Great post! Agree generally but I think it's worth pointing out too aspects that don't carry over to the US (maybe also to other places but I wouldn't know).
1. The (relative) financial advantage of doing the PhD. Most people doing a PhD in the US (at least in STEM) could make >2x more working instead (>5x more if they went to a good undergrad) and there aren't really tax advantages. Plus, the resume boost is not very large AFAIK unless you're gunning for really specific jobs and those tend to be selective enough that they won't take you if you just scrape by during your PhD.
2. Not having to move cities for the PhD is very rare in the US. Some people stay in the same school but the vast majority go somewhere else and most schools are far away from each other.
Thanks! Great points. In general my piece is focused on the UK. While much of what I wrote is true across the board, I definitely grant that there are key differences in the states and I would consider the US PhD to be significantly harder.
On (1), I think this is true everywhere. I discuss the better financial compensation in London vs a PhD in the UK in my essay, and how this incentivises the best and brightest not to do a PhD. It’s not that there is a financial advantage of doing a PhD over a harder job, it’s that it’s better than being an undergrad and comfortable enough to imbue laziness or a lack of agency.
On (2), great point. I suppose for someone doing their undergrad at Cambridge, there are only two other places you’d want to do your PhD at in the UK (London or Oxford), so staying put is much more common. It’s definitely frowned upon but the same is true for the PhD-postdoc transition in the UK, with many staying where they are. There are more good choices in the US which is a great thing.
In my econ PhD we failed a bunch of people out at year one comp exams and then a few more failed out over the next few years. And those people weren’t dumb. My friends in Doctorates of Musical Arts worked much hard than me. There were other fields that were much less rigorous, but those were in “permanent adjunct or unemployment” studies.
Interesting post. I see some pushback in the comments and I note it comes from US experiences. Which matches the observation I was going to make: that UK and US PhDs are quite different. (I am British, with a British PhD, and have worked at both UK and US institutions.)
I agree with a lot of what you say here, but it is quite UK focused. The US PhD is significantly more demanding, and the modal US PhD graduate is ahead of the modal UK one, in terms of skills, experience, independence, etc.
Yes you’re totally right. Just looking at age gives insight here. Average age of PhD applicants is higher in the US, and US PhDs are longer too, so the age at graduation is *much* higher in the US.
That said, I think this discrepancy might normalise at postdoc, with most UK PhDs probably requiring two postdocs before independence (this is just my hunch, haven’t looked at data).
I still think the case can be made for more rigorous application standards in the US too, and also for better assessment criteria that more closely test “ability to contribute to scientific discovery”.
I think I don't know what it would mean for the discrepancy to normalise, but I do think that a typical new US PhD is roughly similar to a typical UK PhD plus two years of fellowship, by which time there has been important challenges of funding acquisition, a growing publishing arc, teaching experience, and so on.
The qualifications themselves aren't equivalent, despite having the same name. Experienced academics know this and take it into account when e.g. recruiting.
Are you American? I have often said the following: that the standards of US UG degrees are low but PG degrees are very challenging. Which is the same pattern you’re describing but in different words
I wonder if, at least in continental Europe, this has to do with education being largely publicly funded. The first year or two are very difficult because they want to minimize waste.
But the final years are relatively easy because if you've gotten that far they don't want to lose their investment.
"This may sound ridiculous, or ridiculously arrogant, but let me clarify my point: I’m not trying to persuade you of my own intellect, I’m arguing that getting a PhD actually doesn’t require particularly high intellect nor does it require particularly hard work—objectively. Yet the societal perception of a PhD is that of a Herculean or even Sisyphean task (almost the opposite is true, as I will explain)."
Well, I'll give you points for honesty, it DOES sound ridiculous AND arrogant! I am not totally sure of your motives writing this, but I certainly hope it is either tongue-in-cheek, or deliberate rage-bait to go viral and grow your new Substack, because the most charitable thing I can say is your experience is very much *NOT* the norm of STEM PhDs in the United States, and from my professional circle, most of the world.
I am not sure how generalizable your quoted 16% drop-out rate is, because many studies suggest the attrition number in engineering, medicine, and the life sciences is closer to 40-50%. Perhaps half of all doctoral students, who already completed a bachelors +/- masters (a small minority of the population already) are just dummies? IMO, no, Occam's razor would suggest the most likely explanation is PhDs are difficult.
"Having been admitted to the PhD programme, what did I have to do to actually get the degree?
I had to attain adequate termly reports from my supervisor
I had to complete a first year report with a viva voce examination
I had to present my work once in four years at an internal departmental seminar
I had to eventually write my thesis, which doesn’t have a minimum word requirement, and then be examined on it in another viva voce."
If every PhD program was like that, I would agree it would be pretty easy. But that is not really the case! For my own PhD in molecular biology at a not-very-prestigious school in the southeastern US, I had to:
Complete 2 full years of additional courses in molecular biology, oncology and statistics (even though I already had another doctorate), some of which was quite rigorous in terms of not only the content, but also the tests and papers required (this was also in the pre-ChatGPT era)
Present my work *annually* at the school, and strongly encouraged at national conferences
Pass preliminary oral AND written exams by committee before beginning dissertation work (I agree that most people pass their prelims, but that is not the rate limiting step of PhDs anywhere, the tough part is the research)
Conduct novel research that would lead to at least two peer-reviewed publications (and ideally more)
I had to write and submit multiple extramural grants. The proposal did not have to be successfully funded, but submission was a requirement
Write up research results in a dissertation (mine was around 200 pages) and defend in front of committee
What makes a PhD difficult, and indeed more challenging than my STEM bachelor's or other doctorate in vetmed, is that nothing is guaranteed. You can end up in a nightmare project that implodes, and your PhD can be over. Some people can eke out a PhD with negative results; most can't. You can have a PI who never stops asking for more experiments, more papers, and other hoops, delaying you endlessly. You can run out of funding. Members of your committee can stonewall you for political or personal reasons (not the most common, but it happens). In contrast, pretty much every other degree out there (including med school!) follows a linear course, and if you check all the boxes and pass the tests, congrats, you get the diploma. Not the case for PhD.
If you want to make the case for more uniformity in graduate training, I'm all for it. But please, don't go shitting on thousands of people's careers just because you had a somewhat easier path. Thanks!
Hi Eric. It seems I have ruffled your feathers, and I apologise for that. It’s also great to get your insight on your experience, thank you for that.
First thing to note is that my experience is representative of the UK, and not the US, which I note in the essay. I think it is absolutely true that US PhDs are harder.
The attrition rate says different things to you and me. To you, a high attrition rate = difficult. To me, a high attrition rate = a lack of resilience. To me, this means the application process could have been more rigorous to prevent these drop outs from wasting years of their lives. I suspect we’ll disagree here and that’s fine.
It seems you do largely agree with me but want to disagree because you found my writing to be offensive to the prestige of your PhD. For example, you agree that I had easy requirements. Would you therefore at least agree with my call for *UK* PhDs to be harder? And if so, then ask yourself if you perceive someone with a PhD from Cambridge or Oxford to be inferior or superior to someone with a PhD from Northeastern Illinois University? And if you perceive Oxford to be superior, then you’ve realised the point of my essay, thank you.
I’m interested in your requirements for degree completion. Was two first author publications a definitive requirement? As in, you fail if you do not achieve this? Writing a grant seems like a great addition and would be welcome in the UK (for my PhD I also had to do this actually to get my Wellcome Trust funding, but I didn’t include it since it’s separate to the PhD itself).
You say “the tough part is the research” and go on to say how what makes a PhD challenging is the uncertainty of scientific research. I totally agree with you!!! I hope I made that clear in the essay (I think I did?!) the toughest part IS the research. NOT the parts that get assessed for the degree (caveat in the differences we’ve discussed for you and me). To boil it down, I’m saying: research is bloody hard and I had a REALLY hard time in my PhD, but I didn’t have to for the degree alone. I chose to because I wanted to make discoveries and contribute to science. All I’m saying is that the ways in which PhDs are assessed are quite unrelated to “contribution to science” and I wish they were more closely aligned especially because “contribution to science” is so difficult, unlike the degree of a PhD.
Hopefully I’ve clarified my position here. I’m not at all shitting on anyone’s career (why would I shit on my own career?!). People who worked really hard during their PhDs (sounds like you did) should find themselves agreeing with my call to make such hard work more appropriately assessed for the degree itself.
Thank you for engaging with my comment instead of reflexive dismissal. You correctly observed that my feathers are ruffled, but your diagnosis is wrong. It is not about "prestige"; I don't even list my PhD in my Substack by-line or bio (and I didn't mention my undergrad institution, but it is prestigious). And I would agree that simply having a PhD does not guarantee one smart.
So why do I consider my PhD one of my proudest accomplishments? Because it took an *ENORMOUS* amount of work over 6 years (while simultaneously doing a medical residency) and contributed to the literature in my tiny field. Then some bloke from the UK who got his degree 5 minutes ago comes along and fires off a post telling the world that PhDs are a joke and barely any work at all. So yeah, color me offended. However, what is most upsetting about this isn't even personal; at a time where public faith in scientists has significantly declined (see reference at the end), the LAST thing we need is misleading media that can justify their prejudice. I don't know if you follow the news much, but in the US we have a hostile new administration that is absolutely gutting what was formerly the crown jewel of biomedical research in the world (the NIH, NSF, and American research university system). Those of us who work in science need to be careful with how our words come across to the public, lest they be coopted and twisted against it.
Side note: You asked about my perception of Cambridge or Oxford. I don't care about "prestige," some of the best research training programs in the US are at public schools like the University of Wisconsin-Madison or the University of Washington in Seattle. Prior to this post, I had no strong views on Cambridge either way, but I have to admit after reading your article, my respect decreased quite a bit.
"I’m interested in your requirements for degree completion. Was two first author publications a definitive requirement? As in, you fail if you do not achieve this?"
The specifics of each program are highly dependent on one's PI and committee. My major professor told me that before he would let me defend my dissertation I needed at least two peer-reviewed publications. I don't know if you would consider not meeting that "failing," but indefinite limbo is essentially the same thing on a long enough timeline (see discussion below).
"The attrition rate says different things to you and me. To you, a high attrition rate = difficult. To me, a high attrition rate = a lack of resilience. To me, this means the application process could have been more rigorous to prevent these drop outs from wasting years of their lives. I suspect we’ll disagree here and that’s fine."
Strong disagree. There are so many reasons someone may have to drop out of a PhD program that do not fit with "lack of resilience." In the US, students often take on massive debt to attend university, and graduate students are paid below minimum wage for many, many years. Grad students also have crappy health insurance (some may totally lack it), and we do not have single-payer healthcare to fall back on. Does someone who cannot handle the financial strain of that "lack resilience"? Several of my colleagues in grad school had to quit because they couldn't make ends meet after so many years. Others worked multiple jobs. I myself drove to another state to work ER shifts to make extra $.
What about someone who is working on a cutting edge project that fails? (Incidentally, the severe consequences of losing funding and/or grad students leads to significant risk averse behavior, and is a major reason why so many projects are incremental instead of groundbreaking innovation). What if their hypothesis IS right, but they are scooped by another research group, and their work is no longer publishable or fundable to the same degree?
What about the many female grad students who have to grapple with sexual harassment from superiors? Or who get pregnant during their studies and don't have accommodating labs? Even if they do, it can delay them by years, and you circle back to the financial strain argument, now with an extra mouth to feed and clothe.
Look, I don't know you, so I don't want to be too harsh, but at minimum, a straight white male who casually bragged they went to undergrad at Cambridge and didn't even have to move for their PhD program seems like someone who has lived a fairly privileged life and may not have contemplated a lot of these challenges. I don't have a study backing this up offhand, but in my experience, the vast majority of grad students do their PhD at a different institution than BS/masters, which necessitates moves, expenses, etc.
"You say “the tough part is the research” and go on to say how what makes a PhD challenging is the uncertainty of scientific research. I totally agree with you!!! I hope I made that clear in the essay (I think I did?!) the toughest part IS the research. NOT the parts that get assessed for the degree (caveat in the differences we’ve discussed for you and me). To boil it down, I’m saying: research is bloody hard and I had a REALLY hard time in my PhD, but I didn’t have to for the degree alone. [...] All I’m saying is that the ways in which PhDs are assessed are quite unrelated to “contribution to science” and I wish they were more closely aligned especially because “contribution to science” is so difficult, unlike the degree of a PhD."
