Home Production Market Production and the Gender Wage

  • Slides: 30
Download presentation
Home Production, Market Production, and the Gender Wage Gap: Incentives and Explanations Stefania Albanesi

Home Production, Market Production, and the Gender Wage Gap: Incentives and Explanations Stefania Albanesi and Claudia Olivetti Presented by: Melissa Mc. Ellin Thursday, February 21, 2008

Presentation Outline n n n n Background Previous Studies Main Issue Model and Assumptions

Presentation Outline n n n n Background Previous Studies Main Issue Model and Assumptions Connecting the Model with Evidence Implications Critique

Background n An emergence and persistence of gender differences, in market wages and in

Background n An emergence and persistence of gender differences, in market wages and in the division of labor within the household: n n 10% differential in female and male wages in the U. S. (O’Neill 2003) Husbands’ home hours are about 1/3 of wives’ (PSID 1976 -2001) n Incentive pay in the labor market

Previous Studies n Francois (1998) n n n Model subjecting only one class of

Previous Studies n Francois (1998) n n n Model subjecting only one class of jobs to incentive problems Efficiency jobs are assigned to men only, in an equilibrium with female wage discrimination Shortcomings: He has the female wage differential stemming from job segregation – His model cannot account for gender differentials within the same occupation

Previous Studies n Gayle and Golan (2006) n n Focus on gender differentials in

Previous Studies n Gayle and Golan (2006) n n Focus on gender differentials in on-the-job human capital accumulation Shortcomings: Discusses only how statistical discrimination leads to occupational sorting and changes in labor market experience between genders

Previous Studies n Goldin (1986) n n n People choose occupations based on their

Previous Studies n Goldin (1986) n n n People choose occupations based on their expected life-cycle labor participation This leads to certain expectations within the labor market which lead to differences in monitoring costs and incentive pay Shortcomings: Introduces incentive pay and expectations, but does not explain occupational arrangements within groupings.

Main Issue n n Joint determinations of gender differentials in earnings and in the

Main Issue n n Joint determinations of gender differentials in earnings and in the household division of labor Identify the source of statistical discrimination, within occupations, in light of the incentive problem on the labor market

Hypothesis n “Incentive problems in the labor market amplify differences in earnings due to

Hypothesis n “Incentive problems in the labor market amplify differences in earnings due to gender differentials in home hours. In turn, gender earnings differentials reinforce the division of labor within the household, leading to a potentially selffulfilling feedback mechanism. ”

Building on Previous Ideas n Builds on 3 existing literatures: n Sexual division of

Building on Previous Ideas n Builds on 3 existing literatures: n Sexual division of labor n Statistical discrimination n Incentive contracts and job design

Model n n n Economy of adult agents, identical except for gender, and identical

Model n n n Economy of adult agents, identical except for gender, and identical firms All agents are married and belong to a household 2 types of good: market good and home good

Assumptions n n Utility cost of labor effort is increasing in home hours Effort

Assumptions n n Utility cost of labor effort is increasing in home hours Effort as well as home hours are private information (unobserved) n n n Unobservability of home hours adverse selection Unobservability of effort moral hazard Households value a public home good produced with the time of both spouses Households optimally choose the contribution of each spouse to home production Workers with high home hours are less attached to market work Male and female workers are equal in labor market productivity, but females are more productive in home hours than males

Model: Labor Contracts n n Since home hours are unobserved, the optimal menu of

Model: Labor Contracts n n Since home hours are unobserved, the optimal menu of contracts will depend on the firms’ belief about the distribution of home hours Since gender is observable, firms can offer different incentive contracts to female vs. male workers – but contract space is unlimited, so they will do this if and only if they believe that the distribution of home hours differs across genders

Model: Labor Contracts n n If firms observe home hours but not effort, they

Model: Labor Contracts n n If firms observe home hours but not effort, they would only face a moral hazard problem If both home hours and effort are unobserved, this introduces adverse selection n High home hour workers will put in less effort, and will thus self sort into jobs with higher salary pay and less incentive pay Low home hour workers will put in more effort at work, so they will sort into jobs with higher incentive pay Incentive pay becomes a combination of output (rather than “effort”) and gender (rather than “home hours”)

Model: Households n n Household production (G) is a function of home hours of

Model: Households n n Household production (G) is a function of home hours of the female (hf), home hours of the male (hm), and the amount of market good used in home production (k) G = g (hf, hm, k) Individual labor earnings can finance purchases of the market good used in home production and can be transferred across spouses

Model: Choice of Home Hours n The optimal allocation of home hours within the

Model: Choice of Home Hours n The optimal allocation of home hours within the household depends on the spouses’ relative opportunity cost of home hours and therefore on the prevailing labor contracts.

Model: Equilibrium n n The household takes labor contracts as being given constraints by

Model: Equilibrium n n The household takes labor contracts as being given constraints by the firms; chooses home hours based on this Ability to bear children is a reason women have an advantage in home hours

Model: Equilibrium n Equilibrium with Ex-Ante Differences Across Genders n n n Assume male

Model: Equilibrium n Equilibrium with Ex-Ante Differences Across Genders n n n Assume male and female workers are equally productive in market work, but women are more productive in home work = the measure of women’s higher relative productivity in home work A high value of , resulting from poor medical knowledge, implies that the only possible equilibrium is one where women are mostly devoted to home production while men specialize in market work Improvements in medical technology reduce the value of , so an ungendered equilibria is possible However, gendered initial conditions and market beliefs or expectations may create a self-fulfilling pattern, so an ungendered equilibria may never prevail

Model: Feedback n Predictions: n n 1. The female/male earnings ratio should be lower

Model: Feedback n Predictions: n n 1. The female/male earnings ratio should be lower when the incentive problem is more severe 2. These effects are stronger when the difference in home hours is greater 3. The wife/husband ratio of home hours should display a negative correlation with the wife/husband earnings ratio 4. The wife/husband ratio of home hours should display a positive correlation with the husband/wife difference in the fraction of incentive pay.

Connecting the Model with the Evidence n n Interpret contracts specifying different levels of

Connecting the Model with the Evidence n n Interpret contracts specifying different levels of effort as corresponding to different positions within a firm Assume that uncertainty associated with a worker’s effort should be higher for management and sales jobs relative to production jobs incentive pay can be used more in these jobs

Connecting the Model with Evidence n Empirical counterparts of predictions 1 & 2: n

Connecting the Model with Evidence n Empirical counterparts of predictions 1 & 2: n n 1. Gender earning differentials should be higher in occupations in which the incentive problem is more severe. 2. These effects should be stronger for married than for never-married workers.

Connecting the Model with the Evidence: Census Data n n Census sample includes all

Connecting the Model with the Evidence: Census Data n n Census sample includes all white individuals between 25 and 54 years of age Individuals employed at least 50 weeks in the past year and usually work at least 30 hours per week Consider 3 occupational categories: sales, management, and production Analyze both married and never married workers

Notes on the Chart n n n Large variation in the ratios of male/female

Notes on the Chart n n n Large variation in the ratios of male/female median earnings across industries conditional on marital status, but always lower for married workers For married workers, the female/male earnings ratio is lowest in management and sales occupations For never married workers, the earnings ratio rankings are reversed

Connecting the Model with the Evidence: PSID n n Unlike the Census, incentive pay

Connecting the Model with the Evidence: PSID n n Unlike the Census, incentive pay and home hours are reported in the PSID Find a strong negative correlation between the female/male earnings ratio and the male/female difference in the fraction of incentive pay Correlation coefficient = -0. 65, significant at the 1% level Empirical counterparts of predictions 3 & 4

Summary of Results n n Incentive problems in the labor market play a major

Summary of Results n n Incentive problems in the labor market play a major role in their model Incentive pay in the labor market amplifies gender earnings differentials These earnings differentials reinforce the division of home hours This can help to explain the substantial and persistent gender earnings gap that remains unexplained by gender differences in schooling, actual experience, and job characteristics

Implications n n n Highlights the importance of incentives and difference in the pay

Implications n n n Highlights the importance of incentives and difference in the pay structure Reveals that expectations of a gender wage gap, which characterize both male and female workers, can have a large impact (Babcock and Laschever 2003) Suggests that equilibria with gender discrimination will be hard to break

Critique: What is Good n n n Addresses within-occupation differences in earnings Examines the

Critique: What is Good n n n Addresses within-occupation differences in earnings Examines the joint decisions of labor hours and home hours Correlates with the empirical evidence on incentive pay well

Critique: What could be improved § Questionable Assumptions: § § § As home hours

Critique: What could be improved § Questionable Assumptions: § § § As home hours increase, labor effort will decrease Utility cost of work effort is increasing in home hours While equal in the labor force, women are more productive at home § Inadequate data set for comparison § Limited analysis stemming from the inability to address selection into different firms or industries § Assumes no efficiency losses for gender discrimination § Does not address how technological changes can affect female labor force participation