Wednesday, June 11, 2014

A Great Reason for Prospective Graduate Students to Pick Princeton

As a Harvard professor, I never thought I would write a post with THAT title!

But after my national survey of all PhD-granting departments (and joint departments) of astronomy or astrophysics (there are 28 of them), Princeton has emerged hands-down as the winner for financially supporting graduate students with young children.

I have written previously about the current state of paid leave for graduate students. Nineteen of the 28 departments (68%) in the survey now offer at least 6 weeks paid leave for graduate students who become parents. Most excitingly, the landscape is changing rapidly, with pending policy changes at several of the institutions that currently offer no paid leave. I would like to see all departments in our community recognize that new parents need at least 3 months (not 6 weeks) of paid leave, and be inclusive of all parents (some benefits are currently restricted to birth mothers). But at least we can assert that paid leave among the 28 US departments of astronomy is now the norm.

But this post isn't about paid leave: Rather, it is about the much larger financial challenge of childcare.


I frequently see colleagues with children trying to get by with far too little childcare, and I have witnessed their frustration when they realize that progress on their research isn't going to happen only at night when the kids "sleep". (It's a lie! Who came up with the expression "sleep like a baby"?)

OK, let's be honest: I am describing myself! I tried to get by with far too little childcare -- and that was as a new assistant professor, with significantly more resources than a graduate student! Graduate school in astrophysics is a full time job. For graduate school parents with no partner, or a partner working full time, the reality is that they will likely need something close to full time childcare. In some cases, a parent or friend can provide this at little or no cost, but for the others, it must be purchased.

Full time childcare for a 3 year old can easily run $15,000 even after the (relatively modest) tax benefit. The number is significantly higher for infants. In some parts of the country the number will be a little less, but regardless it is fair to say that full time childcare would amount to at least 50% of the take home paycheck of a graduate student.

I think that the presidents and deans of many universities would support in principle the idea of subsidizing childcare for graduate students, but the issue is decided not by ideals but rather by cost:  To put this in perspective, let's consider the recent hard-won move to providing 6 weeks of paid leave. For a graduate student receiving a $30,000/year stipend and health benefits, a 6 week paid leave and the retention of subsidized health care amounts might cost $4500 per leave taker. In many cases, external funding and fellowships would cover the leave, so the cost per leave taker is actually less.

Compare this to providing a meaningful subsidy for full time childcare! Leave is a one-time cost per child, but childcare is a standing annual cost. A meaningful subsidy might cover 1/3rd of the annual cost, so $5000 per year per child. It is no mystery why even universities with well-meaning leaders have not provided this subsidy: For a 6 year PhD, the cost would be $30,000, far more than the $4500 cost we have only recently seen on offer to cover the leave. Surely no university could offer such a subsidy?

Actually, some do! Princeton is the leader: Their policy provides $5000/year/child (up to 2 children), providing up to $10,000/year. For a graduate student (even one with a partner who collects a salary), this would represent an enormous change to her or his budget. Several other schools are also putting real money on the table: As best I could determine, the runners up are the University of Michigan, which offers $4000-$8000/year for 1-3 children, and Cornell, which offers $1950/year/child. Importantly, several other schools offer discounted rates at their University-run daycare centers, including the University of Florida, the University of Colorado, and the Ohio State University. Unfortunately, the majority of the 28 schools in the survey offer nothing, but, like paid leave, I believe that landscape is changing rapidly too.

This issue isn't, of course, relevant only for graduate students who become parents during their PhD years. I think that perhaps the greatest impact is for older students who have started down other career paths, begun their families, and then decided to pursue the PhD in astrophysics. If we don't offer meaningful childcare subsidies, we are likely failing to recruit from this pool. Yet this pool may contain some of the best talent: These are individuals who have considered their career options carefully and made a deliberate and likely difficult choice to leave their current track and begin a career in astrophysics. Moreover, I bet (but I have not studied the numbers) that a progressive childcare policy might particularly impact recruitment of women considering the move to astrophysics.

Speaking from our experience at Harvard, our admit list has, at least for the past few years, included individuals in this group: These are prospective students who are slightly older than our typical admit and who are also parents. And each year we have seen these students turn us down -- and yes, most frequently for Princeton!

OK, I admit that Princeton, and Michigan, and Cornell, and the other schools listed above ALSO have great research opportunities. But given similar research prospects, the substantial commitment from some universities to offer real support for childcare may play an important role in graduate school choice. And I can guarantee that whether or not your Dean thinks helping parents is "the right thing to do", he or she is DEFINITELY interested in offering the most effective incentives to recruit the very best students.