036_LOCAL AND REGIONAL DEVELOPMENT_2017_665

MICHAEL DUNFORD

place it can acquire a historical justification the rationality of administrative systems. In
especially if its development is associated addition, it leads to the existence of func-
with the creation of relatively strong regional tionally over- and under-bounded adminis-
identities and with the development of social trative areas with significant repercussions for
movements that press for the preservation of the meaningfulness of widely used statistical
the resulting regional entities. Also possible indicators.
however is the opposite: the creation of new
political and administrative arrangements The NUTS classification
and new regional divisions designed specifi-
cally to make a break with the past.A step of A European regional policy was first put in
this type can occur as part of projects of eco- place at the start of the 1970s. At that point
nomic and political transformation (Dulong, in time a geographical division of the terri-
1975, 1978). An example is the initial crea- tory of the community was required for the
tion of a regional tier of the administration analysis of regional problems, for the design
in Gaullist France in the 1960s where the and implementation of this new policy includ-
creation of Commissions de développement ing decision-making about eligibility for
économique régionale (CODER) was part of regional aid and for the compilation of har-
a programme of state-directed economic and monized regional statistics to inform analysis
social modernization in which regional plan- and policy decisions.The result was the estab-
ning was a vital instrument and where an lishment of the Nomenclature des Unités
important aim was to recompose regional Territoriales pour la Statistique (NUTS). In
elites in established regions with powerful the 1960s what came to be called NUTS
traditional elites considered as obstacles to eco- LEVEL II areas were identified as the frame-
nomic and political modernization. A case in work used for Member State regional poli-
point was Brittany where a new region was cies,whereas NUTS LEVEL I were identified
defined so as not to coincide with earlier as the principal entities for the analysis of
definitions (Dulong, 1975). A longer view community regional issues such as the sub-
of European development would include national impact of customs union and eco-
many striking instances of these two types of nomic integration, and NUTS LEVEL III
change as the processes of political integra- areas were considered as useful in the diagnosis
tion saw the creation of a European nation of regional problems and in identifying where
state system out of an earlier patchwork quilt regional policy measures were required.Today
of political entities, and as the state system the periodic report on the social and economic
was itself successively modified through the situation and development of the regions of
interaction of further projects of integration, the Community, which the Commission is
attempts to preserve the territorial integrity required to prepare every three years draws
of existing states and attempts to preserve mainly on NUTS LEVEL II data.
historical identities. Administrative regions
can coincide with uniform regions, func- The NUTS is intended to provide a single
tional regions or neither. There are reasons uniform breakdown of the territory of the
related to the criteria that an administrative European Union into a hierarchical set of
system should satisfy that suggest that an statistical regions. The main building blocks
administrative region should make functional of the NUTS system are general-purpose
sense. A situation in which administrative administrative divisions of each Member
and functional regionalizations coincide is state. The current NUTS system is a three-
however in practice difficult to achieve level hierarchical classification of regions in
(Parr, 2007; Dunford, 2010), although non- which each Member State is subdivided into
achievement has important consequences for a whole number of NUTS LEVEL I regions,

528

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

each of which is in turn subdivided into Communities, 2003). Other changes such as
a whole number of NUTS LEVEL II regions the creation of new NUTS LEVEL II areas
and each NUTS LEVEL II region is sub- in the UK had to be examined in detail. In
divided into a whole number of NUTS some of these cases negotiations with Member
LEVEL III regions. As EU concern with States were protracted and difficult (Council
areas not derivable from the these three of the European Communities, 2003). The
NUTS levels and especially with smaller reasons why lay in the impreciseness of the
territories (mountainous areas, disadvantaged statistical criteria and the room they left for
agricultural areas, coastal zones, deprived urban manoeuvre in a situation where the change
areas) increased, smaller NUTS LEVEL IV in classification might have an impact on eli-
and NUTS LEVEL V areas were also identi- gibility for Structural Fund support.
fied. At present however former NUTS
LEVEL IV and 5 areas are classed as Local In 2003 a reliance on ‘gentlemen’s agree-
Administrative Units 1 and 2. ments’ between Member States and Eurostat
to establish NUTS classifications came to an
As the administrative systems in the differ- end with the approval of a NUTS Regulation.
ent Member States differ quite significantly The Regulation calls for the use of objective
combining national territorial communities criteria to define regions, the stability of the
into an EU-wide system was far from straight- nomenclature (laying down clear rules for
forward. It was however the path that in the the management of change with a view to
past was chosen most often: issues of data preventing changes in the classification
availability and regional policy implementa- during negotiations over the allocation of
tion required that the NUTS nomenclature regional assistance) and comparability in the
be based primarily on the institutional divi- sizes of the populations of areas at each level
sions currently in force in the Member States. of the hierarchy (see Table 44.1).
One consequence of these differences is that
in many Member States construction of the Assessing the NUTS
first three intra-Member State tiers permitted classification
the use of only two levels of the administrative
system and therefore required construction of As indicated in the last section the administra-
a non-administrative NUTS Level. tive systems in the different Member States
differ sharply.These differences reflect different
The exceptions to the adaptation of NUTS decisions about the division of responsibilities,
classifications to existing administrative estimates of the population sizes required to
arrangements are mainly found in the new meet responsibilities efficiently and effectively
Member States in central and eastern Europe and distinct histories of sub-national govern-
where the establishment of NUTS classifica- ance. Creating a harmonized EU system of
tions was often accompanied by the top-down sub-national territorial communities was con-
imposition of new sub-national administrative sequently an extremely problematic task.
arrangements.
Table 44.1 Threshold population sizes for NUTS
Until relatively recently the NUTS classi- LEVEL I, II and III areas
fication was changed as a result of the initia-
tive of the statistical offices of the individual Level Minimum Maximum
Member States, although subsequent steps in
the procedure were largely determined by NUTS LEVEL I 3 million 7 million
the way in which the classification was com- NUTS LEVEL II 800,000 3 million
piled.Any change in a national administrative NUTS LEVEL III 150,000 800,000
tier used to establish a particular NUTS level
saw an almost automatic change in the NUTS Source: EU (2003)
classification (Council of the European

529

MICHAEL DUNFORD

The reasons for the choice of national by the standard deviation of the populations
administrative arrangements as the founda- of areas at that level in the EU as a whole.
tion for the NUTS classification are abso- Although this provision improves the situa-
lutely clear: on the one hand data is produced tion, it is clear from Figure 44.1 that while the
for these entities at a Member State level; on mean size of NUTS LEVEL II areas standing
the other hand sub-national administrations at 1,831,000 lies between the threshold values
play a role in the design and implementa- of 800,000 and 3 million, a substantial number
tion of EU-funded regional development of NUTS LEVEL II areas lie outside these
programmes. limits. The smallest had a population of just
26,400 while the largest (Ile de France) had a
One consequence is the heterogeneity of population in excess of nearly 11.4 million.
NUTS regions.To some extent this problem
derives from the fact that population is the A more fundamental problem arises from
only criterion for the allocation of national the fact that the features of important geo-
administrative and non-administrative areas graphical distributions do not necessarily
to different levels of the NUTS hierarchy. coincide with administrative boundaries:
Even in narrowly demographic terms how- areas chosen in deciding on territorial break-
ever the areas vary very widely. The reason downs should ideally reflect the geographical
why is that mean size is used to match admin- distribution of the phenomena under inves-
istrative tiers and non-administrative areas to tigation. In relation to many of the issues dealt
particular NUTS levels, although in the case with in Cohesion policy functional areas and
of non-administrative areas changes under in particular travel to work areas would make
the Regulation are only accepted if they more scientific and policy sense. What is
reduce the degree of dispersion measured more the harmonized application of rules for

Frequency Mean = 1830.73
60 Std. Dev. = 1505.59765

N = 267
Minimum = 26.4
Maximum = 11359.6

40

20

0 2000.00 4000.00 6000.00 8000.00 10000.00 12000.00
0.00 Average population in 2004 (’000s)

Figure 44.1 Average population of NUTS LEVEL II areas in 2004.
Source: Author’s elaboration from Eurostat data

530

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

defining functional regions would ensure the The difficulty arises as GDP is usually meas-
international comparability of the regions in ured by place of employment, whereas pop-
the individual Member States. Insofar as ulation is measured by place of residence.
cohesion policy deals mainly with issues to Measuring GDP by place of employment
do with the geography of economic activi- makes sense as regional policy is designed to
ties and employment travel-to-work areas augment the wealth-creating capacities of
make sense.The difficulty is that if a number of economically disadvantaged areas. If however
other distinct subjects require analysis (access GDP per head is calculated for administrative
to schools,for example) the number of poten- areas that are not at the same time travel-to-
tial functional regionalizations increases. work areas the GDP per capita indicator will
be seriously misleading either because the
As far as regional economic development is administrative areas exclude the places of
concerned travel-to-work areas have a strong employment of the people who live there or
rationale. One of the major disputes in the the places of residence of the people who
regional economic policy area concerns the work there. At present, for example, Inner
extent to which differences in employment London has by far the highest GDP per head
rates reflect ‘demand-side’ (differences in of NUTS LEVEL II areas in the EU. This
employment opportunities) or ‘supply-side’ figure is artificially high. The reason why is
(unemployed do not get jobs that exist) that Inner London includes a very large
factors. In this context there is a significant number of places of employment for people
difference between two types of area: areas who do not reside in Inner London, while
with low levels of employment that are not relatively fewer residents work outside of
within easy travelling distance of anywhere Inner London. Inner London in other words
with a tight labour market; and areas with excludes many of the suburbs of London and
low employment rates that are within com- a vitally important commuter zone that lies
muting distance of areas with tight labour beyond the limits of Greater London.
markets. In areas of the first type that are a
part of concentrations of travel-to-work areas The importance of the use of a set of rea-
(TTWAs) with low employment rates, sonably objective criteria in the definition of
demand for labour needs to be stimulated as areas is also highlighted by the fact that meas-
if jobs are not created within the travel-to- ured indicators of disparities and therefore,
work area concerned only with temporary for example, maps of aid eligibility designed to
or permanent migration affords an answer to target disadvantaged areas depend upon the
low employments rates. In this case searching ways in which regions are created. Figure 44.2
for TTWAs and in particular for concentra- (A) and (B) explores a simple hypothetical
tions of TTWAs with low employment rates example (Dunford, 1993). Suppose a country
plays a particularly important role in the is divided into 16 areas (A1,A2, ..., D4) with
diagnosis and design of regional policies. To identical populations but different levels of
this important consideration must be added GDP per head, and that these areas are
another: areas defined as assisted areas should grouped first into four and then into two
not be defined so narrowly as to cut off sup- regions (see Figure 44.2 (A)). The standard
port from nearby functionally interdepend- deviation expressed as a percentage of the
ent areas that are zones of potential growth. mean decreases from 38.5 per cent (16 areas)
The use of administrative areas that are not to 10.6 (four areas: A1..B2, A3..B4, C1..D2
also functional areas raises a number of and C3..D4) and 3.22 (two areas:A1..B4 and
particularly important difficulties in relation C1..D4). It is important to note however that
to one of the most important indicators the choice of regional boundaries can affect
used for EU regional policy purposes: the result. If in Figure 44.2 (A) the 16 areas
Gross Domestic Product (GDP) per head. are divided horizontally rather than vertically

531

MICHAEL DUNFORD

ABCD ABCD
1 5 10 10 5 10 5 10 10 5
2 10 10 10 10 11 10 10 10 10
3 10 15 20 10 12 10 15 20 10
4 5 10 10 5 13 5 10 10 5

(A) (B)

Figure 44.2 Measured inequality and regional division.
Source: Author’s own elaboration

into two groups (A1..D2 and A3..D4) the 2000 agreement.The Agenda 2000 agreement
indicator falls to 9.67 instead of 3.22. was an agreement relating to the EU budget
Alternatively if four areas are identified in for the period 2000–06 that was finally
the manner indicated in Figure 44.2B the reached at the Berlin summit in 1999. This
coefficient of variation will equal 24.7. agreement was profoundly shaped by the
Measured regional disparities depend, there- implications for in particular EU agricultural
fore, not just on the degree of spatial concen- and structural policies of the soon-to-start
tration of economic activities, but also on the eastern enlargement. Also in 1999 in Ireland
regional division of the country: the number two regional assemblies comprising nomi-
of areas and the choice of boundaries affect nated members of indirectly elected regional
the measure of disparity, just as the delimita- authorities were established.As a result of this
tion of electoral districts shapes the outcome division the Border, Midland and Western
of elections. Clearly the ideal solution is to region retained Objective 1 status for the
use functional economic areas which com- purpose of the Structural Funds for the
bine places of work with corresponding period 2000–06. The Southern and Eastern
places of residence, although disparities region qualified for Structural Funds assist-
between politically identified areas are sig- ance under the phasing-out regime for
nificant as determinants of the resources over Objective 1 until December 2005.
which different communities can exercise
political leverage. A more recent example relates to the
German Land of Sachsen-Anhalt which
The pertinence of this simple example is was divided into three NUTS LEVEL II
demonstrated in practice by the ways in areas (Table 44.2). For the period 2007–13
which changes in regional boundaries have Magdeburg and Dessau were identified as
in practice affected eligibility for EU regional Convergence regions as their average GDP
aid. In the case of the Republic of Ireland, for per capita at PPS in 2000–02 was less than
example, there was just one NUTS LEVEL II 75 per cent of the EU15 average. Halle how-
area up to the point in time when the higher ever was identified as a phasing-out area as its
levels of GDP per capita in the more devel- average per capita GDP at PPS exceeded
oped south and east were so high as to raise 75 per cent. As Table 44.2 shows however
the Republic as a whole over the threshold Sachsen-Anhalt as a whole is small enough
for Objective 1 status (GDP per head less to qualify as a NUTS LEVEL II area. Had it
than 75 per cent of the EU average). At that in fact not been subdivided the whole of the
point in time the Irish government negoti- area would have qualified for funding under
ated a division of Ireland into two NUTS the Convergence objective.
LEVEL II areas in the context of the Agenda

532

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

Table 44.2 Statistical indicators for Sachsen-Anhalt, 2000–04

NUTS 2 regions Sachsen-Anhalt

Magdeburg Halle Dessau

Population 1178061 835933 521421 2535415
Demographic change, 2000-03 (%) −3.00 −3.90 −4.90 −3.70
Employees 476971 339396 195632 1011999
Employee change 2000-03 (%) −2.80 −6.40 −4.30 −4.30
GDP change 2000-03 (%) 9.20 3.80 8.40 7.20
GDP per employee in 2004 (E) 44455 44459 44171 44402
GDP per capita at PPS in 2003 (EU25=100) 75.50 77.60 70.90 75.20
GDP per capita at PPS in 2000-02 (EU25=100) 72.27 75.07 65.99 74.54
Unemployment rate in 2003 (%) 17.6 21.3 21.3 19.6

Source: Statlisches Landesamt Sachsen-Anhalt, Eurostat, calculations of the Staatskanzlei Sachsen-Anhalt,Author’s
calculations

It is finally important to recognize that the boundaries especially of relatively large ter-
generation of statistics for territorial entities ritorial entities with one of the geographical
involves a significant degree of information distributions central to many policy areas.
loss.These information losses are particularly Figure 44.3(A) plots population density by
problematic in situations such as the one LAU LEVEL II administrative units (Cubitt,
depicted in Figure 44.3 (A, B, C, D and E) of 2007). Figure 44.3(B), 44.3(C) and 44.3(D)
non-correspondence of the administrative plot the same data by NUTS LEVEL III areas,

(A) (B)

(C)

(E)
(D)

Figure 44.3 Geographies of population density.
Source: Adapted from Cubitt (2007)

533

MICHAEL DUNFORD

NUTS LEVEL II areas and NUTS LEVEL I A wider concept of development implies a
areas respectively. As this figure makes clear, wider set of indicators and, since the scale and
movement up the NUTS hierarchy results in the extent of interdependence of different
quite extraordinary losses of detail, while the phenomena vary, greater complexity in the
average values for NUTS LEVEL I areas in definition of appropriate areas.
particular can potentially be quite mislead-
ing. Figure 44.3(E) finally plots the same data Areas for regional development
using grid squares.The use of geo-referenced assistance in China
data of this kind provides a significantly supe-
rior representation of the underlying distribu- China differs from the European Union in
tion than the NUTS administrative divisions. that it is a sovereign state with, as the top deci-
sion-making institution, the system of the
To these considerations should finally be National People’s Congress and, as the top
added the fact that development is at present executive institution, the State Council. The
defined in a relatively restricted manner.Wider State Council exercises uniform leadership
definitions of the meaning of development over a series of sub-national tiers of adminis-
require consideration of the distribution of tration and determines the specific division of
wealth and income, the ways in which wealth powers and responsibilities. China is divided
and income are used and economic sustainabil- into 22 provinces,five autonomous regions and
ity. In regional development studies some writ- four municipalities directly under the Central
ers are calling for a more explicit consideration Government (Figure 44.4).The provinces and
of foundational principles of development
such as equity and justice (Pike et al., 2007).

