EU Structural Funds and Spain’s Objective 1 Regions: An Analysis Based on the Hermin Model by Simón Sosvilla-Rivero* DOCUMENTO DE TRABAJO 2005-24

October 2005

FEDEA and UCM.

ABSTRACT

This paper presents an empirical evaluation of the economic effects of the Structural Funds received by the Spain’s Objective 1 regions through the Community Support Framework (CSF). The empirical results suggest that the funds have significantly contributed to economic growth, as well as to wealth and employment creation. Compared to a scenario without CSFs, the successive investment programmes from 1989 to 2006 would represent an average increase of 0.56 percentage points in the growth rates of the Spanish recipient regions. This increase would translate to an average increase in per capita income of 425 euros at 1999 prices. In terms of the labour market, the CSF has maintained or generated an increase of 1.46 per cent in employment compared to an alternative scenario without CSF. This in turn would represent an average reduction of 0.74 percentage points in the unemployment rate during the aforementioned period.

Key words: European Union, Structural Funds, Regional Convergence, Spain

JEL codes: C51, R58, O11, O52

1. Introduction

With the growth in the number of members of the European Union (EU), particularly following the accession of relatively poor economies (firstly Ireland and Greece, then Spain and Portugal), the marked disparities in income and prosperity levels between regions acquired increasing importance on the Community agenda, becoming what is today known as the ‘regional question’. The creation of the Single European Market in 1992 and the prospect of Economic and Monetary Union added further to the pressure for a policy to offset some of the potentially negative effects in the outermost parts of Europe and ensure the active participation of the potential losers.

This growing interest in the economic and social approximation of the Member States (the so-called “European cohesion”) led to the reformulation and expansion in 1987 of a broad range of policies aimed at structural reform and economic growth and taking the form of a programming document known as the Community Support Frameworks (CSF) which, as the name suggests, were designed to aid the recovery of the least developed regions (with the real per capita GDP below 75 per cent the average, known as Objective 1). The balance of resources provided by and obtained from the overall community budget could not have been better for Spain within the primary terms of the implied flows, going back to the early 1990s with a net balance close to 6 billion euros per year it its favour. The structural and cohesion aid received by Spain has been of a similar, or even greater, magnitude, and Spain has been the country that has most benefited, in absolute terms, from the community’s cohesion policy.

This paper provides an empirical evaluation of the macroeconomic effects of the CSFs received by Spain’s Objective 1 regions. To that end, using macroeconomic models based on the HERMIN-Spain model for each region, we estimate the demand effects recorded during the expenditure periods as well as the long-term supply-side effects stemming from the increase in public capital, private productive capacity and the human capital experienced by each regional economy as a result of the structural funds received.

The paper is organised as follows. Section 2 gives a brief theoretical overview of the anticipated evolution of the spatial localisation of economic activity in a context of economic integration, as well as the implications this has for regional policy. Section 3 outlines briefly the methodology used, while Section 4 presents the results of the evaluation for the successive programming periods (1989-1993, 1994-1999 and 2000-2006), comparing the effects of EU structural aid with the situation that would have prevailed had the aid not existed (reference scenario). Finally, Section 5 offers some final remarks.

2. Regional integration and localisation of economic activity

The orthodox response offered by economic theory to the evolution over time of regional inequalities comes from neoclassical growth theory: if economies differed purely on their initial capital-labour relations, we should see higher growth in poorer economies than in richer ones. Put another way, the convergence hypothesis would hold true. This would be due to the assumption of decreasing returns for the capital factor: since the marginal productivity of capital would be higher in economies with a lower initial capital-labour relation, these should enjoy a higher rate of growth in their capital-labour relation and, therefore, in per capita output (see, for example, SALA-I-MARTÍ, 1996). However, this would only hold true if, as mentioned, the economies differed purely in terms of their initial capital-human relations. If they differed in other areas also convergence would be merely conditional (‘conditional β- convergence’), in the sense that an economy’s growth rate is directly related to the distance with respect to its stationary state. Furthermore the approach is somewhat sceptical as regards the usefulness of regional policy, given that the convergence speeds obtained are essentially the same both for countries that have adopted such a policy and those that have not. It has even been suggested that public investment should be targeted on richer regions since this would facilitate growth as well as efficiency (SALA-I-MARTÍ, 1997).

From the empirical standpoint, the evidence does not lend itself favourably to the convergence hypothesis for very long time-periods and broad sets of countries. Indeed in the long run a tendency towards the formation of convergence ‘clubs’ would appear to exist: economies would tend to be either very rich or very poor - with fewer countries in the intermediate income bracket - and the gap between rich and poor would grow wider (see QUAH, 1996, for example). The concept of conditional β- convergence has been considered to be of little relevance, since what is really of interest is the relative evolution of per capita output in an economy compared to the output of other countries and not to its own stationary state value, which can differ considerably from one economy to another.

Moreover, recent years have seen a series of contributions whose common denominator is that they endeavour to clarify the factors influencing the localisation of economic activities in space: the so-called New Economic Geography (NEG). Although strictly-speaking not ‘new’ (it includes aspects already addressed in localisation theory, regional science, economic history or international trade theory), NEG is of interest in that it provides a common framework, based on hitherto dispersed contributions, to analyse a phenomenon which has been studied relatively little using orthodox approaches. A simple presentation of the early results of NEG can be found in KRUGMAN (1991a), for a more comprehensive overview see Ottaviano and Puga (1998) or, in much greater detail, FUJITA, KRUGMAN and VENABLES (1999). A critical appraisal of NEG is provided in NEARY (2001).

