Interview with Mr. Johann Van Overtveldt, editor of the Belgian Newspaper "Trends Magazine"

November 1999

Q - You have always analysed unemployment strictly in terms of the labor market. To what extent do aggregate demand and the businesscycle drive the evolution of the unemployment rate?

A - In the short and even the medium run macroeconomic events have an important impact on the evolution of unemployment. However, in the long run, nominal prices and wages adjust to bring unemployment back to some equilibrium level, which reflects many real factors such as microeconomic frictions, the economy's distribution of skills, its underlying growth prospects, real interest rates, and also labor market institutions. So, that level does not only depend on labor market institutions but also on other factors. But it does not depend on past economic flcutuations. In the case of Europe, both labor market rigidities and the slowdown in growth have contributed to increasing unemployment. But the slowdown in growth also occured on the other side of the Atlantic, where it dis increase unemployment but by much less than in Europe. So the difference between Europe and the U.S. is explained by labor market rigidities. In the sixties employment performance in Europe was good despite labor market rigidities, because strong growth was making such rigidities less of a burden. For example employers did not care about employment protection legislation because they could reduce their workforce by just letting their employees leave to another job. The slowdown in growth made rigidities more binding and lifted Europe to a higher unemployment level than the U.S. Another way to put it is by saying that because of rigidities, a growth slowdown is more harmful to employment in Europe than in the U.S. It is also true that macroeconomic policy has been very different. In the U.S. the Fed has pursued a policy of fine tuning, raising interest rates when inflation was picking up but lowering them when unemployment started to be an issue. In Europe macroeconomic policy has been much more restrictive, first in the eighties with the so-called "competitive disinflation", then in the nineties with the Maastricht convergence process. Such restrictive policies have contributed to higher unemployment during a substantial amount of time, but again they do not explain that equilibrium unemployment is permanently higher in Europe than in the U.S. Or maybe only indirectly, as political capital has been spent to meet convergence criteria instead of selling labor market reform to the public.

Q - You analysed unemployment as being largely the result of institutional arrangements that are the result of a political equilibrium, in itself the result of a power game between different interest groups. Could you clarify and give examples.

A - Many labor market institutions (not all, though) help some groups of incumbent workers to increase their bargaining power in wage-setting and protect the rents associated with their jobs. It is as if these workers were collectively organized in a labor union that directly set their wages; but, instead of doing that, they manipulate their own wages indirectly by voting and lobbying for economy-wide regulations that make their bargaining position more comfortable at the individual level. This is why we observe arrangements such as minimum wages, work rules, and employment protection legislation. Even unemployment benefits can be interpreted as a way for incumbent employees to increase their wages. That logic also applies negatively, in the sense that institutions that in principle should help the unemployed rather than incumbent employees end up being designed in the interests of the latter. One example is active labor market policies, which, according to economists, should be designed to bring the long-term unemployed back into active job seeking, thus increasing their moderating impact on wages; in practice, many programs seem to rather keep the long-term unemployed "quiet" into relief jobs or training of doubtful value, with very little pressure to seriously evaluate the programme--that is, the programmes reduce competition for jobs between the employed and the unemployed instead of increasing it. In my view, the coalition that benefits from this set of rigidities is a "middle-class" one. It contains relatively unskilled workers whose wages are increased, and relatively skilled workers who may not gain in terms of wages but buy social peace and reduced redistributive pressure, in part because their "opponents" are split into those who keep their jobs and those who become unemployed. Those who lose include the unemployed, the least skilled and the most skilled, and groups of workers who could benefit from flexibility, such as the young and women. It is quite a heterogeneous group, so it is difficult to put together a stable coalition in favor of labor market reform. Traditional thinking about how to reform as always ignored political constraints and often come up with proposals that turned out to be unfeasible because of political opposition. My approach has implications about which reform design, timing, and economic environment are most likely to be successful. For example, reductions in employment protection legislation that only affect new contracts while leaving the same degree of employment protection for incumbent employees are more likely to be politically viable, because they are designed so as to preserve the interests of the politically important coalitions. Indeed, one can show that reforms made in that way generate unanimity between the employed and the unemployed. And it is in that fashion that actual progress has been made in Europe, where flexibility has been achieved by introducing and/or liberalizing the use of temporary contracts at the margin. Another prediction is that higher exposure of the employed to unemployment makes them more likely to gain from reform, so that times of high exposure should be taken advantage of to implement a reform. This seems to be again compatible with the data, that say that reforms of the type just mentioned typically occur at times of rising unemployment, when insiders are most exposed. By contrast, the economy may be locked into a situation where the employed are so protected and have so little exposure that it is very difficult to implement a reform which pleases them. This seems to be the case of Italy, while in the U.K., where exposure was always higher than in the rest of Europe, reform could indeed take place in the first part of the eighties.

Q - In Belgium a strong belief exists that the unemployment is largely the result of the very high social security taxes, especially those weighing on employers. Please comment.

