Work Is Work

Yeah, that title oversimplifies things a bit, but, as promised, here’s my post on why my treatise on the 40 hour work week probably does apply to success at all levels of science, even at the R1.

The basic premise is this: success at an R1 is not based on being more intelligent, or a harder worker, compared with success at any other level of academic (or, indeed, nonacademic science). Academic science is what I know, so I’m going to stick with that. I hold the firm conviction, based on experience and data (if you want lots of links, please go read this blog post, not one of mine, but absolutely excellent), that the 40-hour work week is coming close to a hard limit for most people, in terms of their ability to be productive. The vast majority of people appear to have a work limit of 45 hours or less. Beyond that, you’re less productive, you start making mistakes, and you start hating your life. (That last bit is my add-on, based on observing people in labs where they’re expected to work 70-hour weeks.)

Really simplistically, let’s say you have a personal work limit of 45 hours. You keep working, and you make a mistake. Let’s say that mistake requires you to do a five-hour experiment over. Well, then, you might as well have stopped at 35 hours, given that you’ve not only wasted time, you’ve cost yourself some.

Now, I concede that bell curves being what they are (bad stand-ins for human variability, mostly), there will be the occasional person with a much longer work limit. Even people in the 70-hour range. But I don’t for a second delude myself into thinking that this particular exception applies to all scientists. Nope, not even the ones who manage to “make it” at an R1.

We have managed (“we” here refers to academic science, but could also refer to American Puritanical Work Until You Die culture) to make ourselves believe that science is a calling, and one that requires your entire being, such that you should not only expect to put in 70 hour weeks, but you should LOVE it. I had one advisor tell me this (the only one who ever said anything like that to me): you can’t do science unless you love it, because it’s hard, and it’s 24/7.

Well, bullshit. For a start, it’s a job. It’s a job you quite possibly do love, and that makes you extraordinarily fortunate. But it is a job. It’s work. Love it or not, you will come to hate it if it’s all you do. (Again, there will be exceptions. Isaac Asimov, who was somewhat famous for his amazing work ethic, used to start writing at 7:30 each morning, and not stop until 10:30 at night, except to eat, get the mail, have lunch with colleagues, visit with family, etc. All of those things were described, by him, as “resented distractions”. He appears to have absolutely loved his work to the exclusion of pretty much everything else. But you’re probably not Isaac Asimov.)

So, it’s a job. But it’s a job we’ve been indoctrinated to think requires you to be at it 24/7. Actual time in lab, for people who have recently complained to me of this, and for those I personally observed in three R1 labs over the course of my own career (including myself until I burned out in my fifth year of grad school): 10 hour days are common, 7 days a week. 70 hours. Every week. For years.

No way is that sustainable. No way are you productive for what may even be half the time you’re at work. That is just wasted time you could be using to read a book, or go to the beach, or to the movies, or take your children to the park, or have sex, or sleep. (Sleep is wonderful. I have two small children, so I didn’t sleep for most of three years. Sleep is now one of my favorite things to do.)

But people buy it. One of the common refrains I hear when I say that you can, indeed, succeed on 40 hours a week is along the lines of, “I believe there are some people who can do it, but they’re rare.” Usually something about brilliance is added here.

All of the above is to get me to this: success at an R1 is a crapshoot. It requires four things: brains, hard work, connections, and luck. Two of those (brains and work) are common to pretty much everyone at every level of academic science. The “connections” aspect is probably stronger among those at R1s. But it doesn’t require that 70-hour week to achieve. The last, luck…is just luck.

Let’s start with publishing. To get that R1 job, you probably do need to publish in Cell, Science, or Nature (CSN). Okay, great. But I’m here to tell you that publishing in CSN is more about luck and who you know that it is about good science. The exact same story might get desk rejected by one editor for not being “sexy” enough, while another loves it. The editor you send it to, if you have any control over it at all, will hopefully be someone you know. Unfair, but true. Connections. The reviewers you get may all be Reviewer 3, or they may be reasonable people. Luck. (Also, possibly connections if they know and respect you.) Ultimately, whether that’s a CSN paper or the next tier down isn’t about how good your science is or how many hours you put in to generate the data.

Next up: funding. Holy shit, such a crapshoot. It’s heavily weighted towards it being easier to get money if you already have money, and it’s also really biased towards people with good “pedigree”. That is, who trained you? Connections. Whether your study section scores you high enough to get funded, given the ever-decreasing percentage of successful grants (okay, at NSF this may not be true. Seems stable recently) is also about luck. Yes, some will be amazing grants, top of the heap. Everything closer to the funding line…really seems to depend on who reviews it. Ask someone who’s resubmitted the same grant after addressing the reviewers’ concerns and gotten a LOWER score with the new panel about that.