Figure 44.4 Administrative divisions in China and the four economic belts.
Source: Author’s research

534

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

autonomous regions are divided into autono- Although the aim was to accelerate growth
mous prefectures.Autonomous prefectures are (or in Marxist terms develop the productive
divided into counties, autonomous counties forces) permitting some areas and some
and cities. The counties and autonomous people to get rich first, the 1980s nonetheless
counties are divided into townships, ethnic saw the first of a set of initiatives to address
townships and towns. The municipalities China’s growing regional disparities and to
directly under the Central Government and support the restructuring and development of
large cities in the provinces and autonomous economically disadvantaged areas.Accordingly,
regions are divided into districts and counties. the ‘Sixth Five-year Plan’ (1981–85) divided
The result is a four-level system, a three-level the whole country into coastal areas and hin-
system if the prefectural level is absent and a terland.The ‘Seventh Five-year Plan’ put for-
two-level system where municipalities under ward the concepts of ‘eastern, central and
the Central Government are divided only western’ regions.The ‘Eighth Five-year Plan’
into districts (see http://www.china.org.cn/ envisaged strategic development trends for
english/Political/28842.htm). seven cross-provincial economic zones. The
‘Ninth Five-year Plan’ strengthened the
Chinese national regional development financial, investment and policy supports
strategies have for the most part operated at a to central and western regions. The ‘Tenth
very large geographical scale. In the 1960s, Five-year Plan’ put forward proposals for an
China was divided into an Eastern, Central overall plan for regional development,involv-
and Western region. After 1964 the priority ing a ‘great western development drive’ (xibu
was the development of a Third Front (or da kaifa), the restructuring of old industries
third line of defence) of strategic industries and industrial areas in north-eastern China
dispersed in mountainous areas in Sichuan, and the rise of central China with coastal areas
Guizhou and Yunnan in south-west China. continuing to take the lead in development.
Essentially the aim was to develop rail and (The Great Western Development Strategy
other infrastructures and to develop strategic was started in 2000. It covered the provinces
industries (such as chemical, metallurgical, of Gansu, Guizhou, Qinghai, Shaanxi,
energy, machine-making, electronics and avi- Sichuan, and Yunnan, five autonomous
ation) away from the north-east of China and regions (Guangxi, Inner Mongolia, Ningxia,
from the coast in the face of fears of conflict Tibet, and Xinjiang), and one municipality
with the Soviet Union and with the United (Chongqing). This region contains 71.4 per
States which at that time was conducting a cent of mainland China’s area, but, as of the
war in Vietnam. In the 1980s the orientation end of 2002, only 28.8 per cent of its popu-
of development changed radically, yet these lation and, as of 2003, just 16.8 per cent of its
industries laid foundations for a more recent total economic output. The programme
drive to develop western China. involved investment in: infrastructure (trans-
port, hydropower plants, energy and tele-
The early 1970s saw the normalization of communications), the enticement of foreign
relations with the United States and a major investment, increased ecological protection
change of course in China with the adop- (such as reforestation), the promotion of
tion of a strategy of modernization (the education, and retention of talent flowing
four modernizations) and in 1978 of reform to richer provinces. As of 2006, a total of
and opening up. This change of course saw 1 trillion yuan had been spent building
a remarkable acceleration in Chinese eco- infrastructure in western China.
nomic growth at the expense of a marked
increase in geographical and social inequali- A largely similar regional division of China
ties in part as externally oriented growth was underpinned the balanced regional develop-
concentrated in Special Economic Zones ment strategy in the Eleventh Five-year Plan
and open cities on the east coast of China.
535

MICHAEL DUNFORD

which called for development that reflects were put in place. In 1994, 592 poverty
the carrying capacity of the environment and counties were identified.A 2001 revision also
the development regional resource endow- identified 592 poverty counties (plus all 73
ments, that addresses the weaknesses of dis- counties in Tibet) removing poverty counties
advantaged areas, and that involves a clear in eight eastern provinces.These areas receive
zoning of economic activities, stronger inter- earmarked funds for enterprise support, con-
regional interaction, an equitable allocation struction and preferential loans and are given
of public services and reduced disparities in preferential treatment in the allocation of
living standards. To these ends it called for: investment subsidies. In addition, a partnership
advancing the development of the Western system pairs each western province (except
Region, revitalizing north-east China and Tibet which is paired with all provinces)
other old industrial bases, promoting the rise with an eastern province which is required to
of the Central Region, Encouraging the support poverty reduction programmes. To
Eastern Region to take a lead in develop- this spatial strategy the Eleventh Plan added a
ment, and supporting the development of strategy for promoting the formation of pri-
old revolutionary bases (the areas from where ority development zones in part to move in
the Chinese Communist Party and the Red the direction of a model of development that
Army drew its strength in the period from is more sustainable from an environmental
the start of the Long March in 1934 to the point of view. Four classes of area were to be
Communist victory in 1948), ethnic minority identified:
areas and border areas.
i) Optimized development zones: regions
Chinese regional development policies with high-density land development
involve several types of action: and a declining resource and environ-
mental carrying capacity.
1 an investment policy under which,
for example, the Central Government ii) Prioritized development zones: regions
provides 29 per cent of resources for with relatively strong resource endow-
drinking water projects in the east and ment and environmental carrying capa-
63 per cent in the west; city as well as favourable conditions for
the agglomeration of economic acti-
2 a tax policy under which corporate vities and people.
income tax stands at 15 per cent in the
west and 30 per cent in the east with iii) Restricted development zones:regions
until recently 15 per cent for multina- with weak resource endowment and
tional and other companies in Special environmental carrying capacity, poor
Economic Zones) and where there are conditions for agglomeration of eco-
special value-added tax arrangements nomic activities and people,and which
for north-east China and selected cities are crucial to wider regional or national
in central China; ecological security.

3 a credit/loan policy under which dis- iv) Finally, prohibited development zones:
advantaged areas get more long-term legally established nature reserves.
credit; and a tax transfer policy under
which some formula-driven elements These zones were to be identified through
operate to the advantage of disadvan- area classification exercises conducted first at
taged areas.The tax policy and invest- a national level and subsequently at a provin-
ment policy area classifications differ. cial level. This classification raises many
important issues. It raises demographic issues
Alongside successive regional development to do with the relocation of people, and
strategies spatial poverty reduction programmes the livelihoods that support them; issues to

536

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

do with the household registration (hukou) Competences that relate to territorial devel-
system; industrial issues to do with the devel- opment are for the most part competences
opment of non-polluting industries in opti- that are shared with its Member States and
mized, restricted and forbidden development a set of sub-national, regional and local
areas; investment issues; environmental issues authorities.
to do with different environmental regula-
tions in different areas; and not least fiscal The EU budget
issues. On this last front, measures to restrict
development were to prove extremely con- In absolute terms the European Union (EU)
troversial due to the negative impact that they budget is large. At present it stands at over
would have on sub-national government E100 billion per year: in 2007 appropriations
revenues in a period in which central govern- stood at E115.5 billion. As a share of EU
ment was asking sub-national administrations income and public expenditure it is however
to invest more in health and education. small, standing at less than 1 per cent of Gross
National Income (GNI) and at less than
Most striking finally are the ways in which 2.5 per cent of public expenditure.As a share
the evolution of regional policy thinking of GNI however it has recently decreased
reflects the evolution of development models in size.
in China. Just as in the European Union
regional policy where regional policy was As in the past the most recent 2007–13
redesigned to reflect more closely the eco- New Financial Framework was largely deter-
nomic growth-oriented goals of the Lisbon mined by national interests (Mrak and Rant,
Agenda and the Sapir Report (Dunford, 2004). The reason why was that national
2005), in China regional policy is changing interests expressed in terms of the global and
in important ways to reflect the goal of har- partial (related to particular issues) net cash
monious development understood as social flows/net budgetary balances (NBB) gave
harmony and harmony with nature and will rise to coalitions that corresponded very
change further to reduce the degree of closely with the actual coalitions that shaped
dependence of China on export-oriented the negotiation of the budget. The underly-
growth. ing data are plotted in Figure 44.5. On the
vertical axis is plotted each Member State’s
Solidarity, cohesion and NBB defined as total expenditure allocated
the allocation of financial to a country less its total contributions com-
resources in the EU prising traditional own resources, the VAT
source and the GNI source plus net receipts
In order to achieve their strategic goals and from the UK rebate. Net contributors have
to meet their responsibilities governments negative NBBs and net recipients positive
require financial resources. The aim of this NBBs. Each column also identifies partial
section is to identify the ways in which in NBBs defined as the net cash flows attribut-
the European Union (EU) and in its con- able to individual issues: Member State
stituent Member States financial resources receipts from the issue minus Member State
are acquired and allocated in particular to contributions to financing of that issue. In
activities relating to regional economic deve- Figure 44.5 NBBs and partial NBBS are all
lopment. The EU is a union of sovereign expressed as shares of GNI.
nation states.The powers of the EU are those
powers that the Member States agree to These data suggest that new Member
confer on it in order for the EU to achieve its States and net recipient old Member States
objectives as set out in successive Treaties. wanted high spending especially on the
Common Agricultural and Cohesion policies,

537

MICHAEL DUNFORD CAP Cohesion Other UK rebate
6.00%
Relative net budgetary balance as a % of GNI5.00%

4.00%

3.00%

2.00%

1.00%

0.00%

BG LU RO LT LV EE PL HU SK CZ PT EL SI MT BE IE ES CY FI DK FR AT UK IT NL SE DE
–1.00%

Figure 44.5 The 2007–13 new financial framework: net budgetary balances, partial budgetary balances
and GNI.
Source: Author’s elaboration from data in Mrak and Rant (2004)

old Member States wanted low spending The new Cohesion Policy architecture iden-
especially on Cohesion policy, and Belgium tified three objectives and three financial
and Luxembourg as well perhaps as the instruments: a convergence objective; a
European Commission probably wanted regional competitiveness and employment
high administrative spending. Mrak and Rant objective; and, a European territorial cooper-
(2004) showed that the main drivers of the ation objective; 81.4 per cent overall financial
New Financial Framework stemmed from resources were allocated to the convergence
(1) the existence of strong and opposing coa- objective 15.8 per cent to the regional com-
litions that prevented major reductions in petitiveness and convergence objective and
cohesion spending, (2) the ability of the Gang 2.5 per cent to the European territorial
of Six to secure low overall spending and cooperation objective.
in the face of the retention of the October
2002 Franco-Germen agreement to extend In spite of the strong concentration of
the Common Agricultural Policy via a lim- resources on convergence areas, aid intensity
ited financial commitment to the Lisbon does not increase as relative national prosper-
objectives. ity decreases. In Figure 44.6 Member States
are ranked according to their GNI at PPS
The financial resources per head in 2003–5, while aggregate aid
for Cohesion policy per capita is recorded on the vertical axis,
using as a denominator national population
As indicated in the last section Cohesion figures. EU12 countries with the exception
Policy was allocated E308,041 million in of the two newest Member States (Bulgaria
2004 prices (E347,410 in current prices) for and Romania) and Cyprus receive between
the period 2007–13. This sum was divided E373 (Czech Republic) and E252 per head
into a financial profile of annual allocations. (Poland).As Figure 44.6 shows there is a clear
tendency for aid per capita to increase at
538 first as GNI per capita increases and only to

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

400 250

350
200

300

250 150
200
150 100

100
50

50

00
Annual aid per capita ( )
GNI at PPS per capita (EU27=100)

CUznieLtcNuehexdtLeDGPhiKRHRSSIitFBBSleeliPoreoEmuernGFACheurronownlloselrtrLbnplllygtuumSvvmIeggaeMuaaogmutaiaaspdoaapetaadtanreagnuaabanunnolnvrnnkranlrlrrniieiicitiiiiciduadddlyykygmcaanassameaaaanae

Figure 44.6 Aid per capita in 2007–13 (current prices) and GNI at PPS per head (EU27 = 100).

decline once GNI per capita reaches around the EU27 average compared with 71.8 per
84 per cent (in the case of Slovenia) of the cent in the case of the Czech Republic.
EU27 average. Estonia’s GNI at PPS was lower at 55.6 per
cent but was not as low as that of the two
If the population figures used in comput- newest Member States.The aid intensity for
ing aid intensities relate to the populations Poland (E166) is close to that for the UK
resident in eligible areas (and not national (E163) and Germany (E155).
population figures as in Figure 44.6) and the
convergence, phasing-out, phasing-in and Criteria for the allocation
competitiveness and employment objectives of financial resources
are considered separately some striking addi-
tional results emerge. Average annual aid The financial allocations are the result of
intensities indicate that annual aid per capita published and unpublished criteria that
to convergence regions which stands at E184 differ from one strand of policy to another.
in assisted areas is not far ahead of that for the The outcomes however were not simply a
phasing-out areas which receive E141. The result of the application of these criteria but
phasing-in areas receive E82 per head per reflected also a set of overarching constraints
year, and the regional competitiveness and (Bachtler et al., 2007: 24) and a series of
employment (RCE) areas receive E21. Also compromises made in particular during
striking is the fact that amongst the conver- European Council negotiations. Of these
gence regions the highest aid intensities are overarching constraints the most important
for Portugal (E344) and Greece (E333). Of was an absorption cap designed to restrict
the new Member States the Czech Republic financial resources to a share of national
(E269) and Estonia (E239) do well.Romania Gross Domestic Product that the recipients
and Bulgaria receive E84 and E81 respec- could spend effectively.This cap is important
tively even though their per capita GNI at
PPS stood at just 32.7 and 34.4 per cent of 539

MICHAEL DUNFORD

mainly as it overrides mechanisms which convergence regions by a premium of E700.
allocate resources according to need and in If, for example, 1,000 people are out of work,
accordance with the original principle of the unemployment rate is 10 per cent and
concentration of resources. Another factor the average rate is 5 per cent, excess unem-
driving down per capita Structural Fund ployment stands at 500 and the region would
flows to the lowest income Member States receive an additional E350,000.
was the setting of the share of the Cohesion
Fund for Member States that joined the The main driver of the allocation of
Union on or after 1 May 2004 at one-third resources is relative GDP per capita and
of their total financial allocation (Structural relative GNI per capita. The unemployment
Funds plus Cohesion Fund). The effect of driver however generates a quite different
this constraint is to increase the relative geographical distribution allocating resources
importance of Cohesion Fund resources in particular to a number of EU15 Member
which in contrast to Structural Fund States (Italy, Germany, Spain and France).
resources are confined to investments in More strikingly the published indicative allo-
transport and environmental infrastructures. cation of resources differs markedly from the
outcome derivable from the application of
As already mentioned the underlying this variant of the Berlin mechanism. The
criteria vary from one strand of policy to main reason why is that the resulting alloca-
another. Consider the case of the convergence tion of resources is inconsistent with the
objective. As is well known, areas eligible for spending caps,and that the resources in excess
the convergence objective are NUTS 2 areas of the caps that were initially allocated to
whose per capita GDP at PPS is less than low-income Member States are re-allocated.
75 per cent of the Union average.The alloca-
tion of resources for each Member State is Government finance in
the sum of the allocations for its individual EU Member States
eligible regions. The way in which regional
allocations are initially derived centres on the In 2007 the EU budget appropriations stood
so-called Berlin mechanism implemented at E115.5 billion, and that as a share of EU
in 2000–06. Three steps are involved. First, income and public expenditure they stood at
each region’s population is multiplied by less than 1 per cent of GNI and at less than
the difference between its GDP per capita 2.5 per cent of public expenditure. It means
measured at PPS and the EU25 average to that it is the Member States and sub-national
derive a sum expressed in E. Second, the sum tiers of national government that account for
derived from the first step is multiplied by a most European public expenditure.The aim
relative prosperity coefficient reflecting the of the first part of this section is to put some
relative GNI at PPS of the Member State in figures to these roles before attention is paid
which the eligible region is situated. As a to some of the mechanisms for fiscal redistri-
result the sum is larger the lower regional bution inside EU Member States.
GDP per capita and the lower is the relative
prosperity of the Member State concerned. In 2007 EU27 general government expend-
The third step involves the computation of iture (excluding the expenditure of public
an additional sum that reflects the existence corporations) stood at 45.8 per cent of GDP
of relatively high unemployment compared (EUROSTAT, 2008). Government revenues
with other eligible areas.This sum is derived were equal to 44.9 per cent of GDP. For the
by multiplying the number of people out EU15 expenditure stood at 46.1 per cent
of work in that region as a result of the fact (0.8 per cent more than government reve-
that the unemployment rate is in excess of nue).This figure was a long way beneath the
the average unemployment rate in all EU 1995 figure of 52.5 per cent.

540

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

Within the EU27 there are very wide In each Member State similar sets of respon-
variations in government expenditure per sibilities are divided up between national and
capita. In 2007 general government expend- sub-national tiers of government. Generally
iture expressed in Euros and adjusted for speaking sub-national government has sole
Purchasing Power Parities (PPP) was highest or shared responsibilities for a wide but vary-
(setting aside Luxembourg) in Sweden ing range of activities: land use planning,
(16,465), Denmark (15,418), Austria (15,320), economic development, infrastructure provi-
the Netherlands (14,907) and France (14,451). sion, and the provision of a range of local
At the other end of the spectrum lie formerly services that may include education, health
Communist new Member States in eastern and employment.
and central Europe. The lowest scores for
Bulgaria (3,580) and Romania (3,723) are The division of responsibilities across dif-
approximately 22 per cent of the figure for ferent tiers of government requires a corre-
Sweden. Expressed in Euros (without the sponding allocation of financial resources
PPP adjustment) the range extended from such that all sub-national authorities can
28,787 in Luxembourg and 21,104 in meet these responsibilities and in particular
Denmark to 1,421 in Bulgaria and 2,083 in can provide citizens with largely comparable
Romania. By far the most important contri- services at similar levels of overall taxation.
bution was made by central government If all tax revenues are collected centrally,
which accounted for 25.1 per cent of EU27 resources can be allocated so as to secure
GDP. States present in only four countries equal service provision making allowance for
accounted for 4.2 per cent and local govern- differences in needs and costs through for-
ment for 11.2 per cent, although these fig- mula-driven methods of resource allocation
ures differed substantially from one country that allocate more resources to areas with
to another. relatively high costs or greater needs.

National governance and fiscal If conversely some taxes are raised at a
equalization in the EU sub-national scale situations will arise in
which there are mismatches between the
As emphasized in the last section the Member revenue-raising capabilities of sub-national
States account for a large share of public governments and the costs of providing sim-
expenditure, and are mainly responsible for ilar services: some areas will have high tax
a wide range of activities. One indication of revenues and low costs and others will have
the scale and scope of Member State activi- low tax revenues and high costs. In this situ-
ties is provided by the functional (as opposed ation ensuring that citizens receive compara-
to government departmental) analysis of ble services at similar levels of taxation
UK public expenditure: in 2006–07 UK requires movement in the direction of fiscal
public expenditure amounted to 41.3 per equalization either through fiscal redistribu-
cent of GDP. Social protection (13.4 per cent tion (horizontally across tiers of government
of GDP), health (7.1 per cent), education or vertically from higher to lower tiers of
(5.5 per cent), general public services (3.6 government) or tax-sharing arrangements
per cent), economic affairs (2.9 per cent) and (where different tiers of government are
defence (2.4 per cent) were the most impor- entitled to fixed shares of specified taxes).
tant areas of activity. Considered in its widest
sense a number of these areas of expenditure In EU Member States equalization meas-
play an important role in local and regional ures of this kind serve at least to reduce these
development. disparities. Some years ago Wishlade et al.
(1996) estimated the size and impact of fiscal
transfers in seven Member States (France,
Germany, Italy, Portugal, Spain, Sweden and
the UK).As this study showed, irrespective of

541

MICHAEL DUNFORD

whether a flow (in which regions is public life as many assets and many enterprises are
money spent) or benefit concept (in which state-owned.
regions do the benefits of public expenditure
accrue) of the distribution of expenditure In China there are significant disparities in
is used, the richer regions in the eight coun- the resources available to sub-national author-
tries studied transfer significant sums to the ities. An unequal distribution of revenue
poorer regions. At an EU scale, however, conflicts with principles of equal access to
whether a region is a part of an economically public services: sub-national governments are
strong or an economically weak Member not able to finance basic public services such
State makes a great difference: areas with as education, medical care and social security.
similar levels of GDP per head are net recip- As Figure 44.8 shows in 2005 per capita fiscal
ients of public expenditure flows in rich revenue varied from RMB9957 in Shanghai
countries such as Germany but net contribu- to 1,202 in Anhui. These variations were a
tors in poorer countries such as Spain. reflection of large variations in fiscal revenue
(RMB7,980 in Shanghai to 435 in Tibet)
Finance for regional that were not invariably rectified by transfers
development in China (RMB6,921 in Tibet to 521 in Fujian).
Although there are very large per capita
In China fiscal revenue is far smaller as a transfers to some provinces with little fiscal
share of GDP than it is in EU Member States. revenue there are also large positive transfers
As Figure 44.7 shows fiscal revenue declined to relatively rich provincial level cities such
from 30.1 per cent of GDP in 1978 to 10.3 as Shanghai and Beijing. This situation is a
per cent in 1995. After 1995 it rose to reach consequence of a number of features of the
18.4 per cent in 2006. Of course this figure Chinese fiscal system.
underplays the role of the state in economic
As Figure 44.9 shows from 1979 onwards
central government expenditure declined as a
share of the total standing at around 30 per cent

Fiscal income as a share of GDP (%)
40

30

20

10

0 1983 1988 1993 1998 2003 2008
1978

Figure 44.7 Fiscal revenue as a share of GDP, 1978–2006.
Source: Author’s elaboration from People’s Republic of China, National Bureau of Statistics. Note that
figures for some years up to 1990 were interpolated

542

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

Tax revenue/Transfers per resident in 2005 (Yuan/RMB)12000Tax revenue/Total population Transfers/Resident population
InnHeGreCiuSSlhahZhoLGMioGJQSXnaJiYihianiuoSNSianuHBiTgniGHacHunAenHgnHiFanhhajjojgaenTzddiiiaenhiungnigueaiungaanjnqjgJhooaiaiainnguinnnhhhbbnibgonngjxllosnnnsnnnnnaaiiiaiaaaauaexexxxeiiiiiiiiituugggugggggnnnannannn10000

8000

6000

4000

2000

0

Figure 44.8 Fiscal revenue and fiscal transfers per inhabitant, 2005.
Source: Author’s elaboration from People’s Republic of China, National Bureau of Statistics

until 2003. The share of local government 1993 to 52.8 per cent in 2006.As the division
increased to reach in the order of 70 per cent. of responsibilities was not changed central
As a result of the introduction of a tax-sharing government came to receive twice as much
system in 1994 central government’s share of revenue as it spent, whereas local government
total fiscal revenue rose from 22 per cent in received about two-thirds of what it spent.

Fiscal revenue and expenditure (% of total)
100

Local government fiscal income
Local government fiscal expenditure

75

50
Central government fiscal expenditure

25
Central government fiscal income

0 1983 1988 1993 1998 2003 2008
1978

Figure 44.9 Fiscal income and expenditure in China.
Source: Author’s elaboration from People’s Republic of China, National Bureau of Statistics. Note that
figures for some years up to 1990 were interpolated

543

MICHAEL DUNFORD

In China there are five levels of government inter-provincial transfers but these trans-
and five levels of public finance with the fers are not unified into an integrated
central government providing near-pure system.
public goods and other responsibilities split
across levels of government: nine years Conclusions
of compulsory education is provided mainly
at a county level, while rural cooperative The aim of this chapter was to consider two
health care involves four levels up to the interconnected issues that generally receive
county. Central government expenditure too little attention in academic discussions of
accounts for 30–35 per cent of the total. local and regional area development policies:
the definition and classification of areas and
Central government revenues derive from the mechanisms and distributional conse-
21 tax items. At a sub-national level there quences of financial resource allocation.Area
are 31 tax items of which the most important development policies by definition involve
is a resource tax. Extra-budget income from, the identification of geographical areas to
for example, land development accounts for which spatial policies will apply and the
one-third to one-half of local revenue. establishment of criteria determining the eli-
Alongside these two sets of tax items there gibility of different areas for different types of
are a number of tax-sharing items.An exam- support. In this chapter I have shown how in
ple is value-added tax of which 75 per cent the EU and in China administrative defini-
goes to central government and 25 per cent tions of regions are the foundation for most
to sub-national government. area development policies. In the EU a set of
rules have been established in an attempt to
The gap between sub-national government harmonize different national administrative
revenue and expenditure is covered by central systems. This NUTS system plays three
government fiscal transfers although their important roles. First, it provided the frame-
contribution to fiscal equalization is limited. work for the development of harmonized
These transfers fall into three groups: regional statistics. Second, it served as the
foundation for the socio-economic analyses
1. General transfers (33 per cent of the of the EU regions.Third, it provides a frame-
total) are mainly compensation for work for EU regional policy and in particular
losses caused by the 1994 reform. Only it is used in deciding on eligibility for regional
10 per cent of this transfer is in reality aid: with the establishment of the Structural
formula-based. In this case relative Funds the classification of areas eligible for
underdevelopment is considered with support under Objective 1 or the Con-
a national average of 60 per cent and, vergence objective was carried out for NUTS
for example, 90 per cent for Tibet. LEVEL II regions, while the classification of
areas eligible under other priority objectives
2. Specific transfers (33 per cent of the has involved the use of NUTS LEVEL II
total) are divided into a first set of areas. In the Chinese case area development
funds earmarked for service provision, policies also rest on administrative divisions
and a second part which does not of the country. Although China has not
require match-funding comprising developed a set of formal rules similar to
210 vertically managed items whose those embodied in the NUTS system the
allocation is based on precedent/quotas different levels of the Chinese administrative
and not science; and system are roughly comparable with tiers of
the NUTS system, although some levels of
3. a tax rebate (33 per cent of the total)
which is a legacy of 1994 reform and
was designed to ensure that the revenue
of sub-national governments did not
fall. In addition, there are sub-national

544

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

the Chinese system are subject to more rapid occurs however only at a Member State level.
and frequent changes. At an EU level there are very wide dispari-
ties. In relation to resources for area develop-
The reasons for the choice of national ment it was pointed out that aid for 2007–13
administrative arrangements as the foundation is not proportional to relative GNI.Although
for area development policies are absolutely the underlying Berlin mechanism allocates
clear: on the one hand data are produced most resources to the most disadvantaged
for administrative entities; on the other hand areas capping mechanisms in particular result
sub-national administrations play a role in the in a situation in which aid at first increases
design and implementation of regional devel- with GNI and only subsequently falls.
opment programmes.Administrative areas are
however especially in the EU heterogeneous. In the Chinese case large disparities in the
A more fundamental problem arises from the availability of fiscal resources per capital were
fact that the features of important geographi- noted. Although the Chinese government
cal distributions do not necessarily coincide compensates for the lack of financial resources
with administrative boundaries. In relation to in some provinces with very large transfers to
many of the issues dealt with within develop- areas in the west of China, it also supports
ment policy functional areas and in particular politically powerful and economically advanced
travel-to-work areas would make more scien- areas. An unequal distribution of resources is
tific and policy sense.The importance of the an impediment to the Chinese government’s
use of a set of reasonably objective criteria in ambitions to improve health, education and
the definition of areas is also highlighted by social security provision. The importance of
the fact that measured indicators of dispari- fiscal reform is accentuated by several other
ties and therefore, for example, maps of aid factors. One is the need to release savings and
eligibility designed to target disadvantages to expand the domestic market to underpin
areas depend upon the ways in which regions China’s future economic growth. Another is
are created. the fact that the prevention or the restriction
of development in ecologically sensitive areas
As far as financial issues are concerned will under the current fiscal system place
emphasis was placed on the importance of limits on revenue generation in these areas.
examining the geographical distribution of Additional transfers will be required therefore
public finance considered as a whole.Generally not just to enable sub-national authorities to
speaking the per capita financial resources meet their health, education and social secu-
available to sub-national government should rity responsibilities but also to compensate
enable the uniform provision of public serv- these areas for ecological protection schemes
ices.As such these resources should be roughly that will improve environmental conditions
proportional to the population served with not only in the areas affected but in other parts
some allowance for variations in the costs of of China.As these considerations also indicate,
equal service provision due to variations in finally, questions of the definition and financ-
need and cost structures. Area development ing of area development intersect in important
resources exist alongside and complement ways with definitions of the meaning of devel-
normal public service provision providing opment and the choices made with respect to
additional resources to deal with economic development models.
adjustment and economic development but
are by comparison relatively small. In the EU References
case attention was paid mainly to EU area
development policies. Equal service provi- Bachtler, J, Wishlade, F and Méndez, C. (2007)
sion is a responsibility at present of Member New Budget, New Regulations, New Strategies:
States and involves the use of a variety of
schemes for fiscal equalization. Equalization 545

MICHAEL DUNFORD

The 2006 Reform of EU Cohesion Policy, Dunford, M. (2010a) ‘Area definition and classifi-
European Policy Research Paper, 63. Glasgow: cation: the case of the European Union’, in
European Policies Research Centre. Graham Meadows and Wang Yiming, Chinese
Commission of the European Communities and European Cohesion and Regional Development
(2007) Commission Regulation (EC) No. Policies, Brussels: European Commission.
105/2007 of 1 February 2007 amending
the annexes to Regulation (EC) No. 1059/ Dunford, Michael (2010b) ‘Financing solidarity:
2003 of the European Parliament and of the matching resources and responsibilities through
Council on the establishment of a common the budget for European Cohesion Policy and
classification of territorial units for statistics Member State financial arrangement’, in
(NUTS). Graham Meadows and Wang Yiming, Chinese
Commission of the European Communities and European Cohesion and Regional Development
(2008) Commission Regulation (EC) No. Policies, Brussels: European Commission.
11/2008 of 8 January 2008 implementing
Regulation (EC) No. 1059/2003 of the European Parliament and Council of the European
European Parliament and of the Council on Communities (2005) Regulation (EC) No.
the establishment of a common classification 1888/2005 of the European Parliament and of
of territorial units for statistics (NUTS) on the the Council of 26 October 2005 amending
transmission of the time series for the new Regulation (EC) No. 1059/2003 on the estab-
regional breakdown. lishment of a common classification of territo-
Council of the European Communities (1993) rial units for statistics (NUTS) by reason of the
Council Regulation (EEC) No. 2082/93 of accession of the Czech Republic, Estonia,
20 July 1993 amending Regulation (EEC) Cyprus, Latvia, Lithuania, Hungary, Malta,
No. 4253/88 laying down provisions for Poland, Slovenia and Slovakia to the European
implementing Regulation (EEC) No. 2052/88 Union.
as regards coordination of the activities of the
different Structural Funds between themselves European Parliament and Council of the European
and with the operations of the European Communities (2008) Regulation (EC) No.
Investment Bank and other existing financial 176/2008 of the European Parliament and of
instruments. Official Journal L 193, 31/07/ the Council of 20 February 2008 amending
1993 P. 0020–0033. Regulation (EC) No. 1059/2003 on the estab-
Council of the European Communities (2003) lishment of a common classification of territo-
Regulation (EC) No. 1059/2003 of the rial units for statistics (NUTS) by reason of the
European Parliament and of the Council of 26 accession of Bulgaria and Romania to the
May 2003 on the establishment of a common European Union.
classification of territorial units for statistics
(NUTS). Official Journal of the European EUROSTAT (2008) Government Finance Statistics,
Union L154/1. 21/6/2003. Luxembourg: Office for Official Publications
Cubitt, R. (2007) European Regional Classification, of the European Communities.
Statistics and Geography, Second High-level EU
China Seminar, European Union Open Days, Mrak, M and Rant,V (2004) Financial perspective
Brussels, 8–9 October. 2007–2013: domination of national interest.
Dulong, R. (1975) La question bretonne, Paris: http://ec.europa.eu/dgs/policy_advisers/
Armand Colin. conference_docs/mrak_m_rant_dom_
Dulong, R. (1976) ‘La crise du rapport Etat/ national_interests.pdf.
société locale vue au travers de la politique
régionale’, in Nicos Poulantzas (ed.) Le crise de Parr, J. (2007) ‘On the spatial structure of
l’Etat, Paris: Presses Universitaires de France, administration’, Environment and Planning A,
209–232. 39(5): 1255–1268.
Dulong, R. (1978) Les régions, l’état et la société
locale, Paris: Presses Universitaires de France. Pike, A. (2007) ‘Editorial: whither Regional
Dunford, M. (1993) ‘Regional disparities in the Studies?’, Regional Studies, 41(9): 1143–1148.
European Community: evidence from the
REGIO databank’, Regional Studies, 27(8): Pike, A., Rodríguez-Pose, A. and Tomaney, J.
727–43. (2007) ‘What kind of local and regional devel-
Dunford, M. (2005) ‘Growth, inequality and cohe- opment and for whom?’, Regional Studies,
sion: a comment on the Sapir Report’, Regional 41(9): 1253–1269.
Studies, 39(7): 972–978.
Wishlade, F., Yuill, D., Taylor, S., Davezies, L.,
546 Nicot, B.H. and Prud’homme, R. (1996)
Economic and Social Cohesion in the European
Union: the Impact of Member States’ Own Policies,
Rapport définitive à la Commission
Européenne (DG XVI), Glasgow: European
Policies Research Centre.

