Real Wages too High?

The Chancellor recently cited evidence that a 1% reduction in real wages could lead to an increase in employment of between 0.5 and 1%. CEPR Research Fellow Steve Nickell challenged this view in a February 19 lunchtime meeting, one of a series of talks in which Research Fellows discuss policy-relevant research. They may also advance specific policy views, but these views are their own and not those of CEPR, which takes no institutional policy positions.

Nickell noted several reasons why unemployment might have risen to its current level. One explanation might lie in a rapid growth of the labour force. Yet in the past large increases had taken place without increased unemployment. Another explanation of unemployment which has been advanced is that real wages are too high: "If only workers were cheaper, firms would hire more of them.' Nickell argued that the issues were more complicated than this argument suggested. The relationship between unemployment and real wages was far from clear. Even if such a relationship were established it was not obvious how the government could influence real wages so as to reduce unemployment.

The level of real wages was not the outcome of labour market bargaining alone. It was essential to bear this in mind when thinking about real wages and unemployment, Nickell argued. Real wages also depend on prices, and analysis of the relationship between unemployment and real wages also must account for the behaviour of prices. One theory of price formation suggests that firms set prices as a mark-up on their costs or wages. This mark up is therefore the ratio of prices to wages; its reciprocal is therefore nothing but the ratio of wages to prices, or the real wage. Real wages seem to depend on the price-setting behaviour of firms and not on the behaviour of the labour market.

Empirical studies had shown, according to Nickell, that the mark- up depended on the pressure of demand and on the rate of wage inflation. In addition, labour market bargaining was influenced by the rate of price inflation, the level of unemployment and the degree of what might be called "wage push'. The interactions were complicated, but the conclusion was clear: the real wage was not decided in the labour market alone.

Nickell then discussed the simulations which had been reported in the Treasury Green Paper as well as those undertaken by the Macroeconomic Modelling Bureau at Warwick. These simulations suggested that with an unchanged level of aggregate demand a reduction in "wage push' would reduce real wages with little effect on employment or inflation. A reduction in wage push together with a higher level of aggregate demand also reduced real wages, while employment rose.
The message conveyed by these simulations was simple, argued Nickell. Policies which reduced wage push while stimulating aggregate demand were clearly desirable; previously these were called "incomes policies'. Nickell argued in conclusion that discussions of the effect of real wages on unemployment served no useful purpose. There was no clear relationship between aggregate real wages and employment and the analysis could be more usefully focussed on the behaviour of "wage push'.

In the lively discussion which followed, some agreed that the real wage was an endogenous variable - the outcome of the workings of the entire economy. It could not serve as a target for government policy, since it was not amenable to government control. Others present challenged Nickell's analysis. Economic theory suggests that as the price of a good falls, demand for it should increase, other things being equal. As real wages fell, demand for workers should rise and unemployment should fall. Nickell responded that this was an analysis based on the workings of a single market. In the aggregate other things were not equal, and the relationship between real wages and unemployment was not straightforward.