V Policy Implications

Antonio Spilimbergo, Eswar Prasad, and Paolo Mauro
Published Date:
May 1999
  • ShareShare
Show Summary Details
Paolo Mauro Eswar Prasad and Antonio Spilimbergo

The international evidence presented in Section II shows that several European countries are characterized not only by high nationwide unemployment, but also by large differences in regional unemployment. Moreover, such differences have persisted for many years. Both Italy and Spain rank high in terms of these undesirable characteristics. A key policy question in these countries is how to promote the movement of capital toward high-unemployment regions—the adjustment mechanism that could play the main role in correcting regional unemployment imbalances—as well as the movement of labor toward low-unemployment regions.1 Using data sets and methodologies that are best suited to each country, the two case studies in Sections III and IV yield consistent, complementary results that point to the following policy implications.

The wage-bargaining system is de facto excessively centralized in both countries. As a result, wages in areas that have experienced persistently high unemployment cannot decline sufficiently in relation to those in low-unemployment areas to correct unemployment imbalances. In fact, the study on Italy finds that wages are similar in high-unemployment and low-unemployment regions, controlling for a number of demographic and educational characteristics of the workers; and the study on Spain finds that both real wages and unit labor costs are similar in high-unemployment and low-unemployment regions or provinces. These findings provide an argument in favor of decentralizing the wage-bargaining system to the firm level, so as to allow market forces to bid wages down in the high-unemployment areas.2 Lower unit labor costs would then attract new firms into the high-unemployment areas; in addition, lower real wages would constitute a further incentive for workers to migrate out of these areas.3

Other labor market institutions and policies that tend to set an artificial floor on wages, thereby hampering the creation of new jobs in high-unemployment areas, include minimum wages and unemployment benefits. The floor set by the minimum wage is likely to have adverse consequences for low-skilled labor force participants in areas where productivity is low, that is, typically areas characterized by high unemployment, although it probably does not constitute a binding constraint on wage setting for most other workers in Italy or Spain.4 Similarly, unemployment benefits tend to reduce job search efforts by the unemployed and to limit downward pressure on wages in areas where unemployment is high. Again, this consideration is more likely to be valid for the low skilled in low-productivity areas. A reduction—or at least the introduction of regional differentiation—in minimum wages and a tightening of eligibility for unemployment compensation could therefore help reduce geographic unemployment differences. These actions could be combined with targeted social welfare programs to protect the truly needy sections of the population.

Special unemployment protection programs often distort incentives, thereby perpetuating regional unemployment differences. The most striking illustration of this is Spain's special unemployment benefit program for temporary agricultural workers in certain depressed areas. This program tends to tie underemployed workers not only to a particular area but also to a certain sector where labor demand is unlikely to rise in the future. The elimination of such programs and their replacement with standard unemployment protection could therefore alleviate regional unemployment dispersion.

Lack of flexibility in the housing market also appears to hamper migration of workers. Restrictions often take the form of high transaction costs (including through taxation) on the purchase and sale of residential real estate, long minimum duration of rental contracts, as well as widespread rent control and public allocation of rental housing.5

The relative inefficiency of formal labor market intermediation mechanisms in facilitating job matching between employers in low-unemployment regions and labor force participants in high-unemployment regions also limits migration. This has been the case in Italy, where formal job intermediation has been restricted to public sector employment agencies that have proved ineffectual at providing cross-regional job listings. In Spain, the statutory monopoly of job placement held by the National Employment Institute (INEM) was abolished and the establishment of non-profit private agencies permitted in 1994. In both countries, the private sector should be allowed to play a much greater role in job matching.

Efforts to improve schooling and retrain workers, particularly in areas with higher unemployment, would also help correct unemployment imbalances. Higher education is usually associated with lower unemployment. Some of the areas in Italy and Spain that are characterized by higher unemployment rates also rank among the worst in terms of levels of education achieved. Moreover, the evidence from Spain suggests that workers with higher education migrate more promptly in response to shocks in regional labor demand.

