Federal Reserve Bank of New York Staff Reports Preference for the Workplace, Investment in Human Capital, and Gender Matthew Wiswall Basit Zafar Staff Report No. 767 March 2016 Revised March 2017 This paper presents preliminary findings and is being distributed to economists and other interested readers solely to stimulate discussion and elicit comments. The views expressed in this paper are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.
Preference for the Workplace, Investment in Human Capital, and Gender Matthew Wiswall and Basit Zafar Federal Reserve Bank of New York Staff Reports, no. 767 March 2016; revised March 2017 JEL classification: J16, J24 Abstract We use a hypothetical choice methodology to estimate preferences for workplace attributes and quantify how much these preferences influence pre-labor-market human capital investments. This method robustly identifies preferences for various job attributes, free from omitted variable bias and free from considering the equilibrium job match. Women on average have a higher willingness to pay (WTP) for jobs with greater work flexibility and job stability, and men have a higher WTP for jobs with higher earnings growth.
These job preferences relate to college major choices and actual job choices, and explain as much as 25 percent of the gender wage gap. Key words: workplace preferences, compensating differentials, human capital, college majors, gender _________________ Zafar: Federal Reserve Bank of New York (e-mail: basit. Wiswall: Arizona State University, University of Wisconsin-Madison (e-mail: matt. The authors thank Ellen Fu and John Conlon for providing excellent research assistance, as well as Joe Altonji for providing feedback on the survey design, Eleanor Dillon, and participants at various seminars and conferences for valuable comments.
This paper is a revised version of NBER Working Paper w22173 (April 2016). The views expressed in this paper are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. 1 Introduction Economists have long recognized that job and occupational choices are not solely determined by expected earnings.1 Although simple models based on earnings maximization abound (see for example the classic Roy, 1951, model), and are quite useful in some applications, it is also clear that individuals have a rich set of preferences for various aspects of jobs beyond expected earnings, including earnings and dismissal risk, and various non-pecuniary aspects such as work hours flexibility and enjoyment of workplace activities. We would expect then that these preferences for various job attributes affect not only job choices, as individuals choose jobs not solely to maximize expected earnings, but also human capital investments, as individuals alter their human capital investment in anticipation of particular future job choices.
However, empirically isolating the role of worker-side preferences for job attributes is difficult. One reason is that the equilibrium allocation of workers to jobs reflects not only the workers’ preferences but the structure of the labor market and firm demands for workers. If one assumes the labor market is perfectly competitive with jobs of all types offered to all workers, then the equilibrium job choices and wages observed directly identify individ- ual preferences–this is the classical model of compensating differentials (Rosen, 1984). If however firms have preferences for some types of workers and offer jobs only to a subset of workers with preferred characteristics (employment discrimination of some form), then the observed job choices do not reflect worker preferences only.
Various kinds of labor market frictions, which prevent workers from matching with their most preferred job types, also break the direct connection between observed job choices and worker preferences. Even when the labor market is perfectly competitive, a second empirical challenge is that because jobs likely vary in many unobserved (to the researcher) characteristics, there exists a familiar omitted variable (selection bias) problem in identifying worker preferences from realized job choices. If the observed characteristics in realized choice data are correlated with the unob- served characteristics, then these estimates of the importance of the observed characteristics in workers’ job choices are biased. To address these empirical challenges, this paper presents a methodology for estimating individual preferences for workplace attributes prior to labor market entry.
We collect data on job attribute preferences through a survey in which we present undergraduate students with a 1 See the famous quote by Adam Smith who lists a number of non-pecuniary job attributes which “make up for a small pecuniary gain in some employments, and counterbalance a great one in others." Wealth of Nations, 1776, Book 1, Chapter 10. 1 series of hypothetical job choice scenarios and elicit their expected future choices across the jobs. The hypothetical job scenarios were constructed to offer students a realistic menu of potential jobs varying in expected earnings and other characteristics such as future earnings growth, dismissal probability, and work hours flexibility. The students’ stated preferences for these jobs allows us to construct a “pure" measure of individual preferences – at the time of the survey – for various job characteristics and estimate, in a simple and robust way, the distribution of their preferences for job attributes.
In this way, our data isolates the preference for workplace attributes, free from making explicit assumptions about the equilibrium job allocation mechanism, preferences of employers, and free from considering key omitted variables. In contrast to our approach, previous work addressing compensating differentials using observed job choices requires generally stronger assumptions about preferences and the firm side of the labor market. Several decades of direct empirical approaches using realized wage differentials to measure compensating differentials have yielded mixed results, some finding large, and others small, compensating differentials for various job attributes (Thaler and Rosen, 1975; Gronberg and Reed, 1994; Van Ommeren et al. More recent work has incorporated compensating differentials into rich models of the labor market, allowing for important features such as search frictions, unobserved job attributes, and dynamic incentives for occupational choices (see for example, Bonhome and Jolivet, 2009; d’Haultfeouille and Maurel, 2013; Bronson, 2015; and Lim, 2015.