Under the neoclassical approach, based on assumptions of perfect competition and constant returns of scale, economic forces operate in an environment that is independent of history and hence the behaviour of an economy depends largely on exogenous elements: tastes, technology and supply of factors. Conversely, the NEG underlines the role of growing returns of scale, whose very existence gives rise to cumulative dynamic processes (see KALDOR, 1972, for a general discussion of the importance of these processes). Thus, the presence of growing returns would render the notion of ‘optimal’ allocation of resources meaningless: the behaviour of the economy would depend on events in previous periods and the role of history and expectations acquires more importance, which in turn could lead economic activity to become concentrated in space.

The starting point for NEG models is the existence of companies that are subject to growing returns of scale and transport costs both for the end goods produced and the intermediate inputs used. The interaction of both elements with the size of the market demand - greater demand entailing easier access to those requiring the end goods and also to the intermediate input suppliers- would condition localisation decisions. Hence, in the context of an integration process a lowering of transport costs should, in principle, lead economic activity to localise in areas with lower production costs. However, it would also lend itself to a concentration of production in one single place - where demand is highest - in order to harness economies of scale. Thus, if we have two countries (‘centre and ‘periphery’), a small reduction in transport costs would cause production to concentrate in the ‘centre’ due to the higher demand and the presence of economies of scale. Equally, a major reduction in transport costs would cause the concentration to occur in the ‘periphery’, due to the lower production costs (KRUGMAN and VENABLES, 1990).

However, the above analysis would be incomplete given that the market size would be exogenous: why would some markets be ‘big’ and others ‘small’? This would lead us to consider the presence of mechanisms of cumulative causation: once a process of concentration of the activity in a given place commences, the situation would tend to become consolidated over time. NEG essentially advocates two mechanisms:

* Drag effects of labour mobility (KRUGMAN, 1991b). A concentration of companies in a given location will cause demand for labour to increase. The ensuing higher salaries would encourage workers to migrate to this area, leading to higher demand on the part of these workers, increased profits, and greater capacity to attract new businesses.

* Drag effects of intermediate inputs in an input-output structure framework (KRUGMAN and VENABLES, 1995; VENABLES, 1996). The concentration of firms in a given location would produce forward drag effects (that is, increased production of intermediate inputs in all sectors in response to the higher end demand from a given sector) and backward drag effects (increased production of inputs by a given sector in response to the increased demand from all sectors).

In both cases there would be a clustering effect – the concentration of firms in a given location will tend to generate an additional concentration. However, there might also be a dispersal effect - the possibility that the tendency towards concentration would reverse and the economic activity would spread geographically due to the presence of immobile production factors, such as natural resources or labour, if the latter were not fully mobile (this would be of particular importance in the case of Europe, in comparison to the United States). Thus, the concentration of production would lead to an increase in the price of the immobile factors compared to areas where the production does not occur. Since this difference is not eliminated because of the immobility of the factors, firms would have an incentive to move away from the congested areas.

Economic activity localisation patterns in space are thus the outcome of these two complementary forces, agglomeration and dispersal. In the event of high transport costs, the desire to supply markets locally would lead firms to locate in different regions. However, with medium-range transport costs agglomeration forces would prevail and the drag effects would lead to a geographical concentration of the activity. With low costs, dispersal forces would prevail and firms would locate according to the price of immobile factors (PUGA, 1999). Moreover, to the extent that an economic integration process would lead to lower transport costs and increased spatial mobility of productive factors, its effects on spatial localisation of the activity would, in principle, be ambivalent. Hence, the interaction between transport costs and factor mobility is crucial when determining production and trade patterns (NORMAN and VENABLES, 1995).

NEG has focused its attention of late on the incorporation of aspects addressed by endogenous growth theory. MARTIN and OTTAVIANO (1999) study the interrelation between companies’ localisation decisions and growth, which is understood as meaning the creation of new companies through R+D generation in a 2-region model (North and South). Among their main findings, the authors note that if the spill-over effects associated with R+D are global (i.e. extending to both regions), localisation would not affect growth, even though higher growth would lead to capital flows towards the South. Conversely, if the spill-over effects associated with R+D are local (i.e. affecting only the region where the innovation takes place), a lowering of transport costs would encourage firms to concentrate in the R+D location, which in turn would increase innovation and growth.

Lastly, NEG has failed as yet to fully develop an in-depth analysis of the implications of the theory for regional policy. In general, however, one can say that to the extent that we cannot know, in the absence of a policy, whether too much or too little agglomeration exists, it would be difficult to predict what direction regional policy should take.

One of the main instruments of EU regional policy is the promotion of public infrastructure, which not only serves to increase economic output but presumably also increases the productivity of private production factors (see ASCHAUER, 1989). PUGA (2002) notes the potentially ambiguous effect of transport infrastructure policies on convergence: although improved connections between two regions with different levels of development would help firms in the poorer region to access inputs and markets in the richer one, they would also help companies in the latter supply the poorer region’s markets from their original location, potentially harming its industrialisation prospects. For their part, MARTIN and ROGERS (1995) point out that, if the aim is to foster convergence by poorer regions, the focus should be placed on funding what the authors call ‘internal’ infrastructure (i.e. which facilitates internal trade) and not ‘international’ infrastructure (i.e. which facilitates international trade), given that if significant growing returns exist the international infrastructure in the poorer region (the one with the poorest internal infrastructure) would propitiate the concentration of production in the rich region.

3. Methodology for evaluating regional effects

3.1. The HERMIN model

As stated above, we have evaluated the macroeconomic effects using an adaptation to Spain’s regions of the HERMIN model.

In this conventional model with Keynesian underpinnings the expenditure and income distribution blocks generate standard income-expenditure mechanisms. However, the model also incorporates various neoclassical characteristics, associated particularly with the supply-side block. Thus, private sector output is not determined solely by demand but rather is influenced also by price and cost competitiveness, in a context of companies striving for the lowest productive costs (BRADLEY and FITZGERALD, 1988). In addition, it uses a constant elasticity of substitution (CES) production function in which the capital/labour ratio is the relative price of both factors. Lastly, the inclusion of a Phillips curve mechanism in the wage negotiation introduces further relative price effects into the model.

