It is commonly thought that workers are reluctant to accept pay cuts and that this causes unemployment, which in turn can only be alleviated by relatively high inflation. But according to the first study of ''downward nominal wage rigidity'' to use pay data for individual workers, in practice, the UK labour market is remarkably flexible. In a paper published in the March 2000 issue of the Economic Journal, Jennifer Smith argues that many workers do accept pay cuts. This has important implications for monetary policy: if there is no downward rigidity, from the point of view of the labour market, there are few reasons why the Bank of England should not aim for an inflation target of 0% rather than the current 2.5%.
In order to prove that nominal wage rigidity is not a problem, Smith had to dig deep into the data. She found that a simple inspection of the data, which apparently shows 9% of UK workers with rigid pay, is misleading. All but 1% can be explained away as arising from long-term contracts, rounding and other measurement errors.
There is continuing debate in monetary policy circles about the correct level for the inflation target. In principle, the inflation target should be set at the ''optimal'' inflation rate for an economy. There are a number of factors that should influence this optimal rate. The amount of nominal wage rigidity in the economy is one of them.
Our understanding of the importance of nominal wage rigidity stretches back at least to Keynes'' General Theory. Keynes emphasised the ''psychological'' fact that workers do not like cuts in their money wage: nominal wages are rigid downwards. This rigidity can lead to unemployment as a result of falls in the demand for labour. But this can be countered at a macroeconomic level by allowing the inflation rate to rise. Essentially, to eliminate unemployment, the real wage – the ratio of the money wage to the price level – needs to fall. The real wage can fall if either the money wage falls or the price level rises. Since workers will not take money wage cuts, the only way to prevent unemployment is to allow prices to rise. Inflation ''greases the wheels'' of the labour market to counteract the ''grit'' of nominal wage rigidity.
If nominal rigidity does exist, there should be many workers with zero pay growth. These are workers who ''should'' have had a pay cut, but have ''stuck'' at zero because of a reluctance to see their pay fall. If the distribution of pay growth is graphed, these workers should form a ''spike'' at zero. Smith found just such a spike: 9% of workers have exactly zero pay growth, which is far higher than at any other growth rate.
But, Smith argues, it would be wrong to conclude immediately that nominal rigidity is serious enough to justify a high inflation target. She conducted further analysis into the true causes of the apparent zero spike and found that most of it can be explained away. Half the spike can be attributed to the fact that many workers are on long-term contracts during which their pay is fixed.
The temporary rigidity in pay during the contract will not have a big impact on unemployment. Furthermore, some of these workers ''should'' have had a pay rise, whereas unemployment will only result from rigid pay among workers who should have had a pay cut. A further 40% of the zero spike can be explained by rounding and other measurement errors. Some workers surveyed round their pay to convenient numbers. Sometimes their pay change will be sufficiently small that the same rounded number is used in consecutive years, which will generate spurious rigidity in pay.
According to Smith''s research, only 1% of workers have zero pay growth that might reflect unemployment-inducing downward nominal rigidity. This very small fraction is not sufficient to justify a high rate of inflation. On the basis of the labour market, the Bank of England might as well aim for an inflation target of zero.
''Nominal Wage Rigidity in the UK'' by Jennifer Smith is published in the March 2000 issue of the Economic Journal. Smith is at the University of Warwick.
