Unions
Swings and roundabouts

Among the most notable features of recent US and European macroeconomic experience have been the steady rise in unemployment rates and the much quicker US recovery from the recessions of the 1970s and early 1980s. Several explanations have been offered. Some point to differences in countries' fiscal and monetary policies; others emphasize 'structural' changes affecting labour force composition and job mismatches; the 'New Classical Macroeconomics' focuses on errors in price expectations; another approach associates the US and European unemployment experiences with the degree to which wages have been 'excessive'.

The 'excessive wage' approach has received considerable attention in recent years, focusing especially on the failure of real wages in Europe to adjust downwards in response to oil and commodity price shocks and the productivity slowdown. But this approach does not yet have a firm theoretical foundation. For example, the conventional Keynesian wage rigidity models fail to explain why real wages are not bid down in the presence of involuntarily unemployed workers. In Discussion Paper No. 114, Assar Lindbeck and Research Fellow Dennis Snower address this issue by examining how the influence of trade unions over wage contracts can make economies less 'resilient', so that swings in unemployment tend to persist, leading to a wage-unemployment 'ratchet' and an upward trend in both real wages and unemployment. Different degrees of corporatism and of union power may therefore explain why the recession of the early 1980s was more protracted in Europe than in the United States.

Lindbeck and Snower assume that unions act mainly on behalf of their employed members, taking little if any account of the interests of the unemployed. In the authors' model, unions exercise power because they can manipulate the labour turnover costs borne by a firm. These costs can take a variety of forms. First, there are the costs of hiring, training, and firing employees. Second, union members may cooperate among themselves but not with non-union members, raising the productivity of union members above that of non-members, and they may even harass non- members. Third, a higher labour force turnover will reduce incentives when current work effort is rewarded wholly or partly in the future.

Because the firm incurs such costs from labour force turnover, incumbent employees have a greater chance of keeping their jobs than the unemployed have of acquiring them. Thus the existence of labour turnover costs gives union members the market power to raise their wages above the level at which non-members would be willing to work, without giving the employer sufficient incentive to replace union members with non-members. In other words, the union members are 'insiders', who have inherent advantages over the unprivileged and frequently unemployed 'outsiders'.
Lindbeck and Snower then consider the behaviour of the economy as a whole. Labour market decisions are made in two stages. First, the wage level is set, in the absence of full information concerning the current state of labour demand. Then, given this wage level, and when more information is available, firms determine employment levels. Unions attempt to maximize the utility of their incumbent members; this depends both on the level of wages negotiated and on the probability of being retained by the firm in the face of shocks to the economy and to the demand for labour. This probability in turn depends on the size of the incumbent labour force and on the wage level.

Suppose a recession occurs, due for example to an external shock such as an oil price rise. Wages will be held at the negotiated level and employment will fall. It is quite likely that the long-run employment prospects of the economy will not be bleak as those in the short run, so although the incumbent workforce will shrink, the remaining insiders will be more confident of being retained than before the shock, the authors argue. They will realize that they can now raise their wages without reducing their job security: higher wages are negotiated, discouraging firms from employing as many workers as they might otherwise have done.

Thus, even if the short-run fall in employment is reversed in the future, employment will be lower than before the shock. Unions have succeeded in negotiating higher wages in the meantime, and fewer workers are hired. Because of this, the authors argue that unfavourable swings in employment tend to persist, through the influence of trade unions on real wages. The more bargaining power unions have, the greater the wage increases insiders will be able to achieve in a slump, accentuating this 'persistence effect'.

In addition, Lindbeck and Snower suggest that cyclical macroeconomic fluctuations may be translated into permanent upward movements in real wages and unemployment. Since laid-off workers often lose influence over wage determination faster than newly hired workers gain such influence, an employment slump may reduce the insider workforce by more than an employment boom of equal magnitude would increase it. Consequently, adverse employment swings lead to wage increases of greater size than the wage reductions generated by favourable employment swings. In the long run, the authors conclude, real wages drift upwards and as a result the level of employment falls.

Lindbeck and Snower infer that union power over wages may therefore hinder an economy recovering from a recession, and that the greater the bargaining strength of unions and the higher the costs of labour turnover, the bleaker an economy's recovery prospects will be. In this light, according to the authors, the more widespread influence of unions in Europe than in the United States may help explain why European economies have found it more difficult to emerge from the recession of the early 1980s than the US economy. The same may also be said of sectors within economies. For example, unions are relatively influential in the US steel and car industries, where employment has suffered comparatively badly. But the unemployment persistence effect also works in reverse, the authors note. Union wage setting tends to perpetuate favourable variations in employment, so that union power will be less harmful in a boom than in a recession.


Union Activity and Economic Resilience
Assar Lindbeck and Dennis Snower

Discussion Paper No. 114, June 1986 (IM/ATE)