AAS 219 Twitter Review: The Postdoc Job Market

by Jessica Lu on January 19, 2012

Continuing with my twitter review of AAS 219 sessions:

The Astrophysics Postdoc Job Market – AAS Employment Committee

This was a panel discussion where panel members included heads of postdoc fellowship programs and successful ex-postdocs who have moved on to faculty, research staff, industry, and education positions. Overall the discussion was interesting but rather open-ended. Here are my tweets, with some additional explanation in a couple of cases.

  • The object of your training is not necessarily to turn you into a faculty member. Be flexible.
  • Consider education, management, policy… you can often have a much bigger impact.
  • Postdoc period is a time of stress and transition, but also a time of learning and searching for your place.
  • AAS Survey 2000-2006, 270 new postdocs available each year and NASA named fellowships 10% of those.
  • 97% of NASA named fellows are still in the field, 85% have permanent positions (not necessarily faculty). — These stats are 2000-2006 before the economic downturn.
  • Our training is incomplete w.r.t. knowing how to interact with people (non-specialists).
  • Research shouldn’t be 100% of your time. Maybe organize a seminar, teach a class, something else.
  • You don’t need a named fellowship for an R1 faculty position. You do need to demonstrate the whole package.
  • Carve out time to demonstrate your independence and the skills that faculty already exhibit.
  • Faculty search committees ask “Are you a match to our department?”
  • The number of postdoc positions has tripled in the last decade.
  • The number of tenure track faculty positions in the last decade has not changed.
  • “Do not assign yourself a flat prior to the probability you will get a faculty job.” – J. Johnson
  • There is the possibility of exploitation in hard economic times. We as postdocs should be aware.
  • There is some precedence for NSF/NASA to incentivize permanent positions. This is highly debated.
  • One suggestion to provide bridge funding for new faculty lines during bad times when faculty aren’t retiring.
  • There is a gigantic information asymmetry when applying for faculty jobs… the rumor mill helps but isn’t perfect.
  • Beware, there can be false information on the rumor mill and self-serving behavior.

If you attended the session, what were your questions and take-away messages?

{ 12 comments… read them below or add one }

1 D January 19, 2012 at 1:34 pm

“Do not assign yourself a flat prior to the probability you will get a faculty job.”
v.
“There is a gigantic information asymmetry when applying for faculty jobs.”

Yeah. Apart from the handful of people who know they’re awesome because they were offered 10 different fellowships, the rest of us don’t have much choice other than to assume we’re “average” and take an uninformed flat prior. If we had all the information we could assign better probabilities, but the whole point of Bayesian analysis is to assign belief probabilities in the face of limited information. I don’t know who J. Johnson is, but either they don’t understand Bayesian analysis, or they don’t understand the job market.

Reply

2 JP January 23, 2012 at 7:51 am

Although there’s a lot of information asymmetry in the process, there’s also a lot of information you can use to work out a prior. If you’ve applied for and not been awarded a named fellowship that immediately tells you that your odds of getting a permanent job have fallen (although plenty of Hubble/Einstein/etc. go to people applying for second postdocs). If you apply to give talks at conferences, but no one wants to hear you then again your odds slip. If you’re working in a huge project, but aren’t the go to person of your little niche, then you’ve got problems. At the end of the day, the really hard thing is to look critically at how you’re doing and compare it to those around you and those you see on the rumour mill getting jobs. To be “average” has to be a killer in getting a faculty job, because the attrition rate is so high.

On a side note, I like the idea of a centralised application system so that the committees could just check a box to send an email with info like “rejected on first cut of all but 20 out of 200 applicants”, “made long list of 20″, “made short list of 5″, “offer” or the like. With >100 applications per position, its unlikely any but the final short list and those that ask specifically are going to get feedback.

3 J January 19, 2012 at 6:36 pm

I have two suggestions with regard to the “information asymmetry” mentioned during the panel.

1. Encourage committees to actually send rejections after they have met.
At this point, there are more fellowships that I have de facto rejections from (i.e., I know others have gotten interviews or offers) than have actually contacted me personally to let me know. It is *extremely* disheartening to know that committees have made their decisions and can’t be bothered to notify those who have not been chosen. I would rather just know and move on with my life, rather than find out second hand from the Rumor Mill.

2. Encourage committees to give some feedback.
I have gotten a few rejections, but very little indication of why I did not make the cut. It would be extremely helpful to know where I fell on the spectrum. Was I in the bottom 50%? Or did I almost make the short list? It would be nice to have some sense of what impression I am making to those who are deciding.

Reply

4 Ann Onymous January 20, 2012 at 6:27 pm

I cannot agree more here. For postdocs and for permanent positions. The lack of information makes things even more crushing for the morale. I am getting strings of rejections for permanent positions as of late (officially or inferred from the rumors mill) and I do not know why. Looking at the list of shortlisted people some have an outstanding CV but some are just normally skilled postdocs like most of us. I do not know what I am doing wrong, or what they are doing better to get shortlisted. Each department has different criteria but at least if we got feedback we could see whether there is a pattern or not.

5 Beth Onymous January 27, 2012 at 12:30 pm

As someone who is on the other side, i.e. hiring a postdoc: believe me, things are infinitely more complicated than they seem from the jobseeker side. It would be inappropriate to say that searching for a postdoc is as stressful as searching for a job, but it is stressful. You need to pick someone who you’ll be working with for the next 3 years and unless a superstar you have only one shot. I started with the position that I will hire the best candidate, but it doesn’t work like that. The best candidates take named fellowship so no point in wasting your time. For the people in the next tranche the noise in references and other people’s opinion is amazing, the same candidate will be described by person A as “really underrated really clever guy” and by person B as a “bordering retardation” (actual quotes). Second, usually the better the person the more independent they are and so often it is better to get a candidate who competently does what he is told rather than someone who might be better in terms of raw ability but will just go in some random direction that she/he wants to pursue. Third, ability really is not 1D space. Some people can do calculations, other write code, some are fast, some are slow but very careful, etc. Then you need to think what will the person actually to for the good of the entire group, what are the synergies, who will she/he talk to. This is why being clearly better does not necessarily translate into more offers.
Finally, when making an offer you are not immediately going to send the rejection letters, the person might not take an offer and the second person might not take an offer either. Finally, one gets 80 applications for a typical job, simply collating these emails and sending a rejection letters is a one hour job that frankly most can’t be bothered to do. But I agree that maybe we should be doing it more promptly. Finally, if you think you really should be shortlisted for a position and you weren’t (i.e. it wasn’t a long shot), I am pretty sure you will get a semi-sincere answer if you just pose this question at someone in the committee.

Reply

6 nick January 27, 2012 at 4:47 pm

“Second, usually the better the person the more independent they are and so often it is better to get a candidate who competently does what he is told rather than someone who might be better in terms of raw ability but will just go in some random direction that she/he wants to pursue.”
So, cut loose the ‘better’ people and string along someone who’s less independent for three more years? This is a failed system.

“one gets 80 applications for a typical job, simply collating these emails and sending a rejection letters is a one hour job that frankly most can’t be bothered to do”
Frankly, that’s a lot of human misery to buy that hour of time.

7 Ann Onymous January 27, 2012 at 7:27 pm

“Finally, if you think you really should be shortlisted for a position and you weren’t (i.e. it wasn’t a long shot), I am pretty sure you will get a semi-sincere answer if you just pose this question at someone in the committee.“

I did that in the past when I was looking for postdocs and I do not think I ever managed to get any useful answer. The most detailed answer was generally something like “your application generated a lot of interest but we had a lot of outstanding applications”. It is polite but useless. More generally it was something like “Sorry, we had so many applications, I cannot give you details about yours.” I wanted people to tell me what they thought were my weaknesses. So in the end I just stopped bothering people about that. Perhaps I should try again for permanent positions as I am really completely in the dark as to what is wrong with my application. I can only infer it is not the scientific output as I am not worse than the mean shortlistee i see on the wiki.

