24
U NIVERSITÀ P OLITECNICA DELLE M ARCHE ______________________________________________________________________________________________________________________________________________________________________________________________ DIPARTIMENTO DI SCIENZE ECONOMICHE E SOCIALI The Long-run Growth Trajectories ANTONIO G. CALAFATI WORKING PAPER n. 366 October 2011

The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

UNIVERSITÀ POLITECNICA DELLE MARCHE ______________________________________________________________________________________________________________________________________________________________________________________________

DIPARTIMENTO DI SCIENZE ECONOMICHE E SOCIALI

The Long-run Growth Trajectories

ANTONIO G. CALAFATI

WORKING PAPER n. 366

October 2011

Page 2: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

2

Università Politecnica delle Marche

Dipartimento di Scienze Economiche e Sociali

Ancona, Italy

Working Papers Series [http://www.dises.univpm.it/WP]

Scientific Committee:

Renato Balducci Marco Gallegati

Alberto Niccoli

Alberto Zazzaro

Editor:

Massimo Tamberi

Author:

Antonio G. Calafati

Università Politecnica delle Marche

Dipartimento di Scienze Economiche e Sociali P.le Martelli, 8 60121 Ancona, Italy

Web site: http://www.antoniocalafati.it

e-mail: [email protected]

Page 3: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

3

The Long-run Growth Trajectories

of Third

ANTONIO G. CALAFATI

Abstract

The paper conducts an empirical exploration of the growth trajectories of the main cities of the

Third Italy a macro-region that has attracted much attention for its economic performances in the past decades. The findings of the extraordinary heterogeneity of city growth trajectories a

why-question still unobserved and unaddressed will be discussed with regard to two

methodological issues: firstly, the claim that macro-regional performances can be explained by macro-regional factors; secondly, the hypothesis, implicit in many studies, that the factors treated in the model as explanatory variables of macro-regional and urban economic performances

remain constant over time. As the paper will show, the results of the empirical analysis question

prompt more general reflection on the theoretical framework to be used to study the long-run growth performances of cities.

Keywords: Third Italy; macro- -run economic performances; macro-regional performances;

JEL Classification: R11, R12, B41; O18; O43

Page 4: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

4

Page 5: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

5

1. Introduction

The paper focuses on the growth trajectories of Third a question that has

been largely evaded in the large body of literature produced since the 1970s on the

economic performances of this macro-

Third Italy a manufacturing territory made up of the administrative regions of Tuscany,

Emilia Romagna, the Marches and Umbria, located in Central Italy, and with a population

of 10.6 million has arisen from three highly-valued features of the growth model that it

has followed since 1951. Firstly, its overall good economic performances driven by

endogenous capital accumulation; secondly, its polycentric economic (and spatial) growth,

with sustained capital accumulation spontaneously taking place in small- and medium-sized

cities, ensuring an enviable degree of territorial cohesion; thirdly, the manifestation within

its boundaries of cases of extraordinarily rapid local trajectories of industrialisation. To

firstly, on the specific configuration of economic, spatial, institutional and political factors

characterising this macro-region1 and, secondly, on the workings of a specific type of local

y successful growth

trajectories.2 Yet, paradoxical as it may appear with hindsight, a comparative analysis of the

long-run economic performances of the main local systems in particular, cities has not

been conducted.

The empirical exploration of citie

necessary after so much attention had been paid to the performances of industrial districts,

Italy like that of rest of Italy

Central Statistical Office (Istat-Irpet, 1987), and there have been neither conceptual nor

practical

1 Among the vast literature addressing the specificity of Third Italy it suffices in this context to

mention Bagnasco (1977, 1988); Brusco (1982); Fuà and Zacchia (1983); Arrighetti and Serravalli,

1999)

2 The neo-Marshallian literature on industrial district began to flourish in Italy in the Eighties:

Becattini (1987, 1989) contain a number of pioneering studies on this type of local system. Recently,

in Becattini et alii (2009) a synthesis of more than thirty years of research was collected (the

bibliography of this volume is an invaluable sources of information on the vast literature on industrial

districts). See also Becattini (2004) for a collection of path-breaking studies on this subject.

