How much does choice of PhD typically matter for a future career in academia?

Question

I have obtained multiple bachelor's degrees and a master's degree. I'm very motivated to pursue an academic career, for which obtaining a PhD seems like a necessity. However, I find it hard to decide on what kind of PhD positions I should be applying to. I often feel like a subject would be interesting enough to work on for multiple years and fits my master's degree quite well, but I really want to be moving in a different direction (specifically, I have two bachelor's, in mathematics and in computer science, and a master's in AI; after my master's I sadly have to conclude that I prefer topics on the border of computer science and mathematics rather than most of what's considered AI nowadays).

I'm having very little success in pursuing those PhD positions that look the most interesting to me. I feel like I would have an easier time if I broadened my search to include any subject I think I could keep up for 4 years, but I worry that this would 'lock me in' to this area of research.

Are my feelings justified given the way things work in academia and would it be justified to say that wanting to pursue an academic career I should not be applying to anything that straightforwardly doesn't fit that career path? Should I be happy to get into any PhD program at all, and accept that my options are rather limited? Or am I worrying needlessly, and would it not be at all unrealistic to make major lateral moves within my academic career, towards mathematics?

Answer 1

Be very selective about your choice of advisor, as much as you can, and then about your choice of field.

I say this because you sound like you are more interested in chasing the research frontier, rather than settling down with a stable technique and deciding how to apply it. At the frontier entire fields emerge and evaporate on the timescale of a PhD. The field you go for when you graduate might not even exist when you enrol, making it impossible for you to study a PhD in it!

That's why you should optimise for your advisor. Their friendship and guidance will be the highlight (or lowlight) of your PhD. A good advisor will mentor you in productivity, wellbeing, networking, multitasking, working with students and admin, finding jobs and grants, and putting your distinctive touch on your research style. They can open doors for you and give you objective feedback. And if you find yourself in a good and healthy research group, you will immediately have a network of future collaborators to give you a leg up.

That's pivotal in choosing where to do your PhD, and you should work as hard as you can to find a good advisor. If it means picking a field you're a little less excited about, so be it.

Answer 2

A metaphor: The dinosaurs were highly specialized and the "fit" for their world was superb. Hence, they grew very large and even more specialized. Then, one day, the world dramatically changed and ...

If you choose today, to specialize for the hottest possible topic today, knowing that your career will only begin in several years, say about seven, then you may find that the world has dramatically changed in the interim. People may be looking in other directions, but also, in the interim period, many (many) other people have joined that hot topic and you may find the zone completely flooded. "We'd love to hire you but we don't expect any openings for several years. Sorry."

This happened to me, actually, and I found a very difficult entry into academia, even in mathematics as a whole, because of economic conditions that changed academia and the possibility of hiring new faculty at any level. The world changed over the course of about two years from a hot hot zone to an empty wasteland (In my case, it was the loss of science funding after the US landed on the moon - a loss no one predicted.)

At some point you will need to specialize to get a doctorate, but I suggest that you don't do that too early, or, at least, that you keep open a lot of options. And, stay open to other options as they arise. And, all the better if the skills you get, even while specializing, are applicable in a wider domain.

At the moment, the STEM fields are better bets than the humanities. Both Math and CS are good choices. That can be expected (predicted) to continue. But, AI may be in a passing phase for its "heat index". Only a very small part of AI as a whole is in the news now. That might pass. The important problems there might get solved in the next few years with lots of people looking at them. Or not. It might open more. It might close. The issues with such things as ChatGPT could, perhaps, renew interest in the humanities. For myself, I'm not looking forward to the day when news is generated by amoral, brain-dead robots, whether you call them intelligent or not.

I find it wiser, for most people, to choose a path that they will enjoy following even if it doesn't necessarily lead to fame and riches at the end. If you love math, then do math, though it may lead you to a job that isn't what it might have been "if only...". But the person I describe here may not be you. Good luck then. I hope you make it.

If you try to make choices that lead to a situation that is all (Full Professor at Harvard or MIT) or nothing, then you might get lucky (and it will take luck beyond your control) or end with nothing. Many people can build a fulfilling life teaching math or CS at a liberal arts college, and it is more likely that there will continue to be demand for that than for specific sorts of high-octane research in a narrow field.