Wrigley, E.A. (1965) ‘Changes in the philosophy
of geography’, in Peter Haggett and Richard J.

A R E A DEF INIT ION AND CL ASSIF ICAT IO N A N D R EGION AL D EV E LOPMEN T FIN A N C E

Chorley (eds) Frontiers in Geographical Teaching, Fan, C.C. (2001) ‘The political economy uneven
London: Methuen, 3–20. development: the case of China’, Progress in
Human Geography, 25(3): 517–519.

Further reading

Fan, C.C. (1995) “Of belts and ladders: state policy
and uneven regional development in post-
Mao China”, Annals of the Association of
American Geographers, 85(3): 421–449.

547



Section VII

Reflections and futures



45

The language of local and
regional development

Phillip O’Neill

Introduction shows how we have used natural regions to
present the world as organised into tracts of
Novel theoretical vocabularies infuse land based on physical characteristics such as
people’s very beliefs and social prac- climate; how we have used nodal regions to
tices. Along with theoretical redescrip- show the role of central places in providing
tions go practical effects such as changed commercial services to their hinterlands;
views about the object of inquiry, administrative regions to show how institu-
altered practices of study, and the estab- tions can divide the world politically or
lishment of new social groupings and bureaucratically; and vernacular or cultural
institutions. regions for showing how romantic imagina-
tions of people can coincide with distinct
(Cutler 1997: 4; cited by bio-physical landscapes.
Barnes 2001b: 548)
Regionalising our world, though, is not a
It should be obvious that there is noth- dispassionate act of convenience or special
ing like an economy out there, unless interest. When we write regions onto our
and until men [sic] construct such an world we are effectively assigning to it our
object. spatial imaginaries, which are the calculations
and desires that we have of, and for, our
(Louis Dumont 1997: 24; world. Jon Murdoch (2006) explains how
cited by Barnes 1998: 99) our spatial imaginaries are enacted by our
spatial deliberations and performances such
I like the fact that both language and the as through our planning activities, our fiscal
region are indeterminate devices. Language spending patterns and through infrastructure
is a choice among an endless list of words and provision. In this way, spatial imaginaries are
combinations of words and symbols.Similarly, our way of choosing order and sequence from
a region is a choice of the way we represent the spatial menu available to us. Murdoch
the world we live in.When we write from a calls this a process of building governmen-
regional perspective we create a way of view- talities into that complex set of interacting
ing the world for a particular purpose, and entities that we summarise with the word
there is a tradition in this. Wishart (2004) ‘space’. Thus Murdoch sees two things

551

PHILLIP O’NEILL

happening when we create spatial imaginaries: language with the study of society and its
first, we “select the spatial attributes thought to politics to show how deeper understandings
be of most significance” in our interaction of our world are possible. The approach to
with the world; and, second, we intervene “in language in this chapter follows closely an
space on the basis of this selection” (2006: approach that Fairclough has termed the
156, emphasis in original). Murdoch’s point critical discourse analysis approach, or CDA.
is important to this chapter. It advises us that As well, in relation to regional development
regionalising our world is more than a con- questions, the chapter draws heavily on the
venient tidying-up of a world that is a bit work of Trevor Barnes, one of human geog-
messy. Rather it is an imposition of a range of raphy’s leading analysts of the role of lan-
ordering desires to create spatial formations guage in the development of geographical
that determine human activity. As such, thought.
regionalising our world is a powerful act that
warrants raised awareness. What does language enable?

The purpose of this chapter, then, is to A first understanding of the role of language
explore the intriguing relationship between is that it sets up the tasks at hand. Different
language and region by focusing on the way language selections enable different types of
language and its devices drive the ways we regional analysis. Hence to describe a region
mobilise the idea of regional development. we draw on unique words and language
The chapter commences with some basic structures to produce a compendium of facts
views of the role of language in framing our and knowledge. For example, we compose
view of the region as a economic entity.This language in a particular way when we ana-
is followed by three case studies of how lan- lyse the dynamism of a region to show
guage has been used to represent the regional changes in industrial sectors through time.
economy in the last half century: first, as a We choose a different portfolio of language
space where neo-classical economic logics to build abstract understandings and models
drive human activity; second, as a site where of regional development processes. And we
income and expenditure flows can be aggre- choose differently again for policy language
gated into discrete Keynesian entities; and, that can justify, for example, certain taxation,
third, drawing on the Marxist language of expenditure or regulatory interventions on
historical materialism, as class-based building behalf of a region.
blocks, or localities. The chapter concludes
with observations of emerging languages of More than words are used in each of these
regional development and an argument for types of regional analysis.There are also maps,
language consciousness as a prerequisite for diagrams, tables of numbers and calculations,
desirable regional development outcomes. mathematical equations and statistical indica-
tors.These are also language devices, depen-
The role of language dent on unique symbols in carrying meaning
to an audience. Throughout this chapter all
The approach taken in this chapter is post- these devices are understood to be part of
structuralist, meaning the adoption of a view what we call language.
that we live unavoidably in a language-
encased and therefore a language-enabled Language enables five things for under-
world.The approach is guided in general by standing regional development. First, it makes
Jon Murdoch (2006) and by the long list of informed observation possible. Language –
works by Norman Fairclough and Bob Jessop words, numbers, symbols, images – brings an
who have systematically joined the study of otherwise disconnected list of unnamed
landscape items into common understanding
552 (see Fairclough 2003). Language is a social

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

agreement on how to name the things we for example, only through the device of
are talking about: a mountain, a motorway, an language. The Chicago School’s powerful
abandoned factory site. Language also helps and enduring representation of the city by a
us group items into useful categories: men concentric zone model, for instance, con-
and women, employed and unemployed, rev- verted observations of a growing and decay-
enue and expenditure.These are simple func- ing mid-western US city in the 1920s into a
tions of language, naming and classifying; yet template for understanding the dynamism of
they draw attention to our inability to say cities worldwide. Language enables abstract
anything at all about our world when a word generalisations.For example,language enables
to capture the presence of a thing or category us to invent the idea of inequality to describe
of things is unavailable or in dispute (see compounded differences in income, educa-
Gottdiener 1995). tion and employment across groups of house-
holds, and to make judgements about the
Second, language drives analysis. Language desirability of these differences.
sets up the idea of the region as having worth
or value, gives a way of expressing this worth Fourth, language enables us to express our
such as through measurement or compari- feelings about a region, to imagine something
son, and guides the monitoring of changes else, a different state of affairs under different
through time such as by providing a way imagined conditions; to chose alternative
of talking about time and of standardising spatial imaginaries. In other words, language
regional conditions from one moment or provides the opportunity for normative think-
period of time to another. Language also ing and judgments. Such thinking capacity
guides us in depicting a region’s strength, opens regional analysis to political debate
its vulnerability and its stability. And lan- about the desirability of what is going on in
guage provides a way of showing a region’s a region and what possibilities there are for
connections and the ways these might pro- change (see Barnes 2001b).
duce strength and autonomy, or dependence.
Often these notions of strength and connec- Fifth, language enables us to provide rela-
tion depend on the use of language meta- tive fixity to relationships between time and
phors, a language tactic we explore in the space. Fairclough (2003, esp. p. 151), citing
next section. Harvey (1996), shows that while space and
time are central concepts in society and soci-
Third, language guides the way we use etal analysis, they are also social constructs.
abstractions, though we mostly do this sub- Through language constructions, space and
consciously since language itself is a process time become core categories in locating con-
of abstraction, a thought event where we ditions and events, showing how they are
convert an observation, thought or feeling changing, positioning people’s reactions to
into a symbol (a written word, an equation them, and creating the parameters for contest,
or an image) or an utterance (a spoken word). conflict and resolution. Obviously, then, space
The arrangement of letters that make up the and time are key language-based concepts for
words ‘thermal power station’, for instance, the study of regions and pivotal to how we
converts a massive industrial installation with understand regional change and development.
coal stockpiles covering many hectares into a
small set of letters on a page capable of con- In summary, then, language-consciousness
veying the same meaning to the reader that is vital to effective regional analysis.As Nigel
would be conveyed by a direct viewing of Thrift (1990) demonstrates, language estab-
the power station in reality. lishes and drives everything that can said
about the region, what the region is, what the
But language also enables the assembly of region is composed of, why the region is an
more complex abstractions. The world can important scale of areal analysis and the nature
be depicted as a human or natural system, of our analytical and policy aspirations.

553

PHILLIP O’NEILL Barnes depicts the language within such set
or stable practices as ‘dead’ language. Not that
Features of language that Barnes is deriding the use of dead language.
deserve attention Rather he notes that when a new approach
to, say, regional analysis is being developed,
Barnes (for example, in Barnes and Duncan there is conflict between the ideas encased
1992) and Fairclough (especially 1992, 1995, in the language forms of the pre-existing or
2003) have shown the importance of a lan- dead language of analysis and the ideas being
guage self-consciousness in regional analysis developed by the new or alternative analysis.
writing and, therefore, of the need for an For the new ideas to become ascendant, by
understanding of basic linguistics and of the definition they must be propelled by a lan-
need for scholars to acquire the basic skills of guage that is more or less new. Prosaically,
language analysis.We now turn to a brief dis- Trevor Barnes argues that the development
cussion of the features of language helpful in of new ideas about the world involves a
developing a language self-consciousness. “redescription of the world in terms of novel
vocabularies”(2001a:164,citing Culler 1997).
We start with the basic unit of language, In other words, the language which postures
the word. A word is an utterance recordable as a replacement language of analysis is a
as a simple collection of letters which repre- living language, alive to new ideas and there-
sent a word as a symbol or sign. A word thus fore to new ways of expressing these ideas.
refers to an object or an idea, the thing that
is being signified by the sign. The signifier– There are many ways that language stabi-
signified link drives linguistic study, while lises in sets of words that carry agreed mean-
the evaluation of the permanence of signifier– ings. Small groups of words in a pattern,
signified relationships is at the heart of the so-called ‘figures of speech’ or ‘common
post-structuralist debate (see Gottdiener 1995). expressions’,are known technically as‘tropes’.
A trope is an important device in regional
Of course, words are usually delivered in analysis. Most regional development con-
sentences so that their meaning can be cepts are expressed as tropes: the rate of eco-
enhanced by being surrounded by other nomic growth, labour force participation
words. The words that are chosen and the rate, environmental sustainability, regulatory
way they are grouped and presented vary environment, industrial cluster, and so on.
according to different contexts. Context
means that both the selection of words and More powerful and more complex than
the meaning of these words vary according tropes are metaphors.Metaphors carry mean-
to who is the composer (the speaker or ing across otherwise stable language worlds.
writer) and who is the audience (the readers They perform descriptive, comparative, expla-
or listeners), as you would know. natory and judgement roles. Technically, a
metaphor is a figure of speech that carries
What I seek to emphasise here is that an idea into a new domain through juxtapos-
words are delivered through language struc- ing separate things with similar characteris-
tures that either systematically reproduce tics. A common metaphor in economic
ways of thinking and understanding, or else development is the biological metaphor
challenge ways of thinking and understanding. of the human body and its development.
This stabilisation or unsettling of meaning is A region might be described as being in a
fundamental to the processes of scholarship youthful, adult or mature stage of develop-
about regions.As we see in the case studies of ment, its progress monitored by growth rates,
regional development below, any approach its component relations describable as its
to regional development has a set of lan- internal metabolism, its money flows seen as
guage expressions and devices that are stabi- having circulatory properties, while periods
lised by their being agreed on by a user
community; meaning that research practice
lives within a stable set of pre-existing practices.

554

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

of depression are seen to need external injec- the field. Beyond words, expressions and
tions, and so on. metaphors, Fairclough (1992) identifies three
other areas for language consciousness. One
Barnes’ work on metaphor and economic is the understanding of the power of lan-
geography over many years shows the impos- guage’s structure and form; that the meaning
sibility of conceiving the region as an eco- language carries is tied into a writer’s or
nomic entity without resorting to the adoption speaker’s language format and approach.
and stylisation of language metaphors in Thus,when a language-user chooses between
building our representations and, therefore, narrative, analytical, inferential or deductive
our understanding of the real world we seek approaches to talk about a region, this
to understand and change. In other words, requires the selection of a matching vocabu-
metaphors are our pathway to shift from lary, metaphorical base, logical sequence and
observation to theorisation (see Barnes 1991, engagement strategy. A second language
esp. p. 112).The adoption of new ways to do understanding advanced by Fairclough is the
regional analysis, then, involves the adoption role of context, being the situation where
of new metaphors as part of taking on new language is authored and targeted, and the
words and language structures. Emphatically, place for discursive practices to be enacted.
Barnes and Duncan (1992: 11) see that new Hence a political speech to a constituent
metaphors “are the jolt, the frisson, that audience about regional disadvantage will
makes us see the world in a different way … contain markedly different language to the
[M]etaphors create new angles on the world language chosen for an academic journal,
… [and then] they gradually acquire a and to that written by a consultant in a report
habitual use.” to a local government authority. A third area
of understanding of the social practices of
Barnes’ analysis of metaphor, and his atten- language, especially the reflexivity of lan-
tion to language consciousness more generally, guage, is the idea that language development
show how language drives our understanding is inseparable from the triangulated relation-
of regions, leads the development of political ship between author, audience and society.
concerns about regional performance, guides As much as an author might try to ignore
(or limits) the exploration of alternative these relationships, authorship is always over-
regional economic pathways, and fosters new determined by the immanence of audience
planning strategies. The way we talk about and society.The idea that authorship is actu-
regions,from day-to-day conversations through ally negotiated with its audience is developed
to sophisticated academic analysis, is simulta- in the body of work emanating from the
neously language-limited and language- Russian Bakhtin writers’ group, an early
enabled. In the case of the metaphors twentieth-century group of linguists and
deployed in all this talk we can see how met- writers which explored the relationship
aphors that have become common and used between the act of authorship and the act of
uncritically – and therefore naturalised or communication, the anticipated conversation
dead – can produce unintended, uniform to come (see Brandist 2002). This awareness
and uncritical decisions and actions.Without is captured by the concept of dialogism, being
language consciousness, we become the slave the idea that all language is reducible to
of the defunct metaphor-maker, warn Barnes utterances that are part of a wider audience
and Duncan (1992: 62). dialogue.