It sounds like PhDs in the UK essentially don't even assess the quantity and quality of a grad student's research. If that characterization is accurate, I am shocked, and feel like the degree should almost be called something else than "PhD," since it has such a different connotation in North America, parts of the EU, and other places around the world. This is likely why tenure-track faculty positions (which are becoming rarer by the year!) require multiple post-docs, since you can't rely on the quality of one's doctoral training.
A. Lupia, D.B. Allison, K.H. Jamieson, J. Heimberg, M. Skipper, & S.M. Wolf, Trends in US public confidence in science and opportunities for progress, Proc. Natl. Acad. Sci. U.S.A. 121 (11) e2319488121, https://doi.org/10.1073/pnas.2319488121 (2024).
Lots I could respond to here. Lot's of assumptions in your reply about who I am and what I know (and don't know). Let me limit myself to a point of agreement:
You point out how it was your supervisor who imposed the stringent two-paper requirement. Many other commenters point out how a supervisor can dictate the difficulty of a PhD. I literally write about that in the essay and don't deny it. I agree that this isn't good. I would perhaps posit that these faculty members are responding to the fact that university-levied requirements are insufficient to drive scientific discovery. I'm arguing for all PhD-granting institutions to make requirements more stringent such that there is less variability between supervisors and the letters "PhD" better reflect actual contribution to science, which supervisors like yours clearly insisted upon on their own accord.
We lost more than fifty percent of my cohort of PhDs before they reached candidacy. I don’t know how many after that rang the bell. They fried us up and ate us for lunch. They beat us within an inch of our intellectual life and then we watched what they did to each other, which was ten times as harsh. United States, so perhaps that matters.
Hard to disagree with you when you write so eloquently! Might I ask where you were and which department? and, phenomenal rhetoric aside, can you elucidate exactly how "they" did this? A high attrition rate can be due to low resilience as much as it can be due to high difficulty. Any insight into objective difficulty of what you had to do would be really good to know here.
Let's see. I can try to think of all the times we cried, maybe, and that will help, although I suspect you already dont believe me, so I'm not sure why I'm engaging.
So, because this is a text-based discipline I am in, the professors (and "they" is the department's professors, Im not sure who else if could be) ran us through rather quite a lot rhetorically and in terms of workload. In the first couple years, classes had you reading between three and six books a week, or their article-length equivalents, on top of whatever work requirements were needed to fund your tuition. I remember being cold-called in class, which was a prestige-building affair of great stress. You were meant to perform, both for each other and the professor, since these people would be your mentors and colleagues. And we did, sometimes with great theatrical flair and often, eventually, joy. But I would not call the grueling preps for those performances easy. Extremely productive, though.
At presentations of research, I was asked the most difficult of questions and occasionally laughed at for being underprepared. And I WAS underprepared. I learned quickly that I was not alone in my fear. I watched a woman from Yale come to give a talk to a full room there. This woman was up and coming, but damn smart, and she was shaking like a leaf. She admitted that she was worried about getting interrogated by this particular group of professors because of their reputation as utterly intolerant of lacunae in your knowledge or poor reasoning. The professors, to their credit, saw her shaking and went easy for the first few rounds of questions.
Orals exams was one hundred or so books in six months, with expectations about being able to summarize orally arguments, subarguments, sources, reception, and, crucially, how the books spoke to one another within the field of study. When we completed the exams, which were punishing, it was not unusual for us (and I did) to cry for a week. That was a big breaking point for a lot of people, and they would tap out at that point with an MA. And then, of course, there was research and writing: the eternal dissatisfaction with whatever chapter you had produced, the amateurishness of your prose, the failure to include x or y, the way you had missed some corner of the research, the working-classness of your affect. Throw out three chapters, start again. Etc. "Have you thought of maybe doing something else?" they often offered. What most of us produced in the end was very high quality. The majority of my cohort (the ones that were left) are teaching and researching in the Ivy League or adjacent. I am not, because I wasn't good enough.
This helped a lot on the job market, though not enough. I interviewed once at an R1 that had me in meetings with everyone under the sun, teaching classes, doing presentations, and being otherwise on for 16 full hours over two days. I prepped for weeks. There were also two "dinners" that were anything but relaxing and were performance pieces (I failed, not making past being in the final three candidates out of several hundred applicants). That was just one interview. But that interview was what the program I was in was training me for.
For what it's worth, I also did an MA in a related discipline at a state school in the US. We all graduated, and it was not hard.
I hope that is sufficient evidence. I could, I think, write for another while.
That's sort of the thing that is happening at the high levels of PhDness in the USA. Some people may find it easy. I did not.
This is going to sound harsh, but I believe it to be true: your programme sounds appropriately difficult for a PhD. That’s what I want everywhere. You yourself admit to how good it was for you and how ultimately fulfilling it was. I really think this is what the PhD is meant to be. If I were you I’d be pissed off that every department of every university doesn’t uphold the same rigour.
Oh, it was the best, most productive thing I've ever done. Hands down. The orals exams, where I cried for a week, were transformative. When I say I'm carved out of wood, I am delighted, not insulted.
Education in America right now is about making sure no one is every uncomfortable enough to have a transformative experience. We are all the poorer (and dumber) for that. I was lucky.
Phds should not be easy. Mine wasn't. I'm sad that some people have found them so.
an interesting essay. as someone who is completing their phd at oxbridge, i think you are generalising your cambridge undergrad + phd bubble to all of academia (or academia in the uk at least). i agree in that uk degrees should have more facets for assessment — and i think this would make the degree easier!! through regular assessment in taught courses, you learn quite quickly how to improve your work; the lack of guidance and feedback throughout the phd makes the process far more arduous. this lack of structure, assessments and accountability (especially for supervisors) causes great variation in phd experiences and four years of your life is pretty dictated by the lab you sign up to (usually from a standpoint of blind naivety).
another point regarding job security. you empahsise, rightly so, how most people who complete a phd end up have to make the switch from academia to industry / civil service — this is a very insecure career pathway! signing up for a career where you have to apply for funding constantly, or move away from family and friends for positions, etc etc. and during the ‘phd years’ you are being paid below market rate (compared to industry positions), without high chance of a big pay off on the horizon (unless maybe you are a math, comp sci, etc major looking to move into finance/tech/the rest of it).
You raise some interesting points. I think I'd push back on the idea that we need more frequent assessment per se. I like the freedom that a PhD provides and think it's necessary to produce scientific discovery. I agree it would be easier as you would have to exert less agency as you'd be kept within tighter guardrails. I think that's bad for science! I just think the assessment at the end could better reflect how well you have contributed to science a bit better, and I don't think a thesis and viva are particularly good at doing that (and they are easy compared to the actual science).
Totally agree with you on the unpredictability of supervisors/projects and the necessarily naive position you are in when you make the decision. Sadly that's never not a fact of science. Changing tack when a project isn't going anywhere by starting something new (defying the sunken cost fallacy) or using brute force to troubleshoot a project to resurrection are important skills for being a scientist. As are people skills. As is resilience to failure. Ultimately, intelligence, agency, and perseverance trump the stochasticity of science. My final suggestion in the essay, whereby failing the PhD requirements isn't terminal but rolls you over to another year, would allow for people to overcome these setbacks. In that instance, an unlucky project equals a long PhD, but at least all PhDs are held up to the same high standard.
On the career instability, yup you're right in the long run, but this backs my point about low agency. I'd say a PhD is an easy and stable option at the time, and the negative career consequences are suffered later. In this way, choosing to do a PhD is kind of like failing the marshmallow test but with regard to career trajectory.
By the way, I don't want to dissuade you at all from your PhD. It's a completely worthwhile endeavour if you make it so. My totally unsolicited advice would be to focus on discovery, then papers, then thesis (within the allowances of your supervisor of course!...)-- both in terms of where you put your effort and where you seek your reward.
I appreciate the provocation here — PhDs are often wrapped in mystique, and poking at credentialism is healthy. But I think the essay confuses structural design for leniency, and that’s where the argument wobbles.
For me, my PhD was brutally difficult — not because the science itself was insurmountable, but because my advisors were abusive, self-serving, and treated students as cheap labor. That’s a story for another day.
Here, I’ll focus on why high pass rates don’t necessarily mean PhDs are “easy” (drawing mostly on the U.S. system).
1. Selection bias.
PhD programs don’t admit people at random. They admit people who already show some mix of tenacity, intellect, and life experience that predicts success (whether or not these are biased predictors is another story). Of course completion rates are high — the funnel into the program itself is doing the heavy lifting.
2. Survivorship bias.
Completion stats only cover those who make it to the end. In the U.S. especially, students face hurdles like qualifying exams, dissertation proposals, and annual reviews where people do get filtered out. You don’t get to the viva if you’re not meeting expectations. That’s not grade inflation — it’s checkpointing. Maybe a student won't get kicked out of a program, but their stipend can be reduced, which then makes it harder to continue.
3. Structure and mentorship.
A viva isn’t a “final exam” in the usual sense. It represents years of guided work with an advisor whose job is to set the student up for success. Unlike a classroom exam where you might cram and fail, the PhD is designed to allow setbacks, course corrections, and gradual growth. For some programs, the supervisors require the students to do multiple rounds of revisions (sometimes over years) on their dissertation before they will even allow a student to set a date to do their final defense. That makes the viva less about gatekeeping and more about demonstrating that mentorship worked.
4. Conflicts of interest.
Yes, supervisors want their students to pass. Failing reflects poorly on them. But that’s the point — if you let someone drift for four years only to fail them at the end, that’s not just a student failure, it’s a supervisory one. High pass rates aren’t proof that PhDs are easy; they’re proof the system is structured to prevent wasting years of someone’s life.
In short, a PhD may feel “easy” if you had the right background, the right PI, and the right system. But that’s less about the inherent triviality of the degree, and more about survivorship, selection, and mentorship working as intended.
My PhD was a truly miserable experience. One useful thing I carried out of it was a blueprint of what not to do. And, that has been far more valuable to my career than any of the "mentorship" itself.
Thanks for your insights. Your story seems to be a really common one, and I think it's the greatest source of the pushback I've got here. I would offer that making the PhD criteria less nebulous, more objective, and more centralised removes the ability of tyrannical supervisors to keep their students from graduating indefinitely perhaps? I say "make PhDs harder" but I mean make the easiest possible path harder, and thereby level the playing field between different supervisors/fields etc.
On 1: I think the US does this better, and your argument holds up. In the UK the admissions process (at least in my experience) is not stringent at all.
On 2: The stats I used (which were specific to the UK) included a 16% dropout rate. Of the 84% that made it to the end, 97% pass. So yes a bit of a survivorship bias going on.
On 3: this is a really good point. It seems there are no solutions, only trade offs (as always!) Making a student refine their thesis before a viva is the supervisor-reputation incentive working well, but I still think there is plenty of scope for this incentive to work badly. It's tricky!
On 4: good point. But (as corroborated by your experience) this seems to give too much power to the supervisor. It shouldn't be the case that a worthy student has a potential career in science nuked because a supervisor was too lazy to pay attention for 4 years, and then they panicked when the viva was drawing near and decided to become draconian all of a sudden. Some kind of objectification or centralisation of standards would surely help to avoid this? These standards would make some PhDs harder, but others easier actually--overall reducing the huge variance in experiences of PhD students, which I think is sorely needed.
I think one issue for all of the pushback here is because PhD programs vary so much, not just between countries, but between institutions, and even between departments.
Whenever the topic of competency exams (comps) comes up, I always ask "what did you need to do for comps?" because, at least in the US, it seems like everybody has had a different experience.
But, back to a central point — supervisors often have way too much power. Not only the power to slow down a student, or be unpredictable, but I've also seen cases where they have the power to pass a student that the committee doesn't feel should pass. Yes, the entire committee has to sign off on the viva, but when the supervisor for the PhD student is also a senior faculty who controls the career trajectories of the other committee members, there can be pressure to just sign off.
Thanks for some interesting conversation here — it definitely provides me some ideas for future posts on my newsletter also!
As someone who, depending on how you look at it, flunked out of a Mathematics PhD in the US, I think what you're saying is at best only applicable to the UK. Do even begin PhD research in math in the US, you have to pass multiple exams which are absolutely brutal. They're comparable to med school entrance exams, and you have to do them multiple times. I'm actually a big proponent of loosening the requirements, at least in math, to even begin. I dropped to just doing a Masters and not a PhD because the exams were insane AND pointless. Anything I do research in I would learn. Why should I prove my knowledge of the fundamentals of several fields of study through an exam format than through coursework?
My department has an incentive to keep you working on your PhD. Once you're given the go ahead to start your dissertation, if you're a teaching assistant you start teaching an actual section of a class instead of providing extra instruction. The TAs are a lot cheaper than adjunct faculty.
The cheap labour incentive is an important one and it’s not good.
