Ask the grumpies: fertility timing

Cloud asks

Have you seen this study? http://people.virginia.edu/~am5by/fertilitytiming_sept2009.pdf

And, more importantly, what do you think of it?

The author of that study was hot hot hot on the market this year.  However, rumor is she’s decided not to move, despite offers.

The Effects of Motherhood Timing on Career Path
Abstract

This paper estimates the causal effects of motherhood timing on female career path, using national panel data from the NLSY79 and biological fertility shocks to instrument for the age at which a woman bears her first child. Motherhood delay leads to a substantial increase in earnings of 9 percent per year of delay, a smaller increase in wage rates of 3 percent, and an increase in hours worked of 6 percent. Supporting a human capital story, the postponement premium is largest for college-educated women and those in professional and managerial occupations. Family leave laws do not significantly influence the premium. Panel estimation reveals evidence of both fixed wage penalties and lower returns to experience for mothers: a “mommy track” is the likely channel for the timing effect.

All righty.  Before we get into the study itself, Ima gonna explain a little bit about empirical econometrics.  Economists are very interested in the idea of causality.  We’re not the only social science that focuses on causality, but we’re more likely to use something we call “natural experiments” rather than actual experiments (as psychologists do) to explore questions of causality (psychologists call these quasi-experiments).  X and Y are correlated, but does X cause Y, does Y cause X or is there a third variable, Q, that causes both?

One of the types of natural experiments we use is something magical called “Instrumental Variables” or IV for short.  IV is really neat because it basically takes a variable Z that we know causes X and does not cause Y (except through its effect on X).  The canonical example of a good IV is the Vietnam draft lottery number to study the effect of Vietnam service on labor market outcomes.  A worse draft number is correlated with actually serving in Vietnam, but because it is randomly assigned, is not related to labor market outcomes except through its effect on Vietnam service.  A “good instrument” will have these two qualities:  Z will be correlated with X, and Z will not cause Y except through the channel of X.  Most instruments are not as good as the Vietnam draft lottery, but we believe that they will tell us something anyway… but generally we don’t think these IV papers give the final word, just additional evidence.

In this paper, Amalia Miller uses several imperfect instruments to look at the effect of motherhood timing on female labor market outcomes.  Remember, in order to be a good instrument, the instrument will have to be strongly correlated with timing of motherhood and not correlated with female labor market outcomes except through the channel of motherhood timing.  The former we can test using statistics:  there are some heuristics we use for the t-stat, and there are some tests for low coefficients or “weak instruments” that are more or less accepted depending on your training.  The latter is where thinking is needed.

Miller’s chosen instruments are:

1.  whether first pregnancy ended in miscarriage

2. whether conception of the first child occurred while using contraception

3. elapsed time from first conception attempt to first birth.

Now think:  Are these related to female labor market outcomes through any means other than pregnancy timing?

I would argue yes:  All three of these are related to maternal health (including obesity), and health is directly related to labor market outcomes.  Miscarriage not only affects timing of motherhood, but also can cause depression, which relates to labor market outcomes.  Different types of people use contraception (because of religion, sense of responsibility etc.) and these personality characteristics may be directly related to labor market outcomes.  And infertility itself can be very time-consuming (I read a paper the other year on just on how much time is spent at the doctors, and how it has to be spent during the working day for the most part, for example).  I’m sure you can think of many more ways that Z, the instrument, relates to Y, the outcome variable through channels other than X.

Does that kill her results?  Well, no.  It just means we can’t be confident in the answer given in this paper (and it took a while to get this paper published and it’s not in one of the standard top econ field journals, probably for the above reasons).  If we attack the same question in many other imperfect ways and get the same results, we’ll feel more confident with the results in this paper.  We don’t really know the answer, but this should update our Bayesian priors to the results that she finds.  If we do come up with a perfect instrument or run a randomized controlled experiment (which we won’t do because we’re not Nazis) or get a much cleaner natural experiment that we can use a cleaner technique on (say, differences-in-differences) and get different results, then those results will trump these results.  In the absence of that, this paper is providing more information, and provides an improvement over the previous literature which is, at best, getting only at correlation.

So that’s a critique of the coefficients she finds– are they really accurate?  Probably not, but they’re probably closer to the “truth” than what the previous literature has found.

Let’s take the results as given.  Are her conclusions merited given the results?

Supporting a human capital story, the postponement premium is largest for college-educated women and those in professional and managerial occupations. Family leave laws do not significantly influence the premium. Panel estimation reveals evidence of both fixed wage penalties and lower returns to experience for mothers: a “mommy track” is the likely channel for the timing effect.

Well, let’s throw that open to the grumpy nation.  What do you think?  What are alternate explanations?

About these ads

18 Responses to “Ask the grumpies: fertility timing”

  1. Cloud Says:

    Thanks for answering my question! I did not know about the use of instrument variables. Why is using those preferred to just looking at maternal age?

    In terms of whether her conclusions are warranted… I’m not sure I understand that last sentence. Is she arguing that women who have their kids earlier are more likely to end up on a mommy track? If so, I suspect that is at least partly because they have less reputation and fewer contacts to fall back on to ensure continued advancement once the kids arrive. I also suspect discrimination- I have seen other studies that seem to show people discriminate against mothers in hiring and promotion decisions.

    But I also feel like I’m not quite understanding this post yet. I’ll have to think on it more. I’m looking forward to reading what everyone else thinks!

    • nicoleandmaggie Says:

      #2 said this post was really confusing, which makes sense as we spend about 2 weeks on instrumental variables in econometrics and many people still don’t fully understand them on the final.

      If you just look at the effect of maternal age on labor market outcomes, you’re getting a correlation, not a causation. Women who have children later have better labor market outcomes, just looking at correlation. The question is, “Does delaying childbearing produce better labor market outcomes?”

      In economics we think of two things that are major “confounds”– reverse causality and omitted variable bias.

      In terms of reverse causality, the idea would be that having better labor market outcomes causes women to delay their childbearing. This could certainly be true– if you’re getting a law degree or PhD or are working 80 hours/week, you may decide to put off childbearing for a while. You may realize you’re likely to be able to afford fertility treatment should you need it as well (Kasey Buckles has a working paper suggesting that, but I’m not sure I buy the argument since I don’t think people delay their childbearing because of the affordability of fertility treatment– I think that’s more likely to be second order).

      In terms of omitted variables bias: Think about the folks you know who had babies in high school and compare them to the folks you know who had their first babies in their late 30s and early 40s. They’re probably different in many ways besides just their child-bearing activity. Religion will be different, where folks were brought up will be different, feelings about education will be different, ability to delay gratification will be different, socioeconomic status of the parents will be different, and so on. So even if it looks like delaying childbearing is producing better labor market outcomes, what it could be that is actually happening is that people who are better at delaying gratification are less likely to have unprotected sex and they’re likely to stay in school longer and be better workers.

      She’s arguing that there’s both a lump sum penalty for women having children and that mothers get lower raises over their life-times than non-mothers and that that is consistent with a mommy-track story– mothers are sorting into jobs that pay less and don’t have as much room for promotion. (Another consistent story would be Becker’s “mothers do all the housework and are less productive at work” story.)

      There’s a good sociology matched pairs resume study that shows that women who put PTA work on their resumes are discriminated against in hiring compared to those who put neighborhood association volunteer work on their resumes.

  2. bogart Says:

    Chirp, chirp!

    I haven’t read the article. Thanks for this overview (side question: what’s the cite for the time-cost (specific to infertility treatment? or medical treatments more generally?) article you mention? I’d be interested to read it.). My one quick thought is to wonder about this point: Family leave laws do not significantly influence the premium.. I buy that that’s what she finds. But given that it’s a US sample, I wonder whether there’s enough variance in our family leave laws to detect anything even if it would exist (in some abstract experimental world). I know some states are better than others, but we pretty much range from “non-existent” to “lousy” by global standards.

    Kudos to you for accurate and effective use of the word “Nazis” in a blog post. This may be the first time that’s ever happened in the whole history of the internet.

    • nicoleandmaggie Says:

      Ugh, do not want to dig up the article. I have it somewhere. Eisenberg et al. (2010) find that infertile
      women seeking treatment spend a median of 51.5 hours on infertility-related activities over an 18-month period. Eisenberg, M. L., Smith, J. F., Saketsky, N., Millstein, S. G., Nachtigall, R. D., & Katz, P. P.
      (2010). The gender gap in the time costs of infertility treatment: Examination of a prospective
      U. S. cohort. Fertility and Sterility, 94, S17.

      That is a very good point on family leave laws. And given that most protection laws do have negative unintended consequences on the people they’re trying to protect, it’s hard to say whether or not that’s a good thing or a bad thing. I’m not 100% sure, but I think the evidence looking directly at family leave laws in the US finds that women with better laws take a wage cut but aren’t negatively affected with hiring. But I’m not positive. I also think that the evidence on European laws is really poorly done (except for the work on Sweden’s mandatory paternity leave) but that they find that more generous leave policies make it more difficult for women to get hired.

  3. Practical Parsimony Says:

    Having had quite a few professor friends and contacts, and encouraged by the feminists in my studies and in conversation, we all mused about when a certain professor would achieve tenure and how that would change her life. Single women married soon after tenure, dumping less accomplished boyfriends who were their support during their educational life.(One guy followed the professor around the country as she pursued her educaton and work. When he was dumped, I really felt sorry for him.) Then the women married a more accomplished man and became pregnant. The married women immediately became pregnant with their first pregnancies. The women who gained tenure immedately became more flamboyant, wearing colors and sexier clothing. They gave up their professional heels for footwear more comfortable. It was too universal to be just accidental or our imagination. An attorney friend was made partner and became pregnant the month after. I commented on her making partner and then getting pregnant, about how clever she was. She smiled as a way of confirming my observation and her intent. All these women seemed to realize motherhood would stifle any professional advancement. Remember, I am in the deep South. Of course, this is all anectdotal but confirmed as intent by many of the women.