Page 6: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

6

undertaken a fact that should be explained from a critical-historical perspective the

in the scientific and policy discourse, and it has not been possible to draw its

methodological and theoretical implications.

The paper will first condu

-

and call attention to their marked

heterogeneity. The paper will then explore some methodological and theoretical

performances question a theoretical framework based on direct causality from macro-

regional factors to macro-regional performances; it also questions the standard way of

framing the influence of macro-

After Section 2 has briefly discussed the issue of the proper units for urban analysis an

unavoidable issue in Italy

system, Sections 3 and 4 analyse the heterogeneity of the long-run growth trajectories of the

methodological and theoretical implications for the study of the relationship between

macro-regional and urban economic performances. Section 6 concludes the paper.

2. The Urban Dimension of the Third Italy

Addressing the issue of the spatial boundaries of cities an issue whose methodological

importance is being increasingly acknowledged (Parr, 2007) is a preliminary step towards

Third Italy in terms of cities de jure would lead to a profoundly distorted conceptualisation

of the territorial organisation of this macro-region and, consequently, to a flawed

interpretation of its urban performances. This is due to the fact that in the past five decades

the territory of the Third Italy similarly to the rest of the Italian territory has undergone

they have reached a given size (and structure), these clusters should be regarded as cities de

facto

Calafati, 2009; Calafati, 2002).

The legal boundaries of Italian cities have not been adapted to the new territorial

organisation. As a consequence, the extant cities de jure no longer capture the urban

dimension, and a shift to cities de facto as units for analysis is required. To undertake this

shift, it is necessary to proceed as in the case of any other functional unit for territorial

analysis with their practical identification. The debate conducted in Italy to date on

territorial coalescence, however, furnishes all the elements with which to do this with

Page 7: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

7

satisfactory approximation. As early as the mid-1980s, an official map of Italy in terms of

-IRPET, 1987) through

the use of a clustering procedure based on the sizes of communes in terms of population

and employment and the direction and amount of travel-to-work flows. This map, which

was subsequently up-

new territorial units with which to express the local dimension in Italy.

The impact of territorial coalescence on the urban dimension of the Italian territory, and

been matters largely evaded in the scientific debate in Italy. The use of local systems as

proposed by ISTAT in order to identify the new urban poles of the Italian territory was the

nly recently been

highlighted (Calafati and Veneri, 2010; Calafati, 2009; Calafati, 2002). Indeed, the local

systems identified by ISTAT capture the new urban organisation of the Italian territory.

ISTAT certainly overestimates the spatial boundaries of cities de facto, but while in some

cases this overestimation may be important in terms of land extension, it is practically never

significant in terms of population and employment these last being the variables on

which this paper focuses. Therefore, ISTAT

Having defined the most appropriate category for territorial analysis that is, the city de

facto

question which is not necessary to address here: the way in which it is conventionally solved

may be followed in order to fulfil the objective of the paper. Indeed, one can resort to the

ISTAT map, according to which the communes of the Third Italy can be clustered into

144 local systems, and identify the urban system by considering those local systems with a

population of more than 80,000 inhabitants in 2009. Following this procedure, the Third

80.4% of valued added (2005) of the entire macro-region.

n system to be noted is the size distribution of

the cities which highlights the paradigmatic polycentrism of this macro-region. The

histograms of Fig. 1 show the size distribution in terms of population. In 2009, with the

exceptions of Bologna and Florence by far the two largest cities, with populations of

773,776 and 709,754 inhabitants, respectively all cities had sizes falling in the range

between 80,000 and 340,000 inhabitants, with the largest number of cities (18) hosting

populations comprised between 80,000 and 120,000 inhabitants. Considering the

28,775 square kilometres have a sufficiently

high degree of territorial autonomy, it follows that this macro-region is a paradigmatic case

of polycentrism.

