Government Contracting Process

7/26/2019 Government Contracting Process

1/21

Journal of Public Administration Research and Theory, Inc., Public Management Research Association and

Oxford University Press are collaborating with JSTOR to digitize, preserve and extend access to Journal of Public

Administration Research and Theory: J-PART.

http://www.jstor.org

Journal of Public Administration Research and Theory Inc.

Public Management Research Association

Privatization and Its Reverse: Explaining the Dynamics of the Government Contracting ProcessAuthor(s): Amir Hefetz and Mildred WarnerSource: Journal of Public Administration Research and Theory: J-PART, Vol. 14, No. 2 (Apr.,2004), pp. 171-190Published by: on behalf of theOxford University Press Public Management Research Association

Stable URL: http://www.jstor.org/stable/3525868Accessed: 16-12-2015 00:25 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of contentin a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.For more information about JSTOR, please contact [email protected].

This content downloaded from 210.187.26.2 on Wed, 16 Dec 2015 00:25:03 UTCAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/http://www.jstor.org/action/showPublisher?publisherCode=ouphttp://www.jstor.org/action/showPublisher?publisherCode=pmrahttp://www.jstor.org/stable/3525868http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3525868http://www.jstor.org/action/showPublisher?publisherCode=pmrahttp://www.jstor.org/action/showPublisher?publisherCode=ouphttp://www.jstor.org/


2/21

Privatization

n d

I t s

R e v e r s e

Explaining

t h

Dynamic s

o f t h overnment

ontract ing

P r o c e s s

Amir

Hefetz

Technion,

srael nstitute

f

Technology

Mildred

Warner

Cornell

University

ABSTRACT

Empirical

vidence hows ocal

governmentontracting

s a

dynamic

rocess

hat ncludes

movements rom

public

delivery

o markets nd frommarket ontracts ack o in-house

delivery.

his

reverse

ontracting

eflectshe

complexity

f

public

ervice

rovision

na

world

wheremarket lternativesreused

along

with

public elivery.

We

develop

methodology

o

link

responses

o

national

urveys

and createa

longitudinal

ata set that

captures

he

dynamics

f

the

contracting

rocess.

We

present

a framework

hat

ncorporatesrincipal

agent

problems,

overnment

management,monitoring

nd citizen

oncerns,

nd

market

structure. ur tatistical

nalysis

inds

governmentmanagement,

monitoring,

nd

principal

agentproblems

o be most

mportant

n

explaining

othnew

contracting

utand

contracting

back-in. rofessional

anagers

ecognize

he

importance

f

monitoring

nd

the need

for

public

ngagement

n he

service

elivery rocess.

heresults

upport

he new

public

ervice

that

argues

public

managers

o more hansteera market

rocess; hey

balance echnical

and

political

oncernso secure

public

alue.

INTRODUCTION

The new

public

management

ncourages overnments

o be

moreefficientand

responsive

by employingmarketapproacheso publicservicedelivery(EggersandO'Leary1995;

Hood

2002;

Kettl

1997;

Osbore and

Gaebler

1992).

Recent

critiques

of the new

public

management

hallenge

he notion hatmarket

ased

approaches

re

adequate

o

capture

he

full set of

political

considerationshatare

part

of local

government

ervice

delivery Box

1999;

Denhardt ndDenhardt

003;

Kaboolian

1998;

Kelly

1998;

Moore

1995;

O'Looney

1993;

Stark

2002).

Privatization nd market

heoriesof

public

management

re driven

by

Funding

for this research

was

providedby

the U.S.

Department

of

Agriculture

National Research

Initiative,

Affinito

Stewart Grants

Program,

Corell

University;

the U.S.

Department

of

Agriculture

Hatch

Program;

nd the

Economic

Policy

Institute.Amir Hefetz was a

PrivatizationFellow with

EPI,

based

at

Cornell

University.

He

is now a Ph.D.

studentat Technion n

Haifa,

Israel. Mildred

Warner

s an

Associate

Professorof

City

&

Regional Planning

at

Cornell

University.

