Elsevier

Social Science & Medicine

Volume 94, October 2013, Pages 17-25
Social Science & Medicine

Gender, family status and physician labour supply

https://doi.org/10.1016/j.socscimed.2013.06.018Get rights and content

Highlights

  • No average male–female physician market hours gap if never married & no children.

  • Marriage and children have opposite associations for male and female labour supply.

  • Children induce much larger increases in work at home for female physicians.

  • The spouses of male and female physicians behave quite differently.

  • Physicians reduce hours of labour market work when married to another physician.

Abstract

With the increasing participation of women in the physician workforce, it is important to understand the sources of differences between male and female physicians' market labour supply for developing effective human resource policies in the health care sector. Gendered associations between family status and physician labour supply are explored in the Canadian labour market, where physicians are paid according to a common fee schedule and have substantial discretion in setting their hours of work. Canadian 1991, 1996, 2001 and 2006 twenty percent census files with 22,407 physician observations are used for the analysis. Although both male and female physicians have statistically indistinguishable hours of market work when never married and without children, married male physicians have higher market hours, and their hours are unchanged or increased with parenthood. In contrast, female physicians have lower market hours when married, and much lower hours when a parent. Little change over time in these patterns is observed for males, but for females two offsetting trends are observed: the magnitude of the marriage-hours effect declined, whereas that for motherhood increased. Preferences and/or social norms induce substantially different labour market outcomes. In terms of work at home, the presence of children is associated with higher hours for male physicians, but for females the hours increase is at least twice as large. A male physician's spouse is much less likely to be employed, and if employed, has lower market hours in the presence of children. In contrast, a female physician's spouse is more likely to be employed if there are three or more children. Both male and female physicians have lower hours of work when married to another physician. Overall, there is no gender difference in physician market labour supply after controlling for family status and demographics.

Introduction

Over the last several decades one dramatic change across OECD countries in the structure of the physician workforce is the increasing participation of women. In the United States, the proportion of female physicians increased from 20% to 30% between 1990 and 2007 (NCHS, 2009). In Canada the share increased from 12% in 1980 to 36% in 2010 (CIHI, 2011). Implications of this trend are noteworthy for health human resource planners and researchers. A key issue is that, on average, female physicians practise fewer hours than their male counterparts. Understanding the sources of differences between male and female physicians' work hours is essential for developing effective human resource policies.

Most economic studies of labour market differences between female and male physicians have focused on understanding gaps in earnings and/or wages. Contributions include Ohsfeldt and Culler (1986), Rizzo and Zeckhauser (2007), and Theurl and Winner (2011). Across developed countries a common finding is that females have lower annual earnings, although there is more debate – see, for example, Baker (1996), and Bashaw and Heywood (2001) – about gaps in hourly wages. Gaps in earnings stem from differences in work hours and/or hourly wages; therefore work hours are central to understanding the mechanics of this issue.

Physician labour supply is, additionally, an important question independent of earnings since physician time is key to service provision. Research looking at trends in hours of work includes Crossley, Hurley, and Jeon (2009) and Sarma, Thind, and Chu (2011), with the latter noting that the presence of children influences female physicians' hours; service provision is explored, for example, by Constant and Léger (2008). These studies observe females providing fewer hours or services per year. Similarly, Watson, Slade, Buske, and Tepper (2006) find that in 2001 average female general practitioners in Canada had paid workloads equivalent to 68% of their male counterparts. Understanding physician labour supply and gender issues is useful for human resource planning purposes in the health care sector given that medical fields are highly regulated and entry is limited. Although not an issue restricted to physicians, the public return also increases with greater physician work time (within safe limits) since training is taxpayer subsidized.

Canada's institutional context is particularly amenable to this study and permits a contribution to the large research literatures regarding gendered labour market differentials surveyed by Bertrand (2011), and the economics of the family discussed by Browning et al. (2014). It has a publicly financed single-payer system (there is virtually no private sector for medically necessary physician services) where physicians have flexibility in choosing their hours of work. Most physicians are not employees but self-employed professionals. Some physicians, such as surgeons, may have aspects of their practice restricted because of limited access to facilities such as operating theatres, but there are not normally limitations on office services. A small number of physicians have salaried positions, but even here those wishing to work extra clinic or office hours may normally do so. Furthermore, for most of our data period there was a perceived physician shortage (Postl, 2006). Thus physicians not only have enormous flexibility in setting their hours, but there has been social and government pressure to increase them and offer after-hour services.