Page 8: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

8

Fig. 1

010

20

30

40

Perc

ent

0 160000 320000 480000 640000 800000

1951 (bin=40,000)

010

20

30

40

Perc

ent

0 160000 320000 480000 640000 800000

2009 (bin=40,000)

The current degree of polycentrism in the Third Italy is the result of two fundamental

respect

polycentrism in 1951 as well (see Fig. 1) Secondly, economic growth was spatially diffused,

manifesting itself in cities belonging to all size classes; and therefore also in cities which

were comparatively small in terms of both population and employment in the

manufacturing sector. These two features are closely interconnected. The distribution of

the industrial base over a large number of cities onomic

is a feature at the origin of the endogenous growth that has marked the

recent economic history of this area.

However, the maintenance of a high degree of polycentrism has been associated with a

significant change in the hierarchical ordering of cities in terms of population, employment

correlation coefficients are 0.87, 0.70 and 0.58 respectively. This has been the consequence

of striking differenc

Page 9: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

9

3. -run growth performances

task that has not been accomplished to date. As already noted, this is paradoxical, since

different social sciences in Italy during the 1970s. The obstacle was certainly not a lack of

data. In fact, quantitative analysis of the long-run growth trajectories followed by the Third

-

2001) on employment (and population), which can be integrated with more recent data on

population (until 2009) and industrial employment (until 2007). The reasons for the lack

search field of

local development economics (see Calafati, 2009).

-run growth performances should be conducted from three

growth rates for each unit of time in this case, for each decade should be considered and

evaluated. Secondly, focusing on long-run performances, the growth rate over the entire

period of analysis (or over some meaningful sub-periods) must be analysed. Thirdly, the

time profile of the growth rates namely, the growth trajectories should be examined. As

the empirical analysis will show, the heterogeneity of the long-run performances of the

l-grounded fact.

The box-plots in Figures 2 and 3 illustrate some fundamental empirical facts concerning

3

3 Data on the public sector are available only since 1981 and that prevent a long-rung analysis of

total employment. Public sector unemployment would become an important factor in the explanation

of differences among cities in terms of total employment when the creation of Regional government in

1970 and later the reinforcing of the administrative role of Provinces boasted the public sector in

some cities.

Page 10: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

10

Fig. 2

-10

010

20

30

Pop

ula

tio

n: g

row

th r

ate

s b

y d

ecade

1951-61 1961-71 1971-81 1981-91 1991-01 2001-09

Fig. 3 sector employment

050

10

015

020

0

Em

plo

ym

en

t in

the

pri

va

te s

ecto

r: g

row

th r

te b

y d

ecade

s

1951-61 1961-71 1971-81 1981-91 1991-01 2001-07

Page 11: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

11

The overall picture, which should be treated as a fundamental why-question, is that of a

-

run economic performances cannot be explained in terms of exceptional cases of local

development against the background of fairly homogenous and positive performances by

other cities. Firstly, cases of decline and stagnation are numerous in each time unit;

secondly, the values of the growth rates are not concentrated but rather distributed over the

entire spectrum within the range.

On observing the values over time of the medians (and means) of the growth rates, two

three

decades of the period examined; b) the phase of extensive growth in the decades 1950-1980

was followed by a relative stagnation (on overage).

-question which is

highly significant in

growth rates.

Fig. 4 -2007

02

46

8

Tota

l gro

wth

rate

s 1

951-2

007

Private sector Manufacturing sector

Let us consider first the box-plots of Fig 4, which clearly express the extraordinary

-run growth performances in the period 1951-2007. In order to

avoid cases of extreme relative growth, only the cities with populations of more than

150,000 inhabitants are considered in the graph. Apart from the outlier (Sassuolo), total

growth rates in employment in the private sector range from a minimum of 63% (the old

industrial town of Terni) to a maximum of 430% (Rimini, the largest seaside resort in

Page 12: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

12

Italy)

urban system as defined in the paper, the dispersion would be even larger: a number of

cities exhibit growth rates considerably higher than that of Rimini, and the corresponding

average is 266%.)

drawn from the scatter plot of Fig. 5, where the x-axis depicts employment in the private

sector in 1951 and the y-axis depicts absolute changes in this variable in the period 1951-

Florence (whose scales are comparatively too large to make the graph manageable). To be

explanatory power (confirmed by a value of the correlation coefficient between the two

variables of 0.55): cities that at the beginning of their development trajectory displayed

local production systems similar in size (employment) have experienced strikingly different

total growth rates. The two vertical lines drawn in the graph help to show this crucial

feature.

