Economics of Gender Research Paper

The research paper begins with an overview of country differences in gender wage and employment gaps. This section discusses how gender differences in employment rates and differences in wage dispersion may relate to the gender wage gap. This is followed by a theoretical overview of gender differences in the labor market covering both supply-side differences in human capital acquisition—the role of gender-specific preferences—and demand-side discrimination. The next section discusses the empirical problem of identifying discrimination in the labor market. The research paper rounds off with two sections covering, in turn, the empirical evidence on discrimination and gender differences in preferences.

Gender Differences in the Labor Market: Overview

Studying differences in labor earnings is of fundamental importance for anyone interested in understanding poverty, social stratification, and the economic incentives facing workers. Labor earnings are the most important component of an employed individual’s income and a major determinant of living standards. This research paper therefore begins with a comparison of gender labor market gaps in a selected number of industrialized countries. First, an overall picture of gender wage gaps and differences between countries is provided, followed by a description of the trends in these gaps during the last 20 years. Country differences in gender wage gaps are related to differences in female employment rates as well as varying wage distributions across countries. Possible explanations for the wage gaps observed, such as productivity differences, occupational gender segregation, preferences, and discrimination are discussed in subsequent sections.

Table 1 reports the gender wage gap, the female employment rate, and the fraction of female employees who work part-time for a selected number of countries during the period from 1989 through 2006. The gender wage gap is defined as 1 minus the ratio of the annual averages of female and male mean hourly wage rates, which can then be interpreted as how much less, as a percentage, women earn per hour relative to men. For example, a gender wage gap of 23% in the United States in 2006 means that the female wage rate is 23% smaller than the male wage rate. In other words, for every $100 that men earn per hour, women earn 23% less—that is, $77. As shown in Table 1, gender differences in pay prevail in all countries, even though the size of the gaps varies considerably across countries. In 2006, the female hourly wage rate in France was 11% smaller than the male hourly wage rate. Thus, wages of employed women in France are closer to men’s wages than in the case of the United States. When one looks at the period from 1994 through 1998, where information on the gender wage gap for the full set of countries is available, one sees that the gender wage gap is 36% in Japan, 10% in France, and 24% in the United States. Table 1 also shows that there seems to be a tendency toward decreasing gender wage gaps over time in most countries since the late 1980s.

An interesting question to ask is why gender differences in labor market outcomes vary across countries. For example, why do women earn less relative to men in a country such as the United States compared with France or Sweden? One reason could be that women have acquired different skill levels across countries and that women in France and Sweden are more qualified than women in United States. Because skills are rewarded through higher pay in the labor market, this is one potential reason for observed country differences in pay. According to Francine Blau and Lawrence Kahn (2000), however, there seems to be little reason to believe that women in the United States are less qualified relative to men than women in other countries. An alternative explanation is that there are country differences in economic returns to skills and therefore differences in economic incentives. Countries with high rewards to skills have wage structures that encourage skill acquisition among workers. This suggests that the wage structure in a country plays an important role in determining the gender wage gap, given that there are gender differences in skills and qualifications. Consider, for example, two countries where women have lower levels of labor market experience than men, but the gender difference in experience is the same in the two countries. If the return to experience is higher in one country, this country will have a larger gender wage gap, all else equal.

Moreover, centralized wage-setting institutions tend to reduce wage dispersion across firms and industries and raise the relative pay of low-wage workers (regardless of gender), which in turn may reduce the gender wage gap. Because most European countries have more centralized wage setting compared with the United States, the degree of centralization of wage setting may be an important explanatory factor behind country differences in the gender wage gap. Empirical evidence suggests that the overall wage structure is of major importance in explaining gender wage gaps, where the higher level of wage inequality in the United States compared with other countries tends to increase U.S. gender differentials relative to those in other countries (see, e.g., Blau & Kahn, 1996a, 1996b).

These explanations of the gender wage gap are based on the assumption that women on average are less qualified than men. It is therefore interesting to see how the gender wage gap has evolved over time as female labor force participation has increased in most countries since the 1970s. Human capital skills, such as level of education and labor market experience, have also increased for females relative to males during this period. Because a stronger attachment to the labor force would increase other labor market skills as well, this suggests that gender wage gaps should shrink over time. According to Table 1, the female employment rate has increased slightly in most countries, together with a decrease in the gender wage gap. Thus, it appears that women to some extent have caught up with men by accruing more skills. It should be stressed, however, that even if women to a large extent have caught up with men in terms of level of education, they systematically chose different types of educations. For example, men to a larger extent than women chose technical educations. If technical educations and the occupations associated with these educations yield higher returns in the labor market than other types of educations, this would improve men’s labor earnings relative to those of women.

Furthermore, because women have lower employment rates than men, women might be a more selected group in the labor market than men with, on average, higher tastes and skills for work compared to the population of women. Such selection into employment makes it difficult to study trends in gender wage gaps because the group of female workers from one time period to another might not be comparable when female employment rates change over time. This also makes across-country comparisons of gender wage gaps difficult because female employment rates are very different across countries, implying that the selection of females who participate in the labor market in one country is potentially very different from the corresponding selection in another country.

