The Intersection

Sex, Work, Children, Trends

Science bloggers are still abuzz over Wednesday’s discussion of women in academia and now Razib (one of my very favorite sciblings) has taken the topic one step further.  He collected data on the mean number of hours worked last week broken down by sex and compared that with the number of children per individual. I always enjoy reading Razib’s amusing analyses (and justified reason to link great flicks from the 80s):

As you can see, the more children women have, the fewer hours they
worked last week (on average). Women are primary care givers, no
surprise. On the other hand, you can see a different trend for males. I
suspect that reflects the reality that losers can’t find mates.

i-5c560948a494d4fd311f981b94168871-hours worked children.png


  1. #1 Angela
    January 23, 2009

    Love the links!

  2. #2 Lilian Nattel
    January 23, 2009

    It looks to me like men work more hours for the first couple of kids and then it declines to a bit above the level of no kids. That might have to do with financial demands (to support kids) and demographics (ie who has more than 5 kids).

  3. #3 Walker
    January 23, 2009

    On the other hand, you can see a different trend for males. I suspect that reflects the reality that losers can’t find mates.

    Huh? The graph is number of children. It has nothing to do with marital status.

    While this may sound very fieldish of me, this is exactly why I don’t trust graphs from biologists without error bars or methodology.

  4. #4 Sheril R. Kirshenbaum
    January 23, 2009

    That’s what I suspect although there are too many variables to actually extrapolate much without more information.

    I doubt Razib meant marriage by ‘mates’ and he’s joking of course…

  5. #5 Isis the Scientist
    January 23, 2009

    Wait, wait, wait, Sheril!!! Stop the freakin’ presses? Is this the data from the 160,000 PhD-holders? And people with no kids are working 39 and 42 hours a week? Am I the only one to who that makes no sense? Where are the error bars on this crap?

    39 hours per week my tight little ass!

  6. #6 Sheril R. Kirshenbaum
    January 23, 2009

    Why of course not Isis…

    It’s data from Razib based on a single week, not limited to science and engineering, and not intended to be taken completely seriously…

    Further, more surprising are those women with 7 children clocking over 30 hours!

  7. #7 Isis the Scientist
    January 23, 2009


  8. #8 ctenotrish
    January 23, 2009

    Umm, if you think a person with one or more children isn’t *working*, you are off your rocker! Maybe not working for salary, but trust me, parents work.

  9. #9 Ashutosh
    January 23, 2009

    As mathematician Paul Halmos said:

    “If your paper is looking vacuous,
    use the first-order functional calculus.
    then it becomes logic,
    and as if by magic,
    the obvious is hailed as miraculous”

    Nothing can kick off a debate on the obvious like statistics.

  10. #10 razib
    January 23, 2009

    to replicate my findings, go here:

    go to the top-left, click “analysis,” then click “comparison of means.”

    dependent: “hrs1”
    row: “childs”
    column: “sex”

    you’ll see ability to filter & stats at the bottom.

    tx for the link sheril 😉

  11. #11 bsci
    January 23, 2009

    WIth regards to academia, I think a more relevant chart is figure 7 at:

    This is a survey of faculty in the University of California system (all campuses). While it does show that women with children do less job-related work than men with or without children, it also shows that part of the issue is that, even at the faculty level, women with children spend much more time than men on caregiving/housework.

  12. #12 amphibious
    January 23, 2009

    Isn’t it more important to discuss how to make academia more hospitable to parents of both genders? (Yes, I’m a guy; my biases: pursuing a doctorate, engaged to a woman with a professional degree, and we’ve discussed this topic frequently.)

    Both men and women freely select their careers and the amount of time they devote to raising children. I think I have the same right to expect a work structure that allows me to be an involved parent as anyone else for the simple reason that I’m human.

    I don’t expect to be able to be as involved with my children as I should be while competing for university tenure – so I’ve adjusted the expectations of my career to one that allows a balance of work achievement and family. In other words, I’m taking myself out of competition for the “top spots” a priori because my CV is less important than my (hypothetical and not yet conceived) children.

    Incidentally, the discussion might benefit from an analysis of veterinarians, particularly at universities. The graduating classes have been heavily skewed towards women lately (Michigan State had an entire cohort of dozens of students, 100% of whom were women).

  13. #13 bsci
    January 23, 2009

    amphibious. I think the point you are overlooking here is that things done to make universities more hospitable to active parents IS more significant for women to men. Whether it makes sense or not, more women than men are significant care-givers and policies that benefit families benefit them more.

    That said, policies to benefit parents in academia should be gender non-specific whenever possible. For example there’s not reason to allow tenure clock slowing for women, but not men when a child is born. On the other hand, things like private spaces so nursing mothers can pump is clearly a female specific benefit with no male comparison.

  14. #14 Isis the Scientist
    January 23, 2009

    Isn’t it more important to discuss how to make academia more hospitable to parents of both genders? (Yes, I’m a guy; my biases: pursuing a doctorate, engaged to a woman with a professional degree, and we’ve discussed this topic frequently.)

    Blah, blah, blah.

  15. #15 leigh
    January 24, 2009

    38 hours… i don’t know whether to laugh hysterically or cry at that. i don’t know when the last time i had a 38 hour week was, but it wasn’t in grad school.

  16. #16 Brenda, or perhaps Carl
    January 24, 2009
  17. #17 Lindsey
    January 24, 2009

    I agree with ‘Brenda, or perhaps Carl’ so reposting the full aforementioned comment:

    I thought the graph at The Intersection blog (see comment #2) was interesting. According to that graph, women with no children work fewer hours than men with no children. This gender difference was as great as that for a woman with no children compared to a woman with 5-6 children (which is a lot of children!).

    It seems to me, based on this data, that children are less of a factor in time spent at work than gender. It might be useful to discuss why.

    Possible explanations below. Some are inflammatory and unrealistic, some otherwise. You decide:

    1) Women are lazier than men.
    2) Women work more efficiently than men.
    3) Women have more unavoidable non-work, non-child-related commitments than men.
    4) The data in the graph are completely bogus for some reason or another.

    If numbers 1 or 3 are true, it indeed behooves an institution to avoid hiring women. If number 2 is true, women should be preferentially hired, especially for positions that pay hourly wages. If number 4 is true, then there is yet no data here supporting the conclusion that women are professionally crippled by a need to care for children.

    Perhaps I missed some possible explanations?

  18. #18 daedalus2u
    January 24, 2009

    The obvious explanation to me is that men lie about how much they work more than women do.

  19. #19 Doormat
    January 25, 2009

    Razib is not being terribly helpful on his blog, but if you dig around, you can find the raw data here: and search for “HRS1”. The exact link (without frames) is

    The question asked is “1a. If working, full or part time: how many hours did you work last week, at all jobs?” Notice how there are large clusters around 30, 60, and especially, 40 and 50 hours. A total of 51020 people answered. The data comes from “General Social Surveys, 1972-2006”: see

    Now, the important point is: This data is from a cross-section of American society. It has nothing to do with academia, and certainly isn’t data taken from people who work in academia, or who necessarily have doctorates etc.

    So, how to interpret the data. It’s well known that women are over-represented in part-time jobs, for example. That alone would explain the fact that women work fewer hours than men, regardless of the number of children they have. On the comment that Lindsey posted, I think this fits under (4), as most other commentators seem to be assuming that this data applies specifically to academics… Actually, bsci makes exactly this point:

New comments have been disabled.