Of course, beyond an understanding of Clearly there is much to be conscious of
trope and metaphor, there is a vast range of in authoring for a regional development
linguistic understandings and analytical skills purpose. However, individual authors rarely
for building a language-informed approach to develop their own sets of vocabularies,
regional analysis. Fairclough (esp. 1992 and
2003) provides a comprehensive guide to 555

PHILLIP O’NEILL

expressions, metaphors and language struc- development of the discipline of regional
tures. Authors tend to belong to schools of science by economist Walter Isard. Central
thought, akin to what Kuhn (1962) called to Isard’s work was the resuscitation of nine-
‘paradigms’, or groups of work that share teenth-century German spatial imaginaries,
common research motivations (often includ- specifically of a spatial economy underpinned
ing both research questions and political ide- by hierarchies of towns and cities, with pat-
ologies), analytical assumptions, investigative terns of rural land use and industrial invest-
strategies (or epistemologies) and anticipated ments explainable by simple, reproducible
audiences.Three of these schools – neo-classical, logics. Barnes (2003b, 2004) shows the links
Keynesian and localities – are now examined between Isard’s seminal volume Location and
as case studies to demonstrate how language Space Economy (1956) and the work of
underpinned their development and mobili- his fellow economists at MIT, especially Paul
sation. Samuelson. Barnes then traces Isard’s work,
and thus the new language of regional analy-
Case studies of language and sis, through the creation of the University
regional development of Pennsylvania’s Department of Regional
Science, its first PhD graduate William Alonso,
Case study of neo-classical models the rise worldwide of the regional science
of regional development discipline, and the incorporation of its work
into regional policy and practitioner prac-
The neo-classical view of the region remains tices. While regional science as an identifi-
a major influence on the study of local and able discipline has declined markedly over
regional development. Perhaps as a conse- the last three decades, regional science tech-
quence, the analysis and critique of the neo- nologies endure in the toolboxes of regional
classical approach has been a major theme of development practitioners, for example, as
Barnes’ prolific writings; and once again we the basis for calculating location efficiencies
draw heavily on Barnes’ work in this section. in GIS distance minimisation models, and
The lineage of the neo-classical tradition in the spatial imaginaries of new economic
in regional development studies is rather geographers such as Paul Krugman and
clear and simple (see Barnes 2001a). The Masahisa Fujita within the economics disci-
neo-classical approach to the region coa- pline, albeit on its margins.
lesced in the 1950s with a concentration
on applied economic theory and modelling. Features of the neo-classical
The timing here is significant because the approach
development of the neo-classical approach
cannot be isolated from the post-Second Barnes (1987) shows how early regional eco-
World War surge in the grand project of nomic theories accepted naturally, as much as
modernity underpinned by widespread accept- uncritically, neo-classical approaches to
ance of the idea that the application of economy. The inclusion of a spatial dimen-
rational, scientific-based knowledge to the sion was designed to display how ordered,
management of human affairs could produce utility-maximising human behaviours played
unproblematic and universally shared advances out on a relatively unproblematic a-social
in the human condition. The neo-classical spatial surface. Subsequently, social theorists
approach was thus, dialogically, propelled dubbed the rational spatial player ‘economic
by scientifically derived understandings and man’ or, humorously, homo economicus. Citing
received by a scientifically enthused audience. Olsson and Gale (1968: 229), Barnes (1987:
301) identifies the analytical ease of play-
Central to the rise of the neo-classical ing with a predictable human population,
treatment of the region in the 1950s was the

556

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

noting that “the players in the spatial eco- than “purely economic”. Weber’s strategy
nomic game are blessed with the attributes involved deploying the extant language and
of economic man”, and so behave rationally analytical techniques of neo-classicism.
and predictably. Ironically,Weber’s heritage is seen as his con-
tribution to a neo-classical spatial imaginary
The starting point for neo-classical spatial rather than to its repudiation.
analysis is the construction of a spatialised
economic surface.There are four famous his- Apparently the work of August Lösch is
torical spatial constructions of the domain of similarly misinterpreted. Barnes (1987) shows
homo economicus (Figure 45.1). The first is how Lösch extended Weber’s analysis through
Johann Heinrich von Thünen’s concentric his 1954 book The Economics of Location. The
model of land use contained in his 1826 book showed how “spatial economic phe-
manuscript The Isolated State. This was the nomena could be expressed in an explicitly
first published demonstration of neo- abstract, formal, and rationalist vocabulary and
economics in a formal spatial setting. Von directly connected to the empirical world”
Thünen used rent theory to show how the (1954: 546). Lösch’s work continued the evo-
occupancy of rural land nearer a town or city lution of regional science as a law-seeking
depends on the user’s capacity to extract high enterprise, the discovery of abstractions and
rates of returns per square metre of land generalisation that played across space irre-
compared to the land productivity of com- spective of any local bumpy bits. Funnily,
peting users. Barnes (2003a) shows how von though, Lösch’s construction of an idealised
Thünen used geometry, relational algebra spatial surface was his way of demonstrating
and calculus to construct a surface of possi- and explaining the idiosyncrasies of the real
bilities for maximising economic returns, world rather then asserting the presence of the
somewhat anticipating neo-classical marginal ideal in actuality.Perhaps this intention derived
economics by half a century.The methodology from Lösch being a student of heterodox
pioneered what later became standard neo- economist Joseph Schumpeter, always one for
classical model-making: commence with elevating the contingency of real human
abstract assumptions that remove spatial contin- behaviour as a prime economic force. Barnes
gency,and then progressively relax them to show (2003a: 81) quotes Derek Gregory (1994: 58):
the effect of movements in individual variables.
Lösch wanted to disclose, to make vis-
The second iconic neo-classical spatial ible, the systematic order or a rational
model is Alfred Weber’s location triangle, economic landscape – what he [Lösch]
published in his Theory of the Location of called “the rational and therefore natu-
Industries in 1909. The triangle shows the ral order” – because he was convinced
choices available to industrial investors if they that such a demonstration held out the
are motivated solely by transportation cost prospect of domesticating the “illogi-
minimisation. Weber (1929, see Barnes cal, irregular, lawless” forces that ravage
2003a) extended von Thünen’s analysis of [a] chaotic reality.
rural land use into models of industrial loca-
tion.Again,though,it was“a projection of pure In other words, Lösch sought to construct a
economics into the spatial domain” (2003a: spatial order in order to determine how the
78).Yet,intriguingly,its purpose was something chaos of the observed world could be bet-
more than a distillation of real life into a “a tered; showing in Barnes’ (2003a: 81) terms, a
disembodied set of logico-mathematical preference for a “logically constructed land-
procedures and diagrams” (Barnes 2003: 78). scape” over “messy real places”. As with
Barnes reveals how Weber sought to expose Weber, this purpose was largely lost in the
the “cultural aspects of modern capitalism” subsequent use of Lösch’s work.
and therefore to expose it as something more
557

(a)

Grazing

Three-field system

Crop farming with fallow and pastur
ntensive crop farmin
I wood and lum e
Fire g ber
Marke t gard ening

1 2 3 4 56

(b) Market

Factory

Raw Material 1 Raw Material 2

Figure 45.1 Neo-classical spatial analysis models
a) Von Thünen’s concentric land use model
b) Weber’s location triangle

558

(c)

A – level centres
A – level market areas
B – level centres
B – level market areas
C – level market areas

(d) Office Price (rent)

Retail

Office

0 Residential
Shopping zone AB

Distance from centre

Commercial 559
(office zone)

Residential zone

Figure 45.1 Continued
c) Christaller’s Central Place theory
d) Alonoso’s bid-rent model of urban land use
Source: Adapted from Alonso (1964) Christaller (1993) Lösch (1954) Von Thünen (1996)

PHILLIP O’NEILL

In any event, Barnes stresses that what was lies a more fundamental variable,
made of Lösch’s work (say, by Berry and economic rationality…[such that] the
others) wasn’t determined only by its lan- extreme, geographical relationship
guage but by the beliefs and social practices among places,and the internal arrange-
that framed its reading. The two – language, ment within them, all get reduced
and beliefs and social practices – proceed to the single logic of rational choice.
hand in glove, dialogically. In this sense, Geographical diversity and complexity
Lösch’s work was very much part of what we are thus explained away.
now call a modernist view of the world, a
“theoretical redescription” (Barnes 2001b: (Barnes 1987: 32)
548), taking a complex, unevenly composed
world and recomposing it as a stylised space It is important to understand that this elimi-
of rational predictable behaviour with, cru- nation of diversity and complexity is bound
cially, a process of “expressing the subject up in the language approaches of the neo-
matter at hand in terms of a new vocabulary classical scholars – which we call their semi-
and syntax” (2001b: 548). osis (see Fairclough 2003) – especially in
the way they used the languages of mathe-
The third iconic spatial model is from matics and geometry and in the way they
Walter Christaller’s 1933 dissertation The approached the task of theory-building. We
Central Place in Southern Germany. In this now turn to an analysis of these language
enduring work Christaller represented the approaches.
settlement pattern of his home region as a
geometric distribution of differently sized The neo-classical approaches described
towns and cities based on the relative capac- above have four common features. First, as
ity of their retailers to attract consumption explained by Barnes (1989), they employ
spending from households spread across a Cartesian perspectives. In other words, the
featureless plain. Christaller’s neatly arranged neo-classical economic geographers reduced
set of nested hexagons has adorned geogra- the earth’s surface to a measurable grid of x
phy and planning textbooks and lecture and y coordinates where the location of
rooms ever since, the elegance of the repre- human activity was determined by the solu-
sentation at least as enticing as the power of tion of linear equations relating, for example,
any explanation on offer. land use intensity to distance from a city
centre; and commercial pull to population
Similar figurative elegance can be found in size and distance from competing centres.
William Alonso’s bid-rent model of land use Non-linear perspectives and behaviours
in a city. Alonso’s (1964) urban land use other than economic rationalist behaviours
theory sees the geographical location pattern were ignored. As Barnes observes (2001b:
of a city as derived from the opportunities 560), it was an approach that reflected a belief
householders have for maximising utility “that the economic landscape was funda-
across “one-dimensional Euclidean space” mentally ordered and could be grasped all on
(Barnes 1989: 305). In other words, the a … sheet of paper” leading, according to
models of Christaller, Lösch and von Thünen Barnes (1989: 299), to an “emaciated view of
imagined the geographical landscape as a place”.
set of “spatial opportunities” (1989: 306) to
be surveyed by rational actors in pursuit of Second, the neo-classical approaches used
commercial activity. Says Barnes, the universalising languages of mathematics,
especially relational algebra, and of statistics.
In summary, neo-classical economic On one hand this gave the neo-classical
geographers assume that beneath the models the appearance of certainty; while, on
heterogeneous economic landscape the other, made the world’s economic geog-
raphy seem as if it were determined by a
560

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

small number of knowable, measurable vari- (Barnes 2003a: 84), emphasising (2003a: 91)
ables. Thus, as Barnes (1989) points out, an that mathematics in particular offered “an
embrace of mathematics steered attention ineluctable principle that would guarantee
towards the enunciation of general principles the truth” (2003a: 91).
without the need for attention to spatial or
temporal subtleties. Likewise, any explora- Third, the neo-classical spatial modellers
tion of regional data was for the verification used utility-maximising, homo economicus
of abstract spatial principles rather than for assumptions, taking these uncritically from
their discovery. their neo-classical economist colleagues. As
noted above, space was seen only as a con-
Importantly, the language strategies of ceptual dimension onto which the predict-
these neo-classical approaches shielded their able economic behaviours of humans could
proponents from unsettling observations and be mapped. Moreover the humans involved
counter-logics. Two periods of history are were impossibly rational, possessing complete
important here. One is the broadening of the information and making decisions solely to
application of mathematics and geometry to maximise their personal circumstances,
fields of scientific inquiry in the nineteenth which everyone did identically. The field of
and early twentieth century. Regional eco- choice was all-knowable remotely and in
nomics was obviously caught up in this.The advance rather than by empirical observa-
other is the quantitative revolution in the tion. Real space had no a priori existence or
social sciences during the 1960s when rap- relevance; and humans were automatons.
idly expanding computing capacities magni-
fied the power of researchers to manipulate Fourth, while the prime purpose of the
large datasets. Regional economics and its neo-classical scholars was to determine the
regional science brother thus became infused underlying reasons for human spatial behav-
with equations, graphs and numbers. iours, these reasons would be common from
one spatial setting to the next, such that
And then paralleling the quantitative revo- “behind the chaos and complexity of the
lution was the appropriation, often uncriti- world there is an order” (Barnes 1987: 473);
cally, of models and metaphors from the and where “diversity, difference and disjunc-
natural and related sciences. Barnes observes ture are excluded by definition” (1987: 473).
that: In other words, the hermeneutic approach of
the neo-classical spatial modellers involved
From physics emerged gravity and the belief that human behaviour was com-
later entropy-maximising models; from monly determined and therefore predictable
sociology and land economics came once key exogenous variables were identi-
models of land use, social physics, fied – in the same way that scientists sought
including the rank-size rule, and urban to uncover an underlying order in the physi-
factorial ecology; and from geometry cal world (see Barnes 1988).Thus the model-
came network and graph theory and ling approach was nomothetic,or law-seeking
the analysis of topological forms that (see Barnes 2003a: 82). On one hand this was
were incorporated into transportation a major departure from the predominantly
studies: descriptive approach of early regional
geographers. On the other, it removed any
(Barnes 2003a: 84, possibility that geographic space and place
citing Pooler 1977) were capable of generating forces of primary
importance to the way people lived, con-
And Barnes adds that, “Marking all these ducted business and built productive land-
different theoretical ventures was the empha- scapes. Paradoxically, economist’s interest in
sis on conceptual precision, deductive logic the spatial landscape, then, showed little
and analytical rigour, that is, rationalism”
561

PHILLIP O’NEILL

interest in real geography and the condition flowed directly from the basic Keynesian
of society at all. income identity,

In contrast, the application of Keynesian Y = C + I + G + (X – M)
principles to the regional economy question
was motivated principally by a determina- where Y is aggregate income
tion to be able to change business and living C is consumption spending
conditions. We now turn to an examination I is investment spending
of the Keynesian region and the language G is net government spending, and
approach that characterised it. (X – M) is export earnings minus
import spending, a territory’s net
Case study of Keynesian regions trade position.

Like most branches of post-Second World Keynes proved mathematically that varia-
War economics, post-war regional econom- tions in income come from net changes in
ics was dominated by the application of the investment, government spending or export
principles of economics developed by British earnings; and the multiplier effects of these
economist John Maynard Keynes. There is income changes are maximised when they
a vast literature explaining the power of are generated externally. For the region, not
Keynesianism in constructing post-Second having its own capacity for fiscal deficits, the
World War national economies as a set of prime source of economic growth was seen
Keynesian economic categories interacting to come from net growth in regional exports.
in defined ways with reasonably predictable This led to the development of ‘export base
outcomes (see Bryan 2001). Importantly, theory’ in regional economics, the simple
Mitchell (2002) argues that Keynes and Keynesian-inspired maxim that a region’s
Keynesianism were instrumental in creating economic performance was ultimately
‘the economy’ as a discrete object of human dependent on the performance of its tradable
management. Thus Keynesian economics goods and services sector.
can be seen as building the idea of a national
economy and then disaggregating it into Second, even though Keynesian economics
spending categories each of which become a had little to say about spatial economic rela-
potential site for Keynesian-style government tionships, the overall Keynesian approach was
intervention. dependent on there being a defined territory
where the basic Keynesian entities are both
These developments were necessary pre- observable and contained in their contribu-
conditions for the idea of the regional econ- tion to the thing known as ‘the economy’.
omy which enabled new ways of depicting As noted, the Keynesian economy in its
and analysing the operation of spatial econo- original formulation is a national economy.
mies at the sub-national level. Common texts In the regional variation of Keynesian eco-
recording the canons of regional Keynesian nomics, though, the region – and not the
economics are Nourse (1968) and Richardson nation – is seen as the scale where the basic
(1972). Four features of the formulation – Keynesian income identity plays out.That this
each steeped in language – are important. scale jump was unexplained in the regional
First, because Keynesian analysis privileged economics literature is surprising since the
demand conditions as the driver of economic key characteristics and policy levers that
growth, the categories that constitute aggre- ensured export earnings held a privileged
gate demand – rather than supply- or pro- growth position in a national economy – the
duction-side drivers – became key attributes existence of a separate national currency,
to be measured and promoted in regional controls over cross-border capital move-
economic management. These attributes ments, the capacity to redistribute incomes

562

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

across sectors, the ability to regulate the credit- In summary, regional Keynesianism, like
creation capacity of the banking sector, and national Keynesianism, provided Western
the capacity to generate fiscal deficits through political systems with an economic develop-
taxation and government spending powers – ment language that maintained the centrality
were absent from the regional scale. Never- of private sector capitalism – during a period
theless, the mathematics of Keynesianism when a third of the world’s population lived
were applied enthusiastically to regional data, under various forms of socialism – by show-
famously in regional input-output models ing how state intervention and regulation
and in the calculation of regional multipliers, could ameliorate capitalism’s cyclical and
still the information cash-cows of boosterist social excesses and inequities. A different,
regional development agencies and consul- radical language of regional economy based
tancy firms worldwide. on class formation processes, however,
emerged with the failure of Keynesian eco-
Third, their national focus notwithstanding, nomics to maintain national economic
Keynesian economics gave regional economics growth in Western economies during the
a territorially based set of economic relations, 1970s. This language, that of the localities
something neo-classical economics struggled school, is the subject of our next case study.
to do. Hence, not only were there definable
Keynesian relations between the non- Case study of localities
household demand categories: investment,
government spending and exports; these were The localities project, as it might be called,
held in tense spatial relations with their amelio- was an idea that was consciously formed and
rating opposites: saving, taxation and imports. developed for the purpose of examining
Like the Keynesian national economy,then,the the spatialised roll-out of an exploitative
regional economy became a spatial imaginary capitalism across a landscape of economic
replete with measurable economic categories regions each with its own unique social and
and a mathematical schema that defined their cultural histories and contemporary circum-
inter relationships. Accordingly, Keynesianism stances. The localities project started with
gave regions an important political argument two concerns. One was the concern of
for the establishment of regional development human geographers that a search for abstract
agencies with budgetary and regulatory capac- and generalised processes across all societies
ities, adopting the same Keynesian logics that by both orthodox regional scientists and rad-
were driving national economic management ical political economists was denying the
practices; even though the power of these possibility of regionally scaled social proc-
capacities was often exaggerated. esses as determinants of social outcomes, and
not just incidental contexts.
Fourth, the application of Keynesian eco-
nomic theory to the regions legitimised gov- Mary Beth Pudup (1988) articulates this
ernment intervention as desirable economic concern in detail. The demands of modern
practice,irrespective of how effective regional social science, says Pudup, displaced a tradi-
Keynesianism proved to be.The elegant dem- tional interest in regional identity and differ-
onstration by Keynes that government action ence. In contrast, she says, the localities
is capable of smoothing economic growth approach sought to attach contingency to
cycles provided also an argument by repre- general social processes by making their exis-
sentatives of labour and capital and by com- tence entirely dependent on an engagement
munity groups in support of ongoing with actually existing regional geographies.
regional development programmes.A regional In addition, Pudup identifies a determination
Keynesianism was thus instilled simultane- to marry an enlivened regional geography
ously in both post-Second World War eco- with a newly invigorated political economy
nomic growth and social equity programmes.
563

PHILLIP O’NEILL

of class, gender and conflict. The historical “the imperatives of the overall process of
context here was the rise of anti-establish- [capitalist] accumulation” (1979: 234) to pro-
ment politics in Britain in the late 1970s duce a differentiated regional landscape,
to counter the emergence of a Margaret which was in fact a“spatial division of labour”
Thatcher-led neoliberalist restructuring.The (1979: 234), exploited in various ways to
localities project sought a genuine engage- maximise profits according to a “series of
ment with class and gender questions. The ‘rounds’ of new investment” (1979: 234).
predominately British project coincided with Massey observes:
an American political economy which was
focused on a differently scaled politics such ‘The economy’ of any given local area
as the city, in the case of David Harvey (e.g. will thus be a complex result of the
1985; see also Barnes 1989: 302–303), and combination of its succession of roles
the corporation in the case of scholars such within the series of wider, national and
as Barnett and Muller (e.g. 1975) and international, spatial divisions of labour.
Bluestone and Harrison (e.g. 1982), and
activists like Ralph Nader. (Massey 1979: 235)

Barnes (1987) observes that the localities The key forces of the UK economy were
project retrieved an aspiration for the region thus identified as being strongly regional in
as the scale where human behaviour displays character. For example, different economic
“richness and diversity” (1987: 305), such sectors were identified as concentrated “in
that in exploring this richness and diversity the areas most propitious in terms of their
one could uncover “the norms, institutions, requirements for production” (Massey 1979:
symbols and social relationships” (1987: 306) 235); while dominant, chiefly exporting
that underpin the geographical world as well industries became “the structuring elements
as “the role of the state at all scales, the role of in the new emerging pattern of regional dif-
local culture and society, the influence of past ferentiation” (1979: 235). Thereafter, the
and present events, the importance of both “regional problem” – Massey’s spatial imagi-
macro- and micro-economic process, and so nary – was produced by the effects on the
on” (Barnes 1988: 487). spatial division of labour within these indus-
tries of the unwinding of the UK’s “imperial
Importantly, Barnes (1988) nominates relationships” as a “dominant world capitalist
Doreen Massey’s work on south Wales as economy” (1979: 235–236), and the rise of a
“a good example of an attempt to under- corporations-based economy reinforced by
stand the broader geographical context in “hierarchies of control” (1979: 236) involv-
which acts [sic] are to be understood” (1988: ing the externalisation of ownership and
488), for Doreen Massey’s article ‘In what control and the separation of research and
sense a regional problem?’ (1979) is com- development functions from direct produc-
monly seen as having led off the localities tion functions. This then produced an eco-
project. In this article Massey proposed the nomic landscape overlain by both a hierarchy
agenda for localities study. And while the of production regions and a hierarchy of
article sought a generalist (historical materi- cities and regions with the tensions between
alist) explanation of spatial economic differ- them generating major and entrenched forms
ence, it proposed the localities project as an of social inequality. In summary, regional
attempt to talk about regional and uneven problems were seen as “the outcome of the
development by melding the language of changing relationships between the require-
class and the geographical language of the ments of private production for profit and the
region. spatial surface” (Massey 1979: 241) with the
regional problem seen not as “a problem
Massey (1979) argued, then, that regional
equality was the result of general processes:

564

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

produced by regions, but by the organisation localities project.The localities work, he said,
of production itself ” (1979: 243). was important as an attempt to “find a vocab-
ulary” that moved “away from ideas consti-
Pudup (1988) claims that Massey’s (and tuted with metaphors like landscapes,
the localities project’s) view of the region topographies and maps” (Thrift 1993: 93),
was a reaction to having learned the social quoting Morris 1987) seeing regions as
process theories of other progressive disci- “frames for varying practices of space, time
plines and sought ways of inscribing these and speed” (Thrift 1993: 94).
onto a spatialised surface. The features of
locality studies, then, were: Thrift thus exposes the localities project as
an attempt to produce new understanding of
i) the core place of regions in Britain’s the spatial economy through the construc-
roll-out of industrial capitalism; tion of an alternate method of analysis and a
set of language tools for its operationalisa-
ii) the assertion that local economies tion. We now turn to the final part of this
could not be understood on their own chapter and consider the importance of lan-
terms; guage in driving the politics of regional
development projects.
iii) requiring reference to processes oper-
ating more broadly, and therefore Conclusion: the politics
more generally; of regional development
language
iv) the primary place of labour and local
labour markets in industrial transfor- Of course there are other regional develop-
mations; and, ment and economic languages that have
steered the way we create and manage local
v) through labour, the inseparability economies. More recently, there are languages
of local cultures and local political associated with post-Fordist approaches to
economies. the region which have evolved through a
range of approaches tagged by labels such
The driving interest for locality studies (and as institutional economics, new economic
its inherent Marxism) was that these inter- geography, competitive regions, and relational
plays explained the UK’s highly spatialised approaches. Like the case studies above, these
“class structure, historical work culture and approaches are also language dependent in
gender relations” (Pudup 1988: 382). the ways they construct particular views, or
spatial imaginations, of regional economies.
Despite the claim to regional and local
contingencies,however,and a class-rootedness In this part of the chapter, we reflect on
discernible at these scales, Thrift (1991) language’s role in the production of theory
showed how locality studies were implicitly and politics, using Norman Fairclough as a
structuralist. According to Thrift, locality guide, and we make some concluding com-
studies stemmed from “a specific need to ments on how language can be mobilised to
understand the economic, social, cultural and secure better regional economic outcomes.
political changes taking place in the UK in
the 1980s” (1991: 459) using the framework Norman Fairclough argues that the “rela-
of “Marxism through the substitution of tively durable social structuring of language”
space for time” (1991: 460). As such, says intersects with the “relatively durable struc-
Thrift, they elevated place to be an essence turing and networking of social practices”
that underpins human activity, a mixer that (2003: 3) such that language singularly pro-
gives flavour and meaning to life, very much duces a “certain commonality and stability in
in a Vidalian tradition, though without pri-
mary causative status. Elsewhere, however, 565
and emphasising its language encasement,
Thrift cedes the spatial integrity of the

PHILLIP O’NEILL

the way the world is represented” (2003: 126). and discursive practices in regional analysis
This view has important implications for the and representation. Powerful skills of lan-
way we understand the process of regional guage analysis are available from study of
development analysis and policy development Fairclough’s texts (e.g. 1992, 1995, 2003).
and, therefore, how we tackle the regional Beyond supplying the social scientist with
analysis task from a language-conscious, or such skills, however, Fairclough (in 2003)
post-structuralist, perspective. Echoing Barnes, delivers a manifesto for critical discourse
Fairclough asserts: analysis, which can be read as a guide to a
politically charged language consciousness in
When different discourses come into regional development studies.For Fairclough,
conflict and particular discourses are semiosis – the broader act of signification and
contested, what is centrally contested communication, including communication
is the power of … preconstructed through visual images as well as language –
semantic systems to generate particular is (at the risk of harping) an “irreducible
versions of the world which may have element of all material social processes”
the performative power to sustain or (2003: 204).
remake the world in their image, so to
speak. In conclusion, consider four final claims.
The first is that there is great political lever-
(Fairclough 2003: 130) age to be gained from allowing both lan-
guage and the region to be indeterminate
Hence, as we have observed, language is critical devices. The strategic purpose of creating
to the processes of social change, especially in such uncertainty is to take advantage of
how we represent them and then bring for- what Fairclough calls an “oscillation” of social
ward new forms of, say, capitalism following practices between a determination by fairly
crisis and envelop these in persuasive ideolo- obstinate social processes and the opportuni-
gies and fill them with everyday social ties afforded by the presence of social action
practices. and agency. Key to understanding this oscil-
lation, says Fairclough, is the way practices
For me,the enduring power of Fairclough’s develop as a stabilised configuration of
work is in its illustration of how language diverse social elements which always include
brings about changes in our knowledge, discourse (Fairclough 2003: 205). As dis-
beliefs and values. Language shapes our cussed, discourse is part of all social activity;
identities; for example, as men and women, it drives all representations; and it figures
as consumers and workers. Language, too, centrally in the shaping of identity, practice
shapes our mental world, our thoughts, feel- and of ways of being (Fairclough 2003: 206).
ings and beliefs. Language also brings into Hence by choosing a language of representa-
being changes in the material world through tion and analysis which exposes the region as
changes in architecture and design, for a framework for observation, understanding
instance; and by changing the ways we and action, we cannot avoid watching the
respond to these changes. Finally, language way our depictions of the region – our spa-
enables us to respond to change as well as to tial imaginaries – shape identity, economic
project and imagine possible worlds different practices and ways of being.An infinite selec-
from the world we currently experience (see tion of language opportunities for an infinite
Fairclough 2003: 8, 124). number of regional analyses, then, doesn’t
paralyse our regional work. Rather, it forces
Beyond its theoretical worth, Fairclough’s us to consciously select the language and the
work is a guide to analysing the presence of regional framework that best match our pur-
language in social practices and social change. poses. And then it beefs up the power of
This chapter has shown the power of under-
standing the role of words, trope, metaphor

566

THE LANGUAGE OF LOCAL AND REGIONAL DEVELOPMENT

social action and agency in the face of obsti- quotation which opens this chapter, Dumont
nate social processes. observes that the economy is very much a
constructed thing. As regional analysts –
The second claim is that language con- through our spatial imaginaries – we are
sciousness is a necessary precursor to achiev- complicit in this construction. A language
ing desirable regional economic outcomes. consciousness in our approach can only help
There is little doubt that the set of motiva- us do our bit of the job better.
tions underlying Norman Fairclough’s per-
sistence in his project to raise language References
consciousness among social scientists includes
his frustration at deteriorating social and Alonso,W. (1964) Location and Land Use:Toward a
regional conditions in his home nation, the General Theory of Land Rent, Cambridge:
UK, under neoliberalist economic manage- Harvard University Press.
ment regimes, his recognition of the role
that language has played in maintaining a Barnes, T. (1987) ‘Homo economicus, physical
hegemony of neoliberalist practices, and his metaphors, and universal models in economic
observation of a paucity of political responses geography’, The Canadian Geographer 31,
from academic social scientists. Fairclough 299–308.
sees that it is necessary to expose the language
strategies of neoliberalism. But, he stresses, Barnes, T. (1988) ‘Rationality and relativism in
this requires an enhanced set of language economic geography: an interpretive review of
analysis skills among social scientists. the homo economicus assumption’, Progress in
Human Geography 12, 473–496.
The third claim,then,is that regional analysis
should always be moving from a language- Barnes, T. (1989) ‘Place, space and theories
limited approach to a language-enabling of economic value: contextualism and essen-
approach; or, in Barnes’ words, to be moving tialism in economic geography’, Transactions of
from dead languages of analysis to new, living the Institute of British Geographers 14, 299–316.
languages. Quite simply, Barnes instructs us,
when our words, tropes, metaphors, models, Barnes, T. (1991) ‘Metaphors and conversations
equations and images are static,then we are lag- in economic geography: Richard Rorty and
ging in matching our analytical words with an the Gravity Model’, Geografiska Annaler 73,
actually existing world which has moved on in 111–120.
time and space.As clumsy as the process often
is, we should always experiment with new Barnes, T. (1998) ‘Political economy III: confes-
ways of saying things in order to build better sions of a political economist’, Progress in
understandings and more desirable outcomes. Human Geography 22, 94–103.

The fourth claim is that the most enlivened Barnes, T. (2001a) ‘On theory, history, and
regional politics comes from a serious anoraks’, Antipode 162–167.
acknowledgement of dialogism and inter-
texuality. Such an acknowledgement turns Barnes, T. (2001b) ‘Retheorizing economic
the idea of communicating academic find- geography: From the quantitative revolution
ings on its head. An engaged language, as we to the “cultural turn” ’, Annals of the Association
have seen, is always a triangulation between of American Geographers 91, 546–565.
author, audience and wider society. When
this engagement is resisted, a failure to com- Barnes,T. (2003a) ‘The place of locational analysis:
municate is one result. Another is that the a selective and interpretive history’, Progress in
author forgoes access to the knowledge and Human Geography 27, 69–95.
experiences that come from the presence of
an audience in the writing process. In the Barnes, T. (2003b) ‘What’s wrong with American
regional science? A view from science studies’,
Canadian Journal of Regional Science 26, 3–18.

Barnes, T. (2004) ‘The rise (and decline) of
American regional science: lessons for the new
economic geography?, Journal of Economic
Geography 4, 107–129.

Barnes,T. J. and Duncan, J. S. (eds) (1992) Writing
Worlds: Discourse, Text and Metaphor in the
Representation of Landscape, London: Routledge.

