ISSUE 2008/02
FEBRUARY 2008
by Indhira Santos SUMMARY The European Union’s budget review, launched last year, is an
Research Fellow at Bruegel
indhira.santos@bruegel.org
bruegelpolicybrief
IS STRUCTURAL SPENDING
ON SOLID FOUNDATIONS?
opportunity critically to examine EU policies and instruments. Structural
Funds are at the heart of the EU cohesion effort, and absorb almost one third
of the EU’s budget. Their declared aims are economic growth and regional
convergence, but these goals do not always complement each other.
Allocation of Structural Funds is not efficient from a pure growth standpoint
and, although with enlargement cross-country transfers have increased sig-
nificantly, on average almost twice as much redistribution still occurs within
regions as opposed to between regions.
POLICY CHALLENGE
Fig 1: Redistribution through Structural
Funds, country average EU15 European Union policies that aim to
100% foster economic growth and reduce
disparities have value. But these
Percentage of total country flows under Structural Funds
objectives need to be targeted with
80% distinct and suitable instruments. The
separation of the growth and
redistribution functions of Structural
60%
Funds is, therefore, desirable. In terms
of redistribution, Structural Funds
40% should aim at fostering cross-country
economic and social convergence,
leaving subnational allocations to
20% member states. In terms of growth,
investment should be primarily target-
ed at energy, trans-national networks
0%
2000-2006 2007-2013 and research. Finally, the EU needs to
Within-country redistribution: intra-regional improve its capacity to assess the
Within-country redistribution: inter-regional
Cross-country redistribution
performance of Structural Funds, and
EU policies more generally.
EU15 minus Denmark and Luxembourg.
Source: Bruegel, based on DG Regio, National Strategic Reference Programmes and Eurostat.
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

IN 2005 the European Council
02 invited the Commission to ‘under-
take a full, wide-ranging review
bruegelpolicybrief
Sections two and three assess SF
in terms of their growth and
regional redistribution goals.
receive structural funding. Each
region is categorised under one of
two 'objectives': 1) Convergence

covering all aspects of EU spend- Finally, we discuss the policy (formerly Objective 1), which
ing, including the common agricul- implications arising from the accounts for 81.5 percent of total
tural policy, and of resources, analysis. SF and focuses on the least devel-
including the UK rebate’1. The oped regions; and 2) Regional and
Commission launched the review 1. STRUCTURAL FUNDS: Competitiveness and Employment
with a consultation paper in OBJECTIVES AND INSTRUMENTS (formerly Objectives 2 and 3),
September 2007, and will deliver which amounts to 16 percent of
conclusions by the end of 2008 or “... the Community shall aim at total funds and targets regions
spring 2009. reducing disparities between the outside the Convergence Objective
levels of development of the to encourage innovation, entrepre-
This review is an opportunity to various regions and the neurship and environmental
rethink the EU’s policy priorities backwardness of the least protection, as well as to accommo-
1
and instruments. In this Policy favoured regions”. date structural changes4.
Financial Perspectives
2007-2013, Council of Brief, we focus on Structural Funds
the EU 15915/05, (SF), the second largest compo- Treaty establishing the European Two instruments provide
19.12.05.
2
nent of the EU budget. Community, Article 158 resources to meet these
To analyse regional
disparities, we focus on
objectives5:
income transfers. SF are the main instrument of the Although reducing welfare differ-
Enough transfers EU’s cohesion policy. Their ration- ences across regions has been • The European Regional
across regions are
needed in order to ale is 1) to encourage economic one of the EU's main objectives Development Fund (ERDF),
foster convergence not growth; and 2) to accelerate the since its foundation, there was at which supports infrastructure
only in terms of
income, but economic
process through which poor the start little EU money to match development and productive
development more regions converge with their richer the goals. Structural spending rep- investment; and
generally. neighbours in terms of economic resented just three percent of the • The European Social Fund
3
Regions are defined at and social development. EU budget in 1970 and ten percent (ESF), which facilitates the
the NUTS II level. NUTS
(Nomenclature of in 1980. In 1986, the Single integration of the unemployed
Territorial Units for We examine the performance of SF European Act recognised regional and disadvantaged into the
Statistics) is used in relative to these goals2. Results policy as an EU-level task, leading labour market.
cohesion policy to sub-
divide national territo- show a disconnection between the to a significant increase in SF.
ries. NUTS II refers to stated objectives and instruments All regions, irrespective of their
areas with a minimum
population of 800,000 of EU structural policy, and indi- EU cohesion policy has two Objective, are targeted by both
inhabitants and a maxi- cate that SF allocation is neither financial arms: Structural Funds instruments. Since there is no one-
mum of three million, efficient for maximising EU growth (SF) and the Cohesion Fund (CF). to-one correspondence between
and largely follows
administrative struc- nor an effective tool for The former is allocated to regions3, Objectives and instruments,
tures in member states. redistribution. There is no while the latter is targeted towards growth and redistribution policies
4
There is a third objec- evidence that the regions attract- countries lagging behind. We focus are served by the same funding
tive, European
Territorial Cooperation, ing the bulk of SF are those where exclusively on SF. instruments. We assess the per-
which represents 2.5 the return on capital is highest. formance of SF successively along
percent of total SF and Furthermore, on average, almost For the 2007-2013 financial these two dimensions.
comprises trans-
national programmes. three quarters of all SF received by period, €278 billion has been allo-
This is excluded from regions – setting aside transfers cated to SF, representing 29 per- 2. DO STRUCTURAL FUNDS
the current study.
5
across countries – are original SF cent of the total EU budget. During MAXIMISE EU GROWTH?
In 2000-2006, there
used to be three more
contributions from taxpayers in 2000-2006, SF accounted for
instruments: EAGGF those same regions. €233 billion or 31 percent of total Existing evaluations of EU struc-
Guarantee (agricul- commitments. tural policy examine if SF promote
ture), EAGGF Guidance
(agriculture) and the Section one of this Policy Brief growth in beneficiary regions, but
FIFG (fisheries). introduces EU structural policy. All EU regions, rich and poor, do not assess if SF are distributed
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