The models used in this paper have all the characteristics of the HERMIN model, in addition to those of the different regional economies, which are treated in all respects here as small economies.

3.2. Supply and demand effects

Given that the study aims ultimately to identify and model the channels through which European aid can affect (and eventually speed up) economic growth in Objective 1 regions, we differentiate between demand and supply effects.

In terms of the demand side, EU-funded projects represent a stimulus for the economy on account of the increased public spending, which is transmitted directly to demand and, consequently, to output. Also stimulated are employment, earnings, prices and salaries. For their part, the supply-side effects operate via costs, productivity and competitiveness, stimulating production, cutting imports and increasing exports. Moreover, the increase in productive capacity helps mitigate inflationist pressure originating on the demand side.

¹ See Herce and SOSVILLA-RIVERO (1995) for a more detailed description of the Spanish version of the model, and HERCE and SOSVILLA-RIVERO (1994) for a discussion of the macroeconomic treatment of CSF funds.

² The Phillips curve describes the negative relationship between unemployment rate and wage inflation: high unemployment is associated with low wage increases while a low unemployment rate is associated with higher wage inflation.

In this work the possible effects are grouped together by the programmes involved:

a) public spending on infrastructure, the main effect of which is a reduction in transport and other communications costs, leading to lower production costs and greater competitiveness, while also stimulating long-term increases in production and employment (GRAMLICH, 2004).

b) spending on human resources: this programme increases the efficiency and productivity of the beneficiaries, reducing existing firms’ costs through learning by doing, increasing product quality and encouraging the creation of new firms to harness the aforementioned increase in efficiency and productivity.

c) aid for productive investment, to foster activities considered important and desirable, leading in the long run to higher levels of production, exports and employment.

We assume that the economic benefits derived from each programme manifest themselves in the form of externalities and we try to capture them by modifying the model’s key equations (primarily, the production and factor demand functions). Specifically, two kinds of externality are taken into account: the increase in private factor productivity and the improved quality of the product supplied by the private sector.

Regarding the first type, if we consider the following CES production function:

\[O = A \left\{\delta \left(\exp \left(\lambda_ {L} t\right) L\right) ^ {- \rho} + (1 - \delta) \left(\exp \left(\lambda_ {K} t\right) K\right) ^ {- \rho} \right\} ^ {- (1 / \rho)}\]

where O, L and K represent value added, employment and capital stock respectively, A is a scale parameter, is the elasticity of substitution, δ is a factor intensity parameter and and are the rates of technical progress incorporated in labour and capital respectively, respectively, the externality can be incorporated by endogenising the scale parameter as follows for investment in public infrastructure (KGINF), human capital (KH) and private sector (K), respectively:

\[A _ {t} = A _ {0} (K G I N F _ {t} / K G I N F _ {0}) ^ {\eta 1} (K H _ {t} / K H _ {0}) ^ {\eta 2} (K _ {t} / K _ {0}) ^ {\eta 3}\]

where subscripts t and 0 represent the cumulative stock with and without European aid, and and represent the corresponding elasticities.

The second type of externality operates directly through the impact each programme has in terms of improving industrial output quality (in turn producing greater external demand for the goods) and also indirectly through increased flows of direct foreign investment, stemming from the availability of more highly-qualified technical and scientific personnel and better infrastructure (PORTER, 1986), and the ensuing modernisation of the equipment and production techniques of the companies and their greater inclination to export. To capture this kind of externality, we relate the growth in infrastructure stock, the increased human capital and greater availability of sectorial private capital with the measure of the external demand used in the HERMIN model (OW, a key variable in determining the level of production of the tradeable sector) as follows:

\[O W X = O W (K G I N F _ {t} / K G I N F _ {0}) ^ {\eta 1} (K H _ {t} / K H _ {0}) ^ {\eta 2} (K _ {t} / K _ {0}) ^ {\eta 3}\]

In the empirical application, the following values are adopted for the different elasticities: for η 1 the value is estimated from an enlarged production function with public capital for each of the regions studied by SOSVILLA-RIVERO and HERCE (2002) (Table 1); for η 2 the value ^0.07 is used ^(based on estimations of the social returns of education and vocational training made by CORUGEDO et al. (1992), and for the value used is 0.10, based on macroecnomic information on CSF 1989-1993 provided by HERCE (1994).

This manner of incorporating supply effects into a conventional economic model is, without doubt, an initial ad hoc attempt but it has been used repeatedly in empirical literature of this type. In order to limit the risks, we have chosen the most moderate of the elasticity values suggested in the literature, while in the simulation carried out we have made the effects mature progressively. Clearly, the results regarding the supply-side effects of European aid will depend on the size of the externalities and the speed with which they mature.

3.3. Simulations

The effects derived from CSFs are so great and extend over such a long time that it is simply impossible to gauge these simultaneously for the whole set of macromagnitudes (real output, per capita income, employment and rate of unemployment), years (1989-2006) and for all Spain’s regions. On the other hand, it is vital to define precisely the scenarios and differential results in order to understand what is being measured

We will use the following criteria and definitions:

a) The Gross Value Added (GVA), employment and population for the period 1989-2006 are established using observed data and official projections, which are completed further by realistic evolution hypotheses. The variation rates for each region and for each magnitude are consistent with the national variation rate, which is a weighted average of these.

^b) We will assume that the GVA and employment projections include the investment effects of the successive CSFs. In other words, had the investment not existed the GVA and employment would be lower by an amount equivalent to such effects. These projections will be called CSF scenario.

^c) From the GVA and employment projections (CSF scenario) we subtract the difference estimated by our models, the total effects (supply and demand) of the investment, to define an alternative reference scenario which we will call the no- CSF scenario.

