decision-maker�s understanding of the outsourcing sit-

uation (object) inevitably comes to play a role (ontology). On the one hand, if

convinced that a true state of a¤airs exists (realism), the decision-maker will decide

on techniques and create methods he believes will reveal its existence while taking

4In the literature, a paradigm is sometimes also referred to as a "skeleton theory" (Laughlin1995) or an "approach" (Arbnor and Bjerke 1997). However, in accordance with (Nørreklit et al.2010), I prefer the term paradigm because it is more structured and less arbitrary.

5The political aspect also forms a part of a paradigm (Arbnor and Bjerke 1997). However, thisis not emphasized in this article.

10

into consideration methodological limitations and the complexity of the state of af-

fairs. As a result, a certain type of knowledge is created. On the other hand, if

the decision-maker believes that the true state of a¤airs exists only in the eyes of

the perceiving mind and can be partly subdued to construction (idealism), it might

leave him with nothing but the interest of collecting information that can underpin

a decision already made. In this case, the models chosen may prove to be very

di¤erent and perhaps in�uenced by such factors as the prevailing norms and beliefs

in the company or the society as a whole. As a result, another type of knowledge is

created.

Consequently, knowledge creation is best seen as a process of interdependent in-

teraction between epistemology and ontology. Figure 3 is thus not to be interpreted

as a deterministic sequence of steps in which the decision-maker initially selects the

epistemology to follow and subsequently chooses his ontology. Instead, the two-sided

arrows in the framework symbolize that the ontology and epistemology repeatedly

in�uence and shape each other in the creation of knowledge for decision-making.

Reminded by the structure in �gure 3, the remainder of this section elaborates

on the elements herein. First, a distinction between beliefs and knowledge is out-

lined. Subsequently, the scope and use of epistemology and ontology in this paper

is explained and theoretically underpinned.

3.1 Beliefs and knowledge

In accord with �gure 3, rationally informed decision-making is ideally to be based

on knowledge that is certain and not �awed by prejudiced believes or mistakes

(Nørreklit 2009, p. 6).6 In this respect, a sharp distinction needs to be made

between what the decision-maker believes to be true and what he knows to be true.

According to Schwitzgebel (2006), a belief is generally taken to refer to the attitude

one has whenever one takes something to be the case or regard it as true. Therefore,

a belief does not necessarily imply any uncertainty albeit it might. Exactly this point

is what distinguishes a belief from knowledge in traditional philosophy, since Plato

in his work, Meno, de�nes knowledge as true justi�ed belief (Pritchard 2007). That

is, for one to know something: (1) it must be true, (2) one must actually believe it,

6Notice that rational decision-making in this paper should not be confused with the epistemo-logical orientation, rationalism, also referred to in this paper. Speci�cally, rational decision-makingonly ascertains that the decision-maker will choose his course of action on the basis of a rationalevaluation of alternative means to given ends whether they are derived from reason (rationalism)or experience (empiricism). In the context of outsourcing, this implies that the decision-maker,given the information available to him, will choose the set of actions that maximizes the company�sexpected long-run pro�t.

11

and (3) it must be justi�ed (there must be su¢ cient evidence for it). Consequently,

what essentially di¤erentiates knowledge from a mere true belief is that is has been

justi�ed and so knowledge is nothing but a special type of beliefs. Although exposed

to criticism by Gettier (1963), this distinction between beliefs and knowledge will

be applied in this paper.

Thus, referring to �gure 3, a decision-maker, who considers the idea (belief) of

outsourcing, but wants to base his decision on knowledge, has to �nd the means

that can justify his mere belief and transform it into knowledge. These means will

be decided by the decision-maker�s epistemology and ontology as explained next.

3.2 Epistemology

In general, epistemology is concerned with the nature7, source(s), and limits8 of the

decision-maker�s knowledge (Markie 2008). However, of particular interest to the

analysis of the management accounting and strategy literature conducted in this pa-

per, is the second issue, which relates to how the decision-maker can gain knowledge

about the outsourcing situation. Clearly he can form true beliefs just by making

lucky guesses about the company and the industry. Still, it is less obvious how he

can acquire justi�ed beliefs (knowledge) regarding the same matter. Furthermore,

to know the world, the decision-maker must think about it, yet it is ambiguous how

he gains the concepts he uses in thought or what assurance, if any, he has that the

ways in which he divide up the world using his concepts correspond to divisions that

actually exist. In this respect, and as hinted in �gure 3, a classical dispute exists

between so called rationalism and empiricism on what makes up the sources of the

decision-maker�s concepts and knowledge (Ryan et al. 2002). In what follows, an

introduction will thus be given to the competing thoughts on either side.

3.2.1 Rationalism

Essentially, rationalism can be traced back to Socrates (469-399 BC) and Plato

(429-347 BC), who argued the existence of abstract forms of knowledge (Ryan et al.

2002). Yet, it was not until the seventeenth and eighteenth centuries that the ideas

now termed as �rationalism�were fully articulated, �rst and foremost by Descartes

7The nature of a decision-maker�s knowledge concerns what contrasts it to his mere beliefs aslaid out in the previous section already (Markie 2008).

8Some aspects of the world may be within the limits of the decision-maker�s thought but beyondthe limits of his knowledge. Faced with competing descriptions of them, he cannot know whichdescription is true. Indeed, some aspects of the world may even be beyond the limits of thedecision-maker�s thought, so that he cannot form intelligible descriptions of them, let alone knowthat a particular description is true (Markie 2008).

12

(1596-1650) and later by Spinoza (1632-1677), and Leibniz (1646-1716) (Hjørland

2005). To understand rationalism, it is important to accentuate the two criteria on

which this epistemology is based; the Intuition/Deduction criterion and the Innate

Knowledge criterion (Markie 2008).9 Thus, to be a rationalist is to adopt at least

one of these two criteria.

Criterion (Intuition/Deduction). Some propositions about the world are know-able by the subject�s intuition alone; still others are knowable by being deduced from

intuited propositions.

Criterion (Innate Knowledge). The subject has knowledge of some truths aboutthe object as a part of his rational nature.

The Intuition/Deduction criterion involves how the subject becomes justi�ed in

believing propositions about the object. In this respect, intuition is a form of rational

insight. Intellectually grasping a proposition, the subject �sees�it to be true in such

a way as to form a true, justi�ed belief in it. Following through on this, deduction

is a process where the subject derives conclusions from intuited premises through

valid arguments, ones in which the conclusion must be true if the premises are

true. Hence, according to rationalism, intuition accompanied by logical deduction

provides the decision-maker with knowledge gained a priori and independently of

sense experience. I will further elaborate on this in section 3.3.1 when realism is

discussed.

The Innate Knowledge criterion asserts this very same thing. Yet, the di¤er-

ence to the Intuition/Deduction criterion rests on the understanding of how a pri-

ori knowledge is gained. Whereas the Intuition/Deduction criterion cites intuition

and subsequent deductive reasoning, the Innate Knowledge criterion emphasizes the

decision-maker�s inherent rational nature. The subject�s innate knowledge is neither

learned through sense experience nor intuition and deduction. It is just a part of

his nature. Experience may trigger a process by which the decision-maker brings

this knowledge to consciousness, but the experience does not provide him with the

knowledge by itself. It has somehow been with the decision-maker all along.

With this in mind, two other closely related standpoints are normally also adopted

by rationalists (Markie 2008). The �rst is the Indispensability of Reason criterion

asserting that experience cannot provide what we gain from reason:9Some philosophers argue that a third thesis exists: The Innate Concept thesis (Markie 2008).

Still, others, including Locke (1690, p. 91), argue that the Innate Concept thesis is included in theInnate Knowledge thesis as a particular instance of knowledge can only be innate if the conceptsthat are contained in the known proposition are also innate. In this paper the author adopts thislatter view.

13

Criterion (Indispensability of Reason). The knowledge the subject gains aboutthe object, by intuition and deduction, as well as the ideas and instances of knowledge

about the object that are innate to the subject, could not have been gained through

sense experience.

The second is that reason is superior to experience as a source of knowledge:

Criterion (Superiority of Reason). The knowledge the subject gains about theobject by intuition and deduction or has innately is superior to any knowledge gained

by sense experience.

The argument for this last criterion is provided by Descartes in his work Regulae ad

directionem ingenii (Rules for the Direction of our Native Intelligence) rule II and

III, as he argues that what is known a priori is certain, beyond even the slightest

doubt, while what one believes, or even knows, on the basis of sense experience is

at least somewhat uncertain (Descartes 1628, p. 1-4).

In conclusion, rationalism (and hence rational knowledge) assumes that a decision-

maker considering outsourcing does not need to look beyond himself to form a

justi�ed true belief about the outsourcing situation. That is, knowledge about what

is can be known a priori and does not have to be perceived. This con�icts with

empiricism as will be explained next.

3.2.2 Empiricism

The thoughts originating within empiricism date back as far as Aristotle (384-322

BC) who challenged the views of Socrates and Plato and argued that knowledge is

gathered by observation and categorization (Ryan et al. 2002). However, modern

empiricism is in general taken to be linked to British thinking and the scienti�c

revolution consummated by Newton (1643-1727) (Ryan et al. 2002; Hjørland 2005).

Though mutually very di¤erent empiricist, such as Locke (1632-1704), Berkeley

(1685-1753), Hume (1711-1776), and Mill (1806-1873), are thus today regarded as

some of the most in�uential classical empiricists (Hjørland 2005).10

For the purpose of this paper, it su¢ ces to consider empiricism in its strongest

form claiming the following (Markie 2008):

10Notice that a certain group of empiricists referred to as hermeneuticists (interpreters) existswho make a distinction between the methods of the natural sciences and those of the social sciences(Arbnor and Bjerke 1997). In particular, they assert that natural science methods, even if modi�ed,are essentially unsuitable in social science domains. Prominent philosophors include Husserl (1859-1938) and Heidegger (1889-1976) (Beyer 2003; Rambjerg and Gjesdal 2005). For the purpose ofthis article, however, I will not distinguish between di¤erent variations of empiricism.

14

Criterion (Empiricism). The subject has no source of knowledge about the objector the concepts he uses about the object other than sense experience.

Hence, when compared to rationalism, empiricism rejects the Intuition/Deduction

criterion and Innate Knowledge criterion. Insofar the decision-maker has knowl-

edge about the outsourcing situation, it is a posterior and dependent upon sense

experience (Hjørland 2005). As a consequence of this, empiricists also reject the Su-

periority of Reason criterion; since reason alone does not bring the decision-maker

any knowledge, it certainly does not give him superior knowledge. With regard

to the Indispensability of Reason criterion, empiricists generally reject this as well,

albeit they need not. The Empiricism criterion does not entail that the decision-

maker has empirical knowledge. It merely implies that knowledge can only be gained

through experience. In fact, modern empiricism in general tends to accept as un-

tenable the idea that knowledge is uniquely determined by experience (Ryan et al.

2002). Instead, proponents hold that experience can represent a justi�cation of the

beliefs that one has about what is known. That is, only through repeated obser-

vation of particulars can the decision-maker begin to form an understanding of the

outsourcing situation. The basic methods of empiricism are thus observation and

induction �a bottom-up strategy in the processing of information (Hjørland 2005):

It is obvious that not all knowledge stems directly from experience.

Hence empiricism always assumes a strati�ed form, in which the low-

est level issues directly from experience, and higher levels are based on

lower levels. It has most commonly been thought by empiricists that be-

liefs at the lowest level simply �read o¤�what is presented in experience.

If a tree is visually presented to me as green I simply �register�this ap-

pearance in forming the belief that the tree is green. Most of our beliefs

�general beliefs for example �do not have this status but, according to

empiricism, are supported by other beliefs in ways that eventually trace

back to experience. Thus the belief that maple trees are bare in winter

is supported by particular perceptual beliefs to the e¤ect that this maple

tree is bare and it is winter (Alston 1998).

To conclude the discussion of rationalism and empiricism, I stress that the descrip-

tions to some extent resemble the extreme views within each epistemology, and so

more moderate versions on either side need not be complete contrasts. Indeed, it is

more suitable to think of the discussion as a continuum that spans from pure ratio-

nalism to pure empiricism. This has already been indicated for modern empiricism

in its acceptance of intuition as a part driver of knowledge creation. Yet, it also holds

15

true for moderate forms of rationalism, which acknowledge the role of observation

(Hjørland 2005). Consequently, the literature study of outsourcing decision-making

in this paper will not apply rationalism and empiricism categorically, but instead try

to point out the dominating characteristics of either epistemology in the scholarly

work analyzed.

3.3 Ontology

Whereas empiricism and rationalism are epistemological distinctions concerned with

how knowledge is obtained, ontology deals with what the decision-maker discerns

as real (Ryan et al. 2002). Referring to �gure 3, a distinction is normally made

between realism and idealism, which is outlined next.

3.3.1 Realism

As the discussion of realism in this paper is constrained to ontology and how the

decision-maker justi�es his belief about the outsourcing situation (the object) on

the basis of sensory experience, focus will be kept on what is known in the literature

as indirect realism (representationalism) (Bonjour 2007). This perspective holds

that the immediate object of experience or intuition represents an external object

in a way that allows one to infer justi�able from such experience or intuition to the

existence of the corresponding external object. For example, if one sees (experiences)

a white table, what one actually sees is not the table itself but a representation

of it. In this way, di¤erences in perception which occur due to changes in light

conditions, positions of the viewer, etc., can be easily explained, as it is not the

object (table) that changes but merely the perception of it. However, the external

object need not have a physical existence. Indeed, the external object can be a

non-physical construct, such as an idea or a law of society, if only its existence is

independent of the decision-maker observing or intuitively grasping it. Historically,

representationalism has been the dominating view within realism and was adopted

by both Descartes and Locke (Bonjour 2007).11 In e¤ect,realism will be taken to be

synonymous with indirect realism henceforth.

Criterion (Realism). The object has a reality that is independent of the subject�sperception. The subject can never "see" the true object itself but only a representation

11Notice that another and more simple form of realism than indirect realism does exist. Thisis called direct realism (naïve realism) and holds that a physical object is directly perceived in away that allegedly avoids the need for any sort of justi�catory inference from sensory experienceto physical reality. Consequently, according to this view, if one sees a white table one is indeedseeing the table and not some representation of it.

16

hereof. Still, in seeing this representation he can infer justi�ably to the object itself.

Thus, at the heart of realism is an underlying understanding of the world as consist-

ing of unalterable circumstances that can be revealed and responded to. Hence, to

exemplify the essence of decision-making within realism and to contrast it with ide-

alism as will be introduced shortly, think of a maze (the object) through which the

subject has to �nd its way. The maze consists of walls that are, by the de�ning prin-

ciples of realism, taken by the subject to be unchangeable. Therefore, the subject�s

decision-problem is reduced to a simple optimization problem, where the optimal

way through the maze is constrained by the unalterable elements �the walls. In this

manner, decision-making based on realism is all about discovering the unchangeable

aspects of the world (the walls) and react accordingly to these (�nd the optimal way).

With the adoption of realism as the ontological standpoint, it is important to have

a notion of how the decision-maker�s belief about the outsourcing situation can be

justi�ed and turned into knowledge available for decision-making. In this respect,

realism is often associated with the correspondence theory of truth (David 2009):

...a belief is true when there is a corresponding fact, and is false when

there is no corresponding fact (Russell 1912, p. 129).

To fully grasp this theory, consider the following illustration:

<Insert �gure 4 here>

Figure 4 mimics indirect realism, since the subject cannot observe the object itself

(the true state of a¤airs) but only experience an appearance of it or intuitively grasp

its existence (some state of a¤airs). Knowing this, the subject checks to see if his

belief corresponds with his experience or intuition. If this is the case, the belief is

justi�ed and becomes knowledge since the subject, according to criterion (realism),

can infer to the object itself directly from the experience or intuition.

To some extent the correspondence theory of truth can potentially cause a prob-

lem in relation to the creation of knowledge since knowledge is, per se, beliefs known

to be true (as clari�ed in section 3.1). Yet, in practice, the use of the correspondence

theory of truth inevitably entails di¤erent degrees of certainty as the actual experi-

ence or intuited premises might di¤er from the external object itself, which makes

inference from experience or intuition to the external object somewhat uncertain.

For example, a measuring error might cause the subject�s experience (observation)

to di¤er from the external object within empiricism. Likewise, within rationalism

17

the external object which the subject tries to intuitively grasp might be so complex

that it becomes impossible to comprehend its full entity. However, the ideal use of

the correspondence theory of truth needs to be seen in relation to decomposition

analysis, coined by Descartes (1628), rule XIII, as follows:

If we perfectly understand a problem we must abstract it from every

super�uous conception, reduce it to its simplest terms and, by means of

an enumeration, divide it up into the smallest possible parts (Descartes

1628).

Thus, the very idea of decomposition analysis is to decompose the problem of interest

into su¢ ciently small parts (fundamental problems) that allows certain and infalli-

ble knowledge to be extracted, and upon which all subsequent knowledge, which is

needed to solve the initial problem, can be constructed. In this respect, the ideal use

of the correspondence theory of truth is meant to study these fundamental problems

where the perception error is non-existing or small enough to ensure the creation of

beliefs known to be true (knowledge). Within rationalism, decomposition analysis

is based on intuition and deduction, as the fundamental problems are initially dis-

cerned and solved by intuition. Subsequently, all problems leading up to the initial

problem of interest can be answered stepwise using deductive reasoning �the oper-

ation on the basis of which the truth of the conclusion is purported to necessarily

follow from or be a logical consequence of conclusions already known to be true.

Within empiricism, decomposition analysis is based on experience and induction.

First, the fundamental problems are discerned and solved through experience as the

perception error is non-existing or su¢ ciently small. Following this, the conclusion

to the initial problem of interest is constructed through a sequence of steps based

on inductive reasoning �the operation where more general conclusions are reached

from speci�c facts (experiences) through logical arguments. Figure 5 below captures

the similarities and di¤erences between decomposition analysis within rationalism

and empiricism, respectively.

<Insert �gure 5 here>

3.3.2 Idealism

The contrast to realism is idealism which in its most simple form represents the view

that reality exists within the mind of the subject (Ryan et al. 2002). Historically,

Berkeley is considered one of the most in�uential proponents of idealism (Downing

2004), and in his much-studied work, ATreatise Concerning the Principles of Human

Knowledge, he states:

18

It is indeed an opinion strangely prevailing amongst men, that houses,

mountains, rivers, and in a word all sensible objects have an existence

natural or real, distinct from their being perceived by the understanding.

But with how great an assurance and acquiescence soever this principle

may be entertained in the world; yet whoever shall �nd in his heart to

call it in question, may, if I mistake not, perceive it to involve a manifest

contradiction. For what are the forementioned objects but the things

we perceive by sense, and what do we perceive besides our own ideas or

sensations; and is it not plainly repugnant that any one of these or any

combination of them should exist unperceived (Berkeley 1710, p. 31)?

Hence, from this quote, Berkeley�s main argument for idealism can be reduced to

the following sequence (see Winkler (1989, p. 138)): "A subject perceives objects

(houses, mountains, etc.) and he can only perceive ideas. Therefore, ordinary ob-

jects are ideas." This line of argumentation has since been heavily disputed (Downing

2004), yet if accepted, it naturally implies that every object perceived is mentally

constructed and only continue to exist in the presence of a perceiving mind (Sprigge

1998).

Criterion (Idealism). The object has no reality independent of the subject�s per-ception. Indeed, the object is nothing but an expression of the subject�s mind.

Falling in line with the traits of idealism and of speci�c interest to this article is con-

structivism, which in its radical form shares the same ontological stance as idealism

(Oxford 1997). However, the weaker form of constructivism does not preclude the

existence of an external reality; it merely claims that each of us construct our own

reality through interpreting perceptual experiences of the world (Jonassen 1991).12

Criterion (Constructivism). The object can have a reality partly independent ofthe subject�s perception. However, some part of reality is constructed through the

subject�s perception.

Hence, returning to the example from section 3.3.1, constructivism, as it is un-

derstood in this paper, holds the view that the maze itself and some of the walls

within the maze cannot be removed nor overcome �they simply represent a part of

the reality that cannot be subdued to construction within the mind of the subject.

Therefore, the subject�s understanding and interpretation of how to get through

the maze will have to take into account these unchangeable circumstances. Still,

12The de�nition of weak constructivism bears close resemblance with constructivist pragmatismas it is introduced in Nørreklit et al. (2006)

19

the rest of the walls within the maze may be less de�nitively given but can indeed

be exposed to creative thinking and interpretation, which might even contradict

with their original intended purpose. For instance, the subject might realize that

new pathways through the maze can be constructed if certain walls are torn down

while others are passed by simple jumping. In other words, constructivism is indeed

about constructing and acting upon new possibilities as objects are combined and

understood in novel ways. Yet, in its weak form it is also about accepting that

certain aspects of reality are unalterable and cannot be subdued to constructivistic

thinking.

For the purpose of this article, a distinction between idealism and constructivism

will not be made. Instead, the ideas from constructivism will be incorporated into

the notion of idealism to stress the fact that reality is not per se given to the decision-

maker but can in fact be shaped �or constructed �either mentally or socially.

Accepting this leads to the question of how beliefs are justi�ed and thus how knowl-

edge is created within idealism. In this respect, the coherence theory of truth is

o¤ered as a pendant to the correspondence theory of truth associated with realism

(Young 2008). Essentially, the coherence theory of truth builds upon the principle

that a belief coheres with a set of other beliefs if and only if it is entailed by members

already in the set. Given this, the justi�cation of a belief can only be found in its

coherence with other beliefs. Truth or falsity of beliefs are therefore checked, not in

terms of their correspondence with reality but, rather, in terms of their �coherence�

either with the other beliefs of the decision-maker or with the beliefs of others (Ryan

et al. 2002; Kirkham 1998).

In conclusion, the discussion of realism and idealism, as it is laid out above, merely

represents extremes on a continuum that spans from pure realism to pure idealism.

In practice, however, these extremes are hardly tenable (Ryan et al. 2002). On

the one hand, the central problem of realism relates to bridging the gap between

the experience or intuited premises of the object and the object itself, as already

discussed in section 3.3.1. On the other hand, the main problem of idealism is that

it takes one to the position that what is true is either what one chooses to believe

is true or what society believes is true.13 Therefore, in keeping with the treatment

of epistemology, the literature study of outsourcing decision-making in this paper

will not apply realism and idealism categorically, but instead try to point out the

dominating characteristics of either ontology in the scholarly work analyzed.

13For further discussion on what can be discerned as real see Nørreklit et al. (2010).

20

4 A brief historical overview of outsourcing in

strategy, management accounting and business

practice

In this section, the emerging ideas on outsourcing in the strategy and management

accounting literature are described as they have inspired and been inspired by ac-

tual business practice. Partly combining the structure from Grant (2005, p. 19)

with Hätonën and Eriksson (2009), the history is divided into three main eras, each

focusing on the most important developments and characteristics. Speci�cally, the

time period indicated for each era is primarily related to the developments in busi-

ness practice and is not to be interpreted too literally �it merely insinuates a rolling

transition from one era to another. Also, since a natural time lag exists between

practice and academia (see the cultural circuit of capitalism in Thrift 2005; Seal

2010) there is not a year-to-year correspondence between the strategic frameworks

being developed in the strategy literature and their di¤usion into practice. Fur-

thermore, this time lag tends to be even longer for accounting as some time passes

before the strategic concepts �nd their way into the discussions in the management

accounting literature.

With respect to the research stream on strategy, focus is kept on the fundamen-

tal frameworks, or theories of the �rm, that have marked the di¤erent periods.

Obviously, many more than these few carefully selected theories reside within the

strategy literature, yet I believe that the "boundary discussion of the �rm" with

which these are concerned are most relevant to the essence and aspects of strategic

outsourcing. Additionally, the fundamental nature of these frameworks imply that

most subsequent literature tend to build on their premises and therefore mainly

represents relatively minor modi�cations to already existing ideas.

Consequently, the selection of management accounting literature mimics the the-

oretical developments in strategy. As such, interest is given to how management

accounting has managed to operationalize the strategic concepts to design models

(i.e. techniques and methods) that, in accordance with the de�nition of management

accounting, can be applied in practice to support the strategic framework through

information provision and evaluation of alternatives (subtask 3 and 4) as laid out

earlier in section 2. In particular, for every theory of the �rm selected, the ques-

tion has been asked whether any models have since been developed in the practical

oriented academic literature on management accounting (the conventional wisdom

literature) to provide decision-support for the strategic framework when carried into

21

business practice.14

Ending the descriptions of each era, the framework from section 3, �gure 3, is

applied to analyze the ontology and epistemology of the strategy and accounting

literature. The objective is to clarify how the understanding of the outsourcing sit-

uation has changed with new strategic frameworks emerging (ontology), and how

the complexity of the constructs considered important for optimal decision-making

within these strategic frameworks can be discerned (epistemology). Analyzing the

same dimensions for the accounting literature makes it possible to de�ne to what

extent the accounting models have been able to support and satisfy the needs raised

by the changing strategic frameworks.

4.1 The era of Corporate Planning: 1950s - 1970s

In the post-war period conglomeration along with horizontal and vertical integration

(insourcing) was widely held among managers as the proper way to organize business

activities (Lonsdale and Cox 2000; Rumelt 1982; Stigler 1951). The prevailing belief

was that internalization of all activities in the value chain would lead to economies

of scale and that bene�ts related to superior market power (horizontal integration),

improved insurance following from increased portfolios of goods and services (con-

glomeration), and better control with the value chain (vertical integration) could be

reaped (Lonsdale and Cox 2000).

As an example of this, 7-Eleven used to deliver its own gasoline, and make its

own candy and ice. In fact, the company even owned the cows that produced the

milk it sold (Gottfredson et al. 2005). Likewise, Ford Motor Company developed

the concept of "Fordism" (Piore and Sabel 1986) which consisted of integrating into

the �rm not just car assembly activities and production of components but even the

extraction of iron ore and car dealerships (Mol 2007).

4.1.1 The characteristics of corporate planning

To control companies vastly growing in size and complexity, senior executives looked

for new methods to systemize planning and decision-making (Grant 2005). However,

strategy, as a formalized science, had yet to emerge. Instead, �nancial budgeting

functioned as the basic technique for annual �nancial planning, while discounted

cash-�ow (DCF) models to capital budgeting o¤ered a new approach to appraising

14Hence, notice that attention will not be given to the theoretical part of the managementaccounting literature studying outsourcing using agency theory or the like.

22

individual investment projects. Consequently, corporate planning (long-range plan-

ning) and the forming of separate corporate planning groups within the company

were devised as a structure for coordinating individual capital investment decisions

and planning the long-term development of the �rm (Wrapp 1957; Payne 1957).

As such, corporate planning at that time normally involved the development of

a �ve-year plan initially laying out the company�s goals and objectives. Based on

these, forecasts of key economic trends (including market demand, the company�s

market share, revenue, costs, and margins) were conducted and used to prioritize

and allocate capital expenditures for the di¤erent products and business areas of

the �rm.15

In general, corporate planning can be seen as a part of a wider enthusiasm among

companies at this time towards using "scienti�c" techniques and methods in rela-

tion to decision-making, including accounting models such as cost-bene�t analysis,

budgeting, and DCF appraisal in combination with mathematical methods such as

linear programming, econometric forecasting, and macro economic demand manage-

ment (Grant 2005). Indeed, several economists even argued that scienti�c decision-

making and rational planning by corporations would supersede markets in allocating

resources (Galbraith 1967). Hence, the use and knowledge of accounting principles

in particular were regarded as essential for company planning and decision-making,

which is captured ever so clearly by Wrapp (1957):

If a company has had considerable experience with market research and

budgeting procedures, the planning group may be able to start by set-

ting �ve-year salesgoals... The point... is that sales and cost forecast-

ing techniques are indispensable to this phase of planning. Unless the

members of the planning group are familiar with such techniques �and

certainly many operating managers are �they will probably be skeptical

of the "crystal gazing" when �rst introduced to it. Because it is foreign

to their own experience, the planning may seem to them like so much

"guesswork" (Wrapp 1957, p. 44).

Understanding corporate planning, therefore, inevitably entails understanding ac-

counting, since corporate planning does not constitute an independent theoretical

framework but merely embraces the fundamental logic in accounting as to how a

15A �ve-year plan was held to be optimal, since a shorter period would draw attention too muchto operating problems and discourage the considerations of planning problems. Likewise, a longerperiod would become too nebulous as a basis for developing "supporting" plans, i.e., detailedstatements of what must be done in order to meet the long-term goals (Wrapp 1957).

23

company can be expressed in terms of a �nancial statement (pro�t, revenue and

costs) and a balance sheet (assets, liabilities and equity). In other words, corpo-

rate planning simply re�ects the general belief that nothing more than knowledge

of accounting particulars is essentially needed for optimal decision-making.

4.1.2 The use of corporate planning for make-or-buy decision-making

The speci�c steps of corporate planning and hence the use of accounting models

in relation to make-or-buy during this period is outlined in Gross (1966). Comple-

menting the description herein with the more general considerations on make-or-buy

in Higgins (1955), provides a clear indication of the understanding of outsourcing

during the era of corporate planning. Both articles are concerned with make-or-

buy decision-making from the perspective of a manufacturing company �a typical

feature of this period. Thus, very characteristically, the object considered for out-

sourcing is a well de�ned physical phenomenon (e.g. simple production component)

that can be bought at arm�s length from available suppliers in the market.

In accordance with the general steps of corporate planning, the starting point for

make-or-buy analysis is an estimation of the company�s expected sales volume on the

basis of a projection of the general market growth (Gross 1966). The expected sales

volume subsequently functions as the foundation for constructing budgets re�ecting

capital investments (physical and human) and costs that can be used for comparison

between the make-or-buy alternative. Indeed, both articles stress the importance of

proper cost comparison, since the lower alternative cost applicable to either make

or buy is believed to have the greatest impact on management�s �nal choice, simply

for the reason that the company�s main objective is to maximize long-term pro�t

(Higgins 1955, p. 112; Gross 1966, p. 750). Hence, costs are allocated to the make

alternative using full cost, with overheads being distributed as a percentage of direct

labor costs. With respect to determining the cost of buying, Gross (1966) simply

de�nes it to be the supplier price. Higgins (1955), however, extends this to also

include cost of incoming freight and to make allowance for costs related to search

and inspection. Thus, although the term is not explicitly used, Higgins (1955) makes

an early attempt at estimating the total cost of ownership16.

Nevertheless, neither Higgins (1955) nor Gross (1966) mention the notion of

opportunity cost and incremental cost in relation to cost allocation. Still, to provide

fairness to the conception of costing during the era of corporate planning, it is

important to stress that the de�nitions of opportunity cost and incremental cost are

indeed introduced and discussed in the accounting literature (McRae 1970). Thus,

16See section 4.2.2 for further information about total cost of ownership.

24

although several important accounting text book of this period, such as Gordon and

Shillinglaw (1964), along with the papers discussed here in relation to outsourcing,

do not adapt these concepts, a growing academic interest revolving around how

to incorporate and interpret opportunity cost and incremental cost in the context

of accounting was emerging (Mathews 1962, pp. 453-454; Horngren 1965, p. 165;

Anthony 1964, p. 569).

Whereas quantitative aspects, such as cost and investment comparisons, consti-

tute the principal factor in make-or-buy decision-making during this period, more

qualitative aspects (intangibles) are not left unnoticed. Indeed, Higgins (1955) states:

Cost studies take time; a month or more is perfectly normal. They

involve expense on your part. They give a snapshot of current conditions,

but they do not show what your costs or your supplier�s costs will be

a year or �ve years hence. They do not re�ect intangibles which may

vitally a¤ect your future production (Higgins 1955, p. 113).

To compensate for this, it is suggested, among other things, to take into account

component quality (Higgins 1955; Gross 1966) and supplier access (Higgins 1955)

when evaluating each alternative. Still, both articles are rather silent as to what

extent these qualitative aspects can actually tip the decision in favor of either al-

ternative. In particular, Gross (1966) seems more concerned with using qualitative

factors to shed extra light on certain aspects of the quantitative factors (expected

sales, costs, and investments) and caution their estimation, than he is with using

the qualitative factors as actual decision variables.17 That is, the lack of strategy as

a formal science comes to be re�ected in the absence of an underlying conceptual

framework for understanding the boundaries of the �rm (i.e. a theory of the �rm)

that can help bind together the qualitative and quantitative factors to form a more

comprehensive assessment of each alternative. Instead, outsourcing decision-making

is somewhat reduced to a checklist of independent qualitative and quantitative issues

to be investigated. Since the quantitative factors by nature are easier to handle than

the qualitative factors, it leaves the impression that quantitative aspects automati-

cally come to play a dominating part, possibly dictating the role and weight given to

more qualitative oriented aspects. In this respect, the make-or-buy decision-making

process merely manifests the general tendency of corporate planning to embrace

the variables comprised in the fundamental accounting logic. Thus, make-or-buy

decision-making during the era of corporate planning, is characterized by relying

17See for instance the discussion on the qualitative aspect of management having the necessaryknow-how to undertake and handle outsourcing in a given industry (Gross 1966, p. 753)

25

heavily on decision variables (expected sales, costs, and investments) which can

be constructed and gauged on the basis of quantitative historical accounting data.

Piecing together the descriptions in Higgins (1955) and Gross (1966), �gure 6 below

sums up the thoughts of this period as to how corporate planning can be structured

and what issues to consider in outsourcing decision-making.

<Insert �gure 6 here>

4.1.3 Analysis

As can be seen from the description of the era of corporate planning, there is an

almost complete congruity between the notion of corporate planning and the treat-

ment of make-or-buy decisions in the management accounting literature. Still, for

the purpose of this article, it is bene�cial to study each of them separately.

Using the framework from section 3, �gure 3, the underlying ontology of corporate

planning seems heavily in�uenced by the thoughts of realism. Speci�cally, there

appears to be a tendency towards a genuine belief in the existence of an objective

and unalterable truth of how the world works (cf. criterion (Realism)); a truth that,

if revealed, will determine whether outsourcing is optimal or not. This naturally

follows from the nature of planning, since planning in essence assumes that some

underlying state of a¤airs subsists in relation to which an optimal sequence of actions

can be constructed and e¤ectuated to ensure that given objectives are met. That is,

planning sees the world as consisting of objective structures that can be discerned

and described, thus, in turn, making it possible to predict and extrapolate future

developments.

Additionally, the epistemology of corporate planning tends towards empiricism

where the most in�uential source of knowledge for outsourcing decision-making is

careful observation of particulars (cf. criterion (Empiricism)). Indeed, corporate

planning is built on a general conception of how the market and the companies

work, yet this conception cannot generate knowledge in itself but merely help clar-

ify which factors empirical evidence need to be provided for. This is most clearly

seen from the checklist of quantitative and qualitative factors to empirically inves-

tigate in relation to make-or-buy decision-making in �gure 6.

Since corporate planning essentially embraces fundamental accounting logic and

most attention is, therefore, given to factors that can be measured by �nancial

numbers calculated from historical data, accounting comes to play a very natural

role as the tool that can reveal the truth about the world and decide in favor of either

26

make or buy. That is, albeit several accounting models used during this period are

overly simplistic and �awed (e.g. cost allocation based on full cost), it is not a con-

cern since they meet the expectation and enthusiasm of corporate planning to apply

scienti�c techniques and methods for optimal decision-making. This point is further

emphasized by the intake of concepts such as opportunity cost and incremental cost

from the economics literature. In this respect, the congruity between the thoughts

in corporate planning and the principles available in accounting implies that em-

piricism and realism remain unchallenged as the innate epistemology and ontology

in the management accounting literature, respectively. Indeed, it is essentially ac-

counting that in�uences corporate planning and not vice versa. Speci�cally, the

accounting treatment of make-or-buy decision-making clearly illustrates the steps

of decomposition analysis (bottom-up strategy) and the use of the correspondence

theory of truth in the processing of information. That is, the analysis of the out-

sourcing problem is broken down into smaller parts to minimize the perception error

from empirical investigation. For instance, the knowledge creation, following from

a valid empirical projection of the general market growth, functions as the starting

point for a company�s speci�c sales forecast, which in turn functions as the basis for

a more complex budget analysis of the costs and investment requirements related to

each sourcing alternative.

In conclusion, the thoughts of corporate planning are fully satis�ed by the ac-

counting models available during the era of corporate planning without stretching

the limits for plausible data collection in the management accounting literature

in order to obtain correspondence to the object under study. For this reason, in

the literature, accounting plays a very dominant role with respect to make-or-buy

decision-making throughout this period.

4.2 The era of the Big Bang: 1980s - Early 1990s

By the end of the 1970s, it was increasingly acknowledged that many large and

diverse corporations underperformed the market. In his seminal paper, Rumelt

(1974) examined the linkage between diversi�cation and performance, and found

that the highest levels of pro�tability were exhibited by companies having a strategy

of diversifying primarily into areas that drew on some common skill or resource. In

addition, the lowest levels of performance were those of vertically integrated �rms

following strategies of diversi�cation into unrelated businesses. Thus, reinforced by

the onset of a global recession in the early 1980s, a consensus slowly emerged to

suggest that �rms should focus on fewer activities (Lonsdale and Cox 2000).

Moreover, the turbulence in the business environment also made it di¢ cult for

27

companies to plan their investments, new product introductions, and personnel re-

quirements three to �ve years ahead (Grant 2005). The result was a shift in emphasis

from corporate planning to strategy making, where the focus was less on the detailed

management of companies�growth paths than on positioning the company in the

markets and in relation to competitors in order to maximize the pro�t potential.

Hence, although outsourcing as a corporate concept had been around for a long

time (Dibbern et al. 2004; Quinn and Hilmer 1994), it was the combination of early

contracting out during the 1980s (Lacity and Hirschheim 1993) and the transition

from corporate planning to strategy making that marked the entrance to the era

of the Big Bang18, and established outsourcing as a viable strategy (Hätonën and

Eriksson 2009). In this respect, companies gradually started to experiment with out-

sourcing non-core business processes to cut operational costs and obtain e¢ ciency.

This has since been termed traditional outsourcing.

4.2.1 Strategy: Transaction Cost Economics

The theoretical foundation underlying the shift in management ideology during the

era of the Big Bang with regard to make-or-buy decision-making can to a great

extent be attributed to Williamson�s Transaction Cost Theory (Williamson 1975;

1979; 1981; 1985), which builds on the ideas put forth in Coase (1937). According

to Williamson, a transaction takes place when:

... a good or service is transferred across a technologically separable in-

terface. One stage of activity terminates and another begins (Williamson

1985, p. 1).

Hence, a transaction does not necessarily involve a transfer between separate legal

entities but can in fact take place within the same company. With this in mind,

Transaction Cost Economics (TCE) in general analyses situations in which trans-

action cost avoidance by �rms can be particularly acute and focuses on exchanges

in which opportunistic potential is signi�cant (Conner 1991). Such potential exists

when three conditions pertain simultaneously: Asset speci�city, small numbers of

potential transactors, and imperfect information in combination with bounded ra-

tionality19. Asset speci�city imposes a condition of dependence upon the owner, A,

of the speci�c asset, because A�s value depends on the presence of another input, B,

18The name Big Bang is used to emphasise the strength and intensity with which this �rststrategic perception of outsourcing gained popularity (Hätonën and Eriksson 2009).19Notice that bounded rationality refers to human behaviour that is intendedly rational but only

limited so (Simon 1961, p. xxiv). Put di¤erently, bounded rationality implies that an agent intendsto behave rationally, yet fails to do so due to his human limitations.

28

to which the former is speci�c. Small numbers reinforces this dependence, because

A cannot costlessly �nd a replacement for B, should B�s services be withdrawn.

Finally, imperfect information means that complete contingent contracts cannot be

written, which in turn implies that a priori knowledge of B�s later actions cannot

be fully incorporated in determining A�s ex ante hire-price. Hence, A cannot nullify

the risk of later opportunism by B through an ex ante adjustment of the price for

A�s services.

