Nickell and Bell's chapter 10 provides valuable insights on a theme that keeps recurring in economic debates in the world at large: ‘can a shift in the relative demand for labour in favour of skilled workers and against unskilled workers explain an increase in the aggregate rate of unemployment?’ As Krugman (1994) and others have previously argued, the answer is clearly ‘yes’ if a wage floor (e.g. minimum wage) prevents the wage rate for unskilled workers from dropping, leading to extra unskilled unemployment that becomes part of aggregate unemployment. The difficulties, as we shall see below, are in finding the quantitative counterpart of the relative demand shift in microeconomic data, and in attributing to that demand shift more than a small part of the overall rise in unemployment that has plagued Europe since the early 1980s.

The title of chapter 10 proclaims that it is about payroll taxes, but it is not, at least after the first page. Any link between the payroll tax and unemployment is dismissed by ‘a glance at Denmark’ and by figure 10.1, which shows no significant relationship between unit labour costs and payroll tax rates. But figure 10.1 is not convincing, because there are many causes of the cross-country variation in unit labour costs displayed there. A role for the payroll tax that is not evident in the simple bivariate regression reported by the authors might emerge in a full-blown multiple regression study.