Spain
Selected Issues

Spain has experienced income convergence consistently in the past decade, despite gradual losses in competitiveness. The empirical evidence indicates that overall EU enlargement offers a range of opportunities for Spain, points to potential pressures in specific sectors, and offers challenges, and tackling the challenges requires a flexible economy. The pattern of Spanish exports is dominated by its off-center location in the Southern part of the EU. Geographical location also plays a central role in determining the origin of foreign direct investment (FDI) flows to Spain.

Abstract

Spain has experienced income convergence consistently in the past decade, despite gradual losses in competitiveness. The empirical evidence indicates that overall EU enlargement offers a range of opportunities for Spain, points to potential pressures in specific sectors, and offers challenges, and tackling the challenges requires a flexible economy. The pattern of Spanish exports is dominated by its off-center location in the Southern part of the EU. Geographical location also plays a central role in determining the origin of foreign direct investment (FDI) flows to Spain.

EU Enlargement and Spain: More Opportunities than Challenges

A. Introduction

1. Since the mid 1990s, Spain has experienced a continuous narrowing of the per capita income differential vis-à-vis the EU. Spain’s per capita income exceeded 85 percent of the EU average in 2002, roughly 10 percentage points higher than a decade ago, and increased further in 2003. Robust growth was spurred by prudent macro economic policies: the large fiscal deficits of the early 1990s gave way to sustained gains in fiscal consolidation that, by allowing real interests to fall and securing early EMU participation, crowded-in private investment.1 At the same time, expenditure control allowed needed reductions on the tax burden. This was complemented by reforms in labor and product markets, resulting in significant pay-offs in employment, particularly evident during the recent slowdown of economic activity when private consumption remained strong. In addition, domestic demand was buoyed by public infrastructure and booming construction activity in this period.

uA01fig01

Spain: Real GDP Growth, 1994-2003

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Sources: Eurostat; and Fund staff estimates.

2. Output growth was accompanied by increasing trade. Openness—the sum of exports and imports as a share of GDP—advanced dramatically in the past decade averaging about 45 percent in recent years, roughly 15 percentage points higher than a decade ago. The ratcheting up of exports mirrors inroads in Spain’s share of EU imports, by far the most important market accounting for over two-thirds of Spanish exports (Figure 1).

Figure 1.
Figure 1.

Spain: Share of EU Imports and Importance of Exports to EU, 1990-2002

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Source: IMF, Direction of Trade.
uA01fig02

Spain: Merchandise Trade, 1990-2002

(in percent of GDP)

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Source: IMF, World Economic Outlook

3. Spain’s ability to preserve these gains, however, is unclear: higher price and cost inflation has been eroding competitiveness and productivity growth has been disappointing. Inflation has outstripped the euro area by a cumulative 6 percentage points since 1997. The real effective exchange rate has appreciated steadily over the past years, in part reflecting the strengthening of the euro against the dollar. In an economy that is catching-up some of the appreciation could be the natural by-product of a convergence process and the Balassa-Samuelson effect, but Spain’s productivity data does not provide much solace in this regard. Moreover, with strong increases in unit labor costs, export profit margins have been narrowing consistently in the past three years. A continuation of these trends could hinder the economy’s ability to continue copine with an environment that promises intensified competition as EU enlargement proceeds.

uA01fig03

Exchanged Rate

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Export Margins

(Year-on-year rate of change)

article image
Source: Bank of Spain.

Assumes a common unit labor cost for the economy.

4. EU enlargement could hasten Spain’s loss of competitiveness. Not only are wages sharply lower in accession countries, but these countries also benefit from a highly skilled workforce. Accession countries’ share of the EU imports has increased rapidly and these countries have been a magnet for foreign direct investment (FDI). FDI from the EU has almost doubled in the past five years, and on average has exceeded FDI in Spain by significant margins. As enlargement proceeds these developments are suggestive of the potential for trade (and FDI) diversion from Spain and to accession countries. Even though the bulk of trade barriers with the accession countries has already been lowered, the continued harmonization of legislation and regulations afford the potential for new trading opportunities for these countries.

Wages and Schooling

article image
Note: Data refers to the manufacturing sector; wage and schooling data are respectively for 2000 and 1999.Sources: ILO, Eurostat, and the European Economy Group.
uA01fig04

Share in EU imports, 1995-2002

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Source: IMF, Direction of Trade.