Consequently, Williamson argues that �rms will come into existence when this

opportunistic potential is signi�cant, since �rms (hierarchies) exhibit a joint pro�t

maximizing attitude that can help cope with opportunism by vertically integrating

the transaction (Williamson 1975). That is, the transaction cost of using the market

may simply be too high and di¢ cult to handle through an arm�s length contractual

arrangement. This in turn favors internalization of the transaction. However, ver-

tical integration is not optimal per se, because �rms lack a clear linkage between

performance and reward which reduces explicit incentives to perform. Therefore,

if opportunistic potential is limited due to low asset speci�city, large numbers, and

perfect information, transactions between autonomous contractors will come to dom-

inate, as the market automatically provides these high incentives through the mar-

ket pricing mechanism. Accordingly, the preferred sourcing alternative is thus the

governance structure that can minimize the production cost and transaction cost

altogether.

Having understood the notion of TCE, it is important to stress that Williamson

is not very explicit on what makes up transaction costs, since he is more concerned

with the transaction cost regulating ability of the di¤erent governance structures

than he is with the absolute size of the transaction costs. Consequently, he con�nes

his characterization of transaction costs to include the costs related to planning,

adapting and monitoring task completion under alternative governance structures

(Williamson 1985, p. 2).

The idea of comparing more than the mere internal production cost and the market

price when deciding on make-or-buy, can also be found in Porter (1980) �the �rst

cornerstone of the famous Industrial View which I will return to more thoroughly

in section 4.3.1 below. On vertical integration, Porter speci�cally writes:

Many vertical integration decisions are framed in terms of the "make

or buy" decision, focusing on the �nancial calculations such a decision

entails. That is, they are preoccupied with estimating the cost savings of

integration and balancing them with the investment required ... The de-

29

cision must go beyond an analysis of costs and investment requirements

to consider the broader strategic issues of integration versus use of mar-

ket transactions, as well as some perplexing administrative problems in

managing a vertically integrated entity that can a¤ect the success of the

integrated �rm (Porter 1980, p. 301).

Clearly, Porter questions the fundamentals of corporate planning and the idea of rely-

ing mainly on a static scienti�c approach to make-or-buy decision-making. Instead,

something extra is needed � something that can handle the competition between

companies. Thus, albeit the term "transaction cost" is not speci�cally stressed in

the above, Porter by and large captures the very essence hereof in his description.

That is, the company needs to look beyond the obvious costs and investment re-

quirements related to each sourcing alternative to also consider the indirect costs and

investments of handling the dynamics/complexity in the market and the hierarchy,

respectively. However, whereas Williamson is chie�y concerned with safeguarding

against opportunistic behavior, Porter focuses on the industry in which the com-

pany operates; the sole purpose is to create a position within the industry that can

be defended against potential competitors. Hence, based on an analysis of the �ve

competitive forces, the company must compare the strategic bene�ts and costs when

deciding on make-or-buy.20

Whether the focus of analysis is opportunism or industry competitiveness, the overall

message in the strategy literature during the era of the Big Bang is unambiguous;

the company can no longer rely on the static scienti�c approach to make-or-buy

decision-making. Instead, it explicitly needs to take into account the costs and

bene�ts of dealing with the market (opportunism and competition) as well as the

hierarchy (low incentives). However, since the Industrial View gains more ground

in the literature in later periods as it is further developed, I will postpone the dis-

cussion of this theory for now and concentrate the descriptions of the management

accounting literature on TCE alone.

20For instance, a strategic bene�t of vertical integration (making) might be that it allows thecompany to o¤set supplier or customer bargaining power. If the company is dealing with suppliersor customers who wield signi�cant bargaining power and reap returns over investment in excess ofthe opportunity cost of capital, it pays for the �rm to integrate even if there are no other savingfrom the integration (Porter 1980, p. 307). Likewise, strategic cost of vertical integration mightbe related to overcoming mobility barriers caused by cost advantages from proprietary technologyor favorable sources of raw materials (Porter 1980, p. 309).

30

4.2.2 Management accounting: Total Cost of Ownership

To transform transaction costs into accounting terms, the concept of Total Cost of

Ownership (TCO) has been suggested as a proxy in the management accounting

literature (Ellram 1993; Ellram 1995; Ellram and Maltz 1995). Indeed, the notion

of TCO was hinted at in terms of life-cyle costing (e.g. Jackson and Ostrom 1980;

Harriman 1982) several years before the concept actually surfaced.21 Yet, with the

emergence of Activity-Based Costing (ABC) during the 1980s (Kaplan 1983; Ka-

plan 1984b; Kaplan 1984a; Kaplan 1985) it became possible to partly capture the

complexity of transaction cost through the notion of TCO.

Broadly de�ned, TCO is an approach that requires the buying �rm to determine

which costs are considered most important or signi�cant in the acquisition, pos-

session, use and subsequent disposition of a good or service (Ellram 1995). One

way to categorize the cost of ownership is based on the order in which the cost el-

ements are incurred in the transaction sequence: Pre-transaction, transaction, and

post transaction (Ellram 1993). In this respect, pre-transaction costs refer to the

costs incurred prior to placing the order, and comprises the costs of investigating

alternative sources, qualifying and educating suppliers regarding the �rm�s systems

and expectations, and adapting to the systems, styles, and delivery methods of new

sources of supply. Moreover, the actual transaction cost elements are those items

related to order placement and receipt, which include the price of the item itself

along with the costs associated with actually placing an order and getting the order

into the �rm or supply chain. In particular, this comprises the costs associated with

preparing and placing the order (EDI, fax, phone, etc.), following up on the order,

receiving, matching receiving data to the invoice, and paying the bill. Finally, post

transaction costs relates to those costs incurred when the item is owned or in the

possession of the company. The actual occurrence of posttransaction costs may be

soon after the order is received, or years later when the purchased item is in use or

being modi�ed, repaired, or disposed of. For this reason such costs may be sepa-

rated from the purchase by a great deal of time and thus di¢ cult to track. More

speci�cally, the cost may for instance include product repair in the �eld, routine and

special maintenance costs, costs associated with replacement part scarcity and/or

obsolescence.

21Life-cycle costing focuses primarily on capital or �xed assets. The aim is to go beyond thepurchase price of an asset, to determine how much it actually costs the organization to use, main-tain, and dispose of that asset during its lifetime. In this respect life-cycle costing is congruentwith TCO but represents only a subset of TCO activity (Ellram and Siferd 1998).

31

Since TCO is a very broad concept, it can be used in several di¤erent ways in

relation to outsourcing decision-making. For the purpose of this paper, however,

focus will be kept on Ellram and Maltz (1995), who have outlined a normative yet

practical model for determining the TCO of each make and buy alternative, respec-

tively. The �rst step in developing a complete understanding of the internal costs

is to construct a �ow chart of the sequence of activities performed in the current

system. Subsequently, key cost drivers are identi�ed for each step in the �ow chart

and costs are allocated accordingly using ABC, thus providing insight into the cost

structure of the make alternative. The same procedure is then followed for the

outsourcing alternative where the company in collaboration with the prospective

supplier(s) try to establish a �ow chart of the steps and costs involved, should the

transaction be handled by a third party.

Based upon the respective cost studies, the idea is to approach the outsourcing

decision as an incremental cost analysis where the key in comparing the current and

proposed alternatives is to examine: (1) what current costs will change or be elimi-

nated, (2) what additional costs will be added. Ellram and Maltz (1995) illustrate

this procedure for a particular case company where a number of key areas are identi-

�ed to deviate between the make and buy alternative. For instance, if outsourcing is

chosen, savings on invested capital can be deployed alternatively (opportunity cost

added to inhouse alternative), transportation might be handled by the supplier (dif-

ferential in transportation costs added to inhouse alternative), orders will arrive in

smaller quantities (di¤erential in obsolescence costs added to inhouse alternative), a

higher price needs to be paid compared to internal production cost alone (di¤erential

in price added to outsourcing alternative). Hence, adding up the cost increments

for each alternative clari�es whether making or buying is cheapest. In accordance

with Williamson (1975; 1979; 1981; 1985) and Porter (1980), an essential point in

Ellram and Maltz (1995) is thus that a simple comparison of internal production

cost with price might lead to the wrong outsourcing-decision, simply for the reason

that important and inevitable extra costs related to each alternative are not taken

into consideration.

Figure 7 below illustrates the model for outsourcing decision-making based on TCO.

<Insert �gure 7 here>

32

4.2.3 Analysis

Exploring the ontology of TCE, this theory builds fundamentally upon realism al-

though certain idealistic traits are incorporated into the framework. Williamson does

not question whether an objective world exists, but merely how much of the world

the decision-maker knows and can comprehend. That is, the assumption of bounded

rationality implies that the decision-maker has an imperfect understanding of the

characteristics of the particular transaction at hand. Yet, had he been provided with

perfect information and could he attain complete comprehension hereof, he would

be able to infer how the world works and thereby decide the optimal governance

structure to put in place. Hence, Williamson believes that an ideal arrangement

for every transaction does exist and will be independent of the decision-maker�s

perception. However, through the assumption of bounded rationality, he simulta-

neously recognizes that this optimal governance structure might never be known to

the decision-maker. Consequently, the optimal governance structure depends on the

decision-maker�s subjective perception of the transaction; two independent agents

facing the same transaction might not perceive the same institutional arrangement

as optimal for the very reason that they comprehend the same information di¤er-

ently. Thus, in TCE the optimal governance structure needs to be seen as a function

of the decision-maker�s bounded rationality.

Turning to epistemology, the source of knowledge to the decision-maker in TCE

is, in every respect, rationalism. Referring to the criterion (Intuition/Deduction),

the theoretical setup originates from intuited premises (assumptions) of how the mar-

ket and hierarchy function. On the one hand, the market is fundamentally believed

to be an e¢ cient governance structure for the intuitive reason that direct incentives

are provided through the market pricing mechanism. Additionally, human behavior

is assumed to be �awed by opportunism, thus intuitively imposing risk on potential,

transaction-speci�c, investments made in the relationship. On the other hand, the

hierarchy is believed to su¤er from inherent ine¢ ciency based on the rational con-

viction that this governance structure is naturally encumbered by a lack of direct

incentives, which in turn causes problems with moral hazard. Following through

on the criterion (Intuition/Deduction), these intuited premises subsequently serve

as the basis from which propositions about the transaction cost minimizing ability

of each governance structure in the TCE framework is deduced. Consequently, the

characteristics of each sourcing alternative available to the decision-maker in TCE

is constructed from decomposition analysis driven by the correspondence theory of

truth. First, knowledge is created based on intuition alone and subsequently knowl-

edge is deduced from these intuited premises using mathematical and hence rational

33

principles.

In total, it is important to understand that the combination of rationalism and

realism makes it credible to use the correspondence theory of truth to justify �

or prove the correctness of � true beliefs about very complex objects such as hu-

man behavior (opportunism) and relational interdependency caused by transaction

speci�c investments. For instance, accepting rationalism as a plausible way to gain

knowledge about the world, correspondence between the decision-maker�s belief that

the supplier is believed to be opportunistic, and the actual object, the supplier is

opportunistic, can be obtained from deduction alone simply by intuitively grasping

that human behavior, under the right circumstances, cannot be any di¤erent. Thus,

contrary to corporate planning, abstract phenomena such as human behavior and

relational interdependency become very important decision-variables for a decision-

maker, who relies on TCE when evaluating the make and buy alternatives.

The question then becomes to what extend TCE challenges the ontology and epis-

temology in accounting. As explained in section 4.2.2, TCO is primarily concerned

with extending the collection of cost information in order to provide more relevant

cost estimates for outsourcing decision-making. Thus, in continuation of account-

ing under corporate planning, TCO implicitly builds on the premise that the truth

about the world can be (partly) discerned if only su¢ ciently precise accounting in-

formation is reported. Therefore, realism remains the underlying ontology of TCO

and is never as such contested.

With regard to epistemology, the descriptions from section 4.2.2 reveals that

the underlying epistemology of TCO is in every respect empiricism. Yet, building

on empiricism, it is less obvious as to what extent TCO can support the ideas

found in TCE. According to section 4.2.1, transaction costs are caused by planning,

adapting, and monitoring task completion, which, at �rst, seems to �t the pre-

transaction cost, transaction cost, and post transaction cost elements in the TCO

model perfectly. However, a fundamental di¤erence between TCE and TCO exists.

TCE is abstract and concerned with the transaction cost minimizing ability of each

governance structure while TCO is concrete and focuses on the size of the transaction

costs. For tractability, the TCO model thus analyses each sourcing alternative as a

static relationship that will remain in place after a given governance structure has

been chosen. The implication is that the transaction costs that are reported upon,

to a large extent, come to resemble the expected logistical and administrative costs

under each alternative, since these costs are the easiest to quantify and measure.

In particular, the outsourcing alternative is analyzed as if the third-party sup-

34

plier will never break the relationship or behave unreasonably opportunistically after

the contract has been established, and so the costs allocated to the outsourcing al-

ternative primarily concern supplier selection, supplier education and transfer of the

good/service from the supplier to the company. However, as explained earlier, the

very essence in TCE when deciding on the optimal governance structure is to con-

sider the potential risk associated with the transaction speci�c investments which

the company carries into the relationship. Therefore, if TCO really builds on TCE,

as it claims to do, it needs to account for the potential loss that the company faces

on its transaction speci�c investments insofar the supplier behaves opportunisti-

cally and perhaps exits from the relation after contract initiation. Still, the cost

of risk is, ceteris paribus, very di¢ cult to measure as it necessitates assessing both

the likelihood of the supplier behaving opportunistically as well as the value of the

transaction speci�c assets in alternative use. The same problem characterizes the

make alternative, where TCO encompasses the expected costs related to produc-

tion, logistics and administration, yet fails to explicitly consider the expected costs

of ine¢ ciency assumed to naturally encumber the hierarchy in TCE.

Realizing these di¢ culties, it is important to stress that these are not conceptual

problems with TCO, since nothing theoretically hinders the expected costs of ine¢ -

ciency and a risk premium to be added to the make and buy alternative, respectively.

Instead, these problems are caused by the underlying ontology and epistemology of

management accounting, as the combination of realism and empiricism makes it

di¢ cult to apply the correspondence theory of truth to justify true beliefs about

the abstract objects represented in TCE. Speci�cally, opportunistic behavior and

relational interdependency caused by transaction speci�c investments, which are

justi�ed on the basis of intuition in TCE, cannot easily be justi�ed directly from

empirical observations (historical accounting data) in the TCO model. Perhaps a

track record of the supplier�s earlier engagements can provide some insight, yet it

will most likely be di¢ cult to use this to validly establish the probability of him

behaving opportunistically in the particular relationship under study. As such, the

problem is that no close correspondence between the given empirical observations

(track record) and the actual object (opportunistic behavior) exists, for which rea-

son management accounting, ceteris paribus, cannot accept the belief as infallible

knowledge. To use the correspondence theory of truth to make a valid empirical as-

sessment requires the problem to be decomposed into smaller parts, possibly needing

a physiological understanding and observation of the supplier�s mind. Since such

information, for obvious reasons, is not available, TCO shies away from the problem

35

by conservatively leaving out the cost of risk in the evaluation of the outsourcing

alternative.

When compared to the era of corporate planning, it is, thus, vital to understand that

management accounting�s role as the most important provider of relevant informa-

tion for outsourcing decision-making has changed and is no longer absolute during

the era of the Big Bang. Instead, a gap has opened where the knowledge deemed

necessary for optimal outsourcing decision-making in the strategy literature cannot

be fully satis�ed by the tools available in the management accounting literature,

because TCO can only partially incorporate the complexity in TCE without com-

promising accounting conservatism as to what can be regarded as valid knowledge.

However, it deserves notice that the gap is not immense, since TCE is fundamen-

tally concerned with minimizing costs for what reason management accounting �in

spite of its short-comings �still has a very central role to play in the provision of

information for outsourcing decision-making.

4.3 The era of the Bandwagon: Early 1990s - Early 2000

Where the 1980s instigated outsourcing as a viable strategy, it was during the 1990s

that contracting out really gained momentum (Morgan 1999). Positive experiences

from early outsourcing arrangements encouraged other companies to follow (Lacity

and Hirschheim 1993) and initiated a new era in outsourcing history, the era of

the bandwagon (Hätonën and Eriksson 2009). The thoughts carried into business

practice were spawned by the competition between a new theory of the �rm, the

Resource-based View (RBV), and the value chain framework �an extension to the

already established Industrial View (IV).

As such, both theories throughout the 1990s played an essential role in shaping

the perception of outsourcing as a means of achieving strategic objectives in practice

and led managers to reassess their business strategy. Cost e¢ ciency was no longer

the prime of outsourcing. Instead, companies started seeking external skills, compe-

tences and knowledge to provide value to more complex and strategically important

organizational processes. As a nomenclature for this, strategic outsourcing emerged

as a new buzzword (Quinn and Hilmer 1994) which, unlike traditional outsourcing,

emphasized that more strategic oriented functions were being outsourced based on

a closer relationship with suppliers.

36

4.3.1 Strategy part 1: The Industrial View and value chain analysis

The fundamentals of IV is laid out in Porter (1980), where Porter argues that the

industry and the relative position the company holds herein are to be considered

as the primary drivers of pro�tability. Speci�cally, �ve structural characteristics

(forces) are identi�ed to explain industry attractiveness, including entry barriers,

threat of substitution, bargaining power of buyers, bargaining power of suppliers,

and rivalry among competitors in the industry. Hence, the strength of these forces

by themselves and in combination will decide the overall industry pro�tability. Fur-

thermore, taking these characteristics into account a �rm can gain valuable insight

into the dynamics of the industry and pinpoint possible competitive positions from

which it can defend itself or in�uence key industry determinants to improve company

pro�tability.

Building on these thoughts, Porter (1985) introduces the value chain framework

�the second cornerstone in IV. In addition to the �ve structural industry forces,

this framework sees the attractive position within an industry as a result of a com-

petitive advantage, which stems from the company�s ability to handle the discrete

activities making up the company value chain, such as products being assembled,

salespeople making sales visits, or orders being processed. Porter stresses that often

complex interdependencies between these activities exist for which reason opportu-

nities for optimization and problems of coordination between activities are formed.

Therefore, the company can create a competitive advantage by following one of two

generic strategies, low cost or di¤erentiation22, since an above normal return can be

obtained either by managing value chain activities at a collectively lower cost or in

a way that creates customer value, thus allowing for a price premium to be charged

(Porter 1991).

The essential question in IV is hence how the company can generate and sustain a

competitive advantage over time, or, put di¤erently, why some �rms are simply able

to perform particular activities at a lower cost or in ways that create superior value

to others.23 To answer this question, Porter introduces the concept of structural

factors, which essentially represent the structural di¤erences among competitors in

the cost of activities.24 As such, many structural factors are identi�ed, yet to illus-

22Porter also introduces a third generic strategy called focus, which is, however, nothing but aspecial version of either low cost or di¤erentiation. For the purpose of this paper, I will thus refrainfrom discussing this strategy any further.23According to Porter (1985, pp. 16-17), the two generic strategies are ceteris paribus not

compatible. That is, if a �rm engages in each generic strategy at the same time there is a greatrisk of being "stuck in the middle".24To be precise Porter (1985, p. 70) refers to structural factors as cost drivers. However, to

37

trate the point, I will single out a few of particular importance to outsourcing. For

instance, economies of scale can be a structural factor, since obtaining economies

of scale in a given activity holds the potential to reduce costs. Likewise, learning

and spillovers can be a structural factor for a given activity, as the company over

time can strive to achieve learning curve e¤ects and thus improve cost e¢ ciency.

Additionally, managing the linkages between activities within the company�s inter-

nal value chain or between the company and the suppliers/buyers (vertical linkages)

might constitute structural factors which need careful attention in order to ensure

a smooth work�ow without costly extra steps or unnecessary interruptions.

Also, Porter (1991) points out that moving to the level of structural factors sheds

light on the important question of sustainability. In general, the mix and signi�cance

of individual structural factors varies by activity, by �rm, and by industry. There-

fore, the sustainability of a competitive advantage vis-à-vis rivals depends on the

number of competitive advantages in the value chain and, especially, the particular

structural factors underlying each one. That is, a sustainable competitive advantage

must be based on those activities in which a �rm has proprietary access to scarce

resources (e.g. skills, patents, assets, distribution networks, etc.). The durability

of an advantage based on learning, for instance, depends on the ability to keep the

learning proprietary, while the sustainability of advantages due to timing of factor

positions depends on factor market imperfections.

4.3.2 Strategy part 2: The Resource-based View

In essence RBV builds upon the work in Penrose (1959) on how �rms make decisions

on what to produce at what price, and how �rms move from one product and market

to another. Hence, in RBV the �rm�s ultimate objective is generally taken to be the

achievement of above-normal returns (e.g. Wernerfelt 1984; Barney 1986). Collect-

ing such returns requires the �rm to either obtain a low cost position in the market

when selling identical products in comparison to competitors (Ricardian rents), or

lowering competition by di¤erentiating �rm products from the rest of the market

(monopoly rents) (Barney 1991; Peteraf 1993). Thus, the critical problem faced by

the �rm is how to maintain the distinctiveness of its product, or, for identical prod-

ucts, its low cost position, while not investing so much in obtaining this di¤erence as

to destroy the above-normal returns (Barney 1986). Distinctiveness in the product

o¤ering or low cost is tied directly to distinctiveness in the inputs �resources �used

to produce the product and include all assets, capabilities, organizational processes,

�rm-attributes, information, knowledge, etc. controlled by a �rm.

avoid confusion with the de�nition of a cost driver in ABC, I will refrain from this usage.

38

Hence, RBV emphasizes resources internal to the �rm as the principal driver of

�rm pro�tability and strategic advantage. It sees the returns earned as resulting pri-

marily from the acumen or luck of the �rm in acquiring, combining, and deploying

resources (Wernerfelt 1984; Prahalad and Hamel 1990; Barney 1991; Conner 1991).

Speci�cally, a resource with the potential to create a sustained competitive advan-

tage needs to ful�l the criteria put forth in Barney (1991). That is, it must enable

the �rm to meet factors critical to success in its business environment (value); it

must be rare among a �rm�s current and potential competitors to minimize identical

exploitation to that of the �rm (rarity); it must not be easily obtainable for competi-

tors (imperfect imitability); and �nally, there cannot exist any strategic substitutes

for this resource that are valuable but neither rare nor imperfectly imitable, since

this will make other companies capable of implementing the same strategy with dif-

ferent resources (no strategic substitutes).

When considering RBVs in�uence on outsourcing practice, particular interest needs

to be devoted to that part of the RBV literature called the core-competence approach

developed by Peters and Waterman (1982) and Prahalad and Hamel (1990) among

others (Lonsdale and Cox 2000). Speci�c to this approach, and undoubtedly one

of the reasons for its great impact on business practice, is the close resemblance to

typical management guru literature characterized by catchy buzzwords and phrases

such as �focus on your core competences, and outsource the rest� (Prahalad and

Hamel 1990). Still, the message of the core-competence approach is essentially no

di¤erent from the rest of the literature on RBV; the company needs to obtain, create

and develop certain capabilities in order to stay competitive (McIvor 2005). That

is, the sources of competitive advantage are to be found in management�s ability to

consolidate corporate-wide technologies and production skills into competences that

empower individual businesses to adapt rapidly to changing business opportunities.

Thus, very similar to the de�nition of a resource that holds the potential to create a

sustained competitive advantage in Barney (1991), Prahalad and Hamel (1990) sug-

gest that a core competence needs to meet the following three tests: it must enable

the organization to provide a fundamental customer bene�t and make a contribution

to customer perceived value (customer value); it must be competitively unique and

substantially superior to other competitors (competitor di¤erentiation); and �nally,

it must provide potential access to a wide variety of markets �i.e. act as a gateway

to future products and services (extendibility). In e¤ect, a core competence concerns

those resources that are fundamental to a company�s strategic position. Instead of

developing a strategy based on thinking only in dominating markets (strategic busi-

39

ness units), the core competence approach holds that it is more bene�cial to think in

terms of core competences, which will segment the organization in a totally di¤erent

way.25

This stands in clear contrast to IV which, as explained in section 4.3.1, holds the

view that the company can only obtain above normal returns by managing value

chain activities at a collectively lower cost or in a way that creates buyer value and

hence allows for a price premium to be charged. In this respect, Porter (1991) ar-

gues that RBV is nothing but a re�nement to IV, since resources cannot be valuable

in and by themselves, but only because they allow �rms to perform activities that

create advantages in a particular market. Hence, a resource is simply seen as a

function of discrete activities and arises either from performing activities over time,

acquiring them from outsiders or some combination of the two. For this reason,

IV holds that the basic unit to be analyzed is the discrete activity known from the

value chain framework.

Indeed, and as explained earlier, proponents of RBV agree that resources (or compe-

tences) are a complex concept and include more than the mere activities conducted

within the company value chain. Yet, they still argue that the basic unit of a com-

petitive advantage is to be found at the resource level and thus that RBV is an

independent theory of the �rm. Speci�cally, they reason that IV simpli�es matters

too much when the process of identifying and developing the required assets to con-

struct advantageous activities involves nothing more than choosing rationally among

a well-de�ned set of investment alternatives; if assets are not already owned, they

can always be bought (Teece et al. 1997). From the description above, RBV is not

in agreement with this and stresses that resource endowments are �sticky�. At least

in the short run, companies are to some degree stuck with what they have. That

is, companies simply lack the capacity to constantly assure a timely development of

the resources necessary to compete �either because the resources needed are highly

complex (Dierickx and Cool 1989), or because they consist of intangible assets, such

as tacit know-how, which cannot be bought in the factor market (Teece 1976; 1980).

25Notice that the core competence approach bears some resemblance to the world class man-ufacturing literature (e.g. Jazayeri and Hopper 1999), since focus essentially is with making thecompany lean (focus on what is core and can be carried out to perfection). However, in this articleinterest will only be with the core competence approach.

40

4.3.3 Management accounting: Strategic Cost Management

The management accounting response to the developments in strategy has been

termed Strategic Cost Management (SCM).26 Originally, the concept dates back to

the early 1980s and the work of Simmonds (1981) albeit the concept has since then

been developed by academics such as Bromwich, Govindarajan, and Shank (see

e.g. Shank and Govindarajan 1989; Bromwich 1990). However, so far no agreed

de�nition of SCM exists, yet Shank and Govindarajan (1994) describes it as:

The blending of the �nancial analysis elements of three themes from the

strategic management literature �value analysis, strategic positioning

analysis, and cost driver analysis (Shank and Govindarajan 1994, p.

xiii).

In general, SCM embraces the idea of incorporating strategic re�ections on competi-

tors, customers, and suppliers into the cost analysis to develop superior strategies

that can help the company gain sustainable competitive advantages in keeping with

the theories from the strategy literature. Nevertheless, in the management account-

ing literature it has, so far, only been sporadically touched upon in case study

illustrations while a comprehensive scholarly work linking together strategy and ac-

counting on a theoretical level has still to emerge (Shank 2007; Lang�eld-Smith

2008). For this reason, it is di¢ cult to give a description that covers all the aspects

of SCM, and hence for the purpose of this paper, focus will be kept on how Porter�s

value chain framework has been treated, as the ideas from RBV have yet to �nd

their way into the SCM literature.

However, nothing has been written speci�cally on outsourcing while the main-

stream treatment of make-or-buy in the management accounting literature, even

today, is often reduced to comparing alternatives on the basis of credible cost al-

location, thus leaving the more profound strategic re�ections for a side note (e.g.

Horngren et al. (2005, pp. 255-260)). Still, instead of only pointing to this obvi-

ous fact, I will try to combine the explanations o¤ered in Porter (1985) and Shank

and Govindarajan (1989) to illustrate how SCM can be incorporated into the value

chain framework in relation to outsourcing decision-making. In this respect, I will

also draw on ABC from Kaplan and Cooper (1998) due to the natural correspon-

dence between the activity level in ABC and the value chain framework.

Following the general concepts of IV, Shank and Govindarajan (1989) illustrate how

26Sometimes SCM is also referred to as strategic management accounting (SMA) (Lang�eld-Smith 2008).

41

the initial step in strategic cost management is a �ve-forces-study of the industry

in which the company operates. Accordingly, management accounting takes on the

role of empirically investigating such factors as the size and number of suppliers and

buyers, average cost of products, types of substituting product, etc. However, man-

agement accounting�s role at this stage is solely to provide information, while the

weighting of the relative importance for decision-making is left to strategy.27 Conse-

quently, the results from the industry analysis may immediately clarify whether low

cost or di¤erentiation is the right strategy to follow, in which case the role of SCM

becomes the identi�cation of how best to implement this generic strategy. Still, if

the results from the industry analysis are ambiguous, SCM normally takes on the

additional challenge of trying to reveal information that can rule in favor of either

alternative. However, to make the description tractable, I will, in what follows next,

centre my thoughts around the case where the company has already decided on low

cost as the appropriate choice of strategy.28 Particularly, I will stick to the outline

provided in Porter (1985) and divide SCM into a sequence of four steps.

First, the company needs to de�ne its value chain. Hence, it can start out with

the generic value chain from �gure 1, fromwhich each generic category can be divided

into discrete value activities speci�c to the company. In particular, de�ning relevant

value activities requires that activities with discrete technologies and economics are

isolated, where broad functions such as manufacturing or marketing are subdivided

into activities. The appropriate degree of disaggregation depends on the economics

of the activities and the purposes for which the value chain is being analyzed. In

this respect, the basic principle is that activities should be isolated and separated

if they (1) have di¤erent economics (i.e. cost behavior), (2) have a high potential

impact on di¤erentiation, or (3) represent a signi�cant or growing proportion of cost.

For outsourcing in particular, it will thus be natural to aim at a successively �ner

disaggregation of the closest activities on either side of the activity considered for

outsourcing, in order to gain accurate insight into how outsourcing will a¤ect the

dynamics of the value chain.

Second, having de�ned the activities in the value chain, assets and operating costs

need to be assigned accordingly. The necessity to assign assets to value activities

re�ects the fact that the amount of assets in an activity and the e¢ ciency of asset

utilization are frequently important to the activity�s cost. Speci�cally, operation

27In fact it can be discussed whether information provision for industry analysis in IV is a taskfor accounting. Yet, in keeping with the very broad de�nition of accounting from section 2, I �ndit appropriate to do so.28What is de�ned as strategic cost analysis in Shank & Govindarajan (1989, pp. 40-41) is

described in relation to a generic strategy based on low cost in Porter (1985, pp. 62-118).

42

costs should be allocated to the activities in which they are incurred, while assets

should be assigned to the activities that employ, control, or most in�uence their use.

Porter (1985) doesn�t specify which costing principles to use29, yet ABC seems like

a natural choice of method, as this particular technique centres around allocating

resource costs to activities using so called resource cost drivers �the �rst basis for

distribution in ABC.

The third step is to identify the structural factors that explain variations in the

cost of the value activities compared to competitors. That is, the company�s cost

position results from the cost behaviour of its value activities, which, in turn, are

fundamentally shaped by a number of underlying structural factors. To understand

and to deal with these factors are thus important. Hence, in keeping with section

4.3.1, the company has to search for such underlying causes of cost di¤erences as

economies of scale, learning and spillovers, and the handling of the linkages in the

value chain.

Having completed the three initial steps, the company, �nally, needs to evaluate

the results of these to decide where a competitive advantage can potentially be gener-

ated either through controlling the structural factors better than the competitors or

by recon�guration of the activities in the value chain. Focusing on outsourcing, the

company must therefore evaluate the activity considered for contracting out on the

basis of the structural factors identi�ed as well as the costs allocated to it through

ABC. That is, the company has to question how outsourcing will a¤ect the other

activities in the value chain both internally and externally. For instance, it might

be the case that the costs allocated to the inhouse activity are higher than the price

o¤er from the third-party supplier. Yet, as a result of linkages between activities (a

structural factor), outsourcing might in�ict extra costs on the rest of the activities

in the internal value chain due to additional inspections, adjustments to the current

activity setup, or the like. As such, calculating these extra costs e¤ectively amounts

to determining TCO as discussed earlier in section 4.2.2. However, SCM goes even

further than this. Shank and Govindarajan (1992) illustrate that it can be essen-

tial for decision-making to also contemplate the external part of a company�s value

chain. For example, in the context of outsourcing, it is not di¢ cult to imagine that

contracting out a given activity in the internal value chain might generate extra costs

in a later activity belonging to one of the company�s customers. In turn, this might

lead the customer to demand a price reduction from the company which should be

taken into account when evaluating the outsourcing alternative.

29Porter merely states that "the costs and assets of shared value activities should be allocatedinitially to the value chain of the business unit using whatever methodology the �rm currentlyemploys, typically based on some allocation formula" (Porter, 1985, p. 66).

43

Figure 8 below sums up the steps in SCM and the interdependency between ac-

counting and strategy when related to outsourcing decision-making.

<Insert �gure 8 here>

4.3.4 Analysis

As can be seen from the description of the era of the bandwagon, two theories of the

�rm, IV and RBV, have dominated the strategy literature, while only IV has been

treated in the management accounting literature. In what follows next, the initial

part will thus study the ontology and epistemology of the two theories of the �rm,

after which attention will be brought to accounting. In relation to this, I will try to

position management accounting�s role within the overall IV framework. Finally, I

will argue why RBV has yet to �nd its way into the management accounting litera-

ture.

When studying the ontology of IV, it becomes clear that a shift has occurred com-

pared to TCE. Essentially, TCE focuses on how the company optimally reacts using

the market or hierarchy when faced with given external conditions. In IV, external

industry factors beyond the company�s control still a¤ects its relative position and

pro�tability, yet the company can to some extent reposition itself by constructing

a certain identity based on either low cost or di¤erentiation. That is, the company

does not only react to face given conditions in�uencing company pro�tability but

can also act upon these circumstances in advance to change the way they a¤ect the

company. Realism is, however, still the underlying ontology of IV, although it is not

the pure form where everything is unalterable. Instead, realism de�nes the scope of

possibilities within which the company can construct and shape its own reality, and

so idealistic traits are present in IV cf. criterion (Constructivism). For instance,

if the industry conditions in every respect favor a low cost position, the ways to

realize this low cost position may still be open-ended. Perhaps the company can

move swiftly in the factor market to accumulate certain scarce production factors

that give the company a �rst mover advantage over competitors in the industry. Or

perhaps the company can create switching costs for the main customers by binding

them to the company through intense marketing and perfection of company routines

(vertical linkages). As a result, the company might be able to realize economies of

scale in production and reduce the cost of throughput. Indeed, the industry con-

ditions may even leave unspeci�ed the optimal generic strategy to follow, and so it

44

might be the case that a low cost strategy can be just as pro�table to pursue as a

strategy based on di¤erentiation.

Turning to the ontology of RBV, the trend towards idealism becomes even more

evident. In RBV, the centre of attention is kept on nurturing and developing valu-

able resources. In particular, what is believed to create sustainable pro�t potential

coincide with the intangible aspects that bind together the company assets and

processes, since the intangible, and thus less understood, spawns uniqueness that

is di¢ cult for competitors to imitate. That is, in accordance with IV, RBV holds

the view that some objective reality of assets and industry conditions subsist, yet

in RBV this reality is merely of limited importance for long-run pro�tability. Only

when the company �perhaps unconsciously �constructs itself by spinning its tan-

gible and intangible assets into a web of valuable resources di¢ cult for competitors

to unravel, can a competitive advantage and sustainable pro�t be obtained. Thus,

when seen in the light of criterion (Idealism) and criterion (Constructivism), the

decision-maker�s very perception of company assets and the creativity he holds as to

how these assets can be combined, let alone developed, are essentially what drives

value creation. Consequently, the greatest di¤erence between the ontology in IV

and RBV is that the role of idealism in RBV has been taken much further.

Moving on to epistemology, the source of knowledge for a decision-maker relying

on IV is primarily rationalism, although empiricism also has a role to play. In par-

ticular, IV provides a theoretical frame for structuring outsourcing decision-making,

which is founded on a sequence of intuitive assertions of how an industry and a

company work. Referring to the criterion (Intuition/Deduction), the basic intuited

premise is that industry pro�t potential is essentially decided from 5 competitive

forces. This in turn leads to the deductive argument that a company can only sus-

tain pro�t in an industry if it holds a competitive advantage by providing customer

value on the basis of di¤erentiation or low cost. Building on this line of assertions,

the theoretical conception of an activity is thus introduced to rationally reason how

a company and its supply chain are best studied as a sequence of interdependent ac-

tivities in order to clarify where true value creation takes place. Having emphasized

the intuited premises on which IV builds, however, it is important to realize that

a decision-maker relying on Porter�s framework is not intended to reach a decision

using intuition and deduction alone. In fact, IV is partly modeled in preparation

for empirical investigation, where data material can be collected for several of the

theoretical constructs making up the industry and value chain framework, respec-

tively. Nevertheless, since this coincides with the treatment of SCM below, I will,

45

in an attempt not to anticipate events, refrain from an in-depth discussion of this

here. Consequently, it su¢ ces to say at this point that the epistemology of IV is

based on both rationalism and empiricism.

Turning to the epistemology of RBV, the source of knowledge for decision-making

is much more clearly founded on rationalism. As already discussed in relation to

ontology, RBV highlights the importance of the intangible and subtle aspects. Em-

pirically, observation of these can be challenging. Thus, rationalism becomes the

viable mechanism for knowledge creation, since the level of abstraction can better

be handled by this epistemology. In particular, building on rationalism and idealism

in combination, the coherence theory of truth functions as the underlying principle

that decides when true beliefs have been justi�ed; in accordance with section 3.3.2,

a belief coheres with a set of other beliefs if, and only if, it is entailed by members

already in the set.

To clarify the use of the coherence theory of truth for an outsourcing situa-

tion in the context of RBV, let me exemplify a possible chain of reasoning for a

�ctional car manufacturing company that considers outsourcing one of its depart-

ments, X, to reduce costs. Imagine that the company relies fundamentally on the

vision that it wants to sustain a leading industry position as a producer of fast cars.

In this respect, the vision becomes the �rst justi�ed true belief. Building on this,

the decision-maker perhaps forms the belief that in order to build fast cars, the

company needs to focus its processes around engine development and production,

as a good engine intuitively seems to be a precondition for a fast car. This makes

engine development and production core competences, and since the belief coheres

with the �rst belief (the vision), it is included in the set of true justi�ed beliefs with-

out further notice. Continuing along these lines, the decision-maker may intuitively

be led to the rationale that department X, which presumably delivers a speci�c

subcomponent to the engine, is of particular importance; not because of physical

characteristics of the component it supplies but because of the tacit know-how that

binds it to the rest of the departments engaged in engine production and develop-

ment. As the impact of this tacit know-how cannot be empirically investigated, the

belief rests on intuition alone and is included in the set of justi�ed true beliefs for the

reason that it coheres with the two other beliefs in the set already. Consequently,

the decision-maker concludes that department X is best kept in-house insofar the

company wants to sustain a leading position in manufacturing fast cars. Hence,

although the chain of reasoning in this case is simpli�ed and perhaps pushed a bit

to the extreme, it clearly illustrates how the coherence theory of truth can be put

to work in RBV, where the intangible aspects are regarded as ever so important, to

46

create knowledge for an outsourcing decision-maker.

Having analyzed the two streams within strategy, I will turn to study the develop-

ments in management accounting. As mentioned earlier, SCM constitutes a central

part of IV, and therefore it is important to be conscious of what adheres under

strategy and management accounting, respectively. Focusing on the descriptions in

section 4.3.3, the ontology of SCM is indeed realism, where the job is to reveal and

learn the current state of a¤airs of the industry and the company�s value chain. In

this respect, SCM during the era of the bandwagon remains faithful to the inher-

ent ontology that has marked management accounting throughout the history of

outsourcing.