Maths is probably a special discipline in that innate ability is far more important than in other fields. Without wishing to insult, it sounds like you probably weren’t of that genius-level calibre that many maths PhDs are. To be honest, I think that’s probably a good thing. Getting a maths PhD is widely accepted as being very difficult because of this. It’s different gravy.
You should take a moment to consider how arrogant and rude you are coming across as. I suspect you did and wrote that qualifying statement as an attempt to avoid blame.
What you just said to me is that I could never cut it as a Mathematics PhD AND that it's a good thing. Think about how you would feel if I told you that you didn't deserve to get your life science PhD because you're not smart enough.
I decided not to do a PhD because the expectations were frankly unfair and I don't care enough about research to deal with them. I have the intention to return to getting a Mathematics PhD some day, but at a different institution because I want to specialize in college level mathematics education.
I've been thinking about your post since my friend sent it to me. I suspect you may have experienced a phenomenon similar to "big fish in a small pond." Instead of being the best in a world of average, you're amongst the worst in a world of exceptional. That doesn't mean you're not exceptional. So your perspective is entirely twisted into believing things one way. You don't even address your bias about PhDs being STEM fields. Lab work isn't applicable to the majority of possible areas to get PhDs.
I really don’t consider it rude to say to someone “I don’t think you are a super genius”. And there is no arrogance involved since I make zero claims about my own intellect. I would absolutely fail a maths undergrad, let alone a PhD!
It can just be true that the maths PhD is hard and that is a good thing, and that this means many people who are very smart but not the absolute smartest can fail them or drop out. The PhD is an elite degree. It seems like maths may be (one of) the only disciplines where this is still true, and my essay highlights how it is not true for life sciences.
Do you think that you deserved a PhD and were unjustly robbed of one? Or do you agree with the fact that those who got theirs in your cohort were more capable than you? It’s a brutally honest discussion and I hope you don’t mind the intrusion. I really don’t mean any ill-will by it. I trust you’ve done your introspection on the matter though and so hopefully won’t mind too much. If it’s still a sore topic then fair enough, we can draw a line under this.
I'm not calling you read for saying I'm not a genius; I'm calling you rude for implying that my decision to not get the PhD is a good thing and because I was incapable in the first place. Even if that's not what you intended to say, it's how you came across. And rudeness is more about what you say than what you mean.
Those in my cohort are more willing to jump through the hoops and insane requirements. I will say, my university seems to be an outlier even in the US based on discussions with other students. Those who are pursuing their PhDs are more interested in research or already had a masters degree when they started with us. (Note that those with masters still had to meet the same coursework and exam requirements as those without.) My decision to just pursue the masters was also based on what topics of research are available at the department—none of them interested me. I do not see my peers as being "more capable" than me. In fact, most of my peers were struggling as much as I was. I just decided it wasn't worth it to go for the PhD at this university.
When multiple of my undergrad professors referred to the first year or two of a math PhD as "hazing," it should show there's a problem. Talking with those who have started their research, the hardest part is passing the qualifying exams. Once one starts research, things become perhaps slightly anti-sisyphean, but not very much. Have you considered that the dropout rate vs fail rate is so high because departments don't want to fail students? Instead, they choose not to let them finish and eventually cut their funding, forcing them to drop out. I have no evidence for this hypothesis, but I suspect the majority of higher education drop outs didn't leave because it was too challenging, but because they deemed it not worth the time and effort, especially if you ignore those who failed out.
Fair enough. Does sound brutal. In the UK (even for maths), that two-year hazing doesn’t exist. It’s just research only. I considered myself a research more than a student. That fact alone probably explain how we had such different impressions.
In brief defence of the UK, the undergrad and masters at Cambridge are very tough (probably explaining why I did so badly at them 😂). The PhD at Cambridge is a notable deceleration, probably akin to the third year of your PhD programme and onwards.
I think Oxbridge is a bit of an outlier here re: the low-agency thing - there's a high concentration of undergrads who are high-IQ / low agency as you say, and I think a lot are quite institutionalised from their backgrounds and/or credentialism chasers with a bit of a background cushion meaning they don't need to be out there earning. Plus the sheer appeal of both universities qua institutions means lots of people just like the idea of being there just a bit longer (you see this even post-PhD, those who hang around as junior archivist, library roles, hangers on of music/chapel stuff, Blackwell's music etc. - no shade intended these are my people!)
I can’t speak for the UK, but in the US there’s in general a clear distinction between people who have PhDs and people who don’t, especially in STEM. Those who do simply think much more clearly, carefully, and quantitatively, and are more skeptical in technical ways rather than in some knee jerk fashion, scrutinizing claims in detail rather than simply accepting or rejecting them based on various heuristics.
I don’t mean to claim that PhDs are infallible by any stretch—rather that, in the US at least, the PhD serves to break us of sloppy thinking. Here, we need more of that, not less.
I do agree with you and many commenters that the PhD process is harder here than your experience (I think you should have put “UK PhD” in your title), as we often are required to (1) submit standardized test scores (general and/or subject GREs), (2) pass difficult coursework—much more rigorous than our undergrad or master’s courses, (3) pass preliminary exams—which you can often take a few times, but must pass a threshold score, (4) present your work at various points, often under the withering lens of various professors in the audience.
In my experience, the result is dual: a pre-selection of generally highly intelligent cadres (at least in mathematics and physics), plus a re-shaping process during the PhD, where they wring the slop out of your thought processes—and, relatedly, the arrogance out of your person. Those who leave with a Master’s, in my experience, don’t attain the slop-wringer benefit, at least not nearly to the same degree.
I wish our undergraduate system were much better, but until then, we need more PhDs, not fewer, so long as the quality doesn’t manifestly deteriorate.
Even in the UK, while the PhD should clearly be more difficult based on your description, I don’t think that should mean that fewer should be awarded. The entire point is to select the very best for faculty positions, wherever the bottleneck happens to occur. There could be good arguments for it to occur earlier on (reduce waste, as you say), or later on (increase the supply of a well-educated workforce, and probably increase innovation overall—certainly in the US that’s why they devised such a pyramid system). But unless I missed it, your argument seemed to lack any such analysis of benefits and drawbacks as to the particular bottleneck location.
Yeah. This article astonishes me. I attended a top-25 pure math program in the US. There were maybe 30 of us to start and 4 of us eventually got PhDs (most definitely not including myself). It was hands down the most brutal competitive environment I have ever participated in.
That is quite brutal! Out of curiosity, do you know how much variation there was year-to-year? Eg, would they only take the top 4/top 13%, or was it, at least in principle, possible for everyone to graduate?
My program moved from a somewhat similar model, where no more than half of the class would pass the preliminary exams (and thus eventually get a PhD) to one where 80%+ did so. As I understand it, the criteria in principle stayed essentially the same (pass the prelims, though the graders obviously could have big impacts), but they decided to become more selective.
I think this was entirely selfish, of course: they just got tired of losing prospective PhD students after they had already done promising research. In many people’s experiences, there was very little correlation between research quality and prelim exam performance
A maths PhD at Cambridge would be similar. I think the outlier here is the field! I don't know many people with maths PhDs at all but I assume they aren't that common. This is kind of a positive control for my thought experiment regarding making PhDs harder. It would be great if more fields were like maths!
I agree the US and UK PhDs are different. I do think UK school takes you to a higher level by age 18, and then undergrad in the UK is therefore more advanced too. I also think that between-field differences are greater than between-country differences. A maths PhD (per Emily's comment below) is substantially harder and more objective than a PhD in exercise science, for example.
Regarding the bottleneck, from the point of view of the individual, I think having it earlier is superior. From the point of view of the university, having the bottleneck later is superior because they get to use lots of graduate students and postdocs for productivity. And more students = more money. The system is just a bit of a pyramid scheme at the moment though--particularly in labour-intensive fields like life sciences.
I definitely agree that the difficulty is often more associated with the discipline than the university, with some variation.
Indeed, there can be quite different approaches within-discipline, even if the majority follow a stereotyped approach. My PhD was in mathematics, and I went to a tier 2 graduate school, but which had one of the best programs in my particular applied niche. They had just recently changed from a model closer to that of Emily’s alma mater—with a large chunk, 50%+—who “failed out” with a Master’s, to one where they selected fewer incoming students who stayed more frequently (moving the bottleneck earlier on). There were ~17 initially in my class, eg, and I believe ~13-14 finished, an attrition rate of perhaps 18%.
But, they still pushed you quite hard. You had I believe three chances to pass the preliminary exams (sometimes called “qualifiers”) after passing your initial classes. The purpose was to ensure a baseline level of mathematics for all who completed their PhD, and this was the main selection criterion. If you jumped this hurdle, you were almost guaranteed to graduate.
However, once you chose your specialty, you had a softer hurdle: comprehensive exams. These ensured you were up-to-speed on the math in your area. I don’t believe people ever failed out over these, but you might have to redo an exam or a problem, and so they still instilled fear into you to avoid the shame of such second attempts.
In contrast, at my graduate school, the physics prelim was satisfied simply by passing the first-year course exams. At other universities, such as my undergrad, similar qualifiers as what we had in math were enforced. At one postdoc institute I worked at in Canada, they had no prelims/qualifiers at all (&, unfortunately, it showed in the quality of at least some of their students).
With respect to the bottleneck location, you could be right! I’m not so sure, though. In terms of income, people with a STEM PhD are likely to earn substantially more than those with bachelor’s or Master’s, though there could be confounding factors difficult to account for in such studies. So long as people understand going in their career path prospects, shouldn’t we let people strive for what they want to? Perhaps we’re not as upfront about those prospects as we should be, but I was fully aware when I applied to graduate school. I’m on my second postdoc, and still am preparing for multiple possible outcomes.
I'm a big fan of qualifying exams or just some decently high and objective hurdle to overcome. Completely absent in the UK (except for undergrad medicine).
I also agree with your notion of just letting people strive for what they want as long as they understand their prospects. I'd even go further and blame the individual if they fail to understand the career prospects before starting down a certain career path. There is certainly no shortage of academic smelfunguses to forewarn prospective PhD students of the perils of academia!
I'm currently in the process of applying to computational biology PhDs in Cambridge and London, and I'm surprised to learn of your background going into your PhD. Maybe I'm mistaken, but I've noticed that the usual profile for candidates at the top PhD programmes seems to involve some combination of a first-class degree, a master's in a relevant subject, and ~1 year or more of post-university research experience, often accompanied by academic prizes. I finished my master's last year, graduated with a first, and am currently doing a second research internship in a similar field - just got my rejection email from EBI this morning, didn't even make it to interview.
Having said that, I don't at all anticipate that the experience of a PhD itself would be particularly gruelling - my experience so far has been actually quite relaxing, though maybe that's because I'm based in a rather tranquil research institute in London. The main drawback, in my opinion, is that I would lose 4 years during which I could be building myself up financially, and gaining momentum in a professional career. I don't see myself staying in academia partly because of the scarcity of positions, but also because it seems to inevitably terminate in a role that seems to mostly consist of meetings, admin, grant applications, managing people, etc.
I'm living hand-to-mouth and part of me feels exhausted at the prospect of continuing on this wage for 5 more years, but I really do love biology, and I don't think I can stomach pivoting into consulting/finance/law/etc in exchange for a higher salary. It just seems a terrible bore. I really wish the UK (or anywhere in continental Europe) had a life sciences industry comparable with the scale and vitality of that in the US.
I had plenty of research experience from my undergrad/masters and interviewed well. One of my cohort was straight out of a 3-year undergrad though with no other experience though tbf. Perhaps our admissions committee was unusually good at seeing through CVs and assessing actual scientific ability.
Competition has likely gotten much tougher since covid as I think more and more people are reaching for the academic lifestyle too.
Why don’t you apply to a PhD in the US? Or look at startups/biotech. The more you lean into tech, the less important having a PhD becomes. I don’t know how good you are or your interests, but why not cold-email some lab directors at the Arc institute?
The financial strain of (1) living in the UK (2) doing a PhD is tough. That said, a PhD from Cambridge will likely be very interesting and you can make it fulfilling and worthwhile for the future. If you don’t want to stay in academia then the case becomes much weaker. Keep applying nonetheless to give yourself options, but I’d definitely be looking at other options if I were you
I did an integrated master's at Durham. I would be the first to admit that it wasn't the most rigorous course, and at least in my experience there were no opportunities to get involved in research during undergrad. I'm a home student though, so that isn't an issue.