This is a standard way of presenting the results of simulations similar to those carried out in this exercise. The scenarios chosen allow us to evaluate the extent to which the successive CSFs bring, for each Spanish region, economic benefits compared to a fictitious situation in which the aid does not exist.

Rather than present extremely detailed results by year, macromagnitude and region, we will summarise the main results based on the averages for the different programming periods and regions (detailed results are available from the author upon request).

4. Empirical data and results

4.1. Data

For the programming periods 1989-93 and 1994-99 we have used the annual data for regions and funds set out in CORREA and MANZANEDO (2002). The data refer to implementation (payments). Given the need to differentiate between basic infrastructure, human resources and the assistance given to the productive sector, we have divided the annual datum per fund for each region among the three categories using the percentage distribution by axis for each of the funds. Under the heading of basic infrastructure we have included all expenditure on transport, telecommunications, energy, water and the environment, and health. Under human resources we have taken into account all spending on worker training and on the promotion of research and technological innovation. Lastly, assistance designed to consolidate productive sectors, promote the development of new sectors or encourage horizontal actions for business diversification has been included in the productive environment category. The results obtained reproduce almost exactly the percentage distribution of the annual average for 1994-99 obtained from the Table entitled ‘Summary of structural and similar expenditure in Objective 1 regions’, included in the ‘CSF 2000-2006’ Document (MINISTRY OF FINANCE, 2001, 211).

For programming period 2000-2006 the total data for the period by region and fund have been used, as contained in the aforementioned ‘CSF 2000-2006’ Document, including the performance reserve (MINISTRY OF FINANCE, 2001, chapter 3). The data have been distributed by years using the same annual distribution envisaged for Spain as a whole (MINISTRY OF FINANCE, 2001, 214). It should be noted that in the case of Cantabria, the application of this criterion produces exactly the same results as those given by the ‘CSF 2000- 2006’ Document for the region, the only one with annual accrual (MINISTRY OF FINANCE, 2001, 204). For the distribution by investment areas, the same procedure was used as for the previous programming periods. The results obtained reproduce almost exactly the percentage distribution of the annual average for 2000-2006 obtained from the Table entitled ‘Summary of structural and similar expenditure in Objective 1 regions’, set out in the ‘CSF 2000-2006’ (MINISTRY OF FINANCE, 2001, 211).

Figure 1 shows the average values of the CSFs received by each of Spain’s Objective 1 regions, expressed in 1999 euros. As can be seen, in all the budget periods Andalusia received most funds (an average of 1,075 million 1999 euros during 1989-2006), followed by Galicia (439 million euros), except between 1989-93, when Castilla y León took second place among the regions receiving structural assistance. The last place in the list are filled invariably by Cantabria (an average of 75 million 1999 euros for the period 1994-2006), Murcia (159 million during 1989-2006) and Asturias (185 million during 1989- 2006).

Figure 2 illustrates the average CSF share in GVA (both expressed in 1999 euros). As shown, the greatest impact is seen in Extremadura where for all the programming periods the CSF share in the region’s real output is highest (averaging 2.72% for the period 1989-2006). Behind Extremadura, in descending order, come Castilla-La Mancha (average of 1.68% during 1989- 2006), Galicia (1.56%), Andalusia (1.55%) and Asturias (1.51%). At the other end of the scale the impact in Valencia, Cantabria and the Canaries is lowest, accounting for just 0.71%, 1.16% and 1.18% of the respective real outputs.

Figure 3 gives the average percentage distribution for each area of investment and region for the entire period studied (1989-2006). As can be seen, infrastructure spending accounts for the single largest portion of aid (45% on average), with Castilla y León and Galicia recording a figure of 50%. Cantabria and Valencia are the regions with the lowest figure for this category (36% and 38%, respectively, of the overall aid). The second largest category corresponds to human capital, which averages 34% although it is of greater importance in regions such as Valencia (41%), Extremadura (39%) and Andalusia (36%), unlike in Cantabria and Galicia, where it amounts to just 31% and 32%. Finally, aid to business averages just 21% of total expenditure, although in some regions, such as Cantabria, this category is more important than human capital (33% vs. 31%). In the regions of Castilla y León and the Canaries it represents the smallest share of total spending (18% and 19% respectively).

4.2. Main results

In this section we will use diagrams to compare the average effects in each region according to the different scenarios (CSF, no-CSF) and for each of the macromagnitudes mentioned above.

4.2.1 Effects on real output and growth rate

Figure 4 presents the results for real output, expressed as the differences in average GVA with CSF and without CSF, in both cases in millions of 1999 euros. As can be seen, the simulation results indicate that all the Objective 1 regions would see increases in their real GVA compared to the situation which would have existed had the regions not received the investment under the various CSFs. The regions to benefit most during the period overall (1989-2006) would be Andalusia (2,785 million 1999 euros), Galicia (1,249 million),

Valencia (1,035 million), Castilla-La Mancha (801 million), Extremadura (713 million) and Castilla y León (703). For their part, the Canaries (687 million), Asturias (455 million), Murcia (393 million) and Cantabria (197 million) would be the ones to benefit least.

As Figure 5 illustrates, the demand or Keynesian effects would increase real GVA on average by 2.73% compared to the reference scenario (no EU aid) during the period 1989-2006. The greatest impact of these effects would be seen in Extremadura (average of 5.07%), Castilla-La Mancha and Andalusia (both 3.09%) and Galicia (2.94%), with the lowest in the Canaries and Valencia (deviations of just 2.19% and 1.33% with respect to the no-CSF scenario). The total effects (Keynesian plus externalities) would average 3.68% in the regions during the period 1989-2006. Once again the greatest impact would be seen in Extremadura (7.00% average), Castilla-La Mancha (4.24%), Andalusia (3.97%) and Galicia (3.90%), while the lowest would be in the Canaries and Valencia (deviations of 2.97% and 1.80% with respect to the no-CSF scenario).