evidence on the impact of SF on

Figure 2: Return on capital and Structural Funds net benefits per capita
8000
growth is mixed7.
03

bruegelpolicybrief
Structural Funds net benefits per capita 2007-2013, PPP, euros

7000
The SF contribution to growth may
6000 be curtailed by several factors. The
5000 literature points to the displace-
ment of national regional aid, or to
4000
regions’ strategic behaviour as
3000
they implement projects that have
2000 other objectives besides growth,
1000 either to pursue rent-seeking activ-
0
ities or to retain their SF eligibility.
0.5 1 1.5 2 2.5 Similarly, the potential lack of
-1000
coherence with national policies is
-2000 also cited as a limitation.
Marginal product of capital relative to EU27 average

Data refers to regions. Each new member state is considered as one region. But three other factors should be
Source: Bruegel, based on DG Regio, National Strategic Reference Programmes, Eurostat and Caselli
and Feyrer (2007).
highlighted. As this Policy Brief
shows, the redistributive role of SF
efficiently to start with. Since – with one instrument creates channels resources away from
resources are limited, more money inefficiency. economic centres – reducing
for one region necessarily means growth potential – instead of
less for others, so the key bench- However, while Structural Funds focusing first on the efficient use
mark is to see if each euro is may not maximise growth, they of funds. Furthermore, with the
invested where it will yield the may still contribute to it. Simple EU’s eastern enlargement, regional
highest benefits for the EU as a correlations show that regional net policy resources are spread thinly.
whole. benefits per capita from SF are not The population of regions where 6
Enderveen et al
strongly related to growth, at least GDP per capita is 75 percent or (2006) and Cappelen
et al (2003) found SF
For transfers to be growth-max- when not conditional on any other less of the EU average increased to be ineffective, except
imising, they need to be allocated factors (Figure 3). This is consis- from 68 million to 116 million (or for countries/regions
with an ‘appropriate’
to regions where capital is most tent with other results in the litera- from 18 to 25 percent of the EU27 institutional framework.
productive, ie where the marginal ture6. However, overall, the population). In this new context, it
7
product of capital (MPK) is high- The literature on this
Figure 3: Regional GDP per capita growth rate by Structural Funds net topic has three main
est. Do SF pass this efficiency limitations. First, the
test? We estimate each region’s benefits per capita effects on growth of
these policies may take
MPK relative to the EU27 average, 9 time. Second, it is diffi-
Average annual growth rate (2003-2004), %