Motivating our approach, Hwang et al. (1992) and Bonhomme and Jolivet (2009) conclude that search frictions can imply small equilibrium wage differentials although there are in fact substantial preferences for non-wage job amenities. In a departure from this previous literature, our hypothetical choice methodology is a kind of “stated choice" analysis, similar to “conjoint analysis" and “contingent valuation" methods, used in fields including marketing, environmental and natural resource economics, and health.2 Because our data collection in essence conducts a kind of “experiment" at the 2 In the marketing and environmental contexts, these methods are often used to identify preferences for new, as yet unavailable, consumer products or for public goods like environmental quality, for which realized choices and markets do not exist. For examples in marketing and consumer choice, see Green and Srinivasan (1978), Beggs, Cardell, and Hausman (1981), Louviere and Woodworth (1983), Manski and Salomon (1987), and Ben- Akiva and Morikawa (1990).
For examples, in environmental and resource economics see Smith (2004), Kling, Phaneuf, Zhao (2012), Carson (2012), and Hausman (2012). Studies using stated choice in health include Scott (2001) who estimates the preferences for pecuniary and non-pecuniary characteristics of workplaces among doctors; see also Lagarde and Blaauw (2009) and de Bekker-Grob. More recently Blass, Lach, and Manski (2010), Delavande and Manski (2015), Ameriks et al. (2015), and Fuster and Zafar (2015) have used this hypothetical choice methodology to analyze preferences for electricity reliability, political candidates, 2 individual student level, the “panel" data generated by our design allows us to estimate the distribution of preferences allowing for unrestricted forms of preference heterogeneity.
We combine this data on job attribute preferences with rich data on students’ educational choices and expectations, including data on how students believe potential college majors relate to the jobs which would be available to them, and test whether the job preferences young adults hold in college in fact affect their human capital investments during college. In our sample of high ability students, we estimate substantial willingness-to-pay (WTP) for pecuniary and non-pecuniary aspects of jobs and considerable heterogeneity in their pref- erences for workplace attributes. We find that students have preferences reflecting, on aver- age, a dis-taste for higher job dismissal potential, and a taste for workplace hours flexibility. We estimate that on average students are willing to give up 2.8% of annual earnings for a job with a percentage point lower probability of job dismissal.
The largest average WTP es- timate is for the availability of a part-time hours option. Individuals, on average, are willing to give up 5.1% of their salary to have a job which offers the option of working part-time hours rather than a job which does not offer this option. When dividing our sample by gen- der, we find that women have a higher average preference for workplace hours flexibility, with an implied willingness-to-pay of 7.3% compared to 1% for men. Women also have a higher average WTP for more secure jobs- they are willing to give up 4% of their salary for a percentage point lower probability of job dismissal (versus a 0.6% WTP for males).
We find similar differences in the gender-specific median WTP estimates for these attributes. On the other hand, men have a higher average WTP for jobs with higher earnings growth: they are willing to give up 3.4% of annual earnings for a job with a percentage point higher earnings growth (the corresponding estimate for women is a statistically insignificant 0. There is, however, substantial heterogeneity in preferences for workplace attributes, even within gender. In addition, the skewness of the estimated preference distributions is significantly different from zero for most attributes, suggesting that parametric distributional assumptions that impose symmetry in preferences are not supported in the data.
A natural question is whether preferences recovered from data on hypothetical choices relate to actual occupational outcomes. Using data on reported job characteristics for a subset of our respondents who are employed roughly four years after our original data collection, nursing home care and asset allocation, and housing demand, respectively. Like these previous studies, we exploit the possibilities that hypothetical data can provide rich variation in product characteristics, allowing for identification of preferences under weak assumptions about the form of preference heterogeneity. However, our primary motivation for collecting hypothetical choice data is not because markets and realized choices do not exist, as is the issue in identifying preferences for new products or public goods, but to resolve problems of endogeneity of realized job choices.
3 reveals a strong and systematic relationship between estimated preferences and later actual workplace characteristics. Students with strong preferences for flexible hours, dis-taste for hours, and other non-pecuniary aspects of jobs were later found to be more likely to be working at jobs with those same preferred characteristics. Although these realized job char- acteristics do not solely reflect preferences (given the issues we raised above), our finding of a correlation between pre-labor market job preferences and later actual job characteristics suggests some added credibility of our research design. Our finding of substantial differences in willingness-to-pay for job amenities between men and women is consistent with prior work noting that the gender segregation of occupa- tions and jobs is such that women are more likely to be found in jobs offering greater work- place flexibility (Goldin and Katz, 2011; Flabbi and Moro, 2012; Goldin, 2014; Wasserman, 2015; Bronson, 2015).
However, the observation that women tend to work in certain job types may not reveal women’s preferences alone, but may also be affected by firm-side de- mands for specific workers and discrimination or be driven by some other job attributes that are unobserved in our datasets (Blau and Kahn, 2006).3 Our innovation is to quantify the willingness-to-pay for job attributes using a flexible and robust methodology.