Figure 6 shows the average real growth rate results, expressed as the differences between cumulative GVA growth with and without CSF. Here too the simulations carried out suggest that all the regions would see growth rate gains compared to the situation which would have prevailed in the no-CSF scenario. The regions presenting the highest growth rate differential between 1989 and 2006 would be Extremadura (0.58 points), Andalusia (0.32 points), Murcia (0.25 points), Castilla-La Mancha (0.25 points) and Galicia (0.25 points). The results indicate that the greatest growth rate differences between the two scenarios would be seen during the first of the programming periods (1989- 1993), with Extremadura and Castilla y León (both 0.99 points) and Valencia (0.86) and Castilla-La Mancha (0.71) heading the list. For the second period (1994-1999) the greatest differences would be found in the two Castilles (cumulative difference between the CSF and no-CSF scenarios of more than 0.80 points). Lastly, for the current programming period (2000-2006) the results obtained suggest that the greatest differences between the growth rates under the two scenarios would be in Castilla-La Mancha and Cantabria (in both cases 0.30 points), followed by Extremadura and Andalusia (0.25 and 0.20 points, respectively).

4.2.2 Effects on per capita income and convergence with EU-15

The average results for real per capita income are given in Figure 7 and are expressed as differences between average per capita GVA with and without CSF. As can be seen, all regions would present per capita income gains compared to the no-CSF scenario. In terms of the overall period studied the greatest differentials would be found in Extremadura (662 euros at 1999 prices),

Castilla-La Mancha (468 euros), Galicia (458 euros), Asturias (428 euros) and the Canaries (409 euros). As the Diagram shows, the per capita income gains would increase gradually from one programming period to the next. For 2000- 2006 the simulations indicate that the population of Extremadura would be 941 euros better off in real terms than if there had been no CSF investment by the EU. Galicia and Castilla-La Mancha would figure next in terms of leading beneficiaries, with a per capita income difference of 646 and 628 euros (at 1999 prices) respectively.

Table 2 illustrates the relative situation of real per capita income in Spain’s Objective 1 regions compared to the EU15 average (100 = EU15 average for each year). As can be seen, at the end of the first programming period the figure for Extremadura would be four points higher than it would have been without CSF investment during the period 1989-1993. For Valencia and Castilla-La Mancha the difference would be three points, while for Asturias, the Canaries and Murcia it would be two. In the case of the second programming period, our simulations suggest that the difference between the CSF and no-CSF scenarios for 1999 would be five points for Extremadura, four for Cantabria, Castilla-La Mancha, Castilla-León, Galicia, Murcia and Valencia, and three points for Andalusia, Asturias and the Canaries. It should be noted that, without the structural aid received under the CSF 1994-1999, per capita income in Cantabria and Castilla-León would not have exceeded 75% of the Community average. With respect to 2002, the results obtained indicate that the difference between the two scenarios would amount to six points for Extremadura and Castilla-La Mancha, five in the case of Castilla-León, Murcia and Valencia and three for Andalusia, Asturias and the Canaries.

From a dynamic perspective the data in Table 2 indicate that, had it not been for the CSF investment, the process of real convergence would have been considerably slower between 1993 and 1999 (Andalusia, Canaries, Castilla y León, Galicia and Murcia) or the divergence seen would have been even greater (Asturias, Cantabria and Extremadura). Similar results are obtained for the period 1993-2002.

4.2.3 Employment and unemployment rate effects

Figure 8 shows the average results for employment, expressed as the difference in the average number of employed persons under the two scenarios (with CSF and without CSF). As can be seen, all the regions would experience employment gains compared to the no-CSF situation. In terms of the overall period studied Andalusia would record the most significant increase in jobs between the two scenarios (36,000), followed by Galicia (14,000 more jobs) and the two Castilles (11,000).

As Figure 9 illustrates, the demand or Keynesian effects would increase employment by an average of 2.58% compared to the reference scenario (no EU aid) during the period 1989-2006. The greatest impact of these demand effects would be seen in Extremadura (average of 4.57%), Castilla-La Mancha (3.10%), Andalusia (2.87%) and Galicia (2.72%), whereas the Canaries and Valencia would be the regions to benefit least in this regard (percentage deviations of 2.04 and 1.26 respectively compared to the no-CSF scenario). On average, the total effects (Keynesian plus externalities) would amount to 1.46% in the regions during the overall period 1989-2006. Again the greatest impact would be seen in Extremadura (average of 2.49%), Castilla-La Mancha (2.01%), Andalusia (1.81%) and Galicia (1.46%), whereas the lowest would be in Valencia and Asturias (deviations of 1.15% and 0.65% with respect to the no-CSF scenario). The lower impact of the total effects compared to the demand effects can be explained by the fact that the greater availability of public infrastructure, private capital and human capital as a result of EU aid would make private production factors more productive and, among other consequences, the same amount of goods and services would be produced with less labour.

Lastly, Figure 10 shows the differences in the unemployment rates recorded under the two scenarios. Specifically, it illustrates the decrease - expressed in percentage points- in the unemployment rate under the CSF scenario compared to the situation that would have prevailed had the frameworks not existed. The regions that would experience the greatest reduction in the rate during the period studied are Extremadura (1.36 points), Andalusia (0.98), Galicia (0.79), and the Canaries and Castilla-La Mancha (both 0.73).

5. Concluding remarks

Regional policy has been viewed as necessary in the European Union to sustain further economic and political integration that could be hindered by wide regional disparities. Consequently, the EU has devoted an increasing share of its budget to its regional policy, being Spain the country that has most benefited, in absolute terms, from such policy.