and calculate their net gains from cult to make a causal

SF (see box on page 8). 7 interpretation of
results, as SF are not
allocated randomly.
5
As Figure 2 shows, SF are not allo- Third, SF-associated
opportunity costs are
cated to regions with the highest 3 not taken into account.
return on capital. This holds true EU taxpayers finance
even when limited to non- 1 SF and the money could
be used elsewhere. The
convergence regions. At best, the -2000 -1000 0 1000 2000 3000 4000 5000 6000 7000 relevant question from
-1
relationship is flat; if not, it is a policy perspective is
if SF boost overall
downward-sloping, meaning that -3 growth above what it
lower-MPK regions actually get Net benefits per capita (2000-2006), euros, PPP would be in their
more money than those with high- FI PT SE AT NL absence. See
IT UK DE BE ES Enderveen, et al
er MPK. The pursuit of conflicting GR FR IE DK LU (2002) for a review of
goals – growth and redistribution Source: Bruegel, based DG Regio, National Strategic Reference Programmes and Eurostat.
this literature.
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

makes more sense to concentrate
04 resources in lagging countries –
not regions – and to let domestic
bruegelpolicybrief
rationale of EU structural policies.
If most of the redistribution takes
place internally within member
Italy and Ireland became net con-
tributors. Today, the main recipi-
ents are the member states that

redistribution take place at the
national level. Finally, as dis-
cussed in the following section,
the nature of the redistribution
that takes place through SF also
limits the growth potential of
structural policies.
3. REDISTRIBUTION
Redistribution through structural
policies takes place on two levels:
1. Across countries, correspon-
ding to net country SF bal-
ances; and
2. Within countries, through:
• ‘Inter-regional redistribution’,
related to the funds received
by a region from other
region(s) in the same mem-
ber state.
• ‘Intra-regional redistribution’,
equal to the SF received by a
region and financed by tax-
payers in that same region.
Whether SF redistribute financial
resources within or across
countries is important for the
states, it is difficult to justify the joined the EU in 2004 and after.
EU's active role. If, on the other
hand, redistribution were largely On average, poorer regions receive
across countries, then SF would be more SF per inhabitant net of their
carrying out the same task as the contribution to the policy (Figure
Cohesion Fund, with the difference 5). This relationship is stronger
that the former is channelled among convergence regions but,
through regions. on average, a regional per capita
income increase of
International transfers ‘Pursuing conflicting €100 is associated
in structural policy
goals – growth and with a decrease of €7
arise because some in net benefits per
countries are net con- redistribution – capita. Regions in
tributors while others creates inefficiency.’ Cohesion Fund
are net beneficiaries countries benefit more
(Figure 4). from structural policies, even if
those regions have similar
But net balances have changed incomes to regions in non-
over time, mainly due to enlarge- Cohesion Fund countries. This is
ment and to differences in eco- also true for regions in countries
nomic growth. Net positions of all that have a more unequal per capi-
old member states worsened ta regional income distribution8.
between 2000-2006 and 2007-
2013, although Luxembourg, Regions, like states, also have net
Belgium and the Netherlands balances. The nature and extent of
remain the three main net per the transfers that take place at
capita SF contributors. Among net sub-national level are, however,
beneficiaries, changes were espe- poorly understood. Using data for
cially significant for Spain (81 per- 2000-2006 and 2007-2013 and
cent fall in net receipts), while thirteen member states9, we

Figure 4: Structural Funds average annual net benefits per capita

400

2000-2006 2007-2013
300

200
8
These results are
Euros PPP

obtained using
100
regression analysis.
9
New member states CY IT DE UK FI FR SE AT DK IE NL BE LU
0
are not included in the CZ EE HU LT SK LV PL PT MT BG SI GR RO ES
analysis since most of
their SF programmes
are national. Countries -100
with only one NUTS II
region – Denmark and
Luxembourg – are also -200
excluded.
Source: DG Budget. 2006 Allocation of Expenditure by Member State (2007), DG Regio and own calculations
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

the sample, redistribution across

Figure 5: Regional Structural Funds net benefits per capita by GDP per capita
7000
regions is limited.
05
Net benefits per capita (2007-2013), euros, PPP

bruegelpolicybrief
Two examples may help interpret
5000
Convergence threshold the results. For Spain, a net benefi-
ciary of SF during the 2007-2013
3000
period, 19 percent of SF received
originates from other EU countries.
1000
Of the remaining 81 percent, two
0 10000 20000 30000 40000 50000 60000 70000
thirds of the money spent in each
-1000 region comes from that same
region. The UK, meanwhile, is a net
-3000 contributor to SF. Of the SF
GDP per capita (2004), euros, PPP
received by each UK region, 97
IE FI PT SE AT NL IT UK DE BE
ES GR FR DK BG CZ EE CY LV percent was simply contributions
from local taxpayers.
Data refers to regions. Each new member state is considered as one region.
Source: Bruegel, based DG Regio, National Strategic Reference Programmes and Eurostat.