Difficulties in comparing trends in gender wage gaps across countries are supported by the striking international variation in female employment rates. It is apparent in Table 1, for example, that France had the lowest gender wage gap in 2006 (11%) but also one of the lowest female employment rates. Only 59% of women were employed in France at that time, compared with 66% of women in the United States. One hypothesis, therefore, is that selection into employment is not random and might affect the size of gender wage gaps. In particular, if employed women have relatively high-wage characteristics, low female employment rates are consistent with lower gender wage gaps because women with low-wage characteristics are not included in the observed wage distribution. This may explain the negative correlation between gender wage and employment gaps that are observed in Table 1.

Table 1   Gender Wage Gaps, Female Employment Rates, and Female Part-Time Work Rates Among the Employed (Percentages)

The pattern of countries with high gender wage gaps tending to have high female employment rates might also be reinforced by differences in the extent to which women work part-time. In France, the fraction of employed females working part-time in 2006 was 30%. The corresponding figure for Sweden, which had a larger gender wage gap as well as a higher female employment rate in 2006, was 40%. Working part-time implies that a smaller amount of work experience is accumulated over time in comparison to full-time workers. It might also be the case that working part-time is associated with lower chances for promotion and wage raises. Taken together, there are, therefore, a number of explanations as to why a higher female employment rate, together with high female parttime rates, might be associated with a higher gender wage gap within a country.

Different patterns of employment selection across countries may in turn stem from a number of factors. First, there may be country differences in the gender role of household work, social norms, or both, affecting labor force participation. Second, labor demand mechanisms, including social attitudes toward female employment and those attitudes’ potential effects on employer choices, may be at work, affecting both the employment rate as well as the level of wage offers by gender. Claudia Olivetti and Barbara Petrongolo (2008) suggest that the international variation in gender employment gaps can indeed shed light on across-country differences in gender wage gaps. This study suggests that sample selection into employment explains nearly one half of the observed negative correlation between gender wage and employment gaps. They also show that while the raw wage gap is much higher in Anglo-Saxon countries than in southern Europe, the reason is probably not to be found in more equal pay treatment for women in the latter group of countries but rather in different selection processes into employment. Female participation rates in southern European countries are low and concentrated among high-wage women. Correcting for lower participation rates in southern European countries widens the wage gap to levels similar to those of other European countries and the United States.

Theoretical Explanations for Gender Differences in the Labor Market

Gender differences in labor market outcomes stem from supply-side differences in productivity, labor supply, or preferences or to demand-side differences in opportunity— that is, gender discrimination in employment, wagesetting, or promotion of equally qualified individuals (see Altonji & Blank, 2003, for an overview). Supply-side differences in productivity and labor supply between men and women are often analyzed within the human capital framework (Mincer & Polachek, 1974). The human capital model postulates that individuals invest in education and training and are rewarded for these investments in the labor market, either via enhanced employability, higher wages, or both (Mincer, 1958; Mincer & Polachek, 1974). Within this framework, gender differences in labor market outcomes are attributable to gender differences in human capital investment.

A number of explanations have been forwarded as to why women historically have invested less in education, skills, and other qualifications valued in the labor market. A partial explanation can be found in traditional gender norms concerning child care and housework within families. If women expect to spend more time out of the labor market, they consequently have less time in the labor market to reap the benefits of their human capital investments. As such, women will invest less in market-oriented human capital and, in addition, will direct their investments toward educations and skills with lower depreciation rates due to time out of the labor market (Polachek, 1975). Notice that gender differences in the types of academic training being invested in, due to differential depreciation, also help to explain patterns of occupational segregation by gender. Shorter duration in the labor market as well as more frequent interruptions for child rearing also influence employers’ willingness to finance on-the-job training for female employees as well as promotion possibilities, reinforcing, over time, gender differences in productivity. Note that there is a literature that questions the degree to which human capital models can explain occupational segregation by gender (Beller, 1982; England, 1982).

Not only are there social norms concerning women’s division of labor between home and market production but there are also norms concerning what constitutes typically female or male pursuits in the labor market (Akerlof & Kranton, 2000). In other words, preferences for certain types of occupations or educations may be due to social norms or preconditioning regarding what is considered appropriate for women and the costs of deviating from these norms (Akerlof & Kranton, 2000; Gundersson, 1989; Polachek, 1984). Differences in type of academic training can also be the result of historical differences in remuneration and access to jobs by gender. If women historically have had lower access to (due to norms or discrimination) or lower payoffs from certain occupations, then investments in education and training for these occupations will also be lower. This implies that historical discrimination in the labor market can lead to subsequent differences in human capital investment and that gender differences in the labor market can become a self-fulfilling prophecy (Darity & Mason, 1998). Claudia Goldin (2006) argues that lower remuneration in typically female occupations has an historical basis in the segregation of jobs as women increasingly entered the labor market in the early 1900s. As white-collar positions opened up for women, policies were instituted at the firm level, creating sex-segregated positions. Jobs were increasingly classified as either female or male, where the majority of female jobs were dead-end, providing little room for advancement to higher positions or earnings growth.

