Deutsches Institut für Wirtschaftsforschung. Discussion Papers. Anne Busch Elke Holst - PDF

Deutsches Institut für Wirtschaftsforschung Discussion Papers 1101 Anne Busch Elke Holst Gender-Specific Occupational Segregation, Glass Ceiling Effects, and Earnings in Managerial Positions:

Please download to get full document.

View again

of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.


Publish on:

Views: 12 | Pages: 20

Extension: PDF | Download: 0

Deutsches Institut für Wirtschaftsforschung Discussion Papers 1101 Anne Busch Elke Holst Gender-Specific Occupational Segregation, Glass Ceiling Effects, and Earnings in Managerial Positions: Results of a Fixed Effects Model Berlin, January 2011 Opinions expressed in this paper are those of the author(s) and do not necessarily reflect views of the institute. IMPRESSUM DIW Berlin, 2011 DIW Berlin German Institute for Economic Research Mohrenstr Berlin Tel. +49 (30) Fax +49 (30) ISSN print edition ISSN electronic edition Papers can be downloaded free of charge from the DIW Berlin website: Discussion Papers of DIW Berlin are indexed in RePEc and SSRN: Gender-Specific Occupational Segregation, Glass Ceiling Effects, and Earnings in Managerial Positions: Results of a Fixed Effects Model Anne Busch* and Elke Holst** Abstract The study analyses the gender pay gap in private-sector management positions in Germany based on data from the German Socio-Economic Panel Study (SOEP) for the years It focuses in particular on gender segregation in the labor market, that is, on the unequal distribution of women and men across different occupations and on the effects of this inequality on earnings levels and gender wage differentials in management positions. Our paper is, to our knowledge, the first in Germany to use time-constant unobserved heterogeneity and gender-specific promotion probabilities to estimate wages and wage differentials for persons in managerial positions. The results of the fixed effects model show that working in a more female job, as opposed to a more male job, affects only women s wages negatively. This result remains stable after controlling for human capital endowments and other effects. Mechanisms of the devaluation of jobs not primarily held by men also negatively affect pay in management positions (evaluative discrimination) and are even more severe for women (allocative discrimination). However, the effect is not linear; the wage penalties for women occur only in integrated (more equally male/female) jobs as opposed to typically male jobs, and not in typically female jobs. The devaluation of occupations that are not primarily held by men becomes even more evident when promotion probabilities are taken into account. An Oaxaca/Blinder decomposition of the wage differential between men and women in management positions shows that the full model explains 65 percent of the gender pay gap. In other words: Thirty-five percent remain unexplained; this portion reflects, for example, time-varying social and cultural conditions, such as discriminatory policies and practices in the labor market. JEL Classification: J31, J16, J24 Keywords: Gender pay gap, managerial positions, gender segregation, glass-ceiling effects, Oaxaca/Blinder decomposition, fixed effects, selection bias * DIW Berlin and Berlin Graduate School of Social Sciences (BGSS) / European PhD in Socio-Economic and Statistical Studies (SESS) ** DIW Berlin and University of Flensburg 1 Introduction Many national and international studies on the gender pay gap show a wage disadvantage for women (Bardasi/Gornick 2008; Kunze 2008; Cohen, Philip N./Huffman 2007; Blau/Kahn 2006; Fitzenberger/Kunze 2005; Blau/Kahn 2003, 2000; Waldfogel 1998; Jacobs/Steinberg 1995a; Kilbourne et al. 1994; Marini 1989). Germany had a (raw) wage pay gap of 23.2 percent in 2008, one of the highest in the European Union (European Commission 2010). However, few articles to date have examined the gender pay gap in management positions in Germany (see for other countries: Kirchmeyer 2002; Bertrand/Hallock 2001; Lausten 2001; see for academics in Germany: Leuze/Strauß 2009). The present study seeks to fill this lacuna by analyzing the gender pay gap in private-sector management positions in Germany based on data from the German Socio-Economic Panel Study (SOEP) (Wagner et al. 2007). We focus in particular on gender-specific labor market segregation the observation that women and men are distributed unequally across occupations and within occupational hierarchies and the effects of this segregation on earnings and gender wage differentials in management positions. In Germany, gender-specific labor market segregation has remained very stable over time (Trappe/Rosenfeld 2004): most women still work in typical women s jobs and most men in typical men s jobs. Segregation is also found in managerial positions in Germany. On the one hand we see vertical segregation: women tend to work at lower hierarchical levels than men even within management the upper echelons of which are mainly occupied by men. The higher the hierarchy is the lower the share of women in