identify the extent to which Figure 6: Level of redistribution through Structural Funds, 2000-2006
redistribution through structural
policy takes place across Belgium
Cross-country
Netherlands
countries, within regions and Within country: inter-regional
UK
between regions. Within country: intra-regional
Austria
Sweden
In 2000-2006, on average, 27 per- France
cent of SF flows were inter-country Germany
transfers. But, of the rest, three Finland
quarters were intra-regional. With Italy
enlargement, inter-country trans- Ireland
fers rise for 2007-2013, but intra- Spain

regional redistribution is still twice Portugal

Greece
as large as redistribution between
-100 -50 0 50 100 150
regions in the same country (see
% total flows under Structural Funds
Figure 1 on the front page)10.
Figure 7: Level of redistribution through Structural Funds, 2007-2013
Results by member state are
shown in Figures 6 and 7. There Netherlands
Cross-country
are some notable differences Ireland
Within country: inter-regional
between countries. For example, in Belgium
Within country: intra-regional
Austria
Italy, which practically breaks
Sweden
even in structural policies and has
UK
major inter-regional disparities,
France
inter-regional redistribution is Finland
relatively greater although still Germany
only 41 percent of SF flows in Italy
2007-2013. For net beneficiaries, Spain
10
most of the resources received are Greece Including co-
financing reduces the
transfers from other countries, Portugal
redistributive effect of
especially in Greece and Portugal. -100 -50 0 50 100 150 the funds. Results are
available from the
But, for the majority of countries in % total flows under Structural Funds
author.
Source for Figs 6 and 7: Bruegel, based DG Regio, National Strategic Reference Programmes and Eurostat.
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

This exercise reveals two addition- two re-emerges here. If most of the ment, cross-country transfers have
06 al points. Two regions with similar funds received by a region origi-
per capita income, but located in nate from that same region, it is
on average increased significantly
to 40 percent of total flows of SF
bruegelpolicybrief

different countries, can benefit hard to imagine that they can have and intra-regional redistribution
very differently from SF. Take, for large growth effects. Behind those decreased. In a more diverse EU,
example, Hainaut (Belgium) and positive growth results is the unre- this is a step in the right direction.
Galicia (Spain), both under the alistic assumption that those Yet, almost three quarters of the
Convergence Objective. Both have resources would have otherwise intra-country redistribution still
a GDP per capita (PPP) of about been put to no productive use. To occurs within regions.
€17,400, but while really measure the
Hainaut is a net con- impact of SF, one The evidence we present points at
tributor (€388 per
‘Two regions with needs to show that the need to rationalise structural
capita), Galicia is a similar income but in they generate bene- policy and echoes proposals put
net beneficiary different countries can fits above and forward in the Sapir Report
(€949 per capita). benefit very differently beyond what those (2004). In the case of growth, we
This difference resources would provide further evidence of the
comes about from Structural Funds .’ have accomplished need to concentrate SF in those
because, while SF if there were no SF. areas with higher growth potential;
benefits depend on regional in the case of convergence, our
income, contributions further The main lesson from this analysis findings strengthen the case for
depend on how much the country is that the inter-regional separating growth from
as a whole contributes to the EU redistribution that SF strive for redistribution in cohesion policy
policy. If a region is very different takes place only to a limited and rethinking the role of the EU in
from the country's regional aver- extent. For net beneficiary subnational redistribution.
age, there can be a mismatch countries, a positive net balance
between what it receives and what reflects redistribution from Rationalise objectives
it has to contribute to SF. regions in richer countries. For Over the years, there has been a
most countries, SF largely redis- long debate about the appropriate
Moreover, even for similar regions tribute resources among individu- role of the EU across policy
in countries with comparable als within regions. This finding domains, especially in
income, there is a variation in the renders questionable the ability of redistribution. On the one hand,
net benefits they derive from SF. EU structural policy to achieve its redistribution from richer to poorer
For instance, Guadeloupe (France) objectives of promoting growth countries is often seen as an EU
and Sicily (Italy) both have a GDP and reducing regional disparities. public good since it fosters
per capita of approximately convergence, creating major bene-
€14,000, yet the former receives 4. POLICY PROPOSALS fits for the rest of Europe in the
– net per capita – 33 percent more form of new and wealthier markets
funds. An immediate implication of This Policy Brief has found that: and steadier democracies.
this result relates to the logic
behind redistribution in SF. If the • SF resources are not spent in a On the other hand, subnational
goal is regional convergence, way that optimises their poten- redistribution is a more divisive
would it be better to determine net tial to generate EU-wide growth; issue. Subsidiarity, horizontal
benefits (and not only benefits) and equity across countries, transac-
on a regional basis? How should • Regions themselves pay for tion and information costs and the
one account for non-income much of the SF they receive, need for coherence with national
factors that disadvantage one limiting the extent of inter- macroeconomic policy, all suggest
region more than others? These regional redistribution. that national governments are
remain open questions. best positioned for this task – pro-
Results highlight the dynamic vided they do not violate competi-
One subject discussed in section nature of the EU. With enlarge- tion rules. New member states are
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