So you get your R1 job by getting lucky enough, and knowing enough people, to publish in high-impact journals. You keep it (read: get tenure) by continuing to publish in high-impact journals and getting funded. You have to be lucky, and you have to know the right people.

That’s not hard work. That’s networking and luck. It doesn’t take 70 hours a week.

The work is the same. A qPCR in my lab is the same as a qPCR in a lab at Harvard (actually, it’s harder and probably takes more time, not less, in my lab. Because undergrads and limited resources/infrastructure).The amount of data required to put together a good story (for a reasonable, not-insane journal) is about the same. Writing a grant is EXACTLY the same. I send my grants to the same study sections people at R1s do, only my PUI grants tend to have a lower funding rate. (There are a few PUI- and minority-serving-specific pools, but the amount of money in them is smaller, so there’s no advantage there.)

Yes, I write fewer grants, because I run a somewhat scaled-down lab, AND because I don’t pay most of my people. Which is a bummer. But, here’s the thing: I make up for it with teaching. I guarantee you most faculty with R01 grants don’t teach a 3-course load. I guarantee they aren’t teaching 20 hours a week.

So, since this all started as a discussion of postdoc hours, let me revisit the numbers I was playing with on Twitter:

In my (undergraduate, so half time at best except in summer) lab, we shoot for each person getting one result a week. There should always be an experiment coming out, one being set up, and, ideally, one underway that isn’t done. The papers I write tend to have about 15 panels. Four to five figures, 3-4 panels per figure. So, that’s 15 weeks. But only about 1/3 of experiments result in publishable data. Make that 45 weeks. Theoretically, if 1/3 of your experiments work, and you have just one result per week, one person can generate data for a paper in a year. Now, you do have to repeat the experiments, so figure at least twice that long.

So, that’s my lab. One competent undergrad or MS student can generate a paper in two years. These folks work about half time. Less during classes, often. Assuming you are a Ph.D. student whose classes are done, or a postdoc working full-time, I’d think you could do it in a year. Two results a week. Maybe not a CSN paper, since those damned things tend to be padded with every experiment and the kitchen sink just to try and avoid the unavoidable year of revisions, but a good paper nonetheless. (By the way, if more than 2/3 of your experiments don’t work, it might be worth asking yourself if you’re working too hard and making mistakes because of it.)

So. Are you telling me that you really don’t believe that a person can generate two results a week in 40 hours, with time left over to read, write, etc.? Because, again, I’m not buying it. I’ve seen it done. If you want to see it done, look to the women in your R1 labs. The ones with kids, especially. They’re doing it because they have to. And they’re often getting shit about their “dedication to science” over it, when they’re the most productive people in your lab, by the hour.

If you say you don’t think it can be done, I want you to step back and ask yourself: have you tried? Has ANYONE you know tried, or has everyone assumed it can’t be done? I say you don’t know. Because we were all too scared to see what would happen if we stopped being in lab 10 hours a day, 7 days a week.

Work is work. It’s not harder, or even much MORE, at an R1. People just want to believe that they can control the uncontrollable by working harder at it.


1 Comment

Filed under Uncategorized

One response to “Work Is Work

  1. Polebean

    I agree with what you are saying, but I think, for some people and some lines of research, it’s not about working hard 70 hours a week that makes them more productive, it’s being available to do anything that needs to be done at odd hours, which usually means 70 hours a week in lab. However, it’s balanced by a lot of procrastination and mindless web surfing while in lab, which seems to prevent burnout. These people don’t work harder, but they can get more results over time.

    How? Well for instance, sometimes you need to do a 10-hour timepoint. If you work 9-5 you’ll find it hard to squeeze that into your schedule. If you’re used to hanging out in lab all the time, you can do it anytime. You might not do any work in those 10 hours, but at least you’ll be there to take care of the sample when the buzzer finally goes off. If you’re not usually available for these odd timepoints, the postponed experiments tend to add up, and it leads to slower progress.

    So I don’t believe that working hard for 70 hours a week is productive in the long run, but being able to be “on call” does seem to have its benefits. Again, depending on the types of experiments your research demands.

    Obviously, being on call is easy to do when you’re younger and single, not so much when you have family.

Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s