Fig. 5 -2001)

0

20

000

40

000

60

000

80

000

10

000

0

Em

plo

ye

d in th

e P

riva

te s

ecto

r: a

bsolu

te c

han

ge 1

951

-200

7

0 10000 20000 30000 40000

Employed in the Private sector (1951)

Page 13: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

13

4. The time profile of urban growth trajectories

The previous analysis has evidenced how large the differences in long-

growth performances have been in the Third Italy, but it has not given information on a

further crucial feature, namely the time profile of the growth rates

trajectories. The importance of the time profile resides in the fact that its analysis brings out

a class of phenomena which remain hidden when the focus is exclusively on total

performances but which may be decisive in explaining total performances.

To explore this issue, let us start by comparing the long-run growth trajectories of

Bologna and Florence, which are by far the most important cities in the Third Italy. As Fig.

6 shows, their overall performances differ greatly. During the period under scrutiny, both

cities experienced remarkable growth in population and employment, but Bologna

significantly outperformed Florence.

Fig. 6 -run growth trajectories

Bologna

Florence

10

000

015

000

020

000

025

000

030

000

0

Em

plo

ym

ent

in th

e P

rivate

secto

r

1951 1961 1971 1981 1991 2001

1951-2001

To be noted is that these growth trajectories have the same profile (or shape). Decade by

decade, both cities recorded positive (and comparatively high) growth rates, but those

exhibited by Bologna were somewhat higher in each decade. The cumulative effect over the

47000 more jobs than Florence in the private sector. Focusing on these two cities, the

Page 14: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

14

relevant why-questions seem to be the following: a) why did Bologna and Florence

experience such notable extensive growths? b) why did Bologna record growth rates

persistently higher than those of Florence?

Given the sizes of Bologna and Florence with respect to those of the other cities in the

Third Italy, a counterfactual exercise is warranted: what if Florence had followed the same

long-run growth trajectory as Bologna? But the counterfactual perspective, which points to

order to explain regional performances, appears unavoidable when exploring the time

growth performances.

Fig. 7 Long-run growth trajectories: selected cities (I)

Modena

Parma

Reggio Emilia

Prato

Livorno

Terni

Breakdowns

0

20

00

04

000

06

000

08

000

01

000

00

12

00

00

14

00

00

Em

plo

ym

ent

in t

he p

riva

te s

ecto

r

1951 1961 1971 1981 1991 2001 2011

1951-2007

Rimini

Sassuolo

Lucca

Pisa

Pisa's breakdown

0

20

00

04

000

06

000

08

000

01

000

00

12

00

00

14

00

00

1951 1961 1971 1981 1991 2001 2011

1951-2007

Fig. 7 depicts a set of growth trajectories that confirm the importance of looking at the

time profile as a source of relevant why-questions. Let us start by focusing on the growth

trajectories of the three largest cities in the Third Italy after Bologna and Florence, namely

Modena, Reggio Emilia and Parma. Since 1951 these cities have grown at similar and very

high growth rates and, given their size, they have contributed greatly to regional and

macro-regional performance. The first aspect to note is that, differently from Bologna and

Florence, they have continued their growth trajectories along the same steep trend without

Page 15: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

15

showing any signs of a slowdown in the past three decades. The remarkable and steady

extensive economic growth of these three cities poses an interesting why-question when

compared to the decreasing growth rates of Bologna and Florence, but also to the much

lower and uneven performances of cities of comparable size.