A - In principle, it does not matter whether the tax is imposed on employees or on employers. What really matters is whether the after-tax wage can adjust downwards in order to offset an increase in social security contributions. I believe this is true of skilled workers but not of unskilled workers, because they are constrained by floors determined by minimum wages, unemployment benefits, and so on. It is therefore clear that a substantial part of unskilled unemployment is due to these taxes, and that if we could reduce these taxes permanently and credibly employment would improve. The question is how to finance such a change. The ideal solution is to reduce social security spending, by making health expenditures more efficient and increasing the age of retirement. Another possibility is to increase taxation on capital. I think it is not a good idea because capital is very mobile and capital taxation is pretty distortionary. It is true that a 1 % tax on capital generates less unemployment than a 1 % tax on unskilled labour. But we need more than a 1 % tax on capital to make up for the tax revenues lost by a 1 % tax cut on unskilled labor. In the end, that option turns out to be unattractive in order to reduce unemployment. Next, one can increase social security contributions on skilled labour, which is O.K. but reduces the incentives to acquire human capital and may generate a "poverty trap". My preferred option, in the case of France, is to cut into the vast array of labour market programmes (training, hiring subsidies, preretirement subsisies, etc.), which represent 200 Billion Francs and are very inefficient, because they are constantly changed with considerabe discretion by the French government.

Q - How do you evaluate the 35-hours-week-initiative in France?

A - This is the ideal recipe for a demagogical politician. You tell everybody that they are going to work less for the same wage and that on top of that unemployment is going to decrease! No wonder both right and left have written it on their electoral programme. The real mystery is why so many French people swallowed such an absurdity. The idea that reducing working time creates jobs is based on the fallacy that total output, or total output growth, is fixed, and consequently so is the total number of working hours. Unfortunately, this fallacy is reinforced by the observation that the labor market has failed to absorb new entrants, giving the false impression that the total amount of work cannot be extended, when this failure is due to the disfunctioning of the labor market and the rigidity of the wage adjustment mechanism. We know that a well functioning labor market can absorb any increase in the supply of labor, as was observed in the nineteenth century, in the sixties, and above all in the U.S. Over the last two decades, this country has absorbed twenty-five million immigrants into its labor market, in a context where working hours were actually increasing rather than decreasing! So, instead of attacking the rigidities that prevent the labor market from adjusting, we're introducing a new constraint on firms, which, I am convinced, will further INCREASE unemployment. The argument is simple: people who work less produce less, and therefore must be paid less. But in order to force them to eventually accept pay reductions, labor market conditions must deteriorate, i.e., unemployment must increase. The harmful effect of the 35 hour week will be reinforced by the fact that France is now an EMU member. The increase in labor costs that it will bring about in the short and medium run will not be offset by a devaluation, as happened in 1982. This will deepen its contractionary effect in France, and at the same time create a large country-specific shock which will create tensions within the EMU and put the cohesion of member countries at a test. The 35-hour week will increase the dual structure of the French labor market, deepening the gap between protected employees that have everything and unprotected ones that have nothing. It will not apply to management, which will feel even greater pressure at work from the reduced availability of employees, nor will it affect firms of less than 10 employees, which will again bear the burden of flexibility and competitiveness. Finally, if there is one sector where employment may increase, it is the public sector. This sector is not exposed to competition and the demand for its services such as health and education is quite inelastic. Thus the burden of government will increase, in a country where it is already record high by international standards.

Q - What consequences will EMU have for the labor market in the participating countries?

A - EMU will increase the difficulty of implementing a large reform of the labor market because such large changes are best accompanied by an exchange rate realignment, which will not be available. Many people argue that there will be incentives for increasing labor market flexibility because this is a substitute for stabilization using active monetary policy. I am not totally convinced by that argument. It is true in some respects, but not others. For example, it is true that having greater real wage flexibility will make employment and output more stable, and that is a substitute for active monetary policy. So EMU should increase the value of reforms that make real wages more flexible, although we are only talking about moderate reforms because of my previous argument about large reforms. On the other hand, I can also come up with examples where it is the incentive for rigidity which is enhanced. For example, employment protection legislation makes the response of employment to shocks more sluggish, so it also has a stabilizing effect. So one can well argue that EMu will make employment protection more stringent. THis can take place very easily: think of a cost shock in one country. The goverment cannot devalue, firms start laying off people, and to prevent that employment protection is reinforced. Overall, then, EMU does not affect reform incentives in a clear direction. But it does make big unilateral moves more costly. The obvious consequence is that there will be more need for coordination of labor market reforms across member countries, with a clear participation of the ECB, which should commit to adopt the adequate path for monetary policy should these reforms be simultaneously adopted. This is far from easy, for two reasons. First, labor market institutions are not that similar, so that which reform is needed differs across countries--this is reinforced by diverging views about the functioning of the economy, cf. the French belief that working time reduction solves it all. Second, the ECB up to now is locked into a mandate that is too narrowly defined -- price stability --, which may prevent it from actively participating in such a coordinated reform.

Q - Against the background of your research, how would you describe/analyze the huge difference between Europe and the USA as far as job creating capability is concerned?

The difference is not as huge as is often stated. It seems that European firms create and destroy a large number of jobs every year, which suggests a more dynamic picture than what we get from looking, say, at unemployment durations and job finding probabilities, that are much higher and lower, respectively, in Europe than in the U.S.. At the same time, while people find jobs very quickly in the U.S., many of these jobs end up being very short-lived, and there is a problem of unemployment recurrence, in the sense that a pool of workers constantly moves between unemployment and bad jobs; in Europe these people, instead of moving around, end up being long-term unemployed. It is clear that the labor market functions very well in the U.S. and poorly in Continental Europe. But I would not label the difference in terms of outcomes as "huge". It is actually smaller than I would have expected.