8 Ann Onymous January 27, 2012 at 7:27 pm

“Finally, if you think you really should be shortlisted for a position and you weren’t (i.e. it wasn’t a long shot), I am pretty sure you will get a semi-sincere answer if you just pose this question at someone in the committee.“

I did that in the past when I was looking for postdocs and I do not think I ever managed to get any useful answer. The most detailed answer was generally something like “your application generated a lot of interest but we had a lot of outstanding applications”. It is polite but useless. More generally it was something like “Sorry, we had so many applications, I cannot give you details about yours.” I wanted people to tell me what they thought were my weaknesses. So in the end I just stopped bothering people about that. Perhaps I should try again for permanent positions as I am really completely in the dark as to what is wrong with my application.

9 Beth Onymous January 27, 2012 at 11:32 pm

” “Second, usually the better the person the more independent they are and so often it is better to get a candidate who competently does what he is told rather than someone who might be better in terms of raw ability but will just go in some random direction that she/he wants to pursue.” ”
“So, cut loose the ‘better’ people and string along someone who’s less independent for three more years? This is a failed system. ”

Well, the people who are independent and good do get named fellowships. But as a young faculty building your “grove”, you get grants to do certain things and then your task is to get someone who is actually interested in doing those particular things. This means that someone with better raw capability might just not make the cut if s/he is much more interested in doing something else. The system is broken, but at the level of overproducing PhDs, rather than PIs seeking people for their projects.

“Perhaps I should try again for permanent positions as I am really completely in the dark as to what is wrong with my application.”

Well, not sure how useful this is, but when I was a postdoc I was really stressing over whether the application is in in time, if they got everything, if the references were sent, if there were stupid typos in the letter, etc. – all these things matter at the 3rd order. Nobody is going to decide whether to give or not give you a job based on a mispelling or one missing reference (they can always ask for it), it would be stupid. What matters is who writes the references, how strong they are and how releveant is what you do for the group. References written by people whom the reader doesn’t know are worthless, they are marginalised over. References written recognisable names but distant collaborators are (those that start with “You should not put too much weight to this letter, because I don’t really know the person”) are worth very little. Regarding field: for a named fellowship what matters is “To whom in that group will I talk?” and “What will I taket to them about?” and “How can I fit into this experiment/telescope time/etc that this group has access to.” and not “Do I have more papers than other candidates.” (papers/citations do matter but only if you are a good fit) For postdocs belonging that go from individual grants, you should be asking “What do I and this person have in common and what papers can we be writing together.”

Reply

10 nick January 28, 2012 at 2:35 am

If control rather than collaboration is what your group is after, that’s fine. But if you hire someone because you think they will competently do what they are told and not be imaginative, they need to be made aware early on that they are on their way off of the academic path. Reference letters to the effect of ‘controllable, competent’ are not exactly a ticket to permanence.

11 Mark February 10, 2012 at 4:42 pm

Why not simply invest in very powerful computer workstations, and pursue your own research projects in Astrophysics from your own home. I wonder how many professional Astronomers are actually “self-funded” and don’t even depend on university positions. What about taking an industrial position in research engineering that has a high salary to pay for self-funded research at home?

Reply

12 Shantanu May 25, 2012 at 5:01 am

Mark, I know someone who took a job in biophysics (basically to pay the bills),
but in the evening/nights has been tirelessly working on topic in GR, cosmology, black holes
etc and even has published papers. Unfortunately despite that all his efforts to find
postdoc or permanent jobs have been in vain. (plus of course he has no mentors).

Reply

Leave a Comment

Previous post:

Next post: