Abstract:
We use plant-level linked employer-employee data from Finland to estimate production functions where also employee characteristics (average age and education, and sex composition) are included. We also estimate similar models for wages to examine whether wages are based on productivity. Our aim is to explain productivity besides manufacturing, also in services. For the service sector plants, no data on capital input, working hours, or value added is available, and productivity has to be measured by sales per employee. We use a stepwise procedure to examine whether the results for manufacturing are affected when less satisfactory data is used. Then we proceed to estimate the final model for manufacturing and services combined. The effect of age on productivity is negative, but wages show strong positive age effects. Higher educational level leads to higher wage, but there is a clear productivity difference between nontechnical and technical education. Some of the productivity effects of technical education are negative. Wage-productivity gaps (relative to the reference group, basic education) are positive for the highest level of technical education, but negative for the highest non-technical education. The share of female workers is negatively related to productivity. Also the wage effect is negative, but smaller in absolute value, leading to a positive female wage-productivity gap. However, the negative productivity effect disappears and the gap is negative if the model is estimated with fixed plant effects. – Productivity ; wages ; education ; age ; gender wage gap ; linked employeremployee data