As for biotech/startups in the UK - is it wrong of me to think that entry-level roles are overwhelmingly likely to be given to a PhD, or at least someone with an exceptional undergrad/Oxbridge background? I will also add that job postings from these companies seem far and few in between, and hiring seems mostly done through recruiters, which makes it hard to get a sense of the scale and diversity of the life sciences industry opportunities in the golden triangle. UK unis produce an excess of PhDs, relative to the number of permanent posts in academia - and I would assume that as a result, any biotech/pharma role is going to have PhDs lining up to apply. It's hard to see how I could continue doing something biology related without a PhD.
Anyway I think if I did do a PhD in Cambridge, it would just be really fun. And like you say it is beautiful. It sounds silly but that means a lot to me too!
Yeah if you want to stay in life sciences and the UK, a PhD may actually pay off the most in the long run. Cambridge is a fantastic city and the student lifestyle is great, so it's not a bad thing to do for four years. Sounds like you're honest about the pros and cons and so you'll probably make a smart decision either way.
For biotech, I think a better route is via direct/personal connection. Many companies create positions for people directly, and it'll get "advertised" for like 24h. Cold emails (how I got my postdoc) or meeting people at conferences/talks is an underrated method.
I agree that PhD students have become glorified technicians and it would be better if programs were smaller and more selective. I also think it would be better if they were less focused on productivity. Now students immediately jump into projects because they need to publish as many papers as possible to be competitive on the job market, and PIs need someone making progress on a funded project. But I think we would get more interesting, creative science if they first spent a couple years deeply learning a field by doing nothing but reading and discussing what they read.
At the same time, I don't agree about coasting along being lazy. I spend hundreds of hours drumming up funding for students, and if someone doesn't accomplish anything, they're not getting any more funding from me and they're on their own for the rest of their degree. (This is for deliberate laziness, not the inevitable setbacks that happen sometimes in research). This tends to be a good motivator. I also don't see the thesis as separate from the papers. My students' theses are usually just their publications copied and pasted into one word doc. If they did enough work to produce 3+ articles publishable in respectable journals, that seems like a perfectly reasonable PhD to me.
Thanks for your insight! I actually really like your idea of reading deeply before starting a project--although hard to monitor / easy to lose motivation and do nothing I suppose. My own PhD actually had a version of this naturally built-in thanks to COVID. My first 6 months were spent just reading and writing a literature review, and I think this was really valuable. Despite this, I still did nowhere enough reading in the first half of my PhD. So much time was wasted doing poorly designed experiments that I could've avoided by reading more literature! I like your approach to the thesis. That's exactly what my supervisor and I decided on too. "Don't waste time on your thesis--nobody reads it" was good advice I received from her.
A lit review is a great way to approach this. There's a fantastic quote that a month in the lab can save an hour in the library, and it's so true. I also feel like I'm barely skimming the surface of the literature these days. I wish we could all take a few reading weeks every semester where there's no classes, no meetings, no admin work, and no new journal issues come out - every academic just reads and talks to their colleagues about it. Dream on!
Interesting post. I think the PhD can feel easy if it's something that you're intrinsically motivated to do, but many who are not super passionate/curious about their topic of choice would find it difficult to complete. Did you feel this way about the lab work? And did you enjoy the experience overall?
This also makes me wonder about the differences between lab-based and humanistic PhDs. I remember chatting with college mates who were putting in 60+ hours a week in their lab, while I was in the Politics department in Cambs and was basically in control of my schedule for the whole 3 years (outside of meetings with my supervisor every few months.)
Funny enough, I played a bit with the CURUFC team during the beginning of my PhD (played college rugby for the most part) and definitely understand your point about rugby being part of the appeal of a PhD.
Yup it’s a tough challenge but not an impossible one I don’t think. There must be ways to set absolute requirements that are better/harder and more uniform across PhDs within a field/department and also adjustable between fields/departments. I think achieving a level of citations which are set relative to the field is perhaps one example. Eg 10 citations in life sciences, 2 in politics (I’m making these numbers up but they should be derived via a unifying calculation based on total number of citations within a field). Can be of one’s thesis or journal articles. The degree cannot be awarded until this point. The degree cannot be withheld after this point (by the supervisor, say).
The whole thesis and viva thing is way too wishy washy for my liking, especially given the skewed incentives to pass students and the hand-picked examiners.
Thanks! I really try in the essay to separate the PhD requirements and the PhD experience, but perhaps I could've added another paragraph or two. The PhD requirements are uniform and easy. The PhD experience varies drastically from person to person and can be very difficult. I would certainly say I worked very hard and found mine difficult, but, as you suggest, I found this hard work fulfilling as I was motivated to do it (I certainly wasn't forced to or expected to by either university requirements or my supervisor). In terms of enjoyment, I absolutely loved it. But it was overcoming the difficulties and frustrations that was itself enjoyable in a sense.
In terms of schedules, I was pretty much in control of mine too. In first year I was very lazy. showing up between 10 and 11am. Disappearing by 4pm on a good day, earlier if I could justify "working from home". I switched gears somewhere in my second year (quitting CURUFC actually due to injury and frustration with the club). By third year my basic routine was 9am until 8pm in the lab (eating lunch and dinner at the lab) Mon-Fri and then 4 or 5 hours on Saturday and Sunday. Sometimes I did more when experiments required it, sometimes less when my social life required it. I was probably around the 60 h/week mark for a good 3 years though (whole thing lasted 4.5 years for me). Other PhD students I know stuck to the 11am to 4pm shift throughout their entire PhD (also lab-based). In fact my lab mate had a friend in the chemistry department who we called "lazy Ben" because he was simply never in the lab and constantly messaged my friend to ask if he wanted to play tennis lol. Lazy Ben will still get a PhD.
I refrain from blaming the lazy Bens of the world though, because they are acting rationally within their incentive structure. The point is that nobody is pushed or motivated by the basic PhD degree requirements. It's always either self-motivated or pushed by a supervisor. I think if universities in general raised the bar of minimum requirements, then it would create less space for nightmarish supervisors and equally less space for lazy free-riders.
Well said about the pleasures/satisfaction of "overcoming the difficulties and frustrations." That experience (and the resilience it instills) is the thing that unites all people who've done a PhD, no matter the subject.
One thing that crossed my mind while reading your response: perhaps the standards for passing are low because it is hard to develop consistent "high" standards when PhDs in different fields are so drastically different. Once at a college bar I met a German zoologist whose PhD was basically about trying to understand the mechanism by which a centipede's feet adhered to the surface the centipede was walking on. It's almost impossible to compare that to my project (in Politics but largely a historical PhD based on extensive archival research) or someone in material sciences, for example. So going with the standards of (a) getting into the program + getting funding (which is highly competitive at Cambs at least) and (b) doing work that the supervisor sees as sufficiently good is an imperfect solution when it's really tough to develop consistent standards across dozens of fields.
Really interesting perspective - I couldn't agree more with your fundamental point.
What I DO disagree on you is that the solution is more bureaucracy when I think the real problem that talent is just not fundamentally valued in the UK (especially in academia).
It may sound ridiculous but I think the solution is actually making it far more encouraged to focus on startups and commercialising research. For example, at the University of Edinburgh where I am, the university can take up to 55% equity in a spinout company, whilst at places like MIT it's closer to 5%.
Instead of adding standardized tests or strict graduation quotas to make PhDs artificially harder, we should make the post-graduate startup environment more rewarding. If UK universities copied the US approach and let founders keep their equity, the quality of PhD candidates would increase naturally. The best students would use the degree as a starting point to create new technology, rather than using it to avoid the regular job market.
tl;dr instead of deterring bad students; encourage those with the most talent with better incentives
Great point. I’m inclined to agree. I remember going to this seminar at the LMB in Cambridge about spinning out research. Greg Winter (inventor of humanized monoclonal antibodies for therapies) was on the panel and gave a very bleak view of the entrepreneurial landscape. Had absolutely nothing good to say about the university nor the MRC. Knowing that strongly informed my decision to do my postdoc in the US.
That said, there are other issues with not deterring/preventing bad students from getting into supposedly elite programmes. People tend to lose ambition when they’re surrounded by others without ambition. You also lose the benefit of talent-concentration (eg, two talented would-be founders never meet because the class size is double what it should be). I expand on these issues here: https://sotaletters.substack.com/p/reclaiming-the-british-university
I do like your preference for carrots instead of sticks though tbh. And the startup environment could be massively improved in UK.
I agree totally, I come from US in CS background and I see basically the same set of issues, even at top schools. The top 1% undergrads probably get a job at a quant firm or an AI lab. The only ones who join a PhD even if they are the top 1% is that they are overly fixated on credentials, which is itself a negative sign.
The US system still kinda works, because there are enough talented people who do a PhD not just for the credentials, but that number is constantly decreasing. I would say it’s around 10-20% of top lab PhDs right now. The other thing to note is complete decimation of non top lab PhDs.
In 1960s, most colleges in US were competitive, they had good researchers who invented stuff. Matlab came from Uni of New Mexico etc. Now that is almost never the case. At best it extends to top 10 universities, rest of them are a wasteland. They publish for the sake of publishing with no advancement in science and even the top schools probably have a majority of such work. It’s all very disheartening and people say CS / AI is still one of the better fields in Academia.
I don’t know how this ship can be righted but making PhDs far more rigorous is one solution. > 70% of the students should fail the quals the first time and only 50% should eventually pass. Masters thesis standards should be increased. Claude Shannon came up with information theory in his masters thesis, now most thesis are worthless (mine included). You should just not be given a thesis if it’s not a real contribution.
Overall, I’m not sure if this ship can be righted from the inside
I used to link a PDF of my masters thesis on my CV. I re-read it recently and have decided to remove the hyperlink lol.
I think you're right. The problem is, any attempts to reform academia is seen as an "attack on science". Standardized testing at undergrad and postgrad levels would probably be the easiest and most effective way to start to right the ship.
To add to the chorus of US PhDs responding to this: I got mine in Canada and it was NOT easy! It took me 6 years to fulfill all the requirements of a four-year PhD (and still finished before many others who started at the same time). As an international student (from the Netherlands) I wasn’t even allowed to start directly in the PhD programme but had to start on the MSc track first and then transfer in the 2nd year (because the Dutch MSc that I did was shorter than the Canadian one). I had to publish one research paper before graduation (and complete a submittable draft of another) so I was also at the mercy of finicky journal reviewers and it meant that my success was partly based on doing at least some experiments that didn’t yield negative results.
At least it taught me that I didn’t like research enough to be dealing with that kind of stress and uncertainty all the time, so I’ve been happily working as a science writer and science communicator ever since.
I go back and forth on whether publications are a good requirement to graduate. On the one hand, it’s basically the one actual output of academic research so it makes sense. On the other, making this a requirement might incentivise paper mills and low quality publications.
In your case, I think it sounds like it worked well minus the time penalty. I agree journals and reviewers make the whole thing much slower and shittier. That’s a whole other issue that should be solved (which I’ve written on too!).
In my case, I’m a year into my postdoc and my PhD supervisor still hasn’t resubmitted my big paper to a journal. So perhaps having the publication as a hard requirement might have given her some more motivation! I think my career has suffered as a consequence because my CV for postdoc fellowships is weak as a result. Anyways.
Science is hard, so it should be the case that PhDs reflect that. The system worked in your case, but many slip through the net. Also—glad you introspected and got off the tram lines. As someone who writes in their spare time, I imagine you’re finding science writing much more fulfilling that what you were doing during your PhD!
I found this very helpful and inspiring. However, I must ask since it reflects my current stage in life. How hard was it to get into Cambridge for your undergrad? I’ve heard they require at least 5 scores of 5 AP classes, 1500 SAT, and phenomenal UCAS scoring as well. I’m in the US considering applying, and I meet about half the requirements, probably needing to retake a few tests to score better. But is it really that intense? In the US, essays usually matter more. Did you and your peers all meet these standards? Or are these overblown and exaggerated?
Sounds pretty tough. I can’t say I’m too familiar with the standards for US applicants. Back when I applied yes the standards were super high, although they varied by course. Medicine, maths, and natural sciences had higher thresholds than other courses. Even if get an interview and do well in it, the actual offer can also be brutal—especially if you went to a good school (you get punished to level things out with other candidates). I had to get 4 A*s for my offer. Most A level students don’t even take 4 subjects lol.
That said, I think you should apply regardless—there’s only one way to find out if you’ll make it. And if you don’t make it, dealing with rejection and failure is good for the soul. So it’s a win-win :)
Great post! Agree generally but I think it's worth pointing out too aspects that don't carry over to the US (maybe also to other places but I wouldn't know).