Barnett, R.J. and Muller, R.E. (1975) Global Reach:
The Power of the Multinational Corporations,
London: Jonathan Cape.

567

PHILLIP O’NEILL

Bluestone, B. and Harrison, B. (1982) The Nourse, H.O. (1968) Regional Economics: A Study
Deindustrialization of America: Plant Closings, in the Economic Structure, Stability and Growth of
Community Abandonment, and the Dismantling of Regions, New York: McGraw Hill:
Basic Industry, New York: Basic Books.
Olsson, G. and Gale, S. (1968) ‘Spatial theory and
Brandist, C. (2002) The Bakhtin Circle: Philosophy, human behaviour, Papers in Regional Science,’
Culture and Politics, London: Pluto Press. 21(1), 229–242.

Bryan, D. (2001) ‘Global accumulation and Pudup, M.B. (1988) ‘Arguments within regional
accounting for national economic identity’, geography’, Progress in Human Geography 12,
Review of Radical Political Economics 33, 57–77. 369–390.

Christaller, W. (1993) Centeral Places in Southern Richardson, H.W. (1972) Regional Economics:
Germany, Jena: Fischer. Location Theory, Urban Structure and Regional
Change, London:Weidenfeld Nicolson.
Cohen,T. (2001) Jacques Derrida and the Humanities:
A Critical Reader, Cambridge: Cambridge Thrift, N. (1990) ‘For a new regional
University Press. geography 1’, Progress in Human Geography
14, 272–279.
Fairclough, N. (1992) Discourse and Social Change,
Cambridge: Polity Press. Thrift, N. (1991) ‘For a new regional
geography 2’, Progress in Human Geography
Fairclough, N. (1995) Critical Discourse Analysis, 15, 456–466.
London: Longman.
Thrift, N. (1993) ‘For a new regional
Fairclough, N. (2003) Analysing Discourse: Textual geography 3’, Progress in Human Geography
Analysis for Social Research, London: Routledge. 17, 92–100.

Fairclough, N., Jessop, B. and Sayer, A. (2003) von Thünen, J.H. (1996) Von Thünen’s Isolated
‘Critical realism and semiosis’, in J.M. Roberts State, an English translation of Der Isolierte Staat
and J. Joseph (eds) Realism, Discourse and (translated by C.M.Wartenberg; edited with an
Deconstruction, London: Routledge, 23–42. introduction by P. Hall), Oxford: Pergamon
Press (originally published 1826).
Gottdiener, M. (1995) Postmodern Semiotics:
Material Culture and the Forms of Postmodern Wishart, D. (2004) ‘Period and region’, Progress in
Life, Oxford and Cambridge: Blackwell. Human Geography 28, 305–319.

Harvey, D. (1985) The Urbanization of Capital, Further reading
Baltimore: John Hopkins University Press.
Aylett, A. and Barnes, T. (2009) ‘Language and
Harvey, D. (1996) Justice, Nature and the Geography research’, in R. Kitchen and N. Thrift, (eds)
of Difference, Oxford: Blackwell. International Encyclopaedia of Human Geography,
Volume 6, 153–158. (On the ways language
Isard, W. (1956) Location and Space-economy; A has shaped human geography and regional
General Theory Relating to Industrial Location, studies.)
Market Areas, Land Use,Trade, and Urban Structure,
Cambridge:Technology Press of Massachusetts Fairclough, N. (2003) Analysing Discourse: Textual
Institute of Technology, and Wiley. Analysis for Social Research, London: Routledge.
(A highly useful explanation and guide for
Kuhn, T.S. (1962) The Structure of Scientific social scientists concerning the techniques of
Revolutions, Chicago: University of Chicago language analysis.)
Press.
Murdoch, J. (2006) Post-structuralist Geography:
Losch, A. (1954) The Economics of Location A Guide to Relational Space, London: Sage.
(2nd edition; translated from the German by (An accessible text which explores post-
W.H. Woglom with the assistance of structuralist approaches to the study of space
W.F. Stolper), New Haven, CT: Yale University and place.)
Press (originally published 1940).

Massey, D. (1979) ‘In what sense a regional prob-
lem?’, Regional Studies 13, 233–243.

Mitchell, T. (2002) Rule of Experts: Egypt, Techno-
Politics, Modernity, Berkeley: University of
California Press.

Murdoch, J. (2006) Post-structuralist Geography:
A Guide to Relational Space, Sage: London.

568

46

The evaluation of local and
regional development policy

Dave Valler

Introduction Yet despite this sustained period of attention

The past 20 years or so have witnessed the there is little substantive basis upon which to
development of a substantial critique of eval-
uative work in local and regional economic frame judgements on some of the most basic
development (LRED), most notably in the
United States. As Clarke and Gaile argued questions regarding LRED. What counts as
in 1992, efforts to evaluate local economic
development policy had become “a quag- the ‘success’ or ‘failure’ of LRED in a locality
mire of good intentions and bad measures”
(1992: 193), and a variety of contributions or region, and why? Is the form of LRED
have since sought to change the bases for
evaluative work. In general these have prob- in a place appropriate given the economic,
lematized traditional evaluative measures and
methods, such as the focus on basic job crea- social, political and cultural challenges being
tion/retention measures, and argued for a
shift from narrow ‘process’ or ‘formative’ eval- faced? Is the programme of LRED sustaina-
uation to a broader concern with the overall
‘outcomes’ and ‘impacts’ of policies (see, for ble? My contention is that the lack of clear
example, Bartik, 2002; Bartik and Bingham,
1995; Reese and Fasenfest, 1997). This has guidance on some of these most basic ques-
also been associated with calls for a much
wider range of criteria in evaluating LRED tions derives from a theoretical tendency to
(Reese and Fasenfest, 1997; Molotch, 1991),
albeit accepting the added complexity which underplay both the broader context of eco-
would inevitably arise.Additionally, a number
of contributions have exposed the essentially nomic and state restructuring and the locally
value-laden nature of evaluative processes
and judgements and the often implicit theo- specific conditions within which economic
retical positions adopted in evaluative work.
development activities are situated, and

within which their distinctive contributions

are defined. Responding to this requires the

elaboration of an alternative theoretical

standpoint sensitive to the particularity of

place, and to the multiple roles that LRED

activity comes

to play across production, reproduction, con-

sumption, ideological, political and cultural

spheres.In this sense evaluation would become

a more complex, comprehensive and contex-

tualised exercise than it has been to date.

In this chapter I briefly outline some

aspects of the critical commentary around

LRED evaluation, together with some of the

569

DAVE VALLER

more important avenues which have emerged context within which demonstrating policy
in response. In spite of the value of these sub- efficacy and value for money became much
sequent contributions, little headway has more significant. Here neoliberal ideology and
been made in terms of situating local and the hegemony of public choice theory and
regional economic development in context, new managerialism underscored an ‘eviden-
or in establishing an explicit theoretical basis tial turn’ (Taylor, 2005: 601; see also OECD,
upon which ‘success’ or ‘failure’ might be 2008) in local policy-making, based on the
judged. Indeed, it can certainly be argued that instrumental use of systematic research and
the theoretical foundations for LRED evalu- scientific knowledge, and emphasising as the
ation remain comparatively underdeveloped ideal form of relevant knowledge “quantita-
in a field dominated by more pragmatic and tive methodologies, empirically-tested and
immediate concerns. As a result, evaluative validated” (Sanderson, 2002: 6). In the UK
exercises have been significantly limited in this was associated with the Blairite mantra
terms of interpreting the real ‘meaning’ of ‘what matters is what works’ and a new era
LRED policy and in appraising the overall of policy-making informed by evidence of
impact of LRED as a strategic response rather what was ‘proven’ to be effective in achieving
than as a series of isolated projects. In light of particular outcomes (Nutley, 2003: 3), while
this I argue for an approach based in regula- the US paralleled ‘new managerialism’ with
tion theory, a conceptual framework intro- total quality management.These more rigor-
duced in France in the 1970s which has ous contexts would, of course, be far less
subsequently come to exert considerable sympathetic to the scattergun “shoot any-
influence in studies of the state and social and thing that flies, claim anything that falls”
economic policy more generally across a approach portrayed by some economic devel-
wide variety of contexts, both in Europe and opment practitioners in the mid-1980s (see
in the US.This theoretical stance can assist sig- Rubin, 1988).Third, particularly in the US, a
nificantly in locating LRED activity in a serious questioning of the basic rationale for,
broader political-economic context,in explain- and performance of, LRED policy effectively
ing the evolution of LRED and its distinctive elevated the importance of evaluation proc-
institutional and policy forms with reference esses and the pressure on individual pro-
to the evolution of the political economy and grammes to justify their existence (see, for
the state, and in establishing a framework example, Peters and Fisher, 2004).
within which to evaluate the scale, nature and
overall contribution of LRED activity. With regard to underlying conceptual
issues, authors have argued that ‘economic
Evaluation: critical themes development’ has cultivated very specific
‘growth’ objectives focused on job creation,
While a broad concern for the measurement business attraction and an enhanced local tax
of policy performance has been apparent base.This is in contrast to any broader notion
throughout the history of LRED policy, it is of ‘development’ and the associated construc-
since the mid-1980s that a detailed and thor- tion of new innovative capacities, increased
oughgoing critique of evaluation per se has adaptability, and engagement with questions
emerged. This has incorporated, on the one of social justice and quality of life (Reese and
hand, a direct examination of the concep- Fasenfest, 1997: 196–198; Wolman and
tual and methodological challenges posed by Spitzley, 1996; see also Beauregard, 1999;
LRED evaluation in particular, a point consid- Foley, 1992). Such a narrow focus is not, of
ered further below. However, it also acknow- course, entirely surprising, and as Beauregard
ledged a changed political and governance (1999) demonstrates there are strong reasons
for an employment focus in evaluations of
570 local economic performance.Yet it betrays an

T HE EVAL UAT ION OF L O C AL AN D R EGION AL D EV E LOPMEN T POLIC Y

underlying value system which effectively overcome or allow for these and other meth-
privileges formal, commercial and business- odological problems. Limitations of space
oriented economic activity at the expense of preclude a detailed rehearsal of all the twists
the informal sector and broader notions of and turns of this debate, and readers are
democracy, equity, and social and environ- directed to relevant literature in Further
mental justice (Reese and Fasenfest, 1997; reading below. However, some of the more
Pike et al., 2007). In addition, it references a important directions can be identified.
somewhat under developed theoretical basis
for economic development activity whereby As a first step, there has been a general
“state and local economic development strat- acceptance, at least in principle, of the need
egies evolve incrementally without (any) under- for a more sophisticated approach to evalua-
lying economic theory except that more jobs tion, along with a clearer focus on outcomes
are good and less jobs are bad” (Beaumont (see, for example, OECD, 2004). This has a
and Hovey, 1985: 328; see also Valler and number of implications including, for exam-
Wood, 2009, forthcoming). In this sense the ple: a concern for unexpected results as well
task of evaluation is immediately coloured by as delivery against predetermined criteria;
a partial and in some senses limited under- a focus on the interaction of programmes
standing of the basic constitution of local and policies across a wide range of sectors;
economies (Beauregard, 1993). the need for longitudinal data; the use of
comparative research and control groups; and
Beyond these foundational concerns the questions over the specification of appropri-
evaluation of LRED activity faces numerous ate evaluation criteria. In practice, of course,
methodological challenges, not least because this raises significant challenges including
evaluation of a policy action requires an those alluded to above, and these may in turn
assessment of what would have happened constrain the evaluative exercise undertaken.
in the absence of such activity. Coulson Yet in many different national contexts,
(1990) outlines the problems of ‘displace- and across both academic and policy circles,
ment’ (where job creation in one place is at it is apparent that the focus on outcomes
the expense of jobs in another),‘deadweight’ evaluation has sharpened. In the UK, for
(where the outcome would have been the example, HM Treasury issued a ‘Framework
same in the absence of policy action), ‘indi- for the Evaluation of Regeneration Projects
rect impacts’ (including multipliers), and and Programmes’ in 1995, setting out a gen-
‘leverage’ (which can lead to accounting dif- eral framework for the ex-post evaluation
ficulties), all of which add to the complexity of expenditure projects and programmes
of evaluation. Further problems include, but with regeneration objectives, covering social,
are certainly not limited to: determining environmental and economic objectives and
which outcomes can in fact be measured, distinguishing between ‘monitoring’ (the col-
and which ‘proximate’ outcomes might be lection of performance data) and ‘evaluation’
appropriate where ultimate outcomes are dif- (the examination of a policy to distinguish
ficult to measure; imprecise or inadequate how effectively and efficiently it delivered
information from respondents; potential biases predetermined objectives). In this sense there
in sampling; significant resource constraints; has been something of a turn towards out-
and the thorny issue of demonstrating clear comes evaluation in local and regional eco-
causal relationships (see, for example, Bartik, nomic development and cognate fields even
1992, 2002; Foley, 1992). Despite these obvi- if, as the OECD admitted in 2004, the core
ous difficulties, however, an enormous litera- task of determining programme effectiveness
ture has emerged around the evaluation of has tended to be ‘somewhat obscured’ by
local economic development programmes, a focus on formative evaluation and the
deploying a range of techniques designed to improvement of programme operations.

571

DAVE VALLER

A second key dimension of the evaluation Feser and Sweeney, 2006) and especially in
debate has been the call for more sophisti- developing and transition economies where
cated and rigorous statistical approaches, par- data problems are often severe. Accepting
ticularly in response to the counter-factual this, though, it is apparent that detailed statis-
problem. How might we identify the precise tical investigation and appropriate modelling
impact of firm- or area-based policies on can offer important insights into the evalua-
levels of economic activity such as business tion of individual economic development
creation/expansion, employment generation, programmes. Already a vast literature exists
and productivity growth? Bartik (1992, 2002), around the economic impacts of enterprise
for example, has argued for experimental zones (e.g. Boarnet, 2001), job-training
comparisons between ‘control’ and ‘treatment’ schemes (e.g. Giloth, 2000), tax increment
firms or areas where such policy differentia- financing schemes (e.g. Dye and Merriman,
tion is practicable, or statistical comparisons 2000), and the effects of state and local taxes
of economic impacts across recipient and on business location and growth (see Bartik,
non-recipient firms or areas where it is not. 1991, 1992; Buss, 2001), amongst many other
These methods would adopt randomized programmes. These studies have gone some
criteria to avoid selection biases and a variety way to produce relatively clear and defensi-
of statistical techniques to allow for differences ble claims of programme impact in a field
in other variables which affect outcomes, characterized by significant uncertainty and
thereby “generating objective quantitative complexity.There are, however, obvious dif-
evidence on the bottom line for the pro- ficulties in this area which effectively limit
gram” (Bartik, 2002: 29). Evaluations should, the development of this type of evaluative
however, examine not only the impacts of work in practice. It is expensive, resource-
policies on local business growth, but also intensive and complex; there are practical
important public benefits, namely: difficulties in identifying appropriate compa-
rator groups; and there may be multiple
fiscal benefits for government, and sources of potential bias. It is telling, for
increased earnings for the unemployed example, that Bartik (2002: 12) is aware of
or underemployed. Fiscal and employ- only a single case of economic development
ment benefits can be estimated using evaluation in the US utilizing data from an
regional econometric models which experiment using random assignment.
are combined with special modules
that consider the structure of local A third – rather British-flavoured – strand
taxes and government budgets, and the in the debate highlights the concern with
local labor market. explanation, and the task of opening up the
‘black box’ of policy mechanics to identify
(Bartik, 2002: iii) more precisely how and why a policy works
or fails (Ho, 1999: 424;Turok, 1989; see also
While a closer examination of these tech- Jones, 1999). In place of basic output mea-
niques is beyond the scope of this chapter, it sures or value-for-money indicators evalua-
is certainly worth noting challenges associ- tion should seek a detailed understanding
ated with the production of local and regional of the causal processes and mechanisms which
economic data which in the UK case has generate specific outcomes in particular
been characterized as limited, historically times and places.As Turok (1989) argues, this
skewed towards manufacturing rather than requires some form of theoretical apparatus
services, and unresponsive in the face of eco- in order to identify key processes and rela-
nomic restructuring (see‘TheAllsop Review’, tions, to situate policy in context, and pro-
HM Treasury, 2004). Similar concerns are vide a convincing explanatory basis for the
frequently voiced in the US (for example, diversity of outcomes. The link between