Let us now turn to Prato whose growth trajectory is also depicted in Fig. 7. This city

attracted much attention in the decades of the Italian industrial take-off because of its very

rapid economic and demographic growth. The third largest manufacturing pole in the

Third Italy in 1951 and one of the largest in Italy Prato was conceptualised as a

remarkable example of endogenous growth and as such much researched so that it

After exhibiting the same trend

-91. Apparently, the crisis was followed by prompt recovery

in the subsequent decade. But this rapid recovery notwithstanding, the collapse experienced

in the 1980s put Prato at a significant distance from the other three cities.

Fig. 8 Long-run growth trajectories: selected cities (II)

Piacenza

Ancona

Ravenna

Perugia

0

20

00

04

000

06

000

08

000

0

Em

plo

ym

ent

in t

he p

riva

te s

ecto

r

1951 1961 1971 1981 1991 2001 2011

1951-2007

Ferrara

Forlì

Arezzo

Carpi

Ferrara's Breakdown

0

20

00

04

000

06

000

08

000

0

1951 1961 1971 1981 1991 2001 2011

1951-2007

Analysis of the trajectories of Ferrara, Terni and Livorno, depicted in Fig. 7, makes the

issue of the heterogeneity more complex. At the beginning of the period, these cities were

important manufacturing centres, not particularly different in size from Modena, Reggio

Page 16: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

16

Emilia, and Parma. Their trajectories, however, differed significantly. Ferrara behaved as a

boom city in the 1950s, but after a dramatic economic breakdown in the 1970s, it followed

a sluggish growth trajectory, making a modest contribution to the extensive economic

growth of the Third Italy. Terni and Livorno, two of the largest manufacturing centres in

1951, entered a trajectory of slow growth during the 1950s, experienced an economic

breakdown in the 1980s, and continued along an even more slow-moving growth path.

The above-

in the 1960s, which was a boom decade for practically all the important cities in the Third

Italy. This means that we should not omit from analysis social, economic or political events

that are city-specific events that, as in the case of Ferrara, may deeply and for long affect

the economic history of a city

breakdown in the 1980s was not an isolated event. In the same period, Livorno and Terni,

among the cities analysed, manifested similar behaviours a finding which suggests that

of stagnation for

the Third Italy and for Italy as well. But, as already underlined, a number of cities and

among them important ones like Modena, Reggio Emilia and Parma continued to move

along the same trend, unaffected by the shocks that instead impacted negatively on other

cities.

Consideration of Figures 7 and 8 raises a further theoretical question concerning the

resilience of cities: that is, the rapidity with which they recover after a crisis. As one can

gather from Figures 7 and 8, some cities recovered very rapidly, others very slowly, whilst

others entered a trajectory of decline from which they were unable to escape.4 This finding

the Third Italy have shown different degrees of resilience (adaptation capacity), not just

different degrees of resistance to external shocks.

5. Macro-regional factors and urban performances

- a set of

contiguous administrative regions in Central Italy in the case examined in the paper

4 The private sector as a whole in cities increases also as a consequence of increase in value added

per employed

Page 17: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

17

Hence, a macro-region is a theoretical construct based on the (hypothetical) identification

of macro- -

regional performances (the causal line [a] in Fig. 9.) In research on the Third Italy,

(informal) institutions characterising it according to the received framework among

others, trust, the redistributive orientation of leading firms within their networks of firms,

the orientation to pooling resources within the family are assigned the status of causal

factors of economic performances. The -regional

manufacturing base is also assigned the status of a causal factor of the diffused

industrialisation observed.

Research on the Third Italy has had two focuses. Besides the above-mentioned causal

relatio

particular, it focused on how macro-regional features manifested their effects on the capital

accumulation rate in this type of local system. The analysis of industrial districts was mainly

conducted in terms of a Weberian ideal-type. In its turn, the industrial district became a

dominate the scientific and policy discourse, prevented comparative analysis of industrial

the largest and economic more important local systems, on the other. As depicted in Fig. 9,

if we want to reconcile local and regional (or macro-regional) analysis, it is necessary to put

the entire urban system at centre stage and, consequently, as the foregoing empirical

-run

growth performances.

