Cyclical Earnings Variation and the Composition of Employment
Social Science Research Network
During the economic upswing in New Zealand between 1999 and 2007, employment increased by over 20 percent and average real earnings increased by 9 percent. It is plausible that many low-skilled and low-paid people were attracted into work over this period, changing the composition of employment and depressing average real earnings. Similarly, the boom may have encouraged the market entry of new firms, typically with lower than average productivity and paying lower than average wages, which
... also have depressed average real earnings. This paper uses Statistics New Zealand's Linked Employer-Employee Database (LEED) to assess the extent and impact of such changes in the employment composition of workers and firms over this period. LEED provides comprehensive coverage of all wage and salary employment since 1999. It enables longitudinal linking of both workers and firms, and also of the jobs that link them. As the LEED data do not directly measure hours worked or hourly wages, we construct a measure of the full-time equivalent (FTE) annual earnings rate associated with each job observed. Our analysis uses a linear model that regresses log(FTE annual earnings) on worker demographics (sex and age) and aggregate male and female time effects, and also controls for the effects of constant unobserved worker and firm specific factors. We use the estimates from this model, together with the employment transition patterns of workers and firms over the period, to investigate in detail the effects of compositional change on average earnings. First, we find that workers who enter employment during the period have 19 percent lower earnings than the average worker over the period. This compares with workers who exit employment during the sample period having 2 percent lower average earnings, and workers who are employed in each of the eight years having 7 percent higher earnings than the average worker over the period. Similarly, we find that entering firms pay 8 percent lower earnings than average. In contrast to workers, exiting firms also pay 8 percent lower earnings on average, and firms who employed workers in each year pay 2 percent higher earnings than average. In contrast to the 9 percent measured increase in real earnings, the model estimates imply that, if there had been no change in the composition of workers and firms over the period, average real earnings would have increased by 15 percent. This 6 percent difference is mainly due to a 5 percent decline in the average earnings premiums associated with workers. About 60 percent of this worker effect is attributed to new entrants over the period being lower-paid on average, 25 percent to existing lowerpaid employees working more hours, and 15 percent to lower-paid intermittently employed workers working more hours. Furthermore, about 60 percent of the entry effect is due to new workers entering low-earnings industries, and 40 percent is due to the new workers earning below-average earnings within industries. The changing employment intensity contribution of continuing workers is largely due to withinindustry intensity changes, rather than across-industry reallocation of those workers. We also estimate that there was a 1 percent decline in average earnings associated with firms over the period. This decline was largely due to new firms entering production having lower earnings premiums than those of existing firms, but is offset to a small degree by an increase in employment by firms with higher earnings premiums. Three industry case studies (manufacturing, construction, and property and business services) look fairly representative of the pattern for all firms. Our results imply that annual rate of growth of composition-adjusted earnings was 2.1 percent over the period, compared with 1.3 percent for measured raw earnings. It is plausible that a similar type of composition change effects also apply to changes in average labour productivity, although the magnitude of the effects on productivity will depend also on factors such as changes in technology and other inputs.