Evidence from matched employer-employee data
Technological changes and skill composition
In this paper we carry out a structural analysis of the changes in relative wages and the effect of skill-biased technical changes on the development of relative labour demand. The empirical model is based on a production function in which capital and two types of labour – high skilled and low skilled – are specified as inputs. The classification of high- and low-skilled workers is based on information extracted from estimated wage equations. We estimate separate wage equations for two narrowly defined Norwegian industries, Electrical and optical equipment and Machinery, using unbalanced employer-employee panel data for the period 1995- 2004. Both a skill premium, given by observed and unobserved individual characteristics, and variables unrelated to skill contribute to the observed wages. Each individual-observation is then divided into skill categories according to the predicted skill premium. The results show that in both industries, the share of highskilled man-hours has increased by 20 per cent from 1996 to 2004 while the relative wage between high skilled and low skilled is rather constant. In Electrical and optical equipment, a high-tech industry, all of the increase in the relative demand for high-skilled is accounted for by skill-biased technical change, compared to about 50 per cent in Machinery (a traditional manufacturing industry). In addition we find a strong link between innovations and investments in the long run, although the speed of adjustment of capital towards the equilibrium path appears quite slow.
About the publication
Technological changes and skill composition. Evidence from matched employer-employee data
Øivind Anti Nilsen, Arvid Raknerud, Marina Rybalka, Terje Skjerpen
- Series and number
Methods and documentation
- ISBN (online)
- ISBN (printed)
- About Documents
Documentation, descriptions of methods, models and standards are published in the series Documents.