However, looking at the epistemology of SCM, it becomes clear that empiricism

is no longer the only source of knowledge. Instead, rationalism has also come to play

a central part as a means to discern industry and company speci�cs. As described

in section 4.3.3, the preliminary step in SCM is to provide data for the industry

analysis, where management accounting�s role is to empirically investigate various

industry aspects of a quantitative and qualitative nature. The second and third

step of SCM concern de�ning the company�s value chain, disaggregating it into

activities, and appropriately allocating costs to each of these activities. While both

the de�nition of the value chain and the disaggregation into activities can be partly

grounded in empirical observations, some important matters might not be readily

observable but can indeed only be identi�ed and distinguished using intuition and

deduction. Hence, during these phases, knowledge is created both on the basis of

rationalism and empiricism. Next, rationalism plays a very important role in the

identi�cation of the structural factors underlying the activities in the value chain,

since aspects such as learning and spillover e¤ects along with vertical linkages can

only be recognized let alone understood on the basis of intuition. Following through

on this, intuition also plays a prominent role during the �nal stage of the "what if"

calculations that SCM supplies the outsourcing decision-maker with; the cost-bene�t

analysis of recon�gurating the value chain or improving the handling of structural

factors calls for inferences based highly on creativity and intuition.

Having said this, however, it is interesting to see how the actual accounting cal-

culations related to IV in Shank and Govindarajan (1992) are somewhat marked by

the inherent conservatism in accounting and only partially succeed in incorporat-

ing the thoughts from Porter (1985).30 That is, although Shank and Govindarajan

30Shank (1992) is the only paper that contains an example of how the actual accounting calcu-lations can take form in the context of IV.

47

(1992) take notice of the importance of intangible assets which span and bind to-

gether several activities, it never becomes a focus area for their calculations. Instead,

their value chain analysis remains rather static and essentially amounts to deciding

between alternatives on the basis of costs allocated to activities in the internal and

external value chain, respectively. Thus, it is never questioned whether structural

factors, such as learning curve e¤ects, can be triggered over time to reduce costs at

given activities thereby changing the entire starting point of the calculations in the

analysis. In this respect, it is as if a discrepancy exists between two of the foremost

proponents of SCM in the management accounting literature, Shank and Govindara-

jan, and the original strategic framework, Porter (1985). The inherent conservatism

in the management accounting literature simply restricts the encompassing of such

considerations into the calculations due to a believed lack of credibility, and so it

can be contested whether the epistemology of SCM in reality re�ects the full scale

of rationalism pleaded by Porter (1985).

Still, accepting that, at least in theory, the epistemology of SCM ought to rely on

both rationalism and empiricism, attention should be brought to management ac-

counting�s role in the overall knowledge creation within the IV framework. Building

on the earlier conclusion that the ontology of IV relies on both realism and ideal-

ism while the epistemology has facets from rationalism and empiricism, knowledge

creation in Porter�s framework can basically be separated into two parts; an initial

accounting part and a subsequent strategic part. As such, SCM represents the part

of IV, where preliminary knowledge for decision-making concerning industry and

company speci�cs are prepared on the basis of intuition together with empirical ob-

servations using the correspondence theory of truth. This knowledge is important,

yet not necessarily decisive for the recon�guration of the value chain and the posi-

tioning of the company in the industry. Instead, it enters into the second part of IV,

the strategic part, where the decision process revolves around the coherence theory

of truth, and where intuition is coupled with idealism to decide on which generic

strategy to choose and which activities to focus on.

Having discussed SCM in the context of IV, it deserves notice why a similar ac-

counting model for RBV has never been outlined. I believe the answer to this

question follows from the discussion of accounting conservatism already mentioned

above. Since RBV is much concerned with conceptualizing the intangible aspects

of a competitive advantage through the theoretical constructs of either valuable re-

sources or competences, the primary objective of RBV is not practical applicability

but rather a thorough theoretical understanding of the underlying characteristics

48

of a competitive advantage. As such, this contrasts somewhat with IV, where the

accounting part, and thus some degree of practical applicability, has been designed

into the framework from the very beginning. Hence, the problems already present in

the actual accounting treatment of IV in respect of incorporating intangible aspects

into the calculations is further emphasized in RBV where subtleties, such as tacit

know-how and the like, are even more predominant. In relation to this, accounting

conservatism prohibits a further intake of rationalism as the source of knowledge.

As a result, accounting essentially cannot provide the information believed to be

relevant for outsourcing decision-making in the context of RBV. Furthermore, the

fundamentally idealistic outlook of RBV implies that accounting information based

on realism can only play a very limited part.

In conclusion to the era of the bandwagon, management accounting�s role as in-

formation provider for outsourcing decision-making has changed. To keep up with

strategy and adapt to the complexity of the strategic frameworks, management ac-

counting has expanded its epistemological basis to partly include rationalism. In

this respect, management accounting models still have a rather important yet not

decisive role to play for a decision-maker relying on IV. Moving on to RBV, however,

the picture changes dramatically, since the epistemology and ontology of manage-

ment accounting in reality clashes with that of strategy where the coherence the-

ory of truth has become the predominant method for knowledge creation. Under

such circumstances, the role of management accounting models in the outsourcing

decision-making process has been severely reduced.

4.4 Connecting the historical dots and looking beyond

Bringing the results from the literature analysis together, reveals that a gap has come

to exist over time between the information that accounting models can supply and

the information deemed relevant for outsourcing decision-making within the strate-

gic frameworks. This is illustrated in �gure 9, below. In particular, this �gure shows,

how the information disparity between accounting and strategy is minimal during the

era of corporate planning as the tangible aspects possible to reveal through empiri-

cal investigation were considered highly important for outsourcing decision-making.

In the course of time, however, and as the strategic frameworks have increasingly

directed focus towards intangible aspects, this disparity is no longer minimal. The

biggest gaps are thus found during the era of the bandwagon. IV emphasizes the

intangible aspects although tangible factors are still given some attention. Further-

more, the framework is partly designed with empirical investigation in mind. The

49

de�nition of SCM from Shank and Govindarajan (1994) thus promises a greater

concern for strategic aspects and a more extensive incorporation of these issues into

the cost analysis. Yet the actual examples of SCM provided in the literature only

partly succeed in incorporating such subtle factors. Hence, there seems to be a dis-

crepancy between the theoretical intentions of strategic cost management, SCM(T),

and its actual execution, SCM(A). Finally, RBV represents the strategic framework

with the highest focus on the intangible aspects of (outsourcing) decision-making.

For this reason, no real treatment of RBV exists in the practically oriented part of

the accounting literature.

<Insert �gure 9 here>

Having outlined this historical gap, it is appropriate to look beyond it and brie�y

touch upon what seems to be the current trend in outsourcing decision-making. In

practice, the di¤usion of outsourcing since the beginning of the new millennium has

created a situation in which outsourcing is no longer regarded as a competitive dif-

ferentiator (Lawton and Michaels 2001). Today, globalization has facilitated trade

on a global basis where markets are no longer distinct entities, protected from each

other by trade and time barriers. Instead, a system has transpired in which national

markets are merging into a single global market (Hill and Jones 2001). Consumers

have become more sophisticated and demanding as they have become more knowl-

edgeable on issues such as price, reliability and availability (McIvor 2005). These

changes represent tremendous challenges to organizations, yet they also provide op-

portunities on a much larger scale than ever seen before. The global resource pool

has simply been made available to all companies, regardless of their industry, geo-

graphical location or company size.

To cope with this, companies to a much larger extend than ever before pursue

�exible and dynamic structures which has caused the boundaries of the �rm to be-

come blurry and fading (Hätonën and Eriksson 2009). In the strategy literature this

trend has given rise to several theories regarding inter-organizational relationship

con�gurations such as virtual organizations (e.g. Davidow and Malone 1992; Gold-

man et al. 1995), network organizations (e.g. Miles and Snow 1986), and shamrock

organizations (Morgan 2003). Common for all of these frameworks is that they can

hardly be de�ned as new theories of the �rm. Instead, they represent a collage of

elements from several distinct theories, including TCE, RBV and IV among others

(Buckley and Lessard 2005). Put together, however, these elements further empha-

size the importance of the intangible and loosely coupled aspects in make-or-buy

decision-making. In particular, Mazzawi (2002) refers to this new form of contract-

50

ing out as transformational outsourcing to stress that outsourcing today is no longer

only about doing the same things a bit better, a bit faster or a bit cheaper. Instead,

transformational outsourcing is about changing the paradigm to something smarter,

more �exible and more streamlined. It is about the continuous strive towards cre-

ating an adaptive enterprise where contracting out can help establish the capacity

and capability to respond rapidly and e¤ectively to the plethora of unpredictable

events that constitutes today�s business environment (Linder et al. 2002; Mazzawi

2002).

Turning to the management accounting literature to look for corresponding de-

velopments in the accounting models, it becomes evident, however, that progress

has stagnated thus con�ning current models to what have already been laid out in

relation to the era of the bandwagon. Hence, even though it is far too early to make

strong conclusions to the developments since year 2000, there seems to be indications

that the gap between the management accounting models and strategy frameworks

has only widened even further and will continue to do so if no appropriate actions

are taken.

5 Conclusion and implications

In summary, the literature analysis in this article has shown that the manage-

ment accounting toolkit has become inadequate as support for strategic outsourcing

decision-making. Where accounting information earlier carried decisive weight in

the �nal decision, a gap has come to exist over time between the information that

accounting models can supply and the information considered relevant for outsourc-

ing decision-making within the strategic frameworks being adopted. The key driver

in this transition has been a growing recognition in the strategy literature of the

important role that intangible aspects play in facilitating long-term company suc-

cess. A recognition that has naturally led to the adoption of rationalism as the

predominant epistemology, since the complexity of the intangible aspects can only

be fully grasped by intuition. Also, it is a recognition, where idealistic traits of con-

structivism follow naturally, as the strategic frameworks are increasingly occupied

with creating �or constructing �new possibilities either by establishing a defendable

position in the market or combining resources in novel ways. Essentially, this has

severely reduced the importance of traditional accounting information in the out-

sourcing decision-making process, since the inherent combination of empiricism and

realism has made it di¢ cult to measure more than the tangible and less complicated

51

aspects.31

The question is then whether this disparity makes up a problem. Have we �nally

reached a stage where management accounting is no longer vital for successful make-

or-buy decision-making which is better left to intuitive non-quanti�able strategic

reasoning? Empirical studies on high failure rates in outsourcing arrangements sug-

gest that this is not the case. For instance, three quarters of U.S. managers surveyed

by the American Management Association reported that outsourcing outcomes had

failed to meet expectations (Bryce and Useem 1998). Additionally, Barthélemy

(2001) reports of a survey based on 50 companies, where approximately 14% of all

outsourcing operations were deemed failures due to hidden costs. Companies said

they entered an outsourcing agreement believing they understood all major costs.

They even agreed that some expenses would be incurred in relation to activities such

as �nding a vendor, drafting the contracts, and managing the e¤ort (i.e. transaction

costs), but they thought these expenses would be negligible. Ex post the outsourc-

ing transition, however, most companies came to experience that these cost were

not negligible �in some cases, they actually halved or cancelled out the company�s

potential savings from outsourcing.

Since company success in the end comes down to pro�t-making, it therefore

seems appropriate that strategic thought is more closely aligned with models of

management accounting than is currently the case. Thus, the real question to ask

is how to minimize the gap between these two areas.

5.1 How strategy can help close the gap

With regard to strategy, especially RBV and the recent developments since year 2000

seem to represent a problematic tendency. In spite of how appealing RBV might

appear from a theoretical perspective, it is a questionable and dangerous path to

follow when model constructs become vaguely de�ned thus leaving much room for

misunderstanding and misuse when carried into practice. Clearly, I do acknowledge

that simpli�ed mappings of reality are needed if we want to make serious re�ections

about the world in order to take appropriate and timely actions, since "without

such intellectual constructs there will be nothing but a blooming buzzing confu-

sion" (Huntington 1998). Still, framework abstraction and complexity are simply

no excuse to shy away from considering how concepts can be operationalized and

31Notice that sporadic attempts of measuring intellectual capital exist in the accounting literature(e.g. Mouritsen 1998; Mouritsen et al. 2001). However, the concepts applied are vaguely de�nedand have yet to emerge as part of a connected whole that can be made operationale. Furthermore,the attempts have mostly been directed towards external reporting or performance evaluation.

52

made tractable. Particularly not when a part of the strategy literature deliberately

relies on seductive buzzwords, such as "core competences", and is written in a cer-

tain action-oriented style to ensure a strong appeal to business practitioners that

bestow the concepts with credibility and make them di¤use quickly into real world

decision-making.

Hence, while abstract theories of the �rm are undoubtedly indispensable for

structuring human thinking and understanding of the outsourcing situation, it is

valid to constantly question whether the current abstraction level can be deceptive

and create a feeling of false security when the frameworks are carried directly into

real world decision-making without further notice. A feeling that at worst can lead

to unfortunate and wrong decisions being made. Therefore, I believe a part of the

RBV literature and the newest trend in outsourcing have taken a wrong turn as

they do not address this issue but instead seek refuge in the academic ivory tower

where concern can be focused on theory and where practice can easily be neglected.

Without taking a stance in the theoretical debate between IV and RBV, it seems

as if IV di¤ers in this respect since the framework has been designed with practical

implementation in mind. Accordingly, accounting models are naturally incorpo-

rated into the overall framework and Porter (1985) already provides some guidance

for the estimation of the many aspects making up the framework. Indeed, this

explicit concern for the concrete use is perhaps all that can possibly be expected

from the strategy literature as the general and long-term perspectives will time

and again carry it away from issues related to actual implementation. Thus, if the

strategy literature could be developed to acknowledge the signi�cance of operational

tractability when thinking up framework constructs and preparing them for further

(empirical) investigation, an important cornerstone will have been put in place. A

cornerstone that will increase the likelihood of �nding common ground for strategy

and management accounting in outsourcing decision-making.

5.2 How management accounting can help close the gap

Where strategy can assist in providing more "friendly" frameworks that allow for

additional analysis, the development of complete instructions as to how the strate-

gic constructs can be made operational should never be a concern for the strategy

literature. Instead, it is better left for management accounting research. For in-

stance, in relation to applying IV in outsourcing decision-making much work still

remains in bridging the gap between theory and practice by re�ning existing models

and developing new techniques for knowledge creation. Nevertheless, and as already

touched upon in section 4.3.3, research into this area has so far been fairly limited

53

and most noticeably amounts to the work done by Shank and Govindarajan dating

back several years. In fact, Shank (2007) points out that SCM in general seems to

su¤er from an "unravelling of the pieces" suggesting a collection of isolated texts

rather than evidence of an emerging discourse:32

By 2000, there was a �fteen-year history of great "beginnings", "pilot"

projects, and "cameo" appearances for SCM, but not much more. It

has been a great topic on the lecture circuit and in cost management

symposia. In military parlance, "it briefs well"! But the topics had not

been gaining traction in mainstream academe or in the corporate world...

(Shank 2007, p. 359).

The reasons for this de�ciency are undoubtedly many, yet a part of the explanation

relates to the results pointed out in this analysis that research on accounting models

in general seems stuck with a paradigm based on empiricism and realism that causes

great di¢ culties in handling the complexity of the intangible aspects found in strat-

egy. Focusing on outsourcing, the inherent conservatism in management accounting

simply makes it di¢ cult to keep pace with the developments in strategy, since only

certain types of information are considered credible for decision-making, which in

turn hampers research on new, and perhaps untraditional, accounting models for

outsourcing decision-making. In other words, management accounting loses terrain

as the decisive tool for strategic outsourcing decision-making, since too much rel-

evance is traded o¤ in exchange for high perceived credibility in the information

provided.

Realizing these serious short-comings, brings up the question whether the current

paradigm is best compromised with traits of rationalism and idealism in order to

keep pace. In my opinion, this is the right path to follow. Yet, it is a path that

should be followed with great caution, as I believe there are limits to how much

a company can construct its identity and/or reposition itself within an industry.

Indeed, management accounting needs to remain the "watchdog" that continuously

checks up on strategic constructs with hard-headed (costing) re�ections to ensure

that the often di¤use strategic objectives at least partially coincide with the real

world. However, in order to ful�l this task, management accounting research needs

to lose its reluctance to deal with matters that do not immediately qualify as in-

controvertible facts derived from empiricism. It simply has to accept that practical

32Seal (2010) takes this even further and stresses that SMA (and hence SCM) has "failed" as anacademic discourse even before it could in�uence the wider managerial discourse.

54

accounting models are no longer meant only for passive information provision but

must take on a new proactive role in the decision-making process, if they are sup-

posed to remain attention directing and carry weight in the �nal decision.

Essentially, what is needed is a more equal relationship between the frameworks

in strategy and the models in management accounting, where claims in strategy

cannot necessarily be forced through on the basis of the argument that accounting

models can simply not comprehend and work with the complexity of the strate-

gic constructs. Hence, when strategy claims something to be true and of high

importance for decision-making, management accounting needs to ask why this is

necessarily the case and try to gauge the realism in the argument on the basis of cal-

culations; calculations that might be somewhat �awed and marked by uncertainty.

Still, involving management accounting on these premises is preferable and will au-

tomatically trigger discussions that can possibly lead to more sober considerations

of each outsourcing alternative, since strategy needs to answer the questions posed

by management accounting and vice versa. Following this path, we will move away

from the current "one-sided" stage, where strategic frameworks a¤ect the models in

management accounting only, to a new "two-sided" stage where accounting models

also in�uence and challenge the thoughts and frameworks in strategy.

In this respect, I believe that academics such as Bromwich, Shank, and Govin-

darajan have taken the critical �rst steps on a journey that can bring management

accounting research towards new and important frontiers. Although SCM is still

somewhat impeded by accounting traditions, the attempt to capture the strategic as-

pects by making management accounting more externally oriented clearly illustrates

how traits of rationalism and idealism can advantageously be build into the account-

ing decision-models to improve the relevance in the accounting information provided

without destroying practical applicability. Management accounting research in this

area thus needs to be restarted and taken to the next level by carefully modifying

the current paradigm with selected traits of rationalism and idealism. Only in this

way can management accounting regain a central position as a relevant information

provider for strategic outsourcing decision-making. If the management accounting

literature is ambitious and strives to make the accountant a business partner in the

decision-making process, it is time to stop speaking about what the hybrid accoun-

tant needs to accomplish and instead start updating his toolkit to assist him in

carrying out his job.

55

6 Further research

Before re�ecting upon further research, it is necessary to point out that the part

of management accounting research concerned with developing decision-models for

practical use studied in the article has so far di¤ered from the rest of the manage-

ment accounting literature. Indeed, both rationalism and idealism thrive in other

areas of management accounting research where the general academic progression of

the �eld over time has been characterized by a steady expansion from accounting�s

initial research links to economics into such areas as sociology, psychology and or-

ganizational studies (Baldvinsdottir et al. 2010). Additional development has been

apparent using the disciplines of mathematical analysis and, perhaps, most notably

philosophy.

Still, where expansions into other research domains at �rst sight might seem

as the obvious way to close the gap between management accounting research and

practice, Baldvinsdottir et al. (2010) argue that the e¤ect will probably be quite

the contrary, since the conclusions drawn from accounting studies so far have, for

the most part, been for further research rather than for practice. Obviously, this

is in every respect problematic if the ultimate purpose of social science research

is to improve life rather than simply to describe and/or understand it. Therefore,

when developing new accounting models for outsourcing decision-making, it is im-

portant that the expansion of the accounting paradigm through further uptake of

rationalism and extension to the current level of realism do not cloud the main re-

search objective; the attainment of models that can be applied to improve real life

decision-making. This is not an easy task to accomplish, yet I believe it is absolutely

crucial.

Keeping this in mind, a �rst step for management accounting research in making ac-

counting models more relevant for practical outsourcing decision-making might thus

be a study of how consultant �rms tackle strategic outsourcing problems. Since

these �rms operate in the mixed zone between strategy and accounting and have

great experience with strategic outsourcing from various situations, such insight will

most properly provide a good preliminary understanding of how and to what extent

the frameworks from the strategy literature are applied in practice.33

Based on this, a second step could be carefully selected case studies of how out-

sourcing decision-making is actually handled in companies �both in relation to the

33According to Shank and Govindarajan (1989) "there is a billion dollar a year market in strategiccost analysis consulting services dominated by such �rms as Bain & Company, Boston ConsultingGroup, Booz Allen & Hamilton Inc., McKinsey & Company, and Monitor, Inc."

56

structure of the decision-making process, the models applied, and the information

collected. In this respect, the objective should not be a simple description of the

procedures followed, if any, but instead a critical examination of how companies ap-

ply and combine the theoretical concepts and techniques. That is, the examination

should seek out answers as to what considerations and assumptions underpin the

outsourcing analysis and which trade-o¤s are made to obtain tractability.

I believe that a better understanding of the strategic frameworks in the light of

practice will assist in elucidating where immediate research into new accounting

models is possible and most crucial to undertake in relation to outsourcing decision-

making. Moreover, a study of practice might help clarify whether the gap between

strategy and management accounting is perceived to be just as big in the real world

as in the academic literature.

Acknowledgements

The author would like to thank professor Hanne Nørreklit, Aarhus School of Busi-

ness, Aarhus University, for her great interest in the project and her valuable criti-

cisms and suggestions along the way. Also, gratitude should go to Professor Falconer

Mitchell, University of Edinburgh, for his insightful comments during the �nal phase

of the project. Any mistakes are entirely the author�s.

References

Alston, W. P. (1998, October). Empiricism. Routledge Encyclopedia of Philoso-phy. http://www.rep.routledge.com/article/P014.

Anderson, P. A. (1983, June). Decision making by objection and the Cuban missilecrisis. Administrative Science Quarterly 28 (2), 201�222.

Anthony, R. A. (1964). Management Accounting. Irwin.

Arbnor, I. and B. Bjerke (1997). Methodology for Creating Business Knowledge(2nd ed.). Thousand Oaks, CA: Sage Publications.

Arnold, U. (2000). New dimensions of outsourcing: a combination of transac-tion cost economics and the core competencies concept. European Journal ofPurchasing and Supply Management 6 (1), 23�29.

Baldvinsdottir, G., F. Mitchell, and H. Nørreklit (2010). Issues in the relation-ship between theory and practice in management accounting. ManagementAccounting Research. Forthcoming.

Barney, J. B. (1986, October). Strategic factor markets: Expectations, luck, andbusiness strategy. Management Science 32 (10), 1231�1242.

57

Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journalof Management 17 (1), 99�120.

Barthélemy, J. (2001). The hidden cost of IT outsourcing.MIT Sloan ManagementReview 42 (3), 60�69.

Berkeley, G. (1710). A Treatise Concerning the Principles of Human Knowledge.New York: Dover Publications, Inc., 2003.

Beyer, C. (2003, February). Edmund Husserl. Stanford Encyclopedia of Philoso-phy. http://plato.stanford.edu/entries/husserl/.

Bonjour, L. (2007, May). Epistemological problems of perception. StanfordEncyclopedia of Philosophy. http://plato.stanford.edu/entries/perception-episprob.

Bromwich, M. (1990). The case for strategic management accounting: The roleof accounting information for strategy in competitive markets. Accounting,Organizations and Society 15 (1/2), 27�46.

Bryce, D. J. and M. Useem (1998). The impact of corporate outsourcing on com-pany value. European Management Journal 16 (6), 635�643.

Buckley, P. J. and D. R. Lessard (2005, November). Regaining the edge for inter-national business research. Journal of International Business Studies 36 (6),595�599.

Burns, J. and G. Baldvinsdottir (2005, December). An institutional perspective ofaccountants�new roles �the interplay of contradictions and praxis�. EuropeanAccounting Review 14 (4), 725�757.

CIMA (2009). Improving decision making in organisations: The opportunity toreinvent �nance business partners. Chartered Institute of Management Ac-countants.

Coase, R. H. (1937, November). The nature of the �rm. Economica 4 (16), 386�405.

Conner, K. R. (1991). A historical comparison of resource-based theory and �veschools of thought within industrial organization economics: Do we have anew theory of the �rm? Journal of Management 17 (1), 121�154.

David, M. (2009, July). The correspondence theory of truth. Stanford Encyclo-pedia of Philosophy. http://plato.stanford.edu/entries/truth-correspondence.

Davidow, W. H. and M. S. Malone (1992). The Virtual Corporation �Structuringand Revitalizing the Corporation for the 21st Century. New York: HarperBusi-ness.

Descartes, R. (1628). Descartes: Selected philosophical writings, Chapter I: Rulesfor the Direction of our Native Intelligence. Cambridge: Cambridge UniversityPress, 1988.

Dibbern, J., T. Goles, R. Hirschheim, and B. Jayatilaka (2004). Information sys-tems outsourcing: A survey and analysis of the literature. Database for Ad-vances in Information Systems 35 (4), 6�102.

58

Dierickx, I. and K. Cool (1989, December). Asset stock accumulation and sustain-ability of competitive advantage. Management Science 35 (12), 1504�1515.

Downing, L. (2004, September). George Berkeley. Stanford Encyclopedia of Phi-losophy. http://plato.stanford.edu/entries/berkeley.

Eisenhardt, K. M. and M. J. Zbaracki (1992). Strategic decision making. Strate-gic Management Journal 13 (Special Issue: Fundamental Themes in StrategyProcess Research), 17�37.

Ellram, L. M. (1993, October). Total cost of ownership elements and implementa-tion. International Journal of Purchasing and Materials Management 29 (4),3�11.

Ellram, L. M. (1995). Total cost of ownership: an analysis approach for pur-chasing. International Journal of Physical Distribution & Logistics Manage-ment 25 (8), 4�23.

Ellram, L. M. and A. B. Maltz (1995). The use of total cost of ownership con-cepts to model the outsourcing decision. The International Journal of LogisticsManagement 6 (2), 55�66.

Ellram, L. M. and S. P. Siferd (1998). Total cost of ownership: A key conceptin strategic cost management decisions. Journal of Business Logistics 19 (1),55�84.

Galbraith, J. K. (1967). New Industrial State. New Jersey: Princeton UniversityPress, 2007.

Gettier, E. L. (1963). Is justi�ed true belief knowledge? Analysis 23, 121�123.

Goldman, S., R. Nagel, and K. Preiss (1995). Agile Competitors and Virtual Or-ganizations: Strategies for Enriching the Customer. New York: Van NostrandReinhold.

Gordon, M. and G. Shillinglaw (1964). Accounting: A Management Approach.Irwin.

Gottfredson, M., R. Puryear, and S. Phillips (2005, February). Strategic sourcing:From periphery to the core. Harvard Business Review 83 (2), 132�139.

Grant, R. M. (2005). Contemporary Strategy Analysis (5th ed.). Blackwell Pub-lishing Ltd.

Gross, H. (1966, October). Make or buy decisions in growing �rms.The AccountingReview 41 (4), 745�753.

Harriman, N. F. (1982). Principles of Scienti�c Purchasing (1st ed.). New York:McGraw-Hill.

Higgins, C. C. (1955, March - April). Make-or-buy re-examined. Harvard BusinessReview 33 (2), 109�119.

Hill, C. W. L. and G. R. Jones (2001). Strategic Management: An IntegratedApproach (5th ed.). New York: Houghton Mi­ in Company.

Hjørland, B. (2005). Empiricism, rationalism and positivism in library and infor-mation science. Journal of Documentation 61 (1), 130�155.

59

Hopper, T. (1980). Role con�icts of management accountants and their positionwithin organisation structures. Accounting, Organizations and Society 5 (4),401�411.

Horngren, C. T. (1965). Accounting for Management Control. New Jersey: Pear-son Prentice Hall.

Horngren, C. T., G. L. Sundem, and W. O. Stratton (2005). Introduction toManagement Accounting (13th ed.). New Jersey: Pearson Prentice Hall.

Hätonën, J. and T. Eriksson (2009). 30+ years of research and practice of out-sourcing �exploring the past and anticipating the future. Journal of Interna-tional Management 15 (2), 142�155.

Huntington, S. P. (1998, January). The Clash of Civilizations and the Remakingof World Order (First ed.). New York: Simon & Schuster Inc.

Jackson, D. W. and L. L. Ostrom (1980, Winter). Life cycle costing in industrialpurchasing. Journal of Purchasing and Materials Management 16, 8�12.

Jazayeri, M. and T. Hopper (1999). Management accounting within world classmanufacturing: a case study. Management Accounting Research 10 (3), 263�301.

Jonassen, D. H. (1991). Objectivism versus constructivism: Do we need anew philosophical paradigm? Educational Technology Research and Devel-opment 39 (3), 5�14.

Kaplan, R. S. (1983). Measuring manufacturing performance: A new challengefor managerial accounting research. The Accounting Review 58 (4), 686�706.

Kaplan, R. S. (1984a). The evolution of management accounting. The AccountingReview 59 (3), 390�419.

Kaplan, R. S. (1984b). Yesterday�s accounting undermines production. HarvardBusiness Review 62 (4), 95�102.

Kaplan, R. S. (1985). The Uneasy Alliance: Managing the Productivity-TechnologyDilemma, Chapter Accounting lag: the obsolescence of cost accounting sys-tems, pp. 195�226. Boston, MA: Harvard Business School Press.

Kaplan, R. S. and R. Cooper (1998). Cost & E¤ect: Using Integrated Cost Sys-tems to Drive Pro�tability and Performance. Boston: Harvard Business SchoolPress.

Kirkham, R. L. (1998, October). Truth, coherence theory of. Routledge Encyclo-pedia of Philosophy. http://www.rep.routledge.com/article/N063SECT1.

Koppelmann, U. (1996). Outsourcing. Stuttgart: Schä¤er-Poeschel Verlag.

Lacity, M. C. and R. Hirschheim (1993). The information systems outsourcingbandwagon. Sloan Management Review 35 (1), 73�87.

Lang�eld-Smith, K. (2008). Strategic management accounting: how far have wecome in 25 years? Accounting, Auditing & Accountability Journal 21 (2), 204�228.

60

Laughlin, R. (1995). Empirical research in accounting: alternative approachesand a case for "middle-range" thinking. Accounting, Auditing & AccountabilityJournal 8 (1), 63�87.

Lawton, T. C. and K. P. Michaels (2001). Advancing to the virtual value chain:Learning from the dell model. Irish Journal of Management 22 (1), 91�112.

Linder, J., M. Cole, and A. Jacobson (2002). Business transformation throughoutsourcing. Strategy and Leadership 30 (4), 23�28.

Locke, J. (1690). An Essay Concerning Human Understanding. Penguin BooksLtd, 1997.

Lonsdale, C. and A. Cox (2000). The historical development of outsourcing: thelatest fad? Industrial Management & Data Systems 100 (9), 444�450.

Markie, P. (2008, August). Rationalism vs. empiricism. Stanford Encyclopedia ofPhilosophy. http://plato.stanford.edu/entries/rationalism-empiricism.

Mathews, R. (1962). Accounting for Economists. Cheshire.

Mazzawi, E. (2002). Transformational outsourcing. Business Strategy Re-view 13 (3), 39�43.

McIvor, R. (2005). The Outsourcing Process: Strategies for Evaluation and Man-agement. Cambridge: Cambridge University Press.

McRae, T. W. (1970, April). Opportunity and incremental cost: An attempt tode�ne in systems terms. The Accounting Review 45 (2), 315�321.

Miles, R. E. and C. C. Snow (1986). Organizations: New concepts for new forms.California Management Review 28 (3), 62�74.

Mintzberg, H. (1978, May). Patterns in strategy formation. Management Sci-ence 24 (9), 934�948.

Mintzberg, H. and J. A. Waters (1985, July - September). Of strategies, deliberateand emergent. Strategic Management Journal 6 (3), 257�272.

Mol, M. J. (2007, September). Outsourcing: Design, Process and Performance(1st ed.). Cambridge: Cambridge University Press.

Morgan, J. (1999). Purchasing at 100: Where it�s been, where it�s headed. Pur-chasing 127 (8), 72�94.

Morgan, R. E. (2003). Outsourcing: Toward the �shamrock organization�. Journalof General Management 29 (2), 35�52.

Mouritsen, J. (1998). Driving growth: Economic value added versus intellectualcapital. Management Accounting Research 9 (4), 461�482.

Mouritsen, J., H. T. Larsen, and P. N. D. Bukh (2001). Intellectual capital and the�capable �rm�: narrating, visualising and numbering for managing knowledge.Accounting, Organizations and Society 26 (7-8), 735�762.

Nørreklit, H., L. Nørreklit, and F. Mitchell (2010). Towards a paradigmatic foun-dation of accounting practice. Accounting, Auditing & Accountability Journal .Forthcoming.

61

Nørreklit, L. (2009). Videnskabsteori �en indføring. Research note.

Nørreklit, L., H. Nørreklit, and P. Israelsen (2006). The validity of managementcontrol topoi: Towards constructivist pragmatism. Management AccountingResearch 17 (1), 42�71.

Nutt, P. C. (1976, April). Models for decision making in organizations and somecontextual variables which stipulate optimal use. Academy of ManagementReview 1 (2), 84�98.

Oxford, R. L. (1997). Constructivism: Shape-shifting, substance, and teachereducation applications. Peabody Journal of Education 72 (1), 35�66.

Payne, B. (1957, March - April). Steps in long-range planning. Harvard BusinessReview 35 (2), 95�106.

Penrose, E. (1959). The Theory of the Growth of the Firm. Oxford: Oxford Uni-versity Press.

Peteraf, M. A. (1993, March). The cornerstones of competitive advantage: Aresource-based view. Strategic Management Journal, 14 (3), 179�191.

Peters, T. J. and R. H. Waterman (1982). In Search of Excellence: Lessons fromAmerica�s Best-Run Companies. New York: Harper & Row.

Piore, M. J. and C. F. Sabel (1986, October). The Second Industrial Divide:Possibilities For Prosperity. Basic Books.

Porter, M. E. (1980). Competitive Strategy: Techniques for Analyzing Industriesand Competitors. New York: Free Press.

Porter, M. E. (1985). Competitive Advantage: Creating and Sustaining SuperiorPerformance. New York: Free Press.

Porter, M. E. (1991). Towards a dynamic theory of strategy. Strategic ManagementJournal 12, 95�117.

Prahalad, C. K. and G. Hamel (1990, May - June). The core competence of thecorporation. Harvard Business Review 68 (3), 79�92.

Pritchard, D. (2007, August). The value of knowledge. Stanford Encyclopedia ofPhilosophy. http://plato.stanford.edu/entries/knowledge-value.

Quinn, J. B. and F. G. Hilmer (1994). Strategic outsourcing. Sloan ManagementReview 35 (4), 43�56.

Rambjerg, B. and K. Gjesdal (2005, November). Hermeneutics. Stanford Ency-clopedia of Philosophy. http://plato.stanford.edu/entries/hermeneutics/.

Rumelt, R. P. (1974). Strategy, Structure, and Economic Performance. Boston,MA: Harvard Business School Press.

Rumelt, R. P. (1982). Diversi�cation strategy and pro�tability. Strategic Manage-ment Journal 3 (4), 359�369.

Russell, B. (1912). The problems of philosophy. Oxford: Oxford University Press,1971.

62

Ryan, B., R. W. Scapens, and M. Theobold (2002). Research Method and Method-ology in Finance and Accounting. Thomson Learning.

Scapens, R. W., M. Ezzamel, J. Burns, and G. Baldvinsdottir (2003). The FutureDirection of UK Management Accounting Practice (1st ed.). CIMAPublishing.

Schwitzgebel, E. (2006, August). Belief. Stanford Encyclopedia of Philosophy.http://plato.stanford.edu/entries/belief.

Seal, W. (2010). Managerial discourse and the link between theory and prac-tice: From ROI to value-based management. Management Accounting Re-search 21 (2), 95�109.

Shank, J. K. (2007). Contemporary Issues in Management Accounting, Chapter16: Strategic cost management: upsizing, downsizing, and right(?) sizing, pp.355�379. Oxford: Oxford University Press.

Shank, J. K. and V. Govindarajan (1989). Strategic Cost Analysis �The Evolutionfrom Managerial to Strategic Accounting, Chapter 3: Concepts in Value ChainAnalysis: The "Famous" Crown Cork and Seal Company Case, pp. 39�59.Irwin.

Shank, J. K. and V. Govindarajan (1992). Strategic cost management: The valuechain perspective. Journal of Management Accounting Research 4, 179�197.

Shank, J. K. and V. Govindarajan (1994). Strategic Cost Management �The NewTool for Competitive Advantage. New York: Free Press.

Siegel, G. H. and J. E. Sorensen (1999, August). Counting more, counting less:Transformation in the management accounting profession. Technical report,Institute of Management Accountants.

Simmonds, K. (1981). Strategic management accounting. Management Account-ing 59 (4), 26�30.

Simon, H. (1961). Administrative Behavior (2nd ed.). New York: Macmillan.

Sprigge, T. L. S. (1998, October). Idealism. Routledge Encyclopedia of Philoso-phy. http://www.rep.routledge.com/article/N027.

Stigler, G. J. (1951, June). The division of labor is limited by the extent of themarket. The Journal of Political Economy 59 (3), 185�193.

Teece, D. J. (1976). The multinational corporation and the resource cost of inter-national technology transfer. Cambridge, MA: Ballinger Pub. Co.

Teece, D. J. (1980). Economies of scope and the scope of the enterprise. Journalof Economic Behavior and Organization 1 (3), 223�247.

Teece, D. J., G. Pisano, and A. Shuen (1997, August). Dynamic capabilities andstrategic management. Strategic Management Journal 18 (7), 509�533.

Thrift, N. (2005). Knowing Capitalism. London: Sage Publications.

Wernerfelt, B. (1984, April - June). A resource-based view of the �rm. StrategicManagement Journal 5 (2), 171�180.

Williamson, O. E. (1975). Markets and Hierarchies: Analysis and Antitrust Im-plications. New York: Free Press.

63

Williamson, O. E. (1979, October). Transaction-cost economics: The governanceof contractual relations. Journal of Law and Economics 22 (2), 233�261.

Williamson, O. E. (1981, November). The economics of organization: The trans-action cost approach. The American Journal of Sociology 87 (3), 548�577.

Williamson, O. E. (1985). The Economic Institutions of Capitalism. New York:Free Press.

Winkler, K. P. (1989). Berkeley: An Interpretation. Oxford: Clarendon Press.

Wrapp, H. E. (1957, January - February). Organization for long-range planning.Harvard Business Review 35 (1), 37�47.

Young, J. O. (2008, September). The coherence theory of truth. Stanford Ency-clopedia of Philosophy. http://plato.stanford.edu/entries/truth-coherence.

64

Inpu

t In

boun

d lo

gist

ics

Ope

ratio

ns

O

utbo

und

logi

stic

s M

arke

ting

& sa

les

Serv

ice

O

utpu

t

Fina

nce

and

plan

ning

H

uman

reso

urce

man

agem

ent

Proc

ess d

evel

opm

ent a

nd d

esig

n Su

pply

cha

in m

anag

emen

t

Prim

ary

supp

ly c

hain

Seco

ndar

y su

pply

cha

in

Supp

ort

activ

ities

Figure1:Supplychainwithprimaryandsupportactivities,respectively.

65

Strategy

Controlling

Accounting

1. Determination of decision alternatives

2. Evaluation of decision alternatives

4. Feedback

3. Selection of decision alternative

Figure 2: The interdependency between strategy and accounting in the phases ofoutsourcing decision-making and implementation, respectively.

66

Subject Decision-maker

Knowledge

Epistemology Source(s) of knowledge

Object (outsourcing situation) World: “true” state of affairs for

company and industry

Appearance of particulars Some state of affairs

Perceptions Uncovering process based on

observation/experience

Reason Uncovering process based

on intuition

Informed decision-making requires knowledge

Knowledge consists of true justified beliefs

Empiricism Rationalism

Realism Object is independent of subject’s

perception

Idealism Object depends on subject’s

perception

Beliefs/Ideas

Beliefs are justified through a conviction of how to uncover the object

Ontology

Figure 3: Framework for analysis of the accounting and strategy literature onoutsourcing.