a case in point. These countries
allocate most SF to thematic
programmes, ie infrastructure or
actually takes place within regions
suggests that the instruments
used to tackle regional disparities
forward in support of the current
system. The first is that the
process of managing SF empowers
07

research, not regional ones. This
does not rule out the possibility of
countries decentralising project
implementation and evaluation. It
rather emphasises the need for
the EU to use country-level criteria
for the allocation of funds.
Find the right level
To promote growth efficiently, SF
should be allocated where the pro-
ductivity of capital is highest and
to projects with clear spillovers.
Currently, allocations appear inef-
ficient and trans-national
programmes – with major poten-
tial EU-wide growth externalities –
represent just 2.5 percent of total
funds. Re-orientation of structural
policies in line with the Lisbon
objectives, as in the last budgetary
negotiations, was a step in the
right direction. Lisbon policies
whose benefits are not limited to
the spending country – namely
transport and energy networks, as
well as research – have a clear EU
dimension and are good candi-
dates for boosting growth.
We have argued that the economic
rationale for EU involvement in
intra-country redistribution is
questionable. But the fact that
most domestic redistribution
bruegelpolicybrief
are not appropriate either. National regional governments. But decen-
governments have their tax and tralised management of SF is com-
transfer systems for redistributing patible with the first-best where
income within their borders. EU within-country redistribution is a
convergence policy would be more member state task. The second
effective if it focused on reducing argument is that SF increase the
gaps between countries instead of visibility of the EU. However,
regions. Cohesion Funds, which allocate EU
funds to countries, also do this, as
Improve matching of goals and does any project generating tangi-
instruments ble benefits for EU citizens. Our
The current structure of SF recog- policy proposals, therefore, do not
nises that different regions have undermine these benefits.
different needs. But it fails to
translate this into an effective Finally, it should be stressed that
one-to-one correspondence of assessing the evidence of SF per-
goals and instruments. formance is ham-
The ability of SF to gen- ‘Structural Funds pered by the lack of
erate growth is limited largely redistribute appropriate informa-
by their parallel redis- tion. Ex-post alloca-
tributive role. resources among tion of SF is reported
individuals within by the Commission
This presents policy- regions.’ only at country level,
makers with a choice. making it difficult to
They can try to quantify trade-offs analyse the use of these funds on
and prioritise, which is difficult. the ground. Moreover, even where
How much should the growth and data is available, there is no clear
convergence objectives count institutional or conceptual frame-
when assessing SF projects? Or work for evaluating EU policies.
they can separate the growth and Addressing these shortcomings
redistribution objectives, and should be a main concern in the
design appropriate instruments future, as resources are sliced
for each separately. thinner and thinner and new
spending priorities emerge.
Two arguments are often put

REFERENCES:
Cappelen A., F. Castellacci, J. Fagerberg, B. Verspagen, 2003, ‘The impact of EU regional support on growth and
convergence in the European Union,’ Journal of Common Market Studies, 41, 4, 621-644.
Caselli, F. and J. Feyrer, 2007, ‘The marginal product of capital’, Quarterly Journal of Economics, 122(2), 535-567.
Ederveen S., J. Gorter, R. de Mooij and R. Nahuis, 2002, ‘Funds and games: the economics of European cohesion policy’,
CPB working paper, 1-103.
Enderveen, S., H. de Groot and R. Nahuis, 2006, ‘Fertile soil for Structural Funds? A panel data analysis of the conditional
effectiveness of European cohesion policy’, KYLOS, Vol. 59, No. 1, 17-42.
Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Viñals and H. Wallace with M. Buti, M. Nava and P.
Smith, 2004, ‘An agenda for a growing Europe: the Sapir report’, Oxford University Press.
 IS STRUCTURAL SPENDING ON SOLID FOUNDATIONS?

