New Evidence on the Labor Market Segmentation Hypothesis
Bibliographic Data
| ID | 9256091 |
|---|---|
| Authors | Kwabena Gyimah-Brempong, Kwabena Gyimah‐Brempong, Paulette Olson |
| Year | 1994 |
| Volume | 52 |
| Issue | 1 |
| Pages | 20-39 |
| Publication date | 1994-03-01 |
| Peer Reviewed | Yes |
| Open Access | No |
| Type | ARTICLE |
| Venue | Review of Social Economy (JOURNAL) |
| Journal identifiers | ISSN: 0034-6764 • E-ISSN: 1470-1162 |
| Publisher | Informa UK Limited (PUBLISHER • GB) |
| DOI | 10.1080/758524599 |
| OpenAlex | W2036926821 |
| Language | EN |
| Citations received | 3 |
| References cited | 9 |
Introduction Since the publication of Glenn Cain's article (1976) on the challenge of segmented labor market theory, there have been numerous empirical tests of the effect of segmentation on earnings. A main concern of the segmentation literature has been the question of how to divide the work force among the various segments of the labor market (Taubman and Wachter, 1986). Earlier approaches used individual attributes or human capital characteristics to segment the labor market. However, this approach was shown to be inappropriate because it suffered from truncation bias (Cain, 1976). The estimates were biased because segments were first divided on the basis of high or low income and then used in equations to estimate the rate of return on education (Schervish, 1983). A number of other studies have used industry or product market characteristics to divide the labor market into various segments (Beck, Horan, and Tolbert 1978; Osberg, Apostle, and Clairmont, 1987). An alternative approach has been to use a switching model that estimates a series of earnings functions and a function that estimates the probability of being in the primary labor market (Dickens and Lang, 1985). Anderson, Butler and Sloan (1987) have criticized this approach and alternatively divide the labor market into segments using a hierarchical clustering model with a parametric stopping rule. Another approach has been to use various ranks of occupational prestige to divide the labor market (McNabb and Psacharopoulos, 1981; Neuman and Ziderman, 1986). Most recently, Boston (1990) found evidence to support the labor market segmentation hypothesis. In particular, he found that there are two distinct labor market segments and that a significant portion of the wage differential between segments was unexplained by differences in worker characteristics. However, his results divide the labor market into only two sectors whereas recent literature on labor market segmentation suggests that there may be as many as four distinct segments (Piore, 1975; Edwards, 1979; Gordon, Edwards, and Reich, 1982; and Gordon, 1986). Moreover, Boston derives his two segments using a one dimensional classification scheme and uses only 44 two digit occupations to derive his segments. Using only two segments may result in too much aggregation and hence the loss of important information. The single dimensional classification scheme which uses only a single criterion, whether workers need specific training or skills, also increases the probability of misclassifying workers than a scheme that is multidimensional and uses multiple criteria. For example, using only two segments and a one dimensional criterion results in classifying secretaries, construction workers, and engineers as part of the primary segment. Similarly, all operators, fabricators, and laborers are classified as being in the secondary labor market (Boston, 1990). In contrast, Gordon (1986) suggests using a multidimensional scheme with multiple criteria to establish four distinct labor market segments using industry and occupational classifications at the three digit level. Using Gordon's classification scheme places secretaries, construction workers, and engineers into three distinct segments. Gordon's scheme also places operators, fabricators, and laborers in the automobile industry -- a core industry -- into a subordinate primary sector whereas operators, fabricators, and laborers in the knitting mill industry -- a peripheral industry -- are in the secondary labor market. The purpose of this paper is twofold. First, the hypothesis put forward by the segmented labor market (SLM) model which divides the labor market into an independent professional and technical segment, an independent craft segment, a subordinate primary segment and a secondary segment will be tested. Second, the differentials between segments will be decomposed into a portion that is explained by differences in worker characteristics and a portion that is unexplained by differences in worker characteristics
Earnings · Econometrics · Economics · Labor market segmentation · Labor relations · Labour economics · Market segmentation · Microeconomics · Nonfarm payrolls · Secondary labor market · Split labor market theory · Gender, Labor, and Family Dynamics · Labor market dynamics and wage inequality · Labour Market and Migration
| Unique citing works | 3 |
|---|---|
| Citations per year | 0,15 |
| Citation span | 2006 - 2011 (6) |
| Citation velocity | historical |
| Highly cited | No |
| Citation types | Neutral: 3 |