67

Subj

ect:

belie

f Ex

perie

nce

/ Int

uitio

n O

bjec

t

Pote

ntia

l err

or d

ue to

pe

rcep

tion

or c

ompl

exity

If th

e be

lief c

orre

spon

ds w

ith

expe

rienc

e or

intu

ition

it is

just

ified

an

d tu

rned

into

kno

wle

dge

(trut

h)

Rea

lism

allo

ws t

he in

fere

nce

that

exp

erie

nce

or in

tuiti

on

corr

espo

nds w

ith th

e ob

ject

Figure4:Therelationshipbetweentheobjectandsubjectinthecorrespondencetheoryoftruth.

68

Exte

rnal

obj

ect (

true

stat

e of

aff

airs

): Im

med

iate

obj

ect (

som

e st

ate

of a

ffai

rs):

Ded

uctio

n D

educ

tion

Ded

uctio

n Intu

ition

In

tuiti

on

Indu

ctio

n In

duct

ion

Indu

ctio

n Expe

rienc

e Ex

perie

nce

Rat

iona

lism

/ re

alis

m

Em

piri

cism

e / r

ealis

m

Figure5:Decompositionalanalysiswithinrationalismandempiricism.

69

Proj

ectio

n of

gen

eral

mar

ket g

row

th

Sale

s for

ecas

t Pr

ojec

ted

com

pany

shar

e of

mar

ket g

row

th (v

olum

e)

Cos

t com

pari

son

• Pr

oduc

tion

cost

at

volu

me

(mak

e)

• Pr

ice

at v

olum

e (b

uy)

Mak

e-or

-buy

dec

isio

n

Wha

t inv

estm

ents

are

nee

ded

to

sust

ain

volu

me?

•

Hum

an c

apita

l: tra

inin

g,

relia

bilit

y, p

rofic

ienc

y, a

nd e

xten

t •

Phys

ical

cap

ital:

prod

uctio

n ca

paci

ty

(Som

e) q

ualit

ativ

e fa

ctor

s •

Prod

uct q

ualit

y ca

n be

co

ntro

lled

in-h

ouse

• Su

pplie

r acc

ess a

nd

relia

bilit

y

• Po

ssib

le re

stric

tions

on

in-h

ouse

indu

stry

kn

owle

dge

may

favo

r bu

ying

Bud

get f

or m

ake-

or-b

uy a

ltern

ativ

e on

the

basi

s of v

olum

e

Figure6:Theaspectsofmake-or-buydecision-makingduringtheeraofcorporateplanning.

70

Act

iviti

es fo

r the

mak

e-or

-buy

alte

rnat

ive

are

iden

tifie

d in

a fl

ow c

hart

and

TC

O a

re

allo

cate

d ac

cord

ingl

y us

ing

AB

C.

Flow

cha

rts a

re c

ompa

red

and

key

activ

ities

diff

eren

tiatin

g th

e m

ake-

or-b

uy a

ltern

ativ

e ar

e id

entif

ied.

TCO

incr

emen

t is c

alcu

late

d fo

r eac

h ke

y ac

tivity

rela

ted

to th

e m

ake-

or-b

uy a

ltern

ativ

e.

Mak

e-or

-buy

dec

isio

n ru

le: t

he a

ltern

ativ

e w

ith th

e lo

wes

t inc

rem

ent i

n TC

O is

cho

sen.

Figure7:StepsindeterminingTCOforoutsourcingdecision-making.

71

Acc

ount

ing

Stra

tegy

• Em

piric

al d

ata

colle

ctio

n fo

r eac

h of

the

five

forc

es

• D

isag

greg

ate

activ

ities

in

the

com

pany

’s v

alue

ch

ain

•

Allo

cate

cos

ts to

the

valu

e ch

ain

activ

ities

• Id

entif

y st

ruct

ural

fact

ors

affe

ctin

g th

e ac

tiviti

es

and

the

inte

rdep

ende

ncy

betw

een

them

Indu

stry

ana

lysi

s D

ecid

e on

the

stre

ngth

of e

ach

of

the

five

forc

es a

nd

Val

ue c

hain

ana

lysi

s A

naly

ze c

osts

, co

nfig

urat

ion

of th

e va

lue

chai

n an

d w

hat

activ

ities

to fo

cus o

n

Gen

eric

stra

tegy

D

ecid

e on

low

cos

t or

diff

eren

tiatio

n

Rel

ativ

e po

sitio

n in

th

e in

dust

ry

Figure8:Stepsinstrategiccostmanagementastheymighttakeplaceinrelationtooutsourcingdecision-making.

72

Cor

p. P

lann

ing

• A

ccou

ntin

g

Ont

.: R

ealis

m

Epi.:

Em

piric

ism

•

Stra

tegy

O

nt.:

Rea

lism

Ep

i.: E

mpi

ricis

m

Big

Ban

g •

Acc

ount

ing

O

nt.:

Rea

lism

Ep

i.: E

mpi

ricis

m

• St

rate

gy

Ont

.: R

ealis

m

Epi.:

Rat

iona

lism

Ban

dwag

on 1

•

Acc

ount

ing

O

nt.:

Rea

lism

Ep

i.: E

mpi

ricis

m

R

atio

nalis

m

• St

rate

gy

Ont

.: R

ealis

m (+

)

Idea

lism

(-)

Epi.:

Em

piric

ism

Rat

iona

lism

Ban

dwag

on 2

•

Acc

ount

ing

O

nt.:

???

Ep

i.: ?

??

• St

rate

gy

Ont

.: R

ealis

m (-

)

Idea

lism

(+)

Epi.:

Rat

iona

lism

?

GA

P: T

he e

xten

t to

whi

ch th

e ac

coun

ting

mod

el c

an su

ppor

t dec

isio

n-m

akin

g w

ithin

the

stra

tegi

c fr

amew

ork TC

E

TCO

IV

SCM

(T)

SCM

(A)

Mis

sing

acc

. tre

atm

ent

RB

V

Tim

e

Tangible Intangible

Stra

tegy

:

Acc

ount

ing:

an

d

Aspects important to decision-making

Figure9:Thehistoricaldevelopmentsinmodelcomplexityforoutsourcingdecisionframeworksinthestrategyandaccounting

literature,respectively.

73

Chapter 2

The outsourcing decision-making process: Two

extensions to the management accounting and

strategy toolkit

The outsourcing decision-making process: Two

extensions to the management accounting and

strategy toolkit

Lars Bråd Nielsen� Hanne Nørreklity

July 27th, 2010

Abstract

The integration of management accounting and strategy in relation to complexdecision-making has long been a severely underdeveloped area in the practice-oriented part of the management accounting literature. Speci�cally, even fora topic as important as outsourcing, the outsourcing decision-making processis still often portrayed as a simpli�ed comparison between internal costs andquotation prices. Puzzled by this lack of research, the current article sets outto study how accounting information interacts with strategy in practice whendeciding on outsourcing. To make this possible, a conceptual framework isinitially developed to distinguish between the areas of strategy and manage-ment accounting, and to investigate the interaction between the two areas inoutsourcing decision-making. The purpose is to pave the way for a quali�eddiscussion of how strategy and management accounting in combination canhelp ensure that the decisions made are both feasible and pro�table. Sub-sequently, the outsourcing decision-making process is analyzed in two casecompanies operating in manufacturing. The cases reveal how strategy framesthe outsourcing decision-making process in accordance with the company�scontextual circumstances and how this framing shapes the accounting infor-mation produced and used. Consequently, the cases show that strategy as wellas management accounting plays a crucial role in outsourcing decision-makingand that they are tightly integrated. Separating the study of accounting infor-mation production and use from the strategic context in relation to (complex)outsourcing decision-making is thus at best simplistic and at worst deceptive.For practice, the results are extensions to the management accounting andstrategy toolkit since the cases can serve as inspiration for how to structurethe outsourcing decision-making process. Also, the article adds to the partof the academic literature on decision theory debating the appropriateness ofstudying the decision-making process as a mechanical system instead of anorganic process based on human intuition. Speci�cally, the results indicatethat both standpoints are legitimate and can be found in practice.

�Aarhus School of Business, Aarhus University, Fuglesangs Allé 4, DK-8210 Aarhus V., Den-mark. E-mail: labn@asb.dk.

yProfessor of Management Accounting, Aarhus School of Business, Aarhus University, Fugle-sangs Allé 4, DK-8210 Aarhus V., Denmark. E-mail: hann@asb.dk.

1 Introduction

There is no question that outsourcing is one of the most sustained trends of cur-

rent business (Fill and Visser 2000). Over time outsourcing has changed the way

in which �rms compete in so diverse industries as automobiles, aerospace, telecom-

munications, computers, pharmaceuticals, chemicals, health care, �nancial services,

energy systems, and software (Carson 2007; Dahan and Hauser 2002; Quinn 2000).

Whereas markets used to be distinct entities, protected from each other by trade

and time barriers, this is no longer the case (Hill and Jones 2001). Instead, a highly

competitive environment �sometimes referred to as an outsourcing economy �has

emerged (Hätonën and Eriksson 2009), in which national markets have merged and

where consumers have become increasingly knowledgeable and demanding about is-

sues such as price, reliability, and availability (Hill and Jones 2001; McIvor 2005).

These changes represent tremendous challenges to companies; yet they also provide

opportunities on a much larger scale than ever seen before.

As the global resource pool is made available, outsourcing holds the alluring

potential of restructuring the organization to reap greater economies of scale, share

investments in R&D and marketing across various markets, or take advantage of

low-cost labor resources for both the manufacturing of products and the delivery of

services (McIvor 2005; Lewin and Peeters 2006). Thus, to meet today�s challenges

and harvest the promised potential, the recent interest in outsourcing suggests fo-

cusing the organization on core activities while simultaneously leveraging external

resources and capabilities (e.g. Peters and Waterman 1982; Prahalad and Hamel

1990). However, while outsourcing may be pro�table for some companies to under-

take, history has shown that it is certainly not the right choice for every company

(Bettis et al. 1992; Bryce and Useem 1998; Barthélemy 2001; Gove 2009). Indeed,

the high failure rate of outsourcing relationships indicates that many companies

base the decision to outsource on insu¢ cient information and a general lack of un-

derstanding.

In view of these shortcomings, we should expect to �nd extensive academic research

on the outsourcing decision-making process within the practice-oriented part of the

management accounting literature. However, so far the literature has mainly been

concerned with studying outsourcing relations while only minor attention has been

given to the actual decision-making process (Widener and Selto 1999; Seal et al.

1999; van der Meer-Kooistra and Vosselman 2000; Tomkins 2001; Lang�eld-Smith

and Smith 2003; Nicholson et al. 2006; Thrane and Hald 2006). Furthermore, even

within the sporadic empirical work (see e.g. Sartorius and Kirsten 2005; Lammin-

76

maki 2008) that currently exists on outsourcing decision-making, the role of strategy

in structuring the decision-making process as well as its e¤ect on the production and

use of accounting information has been neglected.1

In fact, a historical comparison of the outsourcing literature on strategy and

management accounting, made by Nielsen (2010), reveals that a gap has come

to exist between the aspects considered relevant for decision-making within each

area. Whereas over time the fundamental frameworks of strategy concerned with

the boundaries of the �rm (theories of the �rm) have come to emphasize the in-

tangible and subtle aspects of the outsourcing decision, management accounting

research has in general remained focused on the quanti�able aspects, thus failing to

adequately address and debate how, if at all, qualitative strategic factors are to be

incorporated into the accounting analysis. Consequently, the management account-

ing models in the academic literature have lost signi�cant relevance for outsourcing

decision-making, as they can only supply information to the overall strategic frame-

works which are simpli�ed and do not incorporate the full complexity of the decision.

In the light of this, the purpose of the present article is to study the outsourcing

decision-making process in practice. Particularly, the article examines how compa-

nies use outsourcing, how they structure the analysis of the problem, and what types

of information are perceived as relevant for decision-making. The interest is thus

not with the success of the actual outsourcing decision, nor is it with developing

suggestions as to how the concrete analysis could have been improved. Instead, the

objective is to re�ect upon how companies combine soft strategic considerations with

hard-headed accounting analyses when arriving at a decision. Therefore, the article

is best seen as an initial attempt to kick-start management accounting research into

outsourcing decision-making and reduce the prevailing gap between the literature

on management accounting and strategy.

The article draws on interview material from two case companies in manufac-

turing, DEN and EUR, which in very di¤erent ways make use of outsourcing due to

their current situation. Also, to deepen the preliminary understanding of outsourc-

ing decision-making in practice, additional interviews have been conducted with key

employees in three consulting �rms, all of whom hold extensive experience in out-

sourcing.

1Sartorius and Kirsten (2005) is one of the few studies of the outsourcing decision-makingprocess where attention is given to the combined role of management accounting and strategy.Speci�cally, the study demonstrates how transaction cost analysis can assist the company in struc-turing the outsourcing decision. Lamminmaki (2008) also studies the outsourcing decision-makingprocess. However, here management accounting is seen in isolation from strategy.

77

In the next two sections, we will discuss the theoretical framework applied in the

analysis of outsourcing decision-making and explain the research methodology used

for data collection. Following this, the two company cases on outsourcing decisions

are presented and analyzed. On the basis of the analysis, the �nal section concludes

that management accounting and strategy are deeply integrated in the outsourcing

decision-making process. Accordingly, the implications of the results for decision

theory at a more general level are discussed.

2 Framework for analysis of outsourcing decision-

making

This section develops the tools and terminology used for analyzing outsourcing

decision-making in the case companies. Initially, we conceptualize the theoreti-

cal interdependency between the tasks in the decision-making process that reside

within strategy and management accounting, respectively. The purpose is twofold.

First, no clear-cut de�nitions of strategy and management accounting exist, for

which reason it is important to specify how they are to be understood in relation to

outsourcing decision-making in this paper. Second, this makes it possible to clas-

sify the strategic tasks and the management accounting tasks when studying the

case companies. Using this classi�cation, we subsequently draw on the theoretical

aspects of methodology to outline a framework applicable for analyzing how infor-

mation is produced to support outsourcing decision-making in the case companies.

The section closes with some re�ections on business reality. Speci�cally, a frame-

work is constructed to help understand the joint task of strategy and management

accounting of pinpointing decision alternatives that are feasible as well as pro�table

to pursue.

2.1 Conceptualizing strategy and management accounting

in outsourcing decision-making

On strategy in general, Grant (2005, p. 21) notes that several de�nitions exist, yet

all of them revolve around the aspects of where to compete and how to compete.

Mintzberg, in particular, distinguishes between intended strategies (the company

leadership plans and intentions) and realized strategies (the company actions actually

taken) (Mintzberg 1978; Mintzberg and Waters 1985). As such, all strategies can

be compared on the basis of the �t between intentions and realization. When the

78

intended strategy turns out to be the strategy actually realized, it is referred to as

a deliberate strategy, although the match between intention and realization is not

always perfect. Mintzberg thus introduces the concept of an emergent strategy to

describe the situation when the actual strategic actions deviate from the original

intentions and plans or when neither initial intentions nor plans exist at all. Since

the interest of this article is with planned decision-making, we will not focus on

this latter concept but instead concentrate on the notion of a deliberate strategy

where in the initial phase the company plans the strategic actions to be carried out

subsequently. In particular, we thus adopt the de�nition of a strategic plan found in

Horngren et al. (2005), as this fully captures Mintzberg�s conception2 and the two

fundamental aspects pointed out in Grant (2005):

A strategic plan refers to setting the overall objectives for an organization

and outlining how they will be attained (Horngren et al. 2005, p. 299).

Keeping this de�nition in mind, Horngren et al. (2005, p. 11) point out that

decisions in general can be divided into two groups concerned with either (strategic)

planning or control, where the latter refers to the implementation of the (strategic)

plan and the use of feedback to attain objectives. Therefore, albeit both types

of decisions are potentially crucial, strategic decisions tend to involve issues more

important and fundamental to a company than is the case for decisions related to

control.

In continuation of this, Horngren et al. (2005, p. 11) de�ne decision-making as

"the purposeful choice from among a set of alternative courses of action designed to

achieve some objective." Anderson (1983) develops this even further by decompos-

ing the task of making a decision into �ve interdependent subtasks: (1) identifying

the relevant goals; (2) searching for alternative courses of action; (3) predicting the

consequences of following each alternative; (4) evaluating each alternative in terms

of its consequences for goal achievement; and (5) selecting the best alternative for

achieving the goal.3

Hence, focusing on strategic decisions, the use of Anderson�s (1983) decomposi-

tion of subtasks for decision-making shows that the de�nition of strategy essentially

covers subtasks 1, 2, and 5: identi�cation of goals, determination of decision alterna-

tives, and selection of the best alternative. However, to facilitate this �nal selection,

2Notice that the most extreme version of a deliberate strategy is referred to as a planned strategy(Mintzberg and Waters 1985).

3Notice that we use the model of rational action to de�ne decision-making (Eisenhardt andZbaracki 1992). However, other frameworks of decision-making exist; see for instance Nutt (1976).

79

an evaluation of the alternatives and an assessment of the consequences are needed

�and this is where accounting comes into play. As such, Horngren et al. (2005)

de�ne management accounting as:

The branch of accounting that produces information for managers within

an organization. It is the process of identifying, measuring, accumu-

lating, analyzing, preparing, interpreting, and communicating informa-

tion that helps managers ful�ll organizational objectives (Horngren et al.

2005, p. 5).

Consequently, in the context of Anderson (1983), management accounting covers

subtasks 3 and 4: information provision and evaluation of alternatives. Building on

this, management accounting is in this paper interpreted in the sense of strategic

management accounting, as it is outlined in Roslender and Hart (2003), integrating

the traditional internal (cost) focus of the company with external demand factors �

the marketing management aspect. It is thus to be understood as an interdiscipli-

nary approach to decision-making that cuts across traditional functional boundaries

in the company and encompasses techniques and knowledge from both accounting,

marketing, R&D, and supply chain management to provide su¢ ciently organic in-

formation for the evaluation of each alternative.

With these classi�cations in mind, outsourcing decision-making is in this article

conceptualized as an interdependent process of steps that shifts back and forth be-

tween the domains of strategy and management accounting. Speci�cally, it is to be

understood as the encompassing decision process that shapes the company�s entire

supply chain con�guration for a given product and determines whether the individ-

ual parts are to be kept in-house or contracted out. Essentially, the outsourcing

decision is thus a set of interrelated make-buy decisions that need to be processed

in combination. This is illustrated in �gure 1 below.

<Insert �gure 1 here>

To understand and analyze how information is produced and used within the out-

sourcing decision-making process in the case companies, an appropriate conceptual

framework � or methodology � is needed. To this end, we take a paradigmatic

approach, since every outsourcing decision draws on methods for creating business

knowledge, for which reason it is underpinned by two elemental aspects: ontology

and epistemology (Arbnor and Bjerke 1997).

80

The decision-maker�s ontology characterizes the relationship between himself and

the world from which his knowledge stems (Ryan et al. 2002). At it simplest,

it involves whether the decision-maker believes that reality exists objectively and

independently of his mind (realism) or, at least partially, is shaped/constructed

by his own perception (idealism/constructivism) (Nielsen 2010).4 Following this,

the decision-maker�s epistemology refers to the nature, sources, and limits of his

knowledge creation and e¤ectively revolves around how knowledge is created from

sense experience (empiricism) and intuition/deduction (rationalism) (Ryan et al.

2002; Markie 2008; Nielsen 2010).

Together, the ontology and epistemology form a paradigm5 that fundamentally

a¤ects the decision-maker�s understanding of the outsourcing problem and how in-

formation can be produced to rule in favor of one supply chain con�guration over

another. Essentially, the paradigm thus determines the choice and design of the

strategic framework, outlining the central aspects of the decision to be considered.

Accordingly, the strategic framework comes to represent the overall study plan �or

methodic �for analyzing the outsourcing problem and reaching a decision on the

preferred supply chain con�guration.6

Conditioned on the strategic framing of the problem, and in agreement with the

classi�cations established above, the role of management accounting then becomes

to specify and evaluate the consequences of each decision alternative relying on

proper data collection. For this purpose, various accounting techniques are readily

available that, through a methodical procedure7, can be transformed into methods �

or context speci�c accounting decision models �for knowledge creation in relation

to the particular outsourcing problem (Arbnor and Bjerke 1997, p. 17, p. 218).

The information produced and related advice are �nally reported back to the

domain of strategy which uses the information to reach a decision on the preferred

supply chain con�guration. As such, the strategic framework determines whether

the con�guration of the supply chain will be decided as a sequence of smaller steps,

4The literature on philosophy often distinguishes between idealism and constructivism. Still,they share the same ontological stance (Oxford 1997). Thus, for the purpose of this article, we willrefrain from using this distinction and instead combine the ideas from constructivism and idealismto stress the notion that reality per se is not given but can in fact be shaped either mentally orsocially.

5The political aspect also forms a part of a paradigm (Arbnor and Bjerke 1997). However, thisis not emphasized in this article.

6In theory, methodics can exist at several levels, since lower-level methodics can be outlined forthe various parts of a higher-level methodic. However, in this paper, a methodic will only refer tothe overall study plan laid out by the strategic framework.

7A methodical procedure can be de�ned as the way in which the creator of knowledge incor-porates, develops, and/or modi�es some previously given technique in a methodological approach(Arbnor and Bjerke 1997, p. 16).

81

where accounting information is repeatedly produced and brought to the table, or

in one single step based on the information available.

Figure 2 below illustrates this conception of outsourcing decision-making and the

interdependent process between strategy and management accounting. The frame-

work will be applied in the empirical section of this paper to understand and analyze

the outsourcing decision-making process in the case companies.

<Insert �gure 2 here>

2.2 The ontological conception of business reality: Chal-

lenging the epistemology of strategy and management

accounting in outsourcing decision-making

Given the framework in �gure 2, it is appropriate to dwell momentarily on ontology

and re�ect upon business reality in relation to outsourcing decision-making. Ceteris

paribus, companies will only consider outsourcing parts of the supply chain if they

expect to generate pro�t from doing so. Therefore, while companies in principle are

free to choose whatever ontology they want to underpin their approach to supply

chain con�guration, the majority of pro�t-seeking companies will most probably

agree that a business reality exists which �at least partially �needs to be understood

and accounted for in order to provide some assurance for successful decision-making.

In this manner, we presume that a company must take the paradigmatic stance

of pragmatic constructivism, which involves a moderate approach to ontology in-

tegrating dimensions of both realism and idealism/constructivism (Nørreklit et al.

2010). Speci�cally, we buy into the notion of reality since we believe that there are

limits to how much a company can construct itself. Essentially, this means that the

feasibility of the supply chain con�gurations constructed and considered by the com-

pany partially depends on the business reality and thus on the company�s historical

and current resource endowment as well as the complexity of the product intended

for manufacturing (path dependency). Still, within these limits, there is scope for

the company to pursue di¤erent trajectories and create di¤erent con�gurations for

which reason traits of idealism/constructivism are very much present. In �gure 3

below, this is illustrated for a company faced with outsourcing decision-making.

<Insert �gure 3 here>

The inside of the circle formed by the two arrows represents reality �the feasible

supply chain con�gurations that the company can follow. However, not all of these

82

alternatives are pro�table and so only the inside of the cone marked by the fully

drawn lines depicts the con�gurations that are both feasible and pro�table for the

company to pursue. Finally, all supply chain con�gurations outside the circle are

unfeasible and thus by de�nition unpro�table for the company to undertake.

Nevertheless, in practice, only a subset of the total number of decision alter-

natives is considered by the company due to the strategic framework that outlines

how to approach and analyze the outsourcing problem. These supply chain con-

�gurations are represented by the inside of the cone formed by the fully dotted

line. Speci�cally, this cone portrays the idea that some con�gurations which are

intuitively conceivable from the strategic framework need not coincide with what is

either feasible or pro�table �in fact, they might even represent illusions that can

never be realized.

Keeping in mind that this illustration is a simpli�cation of a very complex mat-

ter, it captures the essence of the joint epistemological challenge which strategy and

management accounting face in relation to information production for outsourcing

decision-making as depicted in �gure 2 �an epistemological challenge that consists

in ensuring the feasibility and pro�tability of the decisions made. In turn, this gen-

erates a number of fundamental objectives for both domains to meet in order to

assist in choosing a supply chain con�guration that is feasible as well as pro�table.

At �rst, it is important to ensure that the strategic framework adopted, at least

partially, coincides with business reality, since the decision alternatives conceived

and information collected are fundamentally formed by the strategic aspects con-

sidered. Hence, if the strategic framework leaves out important aspects or over-

emphasizes irrelevant aspects, there is great risk that the supply chain con�gura-

tion chosen will be neither realistic nor pro�table for the company to implement.

However, whether the strategic framework underpinning the outsourcing decision

incorporates the central aspects of the company�s business reality will most often

be di¢ cult to determine in advance. Instead, it is preferable to base it on a learn-

ing process (Nørreklit et al. 2007) in which the company repeatedly evaluates the

success of applying the framework in order to pinpoint critical blind spots or under-

developed areas.

Still, even when the strategic framework adopted coincides with the business re-

ality, it will often be characterized by a broad scope which leaves room for thinking

up unrealistic decision alternatives. As such, management accounting�s task of eval-

uating the supply chain con�gurations proposed is to frame them in both �nancial

and non-�nancial terms. The objectives are, �rst, to assist in reining in the alter-

83

natives which are realistic, and, second, to locate the most pro�table alternative

among these.

For management accounting to accomplish these latter objectives, it is important

to understand that a proposed supply chain con�guration represents nothing more

than a possibility �and a possibility cannot be observed directly because it is some-

thing di¤erent from that which exists (Nørreklit 2011, p. 21). Instead, a proposed

supply chain con�guration is a construct of something that does not presently exist

based on something that does exist. Hence, in order to provide some assurance for

the realism and pro�tability of the con�guration considered, it must in principle be

grounded in accounting facts about the things that make up the company�s business

reality.So what are the things making up the company�s business reality? Essentially,

they are true propositions about the world that exist independently irrespective of

whether the decision-maker is aware of them or not (Nørreklit 2011, p. 19). Thus,

they may take on many di¤erent shapes. For instance, they may have a physical

existence, such as a property, plant or equipment, or they may be institutional

constructs, such as societal accounting laws.

Accepting this, a fact can be de�ned as the decision-maker�s belief about a thing

that is known by him to be true (Nørreklit 2011, p. 20), which in turn raises the

question of how to justify a mere belief as true and thereby transform it into a fact. In

theory, this necessitates establishing a correspondence between the decision-maker�s

belief and the thing itself either from undistorted observation (empiricism) or by

intuitively and infallibly grasping its full existence (rationalism) (David 2009; Ryan

et al. 2002). In practice, however, such correspondence is rarely attainable due

to measurement error or complexity, for which reason only very few beliefs about

things qualify as incontrovertible facts (true justi�ed beliefs). Furthermore, as soon

as more complex beliefs are constructed on the basis of methods (i.e. from induction

or deduction), we move away from facts and into the realm of analysis and re�ection

(Nørreklit 2011, p. 20). The truth of such complex constructs can never be fullycon�rmed but only hinted at in terms of the logical8 structure of the methods applied

and the correspondence or non-correspondence to already established facts.9

8Logic is a rational element inherent in our ability to calculate and reason in a stringent, logicalmanner. It not only concerns the reasoning of mathematics and formal logic, but also reasoningthrough the use and building of concepts (Nørreklit 1987; Matthews 1996, p. 187).

9This idea is more formally developed within Logical atomism by Russell (1918) and Wittgen-stein (1921). Here, the concepts of atomic truths and molecular truths are used. It is arguedthat an atomic truth is true because it corresponds to an atomic fact. Contrary to this, moleculartruths are not assigned any matching facts �strictly speaking, they do not correspond to facts at

84

In accounting terms, this means that the information produced either from di-

rect observation or from processing empirical observations using accounting methods

rarely represents indisputable facts. Believing blindly in the truthfulness of such in-

formation can thus lead to the wrong supply chain con�guration being chosen. Still,

such information is fundamentally tied to something that does exist and hence, al-

though possibly �awed by measurement error or simplistic methodical processing, it

is grounded in the company�s business reality. Therefore, if accounting information

is produced with caution and used with a good deal of thought, it can be of some

help in identifying the supply chain con�gurations that are feasible as well as prof-

itable to pursue.

All told, the empirical part of this paper will focus on analyzing how the case

companies meet the joint epistemological challenge of strategy and management ac-

counting when producing and using information for outsourcing decision-making.

Accordingly, we will only brie�y touch upon ontology to verify that the ontological

stance of the case companies is in fact rooted in pragmatic constructivism. On this

basis, we raise the following more theoretically profound research questions:

� How do companies structure the outsourcing decision-making process (me-

thodic), and how is accounting information produced and used as a basis for

decisions (methods)?

� To which extent do the structure of the outsourcing decision-making processand the accounting information produced underpin the selection of a supply

chain con�guration that is feasible as well as pro�table?

3 Research method

In order to provide answers to the research questions raised, a thorough understand-

ing of the speci�c company and its context is required. Consequently, we draw on a

case-study approach, since it allows us to, at least partially, capture the complexity

of the decision-making process and the managers�way of thinking. In this section,

the choice of case companies and background interviews is initially motivated. Sub-

sequently, the motives for using semi-structured interviews and the design of the

interview guide used are delineated. The section closes with some re�ections on the

di¢ culties encountered along the way.

all. Instead, their truth values are explained in terms of logical structure and the correspondenceor non-correspondence of atomic constituents (David 2009).

85

3.1 Selecting cases and background interviews

The choice of DEN and EUR as the center of attention was motivated by several

factors. First, DEN was chosen for the obvious reason that it is a well-reputed

company of considerable size with an extensive experience in strategic outsourcing

and o¤shoring. At the same time, it has performed relatively well over a long pe-

riod of time, thus indicating an ability to swiftly adapt its supply chain to changing

industry conditions. Second, easy access to detailed company information was con-

sidered important. Speci�cally, we already had a close contact at the company who

helped us identify and gain access to senior sta¤ members with leading roles in the

company�s previous and current outsourcing engagements.

The choice of EUR as the second focal point of the in-depth company analysis

was mainly motivated by the fact that the company in 2010 launched a new corpo-

rate strategy with the intent to clean up the organization and focus the company

on its core activities. Therefore, outsourcing will come to play a decisive role in the

coming periods due to the signi�cant size of the company. Furthermore, it makes

EUR an interesting counterpart to DEN as it is left with a much more fundamental

restructuring problem, where many outsourcing decisions will be directly related to

the core of the business. Consequently, we succeeded in setting up an interview with

key sta¤ members working with outsourcing in relation to this new strategy.

From the very beginning, focus was kept on uncovering the outsourcing decision-

making process in both DEN and EUR. However, to obtain a comprehensive prelim-

inary understanding of outsourcing practice and ensure some degree of generality of

the conclusions arrived at, background interviews were set up with relevant consult-

ing �rms, as we scrutinized the debate on outsourcing in professional magazines as

well as regular newspapers (e.g. Ho-Lanng 2009; Steno 2009; Weiss 2010; Monahan

et al. 2010) to identify important protagonists in the arena.

In this way, contact was established with three �rms, Munk, Andersen & Feilberg

(Mafcon), Trellis, and A.T. Kearney, all of which have signi�cant experience with

outsourcing decision-making from numerous engagements with companies di¤ering

in both size, industry context, and problem type. Furthermore, the selected �rms

all operate in the mixed zone between strategy and accounting, which made them

especially interesting for the current study. Speci�cally, interviews were set up with

partners or key employees at each �rm and succeeded by e-mail correspondence in

order to clarify additional open-ended questions.

A total of 8 interviews were conducted out of which 6 interviews with 8 inter-

viewees were coded and used as either background material or direct input for the

86

two main cases.

3.2 Data collection: The choice of semi-structured inter-

viewing and design of interview guide

As we approached the problem of studying outsourcing decision-making, various

techniques for collecting empirical data were considered, ranging from questionnaires

over structured and semi-structured interviews to dialogues.

In general, the questionnaire and structured interview formats allow researchers

to explore to what extent their own prior theoretical beliefs are re�ected in the

behavior and perceptions of signi�cant actors in the area of interest (Horton et al.

2004, p. 340). However, although certain theoretical dimensions are the focal points

of the analysis, the aim of our study is to explore the research questions in a way

that goes beyond the existing theoretical understandings. Given the complexity of

the phenomena under study, we decided to make use of semi-structured interview-

ing while being aware of the pitfalls related to this technique.10 Speci�cally, semi-

structured interviews were chosen to allow the interviewees a degree of freedom to

explain their thoughts on the outsourcing decision-making process and to emphasize

areas of particular interest or expertise that they believe they held. Moreover, this

form of interviewing made it possible through dialogue to question certain responses

in greater depth to fully capture the complexity of the issue at hand or explore the

underlying motives more directly. Accordingly, we found the semi-structured inter-

view format well-suited for our explorative study of how strategy and accounting in

combination meet the epistemological challenge, as explained earlier.

Before each interview, an interview guide was speci�cally designed for the company

in question on the basis of the public information available. The interview guides

used for consulting �rms interviews were structured identically and consisted of two

overall parts: a generic part including broad questions about outsourcing given to

all interviewees; and a speci�c part containing questions tailored particularly to the

company and person(s) interviewed. Although this latter part added to the personal

nature of the interview and thus minimized the coherence among interviews, we felt

that it was important to unleash the full capacity of semi-structured interviewing

and exploit the expertise of the individual interviewee. For the consulting �rms,

this consisted of questions concerned with general trends or aspects of outsourcing

10Some of the main weaknesses that are often mentioned in relation to semi-structured inter-viewing include response bias and re�exivity (the interviewee says what the interviewer wants tohear) (Yin 2002, p. 86).

87

debated in their publications.11

The generic part for the �rst interviews was mostly based on the prominent the-

ories on outsourcing in the strategy and accounting literature. In particular, the

structure and the questions were inspired by the decision framework presented in

McIvor (2005), as this proposes a rigid, yet theoretical, way of analyzing outsourcing

decision-making based on combining the core issues from di¤erent theories on out-

sourcing. Inspired by this, speci�c questions were identi�ed by one of us, followed by

a review and discussion among the two of us, as we re�ected upon how these issues

should be operationalized in a speci�c context. The resulting questions were �nally

categorized under a few main headings that remained the overall classi�cation for

every subsequent interview conducted with consulting �rms.

Based on the transcriptions of the background interviews, we evaluated the inter-

view process and started designing the interview guides for DEN and EUR. In this

manner, the background interviews helped pinpoint the seemingly relevant issues to

include as we assessed which parts of the theories on outsourcing were carried into

practice. Also, this process gave us an excellent opportunity to re�ne and re�ect

upon the theoretical framework of this article and to phrase the questions for the

two companies in a way that would help illuminate the relevant aspects.

Additionally, the �ow of the background interviews were studied to ensure that

the interview guides for DEN and EUR were carefully directed towards the main

issue of uncovering the speci�c outsourcing decision-making process while simulta-

neously allowing for introductory "ice-breaking" questions to make the interviewees

feel comfortable with the setting. To this end, a structure was chosen where the in-

terviewees were initially asked questions regarding their own career and education,

followed by a general talk about the company. Subsequently, more speci�c questions

were posed in relation to the outsourcing decision-making process, as we tried to

group the questions into two parts: one concerned with the strategic framework; and

one related to the collection and use of accounting information. Finally, by the end

of the interview, the interviewees were deliberately encouraged to raise additional

issues not previously addressed in the interview to make sure that no important

pieces of information were left out.

Appendix 7 provides an example of an interview guide used. It should be stressed

that the questions listed merely functioned as attention directing in order to ascer-

11As an example, Gove (2009) was used as the underlying material for several in-depth questionswhen Trellis was interviewed.

88

tain that every key issue was discussed. Thus, due to the unconstrained nature of

semi-structured interviewing, some questions were addressed in a di¤erent order or

posed in a way that di¤ered somewhat from the written version.

Both the background and case company interviews were conducted at company

sites in a peaceful environment so as to avoid unnecessary disturbances and inter-

ruptions. In particular, they were all informal in nature and of considerable depth as

they lasted from approximately 112to 3 hours. Only the interviews at A.T. Kearney

were slightly shorter and took about 40 minutes. Every interview was recorded with

a Dictaphone and subsequently transcribed.

3.3 Problems encountered

In general, the people contacted have shown great interest in our work. However,

originally our intention was to include interviews with several of the major manage-

ment consulting �rms to form a part of our background material. Unfortunately, it

was only possible to set up interviews with sta¤ members at A.T. Kearney. Fur-

thermore, one interview conducted with a small Danish consulting �rm was neither

coded nor re�ected upon in this article, as the interview was abruptly interrupted

after 30 minutes, for which reason the recorded material consisted of nothing but

abstract re�ections on outsourcing with little practical relevance. Finally, an in-

teresting interview conducted with a large Danish company was long considered for

inclusion as a third case in this article. However, since the interview mainly revolved

around o¤shoring and not outsourcing, it has been left out to maintain a clear focus

on outsourcing decision-making.

4 Cases: Descriptions and analyses

In the following, the outsourcing decision-making process in the case companies,

DEN and EUR, will be described and analyzed. Initially, a short case introduction

will be provided. Then, relying on the methodology from section 2.1 and �gure 2, we

outline, �rst, the strategic framework underpinning and structuring the outsourcing

decision-making process (methodic) and, second, the accounting information pro-

duced and used as a basis for decisions (methods).12 In continuation of this, we

analyze how and to which extent the strategic framework and accounting informa-

12The names of the two case companies have been concealed by request. Furthermore, theirlocation along with the numbers reported have been slightly changed to maintain anonymity.However, the structural descriptions are all intact and have not been modi�ed in any respect.

89

tion produced in combination meet the epistemological challenge (i.e. to ensure

feasibility and pro�tability). In doing this, re�ections draw on the theoretical ter-

minology from section 2.2 and �gure 3.

4.1 DEN

DEN is a large Danish company with approximately 7500 employees selling products

and services to people with medical problems in more than 65 countries. Its primary

shareholder is a family fond, established by the founders of the company, which

currently owns 46% of the shares corresponding to 68% of the total shareholder

votes. The company is structured around four main executive areas: (1) Global

R&D with the responsibility of developing new products and services; (2) Global

Operations comprising all manufacturing activities along with the logistics and the

design, assembly, and technical maintenance of machinery; (3) Global Marketing

with the primary responsibility of obtaining feedback from customers and developing

strategies for future product launches; and (4) Sales Region focusing on sales, export,

and customer services worldwide.

The total sales can be grouped into three main business areas, each of which is

characterized by steady growth and high customer loyalty. The forecasting of future

sales for already established products can thus be done rather precisely. Further-

more, most of the products sold are publicly subsidized, for which reason �erce price

competition is largely avoided.

Manufacturing in DEN takes place on its facilities in Denmark, Hungary, China,

and the US.13 As such, the production facilities in Denmark and the US �known as

Technical Competence Centers (TCC�s) �are primarily intended for research and

testing of new products in small scale when prototype machinery needs to be de-

signed and assembled. Later on, when product development has been completed

and the production processes have been thorougly tested, production is normally

moved to China or Hungary and increased in scale.

In the following, we will describe and analyze how DEN makes use of outsourc-

ing when establishing the supply chain con�guration for a newly developed product.

The descriptions rely on an in-depth interview and subsequent e-mail correspondence

with two key employees in Global Operations along with the company information

with which they have provided us.

13Also, DEN has a production facility in France which operates as an autonomous entity sepa-rated from the rest of the company.

90

4.1.1 Strategic framework

All boundary discussions in DEN are fundamentally shaped by a self-perception

within the company that revolves around the core competence approach and the

idea of being innovation driven. In particular, within each main business area, DEN

has carefully identi�ed a few components that are key to most of the �nal products

produced, and in which the company excels compared to its competitors. Hence,

although protected by intellectual property rights, the production processes and

knowledge tied to the manufacturing of these components are by de�nition kept

in-house. Building on this, a so called Innovation Value Stream (IVS) is put in place

to ensure a structured facilitation of product innovations related to the core compe-

tences and matching the company�s overall strategy. In e¤ect, the IVS functions as

a strategic adjustment mechanism that helps DEN to constantly meet the changing

customer needs or exploit newly identi�ed possibilities in the market.