it (Holst/Wiemer 2010a, b, Holst/Busch 2010). On the other hand, we see horizontal segregation: the majority of men and only a minority of women in management positions work in typically male jobs. While women in management are less segregated than other female employees, the opposite is true for men (Holst/Busch 2010). Further, managerial positions show gender-specific occupational differences in the size of the enterprise, the sector of the economy, and the industry (Bischoff 2010; Kleinert et al. 2007): women more often head smaller firms, and they more frequently work in health care, welfare, and in the private services. Women with a university degree more often choose a field of study that is dominated by women, such as humanities (Leuze/Strauß 2009). In addition, female managers are more often employed in public services than in the private sector (Brader/Lewerenz 2006). Women s occupations are generally characterized by worse employment conditions in terms of wages; many studies analyze the effects of working in a gender-typical or -atypical 1 occupation on wages and wage differentials between women and men (Busch/Holst 2010, 2009; Cohen, Philip N./Huffman 2007; Hinz/Gartner 2005; Cohen, Philip N/Huffman 2003; Jacobs/Steinberg 1995b; England 1992; England et al. 1988). The question that is not fully answered is why this wage penalty even in management still exists where the best of the best are being employed. To control for unobserved time-constant heterogeneity, we use fixed effects panel models to estimate wages and wage differentials of women and men in management positions. As promotion probabilities are highly gender-biased in Germany (Holst/Busch 2010; Fietze et al. 2009) we take gender-specific promotion probabilities into account by employing a special version of Heckman selection (Heckman 1979). We use these strategies to obtain unbiased estimators of the coefficients for the wage effects of working in a gender-segregated occupation. To our knowledge, this is the first time that the gender pay gap in managerial positions has been analyzed in the German context by taking fixed effects and selection bias into account simultaneously (see for the US: England et al. 1988). Finally, we decompose the gender pay differential to explain the extent to which the gender pay differential is related to gender-specific segregation on the labor market and to other components of our wage equation (Blinder 1973; Oaxaca 1973). The study is structured as follows. In Section 2, we introduce the key theories explaining the dependency between gender-specific labor market segregation and wages, discuss the current state of research, and formulate our working hypotheses. In section 3, we present the multivariate method for quantitative analysis of the gender-specific wage differential in management positions. In Section 4, we explain our database and variables, and in Section 5 we present the empirical findings. Finally, in Section 6, we summarize the results and draw conclusions. 2 Theoretical background Human Capital Approach From an economic point of view, the effect of working in a segregated (male or female) job on wages can be explained through different investments in human capital. The different human capital investments of men and women are interpreted as being the result of a rational cost-utility calculation (Becker 1993, 1991): an assumption is that women have stronger preferences for family work than men and that these affect their choices of lower-paid 2 occupations and less successful career paths. Hence, for women, investments in education, work, and on-the-job-training appear less profitable since the accumulated knowledge becomes obsolete during breaks in employment (Blau et al. 2006; Tam 1997; Mincer 1962). As a result, women invest less in education. Human capital theory uses the concept of selfselection to explain the different proportions of women and men in certain occupations and thus the emergence of gender-specific labor market segregation (Polachek 1981). According to this idea, women rationally choose particular jobs that can be combined with family responsibilities for example, jobs that have lower opportunity costs when working part-time or when employment is interrupted. These are mainly lower-paid jobs. The higher the human capital endowment, and thus the higher the opportunity costs, the lower this effect will be. Becker (1985) assumes that even with the same human capital endowment, it is rational for an employer to pay married women less than men in the same job: Since housework is more effort intensive than leisure and other household activities, married women spend less energy on each hour of market work than married men working the same number of hours. As a result, married women have lower hourly earnings than married men with the same human capital, and they economize on the energy expended on market work by seeking less demanding jobs (Becker 1985: 55). Based on these assumptions, we can formulate the following hypotheses on the dependency between segregation and wages in managerial positions: H1: Female occupations pay less than male occupations, even at the management level; this holds for both women and men working in these occupations. However, the implicit given in the human capital approach of gender-specific preferences was criticized early on in a number of studies (e.g. England 1989; for an overview, see Ferber 1987). In the early 1980s, it was also shown that women who planned to interrupt their careers did not, contrary to the hypothesis of self-selection, choose female jobs more frequently than other women (England 1982). In addition, an analysis for West Germany showed that career breaks do indeed have a negative effect on both men s and women s wages, and that this negative effect is particularly strong if the interruption occurred due to family responsibilities (e.g., parental leave) (Beblo/Wolf 2002). Empirical studies show as well that women are trapped from the very start of their career, in the sense of experiencing a lock-in-effect in occupations with lower pay (Fitzenberger/Kunze 2005). 3 Devaluation Approach At this point, we turn to the devaluation approaches to explain gender-specific wage differences. Here, it is assumed that the historically dominant male breadwinner model, in which women are responsible for the unpaid housework and men for the paid work, lead to corresponding gender-specific values and norms internalized by the individuals and thus to gender-specific orientations and needs ( preferences ) for special jobs, as well as to discriminatory practices on the labor market (Gottschall 2000; Beck 1986; Beck-Gernsheim 1980). The internalization of gender roles in values and norms is (re)produced by a doing gender in everyday interaction processes (Ridgeway/Smith-Lovin 1999; Ridgeway 1997; West/Zimmerman 1987): in order to reduce the amount and complexity of information in daily face-to-face interactions, people make gender-specific assumptions about the person with whom they are interacting. These assumptions form the basis for gender stereotypes that are shaped by cultural perceptions about what constitutes male and female. Nelson (1996) argues that individuals tend to think dualistically and to ascribe abilities hierarchically according to gender norms. The characterization rational is usually ascribed to men and valued more in the labor market than emotional ( emotional work don't 'count' , England 1989: 24) which is usually ascribed as a female characteristic developed by providing unpaid family work at home. Accordingly, expectations about potential performance differ by gender and this may result in wage penalties for women. As far as pay is concerned the devaluation hypothesis postulates a general devaluation of female work (Liebeskind 2004; England 1992; Steinberg 1990; England et al. 1988). This devaluation leads to lower pay for female jobs independent of human capital; the higher the percentage of women in a specific job the lower the pay for women as well as for men. This is referred to in the literature as evaluative discrimination (Achatz et al. 2005; Peterson/Saporta 2004). In addition, studies have shown that even within a specific job (i.e., within a female-dominated, male-dominated, or gender-integrated profession), the work of women is devaluated and paid less than that of men. This is labeled allocative discrimination (Achatz et al. 2005; Peterson/Saporta 2004). 1 In line with so-called intergroup conflicts (Blalock 1967), this allocative discrimination may also be due to an increase in perceived threats and perceived competition over scarce resources as more individuals of the 1 The term evaluative indicates that one job is valued less than another solely because it is numerically dominated by one group of persons (men/women), independently of the real tasks and demands of the job (Achatz et al. 2005: 469). Allocative discrimination involves wage disadvantages that result from hiring, promotion, and dismissal or firing, which are difficult to document (Peterson/Saporta 2004: 859). 4 opposite gender enter the workplace. Women as the lower status group on the labor market (Correll/Ridgeway 2006) are disadvantaged in such conflicts because they have less power. As a result, women in men s jobs may be seen by men as a threat, which may have negative consequences especially for their wages. The devaluation hypothesis has been controversial in the literature. It is generally acknowledged that on average wages in typical women s jobs are lower than in typical men s jobs (Olsen/Walby 2004; Jacobs/Steinberg 1995b). But there is no consensus on the reasons for these findings (England et al. 2000; Tam 2000, 1997; Kilbourne et al. 1994; England et al. 1988). However, neither of the aforementioned studies makes a clear distinction between evaluative and allocative discrimination. An explicit analytical distinction between these dimensions of discrimination was made in a German study by Achatz et al. (2005) identifying evaluative as well as allocative discrimination. Wages decreased with an increasing percentage of women in a job cell, 2 and this wage disadvantage was higher for women