5. In addition, Spain is likely to face a substantial decline in European Structural Funds (ESF) in the medium term. Enlargement will reduce the EU average per capita income by more than 10 percent. Spain’s per capita income (as a percentage of the enlarged EU average) will rise above the 90 percent threshold for Cohesion Funds (CF). Thus, from being one of the largest recipients of ESF, Spain would no longer qualify under the guidelines for CF once the current allocation expires at end–2005. Moreover, 12 of its 17 regions would not be eligible for Objective 1 funds—as their per capita income is projected to exceed the 75 percent threshold. By 2005, and assuming that accession countries continue to converge, Spain’s per capita income is expected to remain above 90 percent of the enlarged EU average, and only three regions would be eligible for Objective 1 funds.2

EU Enlargement: Per Capita Income 1/

article image
Source: European Commission, 2003.

Income correspond to estimates for 2003 in 2000 PPP euros.

6. This paper centers on four channels through which EU enlargement can affect Spain:

  • International trade. Enlargement offers the potential for new trade opportunities and, judging by the experience of previous enlargements, trade creation is likely to result. Still, this does not rule out the potential for Spain’s exports to the EU to be displaced by the increased competition from the accession countries.

  • FDI. The low-wage high-skilled labor force of the accession countries would appear to make them attractive locations for foreign investment. As in trade, the opportunities for “FDI creation” are likely to dominate “FDI diversion.” Nonetheless, eastward relocation of production plants is possible, and to some extent this process has already begun in the run up to accession, affecting most notably automobile production plants.

  • European Structural Funds. Spain will continue benefiting from the infrastructure built with the help of past ESF, but under current guidelines eligibility for further funds is in question. With the future of ESF under discussion, quantifying potential adverse effects is somewhat premature and still subject to the particular modifications to eligibility rules and EU budget allocation decisions.

  • Migration. Although citizens of new member countries have the right to reside in any member state upon accession, the EC has signed agreements with these countries delaying free access to EU labor markets for up to seven years. For Spain even in the absence of restrictions, it is unlikely that flows will be of a magnitude as to have a macroeconomic effect, given the distance and lack of cultural ties, including the lack of an established beach head of immigrants from these countries.

Agriculture is an important area that will be affected by EU enlargement, but assessing its impact on Spain lies beyond the scope of this paper, also because of uncertainties regarding the future of the Common Agricultural Policy.

B. Stylized Facts of Spanish Exports and FDI

Before turning to the potential effects of EU enlargement, this section characterizes the geography and structure of Spain’s exports and FDI.3

Spanish exports

7. The pattern of Spanish exports is dominated by its off-center location in the Southern part of the EU:

  • Spain trades primarily with the EU-15, and trade with accession countries remains small, although it has more than doubled in the past decade (Table 1). These trading patterns are consistent with fundamental factors underlying Spain’s trade. Spain trades significantly more with France than Germany, driven in part by the fact that France is geographically closer; sharing a common border acts to boost trade. Despite being closer to Italy and the United Kingdom, exports to these countries are comparable to those to Germany reflecting the smaller economic “mass”—population and output—of the former countries. For its part, trade with AC-10 countries is small both because the average distance is more than 1,000 kilometers greater than the average for the other EU countries, and their mass (measured by GDP) is substantially less.

  • The share of Spain’s exports to accession countries is similar to that of France or the United Kingdom, but significantly smaller than that of Italy and Germany (Table 2).

Table 1.

Geography of Spanish Exports

article image
Sources: IMF, WEO; COMTRADE (United Nations Statistical Division); and U.S. Department of Agriculture.

Average.

Table 2.

EU Exports to Accession Countries

article image
Sources: EIS, Direction of Trade; and U.S. Department of Agriculture.

8. Spanish exports are concentrated in elaborated goods:

  • Defined as the sum of Divisions 6 through 8 (Table 3), these goods represent about 60 percent of Spain’s exports. It is noteworthy that Division 7 (machinery and transport equipment) constitutes more than half of exports, of which Division 78 (road vehicles) is the most important single item.

  • This concentration makes Spain potentially vulnerable to competition from AC-10 countries that have a similar export structure (Table 4). On average, the share of AC-10 countries’ exports in elaborated goods has exceeded 60 percent during the past five years. But this may not be the complete story: a closer look at Division 7 suggests that AC-10 countries’ exports are more diversified and are not concentrated in road vehicles.

Table 3.

Structure of Spanish Exports

article image
Sources: COMTRADE (United Nations Statistical Division); and SITC-Revision 2.
Table 4.

Structure Spanish Exports Compared to Accession Countries

(Share in percent, average 1997-2001)

article image
Sources: COMTRADE (United Nations Statistical Division); and SITC-Revision 2

FDI in Spain

9. Geographical location also plays a central role determining the origin of FDI flows to Spain (Table 5):

  • About two-thirds of FDI in Spain originates in the EU; the United Kingdom appears as the single most important investor. The peculiarities of bilateral FDI reporting, however, probably underlie the anomaly that FDI from Belgium-Luxembourg exceeds that from Germany and France combined. The bulk of FDI in Spain has been in the services sector, of which “real estate and business activities” is the main beneficiary, far exceeding “hotels and restaurants” and “transports and communication.”

  • As with exports, Spain will potentially compete for EU FDI as AC-10 countries also receive a large share of their FDI from the EU. Their relative proximity to Germany is likely to underlie the large share of FDI from that country. While the United Kingdom is a major investor in both the AC-10 and Spain, France is a relatively small player in Spain despite its geographical proximity. Although the bulk of FDI in AC-10 countries is also in the services sector, these flows are concentrated in financial intermediation reflecting the stage of development of financial markets. Also, the high-skilled low-wage workforce may account for the relatively large share of investment in manufacturing.

Table 5.

FDI in Spain and Accession Countries, Average 1999-2001.

article image
Source: OECD

C. The Gravity Model

10. The gravity model is commonly used as a framework to study bilateral trade, with numerous empirical applications dating back to the early 1960s (Frankel and others, 1997; Helliwell, 1998; and references therein). Recent studies have addressed the question of whether the level of trade between two countries is unusual in the sense of being substantially different from what is predicted by the gravity model. For instance, the gravity model has been used in examining the trading patterns of the EC and EFTA (Bayoumi and Eichengreen, 1995), and the extent to which regional trading blocks lead to trade diversion as opposed to trade creation. The model has also been used to explain FDI flows (Di Mauro, 2000; and references therein).

11. In its original form, the gravity model relates some measure of bilateral trade to the economic mass of the trading partners and the distance between them:

TRADEijt=(YitYjt)αDijβeμijt

where TRADEij is bilateral trade—either nominal exports (EXPORTSij) or imports (IMPORTSij), or their sum—from country i to country j; Yk is nominal GDP in country k=i, j; Dij is the distance between the capital cities of countries i and j; and t is a time subscript.4 The model posits that trade increases with economic mass (α>0), and decreases with distance (β<0).μijt is:

μijt=γi+φj+δt+ϵijt

where γi and φj are fixed effects respectively for countries i and j, δt are common time effects, and εijt are well behaved error terms.

12. The model has been extended to include additional fundamental factors—such as cultural ties, common borders, access to sea, and membership in preferential trade arrangement (PTA)—to characterize fully bilateral trade flows. Most of these factors are fixed in time and hence have the characteristic of being a “dummy” variable, taking on the value of one when the specific characteristic holds, or zero otherwise. These variables are included in the regression models to pick-up the effect on trade of sharing cultural ties, a common border, and so on.

13. Of particular interest for this study are PTA dummy variables to measure the impact of EU membership.5 Consider the following extension of the gravity model:

TRADEijt=(Yit.Yjt)α.Dijβ.BOTHijtΦ.ONEijtr.eμijt

where BOTHijt and ONEijt are dummy variables that equal one respectively when both (i and j) or only one of the trading partners (i or j) are EU members at time t.6 Thus, the coefficient ϕ captures the extra trade between EU members once other fundamental factors determining trade flows are accounted for. Similarly, the coefficient γ captures the effect on trade of not belonging to the EU. To examine whether there are country-specific effects on trade of non-EU membership, the model is extended to include the interaction of ONEijt with individual country dummies. Finally, a crude measure of the speed at which “undertrading” vanishes following accession can be gleaned by including a “trend” variable for the countries of interest that counts the number of years since accession; it is zero before accession and for nonaccession countries.7

Estimation results8

Exports

14. Gravity models explain a large share of the bilateral trade flows among EU countries in the past 20 years (Table 6). Fixed-effects estimates suggest that the bilateral export elasticity with respect to the economic mass of the trading partners is about three-fourths. This implies that trade increases more than proportionally with respect to the size of the economies—when both trading partners grow by one percent, their trade increases by about 1½ percent; this estimate is robust to different specifications (see columns 1–7). Distance between trading partners reduces trade roughly one-to-one, although this effect is smaller when the model controls for a common border (columns 5 and 7); sharing a common border boosts trade by about 75 percent.9 Trade is lower if one or both of the trading partners is land-locked.

Table 6.

Characterizing EU-15 Exports

(Dependent varible: log of exports from country i to country j)

article image
Note. All regressions include country and time dummies for each of the EU-15 countries and years 1978-2002. Robust standard errors are shown in parenthesis as an indication of the precision of the estimates; the ratio of the coefficient and the standard error has a non-standard distribution.