Over time, shifts in the norms concerning female participation in the labor market, greater and more continuous time in the labor market, as well as more equitable distribution of housework and child care will increase incentives for women to invest in human capital accumulation and for firms to invest in female employees. Indeed, the salience of gender differences in human capital investment as an explanation for gender differences in labor market gaps has decreased over time as women increasingly have closed the gender gap in education, at least with regard to level of education. However, hard-to-break differences in remuneration attributable to occupational segregation remain due to either cultural devaluation of female jobs or dual labor market and crowding theories, where exclusion from male jobs leads to crowding in female jobs and consequently lower wages as well as lower returns to education (Bergmann, 1974; Doeringer & Piore, 1971). Occupational segregation by gender is therefore both a cause and a consequence of gender differences in pay. Occupational segregation can lead to gender differences in human capital investment and productivity, but it is also a partial explanation for gender wage and income differentials in a society characterized by occupational gender segregation and lower remuneration for female jobs.

On the demand side, two forms of discrimination are commonly discussed within the economics framework: taste-based discrimination and statistical discrimination. Taste-based discrimination arises because of a disutility among employers (customers or coworkers) for interacting with female workers or because of a preference for male workers (Becker, 1971). If tastes for discrimination are large and the demand for preferred male workers is lower than the supply, a wage differential arises between male and female workers, the implication of which is a competitive advantage for firms that hire equally productive women. This suggests that wage gaps will disappear in the long run as nondiscriminating employers enter the market. Frictions to free entry, imperfect information, collective bargaining, search costs, and other infringements to perfect market competition may, however, lead to sustainable wage gaps due to taste-based discrimination over time. There are a number of empirical studies examining the degree to which competition decreases gender discrimination; see, for example, Orley Ashenfelter and Timothy Hannan (1986); Sandra Black and Elizabeth Brainard (2004); Black and Philip Strahan (2001); Judith Hellerstein, David Neumark, and Kenneth Troske (2002); Xin Meng (2004); and the references therein.

Statistical discrimination arises because of the inability to acquire, or costs of acquiring, perfect information about job candidates. Instead, employers use readily available group statistics to assess candidates (Arrow, 1973; Phelps, 1972). This implies that if women on average have lower relevant labor market experience for a given position, are expected to leave the labor market for child rearing, or both, to a larger extent than male candidates, employers will be less inclined to interview and hire female candidates. Notice that statistical discrimination is individual discrimination based on actual group statistics. The individual in question may deviate from mean group characteristics, but employers are unable to access this information, or the costs of doing so are high. Statistical discrimination is also based on the assumption that unbiased group statistics are readily available. This may not be the case, and agents may act on the presumption that their beliefs are statistically correct but actually base decisions on biased information because of erroneous perceptions of group productivity. In the sociological literature, this is termed error discrimination (England, 1992).

The Empirical Problem of Identifying Gender Discrimination in the Labor Market

The most common method of estimating gender differences in labor market outcomes is through wage, earnings, or employment regressions using register or survey data. Typically, variations of the following basic model are estimated:

y = a + B1 female i + XiB2 + Ei

where y is a labor market outcome (employment, wages, income) of individual i, female is a binary (dummy) variable equal to 1 if female and 0 otherwise, and X is a vector of other control variables, typically both demographic and human capital indicators that may influence the outcome variable and systematically differ between men and women. Finally, E is the random error component measuring the impact of unobserved characteristics in the equation that also influence the outcome variable. In estimation of this type of equation, the coefficient for female indicates to what degree labor market outcomes differ between men and women with similar characteristics (the observable characteristics included in the vector X). It does not, however, tell us whether these gender differences are due to discrimination— that is, to unequal treatment of equally qualified individuals. The reason is that it is impossible for researchers to control for all possible supply-side differences in productivity by gender. For a causal interpretation of the effect of gender on a given outcome, the correlation between the female dummy variable and the random error component must be equal to 0. This is called the zero conditional mean assumption, stating that there should be no systematic differences in unobserved factors between men and women that influence the outcome of interest. If there is such a factor—for example, if women have systematically lower on-the-job training than men and this is unobserved in an income equation—then the measure for gender differences in income will be biased if differences in labor market training are not accounted for in estimation. This implies that observed gender differences in labor market outcomes may be due to supply-side differences in productivity by gender, uncontrolled for in estimation (unobserved characteristics) or discrimination.

Note that there is a large literature concerning the decomposition of gender wage gaps into an explained and unexplained component, the so-called Oaxaca-Blinder method, where the unexplained component measures wage differentials that remain after controlling for all observable characteristics in estimation (Blinder, 1973; Oaxaca & Ransom, 1999). (For extensions of this method, see also Juhn, Murphy, & Pierce, 1991.)

Recently, attention has turned to estimating gender differences in the labor market using experimental methods. The reason is that a well-executed controlled experiment can unequivocally identify a causal effect of gender on an outcome of interest. Experiments are based on random assignment of a given population into a treatment group and a control group. Random assignment guarantees that both observable and unobservable individual characteristics are, on average, similar in treatment and control groups. This implies that any differences in outcomes due to treatment are attributable to gender alone. Studies based on experimental methods are often characterized by high internal validity but low external validity. High internal validity implies that experiments are credibly able to identify a causal effect of gender on outcomes of interest. Low external validity implies that results often cannot be generalized to the population at large. This is due to the fact that experiments are often performed on limited subsamples of the population only. Experimental methods have been used to study the presence of gender differences in preferences as well as employer discrimination. The next two sections provide an overview of the recent empirical evidence within these two strands of research on gender and the labor market.

Evidence of Gender Discrimination

Two commonly used methods to study gender discrimination in the labor market are so-called audit and correspondence testing studies (Darity & Mason, 1998; Pager, 2007; Riach & Rich, 2002). Audit studies use actual test persons to apply for jobs and participate in job interviews, while correspondence testing studies send written applications (CVs) to actual job openings and measure differences in callbacks to interviews. Audit studies have the advantage of testing discrimination in the entire application process, from initial contact with employers to actual job offers. On the other hand, audit studies are unable to credibly control that test persons are similar in all characteristics of relevance to employers (Heckman, 1998; Heckman & Siegelman, 1993). Although test persons are trained to behave and dress in a similar manner, their use introduces uncertainty into the experiment. Female test persons may, for example, internalize expected discrimination and subconsciously behave differently than male test persons. These issues are avoided in correspondence testing studies, where interpersonal contact between test subjects and employers is avoided via the use of written applications only. Many studies also suggest that the majority of employer discrimination occurs at this stage—that is, in callbacks to interviews from formal applications (Riach & Rich, 2002). As such, the correspondence testing methodology is equivalent to a randomized controlled experiment where gender is signaled by, for example, the names assigned to CVs.

The first field experiment to study gender discrimination in hiring was carried out by Peter Riach and Judith Rich (1987) in Victoria, Australia, using the correspondence testing methodology. This study finds that women encountered discrimination 40% more often than men. In a more recent work, Neumark (1996) studies the prevalence of gender discrimination in restaurant hiring in Philadelphia (United States) using both audit and correspondence testing methods. Female and male pairs were matched and trained to apply for waitress and waiter jobs in low-, medium-, and high-price restaurants after providing written CVs to restaurant managers. This experiment design therefore metes out possible differences in test person behavior by testing employer responses to both written applications and job interviews. Results showed that high-priced restaurants were significantly more likely to inter-view men than women (gender difference in callbacks to interviews) and significantly more likely to offer men jobs after interviews. Gender differences in hiring were not significant in low- and medium-priced restaurants. As wages and tip earnings are significantly higher in high-priced restaurants relative to low- and medium-priced restaurants, hiring discrimination in high-priced restaurants provides a partial explanation for within-occupation gender wage differentials in the restaurant sector.

Other correspondence tests of gender discrimination in hiring attempt to explicitly link hiring discrimination to statistical discrimination, either via experiments set up to test the effect of employer expectations concerning forthcoming childbirth among young female applicants (Duguet & Petit, 2006) or by signaling unobserved stereotypical personality traits associated with gender (Weichselbaumer, 2004). Emmanuel Duguet and Pascale Petit (2006) find that for qualified job positions, younger childless women have lower callbacks to interviews than men. This difference vanishes for older women, regardless of whether they have children. Doris Weichselbaumer (2004) tests whether women experience less hiring discrimination in traditionally male jobs if they signal stereotypical male personality traits (ambition, assertiveness, competitiveness, dominant individualism), which are normally not observed on written CVs but are nonetheless assumed to be important for the hiring decision. Three applications were sent to each job opening: two female and one male, where the male applicant and one of the female applicants signaled typically male personality traits while the other female applicant signaled traditionally female personality traits. Results showed that women were treated equally, regardless of variation in personality traits. Unfavorable treatment for women applying to masculine occupations was not mitigated for women signaling masculine traits, and preferential treatment for women in feminine occupations was not threatened by masculine traits among female job candidates.

More recent field experiments attempt to link occupational segregation and hiring discrimination. Riach and Rich (2006) use the correspondence testing methodology to test for sexual discrimination in hiring in England, explicitly testing occupations that can be categorized as typically female, male, or mixed. Men were found to be discriminated against in the female occupation (secretary), and women in the male occupation (engineer). However, and somewhat unprecedented in the research literature to date, significant discrimination against men was found within mixed occupations (trainee chartered accountant and computer analyst programmer). Mangus Carlsson and Dan-Olof Rooth (2008) test 13 occupations in Sweden with varying gender composition and find no hiring discrimination against women in male-dominated occupations and a slight preference for women in female-dominated occupations as well as mixed occupations. This study therefore suggests that present hiring discrimination is not likely to explain substantial occupational segregation by gender.

Another innovative field study uses the introduction of blind auditions to analyze potential hiring discrimination against female musicians within symphony orchestras (Goldin & Rouse, 2000). In the 1970s and 1980s, many major U.S. orchestras changed their audition policies and adopted so-called blind auditions—that is, auditions behind a screen preventing the identification of applicants’ gender. Interestingly, many auditions took the added precaution of muffling the sound of footsteps, which may otherwise have betrayed the gender of candidates, either by rolling out a carpet to the stage or by asking candidates to remove their shoes. Results from this study indicate that the adoption of a screen increased by 50% the probability that a female musician advanced beyond preliminary rounds and by several-fold the likelihood that a female musician would win in the final round.

