Thursday, 30 July 2009

Discriminating Against Women in the Job Market

(off global economics)

At the apropos of an FT article about the small rise of the gender pay gap in Britain, I thought I would share an anecdote.

After my years in Cambridge I took a gap year from academia, which turned out to be a gap decade. I was running a smallish research consultancy, seated in Hungary, but covering the macroeconomics and fixed income markets of many emerging markets. Our main client and almost all our smaller clients were in Western Europe. The size of the unit, at its peak, was 25 people. All our admin personnel were local Hungarians, but the analysts were not, we always tried to hire people from countries that we were covering. That gave me an insight into both how the labour market works in a host of emerging markets, but also the kind of decisions an entrepreneur has to face.

One of these was about discrimination.

We tried to be an equal opportunities employer. It so happened that as the research team grew it was entirely made up of white men. The two admin personnel were women, and the IT guys were, well, guys. Very traditional indeed. So, at one point, I thought that we should do something about this. We tried to hunt down a Roma analyst, but the market was illiquid, to say the least. We also decided to hire a female analyst. At the time, we were looking for someone from Turkey, and thus we advertised an analyst position in the Turkish media. We had a lot of applications, but all of them were men. Damn it! After some internal debates, we re-ran the ads, but with the line added that women are especially encouraged to apply. Now we got fewer applications, but all of them were women… (This probably is telling about Turkey as well…) We hired a well qualified woman, a really nice person.

But the story does not end there. In our business, our human resources model was that we were aiming at young macroeconomists or finance professionals who had a couple of years of work experience. Then we’d train them in-house, which was mostly on the job training. We had a rule of thumb that a new analyst would take one and a half years on average to become a fully functional, ‘useful’ part of the team. That took a lot of investment, and obviously only made sense if then our colleague would stay on after that period (which most of them did). Except that the nice Turkish lady started to try to get pregnant right after having moved to Budapest with her, also rather nice, husband. We discussed this, she was forthright about it, and Liz and I gave her advice (we just had our first baby a bit before) with regards to doctors and hospitals. It was all fair and fine.

Except for the fact that at the same time, our small company was pouring resources into her. Her, as an analyst. There was no way that we would ‘punish’ her for having a baby by not teaching her. But she ended up having a complicated pregnancy, and was away a lot from her job. And, after the baby was born, she decided to be with the baby a lot, and not to follow up on the previous plans of becoming a professional investment banking analyst. In other words, she made a set of personal choices: getting pregnant, long absence after the birth, and eventual departure from the career path, coupled by some uncertainty linked to her medical difficulties linked to the pregnancy. All of these were her decisions, and all of these bore costs for our research company. Part of the cost was personal: we had to fill in for the unexpected time she was not there, and thus some of us had to cut down on our holidays. One analyst, and good friend, in particular did not have a summer holiday that year, at all, as a direct consequence. Furthermore, her costs both in terms of her wage, taxes, the relevant overhead, and the time we had spent on teaching her were all lost, as well as the search cost that we had spent on a replacement. In a small company this is a lot. There was considerable amount of tension. We did hire an other female analyst from Turkey afterwards, but only because we could not live with the idea of not doing so. We knew however, that if she was to become pregnant too, our company would be in serious trouble.

I have recounted this story here to highlight an issue that is a taboo when talking about the pay gap. We like the fact that women can chose if and when they have babies. We also like the idea of equal pay. The two are in contradiction, and the cost of this contradiction is being forced onto the companies by our societies. In our story above everybody was a nice person, and all of them I count as my friends still now, six years later. But I cannot deny the fact that it would have made business sense to either hire a man for the same cost, or hire a women at a lower cost instead.

Most of my friends would throw hard objects at me at this point…

There are two observations on top of this. One. Obviously, there will be substantial variation among sectors. Some of this will come from economic rationality along the lines we faced, that is, long on-the-job training period of employees, and some along plain misogyny. One would expect that there would be sectors typically female or male due to economic reasons, and others due to ‘tradition’. When the society forces every employer, in every sector to offer exactly the same conditions for young men and young women, it may effectively counter some prejudices, but may, at the same time, force some good guys and gals into an awfully tricky situation.

Two. Perhaps part of the problem is that the feminist movement that highlighted the gender gap issues tends to have an ideological, political agenda. It seems that it is all too tempting for some proponents of the women’s issues to pour an anti-capitalist, anti-market sauce on labour market discrimination. However, unfortunately, this might weaken rather than strengthen the attack.

A solution suggested. I would think that there are two real questions here. First: who should pay for the kids, and second, how to reduce the cost of the inevitable uncertainty, by insuring the risk.

The current solution seems to be partial cost sharing, and coverage of a part of the risk. In the developed world, it does not make economic sense for the individual to give birth to babies, a marked departure from the past. Therefore, some of the cost is being born by the ‘pleasure parents’ who enjoy having kids. Some cost is being covered by the society, by providing education, playgrounds, healthcare, etc., when it does. There is a host of arguments why it may make sense for the society to invest into the future generation. But some costs fall on the employers. They do not get anything in return. Thus they discriminate. Once there are laws against discrimination, I suspect that there will be silent discrimination. The kind of pay-gap that we see, perhaps. (By the way, it is interesting that the gender pay-gap rises during a recession, perhaps a signal that the policy, at least in its current form, is a societal luxury.)

I would think that an insurance scheme to which everybody contributes up until the end of fertility (mid 40’s), or pension age (if you want to take into account that men have a longer period of fertility), which then compensates the companies once their employees go on a maternity or paternity leave, and then which gives a return of funds to those who did not have kids, would be a much fairer system. Companies would not feel an economic need to discriminate, thus real prejudice-based discrimination could be outed, and people who opted (or were unfortunate enough) not to have kids would not have to pay for them. If the society wanted to ensure future generations, then the investment into public education, health care, and child care in general should be increased (by the way, that is the model that seems to have worked for the Scandinavian countries).

You might think that this solution would essentially punish people for having children. You are right, in a way. It is a personal choice with cost consequences, and as long as I should have a control over it, I should pay for it. If you think that this is silly for then people will not have any kids, then your worry is probably about the future generations, and thus a societal function. You can solve this by increasing the state investment into kids.

However, for this, or any other solution to the opposing values of equal opportunity, personal liberty, and social fairness, to work, one needs to accept that the problem is much more complicated than just traditional prejudice lingering on. And talking about it should be a taboo no longer.


  1. Hi Tamas -
    This is very interesting (and I don't mean in the English sense...!) I often ponder these themes, and your solution is particularly intriguing amid all the discussion in the US right now about health care. It does seem your example is in part about the uncertainties of health in general, not just about gender and parenting choices.
    (EG What about hiring a man with a significant health issue that you know could keep him away from work for long periods of time?)
    But the parenting aspect has the interesting element of choice - and I think that leads to challenges for women in making decisions as well (I'll speak for myself here..!). How are we "supposed" to handle the combination of wanting a career AND wanting children? Particularly if health insurance is largely tied to employment, as it is here? It doesn't feel like many societies have figured out a comfortable answer to that yet, certainly the US has not, and as an individual I don't think I have either.
    And although having children is a "choice" it's also a fairly "natural" thing for many women to want to do (with an increasing sense of urgency, as you approach your 35th birthday!)
    The discrimination part comes in just because while all people who give birth are women (except that one guy last year, and his status was debatable..!) not all women give birth. So it's the classic case of identifying a correlation between an individual trait and an easily recognized group, and then extending assumptions to the whole group. If I get paid less because I'm a woman and thereby "more likely" to have a kid, that isn't very fair if I don't want kids or can't have them and my individual likelihood of "exhibiting the trait" is actually zero.
    And then there's the argument that if I do have kids, maybe the experience will actually improve my skills and the work I can do, even if it means I'll be physically absent for a while... and then there's the greater social value of having children... it seems like we haven't quite figured out how to quantify the economic value of children and parenting and make sense of it in the context of work, so it gets treated as a "gender issue" when it's really more complicated than that. Which I think is your point..!

    Anyway, good discussion, sorry we can't all discuss over a nice glass of wine, but thanks for posting it - I agree it's a topic that gets brushed under the carpet and yet is quite a huge factor in many decisions, both for employers and for individual women.

  2. Hey Jess,

    There is a big difference between the pregnancy question and a general health question. Although your actions affect your health, and getting pregnant is a chance event as well, there seems to be a much stronger randomness to the health problems than to the pregnancy question. Hence an outright insurance scheme, which works to a great extent as a health care finance solution, will not be enough for the pregnancy question.

    Re a man or a woman who is being discriminated on the job market due to a pre-existing health condition: Although, I would agree that it is likely anti-discrimination laws will just make the real life discrimination less visible, you could also argue that the health condition of the employee is due to a previous chance event. Thus it is more a question about how the principle of insurance should be applied rather than whether there is a case for it the first place. The point of my try at a solution is that it does not force people who are not in the risk pool into the pool, while at the same time might be able to deal with the moral hazard aspect.