Unskilled Workers In America Lose Out From Their Abundant Numbers

Low-skilled workers in the United States are worse off than their counterparts in continental Europe. According to new research by Edwin Leuven, Hessel Oosterbeek and Hans van Ophem, this is not because they lack labour market institutions like employment protection and unions. Rather, it is because there are so many of them.

The study, which is published in the April Economic Journal, notes that wage inequality differs substantially across countries, and is larger in the United States than in continental Europe. This is especially true at the bottom of the wage distribution. Economists debate the causes of these differences. Some argue that differences in institutions across countries are the driving force, while others make a case for differences in the market forces of demand and supply.

According to the ''institutions explanation'', high minimum wages, employment protection and a different role for unions are responsible for the relatively favourable labour market position of the low skilled in countries like Germany, the Netherlands, Sweden and Switzerland.

According to the ''demand and supply explanation'', relative wages of low skilled workers in the United States are lower than elsewhere as a result of an abundant supply of low skilled workers in the United States. The difficulty of measuring skills so that they are comparable across countries complicates any testing of the demand and supply explanation. In an influential study, Cornell University economists Francine Blau and Lawrence Kahn measured skills as a composite of years of schooling and years of work experience.

On the basis of this skill measure, they divided the US workforce into three equal sized skill groups: low, intermediate and high. Employees in ten other countries were then also assigned to skill groups based on the cut-offs between the US skill groups. The test is relatively straightforward: if the average wage of low skilled workers compared to intermediate skilled workers is lower in the United States than in Germany, one would expect that low skilled workers are relatively more abundant (compared to intermediate skilled workers) in the United States than in Germany.

Based on this procedure, Blau and Kahn dismissed the demand and supply explanation because they found no relation between the relative wage position of a skill category and its relative supply. Leuven et al''s research argues that Blau and Kahn''s procedure breaks down because the skill measure they construct implicitly assumes that years of schooling and years of working experience are comparable across countries. It therefore ignores differences in education systems and post-school training systems across countries, and results in wrong rankings if these differences matter.

To illustrate the issue, Blau and Kahn''s numbers imply that half the German workforce belongs in the lowest third of the US workforce, a conclusion that is hard to believe. Instead of using education and training outcomes, Leuven and his colleagues'' study relies on information from literacy and numeracy tests that were developed with the explicit aim of being comparable across countries. Employees in fifteen different countries participated in the International Adult Literacy Survey.

Performing the analysis using this direct skill measure completely reverses the conclusion of Blau and Kahn. A larger relative net supply of a skill group is associated with lower relative wages of that same skill group. More specifically, low skilled employees are relatively worse off in the United States than their counterparts in continental Europe not because they lack the protective labour market institutions but because there are so many of them. Reducing the supply of low skilled employees may therefore be an effective means to improve their relative wages after all.

''Explaining International Differences in Male Skill Wage Differentials by Differences in Demand and Supply of Skill'' by Edwin Leuven, Hessel Oosterbeek and Hans van Ophem is published in the April 2004 issue of the Economic Journal. The authors are at the University of Amsterdam and the Tinbergen Institute.