A recent study on discrimination in hiring attributable to the gender composition of evaluation committees ties together the two strands of research discussed in this research paper—namely, gender differences in preferences and discrimination. Manuel Bagues and Berta Esteve-Volart (2008) use unique evidence provided by the public examination of candidates applying for positions within the Spanish Judiciary. Candidates are randomly assigned to evaluation committees with varying gender composition. Results from this study indicate that female candidates are less likely to be hired and male candidates more likely to be hired when randomly allocated to a committee with a larger share of female evaluators. A careful analysis of the data suggests that the quality of male candidates is overestimated in committees with female majorities. If women systematically overestimate the quality of male candidates, this could partially explain female reluctance to compete, as well as relatively poor performance in competitive environments, as noted in several studies that are discussed in the next section on gender differences in preferences.

The studies discussed in this section have spawned a large research interest in testing the presence of gender discrimination in different countries, labor markets, and populations. The consensus in this literature is that gender continues to be a salient factor in hiring within many labor markets and that more research is necessary to understand the extent to which discrimination explains labor market gaps today as well as the extent to which discrimination influences the preferences and human capital acquisition of coming generations.

Evidence on Gender Differences in Preferences

There is a vast literature aimed at explaining gender differences in preferences. Gender differences in preferences are interesting in and of themselves but also have potential importance in explaining gender labor market gaps. Men and women might have different preferences with respect to risk taking and competition, as well as different reactions to competition. For a survey of the research literature on gender differences in preferences, see Rachel Croson and Uri Gneezy (2009). Different degrees of risk taking among individuals can translate into different choices concerning jobs and occupations, according to the individuals’ risk exposure to unemployment, work injuries, and so forth. If workers are compensated for such risks through higher earnings, one explanation for observed gender wage gaps would be gender differences in risk taking. If women are less likely to compete, it not only reduces the number of women who enter competitive environments—for example, in terms of competition for jobs or wage bargaining—but it also decreases the chances for women of succeeding in these competitions.

The general finding from studies on gender differences in risk-taking behavior, most of which are based on laboratory experiments, is that men are more prone to risk taking than women. There are several explanations as to why men are more risk-prone than women. A number of studies, for example, find men to be more overconfident than women. Muriel Niederle and Lise Vesterlund (2007) test this using a controlled laboratory experiment where pairs of two women and two men perform a task, namely adding up sets of five 2-digit numbers for 5 minutes. Participants were first asked to perform the task with piece-rate compensation and thereafter in a tournament setting. After having experienced both compensation schemes, participants were asked to choose one of the two payment schemes in the final round (either piece rate or tournament). Although there were no gender differences in performance in either compensation scheme, Niederle and Vesterlund find that 73% of men prefer the tournament scheme compared with only 35% of women. To elicit participants’ beliefs on their relative tournament performance, at the end of the experiment, participants were asked to guess how their performances ranked relative to those of other participants. While 75% of men thought they were best in their group, only 43% of women held this belief. Participants who were more confident about their relative tournament performance were also more likely to enter a tournament.

Although numerous studies such as this have found men to be more overconfident than women on average, this difference may not hold for all tasks or for selected participants. One example of this is provided by Lena Nekby, Peter Skogman Thoursie, and Lars Vahtrik (2007), who use a large running race to study the behavior of women who choose to compete in a male-dominated setting and the consequences of this behavior on performance. Participants were given the opportunity to self-select into start groups based on individual assessments of running times. Overconfident behavior was measured as self-selection into start groups with lower time intervals than final results in the same race (or in the previous year’s race) would motivate. Only runners who participated in the same race on the same course during the previous year were sampled so that results would not be contaminated by potentially lower knowledge of individual capabilities among women. Results show that there are environments (male dominated) in which the selection of women who participate are more likely to be confident and competitive and that, within this group, performance improves equally for both genders.

This result is important because gender differences in labor outcomes may be underestimated in selective environments, such as among executives. Earlier studies on the gender wage gap, for example, have found a glass ceiling for women in the upper part of the income-wage distribution (see, e.g., Albrecht, Bjorklund, & Vroman, 2003). One interpretation of a glass ceiling is that women have greater difficulties than men in obtaining higher positions for observationally equivalent qualifications due to unobservable differences in competitiveness. If there are women in male settings who are as competitive as men, women who compete for higher positions may be evaluated on average female behavior and statistically discriminated against and prevented from reaching higher positions. Another example of this is found among professional investors. Although men are more risk taking than women on average, there are no differences in risk propensities among professional investors, as indicated by their investment behaviors (see, e.g., Atkinson, Baird, & Frye, 2003). Thus, while fewer women are selected into positions such as investors and executives, those who do choose to enter these professions have similar risk preferences to the men in these positions.