Fig. 9

The observed heterogeneity of the economic performances of the local systems into

Page 18: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

18

which a macro-region can be decomposed requires the attention to shift to explanation of

ces as an intermediate step towards explanation of macro-regional

growth performances (Fig. 9). However, this shift raises a fundamental methodological

question: what effects do macro-

general, on loca

both the static and dynamic dimension of the problem.

The first solution to this methodological problem is to treat the configuration of macro-

, whose effect is to shift upwards the relationship between

-run growth performances within

the same macro-r -specific states of the

normally limited, for methodological

reasons, to the medium period. (See Cirilli and Veneri (2011) for a recent empirical

exercise on Italian cities conducted from this perspective). In these models the focus is on

identification of the class of factors to be introduced into the model a question recently

addressed from a conceptual perspective by Storper (2010). In an explanatory framework of

this kind, macro-

different macro-regions are compared.

A more complex but theoretically well-grounded framework, which may better capture

the variety of contributions by macro-regional features to the explanation of local

development trajectories, can be developed by introducing the -and-

proposed by Hirschman (1994)

The basic idea behind this methodological perspective is best illustrated by way of example.

contracts. Following this causal chain, one finds that low transaction costs are associated

with vertical dis-integration, and vertical dis-integration with networking processes

(Williamson, 1985). A local production system of this kind will then have the observed

intermediate features (flexibility, on-going innovation, low production costs) and the final

economic one (high accumulation rates).

that not all the steps in the causal sequence have a necessary and sufficient condition in the

previous one. For instance, in the above-outlined case, low transaction costs generate

vertical dis-integration only if the production technology makes vertical dis-integration

transaction costs might be, technological factors prevent vertical dis-integration. It is of

equal interest that effective innovation processes in networks of firms require and this is

cognitive

Page 19: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

19

expected causal effect: if vertical dis-integration is impeded by technological factors or by

a new configuration of relative prices

potential actual effects have to be empirically

ascertained with respect to a specific city or a given typology of city.

The observed heterogeneity of urban growth in the Third Italy raises, with regard to the

macro-regional perspective on urban performances, a further methodological question

concerning the dynamic (or evolutionary) dimensio

stressed, a macro-region is a theoretical construct based on a given set of factors treated as

as is typical

of development theory the following question inevitably arises: can these factors be

-run performance? That is: can they be

the long-run is highly questionable or simply wrong (North, 1990; Kuper, 1999). Whether

or not a structural factor changes over time is an empirical question, and it should be

treated as such. To introduce as an heuristic device the hypothesis that the macro-regional

features have remained constant over the period of analysis is warranted in many research

situations, but it remains an hypothesis that should be tested and relaxed when necessary.

Some macro-regional factors may be regarded as continuously re-generated as a side-

effect of the macro-

fundamental feature of polycentrism was re-produced by the diffused nature of its

industrialisation (which, in turn, depended on the spatial diffusion of the industrial base at

the beginning of the period). Yet one cannot rule out the possibility of macro-regional

factors changing over time. Since the 1950s, formal institutions a

and concerning the fiscal regime, labour-market relations, and local policy-making

processes, have undergone dramatic changes in the Third Italy. During the same decades,

fundamental changes driven by economic growth have also oc

while the redistributive orientation within the family has been weakened by modernisation.

The introduction of the phenomenon of change into the configuration of factors

defining a macro-region has important consequences. Firstly, a region may lose its

specificity because of changes in the factors previously identified as region-specific. (But it

can also acquire region-specific factors. -and-off

-regional features may profoundly

the changing configuration of macro-regional factors combines with local factors in

historical time

features, too, change over time. Therefore, neglect of the evolution of the configuration of

Page 20: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

20

casual factors regional and local factors may lead researchers astray in the interpretation

-run growth.

These methodological remarks point to the task of building an evolutionary model of

-run trajectories. Differences in the observed long-run growth trajectories of

and by the diverse modes in which the evolving local factors combine with macro-

regional and other external factors in generating economic performances. In turn, this task

may be pursued by focusing on case- -

But as has been stressed, we do not lack a sound epistemology with which to address the

study of social phenomena from this perspective (Flyvbjerg, 2001). Even when the aim is to

outline a general theory of city growth, lingering on the inner logic of change as manifest in

specific cities appears to be particularly useful.

