May 19, 2002

Thinking About European Unemployment

If you leave to one side the efficiency losses associated with western Europe's Common Agricultural Policy--its farm policy--and the poverty caused by European barriers to developing-country imports, there are two dimensions along which pan-European economic policy over the last generation can be severely criticized: the slowness with which the information-technology revolution has spread to Europe, and the extraordinary persistence of high european unemployment. Let me put the slow coming of the IT revolution to Europe to one side, and think only about unemployment.

Surely the most thoughtful and brilliant macroeconomist trained in the late 1970s is M.I.T.'s Olivier Blanchard. In his Robbins Lectures he turns his mind to summarizing what he knows about the nature, causes, persistence, and prospects of and for European unemployment. [Olivier Blanchard (forthcoming 2002?), The Economics of Unemployment: Shocks, Institutions, and Interactions (Cambridge: MIT Press:).] He gives us THE WORD.

What does he say?

The story he tells is a complicated one: it involves a slowdown in technological progress; a lag in recognition that keeps this slowdown from being accompanied by an immediate slowdown in aspirations for real wage growth; barriers to entry in product markets, and their erosion as a result of product-market deregulation; barriers to labor market flexibility, and their erosion as a result of labor-market deregulation; costs to firms of firing workers, the resulting creation of labor-market 'sclerosis' and high long-term unemployment, and the persistent difficulty in building a satisfactory theory to linke high long-term unemployment to a badly-performing labor market.

The overall lesson is neoconservative: Blanchard believes that policies have been set in motion--reductions in workers' bargaining power--that will yield large decreases in unemployment and large increases in real wages over the next decade or so. But even though the reduction in union power has been ongoing for fifteen years, so far its only large-scale visible effect has been... a reduction in the wage share. The reason is straightforward: in Blanchard's model, reductions in the real wage affect employment principally by inducing new firms to enter the market, and not by changing the number of workers that existing firms want to employ; thus the increases in unemployment that are to follow from reductions in workers' bargaining power are not "short run" but "long run" phenomena, requiring a lengthy process of firm start-up and expansion.

Why these beliefs--that high labor union power is a powerful source of unemployment, and that reducing labor union power increases employment only with very long lags? It turns out to rest on a basic modelling decision: Blanchard models the bargaining process between firms and workers as one in which (a) both sides agree to set the level of employment at whatever maximizes the firm's producer-side surplus, and (b) both sides then split the rents earned according to union power. Thus in Blanchard's model you cannot have a short-run situation in which firms respond to reduced wages and union power by increasing hiring--because union power and the wage level never have any effect on hiring in the first place. The channel through which changes in union power affect unemployment is a long-run one, that requires the entry and establishment of new firms before employment can rise.

I find this ingenious, but unsatisfactory as an account of why European unemployment has remained stubbornly high fifteen years after the wage share of national product began to fall sharply. It seems to me that perhaps too complicated an analytical apparatus has been deployed here. What's wrong with the belief that fifteen years ago a large chunk of European unemployment was "classical"--produced by a real wage that was stuck at too high a level, the result both of insider-outsider union dynamics and unrealistic aspirations for real wage growth--and that today a large chunk of European unemployment is "Keynesian"--the result of the fact that aggregate demand is too low?

I am profoundly ignorant about the European situation--I have to borrow an informed opinion from Andy Rose up at the Business School whenever I need one--but that European unemployment has shifted over the 1990s from primarily classical to primarily Keynesian would be the first hypothesis I would entertain, yet it is not clear to me from his Robbins Lectures why Olivier Blanchard--who is, on this topic and others, Il Maestro di Color che Sanno--dismisses it.

Posted by DeLong at May 19, 2002 02:38 PM | TrackBack

Post a comment