In terms of gender differences in competitive behavior, men seem to choose competitive environments to a larger extent than women. In contrast to Niederle and Vesterlund (2007), there is also some evidence that men perform better than women in competitive environments. For example, in a study by Gneezy, Niederle, and Aldo Rustichini (2003), men and women were asked to solve mazes on a computer for 15 minutes. Participants were paid either according to a piece rate (a dollar amount per maze solved) or according to a winner-take-all tournament. Under the piece-rate scheme, no significant differences in performance between men and women were found. When participants were paid on a competitive basis, however, the performance of men significantly increased, relative to the performance of women. However, similar to this result concerning selective participation, there is evidence that women who choose to compete perform as well as men in these settings (see, e.g., Datta Gupta, Poulsen, & Villeval, 2005).

In a field study, Gneezy and Rustichini (2004) study children’s running performance in a regular school physical education class. Children were asked to run twice over a short track while the teacher measured their speed. First, the children ran alone. Thereafter, the children were paired according to running times. This led to pairs of runners with different gender compositions. When the children ran alone, no gender differences in performance were noted. However, when the children ran in pairs—that is, in com-petition, running times for boys improved by 0.163 seconds on average, while the performance for girls decreased by 0.015 seconds relative to the when they ran alone. Boys were apparently spurred by competition to a larger extent than girls. One objection to these types of studies is the degree to which the produced results can be generalized to actual labor market settings. A similar concern is the extent to which results are task or location specific.

One area where competition has a potentially strong impact on labor market outcomes is in situations concerning bargaining. Deborah Small, Michele Gelfand, Linda Babcock, and Hilary Gettmen (2007) study bargaining using a laboratory setting where participants were told in advance that they would be paid between $3 and $10 for participation at the end of the experiment. After the participants finished their assigned tasks, the experimenter thanked them for their participation and asked, “Here is three dollars. Is three dollars OK?” Only 2.5% of female participants but 23% of male participants requested more money. In another study, Jenny Save-Soderberg (2007) uses data from two surveys of recent law graduates on their transitions from school to work. Results in this study indicate that in wage negotiations, women consistently submitted lower wage bids than men. It was also found that women received lower wage offers, even when women and men submitted the same wage bid. Negotiating for wages was, however, found to lead to higher wage offers than not submitting any wage bid at all.

The discussion thus far has pointed out that there seems to be evidence of gender differences in preferences, but no discussion on the possible reasons why there might be differences in preferences. Explanations for observed gender differences in preferences include possible genetic differences (see Colarelli, Spranger, & Hechanova, 2006). In contrast, Gneezy, Kenneth Leonard, and John List (2009) use an experimental approach to explore whether there are gender differences in selecting into competitive environments across cultures by examining a patriarchal society (the Maasai in Tanzania) and a matrilineal society (the Khasi in India). Similar to the evidence presented for Western societies, Maasai men opt to compete at twice the rate as Maasai women. The opposite result is found among the Khasi, where women choose the competitive environment considerably more often than men. These results suggest that existing societal structures are linked to observed gender differences in competitiveness. Croson and Gneezy (2009) summarize this literature and conclude that there is support for both genetic and cultural explanations for gender differences in competition. The interesting question for further research is what weight to attach to each of these factors in explaining gender labor market gaps.

Taken together, results from studies on gender differences in preferences suggest that men on average are more inclined to take risks than women and that men also prefer competitive environments to a larger extent. These findings potentially explain a portion of observed gender differences in the labor market because risk taking and competitive behavior are likely to be associated with pecuniary awards. To what degree observed gender differences in preferences are genetic or socially determined is an open question for further research.

Conclusion

This research paper has discussed the role of gender in the labor market from an economic perspective. Theoretically, there are potential supply-side differences by gender in skill acquisition and preferences that can explain why women earn less on average than men. On the other hand, there are also potential gender differences in opportunity—that is, discrimination prior to entering the labor market, affecting skill acquisition and working norms, as well as direct discrimination in the labor market, affecting employment and earning. Employers may, for example, expect women, to a larger extent than men, to have longer and more frequent absences from the labor market. Women as a group may then be perceived as a greater risk by employers then men, leading to lower wages and promotion possibilities and also to lower incentives to invest in human capital for future generations.

Evidence from studies on gender discrimination suggests that unequal treatment of equally qualified men and women continues to be important in the labor market. Experimental studies on discrimination tend, however, to be carried out on limited subsamples of the labor market, implying that it is difficult to make any general conclusions about the extent to which discrimination can explain gender gaps in the labor market.

Studies on gender differences in preferences, in turn, suggest that men on average are more inclined to take risks than women and that men also prefer competitive environments to a larger degree than women. These findings potentially explain a proportion of gender differences in the labor market because risk taking and competitive behavior are likely to be positively awarded by employers.

The consensus in the research literature on gender and economics is that both preferences and discrimination matter for labor market outcomes, outcomes that are of crucial importance for lifetime income, living standards, and intergenerational transmissions of income and opportunity, implying that attention must be paid to minimizing the impact these two overriding factors have in creating gender labor market gaps. There are also social changes in force that with time are likely to alter the underlying mechanisms behind these gender differences; among them, changing norms concerning the within-family allocation of labor between household and market, access to day care, and legislation promoting equal opportunity. The question remains to what degree social changes alone will eliminate differential opportunities by gender and to what degree more active measures are necessary to enforce equal opportunity.

Bibliography:

1. Akerlof, G. A., & Kranton, R. E. (2000). Economics and identity. Quarterly Journal of Economics, 115(3), 715-753.
2. Albrect, J., Bjorklund, A., & Vroman, S. (2003). Is there a glass ceiling in Sweden? Journal of Labor Economics, 21(1), 145-177.
3. Altonji, J. G., & Blank, R. (2003). Race and gender in the labor market. In O. C. Ashenfelter & D. Card (Eds.), Handbook of labor economics (Vol. 3C, pp. 3143-3260). New York: Elsevier North-Holland.
4. Arrow, K. (1973). The theory of discrimination. In O. A. Ashenfelter & A. Rees (Eds.), Discrimination in labor markets (pp. 3-33). Princeton, NJ: Princeton University Press.
5. Ashenfelter, O. A., & Hannan, T. (1986). Sex discrimination and product market competition: The case of the banking industry. Quarterly Journal of Economics, 101(1), 149-174.
6. Atkinson, S. M., Baird, S. B., & Frye, M. B. (2003). Do female fund managers manage differently? Journal of Financial Research, 26, 1-18.
7. Bagues, M. F., & Esteve-Volart, B. (2008). Can gender parity break the glass ceiling? Evidence from a repeated randomized experiment (FEDEA Working Paper No. 2007-15). Madrid, Spain: FEDEA.
8. Becker, G. (1971). The economics of discrimination (2nd ed.). Chicago: University of Chicago Press.
9. Becker, G. (1993). Human capital: A theoretical and empirical analysis, with special reference to education (3rd ed.). Chicago: University of Chicago Press.
10. Belter, A. H. (1982). Occupational segregation by sex: Determinants and changes. Journal of Human Resources, 17(3), 371-392.
11. Bergmann, B. (1974). Occupational segregation, wages and profits when employers discriminate by race or sex. Eastern Economic Journal, 1(1-2), 103-110.
12. Black, S. E., & Brainard, E. (2004). Importing equality? The impact of globalization on gender discrimination. Industrial and Labor Relations Review, 57(4), 540-559.
13. Black, S. E., & Strahan, P. E. (2001). The division of spoils: Rent sharing and discrimination in a regulated industry. American Economic Review, 91(4), 814-831.
14. Blau, F. D., Brinton, M. C., & Grusky, D. B. (Eds.). (2006). The declining significance of gender. New York: Russell Sage.
15. Blau, F. D., & Kahn, L. M. (1996a). International differences in male wage inequality: Institutions versus market forces. Journal of Political Economy, 104(4), 791-837.
16. Blau, F. D., & Kahn, L. M. (1996b). Wage structure and gender earnings differentials: An international comparison. Economica, 63, S29-S62.
17. Blau, F. D., & Kahn, L. M. (2000). Gender differences in pay. Journal of Economic Perspectives, 14(4), 75-99.
18. Blinder, A. S. (1973). Wage discrimination: Reduced form and structural estimates. Journal of Human Resources, 8, 436-155.
19. Carlsson, M., & Rooth, D.-O. (2008). An experimental study of sex segregation in the Swedish labour market: Is discrimination the explanation (IZA Discussion Paper No. 3811). Bonn, Germany: Institute for the Study of Labor.
20. Colarelli, S. M., Spranger, J. L., & Hechanova, M. R. (2006). Women, power, and sex composition in small groups: An evolutionary perspective. Journal of Organizational Behavior, 27, 163-184.
21. Croson, R., & Gneezy, U. (2009). Gender differences in preferences. Journal ofEconomic Literature, 47(2), 448-474.
22. Darity, W. A., & Mason, P. L. (1998). Evidence on discrimination in employment: Codes of color, codes of gender. Journal of Economic Perspectives, 12(2), 63-90.
23. Datta Gupta, N., Poulsen, A., & Villeval, M.-C. (2005). Male and female competitive behavior: Experimental evidence (IZA Discussion Paper No. 1833). Bonn, Germany: Institute for the Study of Labor.
24. Dimand, R. W., Forget, E. L., & Nyland, C. (2004). Retrospectives: Gender in classical economics. Journal of Economic Perspectives, 18(1), 229-240.
25. Doeringer, P., & Piore, M. J. (1971). Internal labor markets and manpower adjustment. New York: D. C. Heath.
26. Duguet, E., & Petit, P. (2006). Hiring discrimination in the French financial sector: An econometric analysis on field experiment data. Annals of Economy and Statistics, 78,79-102.
27. England, P. (1982). The failure of human capital theory to explain occupational sex segregation. Journal of Human Resources, 17(3), 358-370.
28. England, P. (1992). Comparable worth: Theories and evidence. New York: Aldine de Gruyter.
29. (2009). The European Union labour force survey. Available at http://epp.eurostat.ec.europa.eu/portal/page/portal/ labour_market/earnings/main_tables
30. Gneezy, U., Leonard, K. L., & List, J. A. (2009). Gender differences in competition: Evidence from a matrilineal and a patriarchal society. Econometrica, 77(3), 1637-1664.