1. The (relative) financial advantage of doing the PhD. Most people doing a PhD in the US (at least in STEM) could make >2x more working instead (>5x more if they went to a good undergrad) and there aren't really tax advantages. Plus, the resume boost is not very large AFAIK unless you're gunning for really specific jobs and those tend to be selective enough that they won't take you if you just scrape by during your PhD.
2. Not having to move cities for the PhD is very rare in the US. Some people stay in the same school but the vast majority go somewhere else and most schools are far away from each other.
Thanks! Great points. In general my piece is focused on the UK. While much of what I wrote is true across the board, I definitely grant that there are key differences in the states and I would consider the US PhD to be significantly harder.
On (1), I think this is true everywhere. I discuss the better financial compensation in London vs a PhD in the UK in my essay, and how this incentivises the best and brightest not to do a PhD. It’s not that there is a financial advantage of doing a PhD over a harder job, it’s that it’s better than being an undergrad and comfortable enough to imbue laziness or a lack of agency.
On (2), great point. I suppose for someone doing their undergrad at Cambridge, there are only two other places you’d want to do your PhD at in the UK (London or Oxford), so staying put is much more common. It’s definitely frowned upon but the same is true for the PhD-postdoc transition in the UK, with many staying where they are. There are more good choices in the US which is a great thing.
In my econ PhD we failed a bunch of people out at year one comp exams and then a few more failed out over the next few years. And those people weren’t dumb. My friends in Doctorates of Musical Arts worked much hard than me. There were other fields that were much less rigorous, but those were in “permanent adjunct or unemployment” studies.
Interesting post. I see some pushback in the comments and I note it comes from US experiences. Which matches the observation I was going to make: that UK and US PhDs are quite different. (I am British, with a British PhD, and have worked at both UK and US institutions.)
I agree with a lot of what you say here, but it is quite UK focused. The US PhD is significantly more demanding, and the modal US PhD graduate is ahead of the modal UK one, in terms of skills, experience, independence, etc.
Yes you’re totally right. Just looking at age gives insight here. Average age of PhD applicants is higher in the US, and US PhDs are longer too, so the age at graduation is *much* higher in the US.
That said, I think this discrepancy might normalise at postdoc, with most UK PhDs probably requiring two postdocs before independence (this is just my hunch, haven’t looked at data).
I still think the case can be made for more rigorous application standards in the US too, and also for better assessment criteria that more closely test “ability to contribute to scientific discovery”.
I think I don't know what it would mean for the discrepancy to normalise, but I do think that a typical new US PhD is roughly similar to a typical UK PhD plus two years of fellowship, by which time there has been important challenges of funding acquisition, a growing publishing arc, teaching experience, and so on.
The qualifications themselves aren't equivalent, despite having the same name. Experienced academics know this and take it into account when e.g. recruiting.
Yup you’ve explained what I was trying to—the difference in ability/experience equalises after some extra years of postdoc by the UK researcher
This seems to be the case in Europe in general. Undergrad is extremely difficult, grad school is (relatively) easy.
Are you American? I have often said the following: that the standards of US UG degrees are low but PG degrees are very challenging. Which is the same pattern you’re describing but in different words
Yes.
I wonder if, at least in continental Europe, this has to do with education being largely publicly funded. The first year or two are very difficult because they want to minimize waste.
But the final years are relatively easy because if you've gotten that far they don't want to lose their investment.
"This may sound ridiculous, or ridiculously arrogant, but let me clarify my point: I’m not trying to persuade you of my own intellect, I’m arguing that getting a PhD actually doesn’t require particularly high intellect nor does it require particularly hard work—objectively. Yet the societal perception of a PhD is that of a Herculean or even Sisyphean task (almost the opposite is true, as I will explain)."
Well, I'll give you points for honesty, it DOES sound ridiculous AND arrogant! I am not totally sure of your motives writing this, but I certainly hope it is either tongue-in-cheek, or deliberate rage-bait to go viral and grow your new Substack, because the most charitable thing I can say is your experience is very much *NOT* the norm of STEM PhDs in the United States, and from my professional circle, most of the world.
I am not sure how generalizable your quoted 16% drop-out rate is, because many studies suggest the attrition number in engineering, medicine, and the life sciences is closer to 40-50%. Perhaps half of all doctoral students, who already completed a bachelors +/- masters (a small minority of the population already) are just dummies? IMO, no, Occam's razor would suggest the most likely explanation is PhDs are difficult.
"Having been admitted to the PhD programme, what did I have to do to actually get the degree?
I had to attain adequate termly reports from my supervisor
I had to complete a first year report with a viva voce examination
I had to present my work once in four years at an internal departmental seminar
I had to eventually write my thesis, which doesn’t have a minimum word requirement, and then be examined on it in another viva voce."
If every PhD program was like that, I would agree it would be pretty easy. But that is not really the case! For my own PhD in molecular biology at a not-very-prestigious school in the southeastern US, I had to:
Complete 2 full years of additional courses in molecular biology, oncology and statistics (even though I already had another doctorate), some of which was quite rigorous in terms of not only the content, but also the tests and papers required (this was also in the pre-ChatGPT era)
Present my work *annually* at the school, and strongly encouraged at national conferences
Pass preliminary oral AND written exams by committee before beginning dissertation work (I agree that most people pass their prelims, but that is not the rate limiting step of PhDs anywhere, the tough part is the research)
Conduct novel research that would lead to at least two peer-reviewed publications (and ideally more)
I had to write and submit multiple extramural grants. The proposal did not have to be successfully funded, but submission was a requirement
Write up research results in a dissertation (mine was around 200 pages) and defend in front of committee
What makes a PhD difficult, and indeed more challenging than my STEM bachelor's or other doctorate in vetmed, is that nothing is guaranteed. You can end up in a nightmare project that implodes, and your PhD can be over. Some people can eke out a PhD with negative results; most can't. You can have a PI who never stops asking for more experiments, more papers, and other hoops, delaying you endlessly. You can run out of funding. Members of your committee can stonewall you for political or personal reasons (not the most common, but it happens). In contrast, pretty much every other degree out there (including med school!) follows a linear course, and if you check all the boxes and pass the tests, congrats, you get the diploma. Not the case for PhD.
If you want to make the case for more uniformity in graduate training, I'm all for it. But please, don't go shitting on thousands of people's careers just because you had a somewhat easier path. Thanks!
Hi Eric. It seems I have ruffled your feathers, and I apologise for that. It’s also great to get your insight on your experience, thank you for that.
First thing to note is that my experience is representative of the UK, and not the US, which I note in the essay. I think it is absolutely true that US PhDs are harder.
The attrition rate says different things to you and me. To you, a high attrition rate = difficult. To me, a high attrition rate = a lack of resilience. To me, this means the application process could have been more rigorous to prevent these drop outs from wasting years of their lives. I suspect we’ll disagree here and that’s fine.
It seems you do largely agree with me but want to disagree because you found my writing to be offensive to the prestige of your PhD. For example, you agree that I had easy requirements. Would you therefore at least agree with my call for *UK* PhDs to be harder? And if so, then ask yourself if you perceive someone with a PhD from Cambridge or Oxford to be inferior or superior to someone with a PhD from Northeastern Illinois University? And if you perceive Oxford to be superior, then you’ve realised the point of my essay, thank you.
I’m interested in your requirements for degree completion. Was two first author publications a definitive requirement? As in, you fail if you do not achieve this? Writing a grant seems like a great addition and would be welcome in the UK (for my PhD I also had to do this actually to get my Wellcome Trust funding, but I didn’t include it since it’s separate to the PhD itself).
You say “the tough part is the research” and go on to say how what makes a PhD challenging is the uncertainty of scientific research. I totally agree with you!!! I hope I made that clear in the essay (I think I did?!) the toughest part IS the research. NOT the parts that get assessed for the degree (caveat in the differences we’ve discussed for you and me). To boil it down, I’m saying: research is bloody hard and I had a REALLY hard time in my PhD, but I didn’t have to for the degree alone. I chose to because I wanted to make discoveries and contribute to science. All I’m saying is that the ways in which PhDs are assessed are quite unrelated to “contribution to science” and I wish they were more closely aligned especially because “contribution to science” is so difficult, unlike the degree of a PhD.
Hopefully I’ve clarified my position here. I’m not at all shitting on anyone’s career (why would I shit on my own career?!). People who worked really hard during their PhDs (sounds like you did) should find themselves agreeing with my call to make such hard work more appropriately assessed for the degree itself.
Thank you for engaging with my comment instead of reflexive dismissal. You correctly observed that my feathers are ruffled, but your diagnosis is wrong. It is not about "prestige"; I don't even list my PhD in my Substack by-line or bio (and I didn't mention my undergrad institution, but it is prestigious). And I would agree that simply having a PhD does not guarantee one smart.
So why do I consider my PhD one of my proudest accomplishments? Because it took an *ENORMOUS* amount of work over 6 years (while simultaneously doing a medical residency) and contributed to the literature in my tiny field. Then some bloke from the UK who got his degree 5 minutes ago comes along and fires off a post telling the world that PhDs are a joke and barely any work at all. So yeah, color me offended. However, what is most upsetting about this isn't even personal; at a time where public faith in scientists has significantly declined (see reference at the end), the LAST thing we need is misleading media that can justify their prejudice. I don't know if you follow the news much, but in the US we have a hostile new administration that is absolutely gutting what was formerly the crown jewel of biomedical research in the world (the NIH, NSF, and American research university system). Those of us who work in science need to be careful with how our words come across to the public, lest they be coopted and twisted against it.
Side note: You asked about my perception of Cambridge or Oxford. I don't care about "prestige," some of the best research training programs in the US are at public schools like the University of Wisconsin-Madison or the University of Washington in Seattle. Prior to this post, I had no strong views on Cambridge either way, but I have to admit after reading your article, my respect decreased quite a bit.
"I’m interested in your requirements for degree completion. Was two first author publications a definitive requirement? As in, you fail if you do not achieve this?"
The specifics of each program are highly dependent on one's PI and committee. My major professor told me that before he would let me defend my dissertation I needed at least two peer-reviewed publications. I don't know if you would consider not meeting that "failing," but indefinite limbo is essentially the same thing on a long enough timeline (see discussion below).
"The attrition rate says different things to you and me. To you, a high attrition rate = difficult. To me, a high attrition rate = a lack of resilience. To me, this means the application process could have been more rigorous to prevent these drop outs from wasting years of their lives. I suspect we’ll disagree here and that’s fine."
Strong disagree. There are so many reasons someone may have to drop out of a PhD program that do not fit with "lack of resilience." In the US, students often take on massive debt to attend university, and graduate students are paid below minimum wage for many, many years. Grad students also have crappy health insurance (some may totally lack it), and we do not have single-payer healthcare to fall back on. Does someone who cannot handle the financial strain of that "lack resilience"? Several of my colleagues in grad school had to quit because they couldn't make ends meet after so many years. Others worked multiple jobs. I myself drove to another state to work ER shifts to make extra $.
What about someone who is working on a cutting edge project that fails? (Incidentally, the severe consequences of losing funding and/or grad students leads to significant risk averse behavior, and is a major reason why so many projects are incremental instead of groundbreaking innovation). What if their hypothesis IS right, but they are scooped by another research group, and their work is no longer publishable or fundable to the same degree?
What about the many female grad students who have to grapple with sexual harassment from superiors? Or who get pregnant during their studies and don't have accommodating labs? Even if they do, it can delay them by years, and you circle back to the financial strain argument, now with an extra mouth to feed and clothe.
Look, I don't know you, so I don't want to be too harsh, but at minimum, a straight white male who casually bragged they went to undergrad at Cambridge and didn't even have to move for their PhD program seems like someone who has lived a fairly privileged life and may not have contemplated a lot of these challenges. I don't have a study backing this up offhand, but in my experience, the vast majority of grad students do their PhD at a different institution than BS/masters, which necessitates moves, expenses, etc.
"You say “the tough part is the research” and go on to say how what makes a PhD challenging is the uncertainty of scientific research. I totally agree with you!!! I hope I made that clear in the essay (I think I did?!) the toughest part IS the research. NOT the parts that get assessed for the degree (caveat in the differences we’ve discussed for you and me). To boil it down, I’m saying: research is bloody hard and I had a REALLY hard time in my PhD, but I didn’t have to for the degree alone. [...] All I’m saying is that the ways in which PhDs are assessed are quite unrelated to “contribution to science” and I wish they were more closely aligned especially because “contribution to science” is so difficult, unlike the degree of a PhD."
It sounds like PhDs in the UK essentially don't even assess the quantity and quality of a grad student's research. If that characterization is accurate, I am shocked, and feel like the degree should almost be called something else than "PhD," since it has such a different connotation in North America, parts of the EU, and other places around the world. This is likely why tenure-track faculty positions (which are becoming rarer by the year!) require multiple post-docs, since you can't rely on the quality of one's doctoral training.