than for men. Busch/Holst (2010) found a similar result for the increasing percentage of women in an occupation focusing on women and men in management positions in Germany 2006 (Busch/Holst 2010). Based on these considerations, the following hypotheses can be formulated: H2: After accounting for human capital effects, there is still a negative wage effect of working in a women s job, which is due to devaluation (evaluative discrimination). H3: The negative wage effect of working in a women s job is stronger for women than for men. This is due to allocative discrimination. The devaluation mechanisms are intensified by stereotyped gender status beliefs, ideas that one gender is more competent and thus higher in status. The result of such beliefs is that, in general, men are seen as justified in holding higher positions of power and privilege (Ridgeway/Smith-Lovin 1999; Ridgeway 1997). This phenomenon is even more prevalent in top positions. An invisible barrier - known as the glass ceiling is preventing women from climbing the career ladder beyond a certain level (International Labour Office 2004; Wirth 2001). Men are expected to possess higher work-related skills and abilities and to show higher 2 Others did not use the percentage of women in the jobs, but the percentage of women in job cells. This was calculated as the percentage of (full-time employed) women in a job per firm (Achatz et al. 2005: 474). This was possible because they used firm-level data instead of individual micro-data. 5 performance and productivity than women (see also Correll/Ridgeway 2006; Ridgeway 2001; Foschi 1996). This results in different career and pay opportunities for men and women since wages reflect the expected productivity of the employee. Employers tend to believe that women fit the management profile less especially the leadership profile; and as a result of these and contrasting assumptions about male leadership qualities employers attribute a higher competence in this area to men (Gmür 2006, 2004; Eagly/Karau 2002; Ridgeway 1997). In addition, according to the homophily principle which states that people interact primarily with others who are similar in given characteristics and build gender-homogeneous networks (McPherson et al. 2001; Ibarra 1997, 1992; McPherson/Smith-Lovin 1987) when making decisions about promotion individuals prefer others who are similar in given characteristics (like gender). Consequently, the predominantly male decision-makers prefer to promote men to management positions (Ridgeway 1997). If, despite the barriers women obtain a managerial position they are highly visible tokens (Kanter 1977) and thus subjected to a more rigorous evaluation of their performance and possible mistakes than men. This increases their probability of being marginalized and demoted from their position. Altogether, women who succeed in climbing the career ladder are probably a highly selected group of women. This might bias their wages upwards and also might bias the coefficients of the independent variables in the wage equations. In gender-segregated occupations in particular this bias might has substantial effects not only on wages but also on the chances of promotion. There is some evidence of a strong negative effect on promotion probabilities for women working in a female occupation (Maume 1999). Again, this can be explained with mechanisms of devaluation of women s jobs: employers provide fewer training opportunities in such jobs and female occupations therefore offer more limited chances of entering management. In terms of allocative discrimination the effect should be stronger for women than for men due to gender status beliefs and different competence expectations. It is therefore assumed that: H4: The wage effect of working in a gender-segregated occupation is even stronger when the gender-specific promotion probability is taken into account. 6 3 Models and Estimation Methods We first estimate a wage equation according to Mincer (Mincer 1974) with additional human capital variables, variables related to gender-specific labor market segregation and variables connected to social structure/family circumstances (see Section 4). We are employing a multiple linear panel regression with fixed effects separately for women and men that, i.e. we only consider within-person changes over time (Allison 2009). One main advantage of this method is that it controls for time-constant unobserved heterogeneity, Therefore, variables that vary between but not within persons (like ability or personality traits) are excluded from the model. However, there might also be another kind of selection bias that is not controlled for in the fixed effects model per se: the selection into managerial positions. Women who receive a management position might be highly selected. This may result in an overestimation of their wages and therefore in biased coefficients. To correct for such a selection bias, we use a special version of Heckman s correction (Heckman 1979) which is applicable for fixed effects regression (E
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks