PhD Article: “Why did you choose economics?”

It is a question I have gotten a lot, from friends and family to new acquaintances. The answer is not—rather unfortunately, perhaps—out of true love for the subject or a desire to push the bounds of theory. Being closer to sociology than theory, I view economics as a tool-kit: one that allows us to tease answers out of complicated issues and bring hard facts to support arguments that at times can seem ephemeral. Economists have the potential—and sometimes the privilege—of being important voices in the debate over social issues.

The past few years have seen an increased focus on these topics, in academia as well as in the public sphere. While substantial gains have been made, especially in terms of LGBT rights, events such as the “locker room talk” scandal during the 2016 US presidential election, the subsequent rollback of transgender rights, and the strenuous pushback against the Black Lives Matter movement, have made it clear how much work has yet to be done.

In economics, the gender wage gap has been the subject of extensive research. A particular subset of this research focuses on what is termed the “motherhood wage gap.” This refers to a phenomenon where women appear to receive a wage penalty after having a child. So what have these decades of research told us? Firstly, that this wage penalty is present in research using a wide variety of functional forms (not simply OLS, but also panel data regression, propensity score matching, non-parametrics, etc). It also resists attempts to account for it through human capital depreciation or work effort theory (which postulates that the demands of caring for a newborn prevent women from putting as much effort into their jobs as they did before becoming mothers). Not only is this phenomenon present in empirical data, but also in lab experiments: mothers are rated as less competent—and thus worthy of less pay—than non-mothers. Estimates for the motherhood wage penalty most often range from 5 to 8%, depending on how many children a woman has and how quickly she returns to her job after having had them.

At the same time, a small but ever-growing strand of literature into the wage effects of sexual orientation has revealed the possible presence of a “lesbian wage premium”. This has proven harder to pin down, for the simple reason that any research on LGBT issues relies on individuals being willing to self-identify as LGBT. The fact that marriage equality for LGBT individuals has only really started to take root in the Western world in the past two decades is a large part of the roadblock in trying to identify LGBT individuals. However, researchers have managed to circumvent these difficulties, in some cases by collecting specialised datasets themselves, or—as is my case—using household data to identify same-sex couples. Estimates of the lesbian wage premium range from 30% at one extreme, to insignificant at the other, but most research has tended to find a wage bonus of perhaps 10-20%.

Identifying the intersection of the two aforementioned phenomena, then, is a tricky business. Not only is it subject to the same difficulties outlined in the paragraph above, but fertility levels among lesbian women are substantially lower than straight women. This makes a very large dataset necessary in order to identify even a few LGBT individuals. In my own dataset, consisting of over 1.3 million observations in total, I have just 1,400 lesbian observations.

Preliminary research on this topic (of which there is precious little) suggested something curious: while heterosexual women receive a motherhood wage penalty, homosexual women appeared to instead receive a wage bonus for having children. Could the ‘lesbian wage premium’ then be explained, in part, by different wage dynamics of motherhood? This question has been the subject of my research for over a year.

I used household microdata from the European Union Standards of Income and Living Conditions (EU-SILC) survey; the particular dataset I used spans from 2006 to 2013. These dates encompass precisely the years in which a majority of European states legalised marriage equality (except for Belgium and Spain who had already legalised it). There is evidence to suggest that this legalisation did result in more individuals being willing to self-identify as LGBT: although just 39% of observations correspond to countries and years with marriage equality, nearly 60% of all lesbian observations are found here. Using a random effects regression with a rich set of controls, I attempted to get to the bottom of the mystery. Here is what I found:

Firstly, there are no significant effects associated with motherhood for lesbian women. There is of course, a negative and significant motherhood coefficient for straight women of around 8%, which is consistent with most of the literature on the subject (if perhaps a little on the high end). Secondly, the motherhood coefficients for lesbian and straight women are, in fact, not significantly different from each other. This means we cannot reject the hypothesis that lesbian women experience exactly the same wage dynamics of motherhood as straight women.

The most interesting results, however, were lying just under the surface. I constructed dummies to split couples into primary and secondary earners.  By interacting this with the motherhood variable I identified which mothers earned more (primary earner) or less (secondary earner) than their partner. The news is fairly grim for straight women who earn less than their husbands: among this cohort, the coefficient of motherhood is a whopping -16%. Those straight women who earn more than their husbands are only moderately better off: though they still experience a statistically significant wage penalty, it is just a quarter of the size, at -4%. For lesbian primary and secondary earners, in contrast, the coefficients are 4% and -24%, respectively. Once more, however, the coefficients for primary straight and lesbian mothers are not significantly different from each other—and neither are the coefficients for secondary straight and lesbian mothers. So again, we cannot say that the wage effects of motherhood are any different for lesbian women, even when we allow for different coefficients depending on earner status. This result is nevertheless important, because women are far more likely to be the secondary earner in a straight couple than in a lesbian one. This implies that what seems at first to be different wage effects is simply a numbers game: a reflection of the fact that straight women are more likely to be secondary earners.

In conclusion: there does not appear to be any evidence that lesbian women experience different wage dynamics of motherhood than straight women. Secondary earner women experience a wage penalty at least four times larger than primary earner ones, and the “lesbian motherhood wage premium” found in preliminary research likely stems from this. There are a few next steps to be taken, however. The next one—which is the current topic of my research—is to investigate fatherhood wage dynamics for gay versus straight men, and whether the primary and secondary earner differentials also apply.

I would be remiss if I did not use my remaining column space to discuss the very real social implications of this research. It implies that, no matter how generous maternity and paternity laws become (Europe has some of the most generous in the world), it may be impossible to eradicate the motherhood wage penalty because it has its root in the gender wage gap itself. Despite decades of research and active legislation, progress to close this gap has ground almost to a halt.

So what part do economists have to play in all this? I believe economists have a duty, not simply to create research or weigh in on economic policy, but to be active participants in the discussion surrounding social issues as well. It is crucial, on social issues especially, not to limit ourselves to passive chroniclers of wage discrimination. It falls on economists first and foremost to communicate to the public clearly and effectively, wherever necessary, that wage discrimination is real and not a myth.

Further legislation has limited scope to reduce wage discrimination. What is needed now is nothing more or less than a sea change at a societal level. An interested observer might find hints of such a change in the rising tide of the #MeToo movement—but the work is far from done. We all have a part to play in society’s progress.

By Julia Martí Hoefer




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s