Essentially, the IVS is a funnel that structures product development from the early

stages of idea generation to the �nal stages of full scale production of one single

product. A new development project starts on the basis of ongoing value proposi-

tions being formed from the needs and potentials identi�ed in the market by Global

Marketing. These value propositions are transformed into Target Product Pro�les

(TPPs) specifying the needs and potentials, which the products are expected to

satisfy/target if manufactured.

Every quarter, an intake of ideas is made among the TPPs, and the ones selected

subsequently enter a so called Accelerate Ideas to Market process (AIM) within IVS,

which divides the product development into manageable stages. Each stage contains

a number of coordinated activities designed to optimize the conceptualization and

manufacturing of the product solution while ful�lling the customer needs as well as

the legal and business requirements. Moreover, to ensure ongoing quality control,

every stage is succeeded by a decision gate, at which the project is reviewed by a

cross-functional group of senior sta¤ members from each executive area. Thus, it is

during the AIM process that outsourcing becomes a relevant tool to consider as the

company plans the future supply chain structure for manufacturing and distribution.

In particular, preliminary re�ections on outsourcing are made already in the early

phase when ideas are transformed into concepts, and they are further substantiated

as the concepts mature into a single product ready for manufacturing. It is therefore

essential to understand that DEN makes use of outsourcing in a very proactive way

due to the continuous strategic adjustment function that the IVS funnel serves.

91

Figure 4 below illustrates the strategic framework in DEN and in particular how

the IVS funnel adjusts the company to the current market trends by fostering inno-

vations related to its core competences.

<Insert �gure 4 here>

4.1.2 Accounting information and models

The con�guration of the supply chain, and hence which parts of production are to

be outsourced, takes place in several stages and relies on information from di¤erent

areas. Essentially, a rough business case is outlined as soon as a TPP is selected for

the AIM process within the IVS funnel. Thus, at the intake for the AIM process

(decision gate 0 ), approximately 10-20 TPPs are selected based on an early eval-

uation of attractiveness focusing mainly on a preliminary individual sales forecast,

the resources available for R&D, and the TPP�s �t with the overall corporate strat-

egy. The chosen TPPs are subsequently raised to the status of "projects" and enter

stage 0 in the AIM process which takes place at the TCCs in Denmark. During

this stage, prototypes of the product concept as well as manufacturing machinery

are developed and tested for each alternative. Accordingly, the results from these

prototype tryouts are used to revise and elaborate upon the business case in order

to describe the revenue and cost side along with the performance expected for each

project.

In order to evaluate the business impact of the product, it is important

that you, already in the early phase of concept development, prepare

a business case that tells you: Marketing says that we can sell so and

so much, and we have calculated these cost and investment estimates.

In total, this gives us a pro�t of this [amount] if we launch the product,

which is then compared to an "as is" situation that describes the business

impact had we not undertaken this project (key employee in Global

Operations).

Speci�cally, a �nancial model is used to calculate the �nancial implications of intro-

ducing the new product14, where three alternatives are normally outlined for a time

period of 5-10 years: (1) the proposal, (2) the "as is" situation, and (3) the best

alternative proposal. In this way, the �nancial model helps to reconcile the calcula-

tions from each executive area and brings about a more holistic information material

for decision-making that partially uncovers the alternative use of the resources.

14E.g. impact on cash �ow, net present value, and rate of return.

92

In determining the revenue side of the business case, Global Marketing cooper-

ates with the sales regions to prepare early country speci�c forecasts of the expected

sales volume if the product concept is brought to market. These volume estimates

are updated as the concept passes through the decision gates in the IVS funnel and

materializes into an actual product. As such, the forecasts are, to the extent pos-

sible, based on the historical information available for similar products since DEN,

in collaboration with its competitors, has installed a system in hospitals around the

world that keeps track of the products sold to customers. Furthermore, a group

of statisticians and market modeling experts repeatedly assess and challenge these

estimates for every country with hard-headed analyses based on historical market

trends for similar products.

That is, they [statisticians and market modeling experts] would say: You

[Global Marketing] have estimated that England can have an uptake of

the product looking like this. If we consider the launch of earlier products

with similar characteristics, we can see that we usually have an uptake

that looks di¤erent. Why is it that this product can have a greater

uptake �do you consider this to be realistic? Hence, Marketing needs to

investigate this and explain why exactly this product will have a better

uptake than the last three products launched within the same area (key

employee in Global Operations).

Also, to ensure that the forecasts provided by Global Marketing are realistic, incen-

tives are provided through the compensation structure so that bonuses will partly

depend on how accurately forecasted sales mimic actual sales. In this way, the

employees in Global Marketing are held accountable for their actions and have no

incentive �ceteris paribus �to deliberately boost the estimates.

Although many aspects are left open-ended as an initial concept is developed

for each project, Global Operations starts to prepare the costing side of the busi-

ness case to provide an early indication of whether the concept at hand can ever

be produced at reasonable costs. The evaluation criterion is a target cost pro�le

determined at a higher level in the organization specifying the contribution margin

that needs to be met in order to ensure pro�tability when additional overheads (e.g.

marketing) are accounted for later on. The cost price for each product concept is

based on all direct variable costs related to materials/components and labor hours

used along with wastage rates. As such, these estimates are largely determined in

collaboration with the engineers at the TCCs who develop the machinery for man-

ufacturing the concepts, since they hold extensive knowledge about the output per

93

hour as well as the number of people needed to operate the machines and handle

the product throughout the manufacturing process. Speci�cally, two di¤erent cost

prices are usually calculated to imitate both the early phase in Denmark, when the

product is manufactured in small numbers, and the �nal phase, when the production

is moved to the facilities in either Hungary or China and increased in scale.

Hence, already at stage 0 in the AIM process, Global Operations simulates sev-

eral con�gurations �or scenarios �of the supply chain to get an indication of which

parts of production can potentially be contracted out. In this respect, the target

cost pro�le and the estimated cost price of the concept play important roles; yet

these are not the only factors considered. Qualitative strategic aspects are also dis-

cussed in an attempt to ensure that no elements of high strategic importance nor

directly related to the core competences are suggested for outsourcing. In relation to

this, attention is also given to intellectual property rights since these might hinder

in-house production of certain elements in the manufacturing process. Additionally,

buying from an external vendor may in certain situations give rise to a further de-

bate on the trade-o¤ between, on the one hand, immediate access to competences

and technology not currently held in-house and, on the other hand, the potential

risk of becoming too dependent on a supplier that might turn into a competitor

later on. Still, although taken very seriously, the prevailing understanding within

Global Operations is that this kind of risk is di¢ cult to quantify for which reason

the considerations are kept very simple.

Risk is evaluated on the basis of a very simple little matrix that measures

business impact and probability. That is, you can invent all kinds of

complicated tools, but the result is never better than the inputs used.

And presumably these inputs are relatively uncertain in either case (key

employee in Global Operations).

This should be considered in connection with the fact that DEN has a long history

of working with the same suppliers on many parts, which naturally reduces the risk

of opportunism. In this manner, several make-or-buy decisions are predetermined

and can be settled already at stage 0.

All told, the business case developed for each concept is assessed at decision

gate 1, after which 1-3 concepts are selected for further maturing within the AIM

process. Essentially, stage 0 is then repeated during stage 1, where the concepts are

re�ned and matured to provide better and more realistic estimates of the aspects

in the business case. Thus, this also entails revisiting the planned con�guration of

94

the supply chain and the parts intended for internal production. Finally, at decision

gate 2 the number of concepts is narrowed down to 1, and the design speci�cations

are frozen. During the succeeding stage 2, the product design is veri�ed from a

technical as well as a business perspective. Additionally, the process speci�cations

for manufacturing are �nalized.

Consequently, at decision gate 3 all product and manufacturing process speci�-

cations are locked and time has come for investing in equipment for in-house pro-

duction and settling the supplier contracts not already established at decision gate

1 or 2. In this respect, decision gate 3 represents the point of no return where the

money is spent. However, to reduce the risk of choosing an improper supply chain

setup scaled for a production level that later on turns out to deviate from the actual

demand, DEN deliberately tries to delay the "full-blown" investment decisions at

gate 3 in order to decrease the uncertainty of the sales forecasts.

We look at what we have, how �exible we can be, and hence how long

we can wait before spending the money. And that is about all that

we can do. This is where we can try to be sensible �that is, to delay

the investments as long as possible while taking into account that if the

predictions [sales forecasts] actually hold � i.e. we have some kind of

blockbuster �we will be able to keep pace. Because this is also impor-

tant; we cannot play too safe and thus be incapable of delivering the

volumes forecasted by Marketing if these estimates actually hold true

(key employee in Global Operations).

Hence, although extensive preparations and analyses precede decision gate 3, DEN

remains cautious and increases the investment level stepwise. For this reason, deci-

sion gate 3 coincides with stage 3 where the actual production capacity is purchased

and the manufacturing of products is started.

Following this, the full launch of the product takes place at decision gate 4

and during stage 4, where production in many cases is o¤shored to the facilities in

either Hungary or China. Finally, the business project is evaluated and closed at

decision gate 5. This gate thus marks the end of the IVS funnel and functions as

a feedback mechanism that can help improve the con�guration of the supply chain

and the outsourcing analysis for future innovations. Consequently, the obtained

market position is evaluated and compared to the forecasted position. Moreover,

information on actual sales, costs, and quality is reported back to the respective

executive areas, where it can be used for future projects or compensation purposes,

as already described for Global Marketing.

95

4.1.3 Analysis

The descriptions in the previous two sections clarify that outsourcing decision-

making in DEN is an integrated part of a greater strategic decision process concerned

with proactive and timely innovation based on the company�s core competences. The

entire setup clearly illustrates that DEN is higly interested in understanding its busi-

ness reality before making any outsourcing decisions and thus shares the ontological

stance of pragmatic constructivism. In the following, the framework from section

2.2 is used to analyze how the strategic framework and accounting information pro-

duced in combination meet the epistemological challenge of determining a supply

chain con�guration that is both feasible and pro�table. Initially, focus will be kept

on the strategic framework and how it underpins realism in the creation of decision

alternatives. Subsequently, the construction of accounting information is studied

to clarify how this assists in establishing a supply chain con�guration that is both

�nancially and non-�nancially grounded in DEN�s business reality.

In order to understand how proposed supply chain con�gurations come into exis-

tence within DEN�s strategic framework, it is necessary to understand how products

possibilities are conceived, since outsourcing decision-making is inextricably linked

with product innovation. Accordingly, product possibilities arise as the market

value propositions are assessed on the basis of the current in-house technology (core

competences). In this way, a feasibility check is by design built into the strategic

framework, as the scope of product possibilities is reduced to those compatible with

the company�s current resource endowment. Consequently, from the very beginning,

the potential supply chains constructible are all motivated yet also constrained by

this product scope.

Acknowledging this, the supply chain con�guration put in place is not selected

instantaneously from a set of pre-de�ned decision alternatives specifying each and

every step of the supply chain. Instead, the con�guration is established and shaped

in small steps as actors from the main executive areas interact with each other and

also with potential suppliers throughout the stages and gates in the IVS funnel.

Thus, the structural design of the strategic framework, with stages and decision

gates, adds to the realism of outsourcing decision-making and underpins the selec-

tion of a feasible supply chain con�guration. This point is further emphasized by

the fact that each successive run is explicitly evaluated at the end of the funnel.

In this manner, a learning dimension is naturally embedded in the IVS structure,

enabling DEN to constantly develop and adjust the procedures used for analysis of

in-house activities as well as for evaluation of suppliers in order to, at least partially,

96

ensure that the decisions arrived at are in accordance with the company�s business

reality.

Having clari�ed that the strategic framework by design underpins the construction

of a feasible supply chain con�guration, it is time to re�ect upon how the accounting

information produced assists in ensuring a con�guration that is both feasible and

pro�table.

Referring to section 4.1.2, accounting information is produced throughout the

IVS funnel as actors from the four main executive areas interact with each other and

use the tools put in place for costing, sales forecasting, and supply chain simulation.

The primary task is to determine not only the cost side but also the revenue side

of the product possibilities, since the revenue side � in terms of sales volume �

implicitly a¤ects the cost structure and thus the optimal supply chain con�guration

for production. For instance, a large volume might justify investing in expensive

fully automated, in-house machinery to bring down the variable cost price while a

semi-manual production facility might do for a smaller volume, thereby increasing

the variable cost price. Or perhaps the sales volume allows a part of the production

to be outsourced to a third-party supplier in return for an even lower quotation

price.

Keeping this in mind, the construction and use of accounting information at the

stages in the IVS funnel basically follows an actor-based epistemology which includes

four integrated processes: subjecti�cation, externalization, objecti�cation, and in-

ternalization (Arbnor and Bjerke 1997, pp. 175-178). Initially, each actor brings

to the table his expertise and experience from previous engagements, which indeed

represent his subjective view on what is feasible and/or pro�table. In interaction

with other actors, a process of externalization subsequently occurs, in which the

various standpoints are communicated and challenged individually and in relation

to each other, as model estimates on sales, costs, and product speci�cations are

brought to the table. In this way, the individual actor�s understanding of the sit-

uation is advanced through re�ection, thus possibly prompting him to reinterpret

what is feasible let alone pro�table. As a result of this process, information is objec-

ti�ed as the set of beliefs and model estimates that correspond with the individual

actor�s expertise and experience as well as cohere across executive areas (social ac-

ceptance). Accordingly, this information can be used for decision-making at the

subsequent gate and as a starting point for the initiatives taken during the next

stage (internalization).

97

Understanding this integrated process, it becomes possible to appreciate the �ex-

ibility of the actor-based epistemology. Speci�cally, the actor-based epistemology

allows DEN to adjust the level of human intervention (subjectivity and external-

ization) in critically assessing model estimates and thus produce information that

�ts the uncertainty level at the various gates in the funnel. This strengthens the

reliability of the make-buy decisions at the gates and contributes to establishing a

supply chain con�guration that is both feasible and pro�table.

At the initial stages of the funnel (stage 0 and partially stage 1), great un-

certainty exists with respect to product speci�cations, thereby leaving many open-

ended questions on both the supplier side and the customer side. Thus, under such

conditions, the process of externalization can come to play a valuable role because

unquanti�able aspects such as the actors�individual intuition and experience from

earlier engagements become excellent information sources. In particular, the inter-

action between actors can increase the truthfulness of the information produced for

decision-making, since it allows for a meaningful reduction in the complexity and

number of decision parameters to be considered, as the parameters are re�ected

upon from di¤erent angles and weighted against each other. For instance, referring

to section 4.1.2, outsourcing a given activity might be ruled out from the very be-

ginning, based on the intuitive argument that it is too tightly linked to the core

competency area and thus of high strategic importance to DEN. Likewise, a given

supplier alternative might be excluded immediately when weighted against the po-

tential risk of supplier dependency. Hence, the process of externalization addresses

the uncertain issues as opposed to automatically removing them by assumption or

simpli�cation in order to make use of rigid (mathematical) models.

Consequently, the information produced might not represent facts in a very strict

sense, since correspondence to already established facts can be di¢ cult to trace and

verify due to the opaqueness of human interaction/intuition and the lack of a phys-

ical product. Nevertheless, through the experience and expertise of the individual

actors and the coherence established across disciplines, the information produced is

broadly anchored in the company�s business reality. It thus provides some assur-

ance in the feasibility and pro�tability of the consequences envisaged, should the

subsequent gate decisions be based upon it. The make-buy decisions made at the

beginning of the funnel are therefore chie�y of strategic character. They essentially

specify which parts of the supply chain must be exempt from further outsourcing

analysis and remain in-house at any price.

At the later stages of the funnel (stage 2 and onwards), most of the product spec-

98

i�cations have been frozen while parts of the supply chain have been put in place.

Accordingly, the complexity and hence uncertainty related to establishing the re-

maining parts of the supply chain are reduced considerably. For this reason, the level

of human intervention (subjectivity and externalization) in critically assessing model

estimates can be toned down, since information production is basically reduced to

mechanically updating cost and sales estimates as the �nal product speci�cations

and supplier quotation prices become known.

The truthfulness of the information produced now mostly depends on the pre-

cision of the models in use, since correspondence to the physical aspects of the

product can promptly be established. In this way, the previously mentioned learn-

ing dimension embedded in the strategic framework proves valuable and reinforces

the internalization process, as the models and databases used for costing and fore-

casting have been repeatedly tested and re�ned by prior projects. For this reason,

the information produced on sales and costs is indeed grounded in the company�s

business reality and can, ceteris paribus, be constructed with rather high precision.

A precision that is further strengthened by the incentive system put in place as it

holds the employees accountable for the estimates produced and encourages them

to strive for realism.

However, the realized sales stay unknown until after the supply chain structure

has been put in place, and since sales a¤ect the cost structure, some uncertainty

remains. Therefore, in dealing with the time lack between costs and revenue, human

intervention still plays a role. This is most clearly illustrated at gate 3 where major

investment decisions may be postponed to further improve the precision of sales

estimates as product samples are tested on customers.

Still, the general trend in the accounting information produced at the later stages

of the IVS funnel suggests an increasing focus on operational issues (i.e. precision of

cost and sales estimates) rather than strategic issues (risk and �exibility). In turn,

this is re�ected in the make-buy decisions which by and large resemble a traditional

comparison of in-house production costs with supplier quotation prices.

Thus, it is fair to conclude that in DEN the strategic framework, as well as the

accounting information produced, holds the potential to ground the company�s out-

sourcing decision in its business reality to ensure the selection of a supply chain

con�guration that is both feasible and pro�table.

99

4.2 EUR

EUR is a family owned European company with approximately 29000 employees. It

develops, manufactures, and sells mechanical and electronic components for several

industries around the world. The company has 101 factories situated in 24 countries

in addition to 139 sales companies and 470 distributors worldwide.

Before 2008, the company was growing rather rapidly due to expanding markets

as well as mergers and acquisitions. However, the �nancial crisis erupting in 2008

severely impacted the company�s �nancial performance reducing growth to only 2%,

followed by a negative growth of 12% in 2009. As a consequence, the company�s

�nancial performance has changed from several years with satisfactory �nancial

results to a loss in the years 2008-2009.

In an attempt to respond to the problems, the organizational structure of EUR

has been changed. Before 2010, the company was organized around three au-

tonomous divisions with a broad scope. However, an evaluation of the company�s

business areas on the basis of current market positions as well as growth potential

has led to the identi�cation and clustering of both core and non-core business areas.

Accordingly, the organization today consists of two segments. The �rst segment

encompasses the business activities deemed to be core areas and is grouped into �ve

separate divisions all of which are focused on developing and manufacturing com-

munication equipment. Following this, the second segment contains the remaining

business activities currently ranked as non-core areas of the company.

Based on this new structure, the intention is to clean up the company and de-

termine which parts of the organization should not be kept in-house in the future.

In charge of overseeing and handling this assessment is the corporate function EUR

Corporate Analytics (ECA), which is governed by a cross-functional board including

the heads of the �ve new divisions within the core segment, along with senior sta¤

experts in innovation, operations (supply chain), and sales and marketing. In this

manner, a very important task of ECA is the development and support of common

tools and processes that can help to systematize the analysis of whether company

processes/activities are to be outsourced or not.

In the following, we will thus describe and analyze how EUR plans to carry out

the clean-up process based on the tools and processes developed by ECA. The de-

scriptions and analysis rely on an in-depth interview and subsequent e-mail corre-

spondence with two key employees in ECA along with publicly available company

information.

100

4.2.1 Strategic framework

Key factors in EUR�s �nancial di¢ culties have been the company�s broad scope

and manufacturing depth, as they have prolonged the ramp-down time of stock

and machinery. Furthermore, due to a high degree of technological enthusiasm in

the company, innovation has been too diversi�ed, without a clear focus on central

customer needs or product pro�tability.

Therefore, as hinted in the case introduction, the main objective of the strategy

launched in 2010 is to focus the company on its core competences as well as reduce

the complexity of the product portfolio. Thus, in contrast to DEN where the core

is already identi�ed and drives innovation along with outsourcing, EUR faces a fun-

damental clean-up process where the boundary discussions are related to the very

heart of the business. In this way, EUR tries to reach the same evolutionary stage

as DEN by transforming the company from being purely manufacturing driven into

being a supply chain company powered by innovation.

To reach this later evolutionary stage, the strategic model underpinning EUR�s

transformation is basically divided into two phases. As described in the case intro-

duction, the �rst phase has already been completed and concerned regrouping the

company�s business areas into segments so as to identify which areas can be con-

sidered as a part of the core and non-core business, respectively. Thus, this phase

has essentially functioned as a very coarse clean-up (elimination) process, where the

product and service portfolio has been heavily reduced, since many non-core busi-

ness areas will subsequently be sold o¤while others will be developed in partnership

with relevant companies.

At the time of writing, the second phase is still in process and focuses on the core

segment. In this phase, a much more meticulous approach is taken on the basis of

the tools developed by ECA. The objective is to evaluate the supply chain con�gura-

tions for each product within the core segment to clarify which processes/activities

are to be contracted out in order to ensure optimal conditions for the core business

areas henceforth (outsourcing decision-making process). Speci�cally, the problem is

thus approached from two perspectives: an in-house view of the company and an

external market view of suppliers and customers.

The idea is to resolve the previous problems of unfocused innovation and inap-

propriate manufacturing depth by �rst identifying which customers are important

for improving business. As a result, ongoing engagements with customers of lit-

tle importance are to be slowly phased out and delicately handed over to relevant

101

suppliers. Following this, EUR�s supply chain con�guration for the product is re-

viewed in the light of the needs identi�ed among the customers classi�ed as impor-

tant. In this respect, a make-or-buy analysis plays a crucial role in identifying the

processes/activities which are potential outsourcing candiates.

To maintain a holistic perspective in the outsourcing decision-making process,

the outsourcing candidates are �nally discussed by the cross-functional board of

ECA. In this way, the board represents an additional management view, ensur-

ing that other relevant (qualitative) strategic aspects are also contemplated before

reaching a decision on the preferred recon�guration of the supply chain.

Figure 5 below illustrates how a mismatch has come to exist between EUR�s cur-

rent competences and the market. The strategic framework portrayed represents

the company�s reaction to these circumstances as it attempts to close the gap by re-

structuring company activities and focusing future innovation on its core activities.

In the next section, �gure 5 will be used as the point of reference when elaborating

on information production at the various stages.

<Insert �gure 5 here>

4.2.2 Accounting information and models

Referring to �gure 5, the core and non-core segments have already been identi�ed

in the �rst phase, as the market potential and pro�tability of the company tech-

nologies have been assessed. Thus, after the regrouping, the revenue of the business

activities classi�ed as core is approximately e 1.7 billion, while the market potential

is around e 22 billion. Furthermore, EBIT / Net Sales is about 15% for the core

segment compared to a company average of 10%.

Entering the second phase, where focus is on the core segment, the revenue size,

turnover ratios, and pro�tability of the customers are evaluated to help identify

which customers are important and should be the future center of attention. Fol-

lowing this, the selected customer base is analyzed from a market view as value

propositions are developed for each customer category to pinpoint the most essen-

tial needs along with the most appreciated features in the current product portfolio.

Thus, product turnover ratio, sales volume, and pro�tability are used as one type of

indicators of purchasing preferences. Furthermore, the various customer categories

are interviewed and asked to complete questionnaires so as to identify the speci�c

needs of the individual customer as well as the common trends among larger clus-

102

ters. In this respect, it is explored why customers buy EUR�s products and what

features they are willing to pay for. Also, e¤ort is spent on uncovering the sales

interdependencies between di¤erent products, since some customers buy packages

of products; that is, the presence of one particular product may not add signi�cant

value to the package at hand, yet if not included, the customers may abstain from

buying the package.

Building on the value propositions prepared, an in-house view of company processes

and activities is taken. In particular, it is examined which processes (e.g. machin-

ing, manufacturing, or logistics) are vital in delivering the perceived customer value.

This includes a qualitative assessment of how tightly the processes are linked to the

value creating features identi�ed in the value propositions. The processes labeled

as non-value creating are immediate candidates for outsourcing whereas the value

creating processes move on to yet another round of analysis.

This time attention is paid to the activities � or competences � within the

processes for the purpose of pinpointing those providing the company with dif-

ferentiation compared to the competitors. Thus, a particular activity within a value

creating process (e.g. engineering) may be �rm-speci�c (e.g. highly specialized en-

gineering skills) while the rest of the activities are rather generic. Accordingly, each

activity is evaluated against the supplier market to determine whether it is readily

available and thus generic. If this is the case, it will be carefully considered for

outsourcing.

As indicated in �gure 5, a make-or-buy analysis involves several aspects, whether

it is related to a non-value creating process or a value creating, yet generic, activ-

ity. Essentially, ECA attempts to model EUR�s supply chain con�guration as it

will come to look under the make-or-buy alternative. Initially, focus is on the cost

side where the Cost of Goods Sold (COGS)15 is estimated if the generic activity is

kept in-house. In this respect, a detailed analysis of the in-house costs is prepared,

specifying the e¤ect on direct labor, raw materials, and �xed costs related to the

production. Hence, this part of the analysis can be carried out with rather high

precision, since the company has established fairly advanced costing principles and

data keeping methods throughout the manufacturing process. Following this, the

market is screened to determine the COGS if the activity is outsourced. Conse-

quently, suitable suppliers are evaluated on the basis of a thorough analysis of the

total cost of ownership (TCO) established from quotation prices and workshops with

15Notice that COGS is here de�ned as the costs attributable to the production of the goods soldby a company. In EUR, it includes both the direct costs and the overheads that can be tied to theproduction of the goods.

103

the suppliers involved.

As a part of the cost analysis when modeling the supply chain con�guration of

the make-or-buy alternative, ECA has developed econometric models for estimating

the cost of being �exible. Speci�cally, this involves analyzing the cost of reaction time

when production needs to be ramped up or down to meet �uctuations in demand.

In this manner, outsourcing is seen as a strategic tool that can help transform parts

of the COGS from �xed into variable costs, since freeing up �nancial resources

currently tied to the labor and capital used for in-house production will make the

company more �exible. For instance, in considering ramping down, ECA takes into

account such elements as severance fees and union costs to employees laid o¤. Also,

attention is directed to the costs of closing a factory, where aspects such as the sunk

costs of the �xed assets are assessed.

However, an outsourcing analysis is not based on costs alone; it also involves

a qualitative assessment of the loss or gain in technology and competences along

with the logistical and manufacturing �ow through the supply chain. In relation to

this, ECA has developed a number of tools for putting the information produced

on these intangible aspects on the same footing as the results of the cost analysis.

For instance, in considering outsourcing of a factory�s activities, employees or other

stakeholders with expert knowledge on relevant issues (e.g. logistics, manufacturing,

sales and marketing, or management) are asked to �ll in a questionnaire concerned

with the importance of the qualitative aspects as well as costs in the concrete out-

sourcing situation. Speci�cally, the questionnaire is based on a Likert scale scoring

system, where the meaning of the numerical levels is carefully de�ned and explained

to the respondents in advance. The results of the questionnaire are subsequently

discussed at a forum where each respondent is asked to justify and explain his par-

ticular answer on given issues in front of the other respondents as well as a selected

group of outsiders invited for the purpose of challenging the scores from an external

and independent point of view. This justi�cation procedure thus adds substance to

the questionnaire and functions as a psychological mechanism making the individ-

ual respondent carefully consider the questions asked. Based upon the results of

the questionnaire, a �nal score is subsequently calculated for each of the intangible

and cost aspects as relative weights are assigned to the individual questions. The

�nal scores thereby indicate the relative importance of the separate aspects in the

concrete make-or-buy situation. For instance, the �nal score on costs might turn out

to be higher than the score on the remaining intangible aspects. Consequently, the

cost issue will carry greater weight in the �nal outsourcing decision when evaluating

whether the factory�s activities/processes are to be contracted out.

104

Finally, an outsourcing analysis includes a risk assessment of outsourcing a par-

ticular activity/process as EUR seeks to anticipate the optimal response strategy.

In this respect, a signi�cant risk of outsourcing is supplier dependency due to op-

portunism. Consequently, EUR attempts to safeguard against this risk by diver-

sifying the outsourcing of a part (e.g. of the factory�s activities/processes from

before) to several suppliers or the knowledge involved in an activity/process to sev-

eral places. However, it also entails keeping certain generic activities or non-value

creating processes in-house just to maintain control of the supply chain. Still, EUR is

walking a tightrope, since reducing the supplier depth is also a priority. Accordingly,

the company looks for suppliers with whom it can build long-term partnerships so

as to enhance productivity as well as reduce the risk of opportunism.

In total, the information produced from the make-or-buy analysis illuminates the

multifaceted consequences for the supply chain con�guration for the particular prod-

uct/process should this activity/process be outsourced. For this reason, the �nal

outsourcing decision on the supply chain con�guration is made by the board of di-

rectors at ECA based on the accumulated material from these make-or-buy analyses

(i.e. the management view in �gure 5). Nevertheless, during the concluding deci-

sion process, it is recognized that any member of the board may bring "special"

knowledge to the table con�icting with the information already produced by ECA.

However, for the board to include this information as a part of the basis for the

decision, the arguments have to be substantial.

4.2.3 Analysis

The descriptions in the previous two sections establish that outsourcing decision-

making is a part of a fundamental clean-up process in EUR that serves the purpose

of bringing the company back on track and focusing it on its core competences. In

this way, the descriptions illustrate that EUR has come to show great interest in

understanding its business reality and thus subscribes to the ontological stance of

pragmatic constructivism. Similarly to the case of DEN, the framework from sec-

tion 2.2 will be applied next to analyze how the strategic framework and accounting

information produced meet the epistemological challenge of determining a supply

chain con�guration that is both feasible and pro�table. Again, focus will initially

be kept on the strategic framework after which attention will be drawn to the pro-

duction and use of accounting information.

Compared to DEN, EUR is not in a situation where outsourcing decision-making

105

is related to the creation of new product possibilities (innovations). Up until 2010,

product innovation was an unstructured and uncontrolled bottom-up process chie�y

driven by the technological enthusiasm in the company (internal focus) but without

much consideration given to pro�tability. Referring to �gure 3, a product portfolio

has thus been created partly consisting of products that might be technologically

feasible yet unpro�table to manufacture. The underlying reasons for this have been

a lack of focus on customer needs (external focus) along with a series of in�exibly

con�gured supply chains where most activities have automatically been kept in-

house on the grounds of a strong and somewhat conceited believe in the company�s

own ability to handle things better than any possible third-party supplier.

In this respect, the strategic framework described in section 4.2.1 does not serve

the purpose of underpinning feasibility in the creation of new possibilities since fea-

sibility is already present. Instead, it is intended to draw up the guidelines for a

comprehensive clean-up in which unpro�table possibilities contained in the portfolio

already are eliminated and the supply chain con�gurations for the remaining possi-

bilities are optimized. In the light of �gure 3, it is about redirecting and narrowing

down the strategic scope to match the part of EUR�s business reality that is both

feasible and pro�table.

To meet this epistemological challenge, the strategic framework directs attention

to each of the blind spots that has resulted in the �nancial di¢ culties. First, the

framework is deliberately structured and controlled top-down so as to avoid trans-

ferring the prior problems of unstructured innovation to the clean-up process as

well. Accordingly, top management has outlined the core competences during the

�rst phase, thereby streamlining the subsequent reduction in the product portfolio

and the company�s activities/processes. Additionally, during the second phase, a

board of directors (management view) with expertise from di¤erent areas is put in

place to oversee the analysis of the core segment and ensure consistency in the deci-

sions made with respect to the con�guration of the supply chain for the particular

product.

Secondly, a clear and explicit focus on the customer needs is incorporated into

the framework (market view) so as to improve the revenue side of the business and

lay the ground for a thorough analysis of the cost side. In the �rst phase, the coarse

clean-up has been based on a market evaluation of current business activities in

order to identify on which activities the company�s limited time and resources are

to be spent. In the second phase, the outsourcing decision-making process is partly

based on an identi�cation of the most important customers followed by an explicit

106

examination of their needs (value propositions).

Finally, a thorough inspection of EUR�s internal processes/activities is built into

the framework (in-house view) in order to enhance �exibility and reduce costs. This

entails a critical analysis of the current supply chain con�gurations for the various

product possibilities to determine whether they are optimally structured and pro-

vide the necessary �exibility for ramping production up and down.

Having clari�ed that the strategic framework seeks to improve company pro�tabil-

ity by addressing the critical blind spots that have caused the �nancial problems in

the �rst place, we can proceed to re�ect upon the production and use of accounting

information in relation to this. Referring to section 4.2.2, accounting information is

produced in both the �rst and second phase of the strategic framework. In the �rst

phase, the primary task concerns establishing the right �nancial key �gures for elim-

inating non-core business areas. Subsequently, it entails preparing information for

outsourcing decision-making empirically grounded in EUR�s business reality. Since

the interest of the present article is with outsourcing, we will in the following concen-

trate the analysis on the production and use of accounting information within the

outsourcing decision-making process in the second phase, as it is outlined in �gure 5.

Underpinning the construction of information for outsourcing decision-making in

EUR seems to be a general appreciation of Porter�s value chain framework (Porter

1980; Porter 1985) along with the thoughts on strategic cost accounting found in

Shank and Govindarajan (1989; 1992). Speci�cally, the construction and use of ac-

counting information in EUR follow an analytical epistemology which includes the

following steps: (1) formulating the problem, (2) designing methods and collecting

data, (3) solving the model (simulation and evaluation of e¤ects), and (4) reporting

results for decision-making (Arbnor and Bjerke 1997, p. 292).16

Formulating the problem means rephrasing the overall problem into a set of

investigative problems. In relation to outsourcing decision-making, this process

starts as the strategic framework breaks down the evaluation of an existing supply

chain con�guration for a product into the problem of identifying value creating

processes based on a market view and an in-house view. It subsequently continues

as the problem of evaluating the value creating processes is further divided into the

problem of categorizing activities as being either generic or capable of providing

di¤erentiation. Finally, the make-or-buy analysis is broken down into a study of

16To be precise, the analytical epistemology described resembles "a plan for a guiding study" inArbnor and Bjerke (1997). However, for the purpose of this article, the steps have been rewrittenslightly to �t EUR.

107

costs, technology, competences, manufacturing and logistical �ow, along with risk.

Based upon the formulation of the problem, methods/measures are designed for

empirical investigation and data is collected. In this respect, there seems to be a

view within EUR that decisions are best founded on quantitative information. Con-

sequently, most of the investigative problems are at least partially studied on the

basis of quantitative methods/measures. For instance, apart from semi-structured

interviews, customer needs are identi�ed from questionnaires as well as quantitative

accounting measures such as turnover ratio and sales volume. Furthermore, in the

make-or-buy analysis, data on in-house costs is extracted from the databases which

automatically register relevant costs using appropriate costing principles (methods)

throughout the manufacturing process. Likewise, questionnaires are used to collect

data on the importance of more subtle aspects such as the loss/gain in technol-

ogy and competences or the logistical and manufacturing �ow through the supply

chain. In this respect, it should be noted that the data collected indeed represents

a construction of information in its own right.

Following this, more complex information is constructed as di¤erent supply chain

con�gurations are simulated and evaluated in the make-or-buy analysis. This in-

cludes allocating costs to the supply chain con�guration considered under the make-

or-buy alternative. It also entails calculating a score from the questionnaires indi-

cating the relative importance of both intangible and cost aspects in the concrete

outsourcing situation. This information is �nally reported to the board of directors

that based on this makes the de�nitive decision on how the supply chain is to be

recon�gured for the given product.

To appreciate the analytical epistemology for information production in relation to

outsourcing decision-making, it is best contrasted with the actor-based epistemol-

ogy found in DEN. Particularly, while the information production in DEN is actively

shaped by the decision-makers through externalization, the decision-makers do not

play an active role in the information production in EUR. Instead, information is

constructed in a predetermined and structured way decided by the formulation of

the problem and the rigid models for data processing put in place. Thus, within this

setup, the actors in the information production are essentially regarded as nothing

more than information automatons from which quantitative data can be extracted

for subsequent modeling purposes.

Keeping this in mind, it is important to realize that outsourcing decision-making

in EUR concerns recon�gurations of existing supply chains which potentially entail

employee layo¤s and a restructuring of work tasks. Hence, under such circum-

108

stances, it is questionable whether the actor-based epistemology from DEN �had it

been put in place in EUR �would be more capable of producing information that

could support optimal outsourcing decisions for the company as a whole. The rea-

son is that structural changes and layo¤s might create organizational inertia which

can encumber the collaboration (externalization) between the actors engaged in the

information production as they try to distort the information produced in order to

serve their own interests (suboptimization). This problem is further strengthened by

the fact that so far innovation has been an unstructured and decentralized process

driven by the individual actors where little attention has been given to EUR�s overall

�nancial performance. Indeed, this might indicate that neither the will nor the com-

petence to collaborate across areas currently exists in the company. Consequently,

under such conditions, the analytical epistemology established seems more capable

of handling the outsourcing decision-making process, since the information produc-

tion and use are centralized to maintain an explicit focus on �nancial performance

while simultaneously ensuring that the information produced still re�ects the vari-

ous areas of expertise within the organization.

However, acknowledging that the analytical epistemology might be the best alter-

native given EUR�s current situation, it is important to understand that the rigid

structure and the explicit focus on the current �nancial performance come at a cost.

Speci�cally, the use of questionnaires to reveal such complex factors as the loss/gain

in competences and technology along with their relative importance compared to

costs can be problematic. Although the way of using the questionnaires in EUR is

quite ingenious, as respondents are confronted with their answers, it is still impeded

by the fundamental problems of the survey technique. That is, only the aspects of

technology and competences which are speci�cally addressed in the questionnaire

will be revealed. In this way, it will most probably be di¢ cult from a sequence

of questions to capture the full impact of outsourcing on in-house technology and

competences.

Indeed, this gives reason to treat the cost estimates produced in the make-or-buy

analysis �however detailed they might be �with a good deal of caution since they

are calculated quite mechanically, based on historical data. Yet, a recon�guration

of the supply chain might have radical implications for competences and technology

as it involves a new supplier relationship with which EUR has little or no previous

experience. Consequently, this might fundamentally change the cost structure of the

supply chain con�guration and render the historical data used for cost allocation and

simulation partially obsolete.

109

Realizing these shortcomings, the analytical epistemology in EUR tries to com-

pensate for the rather mechanical construction of information at the very last step

when the information produced is confronted with the human intuition of the board

of directors in ECA. In principle, this leaves a door open for debating the more

subtle qualitative aspects of the outsourcing problem. However, since the prevailing

notion in EUR seems to be that decisions are best based on quantitative informa-

tion, such arguments will most probably not carry the same weight, if any, in the

de�nitive outsourcing decision.

Still, the accounting information produced on the basis of the analytical episte-

mology grounds the outsourcing decision in EUR�s business reality, since the models

are based on empirical observation of things that actually exist. Hence, even if the

more complex information constructed from the collected data does not represent

facts and thus needs to be used cautiously, it can indeed help select a supply chain

recon�guration for the individual product that is both feasible and pro�table. The

reason is that the information produced illuminates the outsourcing problem from

several angles, all of which are fundamentally shaped by the strategic framework as

the overall problem is formulated into a set of investigative problems in the analyt-

ical epistemology. Therefore, the angles address exactly those blind spots that have

created the unpro�table supply chain con�gurations in the �rst place. Consequently,

there is good reason to believe that the information can assist in determining the

proper corrective actions to be undertaken.

All told, it is fair to conclude that using the strategic framework in combination

with the accounting information produced holds the potential to ground EUR�s

outsourcing decisions in its business reality to ensure the selection of supply chain

recon�gurations for the various products that are both feasible and pro�table.

5 Conclusion and implications

In summary, this article has developed a conceptual framework for studying the in-

terdependency between strategy and accounting in outsourcing decision-making in

order to pave the way for a discussion of feasibility and pro�tability. Following this,

the article has analyzed the outsourcing decision-making process in two case com-

panies in which outsourcing is seen as an important tool for optimizing the supply

chain con�gurations.