Among the possible instruments that this paper has not analyzed in detail, active regional policy, such as government-provided infrastructure, may be the one currently used to the greatest extent, both at the national level and at the EU level. One example is the structural funds provided to regions whose per capita GDP is below 75 percent of the EU average. In principle, active regional policy could help raise productivity in the high-unemployment areas, thereby making them a more attractive location for firms. At the implementation stage, however, it is often difficult to pick the right sectors or the right type of infrastructure. For example, Italy's active regional policy between the early 1960s and the mid 1970s, when large-scale industrial plants were built in the south, failed to engender commensurate spillovers in terms of output and employment. Although an analysis of such policies lies outside the scope of this paper, past experience does not bode well for future attempts in this area.

Finally, regional policies should be consistent over time. Both workers and capital make location decisions based on long-term horizons. For this reason, effective regional policies should guarantee stability over time because temporary measures are likely to have negligible effects. In the context of wage-bargaining decentralization, for instance, long-lasting measures are likely to be more effective. The patti territoriali and contratti d'area, recently introduced in Italy to permit temporary derogations from national wage contracts, have failed to promote the migration of firms to high-unemployment areas partly because they are perceived as providing only temporary wage flexibility.

In addition to policies that could be implemented at the national level, the country studies also provide two key lessons for the euro area as a whole. First, the studies emphasize the importance of relative wage flexibility to correct unemployment imbalances. The expectation that the euro area will be characterized by a low-inflation regime, which—while obviously desirable—will tend to hinder relative wage adjustment, strengthens the case for the policy measures recommended above. Second, a decentralized wage-bargaining system is even more crucial in the euro area, in light of large productivity differences across countries. Calls for “equal work, equal pay” by workers in low-productivity countries, which could well arise once the common currency facilitates comparison of wage levels among euro countries, would tend to lead to higher unemployment in low-productivity countries. The creation of a European wide wage-bargaining system would hamper the ability of individual countries to respond to asymmetric shocks.

The rate of migration also reflects cultural differences across countries, thus Europe could have a significantly lower rate of migration than the United States even with the same economic incentives. For a study on the importance of cultural attitudes versus economic incentives, see Spilimbergo and Ubeda (1996).

Calmfors and Driffill (1988) argue that both highly centralized and highly decentralized wage-bargaining systems perform better than systems with an intermediate degree of centralization, such as those in Italy and Spain. Their arguments, combined with the considerations above, reinforce the need to move to a highly decentralized system.

Hanson and Spilimbergo (forthcoming) find that international migration between Mexico and the United States is highly sensitive to variations in wage differentials even at a monthly frequency and in the presence of border enforcement.

In Spain, the statutory nationwide minimum wage amounts to one-third of average wages for adults. The minimum wage for workers below 18 years of age with training contracts amounts to 85 percent of the minimum wage for adults. In Italy, there is no statutory minimum wage, but the minimum wages set in collective agreements typically amount to 70 percent of the average wage, a higher percentage than in most other European countries.

While it is difficult to compare restrictions across countries, the high share of owner-occupation in the total populations of Italy and Spain—70 percent and 75 percent, respectively—compared with about 60 percent in a sample of more than 20 OECD countries (Oswald, 1998)—is an indication that rental market restrictions are relatively stringent in these countries.


    CalmforsLars and JohnDriffill1988Bargaining Structure Corporatism and Macroeconomic Performance” Economic Policy Vol. 3(April) pp. 1461.

    HansonGordon and AntonioSpilimbergoforthcoming, “Illegal Immigration, Border Enforcement, and Relative Wages: Evidence from Apprehensions at the U.S.- Mexico BorderAmerican Economic Review.

    OswaldAndrew1998A Conjecture on the Explanation for High Unemployment in the Industrialized Nations” Economic Research Papers no. 475(University of Warwick).

    SpilimbergoAntonio and LuisUbeda1996Multiple Equilibria in Geographic Labor Mobility” (Ph.D. dissertation;Cambridge, Massachusetts: Massachusetts Institute of Technology).

    Other Resources Citing This Publication