6. Conclusions

The empirical analysis conducted in the paper has revealed the marked heterogeneity of the

long-

growth model. And in fact, the why- -run

performances has remained unanswered: we do not have an explanation of the dynamic

performances of the key local systems i.e. cities of this macro-region; and a theoretical

framework on which to base urban development policy is simply non-

but without it being possible to connect them to the underlying urban structures. This is a

rather critical state of affairs. With an urban system made up of cities whose economies are

manufacturing-based in most cases, with city-size distribution skewed toward small and

medium sizes, with some cities already in profound social and economic crises and under

the pressure of globalisation and territorial competition the Third Italy is entering a very

difficult phase of its economic history. The lack of a theoretical framework on which to rely

tial growth trajectories may prove a severe obstacle against the

construction of an effective policy-

in this macro-region.

the

received interpretation of the economic development trajectory of this macro-region, and

namely the casual relationship that runs from macro-regional factors to macro-regional

performances. This is a point of methodological relevance that goes beyond the case-study

performances force conceptualisation of macro-regional growth performances as the

. But this raises a fundamental

Page 21: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

21

methodological question: how to frame the effects of macro-

-

city-specific structural features to determine formances. The paper has

-

in the underlying macro-regional

and urban structures must be taken into consideration.

References

Arrighetti A and Seravalli G (1999) Istituzioni intermedie e sviluppo locale. Rome: Donzelli.

Barca F (2009) An Agenda for a Reformed Cohesion Policy. A Place-based Approach to Meeting

European Union Challenges and Expectations, EU-DG Regional Policy, Independent

Report.

Bagnasco A (1977) Tre Italie: la problematica territoriale dello sviluppo economico italiano.

Bologna: il Mulino.

Bagnasco A (1988) La costruzione sociale del mercato. Bologna: il Mulino.

Becattini G (2004) Industrial Districts. A New Approach to Industrial Change. Cheltenham:

Edward Elgar.

Becattini G (ed.) (1989) Modelli locali di sviluppo. Bologna: il Mulino.

Becattini G (ed.) (1987) Mercato e forze locali: il distretto industriale. Bologna: il Mulino.

Becattini G, Bellandi M and De Propris L (eds) (2009) Handbook of Industrial Districts.

Cheltenham: Edward Elgar Publishing.

Blumenthal P, Wolman HL and Hill E (2009) Understanding the Economic Performance

of Metropolitan Areas in the United States. Urban Studies 46(3): 605-627.

Calafati AG (2009) Macro-regions, Local systems, and Cities: The Conceptualisation of

Territory in Italy since 1950. Italian Journal of Regional Science 8 (Special Issue):

11-34.

Calafati A G and Veneri P (2011) Re-defining the Boundaries of Major Italian cities.

Regional Studies, forthcoming

Coombes MG, Dixon JS, Goddard JB, Openshaw S. and Taylor, PJ (1979) Daily urban

systems in Britain: from theory to practice, Environment and Planning A 11: 565-

574.

Rivista Italiana degli Economisti, forthcoming .

Flyvbierg B (2001)Making Social Science Matter. Cambridge: Cambridge University Press.

Fuà G (1983) Industrializzazione nel Nord Est e nel Centro, in Fuà G and Zacchia C (eds),

Industrializzazione senza fratture. Bologna: Il Mulino.

Page 22: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

22

Glaeser EL, Scheinkman JA and Shleifer A (1995) Economic growth in a cross-section of

cities. Journal of Monetary Economics 36(1): 117-143.

Hirschman A O (1994) The On-and-Off Connection Between Political and Economic

Progress. American Economic Review, Papers and Proceedings, May.

Kuper A (1999) Culture. The Anthropologist Account. Harward, Harvard University Press,

Harvard

ISTAT-IRPET (1989) I mercati locali del lavoro. Milan: FrancoAngeli.

ISTAT (1997) I sistemi locali del lavoro 1991. Rome: Istat.