110

In DEN, outsourcing is used proactively to establish new supply chain con�gurations

for product innovations. In this manner, the outsourcing decision-making process

is built around a strategic framework that facilitates an actor-based epistemology

for the production and use of accounting information. Speci�cally, the central el-

ements of the observed actor-based epistemology are: i) the decision-makers are

actively engaged in the information production; ii) intuition (subjecti�cation) and

human interaction (externalization) play important roles in the information produc-

tion (objecti�cation); and iii) the time frame allows the supply chain con�guration

to be determined in steps as information is repeatedly updated and re�ected upon

(internalization).

In general, the strategic framework in DEN fosters ongoing learning. This hap-

pens through internalization and it is made explicit at the very last stage of the IVS

funnel where the success of the outsourcing decision-making process is evaluated for

the bene�t of future product innovations. Consequently, critical blind spots within

the strategic framework itself can be identi�ed, yet the evaluation process also up-

dates and expands the decision-makers�intuition. In this way, the role of accounting

information becomes twofold: it serves as input for the speci�c gate decisions while

simultaneously stimulating the decision-makers�intuition for future decision-making.

Over time, this will �ceteris paribus �gradually improve the decision-makers�ability

to exercise good judgment in the outsourcing decision-making process when dealing

with uncertainty or the complexity of such aspects as the loss/gain in technology

and competences. Altogether, using the strategic framework in combination with the

accounting information produced thus facilitates the construction of supply chain

con�gurations that are both feasible and pro�table.

In EUR, outsourcing is used as a tool for radically cleaning up the organization

and recon�guring the supply chains for the existing product portfolio in reaction

to �nancial di¢ culties. Previously, EUR was also driven by a strategic framework

that facilitated an actor-based epistemology. However, the scope of the underlying

strategic framework was too captivated with the technological aspects of product

innovation without focusing enough on pro�tability. As a result, the outsourcing

decision-making process currently under way is underpinned by a new strategic

framework which facilitates an analytical epistemology for the production and use

of accounting information. More speci�cally, the essential elements of the observed

analytical epistemology are: i) the decision-makers are not actively engaged in the

information production; ii) models and predetermined decision rules play important

roles in the information production where actors are regarded as nothing but infor-

111

mation automatons; and iii) the pressure to react entails a shorter time frame in

which information is brought to the table and one de�nitive decision on the supply

chain con�guration is made.

Hence, the primary role of accounting information is to serve as direct input for

decisions on the appropriate supply chain con�gurations. To this end, the appar-

ently objective nature of the accounting information allows it to be used for the

purpose of removing debatable issues from the outsourcing decision-making process

and turns the decision-making process into a matter of calculation (Hopwood 1974).

It thus seems to be a well-suited immediate response to the situation that EUR faces

since recon�guration of existing supply chains might involve dealing with delicate

questions of layo¤ or the restructuring of work tasks. In sum, using the strate-

gic framework in combination with the accounting information produced can assist

in reestablishing pro�tability to an already feasible product portfolio. However, it

might lead to some potentially pro�table possibilities being killed in the "heat of

battle".

In conclusion, the cases reveal two di¤erent approaches to outsourcing decision-

making, each of which seems appropriate given the speci�c circumstances. Fur-

thermore, the approaches clearly illustrate that management accounting and strat-

egy play important roles in the outsourcing decision-making process, and that they

are tightly integrated as the strategic framework (methodic) shapes the accounting

methods used for information production. Consequently, accounting information

plays a crucial role in grounding the selected supply chain con�gurations in the

company�s business reality in order to ensure feasibility and pro�tability. However,

accounting information can only play a part if the strategic framework captures the

central elements of the company�s business reality. Indeed, the case of EUR illus-

trates that a strategic focus which is too concerned with technological feasibility

without considering pro�tability can distract the company from its business reality

and lead to �nancial di¢ culties. In relation to complex decision problems, it is thus

a joint task of strategy and management accounting to constantly look for ways to

improve the framework by critically searching for signi�cant blind spots.

The results have practical implications since the approaches can serve as inspi-

ration for practitioners in other companies. Speci�cally, the two cases show how

outsourcing decision-making can be structured for companies at di¤erent stages of

development. Relatively speaking, EUR is at an early evolutionary stage and faces

a radical clean-up of company activities in order to locate the core competences and

112

reestablish pro�tability in the product portfolio. To meet this objective, the analyt-

ical epistemology seems to be the appropriate tool for information production and

use in relation to outsourcing decision-making. DEN, on the other hand, is at a later

stage of development and has already established its core competences. Instead, the

company proactively and deliberately creates innovations based on its core compe-

tences to maintain feasibility as well as pro�tability. In handling this incremental

clean-up, the actor-based epistemology seems to be the preferred alternative for

information production and use in relation to outsourcing decision-making.

In this way, the two approaches represent di¤erent extensions to the manage-

ment accounting and strategy toolkit since each of them serves a purpose under the

right circumstances. However, while the analytical epistemology might function as

a "quick short-term �x" to a situation gone wrong, we argue that it can never stand

alone. The reason is that ceteris paribus the recon�guration of an existing supply

chain involves: �rst, high risks caused by such factors as organizational inertia, time

pressure, or the lack of experience with similar situations; and second, unnecessary

costs as a result of investments in property, plant, and equipment that cannot be

fully recouped (sunk cost). Instead, the company should use the momentum from

the fundamental clean-up and fortify its position in relation to future innovations by

establishing a proactive approach to outsourcing decision-making. The argument is

that the strength of the actor-based epistemology increases with the decision-makers�

experience, and so it takes time to build up the necessary intuition as well as col-

laboration among actors. Yet, if established in due course, the learning mechanism

embedded so neatly into the proactive outsourcing decision-making process might

indeed help reduce the risks and costs of identifying appropriate supply chain con-

�gurations.

Recognizing the value of this, the article adds to the management accounting liter-

ature on decision theory which at the general level can be divided into two oppos-

ing camps. Into the �rst camp falls the literature rooted in micro economics (e.g.

statistical decision theory and n-person game theory) which is characterized by a

decision-making process where a choice (decision) is based on an explicit, compre-

hensive understanding and anticipation of the (expected) consequences of following

each alternative (see e.g. Keen and Morton 1978; Demski 1980; Sprague and Carl-

son 1982). Accordingly, the purpose of accounting information is to reduce the risk

a¤ecting the choice of a given alternative.

This view has been heavily criticized by the literature in the other camp partly

rooted in organizational psychology. Speci�cally, proponents from this wing argue

113

that it is possible to make intelligent choices (decisions) without an explicit com-

prehension of the full consequences �for instance, by following intuitions and the

advice of others (externalization) (see e.g. March 1987; Dane and Pratt 2007; Hall

2010). As a result, the value of accounting information is not only related to re-

ducing the risk a¤ecting the choice of a given alternative. Indeed, it also improves

the decision-maker�s understanding of his work environment and thus builds up his

intuition for future decision-making (Hall 2010).

Consequently, the crux of the disagreement between the two camps is whether in-

formation production and decision-making are best studied as a mechanical process

where (mathematical) systems can predict the optimal reaction pattern, or as an or-

ganic process where the human mind in an intelligent yet unsystematic way reaches

a decision. However, without denying the signi�cance of this academic discussion,

the analysis in this article clearly suggests that siding with one of the camps in

relation to decision-making in practice is rather pointless. Instead, we need to un-

derstand under which circumstances each perspective is most appropriately applied

in spite of its shortcomings. Some decision problems might call for an analytical

epistemology in which the system exercises authority over the human mind since for

various reasons the human mind cannot be fully trusted. Other decision problems

might bene�t from an actor-based epistemology by unleashing the full power of the

human mind while toning down the role of the system in order to help comprehend

and deal with the complexity of the decision problem. Indeed, this is in keeping

with the very heart of pragmatic constructivism: it may be less important that the

assumptions of the strategic framework and accounting methods used are a perfect

�t of the organizational complexity, but it is crucial that they capture the essentials

and help managers understand and analyze what is feasible as well as pro�table. In

practice, perfection is not the main objective; it is simply about making it work.

6 Further research

To develop the strategy and management accounting toolkit even further, several

issues need to be examined in greater detail. First, our analysis hints that for

complex decision problems like outsourcing decision-making, it might be useful to

consider the role of management accounting in a broader perspective and as an

interdisciplinary approach encompassing techniques as well as knowledge from both

accounting, marketing, R&D, and supply chain management in order to provide

su¢ ciently organic information for decision-making. However, so far the literature in

management accounting has been somewhat hampered by accounting conservatism

114

often concerned with traditional cost issues without purposefully addressing the

di¢ cult strategic issues. Accordingly, there is an immense scope for picking up

where Shank and Govindarajan (1989; 1992) left o¤ to study further studying the

role of management accounting in strategic decision-making.

Second, this article has not dealt with how DEN in the �rst place has been able

to successfully implement the actor-based epistemology for the production and use

of accounting information. Consequently, this is an area for additional investigation

since the transition phase, in which the individual actors acquire the necessary

comprehensive understanding of their work environment (intuition) and learn to

collaborate with each other, might be cumbersome and risky before the bene�ts can

be fully reaped.

Finally, this study merely represents a snapshot in time of the outsourcing

decision-making process. In order to improve our understanding of the process,

supplementary longitudinal studies need to address how the strategic framework

and the accounting methods applied coevolve and shape each other over time.

7 Appendix

This appendix presents the guide used for interviewing the case companies, DEN

and EUR.

Interviewee�s background

� What is your background?

�Education and work experience?

�How long have you been with the company?

�How long have you worked with outsourcing?

Company background

� How is the company organized?

�What is outsourced?

�What outsourcing destinations are used?

� Does the company have an underlying outsourcing philosophy?

115

� Is outsourcing mainly seen as a strategic tool (access to new competences, closer

alignment with customer needs or close proximity to customer facilities)?

� Is outsourcing primarily seen as a tool to cut costs and improve the logistics?

� Does the company have a long track record of working with outsourcing?

Outsourcing decision-making process

� What does the decision-making process look like?

�What is the time frame?

�How detailed is the analysis - is speed a key issue (trade-o¤)?

�How much weight is the decision-making process given compared to the subsequent

implementation process?

�Do you have a model that frames the way you consider and analyze the outsourcing

situation?

� Is there a speci�c structure for the decision-making process (steps)?

� Who is taking part in the various steps �is it interdisciplinary?

� Suppliers, management, accounting, engineers, etc.

� Strategy model: How do you view the company and its surroundings in relation to the

outsourcing situation?

�Do you have an underlying framework for analyzing the company and the industry?

� Porter: Value chain analysis and industry analysis?

� RBV: Core competences and resources?

�What is driving the process?

� Is it an internal perspective where e¢ ciency determines what should be out-

sourced?

� Is it an external perspective where the customer (market) determines what should

be outsourced?

�What is your reason for deciding to use this exact approach to outsourcing?

� Are there any issues that you do not catch with this approach (blind spots)?

� How do you address issues that you are aware of having left out by approach-ing the problem in this way?

116

� How do you assess the realism of the model?

� Is it based on the company�s own experience from previous outsourcing en-

gagements? Or is it based on what other companies have used with success?

� Have you made wrong outsourcing decisions earlier on where you subse-quently found out that you had made miscalculations?

� Have you incorporated this experience into the analyses that you are con-ducting now (learning)?

� Accounting model: How do you calculate the �nancial implications?

�What dimensions do you analyze, and how are they made operational �what models

are you using?

�What type of data enters into the model �how are these data collected?

� Customer value �revenue:

� Do you quantify customer value?� What data do you rely on?� How do you collect these data?

� Cost of making:

� Production costs� Opportunity costs� How do you collect these data?

� Cost of buying:

� Quotations� Transaction costs� Risk (loss of competence or control of the supply chain)� How do you collect these data?

�How do you weight the �nancial assessment compared to the strategic considerations

not incorporated into the �nancial evaluation?

� Are you conservative when estimating costs?

� Is there anything in relation to the decision-making process � besides the issues already

discussed �that you feel is important to address?

Acknowledgements

The authors would like to thank Professor Lennart Nørreklit, Aalborg University,

for discussions on key issues. Any mistakes are entirely the authors�.

117

References

Anderson, P. A. (1983, June). Decision making by objection and the Cuban missilecrisis. Administrative Science Quarterly 28 (2), 201�222.

Arbnor, I. and B. Bjerke (1997). Methodology for Creating Business Knowledge(2nd ed.). Thousand Oaks, CA: Sage Publications.

Barthélemy, J. (2001). The hidden cost of IT outsourcing.MIT Sloan ManagementReview 42 (3), 60�69.

Bettis, R. A., S. P. Bradley, and G. Hamel (1992). Outsourcing and industrialdecline. Academy of Management Executive 6 (1), 7�22.

Bryce, D. J. and M. Useem (1998). The impact of corporate outsourcing on com-pany value. European Management Journal 16 (6), 635�643.

Carson, S. J. (2007). When to give up control of outsourced new product devel-opment. Journal of Marketing 71 (1), 49�66.

Dahan, E. and J. R. Hauser (2002). Handbook of Marketing, Chapter 8: Productdevelopment: managing a dispersed process, pp. 179�222. Thousand Oaks,CA: Sage Publications.

Dane, E. and M. G. Pratt (2007, January). Exploring intuition and its role inmanagerial decision making. Academy of Management. The Academy of Man-agement Review 32 (1), 33�54.

David, M. (2009, July). The correspondence theory of truth. Stanford Encyclo-pedia of Philosophy. http://plato.stanford.edu/entries/truth-correspondence.

Demski, J. (1980). Information analysis. MA: Addison-Wesley.

Eisenhardt, K. M. and M. J. Zbaracki (1992). Strategic decision making. Strate-gic Management Journal 13 (Special Issue: Fundamental Themes in StrategyProcess Research), 17�37.

Fill, C. and E. Visser (2000). The outsourcing dilemma: a composite approach tothe make or buy decision. Management Decision 38 (1), 43�50.

Gove, K. (2009). Outsourcing in Denmark - across four industries: IT, �nance,manufacturing, and pharma/biotech. Technical report, Trellis.

Grant, R. M. (2005). Contemporary Strategy Analysis (5th ed.). Blackwell Pub-lishing Ltd.

Hall, M. (2010). Accounting information and managerial work. Accounting, Or-ganizations and Society 35 (3), 301�315.

Hill, C. W. L. and G. R. Jones (2001). Strategic Management: An IntegratedApproach (5th ed.). New York: Houghton Mi­ in Company.

Ho-Lanng, K. F. (2009, April). Sæt outsourcing i angrebet. Computer World.http://www.computerworld.dk/art/51126/saet-outsourcing-i-angrebet.

Hopwood, A. (1974). Accounting and Human Behavior. Prentice Hall.

Horngren, C. T., G. L. Sundem, and W. O. Stratton (2005). Introduction toManagement Accounting (13th ed.). New Jersey: Pearson Prentice Hall.

118

Horton, J., R. Macve, and G. Struyven (2004). The Real Life Guide to AccountingResearch: A Behind-the-Scenes View of Using Qualitative Research Methods,Chapter 20: Qualitative Research: Experiences in using semi-structured in-terviews, pp. 339�359. Elsevier Science.

Hätonën, J. and T. Eriksson (2009). 30+ years of research and practice of out-sourcing �exploring the past and anticipating the future. Journal of Interna-tional Management 15 (2), 142�155.

Keen, P. G. W. and M. S. S. Morton (1978). Decision Support Systems: AnOrganizational Perspective. MA: Addison-Wesley.

Lamminmaki, D. (2008). Accounting and the management of outsourcing: Anempirical study in the hotel industry.Management Accounting Research 19 (2),163�181.

Lang�eld-Smith, K. and D. Smith (2003). Management control systems and trustin outsourcing relationships. Management Accounting Research 14 (3), 281�307.

Lewin, A. Y. and C. Peeters (2006). O¤shoring work: Business hype or the onsetof fundamental transformation? Long Range Planning 39 (3), 221�239.

March, J. G. (1987). Ambiguity and accounting: The elusive link between infor-mation and decision making. Accounting, Organizations and Society 12 (2),153�168.

Markie, P. (2008, August). Rationalism vs. empiricism. Stanford Encyclopedia ofPhilosophy. http://plato.stanford.edu/entries/rationalism-empiricism.

Matthews, E. (1996). Twentieth-Century French Philosophy. Oxford: Oxford Uni-versity Press.

McIvor, R. (2005). The Outsourcing Process: Strategies for Evaluation and Man-agement. Cambridge: Cambridge University Press.

Mintzberg, H. (1978, May). Patterns in strategy formation. Management Sci-ence 24 (9), 934�948.

Mintzberg, H. and J. A. Waters (1985, July - September). Of strategies, deliberateand emergent. Strategic Management Journal 6 (3), 257�272.

Monahan, S., P. V. D. Bossche, and F. Tamayo (2010, April). Make vs. buyrevisited: Reassessing your company�s manufacturing strategy. A. T. Kearney.

Nicholson, B., J. Jones, and S. Espenlaub (2006). Transaction costs and controlof outsourced accounting: Case evidence from India. Management AccountingResearch 17 (3), 238�258.

Nielsen, L. B. (2010). A retrospect of management accounting�s role in strategicoutsourcing decision-making. Working paper.

Nørreklit, H., L. Nørreklit, and F. Mitchell (2010). Towards a paradigmatic foun-dation of accounting practice. Accounting, Auditing & Accountability Journal .Forthcoming.

119

Nørreklit, L. (1987). Formal structures in social logic. Aalborg: Aalborg Univer-sity Press.

Nørreklit, L. (2011). An Actor�s Approach to Management, Chapter 1: Actorsand Reality. DJOEF. Forthcoming.

Nørreklit, L., H. Nørreklit, and F. Mitchell (2007). Business Performance Mea-surement, Chapter 9: Conditions for Validity in Accounting Performance Mea-surement. Cambridge: Cambridge University Press.

Nutt, P. C. (1976, April). Models for decision making in organizations and somecontextual variables which stipulate optimal use. Academy of ManagementReview 1 (2), 84�98.

Oxford, R. L. (1997). Constructivism: Shape-shifting, substance, and teachereducation applications. Peabody Journal of Education 72 (1), 35�66.

Peters, T. J. and R. H. Waterman (1982). In Search of Excellence: Lessons fromAmerica�s Best-Run Companies. New York: Harper & Row.

Porter, M. E. (1980). Competitive Strategy: Techniques for Analyzing Industriesand Competitors. New York: Free Press.

Porter, M. E. (1985). Competitive Advantage: Creating and Sustaining SuperiorPerformance. New York: Free Press.

Prahalad, C. K. and G. Hamel (1990, May - June). The core competence of thecorporation. Harvard Business Review 68 (3), 79�92.

Quinn, J. B. (2000). Outsourcing innovation: the new engine of growth. MITSloan Management Review 41 (4), 13�28.

Roslender, R. and S. J. Hart (2003). In search of strategic management account-ing: theoretical and �eld study perspectives. Management Accounting Re-search 14 (3), 255�279.

Russell, B. (1918). Logic and Knowledge: Essays 1901-1950, Chapter 3: The Phi-losophy of Logical Atomism, pp. 177�281. London: George Allen and Unwin,1956.

Ryan, B., R. W. Scapens, and M. Theobold (2002). Research Method and Method-ology in Finance and Accounting. Thomson Learning.

Sartorius, K. and J. Kirsten (2005). The boundaries of the �rm: why dosugar producers outsource sugarcane production.Management Accounting Re-search 16 (1), 81�99.

Seal, W., J. Cullen, A. Dunlop, T. Berry, and M. Ahmed (1999). Enacting aEuropean supply chain: a case study on the role of management accounting.Management Accounting Research 10 (3), 303�322.

Shank, J. K. and V. Govindarajan (1989). Strategic Cost Analysis �The Evolutionfrom Managerial to Strategic Accounting, Chapter 3: Concepts in Value ChainAnalysis: The "Famous" Crown Cork and Seal Company Case, pp. 39�59.Irwin.

120

Shank, J. K. and V. Govindarajan (1992). Strategic cost management: The valuechain perspective. Journal of Management Accounting Research 4, 179�197.

Sprague, R. H. and E. Carlson (1982). Building E¤ective Decision Support Sys-tems. Prentice Hall.

Steno, C. (2009, April). Den nye outsourcingsbølge. Business.dk.http://www.business.dk/bny/den-nye-outsourcingsboelge.

Thrane, S. and K. Hald (2006). The emergence of boundaries and accounting insupply �elds: The dynamics of integration and fragmentation. ManagementAccounting Research 17 (3), 288�314.

Tomkins, C. (2001). Interdependencies, trust and information in relationships,alliances and networks. Accounting, Organizations and Society 26 (2), 161�191.

van der Meer-Kooistra, J. and E. G. J. Vosselman (2000). Management con-trol of inter�rm transactional relationships: the case of industrial renovationand maintenance - design and methods. Accounting, Organizations and Soci-ety 25 (1), 51�77.

Weiss, K. (2010, March). Alt kan outsources. Erhvervs Bladet.http://www.erhvervsbladet.dk/vaekstmaskinen/alt-kan-outsources.

Widener, S. K. and F. H. Selto (1999). Management control systems and bound-aries of the �rm: Why do �rms outsource internal auditing activities? Journalof Management Accounting Research 11, 45�73.

Wittgenstein, L. (1921). Tractatus Logico-Philosophicus. London: Routledge,1961.

Yin, R. K. (2002). Case Study Research: Design and Methods, Volume 5 of AppliedSocial Research Methods Series. Thousand Oaks, CA: Sage Publications.

121

Prod

uct

Supp

ly c

hain

con

figur

atio

n

Third

-par

ty

supp

lier

Mak

e

Third

-par

ty

supp

lier

Buy

Sele

cted

supp

ly c

hain

co

nfig

urat

ion

Alte

rnat

ive

supp

ly

chai

n co

nfig

urat

ion

and

Act

iviti

es/p

roce

sses

Figure1:Theoutsourcingdecisionconcernsthecompletecon�gurationofthesupplychain.

122

Epis

tem

olog

y O

ntol

ogy

Para

digm

A

ccou

ntin

g te

chni

ques

A

ccou

ntin

g m

etho

ds

Out

sour

cing

pro

blem

(s

uppl

y ch

ain

conf

igur

atio

n)

Out

sour

cing

dec

isio

n

1 2.

1

2.2

3

4

6

Stra

tegi

c fr

amew

ork

(met

hodi

c) 5

Production of acc. information

Met

hodi

cal

proc

edur

e

Figure2:Frameworkforanalysisofoutsourcingdecision-making.

123

 

 

Unfeasible 

Unprofitable butfeasible

Profitable and feasible 

Strategic scope

Unfeasible 

Unfeasible  Unfeasible 

Figure 3: Intuitive illustration of the interdependency between the company�sstrategic framework and business reality in the construction of supply chain al-

ternatives.

124

Val

ue p

ropo

sitio

n

Idea

gen

erat

ion

( T

PPs)

Gat

e 0

Pick

TPP

s (1

0-20

TPP

s)

Gat

e 1

Sele

ctio

n of

pr

oduc

t co

ncep

ts

(1-3

pro

ject

s)

Stag

e 0

Con

cept

dev

elop

men

t St

age

1 C

once

pt m

atur

ing St

age

2 Pr

oduc

t ver

ifica

tion

Inta

ke

for A

IM

Stag

e 3

+ 4

Prod

uct r

ealiz

atio

n an

d la

unch

M

arke

t C

ompe

titor

s C

usto

mer

s Su

pplie

rs

Com

pany

DE

N

Cor

e co

mpe

tenc

es

Proa

ctiv

e ad

just

men

t

Gat

e 4

+ 5

Prod

uct l

aunc

h an

d bu

sine

ss

eval

uatio

n

Gat

e 3

Prod

uct a

nd

man

ufac

turin

g pr

oces

ses a

re

f roz

en

Gat

e 2

Prod

uct

conc

epts

and

de

sign

spec

s are

fr

ozen

Stra

tegi

c fr

amew

ork:

Inno

vatio

n V

alue

Str

eam

(IV

S)

Out

sour

cing

de

cisi

on-m

akin

g

proc

ess

Mat

ch

Figure4:ThestrategicframeworkoftheDanishcompanyDEN.

125

Value creating processes

Evaluate business areas (first phase)

Core business areas (second phase)

Evaluate customers Unimportant Suppliers

Important

Market view Value propositions

In-house view Scrutinize processes and

activities

Evaluate activities

Non-value creating processes Evaluate processes

Activities providing differentiation

Generic activities

Core and no outsourcing

Make-or-buy analysis

Management view Outsourcing decision

Make-or-buy analysis

Market Competitors Customers Suppliers

Non-core business areas

Divestment or partnership

Mismatch

Reaction: Fundamental adjustment (clean-up)

Company EUR Competences

Strategic framework

Outsourcing decision-m

aking process Elim

ination process

Repeated for each product

Figure 5: The strategic framework of the European company EUR.

126

Chapter 3

The impact of outsourcing on investments in

firm-specific human capital under different contract

regimes

The impact of outsourcing on investments in

�rm-speci�c human capital under di¤erent

contract regimes

Lars Bråd Nielsen�

July 27th, 2010

Abstract

Investments in �rm-speci�c human capital have come to play a central rolein the value creation for most companies as job tasks have become ever morecomplex and demanding to carry out. Today successful performance of essen-tial job tasks often necessitates highly specialized knowledge and skills, thusrequiring continuous updating of employee competences. This paper developsa two-period agency model to show how the threat of layo¤ (outsourcing of jobtasks to a third-party supplier) can help a company trigger or ease employeeinvestments in �rm-speci�c human capital by creating explicit career concernsfor the individual employee. Results are provided under long-term as well asshort-term contracting regimes. In particular, the paper has relevance forcompanies operating under short-term contracting where investments in �rm-speci�c human capital might be pro�table yet very di¢ cult or even impossibleto induce.

�Aarhus School of Business, Aarhus University, Fuglesangs Allé 4, DK-8210 Aarhus V., Den-mark. E-mail: labn@asb.dk.

1 Introduction

The resource-based view suggests that a key element in ensuring long-term company

pro�tability is an ongoing focus on what resources to use and how to combine these

most e¤ectively to exploit their full potential (Wernerfelt 1984; Barney 1991). In

many companies the most important resources are human by nature, which naturally

relates �rm prosperity to the knowledge and skills held by the employees (Becker

1964; Spender and Grant 1996). Sometimes these capabilities can be combined and

codi�ed into intellectual assets like inventions and technologies that can be claimed

as property by the company. At other times it is impossible, or very costly and

di¢ cult, to materialize human capabilities, for which reason their value-creating po-

tential remains with the individual worker and can only be used at his discretion

(Edvinsson and Sullivan 1996; Kogut and Zander 1993; Nonaka and Takeuchi 1994).

Still, however, such non-codi�ed knowledge and skills can generate value through

reinforcement of essential processes within the company.

Whether knowledge and skills can be codi�ed or not, learning is essential because it

is a precondition for value creation that the employees acquire the necessary know-

how to master and develop the company�s intellectual assets or reinforce company

processes. Some required capabilities are non-idiosyncratic and widely available in

the labor market as they can be acquired by any worker prior to joining a particular

�rm (e.g. school training). Yet, if competition is �erce, competitiveness and thus

pro�tability can best be ensured by focusing company activities on superior resources

that are unique and di¢ cult for competitors to imitate (Prahalad and Hamel 1990;

Barney 1991). That is, only through such resources is it possible to obtain a low

cost position in the market (Ricardian rents) or lower competition by di¤erentiating

company products from the rest of the market (monopoly rents) (Peteraf 1993).

Consequently, �rm-speci�c learning (on-the-job training) needs to take place inside

the company after the worker has been hired. For example, the pro�t potential of

a manufacturing �rm producing a standard good following a low-cost strategy may

depend on the employees�inclination to acquire speci�c know-how about the com-

pany�s production processes that can help perfect critical manufacturing routines,

thus improving productivity. Likewise, the monopoly rent of a high-tech company

having di¤erentiated its business thanks to a patent on one of its intellectual as-

sets may partially depend on certain employees acquiring the knowledge necessary

to master this technology, a knowledge acquisition which by de�nition is company

speci�c due to the technology being patented.

129

More formally, the term human capital can be used to capture the idea that much

skill and knowledge required to do a job can only be acquired if some "investment"

is made in time and resources (Becker 1964). In this respect, there are generally

two types of human capital; �rm-speci�c and regular human capital. Regular (non-

�rm-speci�c) human capital is transferable between jobs, which implies that the

inherent value is independent of a particular employment relationship. Contrary to

this, �rm-speci�c human capital is characterized by having a much higher value in

a given employment relationship than it does in other potential relationships, and

so, just like specialized physical investments, it may enhance productivity beyond

what can be realized using regular human capital.

Hence, where pro�tability can be improved by employees investing time and e¤ort

in tailoring their capabilities to speci�c company operations, companies have a clear

interest in such investments being undertaken. The di¢ culty, however, is that the

inherent asset speci�city complicates the employment relationship by introducing a

fundamental hold-up problem. On the one side, if the �rm has paid for the speci�c

training of a worker who quits to take another job, its capital expenditure will be

partly wasted and no further return can be collected (Becker 1964). However, even

if the cost of education is negligible, as will be assumed in this paper, providing the

facilities will not necessarily do the job since learning can only take place at the

worker�s own discretion. This represents the other side of the hold-up problem and

will in this paper be described as a combination of two e¤ects. First, the total e¤ort

that an employee can exert in a given period might be constrained, which can cause

reluctance to invest in �rm-speci�c human capital since learning is counterproduc-

tive in the short run and reduces the e¤ort available to spend on other short-term

productive activities. Second, the �rm speci�city makes the investment risky as

the acquired abilities only have limited value in alternative use, thereby making the

worker vulnerable to layo¤ or salary cuts before fully recouping his investment.

Since learning cannot be enforced by the company, it can merely try to encour-

age learning implicitly through carefully designed incentive mechanisms that make

it attractive for the employee to acquire �rm-speci�c capabilities. One way to in-

duce such long-term actions is by providing job security for the worker, which can be

attained through long-term contracts where the company commits to a given com-

pensation in future periods at the time of contract initiation (Milgrom and Roberts

1992, p. 363). By doing so, the employee no longer needs to worry about dismissal or

forced salary cuts as a consequence of being locked in when undertaking investments

130

in �rm-speci�c human capital. In particular, if the salary in future periods partly

depends on how productive the worker is, and productivity can be in�uenced posi-

tively by improving skills and knowledge, long-term engagements naturally motivate

the employee to upgrade his competence level.

However, not all companies can operate under long-term contracts but instead

have to rely on short-term engagements. For example, it is not di¢ cult to imagine

a company that for strategic reasons needs to keep the cost structure �exible and

thus contracts with a part of the employees on a short-term basis in order to make

periodical salary adjustments possible. Likewise, a company might be interested

in minimizing employee shirking through the disciplining e¤ect of ongoing perfor-

mance evaluation that, contrary to long-term engagements, is naturally embedded

in short-term contracts (Anderhub et al. 2003). Still, even under such conditions it

may be in the company�s best interest that �rm-speci�c learning takes place, since

company performance may then be enhanced due to increased productivity ex post

the competence investment. The problem is, however, that short-term contracting

does not eliminate the risk related to the worker�s investment in �rm-speci�c human

capital as it makes it non-credible for the company to commit to compensations in

future periods. Thus, if employee competences can be observed, a rational company

will try to exploit the worker�s in�exibility ex post the investment and enforce a

salary cut that leaves the employee with little or no expected value of his compe-

tence improvement. In total, this might eliminate the worker�s incentive to invest

in �rm-speci�c capabilities when taking into consideration the negative e¤ect that

learning has on short-run productivity.

One way to incentivize the worker to take given actions even under short-term

contracting can be found in the literature on labor economics where the disciplining

e¤ect of unemployment has been studied (Calvo 1979; Shapiro and Stiglitz 1984).

The basic idea in Shapiro and Stiglitz (1984) is that if no unemployment exists and

all workers receive the market wage, the worst that can happen to a worker not

exhibiting the right behavior (e.g. shirking) is getting �red. However, since he can

immediately be rehired at the market wage, he pays no penalty for his misdemeanor

and has no incentive to behave well. With unemployment the situation changes;

even if all �rms pay the same wages, a worker has an incentive to behave correctly

because if he is laid o¤, no other job is immediately available. Therefore, the equi-

librium unemployment rate must be large enough to make it worthwhile for workers

to undertake actions that are in the company�s best interest and avoid dismissal.

Another way to provide the worker with incentive to undertake certain actions,

131

which is of particular interest to the present paper, can be found in the literature on

career concerns, also known as implicit market-driven incentives (e.g. Gibbons and

Murphy 1992; Dewatripont et al. 1999a, 1999b; Holmström 1999). Typically, career

concerns arise when the (internal or external) labor market settles future compensa-

tion to re�ect employee ability but neither the worker�s skill level nor his productive

e¤ort is observable. Yet, output can be observed and so the market uses this to

update its belief about the worker�s ability in order to apply this revised expecta-

tion to determine future wages. As a result, the worker wants to take advantage of

e¤ort being unobservable, in an attempt to increase output and thus in�uence the

market�s belief about talent.1

1.1 Contribution

Motivated by the importance of �rm-speci�c human capital in many employment

relationships and intrigued by the use of unemployment and career concerns (at-

tractiveness) as incentive mechanisms, this paper sets out to formally study how

outsourcing can be used as a possible device to overcome incentive problems of

investing in competence improvement under both long-term and short-term con-

tracting. Assuming that capabilities are observable and no rent yielding jobs are

available immediately after layo¤, the simple idea is that the mere threat of out-

sourcing makes the employee concerned with unemployment, which induces him to

invest in �rm-speci�c human capital in order to stay attractive to the company.2

Under long-term contracting (full commitment), it is shown that introducing

an outsourcing alternative is not necessary to induce a certain investment in �rm-

speci�c human capital since the employee�s incentive can be regulated (strengthened

or weakened) through commitment to a given compensation ex post the investment.

Still, it is demonstrated that it might be alluring for the company to bring about

outsourcing �exibility as it comprises two potentially value-creating e¤ects; a real

option e¤ect and a behavioral e¤ect. The real option e¤ect increases the expected

1For models along these lines, see Holmström (1999), which is a re-publication of a paper from1988, and Gibbons and Murphy (1992).

2This partly resembles the idea in Shapiro and Stiglitz (1984) of using unemployment as adisciplining device, although the concern of the present paper is not with analyzing the dynamicsor equilibrium e¤ects in the labor market. Furthermore, the incentive to signal attractiveness bearscertain similarities to the typical career concern model. However, while the underlying mechanismmay be the same, the proposed setup di¤ers from the standard modeling of career concerns in twoimportant ways. First, the worker�s ability is not a constant but can indeed be improved by hisown behavior if he spends time on training. Second, the worker�s competence level will be fullyrevealed to the company, and so what drives future wages is not the observed output but in factthe observed ability, since this partially determines whether the worker will be laid o¤ or allowedto continue production.

132

pro�t since the access to a third-party supplier reduces the dependence on in-house

learning if the company can choose the severance fee paid to the worker at dismissal

without severe constraints. The behavioral e¤ect reduces the incentive cost of mak-

ing the employee undertake a certain investment in �rm-speci�c human capital as

the mere introduction of outsourcing �exibility naturally incentivizes him to invest

in order to signal attractiveness and avoid layo¤.

Under short-term contracting, it is shown that the absence of an outsourcing

alternative leaves the employee with no incentive to invest in �rm-speci�c human

capital. Speci�cally, since the company cannot credibly commit to any future con-

tracts, it will always exploit the �rm-speci�city of the worker�s investment and en-

force a salary cut ex post the competence improvement reducing the value of the

�rm-speci�c investment to zero. As such, this can be harmful to the company if

learning is productive. It is thus demonstrated how the company can bene�t from

introducing outsourcing �exibility into the setting to create explicit career concerns

for the employee and make him undertake long-term actions to signal attractiveness

and maintain production in-house. That is, although future compensation will be

adjusted by the company to o¤set the e¤ect of �rm-speci�c investments under short-

term contracting, there are still rents to be collected by the worker if not dismissed.

Consequently, compared to full commitment the introduction of an outsourcing al-

ternative is absolutely necessary to encourage a positive investment in �rm-speci�c

human capital under short-term contracting.

1.2 Outline

The rest of the paper proceeds as follows: section 2 outlines the model and es-

tablishes �rst-best and second-best under long-term contracting in the respective

cases of no outsourcing and outsourcing. Next, section 3 studies the model under

short-term contracting. In relation to this, a number of example-based re�ections

are presented on the value of outsourcing �exibility to both the company and the

employee. Finally, section 4 concludes the analysis under both contracting regimes

while suggestions for further research are provided in section 5.

2 Full commitment

2.1 The basic model without outsourcing �exibility

In the following, we will consider a company setting lasting for two periods and

consisting of a risk neutral principal (the manager or company) and a risk neutral

133

agent with limited liability (the employee). For the purpose of illustration, we will

let the agent be the company�s sole productive resource and assume that he possesses

two basic capabilities; a physical productive ability and an intellectual capacity to

learn. As such, the physical productive ability makes the employee able to produce

the company�s output while the intellectual capacity provides the employee with

an opportunity to acquire certain �rm-speci�c skills regarding company processes

and best practice that can complement and amplify the e¤ect of using the physical

productive ability.

Under full commitment, the principal at t = 0 contracts with the agent for pro-

duction of the stochastic output ~x1 2 fxL; xHg in period 1 and ~x2 2 fxL; xHg inperiod 2 where �x = xH � xL > 0.3

The output in period 1 is a¤ected by the level and composition of the agent�s

e¤ort. As such, the agent�s e¤ort can take on two levels, e1 2 feL; eHg witheH > eL � 0, which can be further divided into physical productive e¤ort, a � 0,

and skill improvement (investment in �rm-speci�c human capital), h � 0. Mathe-matically, we will assume that the probability of realizing the high output (upstate),

xH , in period 1 is given by '1 (xH j e1; h) = (e1 � h)1=2 while the probability of re-

alizing the low output (downstate) is '1 (xLj e1; h) = 1 � (e1 � h)1=2. Intuitively,this implies that the more e¤ort the agent exerts, the greater is the chance of re-

alizing the high outcome. However, should the agent choose to spend time on skill

improvement, this will diminish the probability of realizing the high outcome, since

investing in �rm-speci�c human capital is assumed non-productive during period

1. Moreover, we will assume that putting forth e¤ort does not come without cost,

for which reason the agent su¤ers a personal cost of �1 = 0 or �1 = �H > 0 when

exerting eL or eH , respectively. Finally, at the end of period 1 the output, x1, is

realized and since this output is the only variable in period 1 that can be contracted

on, this determines the agent�s payment, c1. As such, c1 can never be negative due

to the agent�s limited liability.

All told, the agent�s expected utility in period 1 conditional on e1 and h takes

the following form

E [U1j e1; h] = '1 (xH j e1; h) c1H + (1� '1 (xLj e1; h)) c1L � �1= c1L + (e1 � h)1=2�c1 � �1: (1)

3Notice that all through this article, � represents change in the particular variable of interest.

134

Likewise, if the revenue from production is assumed to be net of production cost,

the principal only explicitly incurs the compensation cost to the agent and faces the

following expected pro�t conditional on e1 and h.

E [�1j e1; h] = '1 (xH j e1; h) (xH � c1H) + '1 (xLj e1; h) (xL � c1L)= (e1 � h)1=2 (xH � c1H) +

�1� (e1 � h)1=2

�(xL � c1L) : (2)

To keep the model tractable in period 2, we will continue the simplicity from period

1 and assume that the agent�s e¤ort can take on the same two levels, e2 2 feL; eHg.This time, however, further skill improvement is pointless for which reason the agent

only directs his e¤ort towards physical production.