DOI:

10.1093/jopart/muh012

Journal

of

Public Administration

Research and

Theory,

Vol.

14,

no.

2,

pp.

171-190

?

2004 Journalof Public Administration

Research and

Theory,

Inc.;

all

rights

reserved.

http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/21

172

Journalof Public

Administration

Research

and

Theory

government

failure

to

meet economic and

organizational

efficiency

(Niskanen

1971;

Savas

2000).

However,

the

need

for

public

goods

provision

arises from

market

failure,

and

contract markets exhibit

many

of these

same failures

(de

Leon and

Denhardt

2000;

Lowery

1998). A theoretical framework is needed which gets beyond the market versus

government

failure

dichotomy

and

explores

the full

range

of

components

in the

public

service

delivery

decision

(Zebre

and

McCurdy

1999).

Some scholars have used transac-

tion cost

approaches

to

compare

the costs of direct

public

delivery

with the

contracting

process (Alexander

2001;

Brown and Potoski

2003;

Nelson

1997).

We

present

a

broader

framework that

also includes

attention to the

uniqueness

of citizen

interaction

and

pro-

fessional

management

in

public

service

provision.

Pragmatic

local

government

managers

use markets

in

a

dynamic approach; they

contract out and

bring

unsuccessful contracts back in-house for

direct

public provision

(Warner

and Hebdon

2001; Warner, Ballard,

and

Hefetz

2003).

They

also mix

public

provision

and

private

contracts

for

the same service

(Miranda

andLerner

1995).

Even with

careful

attention to

monitoring,

local

government experience

with

contracting

has

not

always

been

positive

(Adler

1999;

Sclar

2000).

Local

government practice

has

always

involved

a mix of

public

provision

and

government contracting

(Sonenblum,

Kirlin,

and

Reis

1977).

However,

this behavior was

reinterpreted

as

privatization

in

the 1980s as

part

of a broader

push

for market

approaches

to

government

service

delivery (Henig

1990).

This enthusiasm

for market

solutions

was reflected

in

new data series. The

International

City County Management

Association

(ICMA)

in

1982

and the Census of

Governments

in 1987

began quinquennial surveys

of local

government

contracting,

and

these

remain the

primary

sources

of

national

data. Because the

presumption

of

government

failure was so

strong,

and the

political support

for

market solutions so

widespread,

these

surveys

do

not

attempt

to

capture

the

dynamic

movement between

market and

government

provision typical

of

local

government practice.

Rather,

they

focus on

contracting

as a

one-

way

street,

asking

only

which

services

have been contracted

out.

In

this article we

present

a

methodology

that links

responses

to the

ICMA

surveys

to

create

a

longitudinal

data

set

that enables

us to address the

dynamics

of

the

contracting

process.

We find that from

1992 to

1997,

93

percent

of

the

628

governments

that

responded

in both

survey years newly

contracted out

at least one

service,

and 81

percent

of these

governments

contracted back-in at least

one

service.

Almost

three-quarters

of

governments

engaged

in

both new

contracting

out and

contracting

back-in.

These

findings challenge

the

adequacy of a singular focus on contractingout. On average, governments newly contract

out

six services and contract back-in

four services. That

contracting

back-in would

be so

common-over

half

the

level of

new

contracting out-requires

that researchers

give

attention

to the

dynamics

of

the

contracting process.'

Private

firms that contract out balance concerns

with

efficiency, quality, timing,

reliability, security,

and

internal

capacity

(Nelson

1997).

The new

public

service literature

argues

local

government managers

who use market

approaches

to deliver services must

balance

an even

wider set of

concerns,

including

accountability

and

public

preference

(Blanchard,

Hinnant,

and

Wong

1998;

Feldman

and Khademian

2001;

Sanderson

1998;

Stirton

and

Lodge

2001).

1 For

our

study,

only

services once

provided by government

are

included

in

our definitionof

contracting

back-in.

This could include

contracting

back-in from

nonprofit,

or

profit,

or

other

government

providers.

In their definitionof

contracting-in,

Welch and Bretschneider

1999)

include new

contracting

out such as

intermunicipal ooperation

and

governmentalentrepreneurship.

Our

study distinguishes

new

contracting

out from

contracting

back-in.



4/21

Hefetzand

Warner

Privatization

nd

Its Reverse

What

is

Contracting

Back-in?

The

debate over

privatization

s

shifting

from

an

ideological

debate

(Savas 2000)

to

a

proper

discussion of

the

political,

economic,

and

community

actors that determine

choice of market rgovernment elivery FeigenbaumndHenig1994;Kirkpatrick999;

Lowery

1999;

Sclar

2000).

The same

concernswith

efficiency,quality,

and nnovation hat

motivate

privatization,may promote

contracting

back-in

(Henig

1990;

Warner and

Hebdon

2001).

Contracting

ack-incan

be

several

hings.Typically

t

is assumed o be

competitive

bidding,

where

ocal

employees

contract o

bring

he workback

(Martin1999).

However,

in

an

analysis

of

twenty-six

case

studies of reverse

contracting,

ompetitivebidding

explained

less than half of the cases

(Ballard

and

Warner

2000).

In

all

of the

cases,

disappointment

with

service

quality

or difficulties with

contract

specification

and

monitoring

were

factors

driving

he

decisionto

bring

he contract

back

in-house.

In

most

of the cases, internalprocess improvementby labor managementcooperationwas

associated

with

contracting

ack-in.

Contracting

back-in

may

reflect market success where

competition

increases

efficiency (Lavery

1999).

In

other

cases

it

reflects failure of markets o

meet desired

outcomes

or failure of

government

o

adequatelymanage

and monitorcontracts

Sclar

2000).

It also

may

represent

new form of

partnership

etween

public

and

private

actors

(Freeman

000;

Rhodes

1996)

where

both

cooperate

n

service

provision.

The

debate on

privatization

as been

highly

ideological,

relying primarily

on

case

studies

where

proponents

ind

cost

saving

efficiencies

(Eggers

and

O'Leary

1995;

Savas

2000)

and detractors

ind cost

overruns,

orruption,

nd erosion

n

wages

and

in

citizen

voice

(Hebdon

1995;

Sclar

2000;

Starr

1988).

Although empiricalanalyses

find

only

limited

support

or

market

advantages

Boyne

1998;

Ferris

1986;

Hirsch

1995;

Milne

1997),

the debateremainsdominated

y public

choice

theory

andhas

rarely

addressed he

dynamic

and mixed

(public

and

private)

natureof

local

government

ervice

delivery

(Boyne 2002).

Local

government

ontracting

has been

treatedas a one

way process-

towardmarkets-and most studieshave

only

lookedat

contracting

ut

(Boyne

1998;

Ferris

1986;

Hirsch

1995).

Those studies that have

acknowledged

he

possibility

of reverse

privatization

r

contracts

eing

brought

ack n-house

Kodryzcki

998;

Lopez

de

Silanes,

Shleifer,

and

Vishny

1997)

havenot

given

it

detailed

attention.

Warner

nd

Hebdon

2001)

are the onlyones to empiricallymeasurecontracting ack-indirectly,as another orm of

service

provision,

but their

analysis

s

limited

to

New

York

state.

Reverse

contracting

demonstrates he

complexity

of

service

provision

in

a world

where marketalternativesare used

along

with

public delivery.

It

may

involve market

failure,

government

ailure,

or both.

To

gain

a

better

understanding

f the

contracting

processrequires

an

analysis

of both

contracting

ut and

contracting

ack-in,

which

gives

attention o

economic,

managerial,

nd

politicalperspectives.