08 BOX 1 Data and Methodology regional level is the same as that observed across
countries11.
bruegelpolicybrief

Marginal Product of Capital

The MPK is the additional output resulting from the use of
one more unit of capital. Output (Y) depends on the
inputs used – capital (K) and labour (L) – and total fac-
tor productivity (A). Total factor productivity accounts for
other factors such as technology, education and the qual-
ity of economic institutions. We assume that the
production relationship takes the following form:
α 1−α
Y = AK L (1)
where α is the share of capital in output and 1– α that of
labour. α is assumed to be 0.35, in line with existing esti-
mates.
Using 2004 data from Eurostat, we compute for each
region the ratio of its MPK to that of the average EU27.
According to its definition, the MPK (r) is:
α −1
⎛K⎞
r = α A⎜ ⎟
⎝L⎠ (2)
That is, keeping everything else constant, the MPK of a region
increases with total factor productivity but decreases with
capital per worker due to ‘diminishing returns’ (ie an addition-
al unit of input yields less and less additional output).
Net Benefits Structural Funds
Net benefits, adjusted for differences in purchasing
power, are the difference between the SF received by a
region (‘benefits’) and that region’s input into the
financing of the policy (‘contribution’).
Benefits
The regional allocation of SF is derived from cohesion
policy plans prepared by member states. Funds are dis-
tributed by region, except for multi-regional (such as
Objective 3 in 2000-2006) and national thematic
programmes. On average for 2000-2006, national
programmes with no regional breakdown represented
one tenth of total SF, while for 2007-2013, they are less
than one percent. In these latter cases, we allocate bene-
fits proportionally to population size within the relevant
group of regions.
Our analysis uses programmed instead of actual regional
expenditure, as data on real spend is not readily available.
If poorer regions execute a smaller proportion of the
funds originally assigned to them, this Policy Brief over-
estimates the extent of inter-regional redistribution.

Total factor productivity is estimated from the production Contributions

function (1) as: We use each region’s share of its country’s disposable
income to distribute the budgetary cost of structural
Y
A= α
policies12. In turn, the contribution of each country to SF
⎛K⎞ depends on the share of the overall budget financed by
L⎜ ⎟
11
We use other meth- ⎝L⎠ (3) that member state. For example, in the case of France, on
ods to calculate the average, it paid for 17.2 percent of the EU annual budget
ratio of regional MPK to Based on country estimates (Caselli and Feyrer, 2007), in 2000-2006. As money is fungible, this means that it
that of the average the capital per worker ratio is taken to be 1.1 times the also paid for 17.2 percent of SF policy. This contribution
EU27. The one we ratio of GDPs per capita. The assumption made is that the is then divided among French regions proportionally to
present here is the
most favourable for relationship between K/L and income per capita at the their disposable income.
poor regions as they
have relatively higher
estimates for MPK.
The author is grateful to all those who reviewed drafts of this policy brief and exchanged arguments about its
content, and would like to thank Concetta Mendolicchio at Bruegel for excellent research assistance.
12
We also estimate
regional contributions
using GDP per capita. Bruegel is a European think tank devoted to international economics, which started operations in Brussels in 2005.
This is important for
regions where there is a
It is supported by European governments and international corporations. Bruegel’s aim is to contribute to the quality
large difference of economic policymaking in Europe through open, fact-based and policy-relevant research, analysis and discussion.
between residents’
income and the value of © Bruegel 2008. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted in the original
production due, for language without explicit permission provided that the source is acknowledged. The Bruegel Policy Brief Series is pub-
instance, to a large
lished under the editorial responsibility of Jean Pisani-Ferry, Director. Opinions expressed in this publication are those of
number of commuters
the author(s) alone.
(eg Brussels, inner
London). However, the
country results do not Visit www.bruegel.org for information on Bruegel's activities and publications.
change significantly. Bruegel - Rue de la Charité 33, B-1210 Brussels - phone (+32) 2 227 4210 info@bruegel.org