A. Lupia, D.B. Allison, K.H. Jamieson, J. Heimberg, M. Skipper, & S.M. Wolf, Trends in US public confidence in science and opportunities for progress, Proc. Natl. Acad. Sci. U.S.A. 121 (11) e2319488121, https://doi.org/10.1073/pnas.2319488121 (2024).
Lots I could respond to here. Lot's of assumptions in your reply about who I am and what I know (and don't know). Let me limit myself to a point of agreement:
You point out how it was your supervisor who imposed the stringent two-paper requirement. Many other commenters point out how a supervisor can dictate the difficulty of a PhD. I literally write about that in the essay and don't deny it. I agree that this isn't good. I would perhaps posit that these faculty members are responding to the fact that university-levied requirements are insufficient to drive scientific discovery. I'm arguing for all PhD-granting institutions to make requirements more stringent such that there is less variability between supervisors and the letters "PhD" better reflect actual contribution to science, which supervisors like yours clearly insisted upon on their own accord.
I'm sure this is meaningless to you but just to let you know that I (a random internet stranger) agree's with you.
We lost more than fifty percent of my cohort of PhDs before they reached candidacy. I don’t know how many after that rang the bell. They fried us up and ate us for lunch. They beat us within an inch of our intellectual life and then we watched what they did to each other, which was ten times as harsh. United States, so perhaps that matters.
I’m carved out of wood because of it.
It was not easy.
Hard to disagree with you when you write so eloquently! Might I ask where you were and which department? and, phenomenal rhetoric aside, can you elucidate exactly how "they" did this? A high attrition rate can be due to low resilience as much as it can be due to high difficulty. Any insight into objective difficulty of what you had to do would be really good to know here.
Let's see. I can try to think of all the times we cried, maybe, and that will help, although I suspect you already dont believe me, so I'm not sure why I'm engaging.
So, because this is a text-based discipline I am in, the professors (and "they" is the department's professors, Im not sure who else if could be) ran us through rather quite a lot rhetorically and in terms of workload. In the first couple years, classes had you reading between three and six books a week, or their article-length equivalents, on top of whatever work requirements were needed to fund your tuition. I remember being cold-called in class, which was a prestige-building affair of great stress. You were meant to perform, both for each other and the professor, since these people would be your mentors and colleagues. And we did, sometimes with great theatrical flair and often, eventually, joy. But I would not call the grueling preps for those performances easy. Extremely productive, though.
At presentations of research, I was asked the most difficult of questions and occasionally laughed at for being underprepared. And I WAS underprepared. I learned quickly that I was not alone in my fear. I watched a woman from Yale come to give a talk to a full room there. This woman was up and coming, but damn smart, and she was shaking like a leaf. She admitted that she was worried about getting interrogated by this particular group of professors because of their reputation as utterly intolerant of lacunae in your knowledge or poor reasoning. The professors, to their credit, saw her shaking and went easy for the first few rounds of questions.
Orals exams was one hundred or so books in six months, with expectations about being able to summarize orally arguments, subarguments, sources, reception, and, crucially, how the books spoke to one another within the field of study. When we completed the exams, which were punishing, it was not unusual for us (and I did) to cry for a week. That was a big breaking point for a lot of people, and they would tap out at that point with an MA. And then, of course, there was research and writing: the eternal dissatisfaction with whatever chapter you had produced, the amateurishness of your prose, the failure to include x or y, the way you had missed some corner of the research, the working-classness of your affect. Throw out three chapters, start again. Etc. "Have you thought of maybe doing something else?" they often offered. What most of us produced in the end was very high quality. The majority of my cohort (the ones that were left) are teaching and researching in the Ivy League or adjacent. I am not, because I wasn't good enough.
This helped a lot on the job market, though not enough. I interviewed once at an R1 that had me in meetings with everyone under the sun, teaching classes, doing presentations, and being otherwise on for 16 full hours over two days. I prepped for weeks. There were also two "dinners" that were anything but relaxing and were performance pieces (I failed, not making past being in the final three candidates out of several hundred applicants). That was just one interview. But that interview was what the program I was in was training me for.
For what it's worth, I also did an MA in a related discipline at a state school in the US. We all graduated, and it was not hard.
I hope that is sufficient evidence. I could, I think, write for another while.
That's sort of the thing that is happening at the high levels of PhDness in the USA. Some people may find it easy. I did not.
This is going to sound harsh, but I believe it to be true: your programme sounds appropriately difficult for a PhD. That’s what I want everywhere. You yourself admit to how good it was for you and how ultimately fulfilling it was. I really think this is what the PhD is meant to be. If I were you I’d be pissed off that every department of every university doesn’t uphold the same rigour.
Oh, it was the best, most productive thing I've ever done. Hands down. The orals exams, where I cried for a week, were transformative. When I say I'm carved out of wood, I am delighted, not insulted.
Education in America right now is about making sure no one is every uncomfortable enough to have a transformative experience. We are all the poorer (and dumber) for that. I was lucky.
Phds should not be easy. Mine wasn't. I'm sad that some people have found them so.
This fires me up. Glad we can agree on this!
an interesting essay. as someone who is completing their phd at oxbridge, i think you are generalising your cambridge undergrad + phd bubble to all of academia (or academia in the uk at least). i agree in that uk degrees should have more facets for assessment — and i think this would make the degree easier!! through regular assessment in taught courses, you learn quite quickly how to improve your work; the lack of guidance and feedback throughout the phd makes the process far more arduous. this lack of structure, assessments and accountability (especially for supervisors) causes great variation in phd experiences and four years of your life is pretty dictated by the lab you sign up to (usually from a standpoint of blind naivety).
another point regarding job security. you empahsise, rightly so, how most people who complete a phd end up have to make the switch from academia to industry / civil service — this is a very insecure career pathway! signing up for a career where you have to apply for funding constantly, or move away from family and friends for positions, etc etc. and during the ‘phd years’ you are being paid below market rate (compared to industry positions), without high chance of a big pay off on the horizon (unless maybe you are a math, comp sci, etc major looking to move into finance/tech/the rest of it).
You raise some interesting points. I think I'd push back on the idea that we need more frequent assessment per se. I like the freedom that a PhD provides and think it's necessary to produce scientific discovery. I agree it would be easier as you would have to exert less agency as you'd be kept within tighter guardrails. I think that's bad for science! I just think the assessment at the end could better reflect how well you have contributed to science a bit better, and I don't think a thesis and viva are particularly good at doing that (and they are easy compared to the actual science).
Totally agree with you on the unpredictability of supervisors/projects and the necessarily naive position you are in when you make the decision. Sadly that's never not a fact of science. Changing tack when a project isn't going anywhere by starting something new (defying the sunken cost fallacy) or using brute force to troubleshoot a project to resurrection are important skills for being a scientist. As are people skills. As is resilience to failure. Ultimately, intelligence, agency, and perseverance trump the stochasticity of science. My final suggestion in the essay, whereby failing the PhD requirements isn't terminal but rolls you over to another year, would allow for people to overcome these setbacks. In that instance, an unlucky project equals a long PhD, but at least all PhDs are held up to the same high standard.
On the career instability, yup you're right in the long run, but this backs my point about low agency. I'd say a PhD is an easy and stable option at the time, and the negative career consequences are suffered later. In this way, choosing to do a PhD is kind of like failing the marshmallow test but with regard to career trajectory.
By the way, I don't want to dissuade you at all from your PhD. It's a completely worthwhile endeavour if you make it so. My totally unsolicited advice would be to focus on discovery, then papers, then thesis (within the allowances of your supervisor of course!...)-- both in terms of where you put your effort and where you seek your reward.
I appreciate the provocation here — PhDs are often wrapped in mystique, and poking at credentialism is healthy. But I think the essay confuses structural design for leniency, and that’s where the argument wobbles.
For me, my PhD was brutally difficult — not because the science itself was insurmountable, but because my advisors were abusive, self-serving, and treated students as cheap labor. That’s a story for another day.
Here, I’ll focus on why high pass rates don’t necessarily mean PhDs are “easy” (drawing mostly on the U.S. system).
1. Selection bias.
PhD programs don’t admit people at random. They admit people who already show some mix of tenacity, intellect, and life experience that predicts success (whether or not these are biased predictors is another story). Of course completion rates are high — the funnel into the program itself is doing the heavy lifting.
2. Survivorship bias.
Completion stats only cover those who make it to the end. In the U.S. especially, students face hurdles like qualifying exams, dissertation proposals, and annual reviews where people do get filtered out. You don’t get to the viva if you’re not meeting expectations. That’s not grade inflation — it’s checkpointing. Maybe a student won't get kicked out of a program, but their stipend can be reduced, which then makes it harder to continue.
3. Structure and mentorship.
A viva isn’t a “final exam” in the usual sense. It represents years of guided work with an advisor whose job is to set the student up for success. Unlike a classroom exam where you might cram and fail, the PhD is designed to allow setbacks, course corrections, and gradual growth. For some programs, the supervisors require the students to do multiple rounds of revisions (sometimes over years) on their dissertation before they will even allow a student to set a date to do their final defense. That makes the viva less about gatekeeping and more about demonstrating that mentorship worked.
4. Conflicts of interest.
Yes, supervisors want their students to pass. Failing reflects poorly on them. But that’s the point — if you let someone drift for four years only to fail them at the end, that’s not just a student failure, it’s a supervisory one. High pass rates aren’t proof that PhDs are easy; they’re proof the system is structured to prevent wasting years of someone’s life.
In short, a PhD may feel “easy” if you had the right background, the right PI, and the right system. But that’s less about the inherent triviality of the degree, and more about survivorship, selection, and mentorship working as intended.
My PhD was a truly miserable experience. One useful thing I carried out of it was a blueprint of what not to do. And, that has been far more valuable to my career than any of the "mentorship" itself.
Thanks for your insights. Your story seems to be a really common one, and I think it's the greatest source of the pushback I've got here. I would offer that making the PhD criteria less nebulous, more objective, and more centralised removes the ability of tyrannical supervisors to keep their students from graduating indefinitely perhaps? I say "make PhDs harder" but I mean make the easiest possible path harder, and thereby level the playing field between different supervisors/fields etc.
On 1: I think the US does this better, and your argument holds up. In the UK the admissions process (at least in my experience) is not stringent at all.
On 2: The stats I used (which were specific to the UK) included a 16% dropout rate. Of the 84% that made it to the end, 97% pass. So yes a bit of a survivorship bias going on.
On 3: this is a really good point. It seems there are no solutions, only trade offs (as always!) Making a student refine their thesis before a viva is the supervisor-reputation incentive working well, but I still think there is plenty of scope for this incentive to work badly. It's tricky!
On 4: good point. But (as corroborated by your experience) this seems to give too much power to the supervisor. It shouldn't be the case that a worthy student has a potential career in science nuked because a supervisor was too lazy to pay attention for 4 years, and then they panicked when the viva was drawing near and decided to become draconian all of a sudden. Some kind of objectification or centralisation of standards would surely help to avoid this? These standards would make some PhDs harder, but others easier actually--overall reducing the huge variance in experiences of PhD students, which I think is sorely needed.
Thanks for your thoughts!
I think one issue for all of the pushback here is because PhD programs vary so much, not just between countries, but between institutions, and even between departments.
Whenever the topic of competency exams (comps) comes up, I always ask "what did you need to do for comps?" because, at least in the US, it seems like everybody has had a different experience.
But, back to a central point — supervisors often have way too much power. Not only the power to slow down a student, or be unpredictable, but I've also seen cases where they have the power to pass a student that the committee doesn't feel should pass. Yes, the entire committee has to sign off on the viva, but when the supervisor for the PhD student is also a senior faculty who controls the career trajectories of the other committee members, there can be pressure to just sign off.
Thanks for some interesting conversation here — it definitely provides me some ideas for future posts on my newsletter also!
As someone who, depending on how you look at it, flunked out of a Mathematics PhD in the US, I think what you're saying is at best only applicable to the UK. Do even begin PhD research in math in the US, you have to pass multiple exams which are absolutely brutal. They're comparable to med school entrance exams, and you have to do them multiple times. I'm actually a big proponent of loosening the requirements, at least in math, to even begin. I dropped to just doing a Masters and not a PhD because the exams were insane AND pointless. Anything I do research in I would learn. Why should I prove my knowledge of the fundamentals of several fields of study through an exam format than through coursework?