Beyond flexible hours, gender gaps in market wages have been proposed as one reason that females allocate less time to the labour market. Pre-labour market gender discrimination may exist regarding the allocation of, and/or self-selection into, medical specialties (Gjerberg, 2002), but in Canada, conditional on specialty, the fixed and universal fee schedule for all medically necessary services that is negotiated between each provincial government and its medical association largely eliminates fees as a source of gender discrimination given the single-payer system. Whenever they deliver a service in either the outpatient or inpatient sector, physicians are reimbursed by the government based on the fee schedule. These common fees imply equal gross payments per service regardless of gender – i.e., equal earnings potential. Net payments may, however, vary if physicians have different overhead costs. In particular, physicians who mostly work in hospitals or certain clinics have lower overhead costs, but the fee schedule reflects this to some extent. Payment per hour may also vary with treatment style and/or productivity. Importantly, these production side decisions are endogenous and potential income – payment per service – is gender blind. Beyond medically necessary services, physicians may also receive payments for publically uninsured services such as completing insurance application forms or cosmetic procedures. On average, this source of income is modest and provincial medical associations commonly provide guidance (including a recommended fee schedule) regarding billing for uninsured services. Overall, while pre-labour market gender based barriers may exist in some contexts (e.g., Nomura & Gohchi, 2012), for practicing physicians the Canadian labour market serves as a laboratory allowing us to study how, given current social norms and individual preferences, highly educated females and males respond when equal gross pay per service (reflecting potential earnings) has been realized and workers have substantial discretion in choosing their market hours.

In this paper, we use Canadian census masterfiles to characterize gender differences in the relationship between family structure and labour supply broadly defined; we focus on weekly hours of market work, although we also examine other elements of labour supply including weeks of paid work per year, part-time employment, and hours of non-market work. Family responsibilities, especially child care, have been cited as reasons for female physicians' lower hours of paid work. Few studies have, however, formally examined the relationship between family structure and physician labour supply from a household perspective. Furthermore, there is little evidence on how spousal characteristics are related to it. Jacobson, Nguyen, and Kimball (2004) show that female physicians who are parents have significantly reduced hours of market work. Gjerberg (2003), using Norwegian data, looks at how family obligations are combined with market work finding that specialty choice and the probability of working part-time are affected by the presence of children, and that having a spouse who is a physician improves career outcomes. Although her focus is on earnings, of particularly relevance is Sasser (2005) who examines how much of the gender gap in annual earnings among physicians is due to women's greater family responsibilities using panel data for young US physicians. Comparing before and after status changes, she finds that female physicians earn 11 percent less once married, plus an additional 14 percent less after having one child or 22 percent less after having had two or more children.

The remainder of the paper is organized as follows. Section two presents a short review of the family status-labour supply research literature outside of health economics. The data and econometric methods used are briefly described in the third section; Section four reports descriptive statistics; and regression based empirical results are presented in Section five. The last section discusses the findings and concludes.

Section snippets

Family status and labour supply patterns

Research on the economics of the family regards specialization in production and economies of scale as principal sources of economic gains to co-habitation. Traditionally, if women have a comparative advantage in home production, while men have a comparative advantage in the labour market, then wives specialize more in home production, while husbands concentrate on the labour market (e.g., Becker, 1991; Lundberg & Rose, 2000, 2002). Empirical work in developed countries observes that married

Data and methods

Our study uses pooled Canadian Census masterfiles from 1991, 1996, 2001 and 2006, which are 20% random samples of the population. For this paper's topic, the census is the only available data source. The data are accessed through Statistics Canada's Research Data Centre (RDC) at McMaster University. University ethics review is not required for RDC projects. In addition to being able to credibly identify a large and representative sample of physicians, the census has information on marital