15. This framework suggests that EU enlargements in the past 20 years—Greece in 1981, Spain and Portugal in 1986, and Austria, Finland, and Sweden in 1995—have led to trade creation among EU members (Table 7).10 This is seen from the regression estimates of the model when PTA dummies (BOTH and ONE) are included. Specifically, when both trading partners are EU members their trade is about 20 percent higher than otherwise justified by fundamental trade factors (column 2). This favorable EU trade effect doubles when the model accounts for differences in trade when only one trading partner is in the EU (column 3). In sum, the experience of previous enlargements suggests that trade creation more than offset trade diversion effects, by roughly some 40 percent between EU members.

Table 7.

Exports and Previous EU Enlargements

(Dependent varible: log of exports from country i to country j)

article image
Note: See note in Table 6 for details. The dummy variable BOTH and ONE take the value of one respectively when both or only one of the trading partners is an EU member, it is zero otherwise.

16. The “Southern” enlargement (Greece, Spain, and Portugal) is of particular interest to this study (Table 8). These countries share the off-center location and lower than average income levels that characterize the current group of accession countries. The results for Southern enlargement suggest mat trade was about 11 percent lower than expected once the fundamentals are taken into account (column 1). Moreover, the results suggest that the degree of undertrading of the Southern members was gradually erased in about seven to eight years.11 This enlargement, however, differs from the current one in that significant trade barriers were in place when these three countries joined the EU.

Table 8.

Exports and Southern EU Enlargement

(Dependent varible: log of exports from country i to country j)

article image
Note: Southern enlargement refers to the accession of Greece, Spain, and Portugal; other accession countries (Austria, Finland, and Sweden) are not included in the definition of EU. DUBAC3 is a dummy variable (holding the definition of EU constant) that takes the value of one when trade is between the EU and Southern accession countries; it is zero otherwise. See note in Table 6 for further details.

17. Turning to the AC-10 enlargement, the models suggest that exports are significantly lower than justified by fundamental factors, but not consistently across countries (Table 9). The sample of countries was extended to include the current group of 10 accession countries set to join the EU in 2004. This allows the model to characterize trade with these countries, but shortens the data sample to the ten years ending in 2002.12 As accession has yet to take place, the results focus on the PTA dummy (DEUAC) that is one when only one trading partner is an EU member. The estimated coefficients suggest that trade between the EU-15 and AC-10 is about 30 percent lower than justified by fundamentals (column 3). Moreover, the degree of “undertrading” varies substantially across countries, from above average undertrading for Spain and virtually no significant undertrading for Germany and France, to overtrading for the United Kingdom. Results for Division 7 exports (column 5) are qualitatively similar, with the exception of the United Kingdom that appears not to overtrade with AC-10.

Table 9.

Characterizing “EU-25” Exports

(Dependent varible: log of exports from country i to country j)

article image
Note. Column 1 reproduces column 7 of Table 6 using the sample period 1993-2002; other columns correspond to EU-25, defined as EU-15 plus AC-10 for 1993-2002. Division 7 (machinery and transport equipment) corresponds to the SITC-Revision 2 classification. DEUAC10 is a dummy variable taking the value of one when trade is between and EU-15 and AC-10 country, and zero otherwise.
FDI

18. Gravity models were used to conduct an analogous study of FDI flows, with the following results:

  • The bilateral FDI elasticity with respect to the economic size (mass) implies mat flows increase more than proportionally with respect to the size of the economies (Table 10). Distance between countries reduces FDI, less so when accounting for a common border (columns 5 and 6); sharing a common border boosts FDI by about 50 percent. FDI is lower if either the country of origin or destination is land-locked.

  • EU enlargements in the past 20 years led to FDI creation among EU members (Table 11); when both countries are EU members their FDI is about 45 percent higher than otherwise justified by fundamental factors (column 2); this doubles when both PTA dummies are included in the model (column 3).

  • The “Southern” enlargement suggests that FDI flows were about 35 percent lower than expected (Table 12); results for the speed at which under-FDI reversed to long-run values are puzzling, but maybe interpreted as the reversal of an initial overshooting of FDI flows that followed accession.

  • FDI between the EU-15 and AC-10 is as much as about 70 percent lower than justified by fundamentals (Table 13, columns 3 and 4); the degree of “under-FDI” varies substantially across EU countries, and is significantly greater for Germany and the United Kingdom.

With the additional information required to construct the bilateral flows exasperating the well-known problems in recording FDI flows, caution is needed when interpreting these results.

Table 10.

Characterizing EU-15 FDI.