ISTAT (2005) Distretti industriali e sistemi locali del lavoro 2001. Rome: Istat.

Jacobs J (1969) The Economy of Cities. New York: Vintage.

North DC (1990) Institutions, Institutional Change and Economic Performances. Cambridge:

Cambridge University Press.

Parr J B (2007) Spatial Definitions of the City: Four Perspectives. Urban Studies 44, 381-

392.

Sforzi F (1990) Problemi di definizione dei sistemi urbani. In: Martellato D and Sforzi F

(eds.) Studi sui sistemi urbani. Milan: FrancoAngeli.

Storper M (2010) Why Does a City-Region Grow? Specialisation, Human Capital or

Institutions? Urban Studies 47(10): 2027-2050.

Williamson O E (1985) The Economic Institutions of Capitalism, New York: The Free Press.

Page 23: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

23

Statistical Appendix A Employment 2001

ID Cities de facto Population Number of Total Manufacturing Private Private

2009 communes land sector sector sector

1 Bologna 773776 32 2047 88819 307604 297653

2 Firenze 709754 19 1263 63616 261799 251010

3 Modena 335797 14 641 42759 129862 119235

4 Reggio nell'Emilia 333783 19 869 44918 121946 110533

5 Parma 312751 24 1547 41806 117907 107411

6 Prato 275697 9 409 41941 98042 97271

7 Rimini 226355 8 326 15524 83962 66798

8 Perugia 223321 6 807 16331 67848 59269

9 Ancona 215721 11 404 18103 66783 60826

10 Piacenza 212134 26 1287 19545 70170 60006

11 Ravenna 198368 3 781 13174 68740 55877

12 Terni 190762 18 1133 13492 47996 44686

13 Livorno 182614 5 295 9980 50974 44595

14 Pisa 181714 5 449 7776 49717 44705

15 Ferrara 181392 10 724 12335 51392 49479

16 Sassuolo 167650 11 573 41547 74700 73092

17 Lucca 162190 6 475 18291 56173 52612

18 Forlì 153717 5 462 18296 53052 48268

19 Arezzo 138230 8 798 17317 48295 48043

20 Carpi 133433 6 357 26472 49221 49276

21 Pistoia 130136 4 368 11378 35924 35054

22 Pesaro 129014 16 289 20417 49517 44970

23 Montevarchi 127594 11 718 14250 35338 33461

24 Cesena 124369 6 652 14063 42231 39222

25 Montecatini-Terme 123455 12 303 9943 33715 33471

26 San Benedetto del Tronto 121072 12 300 12623 37186 33202

27 Fano 120988 14 490 15575 39187 32074

28 Viareggio 119414 3 185 7674 33719 28715

29 Pontedera 116373 15 635 15764 37906 34517

30 Ascoli Piceno 114107 17 864 12416 32204 27980

31 Siena 113645 12 1519 7092 34757 31200

32 Empoli 106684 6 341 13549 33790 31705

33 Grosseto 104951 5 1404 3127 26986 22465

34 Lugo 103136 9 480 12611 30084 27850

35 Cesenatico 103035 9 189 10641 38121 29874

36 Santa Croce sull'Arno 102200 6 301 18187 36179 35731

37 Mirandola 102152 10 638 18662 35518 34114

38 Civitanova Marche 99940 5 179 18040 36671 34804

39 Recanati 96076 9 312 17565 32795 30140

40 Imola 93881 8 479 13741 33173 28943

41 Faenza 88005 6 598 8856 25944 23770

42 Fidenza 87122 10 589 10081 27733 25103

43 Foligno 85649 5 522 7157 23080 21140

44 Massa 81371 2 111 5908 22831 19606

45 Jesi 80193 15 368 11643 28731 24467

46 Cento 80053 7 298 11932 23657 21784

Employment 2007

Page 24: The Long-run Growth Trajectories - UNIVPMdocs.dises.univpm.it/web/quaderni/pdf/366.pdfThird Italy a manufacturing territory made up of the administrative regions of Tuscany, Emilia

24