This paper has provided an evaluation of the macroeconomic effects of the Structural Funds received by Spain’s Objective 1 regions, being Spain the country that has most benefited, in absolute terms, from this EU policy. The results obtained suggest that the funds have contributed significantly both to economic growth and to job and wealth creation. The successive investment programmes during the period 1989-2006 have represented an average gain of 0.56 points in the growth rates of the beneficiary regions compared to the situation that would have prevailed without the aid. This gain would translate to an average increase in per capita income of 425 euros (at 1999 prices). In labour market terms, it is estimated that the total effects of CSF investment would generate or maintain, on average for the aforementioned period, 1.46% more employment than would have been the case without the aid, with an average decrease in the unemployment rate of 0.74 points during the period. These results would be in line with those given by DE LA FUENTE (2003) for the period 1994-1999.

Although as in all empirical works the limitations derived from the different assumptions adopted in the study mean that the values obtained have to be taken with some degree of caution (particularly when they are expressed as contrafactual, given that we simply do know what might have happened without EU aid), the results of our simulations lead us to infer that the aid has made a considerable contribution to the progress of the Spanish economy in recent years, providing a crucial boost to job and wealth creation.

Although there is some debate as to their efficiency [see, for example, Boldrín and CANOVA (2001) and MIDELFART-KNARVIK and OVERMAN (2002)], the simulations suggest that the structural aid provided through the CSFs achieved their goal of facilitating income level convergence between the regions of Spain and Europe, given that without such aid Spain’s Objective 1 regions would be even worse off compared to the EU average. In this regard, our results clarify the premature conclusions rejecting convergence which were obtained by BOLDRÍN and CANOVA (2001), given that the period studied by these authors (1980-1996) fails to take account of the full amount of aid received between 1994-1999. Our results would be in line with those presented by MARÍA-DOLORES and GARCÍA-SOLANES (2002) regarding the positive contribution of the Structural Funds to real convergence by Spain’s Objective 1 regions.

In view of the role of structural funds, it is only natural that Spain should be concerned and anxious with respect to the future of this type of aid in the EU’s Financial Perspectives for 2007-2013. Spain will receive less aids than what it has in the past for several reasons, the most important being that some of the Spanish regions that were in the past included in Objective will no longer be a part of this objective due to natural or statistic convergence. Within the natural convergence case we can cite Cantabria, which was already outside of Objective nº 1 for the period 2000-2006, but will afterwards cease to receive phasing-out aids, Valencia, the Canaries, and Castille and Leon (although the Canaries is arguing that it should retain the aids since it is an outermost region). Other regions of Spain, that are currently under Objective nº1, will cease to be present because of so-called “statistical convergence” after the incorporation of other countries that are poorer in 2004 and 2007. Within this type of region we have Asturias, Murcia, Castille-La Mancha, and the Autonomous North-African cities of Ceuta and Melilla. Moreover, were it to profit from a high rate of growth in the coming years, the region of Galicia could also find that it is above the 75 per cent criteria of average community income thus not being eligible any more for structural aids under this title. Thus, under the actual criteria only Andalusia, Extremadura and, likely, Galicia will continue to receive structural aids although the EU Commission states in its most recent proposal to date that the regions that are affected by the statistical effect may continue to receive transitory aid until 2013. Spain, however, will lose the aids from the Cohesion Fund entirely.

An agreement for 2007-2013 is not merely a matter of urgency for budget reasons (it will be extremely difficult to adopt a budget for 2006 or 2007) or for technical reasons (a minimum of 12-18 months are needed to prepare all the legal and financial instruments associated with the Perspectives), but above all for political (to demonstrate that, despite the institutional uncertainty, Europe has not come to a grinding halt but continues to move forward) and economic motives (to restore confidence among economic players and reinforce the commitment to more growth, more employment and more competitiveness).

Ideally, the failure of the Brussels European Council in June 2005 should be harnessed to draw up an even more ambitious cohesion policy, to tackle appropriately not just the widening of the socio-economic disparities that come with enlargement and the restructuring caused by increasing globalisation, but also the spread of a new knowledge-based economy and the demographic challenges posed by the ageing population and growing immigration. For this a larger budget will be required, although the EU is facing a far-reaching review of its finances in a context of slow economic growth and 20 million people unemployed, a context in which six of the leading net contributors (Germany, Britain, Netherlands, France, Sweden and Austria) are pressing for a reduction in the budget.

A number of options remain open, not all of which are mutually exclusive. To begin with the British rebate might be adjusted, freezing the figure at 4,700 million euros in 2007 and gradually reducing the amount every year thereafter. A second option is to renegotiate the Common Agricultural Policy (under which 40% of the Community budget goes to a sector representing barely 4% of Europe’s GDP) and make an extra effort to reduce spending on agriculture. A further option is the introduction of a degree of gradualism in the form of transition measures to enable the countries most affected (Spain among them) to deal more comfortably with the new financial situation. A fourth possibility would be to recycle funds already committed but not yet spent, using the ‘n+2’ rule. A final option would be to acknowledge the national statistics effect (in addition to the regional effect) with a view to gradually easing out of the Cohesion Fund countries exceeding 90% of the average income of the enlarged EU but which remain below that figure for EU-15.

The task is extremely complicated given the need to find new alternatives, new possibilities and new options that afford mutually beneficial opportunities and are preferable to all the initial proposals. We must move forward on the basis of true consensus, namely, one grounded on what everyone stands to gain as opposed to what each stands to lose. An effort will have to be made to reach more flexible positions and focus the negotiation on the European perspective ahead of national interests. Only in this way will the European project be strengthened, through a renovated vision of the community solidarity which is expressed in the form of structural and cohesion aid, all the more so when the projects funded serve to make the benefits of EU membership more visible to the citizens of Europe.

