Abstract
Using data drawn from the Current Population Surveys, this paper provides a consistent explanation for why the presence of a working wife reduces the husband’s wage among managers, but increases the husband’s wage among non-managers. It is not husband’s occupation per se but rather the distribution of husbands’ wage levels that underlies the working spouse penalty or premium. Positive correlations in earnings between married couples that arise from assortative mating make the cross-wage effects of the husbands’ wages on the wives’ hours of work first positive, then negative in cross-sectional data. The phenomenon of a working spouse penalty/premium is simply the flip side of this relationship.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
By using housework data drawn from the National Survey of Families and Households, Hersch and Stratton (2000) are able to carry out a rather direct test of Becker’s explanation for the male marriage wage premium. They find that married men whose wives are employed do not significantly increase their time spent on housework, and that married men’s hours of housework do not reduce their marriage wage premium.
One could preferably argue for a four-equation model by additionally including the husband’s labor supply equation and the wife’s wage equation, which incorporates the simultaneous determination of wages and hours (Moffitt, 1984) in a model of family labor supply (Ashenfelter & Heckman, 1974). In order for my results to be comparable with those of the previous studies on the working spouse penalty, I have adopted the two-equation model.
The variables in X and Z are described in detail in Sect. 4.
This approach, rather than including the wife’s wage directly, is usually taken so as to avoid the censoring problems caused by the absence of wage data for nonworking women. See for example Buchmueller and Valletta (1999); both Jacobsen and Rayack (1996) and Hotchkiss and Moore (1999) treat the wife’s wage in a similar way.
Three possible predicted values can be derived from a Tobit estimation: a simple linear prediction, using the estimated coefficients from the Tobit and the regressors; an unconditional expected value; and an expected value, conditional on positive hours of work. The results reported here are based on an unconditional expected value, which seems to be the instrument commonly used by Jacobsen and Rayack (1996), and Hotchkiss and Moore (1999).
Hotchkiss and Moore (1999) have also used data drawn from the 1993 March Current Population Survey and showed that the effects of the wife’s hours of work on the husband’s wage vary by occupation.
I pooled non-consecutive years of the CPS in order to avoid overlapping observations. Owing to the sampling rotation scheme used in the CPS, approximately half of the observations in the consecutive March surveys overlap.
The group of occupations with three-digit codes of 003-037 in the 1980 Standard Occupation Classification is categorized as managers. Employing a narrower definition of managers used by Hotchkiss and Moore (1999)—the group of occupations with three-digit codes of 007-022—does not change the qualitative results of this paper.
In addition to the reported coefficients, both regressions include the following independent variables: three region dummies; nine industry dummies; seven occupation dummies; and two year dummies.
The same set of instruments has been used by Hotchkiss and Moore (1999). The robustness of the IV estimates is discussed in Sect. 6.
It seems that the working spouse premium among blacks, noted in Table A.2, is caused by differences in the other explanatory variables such as education and union status.
Lam (1988) provides a theoretical analysis that predicts positive assortative mating by spouses’ wages.
Using longitudinal data of the Swedish population, Nakosteen, Westerlund, and Zimmer (2005) find that after controlling for observed traits such as age and education, spouses manifest a positive correlation even in their disability propensities prior to marriage.
The independent variables included in the wife’s wage regression are age and its square; three dummies for race/ethnicity; three dummies for level of education; three region dummies; a SMSA dummy; and two year dummies. The independent variables additionally included in the wife’s sample-selection regression are family non-labor income; a dummy for the presence of a preschool-aged child; and number of children. The husbands wage regression includes the following independent variables: age and its square; two dummies for race/ethnicity; three dummies for level of education; three region dummies; a SMSA dummy; and two year dummies.
In addition to the quadratic specification, I have also tried fitting a cubic specification and a LOESS nonparametric function. The results, not reported here, support those from the quadratic specification.
Note that 35.71/(2*6.32) = 2.82, which is $16.78 in 2001 dollars.
Control variables included in column 2, and in column 3 as well, are wife’s age and its square; three dummies for wife’s race/ethnicity; three dummies for wife’s level of education; family non-labor income; a dummy for the presence of a preschool-aged child; and number of children; a SMSA dummy; three region dummies; and two year dummies.
The turning point of 2.34 is $10.38 in 2001 dollars.
One may argue that because of assortative mating many of these variables are likely to be correlated with the error term in the wife’s hours-of-work equation and thus may not be valid instruments. Still, the union dummy is considered to be a valid instrument because there is no reason why it should be correlated with the error term in the wife’s hours-of-work equation.
Using the March CPS, Juhn and Murphy (1997) graphically show that there was a negative relationship between the husband’s wage and wife’s probability of working in 1969, while this relationship had changed into a quadratic one, where wives of men in the middle of the wage distribution working the most, by 1989. Therefore, the quadratic relationship between the husbands’ wages and the wives’ labor supply might be a recent phenomenon, though it requires further research.
It certainly is possible to have a working spouse penalty (or premium) in the full sample depending on the composition of the full sample. For example, in their IV estimation of the effect of the wife’s hours of work on the husband’s log hourly wage, Daniel (1992), Gray (1997), and Chun and Lee (2001) all have found a working spouse penalty in their full sample. Furthermore, differences in the choice of IV might be another reason for finding a working spouse penalty (or premium), as will be discussed below.
Even though the results of the Durbin–Wu–Hausman test of exogeneity vary by subgroup, I have used the IV estimates for their consistency.
This relationship does not change even if the two groups with a small sample size—black managers and Hispanic managers—are excluded.
