How should the European Union reform its ‘structural funds’ programme? Diego Puga suggests how the ‘new economic geography’ can help design an effective new regional policy.

Over the past 15 years, income differences across EU members have fallen but inequalities between regions within these countries have risen. As a result, regional problems have become less clearly seen as large disadvantaged regions and more as concentrated pockets of poverty and unemployment. EU regional policy has tried to tackle these problems, primarily through the ‘structural funds’ programme. Yet despite continuing resource transfers, many regions persistently fail to catch up.

With the current structural funds arrangement due to expire in 1999 and reform high on the EU agenda, it is useful to go back to the principles of regional policy. One of the main justifications for such policy is to secure ‘equality of opportunity’: the combination of inherent disadvantages and insufficient fiscal capacity in less favoured regions requires EU-wide policy to provide them with the infrastructure to be able to compete on an equal footing with more developed regions.

Recent location theories (the ‘new economic geography’ outlined in the previous article) provide two additional justifications. The first is that the aggregate gains from economic integration are not evenly distributed between the integrating countries. The theory suggests that ultimately the same forces that initially foster divergent regional outcomes will reverse them, but such a process may take a long time. In the meantime, inter-regional transfers enable aggregate gains to be more evenly shared by all.

The second justification provided by recent location theories is that relatively small interventions can have strong and wide effects. Helping a less favoured region attain a critical mass of industrial activity can enable it to take off. But it is important to select projects that have large ‘multiplier’ effects and to deter regions from competing for investments.

Choosing between projects demands an understanding of their full impact on regional inequalities. With infrastructure provision, for example, what are the effects of a road project connecting regions with different levels of industrialization? Some growth economists would treat infrastructure as merely another input into production: the road project will simply be an addition to the infrastructure stock, which will help the regions grow. But this approach misses the role of transport infrastructure in facilitating the movement of goods and people.

A cost-benefit analysis would probably look at the direct impact of a road project in terms of reducing transport costs and inducing changes in the number of journeys undertaken. But ‘spatial spillovers’, both positive and negative, can make the effects of a project both stronger and wider.

On the positive side, for example, better communications can make a less developed region a more attractive location for firms by giving them improved access to the inputs and markets of more developed regions. A rise in the activity of one industry in the region can induce another supplying industry to produce more efficiently, and so on in a process of ‘cumulative causation’. On the negative side, an improved transport infrastructure makes it easier for firms in richer regions to supply poorer regions at a distance, thus discouraging industrialization in the poorer regions.

The overall impact of an infrastructure project depends not only on the nature of the project itself but also on the economic environment. For example, it has beenargued that infrastructure improvements have worsened the convergence prospects of the Italian Mezzogiorno. Lacking the industrial base and market size of the northern regions but with similar factor costs, local firms have lost out to northern competitors as better communications have lowered the natural protection they initially enjoyed.

Since regional inequalities are increasingly within rather than across countries, the European Commission has proposed decentralizing administration of the structural funds to local governments. The rationale is that they have better information on local needs and the costs of meeting them. But regional governments are unlikely to assess accurately the ‘pecuniary externalities’ created by local activities, which have effects beyond regional political boundaries. So the optimal degree of decentralization is a compromise between exploiting better information on local conditions (by delegating the decision to a small local jurisdiction) and taking proper account of broader repercussions (by delegating powers to a larger jurisdiction). The political economy question of which level of government is more likely to yield to the pressures of special interest groups should also be considered.

The issue of decentralization is also affected by new location theories, which suggest that the spatial distribution of economic activity is just one of many possible outcomes. Progress towards a given outcome is punctuated by critical points at which just a small difference can determine which region gets which industrial sector. Hence, regional governments have incentives to compete for sectors that they consider attractive.

All of this suggests that funds that can be used to attract particular investments should be administered centrally and subject to clear rules. Otherwise, we are likely to see firms shopping for aid while regional governments compete for the activities they want. That would not only be inefficient; it would also undermine the credibility of European regional policy.

But controlling the use of EU regional funds is not sufficient. Subsidies offered to firms usually adopt the form of aid from individual members. The EU has set differential ceilings on state aid, but in practice, the less prosperous countries face more restrictive budgetary constraints of their own. State aid levels tend, therefore, to be roughly proportional to national GDP levels. Tighter limits on the richer countries may be needed, even if only to level the playing field.

Further European integration is also likely to intensify the process of regional specialization: firms will move closer to firms in related activities to exploit positive externalities; and they will move away from firms in unrelated activities to avoid having to compete with them for immobile factors and non-tradable goods and services. This will have two consequences for policy.

First, greater regional specialization will increase the need for schemes designed to enable workers to move from locally declining to locally expanding sectors. It will also increase the general need for funding for regions in difficulties because of the nature of their sectoral specialization. But while the need for training is clear, what kind of training is not. If a sector with a large share of a region’s employment suffers a temporary shock, sector-specific skills will increase the region’s comparative advantage, helping it withstand the shock. But if the shock is permanent, then general skills are going to be more desirable to help workers adapt to alternative jobs.

Second, in a more specialized Europe, sector-specific shocks will increasingly become region-specific shocks. This will call for the structural funds to provide some kind of insurance mechanism. The small size of the EU budget and the rigidity of rules determining each member’s contribution towards that budget imply that national governments will remain better placed than EU regional policy to provide insurance against transitory shocks. Yet if too tough fiscal constraints are attached to the final phase of economic and monetary union, it may be increasingly difficult even for national governments to do that job.