Wednesday, August 29, 2012

The Case for Independent Scholarship #3: On Workaholic Scientists

So, Sam Arbesman has a post up at Wired where he discusses a recent study on the work habits of  scientists in around the world. As a proxy for "working," the authors look at the pattern of downloads of papers or book chapters from Springer. The work makes use of a cool real-time mapping of IP addresses accessing those papers. If you want to see what it looks like, check it out here:

They do a more detailed analysis of the top three countries (in terms of total number of downloads in about a week's worth of data), the US, Germany, and China. Correcting for time zone differences, they find these patterns:

On the left side, each line corresponds to a different day, with the more solid lines being weekends, and the thinner ones being weekdays. On the right, the weekends and weekdays are averaged.

A couple of things that will probably come as no surprise to most of the academics out there. 

(1) the daily hump starts picking up around 9 or 10 in the morning, and carries on until nine or ten at night.

(2) the weekday hump is bigger during traditional "working hours", but evening work hours are pretty consistent throughout the week.

Interesting cultural differences pop out that might not be as predictable. It looks like China has longer and/or more simultaneous breaks for lunch and dinner. (Although, given the common practice, at least in the US, of eating lunch at your computer, maybe the lack of those dips in the top panel are somewhat predictable.) Americans seem to be working a lot more in the middle of the night compared with their German and Chinese counterparts, while the Chinese seem to show less of a difference between weekday and weekend work habits.

The authors' conclusion, and one that is echoed in Sam Arbesman's commentary, is that this work pattern is consistent with what most academic scientists would probably tell you. Academia is a full-time job. And not a full-time job in the sense of a forty-hour work week, but a full time job as in, you sleep, eat, and work.

So is that a good thing or a bad thing? Well, on the one hand, you've got all of these highly trained, highly educated people working really hard and getting paid not a whole awful lot on a per-hour basis. Good deal, right?

On the other hand, it leads to a really crappy lifestyle, where the long hours come at the expense of time spent with family, hobbies, or even taking an interest in subjects outside of the very narrow range defined by your research. If you care about a broader definition of human happiness,  one that treats people as an end rather than a means, this is not a great way to structure your industry. 

Beyond that, it is important to remember that science, like all academic fields, is a fundamentally creative enterprise, and working longer hours does not necessarily translate into better results. Creativity has to be fueled by experience, and a broader range of experience can lead to asking more interesting questions and coming up with more original answers to those questions. The pressures that lead people to download papers from Springer from morning till night don't necessarily lead to the best science.

I should note that Sam's coverage also includes a plug for the Ronin Institute, because Sam is a freakin' rock star!

Wang, X. W., Xu, S. M., Peng, L., Wang, Z., Wang, C. L., Zhang, C. B., & Wang, X. B. (2102). Exploring Scientists’ Working Timetable: Do Scientists Often Work Overtime? Journal of Informetrics, 6 (4), 655-660 DOI: 10.1016/j.joi.2012.07.003


  1. This was a significant factor in my decision to leave academia. As a postdoc, I already felt very busy yet not really very productive in any sense that mattered to me, and I could see it would get worse if I became a professor, what with adding teaching and administration to my responsibilities.

  2. While I think the original data is interesting - this inference that an unusually wide spread of working hours across the scientist population reflects unusually high working hours, doesn't seem particularly tenable. While it may well be the case, it could equally be the case that scientists have a lot of flexibility, and vary their working hours more than most.

  3. Charles, I think you're absolutely right. This data really has no bearing on the issue of whether or not scientists work long hours. It also sets aside the question of how hard they're working.

    I think the part of the data that is actually interesting from a scientific point of view is the pattern of cross-cultural differences, although, again, the data as it stands is open to multiple interpretations.