These variables are expected to be uncorrelated with the error term in the husband’s wage equation. And the Tobit estimation results reported in Table 6 show that these variables are highly significant in determining the wife’s hours of work. The wife’s age and its square, three dummies for wife’s race/ethnicity, and three dummies for wife’s level of education also are included, as determinants of the wife’s wage, in predicting the wife’s hours of work. Given the findings in this paper, however, that there is a positive correlation both in wages and in earnings residuals of husbands and wives as a result of assortative mating, this set of variables cannot be considered instruments.
Bound, Jaeger, and Baker (1995) show that IV estimates are biased in the same direction as OLS estimates when the correlation between the instruments and the endogenous explanatory variable is weak. The estimates of the correlation coefficient between the wife’s actual hours and predicted hours are available from the author upon request.
References
Amemiya, T. (1979). The estimation of a simultaneous-equation tobit model. International Economic Review, 20, 169–181.
Ashenfelter, O., & Heckman, J. J. (1974). The estimation of income and substitution effects in a model of family labor supply. Econometrica, 42, 73–86.
Becker, G. S. (1991). A treatise on the family. Cambridge: Harvard University Press.
Blackaby, D., Carlin, P., & Murphy, P. (1998). What a difference a wife makes: The effect of women’s hours of work on husbands’ hourly earnings. Bulletin of Economic Research, 50, 1–18.
Blau, F. D., & Kahn, L. M. (2005). Changes in the labor supply behavior of married women: 1980–2000. NBER Working Paper 11230.
Bound, J., Jaeger, D. A., & Baker, R. M. (1995). Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak. Journal of American Statistical Association, 90, 443–450.
Buchmueller, T. C., & Valletta, R. G. (1999). The effect of health insurance on married female labor supply. Journal of Human Resources, 34, 42–70.
Chun, H., & Lee, I. (2001). Why do married men earn more: Productivity or marriage selection? Economic Inquiry, 39, 307–319.
Daniel, K. (1992). Does marriage make men more productive? Economic Research/NORC, Working Paper #PRC 92-2.
Davidson, R., & MacKinnon, J. G. (1993). Estimation and inference in econometrics. New York: Oxford University Press.
Devereux, P. J. (2004). Changes in relative wages and family labor supply. Journal of Human Resources, 39, 696–722.
Gray, J. S. (1997). The fall in men’s return to marriage. Journal of Human Resources, 32, 481–504.
Hersch, J., & Stratton, L. S. (2000). Household specialization and the male marriage wage premium. Industrial and Labor Relations Review, 54, 78–94.
Hotchkiss, J. L., & Moore, R. E. (1999). On the evidence of a working spouse penalty in the managerial labor market. Industrial and Labor Relations Review, 52, 410–423.
Jacobsen, J. P., & Rayack, W. L. (1996). Do men whose wives work really earn less? American Economic Review, 86, 268–273.
Juhn, C., & Murphy, K. M. (1997). Wage inequality and family labor supply. Journal of Labor Economics, 15, 72–97.
Killingsworth, M. R., & Heckman, J. J. (1986). Female labor supply: A survey. In O. Ashenfelter & L. Richard (Eds.), Handbook of Labor Economics (Vol. 1, pp. 103–204). Amsterdam: North-Holland.
Lam, D. (1988). Marriage markets and assortative mating with household public goods: Theoretical results and empirical implications. Journal of Human Resources, 23, 462–487.
Maddala, G. S. (1983). Limited-dependent and qualitative variables in econometrics. Cambridge: Cambridge University Press.
Mare, R. D. (1991). Five decades of educational assortative mating. American Sociological Review, 56, 15–32.
Moffitt, R. (1984). The estimation of a joint wage-hours labor supply model. Journal of Labor Economics, 2, 550–566.
Nakosteen, R. A., Westerlund, O., & Zimmer, M. A. (2004). Marital matching and earnings: Evidence from the unmarried population. Journal of Human Resources, 39, 1033–1044.
Nakosteen, R. A., Westerlund, O., & Zimmer, M. A. (2005). Health-related disabilities and matching of spouses: Analysis of swedish population data. Journal of Population Economics, 18, 491–507.
Nakosteen, R. A., & Zimmer, M. A. (2001). Spouse selection and earnings: Evidence of marital sorting. Economic Inquiry, 39, 201–213.
Neal, D. (2004). The measured Black-White wage gap among women is too small. Journal of Political Economy, 112, S1–S28.
Pencavel, J. (1998). Assortative mating by schooling and the work behavior of wives and husbands. American Economic Review, 88, 326–329.
Reimers, C. W. (1985). Cultural differences in labor force participation among married women. American Economic Review, 75, 251–255.
Schneer, J. A., & Reitman, F. (1993). Effects of alternate family structure on managerial career paths. Academy of Management Journal, 36, 830–843.
Smith, J. P. (1979). The distribution of family earnings. Journal of Political Economy, 87, S163–S192.
Stroh, L. K., & Brett, J. M. (1996). The dual-earner dad penalty in salary progression. Human Resource Management, 35, 181–201.
Acknowledgments
I wish to thank Hyunbae Chun, Hsien-Hen Lu, Daniel Hamermesh, Deborah Cobb-Clark, Christopher Surfield, Shoshana Grossbard, seminar participants at Union College, Queens College (CUNY), Colgate University, University of Tsukuba, and University at Albany (SUNY) and participants at the 2004 Annual Meetings of the Society of Labor Economists and the 2005 Annual Meetings of the Southern Economic Association for valuable comments.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Rights and permissions
About this article
Cite this article
Song, Y. The working spouse penalty/premium and married women’s labor supply. Rev Econ Household 5, 279–304 (2007). https://doi.org/10.1007/s11150-007-9013-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11150-007-9013-2