  4. chacha1 Says:

    This is way outside my field but I can’t help thinking the “college educated and professional” variables have A LOT more to do with earning power than does timing of pregnancies. I mean, the average person is going to have much better “labor outcomes” with a college education and a professional/managerial career focus than without either of those. Regardless of childbearing.

    • nicoleandmaggie Says:

      Right, and sizing the true effect of education on wages is one of the big, as yet, unanswered questions of labor economics. (The main omitted variable is “ambition”– it leads to both education and earnings and cannot be measured.) If someone can truly nail that, there’s a Nobel prize waiting for them.

      The effect of fertility timing on labor market outcomes is a smaller question, but still relevant.

  5. bogart Says:

    Thank you for the cite. Did the author happen to (or does she plan to) run the same analysis on male respondents (ha! note my word choice — happen to! Like, oh, lalala, I was just walking along and stumbled into this multivariate analysis! Happens all the time … :)).

    • nicoleandmaggie Says:

      Re: the hours spent, I have no idea. But I can tell you that my DH definitely did not spend anywhere near as long as I did at the doctor for infertility stuff. Even if he’d had more wrong with him, it wouldn’t have required as regular treatment (surgery potentially, but not regular trips while on clomid to get 4d u/s to see if an egg had popped yet).

      Re: the Miller paper, I don’t know. The same instruments shouldn’t be available for men in that survey. Also since the correlation tends to go the other direction for men (kids = higher wages), it’s a less interesting question.

      • IF Says:

        yeah, aside from cases that need surgery, the male contribution (even in the cases where the infertility is male-factor) to the IUI or IVF is over & done in 15 minutes. The women needs the constant monitoring, ultrasounds, injectables, blood draws, etc… when we did IUI, we had to go in at 7:30am together (I HAD to be there, since the appointment was under my name) for his contribution. Then he skipped off to work & I had to WAIT for the sperm washing/spinning & then my turn for the IUI which was usually 9:00-10am. So he missed no work, I missed quite a lot.

  6. hush Says:

    Good post – and I bet y’all are fantastic teachers.

    Perhaps a minor point, but Miller’s first (imperfect, yes) instrument was whether the first pregnancy ended in miscarriage *or stillbirth*. Not sure “stillbirth” in the very late term >24 weeks gestation sense nor in the >36 weeks gestation sense nor in the birth accident sense belongs in the same category as “miscarriage” in the first trimester sense. Hmm… Miller indicated a smallish sample size, so this might not be particularly relevant.

    I wish Miller would have analyzed private, industry-specific leave policies instead of just public ones. One of the footnotes referenced women in academia, medicine (A1B while a resident or an attending?), law & accounting (big firms norms might be 12 weeks of paid leave). Huge differences there. Beyond the scope of her analysis perhaps.

    • nicoleandmaggie Says:

      You’re right– stillbirths and miscarriages are going to have potentially different effects (presumably with the stillbirth the company has thought of you as pregnant). But you’re also right that it’s a small dataset and likely if she removed stillbirths it wouldn’t make much of a difference since there are so few. (And I’m not sure if the question allows that to be separated or not.)

      Again, with the small dataset it’s difficult to break down leave policies by industry, and to some extent they’re endogenous (confounded with) to other aspects of those industries and jobs. Women might be making choices because of those policies or because of things correlated with those policies (like choosing to go into K-12 education if you’re planning on having kids young). We can think of state and federal leave policies as potentially exogenous to (not confounded with) women’s labor market and child-bearing choices, so it’s a bit cleaner. But definitely an interesting question.

  7. Comradde PhysioProffe Says:

    I didn’t understand your post at all, but is a reasonable explanation of the observed relationship between age at birth and career path that the younger you are, the greater the ultimate compounded effects going forward of both slower advancement and lower salary starting points for raises?

  8. BigLittleWolf Says:

    Is anyone including divorce as a factor? Hello? Not to mention the state in which divorce (with children) occurred, which dramatically changes the financial outcome for the woman (likely caring for the children)?

    You can can-can those career outcomes, when the post-marital mess hits, and variables no one can predict enter the picture, along with a few that we can…

    • nicoleandmaggie Says:

      Interestingly, fertility and sex of the children affects the probability of divorce.

      There are a lot of other things that state affects, so generally we put state fixed effects (essentially controlling for how states are different from each other) into the regressions.

  9. MutantSupermodel Says:

    I feel completely dumb right now. I think, basically, it’s saying because i didn’t put off having babies I’m doomed to make crap for the rest of my life. Which is just the sort of thing I needed to read today. :/


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 180 other followers

%d bloggers like this: