Vivek Haldar

Computer Science PhD trends

One of the most under-rated gems in professional computer science is the Taulbee Survey put out every year by the Computing Research Association. It is a treasure trove of statistics, hard data and trends about both the input (enrollment) and output (graduation and employment) of the computing education pipeline. The latest one has data for the 2011–2012 time period.

You could spend hours diving into the data and drawing your own conclusions, but these are some of the things that jumped out at me:

  1. 7% of new PhDs got into tenure-track faculty positions.
  2. 47% of new PhDs went to industry.
  3. 57% of PhD enrollments were nonresident aliens (an all-time high), as were nearly 50% of PhD graduates.

What does this mean? Let’s take them in turn.

  1. The tenure-track faculty position is the top of the academic totem pole and coveted by fresh PhDs. The hard truth is that those positions are small in number and getting smaller, leaving the very distant second choice of post-doc as the only viable option for those who want to stick with academia:

    The data indicates tenure-track faculty positions for new Ph.D.’s have declined steadily since 2004 from 224 to 124 in 2011 while the number of postdoc positions greatly increased. In 2011, new Ph.D. graduates accepted twice as many postdoc positions as tenure-track positions. Around 2003 there were approximately 2.5 times more tenure-track positions than postdoc positions.

    Which brings us to…

  2. At least in computer science, the slack has been picked up in large part by industry. This might have been looked down on in the past (“I have a PhD! I can’t work as a money-grubbing corporate stooge!”), but as I’ve written before, the avante-garde tech companies of today offer work challenging and varied enough (and yes, well-paying enough) to keep nearly half of the graduating crop of PhDs happily employed. I expect this fraction to only grow in the future.

    1 and 2 combined should make every prospective PhD student do a good deal of soul-searching to figure out why exactly they want to go through with it. It is a big commitment, and it is not easy, not just in the intellectual and academic sense, but also psychologically. This doesn’t have to be done before you start. Remember that there is an “out” after about two years when you can leave with a master’s degree. I’ve reached the conclusion that there is no material advantage to doing a PhD, and that the only good reasons to do it are philosophical and deeply personal. I will always remember it as the last still, contemplative and reflective time in my life. A font that forever gives.

  3. To anyone who has walked the halls of a CS department (especially at night), this statistic is far from a surprise. To most of the public, it probably is. Yes, the majority of people in CS PhD programs in the US are foreigners. And the vast majority of them get a full scholarship for the six to seven years that it takes to get a PhD. (They repay part of that by keeping the academic engine churning, but still.) The vast majority of that scholarship money comes from government grants (i.e. taxpayer dollars) from agencies like the National Science Foundation, Department of Homeland Security, Office of Naval Research etc. I am loathe to get into politically charged topics, but given this, it seems downright stupid to make legal residence in the US of those graduating PhDs so hard that a non-trivial fraction of them leave. The logical taxpayer should insist that they stay here and contribute back as productive members of society.