(Dependent varible: log of FDI from country i to country j)

article image
Note. See note in Table 6 for details.
Table 11.

FDI and Previous Enlargements

(dependent varible: log of FDI from country i to country j)

article image
Note. For details see notes in Tables 6 and 7.
Table 12.

FDI and Southern EU Enlargement

(Dependent varible: log of FDI from country i to country j)

article image
Note. For details see the notes in Tables 6 and 8.
Table 13.

Characterizing “EU-25” FDI

(Dependent varible: log of FDI from country i to country j)

article image
Note. Column 1 reproduces Column 6 of Table 10 for 1993-2002; other columns correspond to EU-25, defined as EU-15 plus AC-10 for 1993-2002. For further details see the notes to Tables 6 and 9.

D. Third Market Effects

19. To explore directly the challenges that Spain’s exports may face in the EU market from accession countries a dynamic reduced-form model is examined. These (VAR) models have been used to analyze the dynamic effects of shocks in many contexts, and are particularly well suited to summarize the dynamic correlations in the data. In this connection, the following model can be used to gauge the potential challenges:

[MAC-10MRestMSPA]=[a20a30a10]+[a11a12a13a21a22a23a31a32a33][Mt1AC10Mt-1RestMt1SPA]+[μAC-10μRestμSPA]

where Mx is the (log of) EU (excluding Spain) imports from Spain, the combined AC-10 countries, and the rest of the world; μx corresponds to the (reduced-form) shock of EU imports from these three markets.13 Interest centers in characterizing how EU imports (Mx) evolve following an unexpected increase in imports from AC-10; impulse response functions are used to this effect. The model is also used to explore the impact of a permanent increase in imports from AC-10, that could be associated with a “normalization” of the level of imports as the degree of undertrading—measured in the gravity models above—is reabsorbed after accession. This effect is described by step response functions.

20. Spain’s exports to the EU are likely not to suffer overall, but their composition is likely to change. Two models were estimated in a panel VAR context14—one using total imports, the other limited to Division 7 imports—consisting of individual “EU-14” countries’ imports from Spain, AC-10, and rest of world, and used to compute impulse and step response functions associated with an increase in imports from AC-10 countries (Figure 2).15 The estimates suggest that:

  • Impulse responses. Following a historical shock (roughly 9 percent, panels in the left column) in total imports from AC-10 countries, imports from Spain are roughly unchanged on impact, and a very small positive response follows; imports from the rest of the world increase slightly.16 An analogous shock (roughly 15 percent) in Division 7 imports from AC-10 countries is associated with a small “crowding-out” of imports from Spain of some 2 percent on impact, and the adverse effect appears to persist for several years; virtually no response is observed in EU imports from the rest of the world.

  • Step responses. As in previous enlargement experiences accession is likely, however, to lead to a permanent increase in imports from AC-10. If estimates from the gravity models are accurate, the potential permanent increase is just under two times the typical shock (about 30 percent), and reduced-form models suggest that this would lead to a substantial decline in EU imports from Spain (about 5 percent), and a negligible decline in imports from the rest of the world (right column).

Figure 2.
Figure 2.

Third Market Effects: “EU-14” Imports

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Sources: World Integrated Trade Solution; and Fund staff estimates.

21. Turning to FDI, the results suggest that flows to Spain are likely to increase in step with flows to AC-10 countries, albeit less than proportionately (Figure 3). A reduced-form model analogous to those used above was estimated using FDI flows to Spain, AC-10, and the rest of the world originating in the EU (excluding Spain). The typical shock to FDI flows to AC-10 countries (about 80 percent) is associated with an increase of FDI flows to Spain (between 10 and 15 percent in the next four years). A sustained doubling in FDI flows to AC-10 countries is associated with a substantial increase in FDI flows to Spain (about 45 percent) and to the rest of the world (about 55 percent); these results should be taken with a grain of salt.

Figure 3.
Figure 3.

Third Market Effects:“EU-14” Outward FDI Flows

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Sources: World Integrated Trade Solution; and Fund staff estimates.

E. European Structural Funds

22. Spain is the largest recipient of European Structural Funds (ESF). These funds were established to reduce the economic differences across Europe (Box 1). Spain has received an increasing share of ESF, projected to exceed 25 percent of the total in the current five year period (Figure 4). Of these funds, Objective 1 funds—aimed at assisting a country’s lagging regions develop basic infrastructure and foster private investment—account for the bulk of the funds destined to Spain (Greece and Portugal have been larger recipients of ESF as a share of GDP or EU population). Moreover, Spain enjoys the highest cofinancing rate, averaging about 65 percent compared with less than about 40 percent on average.

Figure 4.
Figure 4.