Additionally, the agent bene�ts from his investment in �rm-speci�c human cap-

ital; the probability of receiving the high output in period 2 is '2 (xH j e2; h) =e1=22 (1 + �h) while the probability of receiving the low output is '2 (xLj e2; h) =1 � e1=22 (1 + �h) for any given h � 0. With regard to the speci�c structure of theprobability function, the parameter � determines how productive �rm-speci�c hu-

man capital is in period 2 and thus howmuch investing in �rm-speci�c human capital

increases the likelihood of realizing the upstate. Moreover, for h = 0 the probabili-

ties in period 1 and 2 are identical; if no investment is made in �rm-speci�c human

capital and the same amount of e¤ort is put forth in each period, the probability of

realizing the upstate is the same in both periods.

In continuation of this, we will make the special assumption that the principal

can only establish a compensation programme in period 2 that is independent of the

realized period 1 compensation. Again, we will assume that the agent su¤ers a per-

sonal cost in period 2 of �2 = 0 or �2 = �H > 0 when exerting eL or eH , respectively.

In total, this provides the agent with the following expected utility, E�UNO2

�� e2; h�in period 2 when no outsourcing �exibility exists (NO)

E�UNO2

�� e2; h� = '2 (xH j e2; h) c2H + '2 (xLj e2; h) c2L � �2= c2L + e

1=22 (1 + �h)�c2 � �2; (3)

while the principal�s pro�t under full contract commitment and no outsourcing in

period 2 is given by

E��NO2

�� e2; h� = '2 (xH j e2; h) (xH � c2H) + '2 (xLj e2; h) (xL � c2L)

= e1=22 (1 + �h) (xH � c2H) +

�1� e1=22 (1 + �h)

�(xL � c2L) :(4)

135

Figure 1 below outlines the sequence of events in the model.

<Insert �gure 1 here>

2.2 Analysis under full commitment and no outsourcing

�exibility

Initially, we will consider the benchmark situation where outsourcing is not possible

and long-term contracts can be established. It is assumed that �x is su¢ ciently

large to provide the principal with incentive to induce the high e¤ort level in each

period, i.e. (e1; e2) = (eH ; eH). First, the optimal level of �rm-speci�c human capital

is derived in �rst-best. Subsequently, we will determine how the principal selects

the compensation structure in each period to maximize his pro�t in second-best.

2.2.1 First-best

In �rst-best (FB) no incentive problem exists for which reason the principal at t = 0

maximizes his pro�t by: �rst, setting c1L = c1H = c2L = c2H = �H to satisfy the

agent�s participation constraint, E�UNO

�= E

�UNO1

�+ E

�UNO2

�= 0; and second,

choosing hNOFB properly. From (2) and (4) we get

maxhNOFB

E��NO

�= E [�1j eH ; h] + E

��NO2

�� eH ; h�= (eH � h)1=2 xH +

�1� (eH � h)1=2

�xL

+e1=2H (1 + �h)xH +

�1� e1=2H (1 + �h)

�xL:

The principal only wants the agent to invest in �rm-speci�c human capital (i.e.

take long-term action) if it increases the expected pro�t and so we initially evalu-

ate @E��NO

�=@h at h = 0 to determine what makes �rm-speci�c human capital

valuable@E��NO

�@h

�����h=0

= e�1=2H �x

�eH��

1

2

�> 0, eH� > 1=2:

That is, eH� > 1=2 represents the necessary condition to make learning favorable

to the principal. Conversely, this also implies that investing in �rm-speci�c human

capital can be a bad to the principal when eH� � 1=2.

Since @2E��NO

�=@h2 = �1=4 (eH � h)�3=2 < 0, the �rst-best problem,maxE

��NO

�

136

wrt. hNOFB , is concave with the following solution4

@E��NO

�@h

= �12(eH � h)�1=2�x+ e1=2H ��x = 0)

hNOFB = max

(0; eH �

�1

2�

�21

eH

): (5)

Thus, in �rst-best the principal induces the agent to improve his skill level, hNOFB > 0,

if this increases the total likelihood of a good outcome (eH� > 1=2). On the other

hand, if learning is unproductive (eH� � 1=2), the principal selects hNOFB = 0 and

induces the agent to only focus on exerting short-term productive e¤ort. In relation

to this, we also notice that the more productive �rm-speci�c human capital is (the

higher � is), the more the principal wants the agent to invest in skill improvement.

2.2.2 Second-best

Having determined the �rst-best investment in �rm-speci�c human capital, we turn

to the second-best situation (SB) where information asymmetry between the prin-

cipal and the agent exists. In second-best, the principal at t = 0 chooses the

compensation plan in period 1 and 2 to ensure that the agent�s participation con-

straint (PC) and incentive compatibility constraints (IC�s) are all satis�ed and lead

the agent to pick the levels of e¤ort and �rm-speci�c human capital that maximize

pro�t.

To study this optimization problem, we will initially establish how the com-

pensation programme in period 1 and 2 can be set to always satisfy the PC and

IC�s. The reason is that this simpli�es the notation in relation to the subsequent

derivations. Following this, focus will be kept on the agent to show how he can

be given incentive to invest in �rm-speci�c human capital. Moreover, it will be

derived how he settles his investment in �rm-speci�c human capital in second-best

(response function). Finally, the principal�s actual maximization problem is studied.

Consider �rst E�UNOHH

�where the agent picks (e1; e2) = (eH ; eH) and E

�UNOe1e2

�that

4The max f�g function indicates that h can never be negative.

137

in this case represents the expected utility for all other combinations of e1 and e25

E�UNOHH

�= E

�U1j eH ; hNOHH

�+ E

�UNO2

�� eH ; hNOHH� (6)

=�eH � hNOHH

�1=2�c1 + c1L + e

1=2H

�1 + �hNOHH

��c2 + c2L � 2�H

E�UNOe1e2

�= E

�U1j e1; hNOe1e2

�+ E

�UNO2

�� e2; hNOe1e2� (7)

=�e1 � hNOe1e2

�1=2�c1 + c1L + e

1=22

�1 + �hNOe1e2

��c2 + c2L � �2:

From (6) and (7), it is evident that the principal can always select c1L = c2L = 0

and pick c1H and c2H to satisfy E�UNOHH

�� E

�UNOe1e2

�and thus the IC�s, since the

agent cares only about the di¤erence in compensation in each period (the bonus).

Moreover, by doing so, the PC is automatically satis�ed as E�UNOHH

�� E

�UNOLL

�=�

e1 � hNOLL�1=2

c1H + e1=22

�1 + �hNOLL

�c2H > 0.

With this in mind, we will turn to investigate how the agent�s investment in �rm-

speci�c human capital changes for given levels of e¤ort and compensation, since

this a¤ects the principal�s actual choice of compensation and allows for comparison

with the �rst-best level of �rm-speci�c human capital. Hence, the agent�s response

function for h as a function of e1; e2; c1H and c2H will be derived. The results are

summarized in proposition 1 below.

Proposition 1 (Worker) (I) If the company can commit to the second-period con-tract, the worker can be given incentives to invest in �rm-speci�c human capital. (II)

Increasing (Decreasing) the bonus paid while skill-improvement takes place, c1H ,

weakens (strengthens) the incentive to invest in �rm-speci�c human capital. In-

creasing (decreasing) the bonus paid after productive skills have been acquired, c2H ,

strengthens (weakens) the incentive to invest in �rm-speci�c human capital.

Proof. (I): Knowing the compensation in both periods, the agent maximizes hisexpected utility by choosing hNOe1e2 at t = 0 for any combination of e¤orts (e1; e2) and

compensations (c1H ; c2H)

maxhNOe1e2

E�UNO

�= E [U1j e1; h] + E [U2j e2; h] :

Evaluating the derivative at h = 0 provides insight into the agent�s incentive to

5Notice that the notation hNOHH represents the level of �rm-speci�c human capital under nooutsourcing (NO) when the agent picks the high e¤ort level, eH , in both periods (HH). Moreover,if the levels of e¤ort in either period, (e1; e2), are left unspeci�ed, this will be denoted hNOe1e2 .

138

invest in h

@E�UNO

�@h

�����h=0

= �12e�1=21 c1H + e

1=22 �c2H > 0) c2H >

1

2�e�1=21 e

�1=22 c1H :

That is, if the principal at t = 0 can commit to a compensation structure where

c2H > 1= (2�) e�1=21 e

�1=22 c1H , the agent can be given incentive to invest in �rm-

speci�c human capital. This completes the argument for (I).

(II): Since @2E�UNO

�=@h2 = �1=4 (e1 � h)�3=2 < 0, maxE

�UNO

�wrt. hNOe1e2 rep-

resents a concave problem with the following solution

@E�UNO

�@h

= �12(e1 � h)�1=2 c1H + e1=22 �c2H = 0)

hNOe1e2 = max

(0; e1 �

�c1H2�c2H

�21

e2

)(8)

Intuitively, hNOe1e2 characterizes the agent�s investment in �rm-speci�c human capital

for given combinations of e¤orts and compensations in each period. The higher the

compensation in period 1 is, the lower will be the agent�s investment in learning.

This is due to the fact that increasing the salary in period 1 makes it less lucrative

to invest in h, since skill improvement is counterproductive in period 1 and reduces

the probability of realizing the high outcome. Increasing the compensation in period

2, however, has the opposite e¤ect as h is productive in period 2 and makes it more

likely to achieve the high outcome. This concludes the argument for (II).

Given these results for the agent, we are ready to state the principal�s overall con-

strained maximization problem in second-best to analyze how the compensation in

139

each period will be settled

maxhNOHH ;c1H ;c2H

E��NO

�= (eH � h)1=2 (xH � c1H) +

�1� (eH � h)1=2

�xL (9)

+e1=2H (1 + �h) (xH � c2H) +

�1� e1=2H (1 + �h)

�xL

s.t.�eH � hNOHH

�1=2c1H + e

1=2H

�1 + �hNOHH

�c2H � 2�H � 0 (PC)�

eH � hNOHH�1=2

c1H + e1=2H

�1 + �hNOHH

�c2H � 2�H �

�eL � hNOLH

�1=2c1H

+e1=2H

�1 + �hNOLH

�c2H

��H (IC1)�eH � hNOHH

�1=2c1H + e

1=2H

�1 + �hNOHH

�c2H � 2�H �

�eH � hNOHL

�1=2c1H

+e1=2L

�1 + �hNOHL

�c2H

��H (IC2)�eH � hNOHH

�1=2c1H + e

1=2H

�1 + �hNOHH

�c2H � 2�H �

�eL � hNOLL

�1=2c1H

+e1=2L

�1 + �hNOLL

�c2H ; (IC3)

where hNOHH , hNOLH , h

NOHL and h

NOLL can all be calculated on the basis of equation (8)

for respective combinations of e¤ort levels e1; e2 and compensation levels c1H ; c2H .

On the basis of (9), proposition 2 below will be proven.

Proposition 2 (Company) (I) If �rm-speci�c human capital is counterproductiveto the company, it equates the �rst period bonus, c1H , with the second period bonus,

c2H , and realizes no investment in skill improvement (�rst-best), hNOHH = 0. (II)

If �rm-speci�c human capital is productive to the company, it sets the �rst period

bonus, c1H , lower than the second period bonus, c2H , and realizes an investment in

�rm-speci�c human capital that is larger than �rst-best, hNOHH > hFB.

Proof. (I): If long-term actions are unproductive (eH� � 1=2), the principal wantsto induce hNOHH = 0 all things being equal. From (8) this can be ensured if the

principal selects identical compensations in both periods, c�1H = c�2H . The question

then becomes whether this compensation policy maximizes the principal�s pro�t and

induces the agent to exert the high e¤ort level in both periods. To establish that

this in fact is the case, we check to see if the IC�s are all satis�ed. As such, it can

be noticed that hNOHH = 0 implies hNOHL = h

NOLH = hNOLL = 0 since it follows from (8)

that hNOLL ; hNOLH ; h

NOHL � hNOHH . Assuming that IC1 is binding (and thus satis�ed), we

140

�nd for c�1H = c�2H

e1=2H c�1H + e

1=2H c�1H � 2�H = e

1=2L c�1H + e

1=2H c�1H � �H ) c�1H =

�H

e1=2H � e1=2L

:

Rearranging IC2 for c�1H = c�2H gives

e1=2H c�1H + e

1=2H c�1H � 2�H � e

1=2H c�1H + e

1=2L c�1H � �H ) c�1H �

�H

e1=2H � e1=2L

;

which is satis�ed for c�1H = �H=�e1=2H � e1=2L

�in particular. Finally, rearranging IC3

and inserting c�1H = c�2H = �H=

�e1=2H � e1=2L

�show that IC3 is satis�ed

e1=2H c�1H + e

1=2H c�2H � 2�H � e

1=2L c�1H + e

1=2L c�2H )�

e1=2H � e1=2L

�c1H +

�e1=2H � e1=2L

�c2H � 2�H � 0)�

e1=2H � e1=2L

� �H

e1=2H � e1=2L

+�e1=2H � e1=2L

� �H

e1=2H � e1=2L

� 2�H = 0:

To see that c�1H = c�2H also maximizes the principal�s pro�t, we only need to ob-

serve that choosing c�2H > c�1H in (8) potentially induces the agent to pick h

NOHH > 0;

which can never be optimal given that �rm-speci�c human capital is assumed coun-

terproductive. Similarly, selecting c�1H > c�2H only reduces the agent�s incentive to

undertake learning even further, which can never be optimal since c�1H = c�2H al-

ready ensures hNOHH = 0. Consequently, when long-term actions are unfavorable to

the principal, he selects c�1H = c�2H = �H=

�e1=2H � e1=2L

�and arrives at the �rst-best

investment in �rm-speci�c human capital, hNOHH = hNOFB = 0. This concludes the

proof of (I).

(II): If long-term actions are productive (eH� > 1=2) to the principal, no closed

formed solution to the maximization problem in (9) can be derived. Yet, the follow-

ing argument can be made: assume that c�1H � c�2H and look at the compensationvariation cvar1H � c�1H � "; c�2H . Inserting this variation into (8) provides us with thefollowing level of �rm-speci�c human capital

hNOe1e2 = e1 ��c�1H � "2�c�2H

�21

e2: (10)

141

Inserting (10) into the principal�s pro�t (9) and rearranging the terms give

E��NO

�=

�c�1H � "2�c�2H

�1

e1=22

!(�x� c�1H + ")

+e1=22

1 + �

e1 �

�c�1H � "2�c�2H

�21

e2

!!(�x� c�2H) + 2xL:

Taking the �rst order derivative of E��NO

�wrt. " and evaluating this at " = 0 give

(note that the positive sign follows from c�1H � c�2H)

@E��NO

�@"

�����"=0

=

�c�1Hc�2H

� 1��

1

2�c�2H

�1

e1=22

�x+

�c�1H2�c�2H

�1

e1=22

� 0:

Intuitively, if " increases and hence the period 1 compensation, cvar1H , decreases, the

principal�s pro�t, E��NO

�, will increase.

Next, consider the agent�s utility, E�UNOe1e2

�and insert (10)

E�UNOe1e2

�=

�1

2�

�(c�1H � ")

2

c�2H

1

e1=22

+e1=22

1 + �

e1 �

�1

2�

�2�c�1H � "c�2H

�21

e2

!!c�2H

��1 � �2:

Taking the �rst order derivative E�UNOe1e2

�wrt " gives

@E�UNOe1e2

�@"

= ��1

�

��(c�1H � ")c�2H

�1

e1=22

+1

2�

�c�1H � "c�2H

�1

e1=22

= � 1

2�e1=22

(c�1H � ")c�2H

(11)

< 0:

Rewriting (11) for each combination of e¤ort (e1; e2) and evaluating this at " = 0

142

give

@E [UeL;eH ]

@"

����"=0

= � 1

2�e1=2H

c�1Hc�2H

@E [UeH ;eL ]

@"

����"=0

= � 1

2�e1=2L

c�1Hc�2H

@E [UeL;eL ]

@"

����"=0

= � 1

2�e1=2L

c�1Hc�2H

@E [UeH ;eH ]

@"

����"=0

= � 1

2�e1=2H

c�1Hc�2H

:

Hence, we have

@E [UeL;eL ]

@"

����"=0

=@E [UeH ;eL ]

@"

����"=0

� @E [UeL;eH ]

@"

����"=0

=@E [UeH ;eH ]

@"

����"=0

< 0:

Consequently, when " increases and thus the period 1 compensation, cvar1H , decreases,

the decrease in E [UeH ;eH ] will always be smaller than or equal to the decrease in

E [UeL;eL ], E [UeL;eH ] ; E [UeH ;eL ]. That is, if the IC�s are satis�ed for c�1H � c�2H , this

will also be the case for cvar1H < c�2H .

Since the principal�s pro�t increases when cvar1H decreases and the IC�s are still satis-

�ed, the principal will select cvar1H < c�2H . As a result of this, the level of �rm-speci�c

human capital will increase above �rst-best

hFB = e1 ��1

2�

�21

e2< e1 �

�1

2�

�2�cvar1Hc�2H

�21

e2= hcvar1H c2H

:

This concludes the proof of (II).

In sum, the analysis shows that under full commitment the agent can be given

incentive to invest in �rm-speci�c human capital by adjusting the compensation

in period 1 and 2 (c1H and c2H) properly (proposition 1). Additionally, we have

proved that the principal can always arrive at the �rst-best level of �rm-speci�c

human capital by equating c1H and c2H . However, he will only do so in second-best

if �rm-speci�c human capital is counterproductive since this is the cheapest way of

ensuring hNOHH = 0. If �rm-speci�c human capital is productive, it will always be

optimal for the principal to choose c1H < c2H and realize a higher investment in �rm-

speci�c human capital in second-best than in �rst-best, hNOHH > hFB (proposition 2).

The intuition is that the principal expectedly saves more from lowering the period

1 compensation than is lost from inducing a �rm-speci�c investment that is above

143

�rst-best when taking into consideration the agent�s reaction pattern.

2.3 Extended model: Introducing outsourcing �exibility

Having characterized the benchmark situation, we extend the basic setup and in-

troduce an external third-party supplier with whom the company can contract at

t = 1 for production in period 2 if it sees �t to do so. The reason for the company

establishing an alternative to internal production can be grounded in two distinct

potentially value-creating e¤ects. First, outsourcing �exibility naturally represents

a real option to the company as it enables the company to avoid situations where

expected in-house production in period 2 is bad; the more lucrative the outsourcing

alternative is the higher the value of the real option is. Second, outsourcing �exibil-

ity may make it easier and thereby cheaper to incentivize the worker to undertake

long-term actions and invest in �rm-speci�c human capital. These two e¤ects will

thus be the centre of analysis in the following where special attention will be given

to the employee�s problem since the complexity of the company�s constrained max-

imization problem leaves no room for meaningful comparative statics.

Formally, we will suppose that the principal at t = 1 is given the �exibility to

dismiss the worker in return for a proper severance fee, , settled at t = 0, if the

expected pro�tability can be improved by contracting with the third-party supplier

in period 2. For this reason, it is assumed that the external contractor o¤ers a price

~p � U (0; 1) so that the distribution hereof is known to both the principal and theagent at t = 0 while the actual value is not realized and observed before t = 1.6

Additionally, the principal conducts an interview with the agent at t = 1 to learn

his type (i.e. his current skill level), h, since this provides the principal with insight

about the expected pro�tability of maintaining production in-house.

To make the model tractable, we continue the simplicity from the basic setup

and extend the notation with an outsourcing scenario only. As such, the probability

of realizing the high output if production is contracted out is not a¤ected by any

factors inside the model but simply given by the constant �. Altogether, this implies

that the agent�s expected utility in period 2 from outsourcing (O) is

E�UO2�= (12)

6Notice that it can be problematic to assume that p � U (0; 1) if the output levels in period2, x2L and x2H , are very high since this might create a situation where outsourcing is always thepreferred alternative. A solution to this could be to model p � U (0; �p). However, to maintainsimplicity this has been abandoned. Instead, it is assumed that the output levels in period 2 arenever so high as to always render the outsourcing alternative favorable.

144

while the expected pro�t to the principal can be expressed as

E��O2�= �xH + (1� �)xL � p� : (13)

Figure 2 below outlines the sequence of events in the model when outsourcing is

introduced.

<Insert �gure 2 here>

2.4 The probability of outsourcing: The principal�s decision

rule

To determine what triggers the principal to contract out production at t = 1, it is

important to understand how E��NO2

�� eH ; h� and E ��O2 � behave as a function ofh and p, respectively. For this reason, we start out by rewriting (4) using c2L = 0

from (6)

E��NO2

�� eH ; h� = e1=2H (1 + �h) (xH � c2H) +

�1� e1=2H (1 + �h)

�xL

= e1=2H � (�x� c2H)h+ xL + e1=2H (�x� c2H) : (14)

Equation (14) depicts a straight line in h with interception xL + e1=2H (�x� c2H)

and a positive slope of e1=2H � (�x� c2H). Similarly, from (13) it can be seen that

E��O2�depicts a straight line in p with interception �xH + (1� �)xL � and slope

�1.On the basis of observing h and p, the decision rule for the principal is to

select the third-party supplier if the expected payo¤ from outsourcing the activities

exceeds the expected payo¤ from internalizing period 2 production, i.e. E��O2�� p� >

E��NO2

�� eH ; h�. Moreover, for every observed h there must exist a p0 such thatE��O2�� p0� = E ��NO2 �� eH ; h�. This implies

p0 = �xH + (1� �)xL � � E��NO2

�� eH ; h�= ��x� � e1=22 (1 + �h) (�x� c2H) : (15)

Hence, the probability at t = 0 of the principal selecting outsourcing at t = 1 can now

be expressed as the probability at t = 0 of the observed price o¤er p from the third-

party supplier at t = 1 being less than p0, i.e. P�E��NO2

�� eH ; h� < E ��O2 �� p�� =

145

P (p < p0).

P (p < p0) =

Z p0

0

1

1� 0dp

= p0: (16)

Figure 3 below illustrates these arguments.

<Insert �gure 3 here>

2.5 Analysis under full commitment and outsourcing �exi-

bility

Understanding what makes the principal choose outsourcing at t = 1, we are ready

to study how the outsourcing �exibility a¤ects the principals expected pro�t directly

(real option e¤ect) as well as indirectly through changes in the agent�s incentive to

invest in �rm-speci�c human capital (behavioral e¤ect).

Focusing �rst on the principal�s expected pro�t, the objective function can be de-

rived as

E��O�= E [�1jh; eH ] +

Z 1

0

E [�2jh; eH ] dp

= E [�1jh; eH ] +Z p0

0

E��O2�� eH� dp+ Z 1

p0

E��NO2

��h; eH� dp= E [�1jh; eH ] +

Z p0

0

(�x2H + (1� �)x2L � p� ) dp

+(1� p0)E��NO2

��h; eH�= E [�1jh; eH ] +

Z p0

0

(�x2H + (1� �)x2L � ) dp

�Z p0

0

pdp+ (1� p0)E��NO2

��h; eH�= E [�1jh; eH ] + p0 (�x2H + (1� �)x2L � )

�12p20 + (1� p0)E

��NO2

��h; eH�= E [�1jh; eH ] + p0

��x2H + (1� �)x2L � � E

��NO2

��h; eH���12p20 + E

��NO2

��h; eH�= E [�1jh; eH ] +

1

2p20 + E

��NO2

��h; eH� : (17)

146

Intuitively, 1=2p20 in (17) represents the real option value (e¤ect) of the outsourc-

ing �exibility compared to the case of no outsourcing, E��NO

�= E [�1jh; eH ] +

E��NO2

��h; eH�; if the principal can choose the severance fee, , unreservedly hecan always arrive at the scenario without the �exibility by choosing so large that

p0 = 0.

Focusing next on the agent�s incentive to invest in �rm-speci�c human capital, the

analysis will be divided into two parts, �rst-best and second-best, in keeping with

the treatment of no outsourcing.

2.5.1 First-best

In �rst-best the principal selects c1L = c1H = c2L = c2H = �H and = 0 and

maximizes (17) by choosing hOFB7

maxhOFB

E��O�= (eH � h)1=2�x+

1

2

�pFB0

�2+ e

1=2H (1 + �h)�x+ 2xL:

The �rst-order condition wrt. h is

@E��O�

@h= � 1

2(eH � h)�1=2�x| {z }

Expected decrease in

period 1 pro�t

+ pFB0@pFB0@h

+ e1=2H ��x| {z }

Expected increase in

period 2 pro�t

= 0; (18)

where @pFB0 =@h = �e1=2H ��x < 0. The explicit solution to hOFB is complicated and

does not bring much insight. Instead, the intuition is clear from (18); it is optimal

for the principal to increase the agent�s investment in �rm-speci�c human capital as

long as the expected decrease in period 1 rent is more than countered by an increase

in the period 2 rent.

Furthermore, comparing (18) to �rst-best under no outsourcing in (5),

@E��NO

�=@h = �1

2(eH � h)�1=2�x + e1=2H ��x, reveals how the real option e¤ect

of the outsourcing �exibility a¤ects the level of �rm-speci�c human capital. That is,

pFB0�@pFB0 =@h

�< 0 represents the reduction in the real option e¤ect as h increases

and in-house production becomes more attractive, which in turn makes the prin-

cipal reduce the �rst-best investment, hOFB < hNOFB , for h productive (eH� > 1=2).

The intuition is that the outsourcing �exibility reduces the downside of operating

7Notice that pFB0 represents the probability of outsourcing when �rst-best conditions exist,c2H = 0 and = 0; i.e. pFB0 � p0 (h).

147

under a low level of �rm-speci�c human capital in period 2. Consequently, this

makes it optimal to induce the agent to exert short-term productive e¤ort in period

1 since this increases the expected pro�t in period 1 (i.e. the probability of pro�ting

from learning in period 2 is reduced). Finally, if long-term actions are unproductive

(eH� � 1=2) to the principal, he will continue to select hOFB = hNOFB = 0.

2.5.2 Second-best

Under second-best, the incentives are no longer necessarily aligned which requires

the agent�s incentive compatibility constraints and participation constraint to be

satis�ed. Thus, at t = 0 the agent has to decide how much e¤ort to exert in each

period and how much e¤ort should be concentrated on skill improvement in period

1 when taking into consideration that production might be outsourced at t = 1. As

such, the worker faces the following maximization problem for given e¤ort choices

(e1; e2) and compensations (c1H ; c2H)

maxhOe1e2

E�UO�= (e1 � h)1=2 c1H � �1 + p0 + (1� p0)

�e1=22 (1 + �h) c2H � �2

�:

Noting that @p0=@h = �e1=22 � (�x� c2H) < 0 gives the following �rst order condi-tion wrt. h

@E�UO�

@h= � 1

2(e1 � h)�1=2 c1H +

@p0@h

� �

�e1=22 (1 + �h) c2H � �2

��+(1� p0) e1=22 �c2H

= 0: (19)

The �rst term in (19) re�ects that investing in h in period 1 reduces time available

for short-term productive e¤ort, which in turn decreases the probability of realizing

the high outcome and hence the expected rent in period 1. At the same time, how-

ever, investing in h reduces the probability of outsourcing and thus increases the

agent�s chance to stay on and earn a rent during period 2. This is captured by the

last two terms in (19). Therefore, (19) partly re�ects the same dynamic as under

no outsourcing, yet complicated by factors related to the risk of outsourcing.

Using (19) proposition 3 below can be derived to capture the behavioral e¤ect of the

outsourcing �exibility under long-term contracting.

Proposition 3 (I �Cost dominates attractiveness) If the cost of layo¤ is high andthe severance fee exceeds the expected compensation from continued in-house pro-

148

duction evaluated at a level of �rm-speci�c human capital analogous to that of no

outsourcing, hNOe1e2, the worker decreases his investment in �rm-speci�c human capital

under outsourcing compared to no outsourcing, hOe1e2 < hNOe1e2. (II �Attractiveness

dominates cost) If the cost of layo¤ is low and the severance fee falls below the

expected compensation from continued in-house production evaluated at a level of

�rm-speci�c human capital analogous to that of no outsourcing, hNOe1e2, the worker

increases his investment in �rm-speci�c human capital under outsourcing compared

to no outsourcing, hOe1e2 > hNOe1e2.

Proof. Assume as a point of reference that the current choice of investment in�rm-speci�c human capital under outsourcing is equal to that of no outsourcing,

i.e. hOe1e2 = hNOe1e2. Furthermore, from the case of no outsourcing in (8), we know

that @E�UNO

�=@h = �1

2(e1 � h)�1=2 c1H + e1=22 �c2H = 0. Using this, the derivative

from (19) can be rewritten as

@E�UO�

@h

�����h=hNOe1e2

= � 12

�e1 � hNOe1e2

��1=2c1H + e

1=22 �c2H � p0

�hNOe1e2

�e1=22 �c2H

+@p0@h

� �

�e1=22

�1 + �hNOe1e2

�c2H � �2

��=

@E�UNO

�@h

�����h=hNOe1e2

� p0�hNOe1e2

�e1=22 �c2H

+@p0@h

� �

�e1=22

�1 + �hNOe1e2

�c2H � �2

��= � p0

�hNOe1e2

�e1=22 �c2H| {z }

Cost: cost of layo¤

+@p0@h

� �

�e1=22

�1 + �hNOe1e2

�c2H � �2

��| {z }

Attractiveness

: (20)

Equation (20) isolates the e¤ects related to the risk of outsourcing; the cost term

can be interpreted as measuring the cost of layo¤ to the agent as it represents the

probability of outsourcing, evaluated at a current level of �rm-speci�c human capital

hNOe1e2 multiplied by the derivative of the rent lost if h is increased and outsourcing is

chosen. This term will thus always be negative. Similarly, the attractiveness term

can be interpreted as the agent�s incentive to stay attractive to the principal by fur-

thering his investment in �rm-speci�c human capital, since it measures how much

the probability of outsourcing is reduced for h increased, multiplied by the agent�s

gain/loss realized if outsourcing is chosen. Depending on the situation at hand, this

term can thus be either positive or negative. On the basis of (20), we can formulate

149

the arguments for (I) and (II).

(I): If the risk of layo¤ evaluated at hNOe1e2 , p0�hNOe1e2

�, is relatively high as a result of a

large �, and the di¤erence between the severance fee and the expected rent in period

2 is positive for hNOe1e2, ��e1=22

�1 + �hNOe1e2

�c2H � �2

�> 0, then (20) will in total

be negative. This in turn leads the agent to reduce his investment in �rm-speci�c

human capital compared to what would have been the case had the outsourcing �ex-

ibility not existed, i.e. hOe1e2 < hNOe1e2. The reason is that the outsourcing conditions

are vastly favorable to the principal, which makes the risk of layo¤ imminent to the

agent (high cost of layo¤). Thus, combined with the fairly lucrative severance fee

paid out at dismissal (low attractiveness), this in total reduces the agent�s incentive

to invest in h. This completes the argument for (I).

(II): However, if the probability of outsourcing, p0�hNOe1e2

�, is relatively low as a

result of a small �, and the di¤erence between the severance fee and the expected

rent in period 2 is negative for hNOe1e2 , ��e1=22

�1 + �hNOe1e2

�c2H � �2

�< 0, then

a situation may arise where the agent will increase his investment in h compared

to what would have been the case had the outsourcing �exibility not existed, i.e.

hOe1e2 > hNOe1e2. The reason is that the agent has incentive to realize the rent in pe-

riod 2, e1=22 (1 + �h) c2H � �2, instead of the severance fee, , which can be doneby investing more in h to increase attractiveness to the principal and reduce the

probability of outsourcing. This concludes the argument for (II).

Fundamentally, the cost and the attractiveness depend on and c2H as seen from

the terms in (20). Now, imagine a situation where the introduction of outsourc-

ing implies that the cost dominates attractiveness for existing contracts established

under no outsourcing. From proposition 3 (I), we know that this leads the agent

to reduce his investment in �rm-speci�c human capital compared to no outsourc-

ing, hOe1e2 < hNOe1e2, which is applicable to the principal only if h is unproductive

(eH� � 1=2). If, however, h is productive (eH� > 1=2) it may be in the principal�sbest interest to limit this reduction by properly combining and c2H .

We will thus examine how these two variables a¤ect the agent�s incentive to invest

in h, although the analysis has to be conducted indirectly on the basis of (19) since

the explicit solution to hOe1e2 is too complex and provides no meaningful insights.

Consequently, given that @p0=@c2H = e1=22 (1 + �h) > 0 and @ (@p0=@h) =@c2H =

150

e1=22 � > 0, we �nd the following �rst order derivatives of (19) wrt. c2H and

@�@E�UO�=@h

�@c2H

= e1=22 �

0@ 2� � 2e1=22 (1 + �h) c2H

�+�2e1=22 (1 + �h)� �

��x+ 1 + �2

1A (21)

@�@E�UO�=@h

�@

=@p0@h

+ e1=22 �c2H = e

1=22 � (2c2H ��x) : (22)

Intuitively, @�@E�UO�=@h

�=@c2H measures the e¤ect on the expected utility of in-

creasing h when c2H is increased. Likewise, @�@E�UO�=@h

�=@ re�ects the e¤ect

on the expected utility of increasing h when is increased. Therefore, both compar-

ative statics indirectly provide information about the agent�s incentive to decrease

or increase his investment in �rm-speci�c human capital. The �ndings are captured

in proposition 4 below.

Proposition 4 (Incentive mechanisms) A combination of long-term contracts

and a severance fee can regulate the worker�s incentive to invest in �rm-speci�c

human capital. (I) Promising a higher bonus in period 2, c2H , has two opposite

directed e¤ects; it incentivizes the worker to further his investment in �rm-speci�c

human capital, h, but it also increases the risk of layo¤, p0, which naturally reduces

the incentive to undertake the investment. (II) Fortunately, if increasing the level of

�rm-speci�c human capital is productive to the company, the increased risk of layo¤

can be mitigated by raising the severance fee, .

Proof. (I): Assume = 0 and consider (21).

For c2H ! 0 (21) reduces to

@�@E�UO�=@h

�@c2H

= e1=22 �

��e1=22 (1 + �h)� �

��x+ 1 + �2

�:

Thus, if � ! 1 the external supplier is relatively productive, which might cre-

ate a situation where the outsourcing o¤er will always be the preferred alternative

and the agent is left with no incentive to invest in h, if c2H is increased above 0,

@�@�@E�UO�=@h

��=@c2H < 0.

Moreover, if � ! 0 the external supplier is relatively unproductive, which might in-

centivize the agent to invest in h, if c2H is increased, @�@�@E�UO�=@h

��=@c2H � 0.

Consequently, c2H can be used to incentivize the agent if the current level of c2H is

151

relatively low (! 0).

For c2H !1 (21) will grow negative. That is,

@�@E�UO�=@h

�@c2H

< 0:

Hence, if c2H is increased from a current level that is relatively high (!1), theagent�s incentive to invest in h is reduced as the probability of outsourcing, p0, is

increased. This concludes the argument for (I).

(II): Assume = 0 and consider (22).

If c2H < 1=2�x then @�@E�UO�=@h

�=@ < 0. Hence, when c2H is set too low

(< 1=2�x), increasing will only make the agent more reluctant to invest in �rm-

speci�c human capital, h, to avoid layo¤.

If c2H � 1=2�x then @�@E�UO�=@h

�=@ � 0. Hence, when c2H is set su¢ ciently

high (� 1=2�x), increasing will have a positive e¤ect on the agent�s incentive toinvest in h as this reduces the probability of outsourcing. Under such circumstances

can function as a moderator. This completes the argument for (II).

Thus, proposition 4 reveals how the combined use of c2H and can incentivize the

agent to invest in �rm-speci�c human capital. The basic idea for the principal is

to provide "pure" incentive to invest in h through c2H . Yet, c2H can only be used

single-handedly to a certain extent since increasing c2H also increases the probability

of outsourcing, p0. However, if c2H � 1=2�x, can function as a moderator that

reduces the probability of outsourcing and provides further scope for strengthening

the agent�s incentive to invest in h through c2H .

On the basis of this, we can elaborate on the principal�s cost of initiating learning by

bringing together proposition 3 and 4. Speci�cally, proposition 4 implies that when

�rm-speci�c human capital is productive to the principal (eH� > 1=2), the combined

use of c2H and under outsourcing may ensure that the agent�s incentive from

attractiveness dominates the cost when undertaking learning. Consequently, the

�exibility to outsource holds the potential to make the agent invest in �rm-speci�c

human capital at a lower cost, since proposition 3 (II) asserts that the behavioral

e¤ect automatically leads the agent to invest more in �rm-speci�c human capital

under outsourcing than under no outsourcing for a given c2H . Thus, establishing

152

the same investment under outsourcing as under no outsourcing, hNOe1e2 = hOe1e2, can

be weakly cheaper for the principal as it may require a lower c2H , possibly in com-

bination with a properly chosen . However, this will only be true if the agent does

not deviate from e1 = e2 = eH when c2H is changed and is introduced.

3 Short-term contracting

Having analyzed how incentives to invest in �rm-speci�c human capital can be

provided under full contract commitment, we will now assume that the principal

can no longer commit to long-term contracts but is instead forced to make use of

short-term engagements. In this respect, it is important to understand that under

short-term contracting the principal cannot commit to the second period contract

at t = 0, meaning that this contract is e¤ectively established and negotiated for

the �rst time at t = 1. In the following, the setup and notation from the previous

section are kept the same. The only modi�cation is that the second period contract

will not be initialized before t = 1.

<Insert �gure 4 here>

3.1 No outsourcing �exibility

In keeping with the structure from section 2, we will return to a situation where

no outsourcing �exibility is provided to the principal. Still, however, the principal

wants to induce eH in both periods, and so he needs to provide the agent with

a compensation scheme that encourages such behavior. Focusing �rst on period

2 and working our way backwards, this implies that the principal�s only interest

under no contract commitment after observing h at t = 1 is to o¤er a su¢ ciently

high compensation to ensure that the agent takes on the high e¤ort level in period 2.

The reason is that e¤ort will automatically be channeled towards physical productive

e¤ort which is the agent�s only alternative in period 2. Formally, using (3) this

requires the following inequality to be satis�ed for any given h observed

E�UNO2

�� eH ; h� � E �UNO2 �� eL; h�) �c2 ��H�

e1=2H � e1=2L

�(1 + �h)

:

Hence, only the di¤erence in compensation a¤ects the agent�s choice of e¤ort level,

leading the principal to select the minimum compensation programme in period 2,

153

(c�2L; c�2H), as follows

c�2L = 0 and c�2H =�H�

e1=2H � e1=2L

�(1 + �h)

: (23)

With this in mind, the agent�s incentive to invest in �rm-speci�c human capital can

now be analyzed. Initially, we notice that the �rst-best level of �rm-speci�c human

capital under short-term contracting has not changed and is still given by (5). That

is, the principal, ceteris paribus, wants to induce a positive level, hNOFB > 0. However,

since it is always optimal for the principal, after observing h at t = 1, to o¤er the

agent the minimum compensation programme from (23) ; the expected utility in

period 2 can be expressed as follows

E�UNO2

�� eH ; h� = e1=2H (1 + �h)

�H�e1=2H � e1=2L

�(1 + �h)

� �H

= e1=2H

�H

e1=2H � e1=2L

� �H (24)

� E�UNO2

�:

Equation (24) reveals that the expected utility (expected compensation component)

does not depend on the investment in skill improvement in period 1 and implies that

the minimum compensation programme in period 2 can never be used to provide

incentives to invest in h at t = 0. Put di¤erently, it will never be credible for

the principal to promise a higher salary in period 2 at t = 0 than the minimum

compensation programme due to contracts being short-term and renegotiated at

t = 1. This has implications for the agent�s incentive to invest in h as the following

proposition 5 reveals.

Proposition 5 (Worker) In the absence of outsourcing possibilities, the agent willnever invest in �rm-speci�c human capital if contracts are renegotiated after each

period, hNOe1H = 0.

154

Proof. The absence of outsourcing possibilities leaves the agent with the followingmaximization problem at t = 08

maxhNOe1H

E�UNO

�= (e1 � h)1=2 c1H � �H + E

�UNO2

�| {z } :Independent

of h

The �rst order condition with respect to h is

@E�UNO

�@h

= � (e1 � h)�1=2 c1H < 0: (25)

Intuitively, this means that marginally increasing the investment in �rm-speci�c

human capital decreases the expected utility, thereby leading the agent to always

choose hNOe1H = 0 and only exert physical productive e¤ort a.