My department has an incentive to keep you working on your PhD. Once you're given the go ahead to start your dissertation, if you're a teaching assistant you start teaching an actual section of a class instead of providing extra instruction. The TAs are a lot cheaper than adjunct faculty.
The cheap labour incentive is an important one and it’s not good.
Maths is probably a special discipline in that innate ability is far more important than in other fields. Without wishing to insult, it sounds like you probably weren’t of that genius-level calibre that many maths PhDs are. To be honest, I think that’s probably a good thing. Getting a maths PhD is widely accepted as being very difficult because of this. It’s different gravy.
"Without wishing to insult"
You should take a moment to consider how arrogant and rude you are coming across as. I suspect you did and wrote that qualifying statement as an attempt to avoid blame.
What you just said to me is that I could never cut it as a Mathematics PhD AND that it's a good thing. Think about how you would feel if I told you that you didn't deserve to get your life science PhD because you're not smart enough.
I decided not to do a PhD because the expectations were frankly unfair and I don't care enough about research to deal with them. I have the intention to return to getting a Mathematics PhD some day, but at a different institution because I want to specialize in college level mathematics education.
I've been thinking about your post since my friend sent it to me. I suspect you may have experienced a phenomenon similar to "big fish in a small pond." Instead of being the best in a world of average, you're amongst the worst in a world of exceptional. That doesn't mean you're not exceptional. So your perspective is entirely twisted into believing things one way. You don't even address your bias about PhDs being STEM fields. Lab work isn't applicable to the majority of possible areas to get PhDs.
I really don’t consider it rude to say to someone “I don’t think you are a super genius”. And there is no arrogance involved since I make zero claims about my own intellect. I would absolutely fail a maths undergrad, let alone a PhD!
It can just be true that the maths PhD is hard and that is a good thing, and that this means many people who are very smart but not the absolute smartest can fail them or drop out. The PhD is an elite degree. It seems like maths may be (one of) the only disciplines where this is still true, and my essay highlights how it is not true for life sciences.
Do you think that you deserved a PhD and were unjustly robbed of one? Or do you agree with the fact that those who got theirs in your cohort were more capable than you? It’s a brutally honest discussion and I hope you don’t mind the intrusion. I really don’t mean any ill-will by it. I trust you’ve done your introspection on the matter though and so hopefully won’t mind too much. If it’s still a sore topic then fair enough, we can draw a line under this.
I'm not calling you read for saying I'm not a genius; I'm calling you rude for implying that my decision to not get the PhD is a good thing and because I was incapable in the first place. Even if that's not what you intended to say, it's how you came across. And rudeness is more about what you say than what you mean.
Those in my cohort are more willing to jump through the hoops and insane requirements. I will say, my university seems to be an outlier even in the US based on discussions with other students. Those who are pursuing their PhDs are more interested in research or already had a masters degree when they started with us. (Note that those with masters still had to meet the same coursework and exam requirements as those without.) My decision to just pursue the masters was also based on what topics of research are available at the department—none of them interested me. I do not see my peers as being "more capable" than me. In fact, most of my peers were struggling as much as I was. I just decided it wasn't worth it to go for the PhD at this university.
When multiple of my undergrad professors referred to the first year or two of a math PhD as "hazing," it should show there's a problem. Talking with those who have started their research, the hardest part is passing the qualifying exams. Once one starts research, things become perhaps slightly anti-sisyphean, but not very much. Have you considered that the dropout rate vs fail rate is so high because departments don't want to fail students? Instead, they choose not to let them finish and eventually cut their funding, forcing them to drop out. I have no evidence for this hypothesis, but I suspect the majority of higher education drop outs didn't leave because it was too challenging, but because they deemed it not worth the time and effort, especially if you ignore those who failed out.
Fair enough. Does sound brutal. In the UK (even for maths), that two-year hazing doesn’t exist. It’s just research only. I considered myself a research more than a student. That fact alone probably explain how we had such different impressions.
In brief defence of the UK, the undergrad and masters at Cambridge are very tough (probably explaining why I did so badly at them 😂). The PhD at Cambridge is a notable deceleration, probably akin to the third year of your PhD programme and onwards.
I think Oxbridge is a bit of an outlier here re: the low-agency thing - there's a high concentration of undergrads who are high-IQ / low agency as you say, and I think a lot are quite institutionalised from their backgrounds and/or credentialism chasers with a bit of a background cushion meaning they don't need to be out there earning. Plus the sheer appeal of both universities qua institutions means lots of people just like the idea of being there just a bit longer (you see this even post-PhD, those who hang around as junior archivist, library roles, hangers on of music/chapel stuff, Blackwell's music etc. - no shade intended these are my people!)
Well put
I can’t speak for the UK, but in the US there’s in general a clear distinction between people who have PhDs and people who don’t, especially in STEM. Those who do simply think much more clearly, carefully, and quantitatively, and are more skeptical in technical ways rather than in some knee jerk fashion, scrutinizing claims in detail rather than simply accepting or rejecting them based on various heuristics.
I don’t mean to claim that PhDs are infallible by any stretch—rather that, in the US at least, the PhD serves to break us of sloppy thinking. Here, we need more of that, not less.
I do agree with you and many commenters that the PhD process is harder here than your experience (I think you should have put “UK PhD” in your title), as we often are required to (1) submit standardized test scores (general and/or subject GREs), (2) pass difficult coursework—much more rigorous than our undergrad or master’s courses, (3) pass preliminary exams—which you can often take a few times, but must pass a threshold score, (4) present your work at various points, often under the withering lens of various professors in the audience.
In my experience, the result is dual: a pre-selection of generally highly intelligent cadres (at least in mathematics and physics), plus a re-shaping process during the PhD, where they wring the slop out of your thought processes—and, relatedly, the arrogance out of your person. Those who leave with a Master’s, in my experience, don’t attain the slop-wringer benefit, at least not nearly to the same degree.
I wish our undergraduate system were much better, but until then, we need more PhDs, not fewer, so long as the quality doesn’t manifestly deteriorate.
Even in the UK, while the PhD should clearly be more difficult based on your description, I don’t think that should mean that fewer should be awarded. The entire point is to select the very best for faculty positions, wherever the bottleneck happens to occur. There could be good arguments for it to occur earlier on (reduce waste, as you say), or later on (increase the supply of a well-educated workforce, and probably increase innovation overall—certainly in the US that’s why they devised such a pyramid system). But unless I missed it, your argument seemed to lack any such analysis of benefits and drawbacks as to the particular bottleneck location.
Yeah. This article astonishes me. I attended a top-25 pure math program in the US. There were maybe 30 of us to start and 4 of us eventually got PhDs (most definitely not including myself). It was hands down the most brutal competitive environment I have ever participated in.
That is quite brutal! Out of curiosity, do you know how much variation there was year-to-year? Eg, would they only take the top 4/top 13%, or was it, at least in principle, possible for everyone to graduate?
My program moved from a somewhat similar model, where no more than half of the class would pass the preliminary exams (and thus eventually get a PhD) to one where 80%+ did so. As I understand it, the criteria in principle stayed essentially the same (pass the prelims, though the graders obviously could have big impacts), but they decided to become more selective.
I think this was entirely selfish, of course: they just got tired of losing prospective PhD students after they had already done promising research. In many people’s experiences, there was very little correlation between research quality and prelim exam performance
A maths PhD at Cambridge would be similar. I think the outlier here is the field! I don't know many people with maths PhDs at all but I assume they aren't that common. This is kind of a positive control for my thought experiment regarding making PhDs harder. It would be great if more fields were like maths!
I agree the US and UK PhDs are different. I do think UK school takes you to a higher level by age 18, and then undergrad in the UK is therefore more advanced too. I also think that between-field differences are greater than between-country differences. A maths PhD (per Emily's comment below) is substantially harder and more objective than a PhD in exercise science, for example.
Regarding the bottleneck, from the point of view of the individual, I think having it earlier is superior. From the point of view of the university, having the bottleneck later is superior because they get to use lots of graduate students and postdocs for productivity. And more students = more money. The system is just a bit of a pyramid scheme at the moment though--particularly in labour-intensive fields like life sciences.
I definitely agree that the difficulty is often more associated with the discipline than the university, with some variation.
Indeed, there can be quite different approaches within-discipline, even if the majority follow a stereotyped approach. My PhD was in mathematics, and I went to a tier 2 graduate school, but which had one of the best programs in my particular applied niche. They had just recently changed from a model closer to that of Emily’s alma mater—with a large chunk, 50%+—who “failed out” with a Master’s, to one where they selected fewer incoming students who stayed more frequently (moving the bottleneck earlier on). There were ~17 initially in my class, eg, and I believe ~13-14 finished, an attrition rate of perhaps 18%.
But, they still pushed you quite hard. You had I believe three chances to pass the preliminary exams (sometimes called “qualifiers”) after passing your initial classes. The purpose was to ensure a baseline level of mathematics for all who completed their PhD, and this was the main selection criterion. If you jumped this hurdle, you were almost guaranteed to graduate.
However, once you chose your specialty, you had a softer hurdle: comprehensive exams. These ensured you were up-to-speed on the math in your area. I don’t believe people ever failed out over these, but you might have to redo an exam or a problem, and so they still instilled fear into you to avoid the shame of such second attempts.
In contrast, at my graduate school, the physics prelim was satisfied simply by passing the first-year course exams. At other universities, such as my undergrad, similar qualifiers as what we had in math were enforced. At one postdoc institute I worked at in Canada, they had no prelims/qualifiers at all (&, unfortunately, it showed in the quality of at least some of their students).
With respect to the bottleneck location, you could be right! I’m not so sure, though. In terms of income, people with a STEM PhD are likely to earn substantially more than those with bachelor’s or Master’s, though there could be confounding factors difficult to account for in such studies. So long as people understand going in their career path prospects, shouldn’t we let people strive for what they want to? Perhaps we’re not as upfront about those prospects as we should be, but I was fully aware when I applied to graduate school. I’m on my second postdoc, and still am preparing for multiple possible outcomes.
I'm a big fan of qualifying exams or just some decently high and objective hurdle to overcome. Completely absent in the UK (except for undergrad medicine).
I also agree with your notion of just letting people strive for what they want as long as they understand their prospects. I'd even go further and blame the individual if they fail to understand the career prospects before starting down a certain career path. There is certainly no shortage of academic smelfunguses to forewarn prospective PhD students of the perils of academia!
I'm currently in the process of applying to computational biology PhDs in Cambridge and London, and I'm surprised to learn of your background going into your PhD. Maybe I'm mistaken, but I've noticed that the usual profile for candidates at the top PhD programmes seems to involve some combination of a first-class degree, a master's in a relevant subject, and ~1 year or more of post-university research experience, often accompanied by academic prizes. I finished my master's last year, graduated with a first, and am currently doing a second research internship in a similar field - just got my rejection email from EBI this morning, didn't even make it to interview.
Having said that, I don't at all anticipate that the experience of a PhD itself would be particularly gruelling - my experience so far has been actually quite relaxing, though maybe that's because I'm based in a rather tranquil research institute in London. The main drawback, in my opinion, is that I would lose 4 years during which I could be building myself up financially, and gaining momentum in a professional career. I don't see myself staying in academia partly because of the scarcity of positions, but also because it seems to inevitably terminate in a role that seems to mostly consist of meetings, admin, grant applications, managing people, etc.
I'm living hand-to-mouth and part of me feels exhausted at the prospect of continuing on this wage for 5 more years, but I really do love biology, and I don't think I can stomach pivoting into consulting/finance/law/etc in exchange for a higher salary. It just seems a terrible bore. I really wish the UK (or anywhere in continental Europe) had a life sciences industry comparable with the scale and vitality of that in the US.
Where were your undergrad and masters?
I had plenty of research experience from my undergrad/masters and interviewed well. One of my cohort was straight out of a 3-year undergrad though with no other experience though tbf. Perhaps our admissions committee was unusually good at seeing through CVs and assessing actual scientific ability.
Competition has likely gotten much tougher since covid as I think more and more people are reaching for the academic lifestyle too.
Why don’t you apply to a PhD in the US? Or look at startups/biotech. The more you lean into tech, the less important having a PhD becomes. I don’t know how good you are or your interests, but why not cold-email some lab directors at the Arc institute?
The financial strain of (1) living in the UK (2) doing a PhD is tough. That said, a PhD from Cambridge will likely be very interesting and you can make it fulfilling and worthwhile for the future. If you don’t want to stay in academia then the case becomes much weaker. Keep applying nonetheless to give yourself options, but I’d definitely be looking at other options if I were you
I did an integrated master's at Durham. I would be the first to admit that it wasn't the most rigorous course, and at least in my experience there were no opportunities to get involved in research during undergrad. I'm a home student though, so that isn't an issue.
As for biotech/startups in the UK - is it wrong of me to think that entry-level roles are overwhelmingly likely to be given to a PhD, or at least someone with an exceptional undergrad/Oxbridge background? I will also add that job postings from these companies seem far and few in between, and hiring seems mostly done through recruiters, which makes it hard to get a sense of the scale and diversity of the life sciences industry opportunities in the golden triangle. UK unis produce an excess of PhDs, relative to the number of permanent posts in academia - and I would assume that as a result, any biotech/pharma role is going to have PhDs lining up to apply. It's hard to see how I could continue doing something biology related without a PhD.
Anyway I think if I did do a PhD in Cambridge, it would just be really fun. And like you say it is beautiful. It sounds silly but that means a lot to me too!
Yeah if you want to stay in life sciences and the UK, a PhD may actually pay off the most in the long run. Cambridge is a fantastic city and the student lifestyle is great, so it's not a bad thing to do for four years. Sounds like you're honest about the pros and cons and so you'll probably make a smart decision either way.
For biotech, I think a better route is via direct/personal connection. Many companies create positions for people directly, and it'll get "advertised" for like 24h. Cold emails (how I got my postdoc) or meeting people at conferences/talks is an underrated method.
thanks for your advice! I appreciate it, fingers crossed
I agree that PhD students have become glorified technicians and it would be better if programs were smaller and more selective. I also think it would be better if they were less focused on productivity. Now students immediately jump into projects because they need to publish as many papers as possible to be competitive on the job market, and PIs need someone making progress on a funded project. But I think we would get more interesting, creative science if they first spent a couple years deeply learning a field by doing nothing but reading and discussing what they read.
At the same time, I don't agree about coasting along being lazy. I spend hundreds of hours drumming up funding for students, and if someone doesn't accomplish anything, they're not getting any more funding from me and they're on their own for the rest of their degree. (This is for deliberate laziness, not the inevitable setbacks that happen sometimes in research). This tends to be a good motivator. I also don't see the thesis as separate from the papers. My students' theses are usually just their publications copied and pasted into one word doc. If they did enough work to produce 3+ articles publishable in respectable journals, that seems like a perfectly reasonable PhD to me.
Thanks for your insight! I actually really like your idea of reading deeply before starting a project--although hard to monitor / easy to lose motivation and do nothing I suppose. My own PhD actually had a version of this naturally built-in thanks to COVID. My first 6 months were spent just reading and writing a literature review, and I think this was really valuable. Despite this, I still did nowhere enough reading in the first half of my PhD. So much time was wasted doing poorly designed experiments that I could've avoided by reading more literature! I like your approach to the thesis. That's exactly what my supervisor and I decided on too. "Don't waste time on your thesis--nobody reads it" was good advice I received from her.
A lit review is a great way to approach this. There's a fantastic quote that a month in the lab can save an hour in the library, and it's so true. I also feel like I'm barely skimming the surface of the literature these days. I wish we could all take a few reading weeks every semester where there's no classes, no meetings, no admin work, and no new journal issues come out - every academic just reads and talks to their colleagues about it. Dream on!
Interesting post. I think the PhD can feel easy if it's something that you're intrinsically motivated to do, but many who are not super passionate/curious about their topic of choice would find it difficult to complete. Did you feel this way about the lab work? And did you enjoy the experience overall?
This also makes me wonder about the differences between lab-based and humanistic PhDs. I remember chatting with college mates who were putting in 60+ hours a week in their lab, while I was in the Politics department in Cambs and was basically in control of my schedule for the whole 3 years (outside of meetings with my supervisor every few months.)
Funny enough, I played a bit with the CURUFC team during the beginning of my PhD (played college rugby for the most part) and definitely understand your point about rugby being part of the appeal of a PhD.
Yup it’s a tough challenge but not an impossible one I don’t think. There must be ways to set absolute requirements that are better/harder and more uniform across PhDs within a field/department and also adjustable between fields/departments. I think achieving a level of citations which are set relative to the field is perhaps one example. Eg 10 citations in life sciences, 2 in politics (I’m making these numbers up but they should be derived via a unifying calculation based on total number of citations within a field). Can be of one’s thesis or journal articles. The degree cannot be awarded until this point. The degree cannot be withheld after this point (by the supervisor, say).
The whole thesis and viva thing is way too wishy washy for my liking, especially given the skewed incentives to pass students and the hand-picked examiners.
Thanks! I really try in the essay to separate the PhD requirements and the PhD experience, but perhaps I could've added another paragraph or two. The PhD requirements are uniform and easy. The PhD experience varies drastically from person to person and can be very difficult. I would certainly say I worked very hard and found mine difficult, but, as you suggest, I found this hard work fulfilling as I was motivated to do it (I certainly wasn't forced to or expected to by either university requirements or my supervisor). In terms of enjoyment, I absolutely loved it. But it was overcoming the difficulties and frustrations that was itself enjoyable in a sense.
In terms of schedules, I was pretty much in control of mine too. In first year I was very lazy. showing up between 10 and 11am. Disappearing by 4pm on a good day, earlier if I could justify "working from home". I switched gears somewhere in my second year (quitting CURUFC actually due to injury and frustration with the club). By third year my basic routine was 9am until 8pm in the lab (eating lunch and dinner at the lab) Mon-Fri and then 4 or 5 hours on Saturday and Sunday. Sometimes I did more when experiments required it, sometimes less when my social life required it. I was probably around the 60 h/week mark for a good 3 years though (whole thing lasted 4.5 years for me). Other PhD students I know stuck to the 11am to 4pm shift throughout their entire PhD (also lab-based). In fact my lab mate had a friend in the chemistry department who we called "lazy Ben" because he was simply never in the lab and constantly messaged my friend to ask if he wanted to play tennis lol. Lazy Ben will still get a PhD.
I refrain from blaming the lazy Bens of the world though, because they are acting rationally within their incentive structure. The point is that nobody is pushed or motivated by the basic PhD degree requirements. It's always either self-motivated or pushed by a supervisor. I think if universities in general raised the bar of minimum requirements, then it would create less space for nightmarish supervisors and equally less space for lazy free-riders.
Well said about the pleasures/satisfaction of "overcoming the difficulties and frustrations." That experience (and the resilience it instills) is the thing that unites all people who've done a PhD, no matter the subject.
One thing that crossed my mind while reading your response: perhaps the standards for passing are low because it is hard to develop consistent "high" standards when PhDs in different fields are so drastically different. Once at a college bar I met a German zoologist whose PhD was basically about trying to understand the mechanism by which a centipede's feet adhered to the surface the centipede was walking on. It's almost impossible to compare that to my project (in Politics but largely a historical PhD based on extensive archival research) or someone in material sciences, for example. So going with the standards of (a) getting into the program + getting funding (which is highly competitive at Cambs at least) and (b) doing work that the supervisor sees as sufficiently good is an imperfect solution when it's really tough to develop consistent standards across dozens of fields.
Really interesting perspective - I couldn't agree more with your fundamental point.
What I DO disagree on you is that the solution is more bureaucracy when I think the real problem that talent is just not fundamentally valued in the UK (especially in academia).
It may sound ridiculous but I think the solution is actually making it far more encouraged to focus on startups and commercialising research. For example, at the University of Edinburgh where I am, the university can take up to 55% equity in a spinout company, whilst at places like MIT it's closer to 5%.
Instead of adding standardized tests or strict graduation quotas to make PhDs artificially harder, we should make the post-graduate startup environment more rewarding. If UK universities copied the US approach and let founders keep their equity, the quality of PhD candidates would increase naturally. The best students would use the degree as a starting point to create new technology, rather than using it to avoid the regular job market.
tl;dr instead of deterring bad students; encourage those with the most talent with better incentives
Great point. I’m inclined to agree. I remember going to this seminar at the LMB in Cambridge about spinning out research. Greg Winter (inventor of humanized monoclonal antibodies for therapies) was on the panel and gave a very bleak view of the entrepreneurial landscape. Had absolutely nothing good to say about the university nor the MRC. Knowing that strongly informed my decision to do my postdoc in the US.
That said, there are other issues with not deterring/preventing bad students from getting into supposedly elite programmes. People tend to lose ambition when they’re surrounded by others without ambition. You also lose the benefit of talent-concentration (eg, two talented would-be founders never meet because the class size is double what it should be). I expand on these issues here: https://sotaletters.substack.com/p/reclaiming-the-british-university
I do like your preference for carrots instead of sticks though tbh. And the startup environment could be massively improved in UK.
I agree totally, I come from US in CS background and I see basically the same set of issues, even at top schools. The top 1% undergrads probably get a job at a quant firm or an AI lab. The only ones who join a PhD even if they are the top 1% is that they are overly fixated on credentials, which is itself a negative sign.
The US system still kinda works, because there are enough talented people who do a PhD not just for the credentials, but that number is constantly decreasing. I would say it’s around 10-20% of top lab PhDs right now. The other thing to note is complete decimation of non top lab PhDs.
In 1960s, most colleges in US were competitive, they had good researchers who invented stuff. Matlab came from Uni of New Mexico etc. Now that is almost never the case. At best it extends to top 10 universities, rest of them are a wasteland. They publish for the sake of publishing with no advancement in science and even the top schools probably have a majority of such work. It’s all very disheartening and people say CS / AI is still one of the better fields in Academia.
I don’t know how this ship can be righted but making PhDs far more rigorous is one solution. > 70% of the students should fail the quals the first time and only 50% should eventually pass. Masters thesis standards should be increased. Claude Shannon came up with information theory in his masters thesis, now most thesis are worthless (mine included). You should just not be given a thesis if it’s not a real contribution.
Overall, I’m not sure if this ship can be righted from the inside
I used to link a PDF of my masters thesis on my CV. I re-read it recently and have decided to remove the hyperlink lol.
I think you're right. The problem is, any attempts to reform academia is seen as an "attack on science". Standardized testing at undergrad and postgrad levels would probably be the easiest and most effective way to start to right the ship.
To add to the chorus of US PhDs responding to this: I got mine in Canada and it was NOT easy! It took me 6 years to fulfill all the requirements of a four-year PhD (and still finished before many others who started at the same time). As an international student (from the Netherlands) I wasn’t even allowed to start directly in the PhD programme but had to start on the MSc track first and then transfer in the 2nd year (because the Dutch MSc that I did was shorter than the Canadian one). I had to publish one research paper before graduation (and complete a submittable draft of another) so I was also at the mercy of finicky journal reviewers and it meant that my success was partly based on doing at least some experiments that didn’t yield negative results.
At least it taught me that I didn’t like research enough to be dealing with that kind of stress and uncertainty all the time, so I’ve been happily working as a science writer and science communicator ever since.
I go back and forth on whether publications are a good requirement to graduate. On the one hand, it’s basically the one actual output of academic research so it makes sense. On the other, making this a requirement might incentivise paper mills and low quality publications.
In your case, I think it sounds like it worked well minus the time penalty. I agree journals and reviewers make the whole thing much slower and shittier. That’s a whole other issue that should be solved (which I’ve written on too!).
In my case, I’m a year into my postdoc and my PhD supervisor still hasn’t resubmitted my big paper to a journal. So perhaps having the publication as a hard requirement might have given her some more motivation! I think my career has suffered as a consequence because my CV for postdoc fellowships is weak as a result. Anyways.
Science is hard, so it should be the case that PhDs reflect that. The system worked in your case, but many slip through the net. Also—glad you introspected and got off the tram lines. As someone who writes in their spare time, I imagine you’re finding science writing much more fulfilling that what you were doing during your PhD!
I found this very helpful and inspiring. However, I must ask since it reflects my current stage in life. How hard was it to get into Cambridge for your undergrad? I’ve heard they require at least 5 scores of 5 AP classes, 1500 SAT, and phenomenal UCAS scoring as well. I’m in the US considering applying, and I meet about half the requirements, probably needing to retake a few tests to score better. But is it really that intense? In the US, essays usually matter more. Did you and your peers all meet these standards? Or are these overblown and exaggerated?
Sounds pretty tough. I can’t say I’m too familiar with the standards for US applicants. Back when I applied yes the standards were super high, although they varied by course. Medicine, maths, and natural sciences had higher thresholds than other courses. Even if get an interview and do well in it, the actual offer can also be brutal—especially if you went to a good school (you get punished to level things out with other candidates). I had to get 4 A*s for my offer. Most A level students don’t even take 4 subjects lol.
That said, I think you should apply regardless—there’s only one way to find out if you’ll make it. And if you don’t make it, dealing with rejection and failure is good for the soul. So it’s a win-win :)