Ebook Wages and the Size of Firms in Dynamic Matching Models
Wage differentials across observationally equivalent workers are both sizable and persistent in reality, and empirical work has given stylized fact status to several relationships between worker wages and employer characteristics. The employer size5wage effect is perhaps the strongest such stylized fact: everything else equal, larger firms or plants pay higher wages, as noted by Krueger and Summers (1988) and exhaustively documented by Brown and Medoff (1989) and Oi (1999), who argues that the relationship between wage and employer size was already observed some hundred years ago. An apparent relationship between wages and employer size might be spurious in raw data, reflecting for example the fact that larger employers hire higher quality workers, or pay higher wages to compensate for inferior working conditions (Masters, 1969), or face higher risk of unionization (Podgursky, 1986), share with workers their above normal profits (Weiss, 1966; Mellow, 1982), or need to offset a lower applicant to job vacancy ratio (Weiss and Landau, 1984) or reduce costly monitoring (Oi, 1983). The employers size, however, remains significant in fixed effect wage equations even when they include all variables relevant to such alternative explanations.
Empirical evidence also indicates that wage dispersion across production units is also related to employment levels. Davis and Halti sector wage dispersion falls sharply with establishment size for nonproduction workers, and mildly for production workers. Like the mean effects, this second moment wage5size effect is robust to the inclusion of control variables such as establishment level union density and indicators of worker quality.
This paper explores the wage structure implications of matching models of the labor market, where job creation and destruction take place at firms of endogenously heterogeneous size, and shows that the joint distribution of wages and employment generated by such models can be consistent with the evidence on both the first and second moment of the wage size distribution.
We are certainly not the first to note that imperfect matching or "search" can rationalize wage differentials across identical workers employed by different firms. In the Burdett and Mortensen (1998) model, on the job search draws workers from low pay to high pay jobs and, in a "wage posting" equilibrium where vacancies are costly and wages attached to a given job are never renegotiated, exante identical employers can earn similar profits through different combinations of wage and employment levels. Consistently with evidence of wage size effects, high wage jobs are more numerous than lower wage jobs in such an equilibrium. However, in the original Burdett Mortensen model and in recent extensions where optimal adjustment of match5level capital yields single peaked wage offer distributions (see Mortensen, 1998) jobs could be spread in the economy at large, rather than bundled in "firms". In fact, since no renegotiation occurs in a wage posting setting, jobs paying different wages could coexist within any legal entity in the absence of unmodeled additional constraints on within firm wage dispersion.
In general, the association of high wages to a specific firms circumstances is lost in a constant returns environment. Matching models with continuous renegotiation of wage rates also typically focus on job level phenomena, and are similarly illsuited to a study of firm level phenomena (see Mortensen and Pissarides, 1999, for a survey). Observable characteristics of real life firms, however, do appear empirically relevant. Dickens and Katz (1987), Katz and Summers (1989), and other references in Blanch flower et al. (1996) find that persistently high wages are associated not only to employment levels and rates of change, but also to larger profits and other indicators of employers "ability to pay". Further, Davis et al. (1996) show that job reallocation, the sum of gross job creation and destruction, monotonically falls with employer size.
Download
PDF Ebook Wages and the Size of Firms in Dynamic Matching Models
Posted in :