Distribution of Structural Funds Among EU Members, 1989-2006

Citation: IMF Staff Country Reports 2004, 090; 10.5089/9781451812107.002.A001

Sources: Martin and others (2002); and Spanish Representation at the European Commission (2002).

23. Objective 1 regions have benefited from the ESF, but with few expectations convergence in these regions has been less than the overall convergence for Spain as a whole. Covering about three-fourths of the Spanish territory, and home to almost 60 percent of the population, Objective 1 regions receive about 70 percent of Spain’s share of ESF.17 Per capita income in these regions has increased on average 8 percentage points relative to the EU average since 1986. This is less than the overall increase in Spain, and only two Objective 1 regions have converged faster than Spain, namely Cantabria and Castilla-La Mancha.18 Indeed, the former graduated from the program, and its funds are to be phased out by 2006.

Per Capita Income Convergence: Objective 1 Regions and Spain

(EU l5=100)

article image
Sources: Spanish Representation at the EC (2002); Martin and others (2002); and EC (2002b) and (2003).

For 2000-06 Financial Perspectives, the eligibility for Objective 1 regions was determined using the average per capital income for 1994-96.

European Structural Funds

The EU Cohesion Policy is carried out through four European Structural Funds (ESF). Specifically: (i) the European Regional Development Fund covering infrastructure, job-creating investment, local development, and aid to small firms; (ii) the European Social Fund pertaining to training and recruitment aid for unemployed and handicapped workers; (iii) the “Guidance” Section of the European Agricultural Guidance and Guarantee Fund contributing to structural reform of the agricultural sector and the development of rural areas; and (iv) the Financial Instrument for Fisheries Guidance dealing with fishing industry restructuring. The Cohesion Fund finances major projects in environment and transport in countries with per capita income below 90 percent of the EU average (Spain, Portugal, Greece, and Ireland).

The ESF provide partial funding (cofinancing) for regional development plans, with 94 percent concentrated in well-defined territories or populations, according to three objectives:

Objective 1 Provides lagging regions with basic infrastructure. Eligible regions must have GDP per capita at or below 75 percent of the EU average; there are 50 Objective 1 regions in the EU-15 (comprising about 22 percent of EU population and 70 percent of ESF and Cohesion Fund; financed with all four ESF).

Objective 2 Aims at supporting economic and social restructuring in industrial, rural, urban or fisheries-dependent areas facing structural difficulties. Eligible areas are determined jointly by the European Commission and the national authorities according to labor market and socio-demographic indicators (18 percent of EU population and 11.5 percent of funds; financed with items i and ii above).

Objective 3 Focuses on training and promoting employment systems. All EU regions, except Objective 1, can be defined as Objective 3, within the limit of the available resources (12.3 percent of funds; financed with item ii).

24. The available empirical evidence, nonetheless, suggests a positive impact of Objective 1 funds. The European Commission (2001a) estimates—using the Spanish Finance Ministry’s MOISEES model—that GDP is expected to be about 0.9 percent higher than otherwise by 2006; employment creation is projected to be greater by about 120 thousand annually. The EC (2002a) also finds a comparable effect when using a dynamic input-output model. Moreover, they estimate that if all Objective 1 activities (EU, Spanish, and private sector) were phased out in 2006, and not substituted by other expenditures, the level of GDP would decline by 1¾ percent, reflecting the impact of reduced investment (3.2 percent of total investment), 210 thousand fewer jobs, and lower labor productivity (about 1 percent annually) and Spanish imports from the EU.

25. Recent research suggests that the impact of Objective 1 funds may have been smaller, and could have been greater had efficiency considerations received more emphasis. Based on regional production functions and employment equations, De la Fuente (2003) concludes that the 1994-99 regional development plan for Objective 1 regions added around 0.5 percentage points per year to GDP growth and 0.3 percent points per year to employment growth (or 27 thousand new jobs). The part of the regional development plan financed by ESF—excluding national cofinancing—accounts for 15 percent of growth from 1993 to 2000 and helped to reduce by 20 percentage points the initial gap in income per capita between the Objective 1 regions and the rest of Spain. This study also estimates that the average social return of the total public spending (infrastructure spending) included in the regional development plan for Objective 1 regions was 30 (50) percent. However, since the richer an Objective 1 region is, the higher the estimated social return on public spending. De la Fuente (2003) argues that the overall impact of EU aid could have been higher, and Spain’s income convergence faster, if efficiency considerations had been given greater weight in the allocation of the ESF.

Empirical Evidence of the Effect of Structural Funds in Spain

article image

Contribution to growth expressed as a percent of the increase in GDP in the period.

