Inequalities in the Labor market
A few weeks ago, GRAPE team presented an article on inequalities in the United States, where the researchers questioned the idea of the country as a land of opportunities. Today, we present another ASSA article dealing with a similar problem: the relation between the labor market and the rise in inequalities, two topics that we care about in GRAPE. Hyeon-Kyeong Kim and Peter Scott (2013: working paper) compare the earnings of employees enjoying a full-time, open-ended contract with the earnings of employees having either a fixed- or a part-time contract (in other words, non-regular contracts). Their analysis is timely, as EU countries experience a surge in both fixed and part time employment. According to Eurostat, the percentage of part-time workers in the EU increased 25% in the last ten years; while full time employment grew by less than 1%. Poland is somehow off this tendency as it is one of the four countries where the proportion of part-time employees fell. Temporary employment also increased over the period: 15% for the EU and an astonishing 71% for Poland.
Kim and Scott presented a modified version of the wage efficiency model. In this model, people can obtain a full-time employment only after an evaluation period (for instance an internship). This period can be considered a part of a screening device, which allows employers to know employees' productivity before engaging in a long term economic relation. Employees, who understand the mechanism, willingly make additional effort during the internship expecting that the employers will reward them with a permanent position. Hence, during this period, employees make an effort which equals the current salary plus the discounted stream of profits from full-time employment (given a certain probability of getting hired). The screening device serves then another purpose, it can be used to maximize the revenue of the employer, as productivity exceeds wage during the period. Moreover, if the employer wants to increase the effort exerted by the whole staff, she only needs to offer a raise to full-time employees, which at the same time increases the expected earnings of part/fixed time employees.
Though the model was not estimated empirically by the authors, they did provide simulations for the Korean economy. In them, the authors showed the effect of the labor reforms introduced in the country since the mid 90's. These reforms favoured labor flexibility, as they facilitated workers dismissal and increased the number of occupations were temporary employment was allowed. The simulations showed that the introduction of the reforms resulted in larger differences in payment between full-time employment and temporary employees; and in an increase in the number of temporary employees. Finally, the model also predicts that the effort exerted during the screening period is larger than after having a permanent contract.
The conclusions from their paper should also turn a warning sign on in the European Union, where these types of employment have grown significantly over the last decade. Moreover, the differences in pay are already visible. The Structure of Earnings Survey from the EU (2010) provides the median wages according to each type of contract (permanent, fixed or apprenticeship). We focus on the differences between permanent and fixed duration. We find a gap in hourly pay of 1,47 at the Union level, in favour of open-ended contracts, with differences among the country members. Within the EU, Poland presents the highest value of the ratio - 1,59 on average. The value is somehow larger for women and in some specific sectors, such as the business economy (1,66). More importantly, while in the EU the difference decreased (the ratio of wages was of 1,5 in 2006) in Poland it grew slightly between those dates.
A potential extension of the model is to include unemployment, in at least two ways. First, because the growth in temporary employment might have resulted from a decrease in unemployment, in which case the introduction of part time contracts reduced inequalities. Second, the model assumes that the outside costs are constant. This assumption probably fails to hold when outside options are limited, as it is the case when unemployment is large.