572

T HE EVAL UAT ION OF L O C AL AN D R EGION AL D EV E LOPMEN T POLIC Y

policy and outcome in LRED, for example, and differentiation can often throw causal
is often indirect and complex, requiring care- dynamics into sharp relief (see also Hudson
ful theoretical analysis and detailed empirical et al., 1997; McCann, 2008).
investigation. Additionally, local economic
policy operates in a much broader context of Finally here, a number of authors have
economic, social, and political forces as well argued for evaluation to incorporate broader
as distinct local conditions all of which are social, welfare and equity impacts which go
integral to the production of outcomes and well beyond traditional economic measures,
cannot be detached from evaluative analysis. a point which chimes with contemporary
In this sense evaluation necessarily requires a calls for a broader and more holistic defini-
convincing theoretical and empirical account tion of ‘development’ per se (see Pike et al.,
of specific outcomes, which would in turn 2007) and is consistent with the theoretical
enable policy learning and the build up of standpoint we elaborate below. Such impacts
cumulative knowledge in contrast to the might include, for example, quality-of-life
shifting fashions and occasional leaps of faith indicators such as life expectancy, educational
which have tended to characterize LRED attainment, health and income disparity,
policy. housing quality and the like, alongside fur-
ther qualitative information regarding the
Developments along these lines have nature of employment and economic activity
yielded fruitful theoretical insights, not least generated (Reese and Fasenfest, 1997).These
around the formulation of a ‘realist’ approach measures potentially offer a more rounded
to evaluation (Ho, 1999, 2003). Here the picture of LRED outcomes, though there are
focus is thrown on the causal properties and clearly serious difficulties involved in select-
liabilities inherent in policy forms, properties ing appropriate indictors and attributing cau-
which may or may not be realized depending sality. An increasingly important concern is
on the diverse contexts in which they are the question of political empowerment and
situated.This then requires an explicit theo- the need to engage recipient communities in
rization of the proposed policy response, the design and conduct of LRED policies,
setting out the mechanisms through which giving rise to ‘participatory’ and ‘empower-
specific programme elements would counter ment’ evaluation focused on stakeholder
perceived problems and a consideration of participation, capacity-building, and active
context in relation to the measures adopted, learning rather than judgement or account-
in order to explain diverse policy outcomes. ability (Brooks, 2008: 28; see Fetterman,
However, despite the obvious importance of 2001).This has, in some instances, supported
such understanding, in practice these insights a range of innovative participatory proce-
have been limited in terms of application dures and a significant re-examination of the
and unfortunately somewhat marginalized respective roles of stakeholders, evaluators,
in the face of more instrumental concerns. and communities in evaluation processes.
In part this probably signifies the need for It has thereby underscored a more reflexive
further middle-range theoretical develop- examination of the evaluative task, and gen-
ment in order to ground these more general erated an enhanced sensitivity to the com-
philosophical insights, and to demonstrate plex and contested nature of LRED policy.
the value of such explanation in comparison It has also found particular resonance in
to more basic measures. It also suggests the developing economies where urban expan-
potential value of analyses of cross-national sion and population growth, infrastructure
policy transfer and learning, given the con- shortfalls and burgeoning urban poverty may
tingent and uneven outcomes created by be accompanied by ‘back-to-front’ develop-
‘policies in motion’ (Peck, 2003: 229). Such ment taking place in advance of formal plan-
processes of international experimentation ning and regulatory procedures. In these

573

DAVE VALLER

contexts participatory approaches may go together to derive a contingent and necessar-
some way to recognize and validate the role ily temporary stability in capitalist accumula-
of the urban poor in developing towns tion processes (Aglietta, 1979, 1998; Boyer,
and cities (Majale, 2008: 273; see also Atav, 1990; Jessop, 1990, 1997; Lipietz, 1987; Peck
2007; Hordijk and Baud, 2006). However, and Tickell, 1992, 1995). This establishes a
the scale of participatory activity overall has distinctive theoretical starting point in which
been marginal in comparison to the empha- the economy tends neither to an automatic
sis on formal performance measures, targets equilibrium as in conventional economics,
and outputs which dominate contemporary nor to an inevitable breakdown as in Marxist
governance arrangements. As the OECD theory (Friedman, 2000: 61). Rather, an
(2004: 4) states, the conventional approach ‘accumulation system’, or characteristic set of
has been to: relationships between production and con-
sumption, emerges to a position of domi-
evaluate individual policy instruments nance within an economic arena and may
and programmes against their explic- come to be sustained in the medium term by
itly stated objectives. In this way, a ‘mode of regulation’ or collection of social
programme evaluations tend to produce and institutional supports which together
isolated and often disappointing provide a degree of coherence and stability to
findings, without due regard to the an overall ‘regime of accumulation’. Such sta-
interaction and cumulative impact of bility is necessarily temporary, acting only to
policies that, by design or not, work in mute or disguise both the inherent contradic-
a ‘target-oriented’ way. tions of capitalist production and the tensions
which necessarily exist within institutional
(OECD, 2004: 4) forms. Each regime of accumulation thus
contains the seeds of its own destruction,
It is in this context that we seek a more con- beyond which capitalist production and regu-
vincing and comprehensive theoretical lation must be thoroughly transformed to
account which might facilitate broader and secure its future survival. Capitalism therefore
more satisfying judgement. This requires proceeds historically through periods of sta-
linkages to be drawn between the nature and bility and growth, when accumulation and
conduct of LRED activity and the broader reproduction are relatively steady, and periods
context of political-economic dynamics and of crisis, when the conditions for capitalist
state restructuring. For this we turn to the social reproduction are found wanting.
broad insights of regulation theory as the
basis for an overarching conceptual frame for Within this overarching framework regu-
the evaluation of LRED activity (see also lationists have offered specific accounts of
Valler and Wood, 2010, forthcoming). political-economic restructuring, most nota-
bly around the emergence and consolidation
A regulationist contribution? of Fordism (Aglietta, 1979) and the putative
transition to post-Fordism (Lipietz, 1987).
The regulation approach Much of this early literature is well known
and there is no need to rehearse it here.
By now the contours of the original regula- Additionally, a wide range of authors have
tionist account are well known. Beginning deployed regulationist insights to position
from an assertion of the contradictory and and inform their analyses across a variety of
crisis-prone nature of capitalist society, regu- spheres, including, for example, UK urban
lation theory emphasizes the social norms, politics (Goodwin et al., 1993), UK urban
mechanisms and institutions which may come and regional development (Peck and Tickell,
1992; 1995), US urban policy (Florida and
574

T HE EVAL UAT ION OF L O C AL AN D R EGION AL D EV E LOPMEN T POLIC Y

Jonas, 1991), and US housing (Florida and ‘post-Fordist’ mode of regulation (Goodwin,
Feldman, 1988). Such accounts have achieved 2001: 74). On the one hand the structures
notable advances in these various spheres and practices of economic development
by contextualizing particular policy fields in activity at sub-national scales will be criti-
a broader political-economic and institu- cally influenced by broader processes of
tional arena, and making critical connections restructuring in state–market relations, inno-
between economic, social, political, cultural vation and competition systems, formal reg-
and institutional practices. Regulationists have ulatory frameworks, governance structures,
also elaborated on a number of apparent patterns of social organization, political and
weaknesses in initial formulations of the regu- ideological commitments, and economic,
lationist approach, most notably in responding social and cultural policy. Clearly national
to accusations of structuralism and economic policy shifts will impact directly on institu-
determinism.Yet the central theoretical and tional forms and policy development and
methodological contribution of regulation implementation in localities and regions.Yet
theory has proved notably robust, and it this cannot in any sense be seen as a simplis-
might be argued that as “a perspective and tic or undifferentiated projection of national
form of analysis” (Goodwin and Painter, changes onto localities. Rather, regulatory
1996) regulation theory has come to exert practices operating at a variety of scales nec-
significant influence. Though concern for essarily find expression locally, both reflect-
the underlying “generative structures and ing the pre-existing character of uneven
mechanisms that shape the actual movement development, and being actively constituted
of social forces” (James, 2009: 184) was largely locally (Goodwin et al., 1993: 69).
superseded after the late 1990s by a more
concretized focus on ‘governance’ forms, the Additionally, of course, localities and
basic importance of sensitivity to macro- regions come to play different roles, with dif-
economic change and political strategy and ferent degrees of success or failure, within
struggle continues to resonate (see especially overall national regimes of accumulation. For
James, 2009). Indeed, despite the relative it is apparent that regions and localities are
absence of explicitly regulationist accounts characterized by distinctive accumulation-
of policy and institutional change since the regulation couplings and will be inserted dif-
late 1990s such insights offer a distinctively ferentially into both wider spatial divisions of
holistic and wide-ranging analytical frame labour and regulatory structures.As Peck and
within which to account for such change, Tickell point out (1992: 352),“some regional
albeit in combination with appropriate economies … will be favoured by national
meso-level theorizations. Additionally, as we accumulation strategies while others will not”.
set out below, the approach also provides a But it is also clear that different regulatory
foundation for an alternative, theoretically functions operate at different spatial scales
nuanced approach to evaluation, though this (Goodwin, 2001: 78; Peck and Tickell, 1992:
point has been little recognized to date. 352), and that a whole variety of social, cul-
tural and institutional forms contribute to
Regulation theory and LRED distinctively local or regional regulatory
effects. Peck and Tickell, for example, high-
From a regulationist standpoint contempo- light the potential role of local growth coali-
rary LRED activity is best viewed within the tions, inter-firm networks, labour market
context of overall patterns of institutional structures and institutions, housing markets,
change and policy experimentation directed venture capital arrangements, forms of local
towards the establishment of some form of governance, local economic policies and
relations in civil society (1992: 353). As a
result economic development activity will

575

DAVE VALLER

also reflect the search for a ‘regulatory fix’ at and regions come to find a degree of “local
the local or regional scale, as localities with economic integrity” (Eisenschitz and Gough,
distinctive economic, social, political, cultural 1993) or “structured coherence” (Harvey,
and institutional histories seek to position 1985) which temporarily stabilizes capitalist
themselves within the context of a broader reproduction in particular times and places.
accumulation strategy (Goodwin, 2001: 78). This emerges from the combination of a
wide variety of processes and practices oper-
Overall, then, regulation theory promises ating at multiple scales, each with their own
significant advances both in contextualizing geographies and spatial structures, and will
LRED activities broadly within the search therefore vary spatially (Painter and Goodwin,
for a resolution to the crisis Fordism and the 1995: 335). For Harvey:
experimentation over new post-Fordist reg-
ulatory forms, and in approaching the diverse At the heart of that coherence lies a
mechanisms and forms of regulation operating particular technological mix – under-
in particular (sub-national) spatial contexts. stood not simply as hardware but also
This has important theoretical implications as organizational forms – and a domi-
for evaluative work. In particular we can nant set of social relations. Together
derive two very broad basic criteria upon these define models of consumption
which to construct evaluations of changing as well as of the labour process. The
policy forms and institutional frameworks. coherence embraces the standard of
These relate to the specific contributions of living, the qualities and style of life,
economic development activities in (i) man- work satisfactions (or lack thereof),
aging ongoing crisis tendencies, and (ii) facil- social hierarchies (authority structures
itating the reproduction of capitalist social in the workplace, status systems of
relations. Clearly, this is not to suggest that consumption), and a whole set of soci-
individual economic development initiatives ological and psychological attitudes
could in any way be evaluated in these towards living, enjoying, entertaining
abstract terms per se. Rather, LRED activity, and the like.
operating in conjunction with an array of
socio-political, institutional and cultural ( Harvey, 1985: 140)
forces and processes, may come to play dis-
tinctive roles in the production of ‘regulatory This, in turn, suggests a distinctive approach
effects’, which are themselves necessarily to evaluation focusing on the extent to
“greater than, or qualitatively different from, which, and the ways in which, LRED activ-
the sum of individual effects” (Painter and ity reinforces, reshapes or even undermines
Goodwin, 1995: 335). Such effects are emer- such coherence in particular places. Clearly,
gent properties of a social system and are LRED activity may be implicated in many
inherently relational. They may incorporate aspects of such coherence, ranging across
specific political compromises, institutional production and consumption relations, social
forms, social expectations or fiscal and orga- and political forms, and cultures and atti-
nizational arrangements, but it is important tudes. Indeed LRED policy may come to
to note that regulatory effects do not emerge play a variety of roles – simultaneously eco-
directly from these individual elements, but nomic, political, institutional and cultural –
through critical interrelations amongst these in the construction of successful regulatory
and other forms, and may therefore be speci- processes.
fied only in terms of the broader system.
Further, as we have outlined earlier, locali-
In this context the starting point for ties and regions are inserted differently in
understanding the ‘meaning’ of LRED activ- wider spatial divisions of labour and regu-
ity is in defining the ways in which localities latory structures, and are characterized by

576

T HE EVAL UAT ION OF L O C AL AN D R EGION AL D EV E LOPMEN T POLIC Y

distinctive social, cultural and institutional Conclusion
forms. Patterns of structured coherence
therefore vary widely between places, and A call for further attention to the theoretical
the contribution of LRED activities to such foundations for evaluation must, of course,
coherence will be similarly diverse.Evaluation be tempered with some degree of caution.
must in turn be a more complex and com- Plainly it would shift the focus away from
prehensive exercise than it has been to date. the basic quantifiable output measures which
For if regulatory processes and effects are dominate contemporary policy discourse,
different in different places, evaluating the and require both policy-makers and academic
contribution of any one particular type of commentators to look beyond immediate
institution or activity across these places may policy structures and prerogatives. It would
involve measuring different things (see Valler generate additional complexity for evaluators
et al., 2000). For example, in considering the in thinking through the emergent and dif-
potential contributions of workfare pro- ferentiated regulatory affects which charac-
grammes to distinct forms of structured terize specific places, and which are far less
coherence in different places we might reflect amenable to communication through basic,
on how the key emphases of workfare-based mechanistic data. In this sense there would
policy, namely “flexibility for enterprise; geo- be no simple off-the-shelf method for evalu-
graphic rescaling of economic and social ation in the broadest sense. There are also
intervention; replacement of entitlements critical conceptual challenges, not least the
with obligations on the part of citizens; and question of how to operationalize a focus on
coalitional power-holding spanning govern- structured coherence, given that the notion
mental, civil society, and profit motivated is based on “loose and heterogenous” foun-
actors” (Krinsky, 2006: 158) might perform dations ( Jessop, 2006: 147).Yet if we wish to
differentiated roles in, for example, respond- move away from “isolated and often disap-
ing to fiscal crisis, channeling and containing pointing findings” we must recognize that
political opposition,enhancing labour market different types of question require different
flexibility, overcoming important institu- types of evaluation, and that ‘success’ or
tional scleroses and managing social polariza- ‘failure’ can only be evaluated in context.
tion (see Valler and Wood, (2009) for further
related discussion). The key point is that the Taking this forward requires significant
contribution of these diverse roles in estab- theoretical and methodological development
lishing or reinforcing effective regulation in to refine high-level abstractions into more
particular places will be distinctive. From this specific middle-range concepts and concrete
viewpoint it is only once we have defined claims. As a starting point, approaching the
the form of structured coherence in a place tendency to structured coherence in particu-
that we may distil the contribution of LRED lar territories would require, first, a detailed
activities in these particular arrangements. and frank assessment of the position of spe-
In turn the broadest evaluative questions cific places within the wider global, national
around LRED activities – of their ‘success’ or and regional division of labor; second, a clear
‘failure’, ‘appropriateness’ or ‘sustainability’ – understanding of the regulatory challenges
can only be determined in context, that is and tensions posed by the location of a par-
with regard to their specific contributions to ticular place – with its distinctive political-
particular accumulation–regulation cou- economic history and socio-cultural and
plings in localities and regions. Evaluative institutional forms – within this broader
work should therefore reflect the particular- environment; and third, a focus on the key
ity of place and the distinctive character of interrelationships between economic, social,
regulation in any given case. political, cultural and institutional relations
which act to mitigate such crisis tendencies.

577