Whatever the outcome, Spain and Spain’s regions must continue to work towards economic policies that seek to resolve structural macroeconomic imbalances, liberalise product and factor markets, and develop (and maintain) adequate productive and human capitalisation in order to ensure that a reduction in or withdrawal of EU aid will not be felt. In this demanding challenge Spain would do well to devote at least the same effort to readying itself to saying farewell to EU aid as that currently being investing in attempts to maintain the aid.

References

1. Aschauer, D. (1989): Is public expenditure productive?, Journal of Monetary Economics 23, 177-200.
2. Avilés Zugasti, A., Gómez García, R. and Sánchez Maldonado, J. (2001): The effects of public infrastructure on the cost structure of Spanish industries, Spanish Economic Review 3, 131-150.
3. Boldrin, M. and Canova, F. (2001): Inequality and convergence: Reconsidering European regional policies, Economic Policy 32, 205-253.
4. Bradley, J. and Fitzgerald, J. (1988): Industrial output and factor input determination in an econometric model of a small open economy, European Economic Review 32, 1227-1241.
5. Correa, M. D. and Manzanedo, J. (2002): Política regional española y europea, Working Paper SGFCC-2002-05, Dirección General de Presupuestos, Ministry of Finance, Madrid.
6. Corugedo, I., García, E. and Martínez, J (1992): Educación y rentas. Una aplicación a la enseñanza media en España: Una nota, Investigaciones Económicas 16, 299-304.
7. De la Fuente, A. (2003): The effect of Structural Fund spending on the Spanish regions: An assessment of the 1994-99 Objective 1 CSF, Working Paper 2003-11, FEDEA, Madrid.
8. Fujita, M., Krugman, P. and Venables, A. J. (1999): The spatial economy: Cities, regions, and international trade, The MIT Press, Cambridge, MA
9. Gil, C., Pascual, P. and Rapún, M. (1998): Public capital, regional productivity and spatial spillovers, Working Paper 9811, Universidad Pública de Navarra, Pamplona.
10. Gramlich, E. (1994): Infrastructure investment: A review essay, Journal of Economic Literature 32, 1176-1196.
11. Herce, J. A. (1994): Evaluación del Marco de Apoyo Comunitario 1989-93, FEDEA, Madrid.
12. Herce, J. A. and Sosvilla-Rivero, S. (1994): The effects of the Community Support Framework 1994-99 on the Spanish economy: An analysis based on the HERMIN model, Working Paper 94-10R, FEDEA. Madrid.
13. Herce, J. A. and Sosvilla-Rivero, S. (1995): HERMIN Spain, Economic Modelling 12, 295-311.
14. Kaldor, N. (1972): The irrelevance of equilibrium economics, Economic Journal 82, 1237-1255.
15. Krugman, P. (1991a): Geography and trade, The MIT Press, Cambridge, MA
16. Krugman, P. (1991b): Increasing returns and economic geography, Journal of Political Economy 99, 483-499.
17. Krugman, P. and Venables, A. J. (1990): Integration and the competitiveness of peripheral industry, in Bliss, C. and Braga de Macedo, J. (eds.): Unity
18. with diversity in the European economy: The Community’s Southern frontier, Cambridge University Press, Cambridge, 56-75.
19. María-Dolores, R. and García-Solanes, J. (2002): The impact of European Structural funds on economic convergence in European countries and regions, in MEEUSEN, W. and VILLAVERDE, J. (eds.) Convergence Issues in the European Union, Edward Elgar, London, Chapter 5.
20. Martin, P. and Ottaviano, G. I. P. (1999): Growing locations: Industry location in a model of endogenous growth, European Economic Review 43, 281- 302.
21. Martin, P. and Rogers, C. A. (1995): Industrial location and public infrastructure, Journal of International Economics 39, 335-351.
22. Martínez, D. (2000): Is there any relationship between public investment and economic growth in the Spanish regions?, Working Paper 0002/13, Departamento de Economía Aplicada, Universidad de Jaén.
23. Más, M., Maudos, J., Pérez, F and Uriel, E.. (1996): Infrastructures and productivity in the Spanish regions, Regional Studies 307, 641-649.
24. Midelfart-Knarvik, K. H. and Overman, H. G. (2002): Delocation and European integration: Is structural spending justified?, Economic Policy 35, 321- 359.
25. Ministry of Finance (2001): Marco Comunitario de Apoyo (2000-2006 para las Regiones Españolas del Objetivo 1, Ministry of Finance, Madrid.
26. Neary, J. P. (2001): On hype and hyperbolas: Introducing the New Economic Geography, Journal of Economic Literature 39, 536-561.
27. Norman, V. D. and Venables, A. J. (1995): International trade, factor mobility, and trade costs, Economic Journal 105, 1488-1504.
28. Ottaviano, G. I. P. and Puga, D. (1998): Agglomeration in the global economy: A survey of the ‘New Economic Geography’, World Economy 21, 707-731.
29. Porter, M. E. (1986): Competition in global industries: A conceptual framework, in Porter, M. E. (ed.), Competition in Global Industries, Boston: Harvard Business School Press, 15-60.
30. Puga, D. (1999): The rise and fall of regional inequalities, European Economic Review 43, 303-334.
31. Puga, D. (2002): European regional policies in light of recent location theories, Journal of Economic Geography 2, 372-406.
32. Quah, D. T. (1996): Empirics for economic growth and convergence, European Economic Review 40, 1353-1375.
33. Sala-i-Marti, X. (1996): The classical approach to convergence analysis, Economic Journal 106, 1019-1036.
34. Sala-i-Marti, X. (1997): És bo que el govern inverteixi “sempre” a les regions menys desenvolupades?, Nota d’Economia 57, 123-157.
35. Sosvilla-Rivero, S. and Herce, J. A. (2002): Efectos del Plan de Infraestructuras de Transporte 2000-2007 sobre las Comunidades Autónomas, FEDEA, Madrid.
36. Venables, A. J. (1996): Equilibrium locations of vertically linked industries, International Economic Review 37, 341-359.