Fraction of the original income per capita differential with the average income per capita of Spain, assuming constant differential in population

Includes the effect of increased taxes required to finance the Spanish contribution to the projects.

Reflects the effect of only the grant element of the Objective 1 funds.

Modelo de Investigación y Simulación de la Economía Española.

F. Migration

26. Immigrants from accession countries are a very small fraction of the Spanish population, but have recently increased substantially.19 Of the roughly 850 thousand citizens of accession countries that reside in the EU-15—roughly 0.2 (0.5) percent of the EU population (accession)—about 60 thousand reside in Spain, which is roughly the same share as in the EU. In Spain, only three accession countries account for the bulk of these immigrants: Romania (about half), Poland (21 percent), and Bulgaria (19). In general, migration flows from AC-10 countries to the EU occurred mostly between the fall of the Iron Curtain and EU recession in 1993; thereafter migration faced increased restrictions.20 Ever since, migration is limited to short-term stays to bordering regions and neighboring countries (often seasonal labor or border commuter), and regulated by bilateral agreements (OECD, 2001 and European Commission, 2001b); however, the effect of networking is important.21 In Spain, however, the official statistics suggest a very large flow in 2002, most probably reflecting earlier flows that had not been captured before in the official statistics; the data remain nonetheless somewhat puzzling.

Foreign Residents in Spain

(In thousands)

article image
Source: EUROSTAT

27. The experience of the Southern enlargement suggests that the likelihood of mass migration is low. Net migration flows from Spain and Portugal to the EU following accession were virtually nil in the second half of the 1980s. During this period, there were restrictions on migration for seven years. Even after these were lifted—coinciding with the recession of the early 1990s—there was hardly any increase in the migration flows from either country. In general, strong economic growth (and real convergence) experienced by Southern European countries as a result of joining the European Union, not only discouraged emigration but encouraged the return of citizens to their countries of origin. To the extent that policies in the AC-10 result in above average growth, migration is likely to be moderate.

28. Available empirical studies predict a modest increase in migratory flows from accession countries in the short-run, followed by long-lived flows needed for migration stocks to reach their new long-run equilibrium. This is due to the high costs of migration, and limited capacities of labor markets to absorb these flows, in particular,

  • About 335 thousand (or 0.1 percent of the EU-15 population) were estimated to migrate, assuming that all of the barriers would have been removed in 2002, followed by increasing flows (peaking at about 1.1 percent of EU-15 population) over the following 30 years (Box 2). Of these, Germany was expected to receive about 220 thousand at first with the stock of immigrants reaching 3.5 percent of the German population. In contrast, Spain was expected to receive about 4 thousand in 2002 with a very small effect on its stock.

  • A recent report for Germany (Boeri and others, 2002) suggests that cumulative net migration or long-run migration stocks from the new member states (attained 15–20 years after the free movement of labor has been introduced) will amount to between 2 and 3 percent of the population.

  • The European Commission (2001b) estimates that annual flows to the EU-15 will increase from about 120 to a maximum of 215 thousand four years after accession. By 2009, migration is estimated to be about 900 thousand, roughly 1.2 percent of AC-10 countries’ combined population or some 0.4 percent of the projected EU working-age population.

29. Suggestive evidence from gravity models, as in the case of exports, indicate that migration flows are significantly lower than expected but not consistently across countries (Table 14). Although migration is likely to respond to the same factors as trade (indeed economic theory suggests international trade can lead to similar effects as international labor mobility), other elements noted earlier are likely to play. Keeping in mind the tentative nature of the evidence, the estimates suggest the following regarding migratory flows between EU-15 and AC-10 countries:

Table 14.

Characterizing “EU-25” Migration

(Dependent varible: log of migration from country i to country j)

article image
Note. Column 1 reproduces column 6 of Table 10 for the sample period 1993-2002; other columns correspond to EU-25, defined as EU-15 plus AC-10 for 1993-2002. For further details see the notes to Tables 6 and 9.
  • The bilateral migration elasticity with respect to the economic mass implies that migration increases less than proportionally with respect to the size of the economies.

  • The distance between countries reduces migration substantially; sharing a common border boosts migration by about 50 percent; and migration is lower if either of the countries is land-locked.

  • Migration between the EU-15 and AC-10 is as much as about 50 percent lower than justified by fundamentals; the degree of “undermigration” varies substantially across countries, with Spain exhibiting a greater degree of undermigration than other large EU countries.