Consequently, proposition 5 stands in clear contrast to proposition 1 under full

commitment where explicit incentives to invest in �rm-speci�c human capital can

be provided through the compensation programmes. Proposition 5 thus shows that

other means are needed if the principal wants to induce the agent to undertake

learning.

To �nish the analysis under no outsourcing, we can calculate the principal�s expected

pro�t. From proposition 5 it immediately follows that the minimum compensation

programme in period 2 from (23) reduces to

c�2H =�H

e1=2H � e1=2L

:

Furthermore, the lack of investments in skill improvement entails that the condition

ensuring that the agent picks eH in period 1 is given by

E�U1j eH ; hNOHH = 0

�� E

�U1j eL; hNOLH = 0

�) �c1 �

�H

e1=2H � e1=2L

:

Hence, only the di¤erence in compensation a¤ects the agent�s choice of e¤ort, e1, and

so the principal can induce eH by arranging the minimum compensation programme,

8Notice that he1H denotes an unspeci�ed e¤ort level, e1, in period 1 and the high e¤ort level(H) in period 2 has been chosen.

155

(c�1L; c�1H), in period 1 as follows

c�1L = 0 and c�1H =�H

e1=2H � e1=2L

:

Consequently, the minimum compensation structures ensuring eH in each period,

(c�1L; c�1H) and (c

�2L; c

�2H), reduce to identical expressions. In total, this implies that

the principal�s expected pro�t can be calculated as

E��NO

�= e

1=2H (xH � c�1H) + e

1=2H (xH � c�2H)

= 2e1=2H

xH �

�H

e1=2H � e1=2L

!:

3.2 Introducing outsourcing �exibility

Given that the agent underinvests in �rm-speci�c human capital compared to �rst-

best, we will once again introduce the outsourcing �exibility to study how this a¤ects

the situation. With the possibility of outsourcing, the agent�s maximization problem

at t = 0 takes the following form

maxhOHH

E�UO�= (eH � h)1=2 c1H � �H + p0 + (1� p0)E

�UNO2

�and results in the following �rst order condition wrt. h

@E�UO�

@h= � 1

2(eH � h)�1=2 c1H| {z }Cost: expected loss

in period 1 rent

+@p0@h

� � E

�UNO2

��| {z }

Attractiveness

= 0; (26)

where @p0=@h = �e1=2H ��x < 0. The cost term represents the agent�s expected loss

in period 1 rent from increasing learning. Furthermore, similarly to full commitment,

the last term in (26) can be interpreted as the agent�s incentive to stay attractive

to the principal by investing more in h; @p0=@h < 0 represents the reduction in the

risk of layo¤ when h increases while � E�UNO2

�denotes the gain/loss realized in

period 2 when the probability of outsourcing decreases.

Using (26) proposition 6 below can be derived to capture the behavioral e¤ect of the

outsourcing �exibility under short-term contracting.

156

Proposition 6 (I �Cost dominates attractiveness) If continued in-house produc-tion in period 2 is not lucrative, � E

�UNO2

�, the agent will continue to invest

nothing in �rm-speci�c human capital, hOHH = 0, even if outsourcing �exibility is

introduced. (II �Attractiveness dominates cost) If the expected gain from continued

in-house production in period 2 is lucrative, << E�UNO2

�, the risk of layo¤ is re-

sponsive to skill improvement, @p0=@h << 0, and the expected loss in period 1 rent

from skill improvement is small, 1=2e�1=2H c1H � 0, the worker can have an incentiveto invest in �rm-speci�c human capital, hOHH > 0, to stay attractive to the company.

Proof. (I): Assume � E�UNO2

�. It then follows from (26) that (notice @p0=@h =

�e1=2H ��x < 0)

@E�UO�

@h

�����h=0

= �12e�1=2H c1H +

@p0@h

� � E

�UNO2

��< 0:

Hence, under such circumstances the agent will continue to invest nothing in �rm-

speci�c human capital, hOHH = 0, as he receives more from dismissal than from

continued in-house production in period 2. That is, < E�UNO2

�is a necessary

but not su¢ cient condition for the outsourcing �exibility to incentivize the agent to

invest in hOHH > 0. This completes the argument for (I).

(II): Assume next < E�UNO2

�and consider

@E�UO�

@h

�����h=0

= �12e�1=2H c1H +

@p0@h

� � E

�UNO2

��: (27)

If � E�UNO2

�(in-house production is not lucrative) and/or @p0=@h = �e1=2H ��x �

0 (risk of layo¤ is not responsive to skill improvement) and/or 1=2e�1=2H c1H >> 0

(big expected loss in period 1 rent from skill improvement), a situation can happen

where @p0=@h� � E

�UNO2

��< 1

2e�1=2H c1H ; thus making (27) negative. In this case

the agent�s incentive to signal attractiveness is dominated by the loss in period 1

rent and so he invests nothing in �rm-speci�c human capital, hOHH = 0.

If << E�UNO2

�(in-house production is lucrative) and/or @p0=@h = �e1=2H ��x <<

0 (risk of layo¤ is responsive to skill improvement) and/or 1=2e�1=2H c1H � 0 (smallexpected loss in period 1 rent from skill improvement), a situation can happen

where @p0=@h� � E

�UNO2

��> 1

2e�1=2H c1H , thus making (27) positive. In this case

the agent has incentive to invest in �rm-speci�c human capital, hOHH > 0. That is,

157

in-house production is so lucrative to the agent that even though the expected rent

in period 2 does not explicitly depend on h, the outsourcing �exibility induces the

agent to invest in h for the very reason that it makes him more attractive to the

principal and reduces the risk of layo¤. This concludes the argument for (II).

Consequently, comparing proposition 6 to proposition 3 under full commitment, we

see that the outsourcing �exibility plays a more crucial role under short-term con-

tracting than is the case under long-term contracting. Speci�cally, the outsourcing

�exibility holds the potential to trigger investments in h under short-term contract-

ing whereas it only serves the purpose of easing incentives to invest in h under

long-term contracting.

As a curiosum to the comparison of proposition 6 and 3, it is perhaps surpris-

ing that (26) does not depend on the risk of layo¤, p0, explicitly but only on the

partial derivative, @p0=@h, since this implies that the ability of the third-party sup-

plier, modeled through �, has no in�uence on the agent�s decision to invest in h.

That is, two separate cases where the external supplier is either highly skilled (�

high) or poorly skilled (� low) can lead the agent to undertake the same level of

skill improvement. This is due to the already mentioned fact that the expected

in-house compensation in period 2 is not a¤ected by the agent�s investment, and so,

as explained above, the agent�s decision to take long-term actions boils down to how

much this reduces expected rent in period 1 compared to the proportional gain in

period 2.

Clearly, this stands in contrast to the full commitment case in (19) where

@E�UO�=@h depends explicitly on p0 and thus � for the very reason that the ex-

pected compensation in period 2 is in�uenced by h. Therefore, under such circum-

stances the two separate scenarios where the external supplier is either highly skilled

(� high) or poorly skilled (� low) can never lead the agent to undertake the same

level of skill improvement, all else equal.

Having clari�ed that the agent can be incentivized to invest in �rm-speci�c hu-

man capital under short-term contracting by the introduction of an outsourcing

alternative, we will, in keeping with the treatment of the full commitment case in

section 2.5.2, study how c2H and a¤ect the incentive. The �ndings are captured

in the following proposition.

Proposition 7 (Incentive mechanisms) A combination of short-term contracts

and a severance fee cannot be used to incentivize the worker to invest in �rm-speci�c

158

human capital. (I) Due to contract commitment not being credible, the bonus o¤ered

after long-term actions have been taken, c2H , will always be reduced by the company

to eliminate the bene�t to the worker of undertaking the investment. (II) Addi-

tionally, the severance fee, , reduces the risk of layo¤ and hence the size of the

investment in �rm-speci�c human capital required to stay attractive to the company.

Proof. (I): Consider the period 2 upstate compensation set at t = 1 from (23)

c�2H =�H�

e1=2H � e1=2L

�(1 + �h)

:

For h ! 1, c�2H ! 0. Hence, the principal takes advantage of the agent�s �rm-

speci�c investment and reduces the bonus in period 2 (c�2H ! 0) according to the

level of investment that he observes (h!1). In consequence, this eliminates thebene�t to the worker of undertaking the investment to increase productivity in pe-

riod 2. This concludes the argument for (I).

(II): Consider next the �rst derivative of (15) wrt.

@p0@

= �1:

Hence, increasing decreases the risk of layo¤ p0, which automatically leads the

agent to reduce his investment in h. This can explicitly be seen from (26) by taking

the �rst derivative wrt.

@(@E�UO�=@h)

@ =@p0@h

= �e1=2H ��x < 0:

That is, a lower skill level is needed for the agent to stay attractive to the principal

and keep period 2 production in-house. This completes the argument for (II).

Thus, comparing proposition 7 to proposition 4 under full commitment, reveals that

c2H and play a very di¤erent role under long-term contracting than is the case

under short-term contracting. Under short-term contracting c2H cannot be used to

incentivize the agent to invest in �rm-speci�c human capital, since he cannot credibly

commit to the period 2 contract at t = 0 and hence will always exploit the �rm-

speci�city of the agent�s investment ex post the investment has been undertaken. In

relation to this, cannot play the role of a moderator that reduces the probability

of outsourcing as it increases when c2H is raised. Consequently, neither c2H nor

159

can be used to strengthen the agent�s incentive to invest in h under short-term

contracting, which stands in clear contrast to the case of full commitment.

3.3 Examples under short-term contracting

Given the analysis above, the natural question is under what circumstances introduc-

ing an outsourcing alternative adds or destroys value to the principal and the agent.

To discuss this, the principal�s constrained maximization problem under outsourc-

ing will initially be established. Subsequently, the �ndings will be illustrated with

carefully chosen examples as no closed form solutions can be derived for the problem.

Since @2E�UO�=@h2 = �1

4(e1 � h)�3=2 c1H < 0, (26) is a concave function with

a maximum derived from the following �rst-order condition wrt. h

@E�UO�

@h= 0) @p0

@h

� � E

�UNO2

��=1

2(eH � h)1=2 :

Since 1=2 (eH � h)1=2 � 0 and @p0=@h = �e1=2H ��x < 0, this implies that E�UNO2

�needs to be greater than , E

�UNO2

�� , in order to ensure the existence of a

maximum. Assuming this is the case provides the following optimal level of �rm-

speci�c human capital

hOe1H = e1 �c21H

4�@p0@h( � E [UNO2 ])

�2 :Hence, for each e¤ort level available at t = 0, the optimal investment in �rm-speci�c

human capital can be determined as follows

hOLH = eL �c21H

4�@p0@h( � E [UNO2 ])

�2 and

hOHH = eH �c21H

4�@p0@h( � E [UNO2 ])

�2 : (28)

Given hOLH and hOHH , it is possible to establish the total expected utility for each

e¤ort level

E�UOLH

�=

�eL � hOLH

�1=2c1H + p0

�hOLH

� +

�1� p0

�hOLH

��E�UNO2

�E�UOHH

�=

�eH � hOHH

�1=2c1H + p0

�hOHH

� +

�1� p0

�hOHH

��E�UNO2

�:

160

Finally, on the basis of these expressions the principal�s constrained maximization

problem can be stated as

maxc1H ;

E��O�=�eH � hOHH

�1=2(xH � c1H) + e1=2H

�1 + �hOHH

�xH

� e1=2H�H

e1=2H � e1=2L

+1

2

�p0�hOHH

��2(29)

s.t.

E�UOHH

�� 0 (PC)

E�UOHH

�� E

�UOLH

�: (IC)

3.3.1 Does outsourcing �exibility bring value to the company undershort-term contracting?

To clarify whether the outsourcing �exibility is of value to the principal, we im-

mediately observe that when can be chosen unconstrained or is bounded from

below, 2 [ min;1[ for min � 0, the outsourcing �exibility can never be a bad

to the principal, as he can simply pick arbitrarily large to ensure p0 = 0 and

E��O�= E

��NO

�if necessary. Thus, the �rst interesting question is if the princi-

pal will ever select an optimal severance fee, � < 1, that makes the outsourcing�exibility valuable compared to the case of no outsourcing. To illustrate that this

might happen, the following example can be constructed.

<Insert table 1 here>

From table 1 it is seen that the principal uses the severance fee to limit the agent�s

incentive to invest in �rm-speci�c human capital ( � � 0:02). At �rst this might

seem counterintuitive as h is productive (eH� = 0:595 > 1=2). However, the out-

sourcing o¤er is relatively lucrative (� = 0:78) and so ex ante the principal is better

o¤ incentivizing the agent to exert physical e¤ort, a, in period 1 only. In total, the

principal strictly bene�ts from the outsourcing possibility since E��O�� 14:61 >

14:46 � E��NO

�due to the real option e¤ect, 1=2p20 � 0:15, while the agent is worse

o¤ as E�UOHH

�� 0:21 < 0:27 � E

�UNOHH

�.

3.3.2 Can outsourcing �exibility be a bad to the company under short-term contracting?

The next question in line is whether the outsourcing �exibility can ever make the

principal worse o¤. Based on the analysis in section 3.3.1 above, only a scenario

where the severance fee is capped, 2 [0; max] with 0 � max < E�UNO2

�; is of

161

interest in this respect and can make it ambiguous as to whether the outsourcing

�exibility is of value to the principal.

Intuitively, an upper bound on the severance fee might exist for several reasons

not modeled in the current setup. Perhaps the principal is interested in maintaining

some �exibility with regard to future employment, as he expects di¤erent employee

skills to be needed for future production. Indeed, this may constrain the severance

fee that the principal is willing to pay the current worker. Moreover, a too lucrative

severance fee might incentivize the agent to behave non-optimally �possibly even

destructively �in order to provoke dismissal. Consequently, an upper bound on the

severance fee is not unrealistic and will be assumed in what follows next.

If �rm-speci�c human capital is unproductive (eH� � 1=2), the principal wants

to induce hOHH = hFB = 0 which matches with the agent�s non-existing incentive to

invest in h under short-term contracting, cf. proposition 5. Therefore, the principal

will only introduce the outsourcing �exibility if the related real option e¤ect is lu-

crative. Yet, if max << E�UNO2

�, the bene�t to the agent from continued in-house

production is large, and so a situation may arise where the agent, in spite of be-

ing set at max, overinvests heavily in �rm-speci�c human capital in order to make

himself su¢ ciently attractive to the principal and reduce the risk of layo¤ (recall

(28)). Under such circumstances, the only way the principal can reduce the agent�s

incentive to invest in h is by increasing c1H �but if the agent overreacts too much

it may be disproportionately expensive to induce a su¢ ciently large reduction in

h. Put di¤erently, when �rm-speci�c human capital is unproductive, the behavioral

e¤ect (i.e. the agent�s investment in h) is always negative and so the outsourcing

�exibility is only valuable to the principal when the real option e¤ect exceeds the

value destruction caused by the behavioral e¤ect.

The same line of reasoning can be applied when �rm-speci�c human capital is pro-

ductive (eH� > 1=2) and the principal, all things being equal, wants to induce h > 0

�although matters are slightly complicated. From proposition 5 and 6, we know that

h > 0 can never be implemented given no outsourcing but can be ensured under

outsourcing. Hence, for small levels of h, the behavioral e¤ect generates value to

the principal. Yet, even when �rm-speci�c human capital is productive, a situation

may arise where the agent due to max << E�UNO2

�overinvests dramatically in h

to stay attractive, thereby destroying value. Consequently, determining the value

of the outsourcing �exibility to the principal boils down to comparing the value of

the behavioral e¤ect against the real option e¤ect. If the agent invests reasonably

162

in �rm-speci�c human capital and the value of the behavioral e¤ect is positive, the

outsourcing �exibility is per de�nition of value to the principal. If the agent over-

invests in �rm-speci�c human capital, thus destroying value to the principal, the

outsourcing �exibility is only bene�cial to the principal when the real option e¤ect

exceeds the value destruction caused by the behavioral e¤ect. This is illustrated in

�gure 5 below for a hypothetical case.

<Insert �gure 5 here>

These mechanisms can be illustrated with appropriately chosen examples, and so

we will initially exemplify that a capped can lead the agent to overinvest in h,

thereby turning the outsourcing �exibility into a bad. The focus will be kept on

the case where h is productive (eH� > 1=2) since evidence provided under these

conditions immediately implies that a similar example can be constructed when h is

unproductive (eH� � 1=2) and the behavioral e¤ect is always negative. To make theillustration more obvious, the cap on the severance fee will be chosen so max = 0:2.

<Insert table 2 here>

Table 2 demonstrates that introducing the outsourcing �exibility is advantageous to

the principal when � is not constrained. Under such circumstances, the principal

will choose � = 0:8 and realize a pro�t of E��O�� 12:41 > 11:91 � E

��NO

�, since

the agent�s investment in �rm-speci�c human capital is reduced to hOHH = 0. Again,

this large reduction in h is due to the outsourcing o¤er being relatively lucrative

(� = 0:8), for which reason the principal prefers to incentivize the agent to spend

time on physical productive e¤ort in period 1 only. However, if � is constrained

by max = 0:2, the principal cannot su¢ ciently reduce the agent�s overinvestment

in �rm-speci�c human capital (hOHH � 0:43) through the period 1 compensation, inspite of this being much higher than under no outsourcing , cO1H � 4:63 > 2:88 �cNO1H . Consequently, the overinvestment causes a reduction in expected pro�t un-

der outsourcing compared to no outsourcing, E��O�� 10:63 < 11:91 � E

��NO

�.

Hence, table 2 illustrates that the outsourcing �exibility need not be bene�cial to

the principal when the severance fee is bounded from above.

To show that the possibility of outsourcing can also be a good to the principal

even when the severance fee is capped, we only need to set the cap arbitrarily close

yet below 0.8. Thus, assume that max = 0:6 and keep the rest of the parameters

from the example above unchanged.

163

<Insert table 3 here>

Comparing table 3 to 2 demonstrates that max = 0:6 makes the outsourcing �exi-

bility a good to the principal since E��O�� 12:35 > 11:91 � E

��NO

�. The reason

is that the relatively high bound on the severance fee makes it possible for the prin-

cipal to reduce the agent�s overinvestment in h to a reasonable level by increasing

cO1H � 3:12 > 2:88 � cNO1H .

Conclusively, it is ambiguous whether the outsourcing �exibility makes the prin-

cipal better o¤ when the severance fee is bounded from above and the principal

cannot commit to the second period contract. However, what is essential here is

that the outsourcing �exibility can in fact turn out to be a bad to the principal even

when �rm-speci�c human capital is productive. This will happen when the agent

overinvests in h and causes a value destruction that exceeds the real option e¤ect of

the outsourcing �exibility.

3.3.3 Can outsourcing �exibility create a complete win-win situation forthe company and the worker?

Having clari�ed the potential value creation as well as value destruction of the

outsourcing �exibility from the principal�s perspective, we will in closing take a

closer look at the agent�s situation. What the examples provided in table 1, 2, and

3 have in common is that the agent is worse o¤ in all cases from the introduction of

the outsourcing alternative. Thus, the interesting question is whether introducing

an outsourcing alternative can create a complete win-win situation for the principal

and the agent. To answer this question, the following example can be constructed.

<Insert table 4 here>

Table 4 demonstrates how the outsourcing �exibility ex ante makes the principal as

well as the agent better o¤. To explain the intuition and dynamics, let us decompose

the expected utilities using the parameter values from table 4.

E�UNOHH

�� 0:500| {z }

e1=2H

c1H��H

+ 0:500| {z }e1=2H

c2H��H

� 1 (30)

E�UOHH

�� 0:507| {z }

(eH�hOHH)

1=2c1H��H

+ 0|{z}p0

+ 0:500| {z }(1�p0)[e1=2H (1+�hOHH)c2H��H]

� 1:07:

(31)

164

From (30) it can initially be observed that E�UNO2

�� 0:5 is greater than the opti-

mal severance fee � � 0:39, for which reason the agent is incentivized to invest inhOHH � 0:06. Consequently, it is seen from (30) and (31) that even though the agentinvests in �rm-speci�c human capital, and thus lowers the probability of realizing

the high outcome, xH = 10, the expected rent in period 1 is still increased under

outsourcing compared to no outsourcing. The reason is that the principal, in order

to avoid/limit the agent�s incentive to overinvest in h, increases the compensation

under outsourcing compared to no outsourcing, cO1H � 1:65 > 1:58 � cNO1H . This

means that in spite of the positive investment in �rm-speci�c human capital, the

expected utility in period 1 increases under outsourcing. Furthermore, given the

fact that the principal cannot commit to the contract in period 2, the investment

in h leads the principal to choose cO2H � 1:50 < 1:58 � cNO2H , so the agent expectedlywill not bene�t from his increased productivity in period 2. Consequently, the agent

will receive the same expected utility in period 2 as under no outsourcing. In total,

this implies that he is better o¤under outsourcing, E�UOHH

�� 1:07 > 1 � E

�UNOHH

�.

Next, breaking up the principal�s expected pro�t in each scenario provides the fol-

lowing insights

E��NO

�� 7:99|{z}

e1=2H (xH�c1H)

+ 7:99|{z}e1=2H (xH�c2H)

� 15:98 (32)

E��O�� 7:64|{z}

(eH�h)1=2(xH�c1H)

+ 8:50|{z}e1=2H

(1+�h)(xH�c2H)

+ 0|{z}12 p20

� 16:14: (33)

Calculation (32) and (33) clarify that the increased compensation in period 1, cO1H ,

together with the reduction in the probability of the high outcome, reduces the ex-

pected pro�t in period 1 under outsourcing compared to no outsourcing. Yet, at the

same time the increased investment in h increases the probability of realizing the

high outcome in period 2 when production is kept in-house, which in combination

with a lower period 2 compensation, cO2H , increases the pro�t from continued in-house

production under outsourcing compared to no outsourcing. Adding to this that the

probability of outsourcing is 0 implies that the expected pro�t in total increases

when the outsourcing alternative is introduced, E��O�� 16:14 > 15:98 � E

��NO

�.

Therefore, this example clearly demonstrates that scenarios do exist where the out-

sourcing �exibility can create an ex ante win-win situation for both the principal and

the agent altogether. This is interesting since outsourcing is normally considered as

a bene�t to the company only, while rarely regarded as potentially valuable to the

165

employee as well. However, it is important to stress that the value to the agent only

exists expectedly and so what seems lucrative ex ante, might ex post turn out to be

a bad if layo¤ is realized.

4 Conclusion

Shapiro and Stiglitz (1984) show how the fear of layo¤ (unemployment) can dis-

cipline employees not to shirk. The present article has elaborated on this idea by

speci�cally focusing on employee incentives to invest in �rm-speci�c human capital

when threatened by layo¤ (outsourcing). However, contrary to Shapiro and Stiglitz

(1984), the rate of unemployment is not the model�s essential disciplining device for

correcting behavior. What drives the incentive is instead the interdependency be-

tween the �rm-speci�city of the employee�s investment and the risk of layo¤, and so

the model also bears some resemblance to the literature on career concerns. Speci�-

cally, on the one hand, an investment in �rm-speci�c human capital increases future

in-house productivity and makes the employee more attractive to the company since

the investment is observable. On the other hand, the inherent �rm-speci�city of the

investment leaves the employee more exposed to layo¤ as the value of alternative

uses is minimal. The analysis of this incentive mechanism has been conducted both

under full commitment and short-term contracting; however, the main results are

related to the latter.

Most importantly, it has been shown that a company forced to operate under short-

term contracting will always exploit the �rm-speci�city of the employee�s compe-

tence investment and enforce a salary cut ex post the improvement. Knowing this

in advance, the worker does not upgrade his skills, which reduces expected company

pro�tability if �rm-speci�c human capital is productive. To overcome this problem,

the idea of the company introducing an outsourcing alternative has been proposed.

Contrary to the obvious assumption that the threat of outsourcing only makes the

employee more reluctant to carry out long-term actions, it has been shown how the

threat of layo¤ can in fact help trigger competence improvements. That is, the risk

of discharge makes the employee concerned with realizing future rents for which

reason he invests in �rm-speci�c human capital to stay attractive to the company

(career concerns).

More generally, the analysis has clari�ed how outsourcing comprises two potentially

value-creating e¤ects to the company; a real option e¤ect that increases with the

166

attractiveness of the outsourcing alternative and a behavioral e¤ect that mimics the

changes in employee behavior caused by the risk of layo¤. Contrary to short-term

contracting, outsourcing �exibility is not needed under full commitment to incen-

tivize learning since incentives can be regulated through commitment to a certain

compensation ex post the investment. Still, based on comparative statics it has

been explained how outsourcing may be bene�cial for the company to bring about

even under full commitment, partly due to the real option e¤ect, but also due to

the behavioral e¤ect which reduces the cost of inducing �rm-speci�c investments.

In relation to this, it has been described how increasing the severance fee under full

commitment can help further the employee�s incentive to undertake learning when

appropriately combined with a compensation ex post the �rm-speci�c investment.

The reason is that the severance fee moderates the probability of outsourcing while

a higher compensation ex post the investment makes continued in-house production

lucrative to the worker. Under short-term contracting the lack of contract com-

mitment destroys the moderating e¤ect of the severance fee and induces the worker

to reduce his investment in �rm-speci�c human capital as the severance fee increases.

In practice, and perhaps in particular for companies operating under short-term

contracting, the idea of employing outsourcing as an incentive mechanism to ease or

trigger �rm-speci�c learning inside the company may thus prove useful. Even when

employees are hired continuously on short-term contracts, developing �rm-speci�c

skills may be important for company pro�tability. As such, the model shows that

if outsourcing �exibility is managed properly, the use of short-term contracting is

not necessarily restricted to jobs requiring low-skilled labor but can in fact be ap-

plied to high-skilled work as well. The real beauty of this is that the competence

improvement is private to the employee and comes at no extra cost to the company,

since it only hinges on the introduction, but not necessarily the use, of an out-

sourcing alternative. Actually, not even the employee is automatically harmed by

the threat of layo¤. In fact, the analysis illustrates that outsourcing �exibility can

create an ex ante win-win situation where both parties bene�t from the introduction.

Also, considering the discussion in recent years on privatization of tasks performed

by public companies operating in monopoly-like markets characterized by long-term

contracts, this article o¤ers an interesting re�ection. Proponents of privatization

usually substantiate their arguments with the need to minimize the bureaucracy

and ine¢ ciency believed to naturally hamper public companies (Hatry 1985; Mor-

gan and England 1988); in the setup proposed in this article, this could be modeled

167

as being too costly to provide the employee with su¢ cient incentive to undertake the

necessary investment in �rm-speci�c human capital through the ex post investment

compensation. Yet, automatically favoring privatization, and thus simply seeking

to replace public companies with private alternatives, may overlook the potential

improvement in e¢ ciency that comes about when public employees change their be-

havior and invest in �rm-speci�c human capital to meet the threat of replacement.

Instead, the analysis suggests that it may be advantageous to put tasks out to tender

so that public and private companies can compete on equal terms.

Having argued for the potential bene�ts of introducing a third-party supplier, how-

ever, it should be stressed that this is not in all instances the solution to problems

regarding long-term actions, and so companies need to apply the principle with

caution. Speci�cally, it has been illustrated that a severance fee encumbered by

an upper bound can turn outsourcing �exibility into a bad for the company if the

employee overinvests signi�cantly in �rm-speci�c human capital in order to stay

attractive and avoid layo¤. Put di¤erently, when the behavioral e¤ect is highly neg-

ative and dominates the real option e¤ect, the outsourcing alternative will harm the

company.

On a �nal note, the modeling of investments in �rm-speci�c human capital provides

an interesting insight in relation to the continuing debate on fair value accounting

(mark-to-market valuation) versus historical cost accounting (original cost) and the

trade-o¤ between credibility and relevance (Bleck and Liu 2007). To brie�y explain

this, let us consider the valuation of an ongoing project that requires no initial mone-

tary investment by the company but only private investments in �rm-speci�c human

capital by the employees. Under fair value accounting, the valuation of the project,

and thus the accounting information reported, will come to re�ect the employee in-

vestment in skill improvement, since the expected cash �ow to be discounted re�ects

the investments of �rm-speci�c human capital through the probabilities. That is,

the accounting information provided under fair value accounting (partly) reveals the

company�s pool of talent. Contrary to this, only the explicit monetary investment of

zero will be re�ected in the accounting information reported under historical cost ac-

counting, which in consequence implies that the company�s pool of talent will never

be revealed using this regime. Hence, if expected cash �ows are highly responsive

to employee investments in �rm-speci�c human capital, this increases the need for

relevance and points towards the use of fair value accounting.

168

5 Further research

In relation to the setup proposed in this article, several aspects could be studied in

further detail. First, the assumption about the agent�s investment in �rm-speci�c

human capital being explicitly observable may be relaxed or changed. In keeping

with the traditional literature on career concerns, this could instead be indirectly

re�ected through the output level to study how it a¤ects the agent incentive to

invest. Second, the model presented here is restricted to only two periods. Indeed,

it could be interesting to generalize the results on the basis of a multi-period model

in which the agent, if laid o¤, can be rehired in later periods. Third, the concern of

this paper has been with the agent�s incentive to undertake �rm-speci�c learning.

However, at a more general level the article is about inducing long-term actions.

The mechanisms analyzed could thus be applied to many other settings in which

the value of the current actions taken might neither be realized nor known until

much later. Finally, on the empirical side, additional case studies will have to reveal

how e¤ective the threat of layo¤ in reality is in inducing employees to undertake

long-term actions.

Acknowledgements

The author would like to thank Professor Hans Frimor, School of Economics and

Management, Aarhus University, for his valuable criticisms and suggestions. Further,

the comments received from the participants at the European Accounting Associa-

tion 33rd Annual Congress 2010 have been helpful. Any mistakes are entirely the

author�s.

References

Anderhub, V., M. Königstein, and D. Kübler (2003, August). Long-term work con-tracts versus sequential spot markets: experimental evidence on �rm-speci�cinvestment. Labour Economics 10 (4), 407�425.

Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journalof Management 17 (1), 99�120.

Becker, G. S. (1964). Human Capital: A Theoretical and Empirical Analysis,with Special Reference to Education. New York: National Bureau of EconomicResearch.

Bleck, A. and X. Liu (2007, May). Market transparency and the accountingregime. Journal of Accounting Research 45 (2), 229�256.

169

Calvo, G. (1979, May). Quasi-walrasian theories of unemployment. The AmericanEconomic Review 69 (2), 102�107.

Dewatripont, M., I. Jewitt, and J. Tirole (1999a, January). The economics ofcareer concerns, part i: Comparing information structures. The Review ofEconomic Studies 66 (1), 183�198.

Dewatripont, M., I. Jewitt, and J. Tirole (1999b, January). The economics ofcareer concerns, part ii: Application to missions and accountability of govern-ment agencies. The Review of Economic Studies 66 (1), 199�217.

Edvinsson, L. and P. Sullivan (1996, August). Developing a model for managingintellectual capital. European Management Journal 14 (4), 356�364.

Gibbons, R. and K. J. Murphy (1992, June). Optimal incentive contracts in thepresence of career concerns: Theory and evidence. The Journal of PoliticalEconomy 100 (3), 468�506.

Hatry, H. P. (1985, September). Privatization pros and cons and ways to makegovernment agencies more competitive. Urban Resources 4, 15�17.

Holmström, B. (1999, January). Managerial incentive problems: A dynamic per-spective. The Review of Economic Studies 66 (1), 169�182.

Kogut, B. and U. Zander (1993). Knowledge of the �rm and the evolutionarytheory of the multinational corporation. Journal of International BusinessStudies 24 (4), 625�646.

Milgrom, P. and J. Roberts (1992). Economics, Organization & Management.Prentice Hall, International Edition.

Morgan, D. R. and R. E. England (1988, November - December). The two facesof privatization. Public Administration Review 48 (6), 979�987.

Nonaka, I. and H. Takeuchi (1994). The Knowledge Creating Company: HowJapanese Companies Create the Dynamics of Innovation. Oxford: Oxford Uni-versity Press.

Peteraf, M. A. (1993, March). The cornerstones of competitive advantage: Aresource-based view. Strategic Management Journal, 14 (3), 179�191.

Prahalad, C. K. and G. Hamel (1990, May - June). The core competence of thecorporation. Harvard Business Review 68 (3), 79�92.

Shapiro, C. and J. E. Stiglitz (1984, June). Equilibrium unemployment as a workerdiscipline device. The American Economic Review 74 (3), 433�444.

Spender, J. C. and R. M. Grant (1996). Knowledge and the �rm: Overview.Strategic Management Journal 17 (Special Issue: Knowledge and the Firm(Winter)), 5�9.

Wernerfelt, B. (1984, April - June). A resource-based view of the �rm. StrategicManagement Journal 5 (2), 171�180.

170

• A

gent

cho

oses

1ean

d h

•

Long

-term

con

tract

s ar

e es

tabl

ishe

d,

()(

)H

LH

Lc

cc

c2

21

1,

,,

•

Out

put

1x is

real

ized

•

1cis

pai

d ou

t •

h is

obs

erve

d by

the

prin

cipa

l •

Age

nt c

hoos

es

2e

• O

utpu

t 2xis

real

ized

•

2cis

pai

d ou

t

0=t

1

=t

2=t

Figure1:Sequenceofeventsinthebasicmodel.

171

• A

gent

cho

oses

1ean

d h

•

γ is

det

erm

ined

•

Dis

tribu

tion

of p

rice

offe

r is

kno

wn

() 1,0

~~

Up

•

Long

-term

con

tract

s are

es

tabl

ishe

d,

()(

)H

LH

Lc

cc

c2

21

1,0

,,0

==

•

Out

put

1x is

real

ized

•

1cis

pai

d ou

t •

h is

obs

erve

d by

the

prin

cipa

l •

Pric

e of

fer f

rom

third

par

ty

cont

ract

orp

is o

bser

ved

• Th

e pr

inci

pal d

ecid

es w

heth

er

to o

utso

urce

or n

ot

• O

utso

urci

ng:γ

is p

aid

out

• N

o ou

tsou

rcin

g: a

gent

cho

oses

2e

• O

utpu

t 2xis

re

aliz

ed

• O

utso

urci

ng:

pis

pai

d ou

t •

No

outs

ourc

ing:

2c

is p

aid

out

0=t

1

=t

2=t

Figure2:Sequenceofeventsunderfullcommitmentwhenanoutsourcingalternativeexists.

172

ONO EE 22 / ππ

h 0p ph /

( )pE O2π

p p

( )hE NO2π

Figure 3: Illustration of how the probability of outsourcing, p0, is decided on thebasis of E[�NO2 ] and E[�O2 ]. For a given h a p0 exists so E[�

NO2 ] = E[�O2 ]. Hence, for

p < p0 observed, the principal will choose outsourcing since E�O2 (p) > E�NO2 (h).

Contrarily, for p � p0 observed, the principal will choose in-house production sinceE�O2 (p) � E�NO2 (h).

173

• A

gent

cho

oses

1ean

d h

•

γ is

det

erm

ined

•

Dis

tribu

tion

of p

rice

offe

r is k

now

n (

) 1,0~

~U

p

• Sh

ort-t

erm

per

iod

1 co

ntra

ct is

est

ablis

hed,

(

)H

Lc

c1

1,

•

Out

put

1x is

real

ized

•

1cis

pai

d ou

t •

h is

obs

erve

d by

the

prin

cipa

l •

Pric

e of

fer f

rom

third

par

ty

cont

ract

orp

is o

bser

ved

• Pr

inci

pal d

ecid

es w

heth

er to

ou

tsou

rce

or n

ot (c

ontra

ct

rene

gotia

tion)

•

Out

sour

cing

: γ is

pai

d ou

t •

No

outs

ourc

ing:

o

Sho

rt-te

rm p

erio

d 2

cont

ract

is

est

ablis

hed,

()

HL

cc

22

,

o A

gent

cho

oses

2e

• O

utpu

t 2xis

re

aliz

ed

• O

utso

urci

ng:

pis

pai

d ou

t •

No

outs

ourc

ing:

2c

is p

aid

out

0=t

1

=t

2=t

Figure4:Sequenceofeventsundershort-term

contracting.

174

h h 0

= 0

NO

Eπ

OEπ

Real option effect

πE

h 1h

h 0 =

0

NO

Eπ

OEπ

Real option effect

πE

h 1

Behavioral effect

Behavioral effect

Figure5:Illustrationsofthebehaviorale¤ectandtherealoptione¤ectwhen iscappedandhisproductive.Left:Outsourcing

�exibilityisagood.Right:Outsourcing�exibilityisabad.

175

Outsourcing

�=0:021674

cO 1H

cO 2H

�H

xH

xL

eH

eL

��

hO HH

0p 0� hO HH

� 0:547474

1:357478

1:357478

110

00:70:01

0:85

0:78

hO LH

0p 0� hO LH

� 0:547474

E� UO HH

� 0:209043@p 0=@h

�7:111610

E� �O�

=14:611569

E� UO LH

� 0:209043e H�

0:595

NoOutsourcing

hNO

HH

cNO

1H

cNO

2H

�H

xH

xL

e He L

�E� �NO

� E� UNO HH

�0

1:357478

1:357478

110

00:7

0:01

0:85

14:4617050:271500

Table1:Theprincipalchoosesa �<1tomaximisehispro�tunderoutsourcingcomparedtonooutsourcing.

176

Outsourcing

��0:2= max

cO 1H

cO 2H

�H

xH

xL

e He L

��

hO HH

0:430741

p 0� hO HH

� 0:211438

4:632998

2:178456

1:510

00:70:1

0:75

0:8

hO LH

0p 0� hO LH

� 1E� UO HH

� 1:665083@p 0=@h

�6:274950

E� �O�

=10:628842

E� UO LH

� 1:665083e H�

0:525

�=0:8>0:2= max

cO 1H

cO 2H

�H

xH

xL

e He L

��

hO HH

0p 0� hO HH

� 12:882219

2:882219

1:510

00:70

0:75

0:8

hO LH

0p 0� hO LH

� 1E� UO HH

� 1:711438@p 0=@h

�6:274950

E� �O�

=12:410325

E� UO LH

� 1:711438e H�

0:525

NoOutsourcing

hNO

HH

cNO

1H

cNO

2H

�H

xH

xL

e He L

�E� �NO

� E� UNO HH

�0

2:882219

2:882219

1:5

100

0:70:10:75

11:9103251:822876

Table2:Acapped leadstheagenttooverinvestinhandturnsoutsourcing�exibilityintoabad.

177

Outsourcing

��0:6= max

cO 1H

cO 2H

�H

xH

xL

e He L

��

hO HH

0:064040

p 0� hO HH

� 13:116927

2:750131

1:510

00:70:10:75

0:8

hO LH

0p 0� hO LH

� 1E� UO HH

� 1:585659@p 0=@h

�6:274950

E� �O�

=12:346062

E� UO LH

� 1:585659e H�

0:525

Table3:Acapped leadstheagenttoinvestreasonablyinhandturnsoutsourcing�exibilityintoagood.

178

Outsourcing

�=0:388260

cO 1H

cO 2H

�H

xH

xL

e He L

��

hO HH

0:063638

p 0� hO HH

� 01:648073

1:500000

110

00:90:1

0:85

0:85

hO LH

0p 0� hO LH

� 0:124907

E� UO HH

� 1:007209@p 0=@h

�8:063808

E� �O�

=16:138072

E� UO LH

� 1:007209e H�

0:765

NoOutsourcing

hNO

HH

cNO

1H

cNO

2H

�H

xH

xL

e He L

�E� �NO

� E� UNO HH

�0

1:581139

1:581139

110

00:90:10:85

15:9736661:000000

Table4:Boththeagentandtheprincipalbene�texantefrom

introducingoutsourcing�exibility.

179