Changes in UK employment and non-employment over the 1980s have occurred across three main dimensions; gender, skill and type of work. In Discussion Paper No. 1356, Research Fellow Jonathan Haskel focuses on the changes in skilled employment with analysis based on a new data set which measures industry skill mixes using detailed data on individual occupations. This is in contrast to much of the previous literature which uses data on non-manual employment as a proxy for the skill mix in a sector. An analysis based on skills measured in terms of education and experience is more relevant to the debate about the impact of education and training, where programmes are usually designed to raise skill levels rather than turn individuals into non-manual workers. Haskel then combines this information with other data sources on trade, the use of computers, product and labour market information.

Using these data the author seeks to document and explain the rise in skilled relative to unskilled employment in UK manufacturing over the 1980s. In particular, he investigates the two main hypotheses that have been proposed to explain the fall in demand for unskilled labour. The first centres on trade. The argument here is that developing countries, who have a predominantly unskilled labour-force, specialize in unskilled-intensive goods. As trade increases, the demand for unskilled workers in developed countries falls and the industry mix becomes more skill-intensive. The second hypothesis centres on technology. Here it is suggested that technological change has increased the demand for skilled labour relative to unskilled over the 1980s. The major finding is that technology seems to be of greater importance, with trade playing a minor role. Haskel’s calculations of the effect of trade and of computers on relative skill demand within industries indicate that between 1981 and 1989 computers raised the skilled/unskilled employment ratio by around 1.5% (the ratio rose by 4.4% over this period). This calculation turns out to be robust to the use of alternative statistical techniques and the inclusion of other variables. It is also consistent with evidence from case studies and other econometric work.