Estimating the Long-Term Equilibrium of Migrants from Eastern Europe

Using long time-series for the German experience, Boeri and Brucker (2000) estimate an error correction model to quantify the long-term migration from AC-10 countries, and the expected length of time needed to reach the new equilibrium. The model posits that migration decision is taken as an investment in human capital whose returns are determined by expectations regarding future income. These expectations are assumed to be based on past values of the differential of per capita income and employment opportunities between the host and the home country, and is formalized with the following equation:

Δmsth.t=β1×Δln(yf/yh)t+β2×Δln(ef)t+β3×Δln(eh)t+β4×ln(yf/yh)t-1+β5×β6×ln(eh)t1+β7×msth.t-1+β8×FREE+β9×GUEST+β10×COU1NTRY

where mst h.t denotes the ratio of the stock of migrants to the population in the home country, y the per capita income (purchasing power parities), and e the employment rate; subscripts f and h denote respectively the foreign and home countries. FREE and GUEST are dummy variables respectively reflecting whether labor mobility between Germany and the accession country is unrestricted, or that a guest worker agreement exist; COUNTRY are country specific constants.

Using the estimated coefficients to extrapolate the migratory flows for other EU countries and assuming the introduction of free movement of labor in 2002, the model suggests that the initial net migration from AC-10 countries is expected to be around 335 thousand—with Germany and Austria receiving respectively about 65 and 12 percent of this flow. Also, estimates suggest that the steady state stock of migrants—some 3.9 million or 1.1 percent of the population of the EU—is reached in about 30 years.

Migration flows from Accession Countries to the EU-15

(In thousands)

article image
Source: Boeri T. and Brucker. H. (2001: EUROSTAT.

These data were not available at the time of the study.

Comparing the projections to the most recent data suggest that the projections for Spain were off by a factor often in 2002: only 4.1 thousand were expected to migrate to Spain compared with 43 thousand. This deviation and that of Germany highlight the difficulties in estimating these flows with any degree of precision.

G. Summary and Policy Conclusions

30. Spain has experienced income convergence consistently in the past decade, despite gradual losses in competitiveness. The stability orientation of macroeconomic policies and reforms in the labor and product markets were the two key pillars of this performance. Robust growth was accompanied by sizeable increases in trade flows, with the EU as its principal trading partner. But a persistent inflation differential with the euro area—not matched by a correspondingly strong productivity developments in the available data—indicate a gradual erosion of competitiveness relative to other euro area countries which is also revealed in a relatively faster rise in unit labor costs. Although exports have held up so far, export margins have been under pressure and EU enlargement holds the potential for increased competition for Spain’s share of the EU import market.

31. The empirical evidence indicates that overall EU enlargement offers a range of opportunities for Spain, and points to potential pressures in specific sectors. Specifically,

  • International trade. The existing level of trade between EU-15 and AC-10 countries bodes well for significant trade creation opportunities; these are estimated to be at least as large as those experienced in previous EU enlargements. Overall, Spain’s exports would not appear to come under additional pressure from the EU enlargement But there is a clear potential for specific sectors—notably machinery and equipment—to be displaced over time by the increased competition from AC-10 countries.

  • FDI. Likewise, the existing level of FDI suggests substantial potential for increased FDI. Overall, FDI in Spain would not appear to suffer following enlargement; taken literally the evidence suggests a small increase is possible. However, this must be taken with some caution given the inherent problems with FDI data, and anecdotal information pointing in the other direction.

  • European Structural Funds. Enlargement will reduce funds flowing into Spain over time. Although available estimates of the benefits to Spain vary, their positive contribution is likely to outlive them to the extent that these boosted infrastructure. With the future of the ESF currently under discussion, it is premature to access the impact that enlargement may have, but this is likely to be borne primarily by specific regions in the medium- and long-run

  • Migration. Available estimates in the literature and gravity model estimates indicate that migratory flows from the AC-10 countries are likely to remain small, particularly into countries that are further away, such as Portugal and Spain. In Spain, given the lack of cultural ties and an established beach head of immigrants from these countries, it is unlikely that migratory flows will have an economy-wide effect. Over time, some migration could materialize in specific labor market segments with limited overall impact.

32. In sum, EU enlargement offers Spain more opportunities than challenges; tackling the latter requires increasing the flexibility of the economy. Be it through increased trade or FDI, the EU-25 will offer new opportunities for exporters and firms, particularly in an environment where the stability orientation of policies continues to be maintained. As the effects of enlargement are felt to different degrees in different sectors over time, the importance of increased flexibility will become all the more evident. Reducing the adjustment costs involved in the needed reallocation of resources will thus require continued efforts to reform the economy, particularly labor and product markets.

APPENDIX: Data Definitions and Sources

article image

Regional Grouping, and Other Country Characteristics.

article image