Descriptive analysis

For the 1991 and 2006 Censuses, Table 1 shows sample means and, for the dependent variables only, standard deviations. Though the means are high for physicians compared to the population, average labour market weeks and hours are lower for female than male physicians, and declined slightly for both sexes between 1991 and 2006. Females' hours of work at home, however, are substantially greater than those for males. Part-time status is much more common for females. Interestingly, the sum of

Family status and physician labour supply

OLS regression results with labour market hours as the dependent variable are shown in the first two columns of Table 3; those for hours of work at home are displayed in column (3), annual weeks of market work in (4), and marginal effects from a probit model of part-time status in (5). As reflected for the Female coefficient, even in our large sample with commensurate small standard errors, never-married male and female physicians with no children have market labour supply outcomes that are

Conclusion and discussion

Building on previous studies that find a relationship between the presence of young children and physician market labour supply, such as Watson et al. (2006) and Sarma et al. (2011), this analysis explores relationships between aspects of family structure and various dimensions of labour supply. Moreover, it extends the literature by providing evidence on work at home.

Family status has a quantitatively important relationship with labour supply for both male and female physicians. On average,

Acknowledgements

We would like to thank Jeremiah Hurley, Lonnie Magee, two anonymous referees and the editor, and participants at the Canadian Economics Association and the International Health Economics Association meetings for helpful comments. Support from the Canadian Labour Market and Skills Researcher Network (CLSRN), and the Ontario Health Human Resources Research Network (OHHRRN), is gratefully acknowledged; the funders had no role in the research. Opinions expressed in this paper do not necessarily

References (28)

  • K. Antonovics et al.

    Are all the good men married? Uncovering sources of the marital wage premium

    American Economic Review Papers and Proceedings

    (2004)
  • L. Baker

    Differences in earnings between male and female physicians

    New England Journal of Medicine

    (1996)
  • D.J. Bashaw et al.

    The gender earnings gap for US physicians: has equality been achieved?

    Labour

    (2001)
  • G.S. Becker

    A treatise on the family

    (1991)
  • Cited by (44)

    • The gender earnings gap in medicine: Evidence from Canada

      2022, Health Policy
      Citation Excerpt :

      However, these studies have focused only on specific specialties (i.e., surgical) and/or provinces, rather than the entire Canadian medical profession. Furthermore, they are limited by focusing on gross revenues or billings rather than incomes/earnings net of practice expenses, and most do not adjust for labour supply which is seen to differ by gender [24]. As noted by others, beyond little being known about the gender pay gap among Canadian physicians, an advantage of analyzing physicians exclusively is that heterogeneity, in terms of both observed human capital investments and unobserved characteristics such as work commitment, is diminished as these individuals have selected to invest many years (10 to 13 on average) in post-secondary education to complete their medical training [16].

    • How do gender differences in family responsibilities affect doctors' labour supply? Evidence from Australian panel data

      2020, Social Science and Medicine
      Citation Excerpt :

      Second, compared with prior studies, our study examines a more extensive set of family covariates describing not only the presence and number of children, but also the age of a doctor's youngest child, as well as spousal employment status. Gjerberg (2003) and Wang and Sweetman (2013) analyse the effects of children and the number of children, both using cross-sectional data. Recent work by Cao and Rammohan (2020), done independently of ours using data from the MABEL survey, considers the presence of dependent children but not the effect of children's age.

    • Gender differences in the incomes of self-employed French physicians: The role of family structure

      2019, Health Policy
      Citation Excerpt :

      Clearly, family status and the presence of children quantitatively influence male and female doctors’ labour supply: while there is little difference in labour supply depending on gender for doctors who have never been married or have no children, marriage – through the fact that a spouse's income offers greater flexibility [27] – and the presence of children both have a strong influence on the number of hours worked, depending on gender and on the number of children. While female doctors usually reduce the number of hours worked after a birth, male doctors tend to increase their working hours [27,43]. The increasing feminisation of the medical workforce is therefore of growing concern for regulators, particularly in countries that offer social policies to promote birth rates.

    View all citing articles on Scopus
    View full text