31. Gneezy, U., Niederle, M., & Rustichini, A. (2003, August). Performance in competitive environments: Gender differences. Quarterly Journal of Economics, 1049-1074.
32. Gneezy, U., & Rustichini, A. (2004, May). Gender and competition at a young age. American Economic Review Paper and Proceedings, 377-381.
33. Goldin, C. (2006). The rising (and then declining) significance of gender. In F. D. Blau, M. C. Brinton, & D. B. Grusky (Eds.), The declining significance of gender? New York: Russell Sage.
34. Goldin, C., & Rouse, C. (2000). Orchestrating impartiality: The impact of “blind” auditions on female musicians. American Economic Review, 90(4), 715-741.
35. Gundersson, M. (1989). Male-female wage differentials and policy responses. Journal of Economic Literature, 27(1), 46-72.
36. Heckman, J. (1998). Detecting discrimination. Journal of Economic Perspectives, 12, 101-116.
37. Heckman, J., & Siegelman, P. (1993). The Urban Institute audit studies: Their methods and findings. In M. Fix & R. J. Struyk (Eds.), Clear and convincing evidence: Measurement of discrimination in America (pp. 187-258). Washington, DC: Urban Institute Press.
38. Hellerstein, J., Neumark, D., & Troske, K. R. (2002). Market forces and sex discrimination. Journal of Human Resources, 37(2), 353-380.
39. Institute for Women’s Policy Research. (2009). The gender wage gap: 2008. Available at http://www.iwpr.org/pdf/C350.pdf
40. Juhn, C., Murphy, K. M., & Pierce, B. (1991). Accounting for the slowdown in black-white wage convergence. In M. Kosters (Ed.), Workers and their wages: Changing patterns in the United States (pp. 107-143). Washington, DC: AEI Press.
41. Meng, X. (2004). Gender earning gap: The role of firm specific effects. Labour Economics, 11(5), 555-573.
42. Mincer, J. (1958). Investment in human capital and personal income distribution. Journal of Political Economy, 66(4), 281-302.
43. Mincer, J., & Polachek, S. (1974). Family investments in human capital: Earnings of women. Journal of Political Economy, 82(2), S76-S108.
44. Nekby, L., Skogman Thoursie, P., & Vahtrik, L. (2007). Gender and self-selection into a competitive environment: Are women more overconfident than men? Economics Letters, 100, 405-407.
45. Neumark, D. M. (1996). Sex discrimination in restaurant hiring: An audit study. Quarterly Journal ofEconomics, 111(3), 915-941.
46. Niederle, M., & Vesterlund, L. (2007). Do women shy away from competition? Do men compete too much? Quarterly Journal of Economics, 122(3), 1067-1101.
47. Oaxaca, R. L., & Ransom, M. L. (1999). Identification in detailed wage decompositions. Review of Economics and Statistics, 81, 154-157.
48. Olivetti, C., & Petrongolo, B. (2008). Unequal pay or unequal employment? A cross-country analysis of gender gaps. Journal of Labor Economics, 26(4), 621-654.
49. Pager, D. (2007). The use of field experiments for studies of employment discrimination: Contributions, critiques and directions for the future. ANNALS of the American Academy of Political and Social Science, 609, 104.
50. Phelps, E. S. (1972). The statistical theory of racism and sexism. American Economic Review, 62, 659-661.
51. Polacheck, S. W. (1975). Discontinuous labor force participation and its effect on women’s market earnings. In C. B. Lloyd (Ed.), Sex, discrimination, and the division of labor (pp. 90-124). New York: Columbia University Press.
52. Polachek, S. W. (1984). Women in the economy: Perspectives on gender inequality. In U.S. Civil Rights Commission (Ed.), Comparable worth: Issue for the 80s (Vol. 1, pp. 34-53). Washington, DC: Government Printing Office.
53. Reskin, B. F., & Bielby, D. D. (2005). A sociological perspective on gender and career outcomes. Journal of Economic Perspectives, 19(1), 71-86.
54. Riach, P. A., & Rich, J. (1987). Testing for sexual discrimination in the labour market. Australian Economic Papers, 26, 165-178.
55. Riach, P. A., & Rich, J. (2002). Field experiments of discrimination in the market place. Economic Journal, 112, F480-F518.
56. Riach, P. A., & Rich, J. (2006). An experimental investigation of sexual discrimination in hiring in the English labor market. B.E. Journal of Economic Analysis & Policy, 6(2), Article 1. http://www.bepress.com/bejeap/advances/vol6/iss2/art1
57. Save-Soderbergh, J. (2007). Are women asking for low wages? An empirical analysis of individual wage bargaining and ability signaling (Working Paper No. 7/2007). Stockholm: Swedish Institute for Social Research.
58. Small, D., Gelfand, M., Babcock, L., & Gettman, H. (2007). Who goes to the bargaining table? Journal of Personality and Social Psychology, 93, 600-613.
59. Weiselbaumer, D. (2004, Spring). Is it sex or personality? The impact of sex-stereotypes on discrimination in applicant selection. Eastern Economic Journal, 30, 159-186.