A

Framework or

Understanding

he

Complexity

of

Government

Service

Delivery

Both theories of

government

ailureand market ailure

rely

on the market

paradigm

o

definethe

necessary

evel

of

government

ntervention.

ublicchoice

theory

argues

iberal

policies

that allow market-like

olutions

may

replace

central

planning

and

improve

he

efficiency

of

the

political process (Buchanan

nd Tullock

1974). Throughprivatization,

173



5/21

174

Journalof Public

Administration

Researchand

Theory

private

irms

compete

for

public

service

delivery

which

may

lead to

cost

savings,higher

efficiency,

and less

government

nvolvement

Domberger

nd Jensen

1997;

Savas

2000).

Critics of

public

choice

challenge

the

separation

of

provision

from

production,

he

methodologicalndividualism f aggregatingndividual onsumerpreferenceso achieve

the collective

public good,

and

competition

as

a

substitute

or

planning

n

local

public

service markets

Lowery

1998,

1999;

Starr

1988).

The social

values

inherent

n

public

services

may

not be

adequately

ddressed

y

the

economic

efficiency

calculusof markets

Kelly

1998;

Starr

1988).

The

consumer otion

of

citizenship

does not

adequately

ddress

he

broader

ocial

concerns

of the citizen

(deLeon

and Denhardt

2000).

Likewise,

governments

are

more than a

business;

they

reflect

collective

identity,

espond

o

diversity,

and

promote

ocial

equity

Box 1999).

Increasing

attention

is

being given

to the

intrinsic value

of

interactionbetween citizens and

government

n

the

public

service

delivery

process

to

promotedemocracy,community

building,and a moresocially equitable ystemof urban erviceprovision Denhardt nd

Denhardt

000;

Frug

1998;

Marmolo

1998;

Potapchuck,

Crocker,

nd Schechter

1998).

To

provide

a better

understanding

f

the

dynamicprocess

of

government

ervice

delivery restructuring,

we

present

a

framework that

includes attention to both

organizational

ehaviorand market tructure.

Principal Agent

Problems

Transaction cost

theory

in

public organizations

combines both

individual

and

organizational

ehavior o address

principal

agent

problems

n

government

rganization

(Williamson1996).

Goal

incongruence

etween

governments

nd their contractors

may

reduce

privatization

r increase

contracting

ack-in.Political

pressure

rom

government

employees,

unions,

and

budget-maximizing

ureaucrats lso

may

limit

private

contracts

(Savas

2000;

Siegel 1999).

However,

empirical

tudieshave found

conflicting

videnceof

the

impact

of labor

opposition.

n

some cases it is shown o reduce

privatization

Chandler

and

Feuille

1991;

Ferris

1986;

Hirsch

and

Osbore

2000);

in

others

t has

limited

or

no

effect

(O'Brien

1994;

Warner

nd Hebdon

2001).

Recent

studies

of

public

managers

ave

challenged

he self-interest laims of

public

choice

theory

and

emphasized

professional

motivation

or

public

service

(DiIulio

1994;

Francois

2000;

Moore

1995).

Management

More

professional ity

management,

s

found

n

the council

manager

orm

of

government,

may

minimize the

effect

of

politics

on

local decisions and lead to a

higher

level of

privatization

Feiock

and

Kim

2000;

Moon and deLeon

2001).

The

government

manager

must

decidewhether

o

makeor

buy

a

service

subject

o market

onditions,

nternal

res-

sure,

and state restrictions n

debt

limit

and taxes

(Hirsch

1995, 1970;

Lopez-de-Silanes,

Shleifer,

and

Vishny

1997;

Nelson

1997).

Political climate also

may

affect

managers'

decisions

Savas2001).

Politics

s

part

of

public

management,

nd

good

managers

o not

ust

make

technicaldecisions.

They

interact

n

the

political process (Nalbandian

1999)

and

facilitate

publicengagement Feldman

ndKhademian

001;

Svara

1998).

Market

Structure

Theoretically ompetition

s

the

key

to

ensuring

avings

from

privatization

Boyne

1998;

Savas

2000).

Publicchoice

theory mplies

competition

will

be

greatest

n

fragmented

ocal



6/21

Hefetzand

Warner Privatization nd Its Reverse

government

ettings

(Tiebout 1956).

Lack of

alternative

uppliers

has been

shown

to be

a barriero

privatization

n

both nner

cities

(Hirsch1995)

andrural reas

Kodrzycki

994;

Warnerand

Hefetz

2003).

In

order o create

competition,

ocal

governments

ngage

in

mixedpublicandprivatedelivery nthe same servicearea Miranda ndLerner1995)and

competitivebidding

(Martin1999).

These movements rom

public delivery

to contracts

andfrom

contracts ack

to

publicdelivery

llustrate he

dynamic

nature f service

delivery

and the

pragmatism

f

local

governmentmanagers.

Monitoring

and

Citizen

Engagement

Contractingequires

lear

specification

f all the attributes f a

good

or

service

even

latent

ones associatedwith

citizen

voice).

In

addition,

t

requires

clear criteria

or

evaluating

performanceEggers

1997;

Poisterand

Strieb

1999).

Performancemeasurement emands

a

professional

monitoring

ystem (Perrin1998)

to enhancethe likelihoodof successful

contracting.ncontrast,heabsenceof a good monitoring ystemmayleadtohigher evels

of

contracting

ack-in.

However,

performance

measurement as been criticized

or

failing

to measure some of

the

most

important spects

of citizen voice and

engagement

hat

are

core to the

governance

process (Sanderson

1998).

A recent

study

measuring

itizen

voice and

privatization

ound that

governments ngaging

in

higher

levels

of

privatiza-

tion

gave

more attention

o

technical

efficiency

measuresbut

not to citizen voice

(Warner

and

Hefetz

2002).

Studies

of

contracting

how that when contract

specification

and

monitoring

are

difficult,

ervicesaremore

ikely

to be

provided

n-house

Nelson

1997).

The

complex

and

changing

nature

of

many public

services

requires

governments

o

engage

in

relational

contracting o thatmonitoring an be an ongoing process(Kavanagh nd Parker1999;

Sclar

2000).

Monitoring

s

costly

in

terms

of time and nformation.Casestudieshave found

the cost of

monitoring

o

average

nearly

20

percent

of contract osts

(Pack

1989;

Prager

1994).

Paroushand

Prager

1999)

suggest public managers

decide whether o

contract

subject

o

the costs of

in-house

production

ersus

contracting

nd

monitoring

osts

given

the

possibility

the

contractor

may

deceive.

The

monitoringprocess

startswith service

specification

ndcontinues

with

government

versight

andassessment

f

service

delivery.

Weak

monitoring

ontrolwill

yield

lower

han

expected avings

Sclar

2000;

Siegal

1999).

We

hypothesize

hat

the

levels

of new

contracting

ut and

contracting

ack-inwill

be affected

by

a combinationof

principal-agent

roblems,

governmentmanagement,

economic and marketconsiderations, nd monitoringand citizen concerns.Typically

studies

of

the decisionto

make or

buy

look at a

single

point

in

time decision

(Coles

and

Hesterly

1998;

Nelson

1997).

We believe that

contracting

s an

ongoingprocess

and that

monitoring

s a

key

feedbackmechanism.

By

looking

over

two time

periods,

we

capture

some

of those

dynamics.

DATA

AND METHODS

Longitudinal

Construction of the Data

Set

This researchcomparesnew contractingout with contractingback-in on managerial,

political,

and economic

grounds.

Data for

this

study

are drawn romthe

ICMA 1992 and

1997

Surveys

of

Alternative

Service

Delivery

Approaches,

he

Census

of

Government

finance iles for

1992,

and he

Census

of

Population

nd

Housing

or 1990.

Every

ive

years

175



7/21

176 Journalof

Public

Administration

Researchand

Theory

since

1982,

ICMAhas

surveyed

all

counties

with

morethan

25,000

population

roughly

1,600

out of

3,100

total)

and

cities over

10,000

population

roughly

3,300).

Roughly

a

third

of

all

governments

ontacted

respond

(31

percent

for

1992

and 32

percent

for

1997),

yieldingasampleof roughly1,400governmentsachyear.TheICMAdatacover64public

services

and

six

alternative ervice

delivery

approaches:

ontracting

o

for-profit

irms,

nonprofit

irms,

or to

another

government

or

public

authority;

elling

franchises,

using

volunteers,

and

providing

subsidies for

the

service.

In this

study

any

alternative

o

delivering

ervices

entirelyby government

mployees

s

considered case of

contracting

out.

The

main

alternatives o

internal

government

rovision

are

contracts

o

private

or-

profit

firms and

intergovernmental

ontracting.

Although

ocal

governments

have

used

contracting

or

a

long

time,

on

average

the

majority

of

services

(58

percent)

are

still

provided

directly

by

in-house

provision

in

1997.

The

surveys

also

ask

managers

68

questions

about factors

that are

motivators,

obstacles,

or

important

n

implementing

alternative

ervice

delivery.

We

paired

esponses

rom

he

1992

and 1997

surveys

o

create

a

longitudinal

ata

set

with

628

responding

municipalities.

Problems

with

changes

n

survey

design

in

1988

do

not

allow

further

ongitudinal

ombination.

Places

responding

o the

survey

n

both

1992

and

1997

exhibit

higher

management

apacity

and

higher

evels

of

contracting.2

The

ICMA

surveys

ask

only

how

the

service

s

provided

urrently,

ot

whether

his is

a

new

contract

or

longstanding rocedure.

To

determine

how

much

new

contracting

was

actually

occurring,

we

neededa

method

o track

changes

n

formsof

service

delivery

for

every

service

for each

government.

We

coded

the data

nto

four

exclusive

categories.

Our

method

distinguished

whether

a

service s

provided

ntirelyby government

mployees,

by

mixed

publicprovision

and

privatecontracts,by

contract

xclusively,

or

not

provided

at

all. We

combined

hese

exclusive

alternatives

ver time

to

createa

transition

matrix

hat

allows us

to

track

changes

n

service

delivery

methods.

See

figure

1.

This

matrix

method

enablesus

to

compare

tability

n

formof

service

delivery

and

to

assess

shifts in

the

direction of

provision-towards

direct

public

delivery

or

towards

contracts.

One

challenge

with

this

pairingapproach

s

that

respondents

may

not

classify

their

service

provision

n the

same

manner n

each

survey

year.

In

order o

increase

he

reliability

of

the

analysis,

we

combined

he two

contracting

lternatives,

mixed

public/

private

and

totally

contracted ut

into

contracting

ut.

We

excluded

he

not

provided

option

from

our

analysis

because

such

events

are not

relevant

o our

analysis.

The

reduced

matrix

has nine

possible

combinations

ver

the two

time

periods

which

we

group

nto four

categories:

table

public

delivery,

table

contracts,

ew

contracting

ut,

and

contracting

ack-in.

Stable

public

delivery

is

found in

section 1

where

government

delivers

he

service

via its

own

employees

n

both

ime

periods.

Stable

contracting

s

found

in

section

2

where

he

government

s

either

contracting

he

service

n

part

or

completely

o

an

alternative

service

provider

in

both

time

periods.

Although

this

option

contains

movements

rom

mixed

public

and

private

delivery

o

totally

contracted ut

and

vice

versa,

we

decided

o

include hese

movements n

the

stable

contracting ategory

ince

contracting

continued.

This

minimizes

chancesof

recall

error

rom

survey

respondents

ho

may

report

2

In the pairedsurveys,the numberof places with a council

manager

orm of

government

s

significantlygreater

(72

percent)

han n

either

of the full

surveys

1992

(62

percent)

or

1997

(64

percent)

X2

=

20.5,p

Documents

Government Contracting Process