Table 1 Estimation of the elasticity of regional output with respect to public infrastructure stock in Spain’s Objective 1 regions

Source: SOSVILLA-RIVERO and HERCE (2002)

Table 2 Per capita income of Spain’s Objective 1 regions in terms of purchasing power parity (PPP)

Sources : Eurostat Spanish Regional Accounts and simulations Notes : ( 1 ) EU- 1 5 = 1 00 ; (2) Percentage points ; (3 ) Cantabria did not receive funds between 1 9 89-93

Figure 1 Average value of CSFs received by Spain’s Objective 1 regions (in 1999 euros)

Figure 2 CSF share in the real output of Spain’s Objective 1 regions

89-93 94-99 00-06 89-06

Figure 3 Average percentage distribution according to investment categories of Spain’s Objective 1 region CSFs (1989-2006)

Infraestructure Aid to business Human capital

Figure 4 Effects of CSFs on real output in Spain’s Objective 1 regions Note: Difference in real GVA between the CSF and non-CSF scenarios (expressed in millions of 1999 euros)

Figure 5 Breakdown of CSF effects on real output in Spain’s Objective 1 regions Note: Average percentage deviation with respect to no-CSF scenario (1989-2006)

Figure 6 Effects of CSFs on real growth rate in Spain’s Objective 1 regions Note: Difference in the cumulative real GVA growth rate between the CSF and no-CSF scenarios

Figure 7 Effects of CSFs on real per capita income in Spain’s Objective 1 regions Note: Difference in real per capita GVA between the CSF and no-CSF scenarios (expressed in 1999 euros)

Figure 8 Effects of CSFs on employment in Spain’s Objective 1 regions 89-93 94-99 00-06 89-06 Note: Difference between the two scenarios (CSF and no CSF)in terms of the number of employed persons (thousands)

Figure 9 Breakdown of CSF effects on employment in Spain’s Objective 1 regions Note: Percentage deviation with respect to the no-CSF scenario

Figure 10 Effects of CSFs on the unemployment rate in Spain’s Objective 1 regions Note: Reduction in unemployment rate under the CSF scenario compared to the no-CSF scenario (percentage points)

DOCUMENTOS DE TRABAJO

References

1. 2005-24: “EU Structural Funds and Spain’s Objective 1 Regions: An Analysis Based on the Hermin Model”, Simón Sosvilla-Rivero.

References

1. 2005-23: “A sequential model for older workers’ labor transitions after a health shock”, Sergi Jiménez-Martín, José M. Labeaga y Cristina Vilaplana Prieto.

References

1. 2005-22: “Price Convergence in the European Car Market”, Salvador Gil-Pareja y Simón Sosvilla-Rivero.

References

1. 2005-21: “Implicit regimes for the Spanish Peseta/Deutschmark exchange rate”, Francisco Ledesma-Rodríguez, Manuel Navarro-Ibáñez, Jorge Pérez-Rodríguez y Simón Sosvilla-Rivero.

References

1. 2005-20: “A Projection of Spanish Pension System under Demographic Uncertainty”, Namkee Ahn, Javier Alonso-Meseguer y Juan Ramón García.

References

1. 2005-19: “The Dynamic of temporary jobs: Theory and Some Evidence for Spain (The Role of Skill)”, Elena Casquel y Antoni Cunyat.

References

1. 2005-18: “The Welfare Cost of Business Cycles in an Economy with Nonclearing Markets”, Luis A. Puch

References

1. 2005-17: “Life Satisfaction among Spanish Workers: Importance of Intangible Job Characteristics”, Namkee Ahn.

References

1. 2005-16: “Persistence and ability in the innovation decisions”, José M. Labeaga y Ester Martínez-Ros.

References

1. 2004-15: “Measuring Changes in Health Capital”, Néboa Zozaya, Juan Oliva y Rubén Osuna.

References

1. 2005-14: “Discrete choice models of labour Supply, behavioural microsimulation and the Spanish tax reforms”, José M. Labeaga, Xisco Oliver y Amedeo Spadaro.

References

1. 2005-13: “A Closer Look at the Comparative Statics in Competitive Markets”, J. R. Ruiz-Tamarit y Manuel Sánchez-Moreno.

References

1. 2005-12: “Wellbeing and dependency among European elderly: The role of social integration”, Corinne Mette.

References

1. 2005-11: “Demand for life annuities from married couples with a bequest motive”, Carlos Vidal-Meliá y Ana Lejárraga-García.

References

1. 2005-10: “Air Pollution and the Macroeconomy across European Countries”, Francisco Álvarez, Gustavo A. Marrero y Luis A. Puch.

References

1. 2005-09: “The excess burden associated to characteristics of the goods: application to housing demand”, Amelia Bilbao, Celia Bilbao y José M. Labeaga.

References

1. 2005-08: “La situación laboral de los inmigrantes en España: Un análisis descriptivo”, Ana Carolina Ortega Masagué

References

1. 2005-07: “Demographic Uncertainty and Health Care Expenditure in Spain”, Namkee Ahn, Juan Ramón García y José A. Herce.

References

1. 2005-06: “EL NO-MAGREB. Implicaciones económicas para (y más allá de) la región”, José A. Herce y Simón Sosvilla Rivero.

References

1. 2005-05: “The real picture: Industry specific exchange rates for the euro area”, Simón Sosvilla-Rivero y Sonia Pangusión.

References

1. 2005-04: “A Residential Energy Demand System for Spain”, Xavier Labandeira, José M. Labeaga y Miguel Rodríguez.

References

1. 2005-03: “The Evolution of Retirement”, J. Ignacio Conde-Ruiz., Vincenzo Galasso y Paola Profeta.

TEXTOS EXPRESS

2004-02: “¿Cuán diferentes son las economías europea y americana?”, José A. Herce.