Prediction is hard. Especially about the future.

Answer 3

Ideally, I would tell you to follow your passion. Instead, I suggest that you follow your passion while also looking at what you might do when you are done.

More realistically, I would submit applications to applied fields that interest me. I give this advice for two reasons. First, academic jobs are hard to get. This will give you a good backup career outside of academia. Second, there is less competition because people have other career options outside of academia. Thus, all the graduates from your program will not be competing for a limited number of positions.

For example, pick engineering over (pure) physics; statistics/data science/applied mathematics over pure math; fisheries science over biology; or clinical psychology over academic psychology.

Look at where the alumni from the programs you are looking at end up. They are the best predictors of your options when you graduate from a program.

Answer 4

You are making a good point and you are asking the right questions. However, instead of answering them for you, I'd rather provide some insight based on my personal experience that I think are in line with your concerns because, after all, you are facing a very important decision and in the end, you will have to make a decision yourself for which you will be the one who will face the (hopefully pleasant) consequences.

Knowing that you have already made your mind to pursue a career in academia, I can assume that you already know that research can be hard and also, I assume that you are already familiar with all the things that could be or are wrong/unpleasant in/with academia. To name a few, I could point out to the publish or perish atmosphere, the limited number of opportunities, high-level of competition and etc. So, in order to make your decision, I would recommend to keep these factors in mind as well. Because, starting a PhD is going to be just the beginning of the journey, and for most part of the journey, you will be on your own. Therefore, because of this, I'd suggest to keep the following two points in mind too, among others:

1. Select a field/subject that you truly enjoy thinking about and working on. As I said, doing research, both during and after finishing your PhD, is going to be hard. There will be so many long hours of feeling lost, getting bad results, frustrations and making little, if any, progress. And to make things worse, for the majority of times, there will be little help from anyone else. This is mostly because, doing research means working on a problem that has not been solved before. So naturally, there are not many people who even may have heard about it. Even if you are lucky to be with a very supportive supervisor, the chances are that they will not be of great help when it comes to solving the problem except providing moral support and pointing you to some of their previous research that may be of relevance to your work. Therefore, research can be hard as it is even if you love your topic but it will become easily unbearable if you do not.
2. Select a field/subject that you know you can be prolific at. This point comes after the previous one but nevertheless, it is very important because, to be successful in the academia, you will need to be prolific. As I mentioned earlier, in the academia the opportunities are limited and it is still suffering from the publish or perish culture. Therefore, even if you enjoy your research a lot but do not produce results (the kind of results that academia deems acceptable), you will have a difficult time finding post-doc positions, securing a tenure-track position and receiving grants.

Answer 5

There already has been quite a bit of advice, but as an addendum to all of the other posts about the primary importance being your choice of advisor, I would say to consider what you plan on doing post-PhD. and make sure your decisions map onto something that will contribute towards that. Namely, I would consider:

1. Do you want to work in industry?
2. If you stay in academia, do you want to be more teaching-oriented or research-oriented (and from 100% in one direction and 100% in the other, what do you feel would be the best balance)?
3. If you stay in academia, would you want to work in a CS department or a math department?

As others mentioned, some advisors will be better at some of these than the others. If you want to be research-focused and go to an R1, you should get a rockstar advisor with a track record of getting students into R1s. If you want to be more teaching-oriented, you should have an advisor who isn't clingy and lets you do some education research on the side (and select an institution that lets PhD students teach courses). And if you want to go to industry, you should get an advisor who will be happy to let you go to industry internships each summer (and ideally a track record of students going on to industry).

However, all else being equal (ie: having two advisors who both are good for your goals without anything strong pushing you towards one or the other), I would in general suggest going for the CS degree. In general, CS is "easier" to move into industry, and also easier to get an academic job (whether it is at an R1, a PUI, or any other type of institution).

However, it is generally rare for CS degrees to be in a math department, while the other direction isn't that uncommon. So if for whatever reason you are set out on being in a math department, then the tiebreaker might go to the math option. But otherwise, the tiebreaker should be towards CS.