Why I'm doing a PhD
I’ve just started a PhD in Behavioural Science at Warwick Business School. People who have thought hard about how to make a difference seem to disagree about the value of PhDs. Having thought about this quite thoroughly for my own situation, I’ve decided to write up my decision process. Hopefully some of the considerations that were relevant for me will be generally applicable and useful to others making similar decisions.
Essentially, I’m doing a PhD because:
I want to use my career to do as much good as I can. However, I’m quite uncertain about which causes are most important and what I should do with my career long term. This means I want to spend the next few years learning and building “career capital” to keep my options open for whatever is highest impact later.
I believe that the PhD I’m doing is the best way for me to do this right now because:
It gives me the opportunity to build skills across a variety of disciplines/areas, whilst expanding my network and also giving me credentials that will help me later
At the same time, the research itself could be valuable - I’ve got a lot of flexibility with what I focus on, within an area that has the potential to be very important and useful (improving rationality/decision making)
I’m fairly confident I’ll be able to work on other high impact projects during the next few years alongside my PhD - volunteering for 80,000 Hours being just one example
I graduated from Oxford University last June with a Masters in Mathematics and Philosophy. I’ve spent the past year working full time for 80,000 Hours. During that time I’ve come to realise two things. First, that maximising the positive impact I have on the world during my career is much more important to me than I’d previously realised. Second, that my personal preferences and interests are much more malleable than I thought.
Together these two realisations mean that I’m basing my career decisions mostly on what I expect to be of most value to the world. This isn’t me being incredibly self-sacrificial: it’s because I care both about my impact on the world and my own happiness, and I expect that focusing on the former is actually also a pretty good way to achieve the latter. I’ll outline my reasoning for this in more detail in a later post.
How I’ve structured this decision
I’m assessing my career decisions both on the extent to which they allow me to do good immediately, and on the extent to which they might enable me to have more impact later. This second consideration is what at 80,000 Hours we call career capital, and is roughly broken down into the skills and knowledge that you build, networks that you create, and credentialization (or “CV points”) that you gain, all which will help you to take high impact opportunities later. There’s a more thorough explanation of this framework for assessing the impact of a career in a recent blog post on assessing the impact of a career.
I’m also thinking about my career on a number of different levels. I think too often we think of career choice as a one-dimensional process, a single “choice” in which we decide what to do for the rest of our lives. As I’ve written about previously, I think this is a mistake, and it’s important to think about your career on various time horizons and varying levels of abstraction. The starting point for me in thinking about my career was therefore to clarify somewhat my standards for impact: what do I ultimately care about, and a little more concretely, what problems or issues in the world do I want to work towards solving? From here I started thinking about what I might be aiming to achieve with my career long-term, which then provides a basis from which to evaluate the different “next step” options available to me. In this post I’ll talk a bit about these higher-level considerations: what I am working towards and want to achieve long-term, as it provides some important context for understanding my near-term career decisions.
What do I ultimately care about? What problems or causes do I want to dedicate my career towards?
Understanding what you ultimately value seems incredibly complex, and I’m not going to get into the nitty-gritty of various different ethical theories here. I’ll just say that what I roughly care about is people and animals being better off, and it doesn’t matter who they are, where they live, or whether they’re even alive at the same time as me. I’m sceptical about how useful fleshing out these views in more detail is, but I’m also happy to be challenged on this if you think I should!
What I find more useful is thinking about what causes or problems I think are most important and effective, and thus where I should be aiming to focus the impact of my career. Thinking about these kinds of questions often ends up forcing you to think about the higher-level questions about what you value anyway, so can be a good way to flesh out these views. Should I be worried about humans going extinct in the future? Well, it depends whether I value people who are going to be alive in thousands of years time as much as those who are alive now, amongst other things.
I’m sympathetic to the argument put forward by Nick Beckstead that we should be concerned with the impact of our actions on the far future, rather than focusing purely on the near-term effects of our actions. I also think it’s plausible that investing more research in understanding which causes are most effective is one of the most important things we can be doing right now (although I remain fairly uncertain about how tractable this is.)
In general, whilst I think it’s useful and important to think about causes when making career decisions, I’m very wary of fixing my sights on a specific cause early on that might narrow the path I take. Cause selection seems to be a relatively new and unexplored area, and I’m not confident that we know enough now about which causes are most effective to work on, let alone which will be most effective in 10 or 20 years time. This uncertainty means two things for my career in the short term. First, I’m likely to do work that focuses on “meta” causes, which are more robust under uncertainty. This potentially includes working on things like improving peoples’ rationality, altruism or ability to cooperate, spreading the ideas of effective altruism. Second, it means I’m weighing “career capital” substantially higher than immediate impact right now, and focusing on doing things which keep a number of options open, so I’m in the best position to work directly on whatever I think is most important later on. 1
What might I be aiming to do with my career long term?
I’m considering a range of things: working for 80,000 Hours or similar organisations, founding or co-founding a startup or nonprofit, going into academia and trying to become a public intellectual, or perhaps working in a foundation to influence the allocation of resources - but am keeping my mind open to a number of possibilities. Like with causes, I’m wary of narrowing too early based on too little information - it seems like the world of work will look very different in 10 or 20 years time and it’s very difficult to predict which jobs will make the most difference.
Again, this means that keeping my options open and building generally useful skills, experience and credentials is a high priority for me at the moment. It also means that being in an environment where I’m able to think and learn more about what I should do longer term is incredibly important.
So why the PhD?
The biggest competitor to doing the PhD would be to stay and work at 80,000 Hours.
As I wrote about in this post, I think working for 80,000 Hours is a really great career opportunity if you want to make a difference - both in terms of building career capital and immediate impact - so this was definitely a non-trivial comparison. In what follows I’ll go through a number of factors which I used to compare these two options and talk briefly about how each scores on each, before summarising my decision by integrating these different considerations.
My immediate impact
Working for 80,000 Hours has fairly clear immediate impact: I’d be directly helping others to have more impact with their careers. Doing a PhD, on the other hand, seems to be mostly valuable insofar as it prepares me to have more impact later, be it in research or elsewhere.
One really difficult question a number of people seem to face is how to trade off impact now against building capacity to have impact later - there’s no clear answer to this question yet. I outlined some reasons above why I personally lean towards focusing on career capital above immediate impact. But of course, you can have both, and I think working for 80,000 Hours is actually a pretty good way to get both. However, I also think there may be more potential for me to have immediate impact during my PhD than average, for the following reasons:
I’ve got a fair amount of flexibility with what I focus my research on, and have been having conversations with a number of people to decide what the most valuable thing for me to do might be. I’m optimistic that there are areas of research within my department and funding that could be valuable - improving rationality and decision making, altruistic behaviour or coordination/group decisions, for example. One of my main priorities now is thinking about this and talking to people more to clarify my direction. (If you think you might be able to help me with this, please get in touch!)
I’m confident that if I manage my time well, I’ll be able to volunteer and/or work on side projects during the time I’m doing my PhD Initially, I expect to continue working for 80,000 Hours in some capacity - likely continuing doing case studies and/or research. Later on, I might want to get involved with and/or even start my own high impact projects.
Overall, I think that 80,000 Hours is probably better on immediate impact, but that the difference might not be as big as it first seems.
Building career capital
I initially thought skill building was one of the biggest reasons for doing a PhD. In academia, your time seems to be explicitly focused on learning and building skills, whereas in most jobs often day-to-day work takes over from explicitly working on skill development. My department is particularly interdisciplinary, meaning my programme should give me the opportunity to build skills across a number of fields including psychology, economics, programming, statistics and business.
However, I’ve since become more uncertain about this. Obviously any assessment of skill building needs to ask the question, “What are these skills useful for?” I’m concerned that a lot of the skills you focus on during a PhD might not be that useful for things outside of academia: and I’m by no means certain that I want to be an academic. It seems plausible that the skills you’d be developing in some jobs might be more broadly useful. Over the past year working at 80,000 Hours, I’ve been able to develop some really valuable skills I might not have got in an academic setting, such as management and public speaking.
It’s not clear to me at this stage whether doing a PhD or working at 80,000 Hours would be better for my skill development. I think there are a number of useful skills I’ll be able to learn during my PhD - designing studies, writing academic papers, and analysing data for example - but it’s also possible that the more “practical” skills I’d learn at 80,000 Hours like project management and leadership would be more valuable, especially if I decide not to stay in academia. Given that I’m not convinced that skill building alone provides a compelling argument for doing a PhD, I think it’s important I think more deliberately about using my time (both inside and outside of my studies) to build valuable skills. The advantage of graduate school over working for an organisation is that how I spend my time will probably be more under my control, so it should be easier to explicitly set aside time for skill development outside of my studies.
I’m often wary of using the word “networking” as it has some negative associations - of schmoozing and using people. But meeting and interacting with people who are doing impressive and interesting things, from whom you can learn and find new opportunities, is clearly a really important thing to consider in your early career.
I’ll first note that as I outlined here I actually think 80,000 Hours is really goodfor this. I’ve met more impressive people doing exciting things in this past year than the rest of my life combined and I now feel I have a much better network of people to ask for advice and learn from than I could possibly have hoped for.
I am genuinely concerned about leaving 80,000 Hours and potentially losing this network. However, I also believe that my PhD will give me the opportunity to meet new, exciting, and possibly more experienced people, which will benefit me hugely. Since I’m in a highly interdisciplinary department based in the business school, I’ll be able to interact with academics from a number of different departments, as well as students and professors from the business school. There are also a number of links with policy which could provide some interesting connections. I believe that if I make the effort, I will be able to both retain and expand my existing connections over the next few years, as well as expanding and diversifying the people I interact with through the PhD. 2
These considerations also make me lean more heavily towards doing the PhD over staying at 80,000 Hours, although there is substantial worry about losing or weakening ties my existing ties. Overall I think doing a PhD will be better for developing my network, though only weakly so.
I think that general credentialization is a strong reason in favour of doing a PhD. People often use qualifications and achievements as a heuristic for assessing your competence, and having a PhD simply sounds impressive and makes people take you more seriously. It also may open up some later career options that would be much harder or even impossible to pursue otherwise. Becoming an academic, working for certain foundations, or working in research management are just a few examples.
On the other hand, working for 80,000 Hours might give me better credentials for working in a startup or nonprofit, or working in fields where practical knowledge is more valued than academic skills. I’m concerned nonetheless that as a young organisation, the lack of recognition might narrow down my options later. Additionally, I’ve arguably already got some of these credentials from working for 80,000 Hours over the past year.
It seems like the PhD provides me with better credentials for some things I might want to do later: becoming a public intellectual or influencing the allocation of resources in a foundation, whereas working at 80,000 Hours might be better for others: working for a startup or organisations like 80,000 Hours. This seems like a fairly even split, but I think the balance is tipped slightly in favour of the PhD. This is because, of all the options I’m considering, I think the PhD rules out the least: even if 80,000 Hours might be better for some. It doesn’t seem like doing a PhD would rule out the option of working for a startup or 80,000 Hours, whereas not doing one and staying at 80,000 Hours would make it very difficult for me to pursue a career in academia or a foundation.
Career capital overall
I think credentials and diverse networks are sufficiently important that the PhD wins overall in terms of building career capital. However, there is certainly a risk of weakening ties within my existing network, and of spending time building skills that may not be all that useful for things I want to do later, which I need to try to compensate for by focusing my research and spending my time outside of my studies wisely.
Skill and interest
I said earlier that I think my skills and interests are actually pretty malleable, so admittedly I didn’t give them a huge amount of weight when making this decision. That said, how well skilled I currently am for each option, and how much they interest me, do both seem relevant to how well I do my job, so seem worth considering. In this case, though, I think there’s very little difference: my background and experience probably make me about equally skilled for both, and I don’t feel a strong motivation or interest for one over the other.
I’m uncertain about a number of things related to doing a PhD: will I be good at/enjoy research? Will I have as much flexibility with my research as I expect at the moment, and I will I actually be able to do research that is itself valuable? Will I have enough time to focus on other things beyond my PhD for the next few years?
All of these questions seem hard to answer beyond an initial point without just starting the PhD and finding out.
Remaining questions and uncertainties
Very few decisions in our lives are made with anything close to perfect certainty: essentially we reach a point where we have to decide whether it’s worth getting more information, or to just make the decision using what information we have now. I reached the point with my decision where I felt there were probably diminishing returns on analysing it further, but this doesn’t mean there weren’t still uncertainties. The biggest remaining ones include:
To what extent I’ll actually have the flexibility or ability to do valuable research during my PhD
How much free time I’ll have during my PhD to work on other projects
How replaceable I am at 80,000 Hours
When thinking about our careers, there are several different levels of decision to be made. “Should I do a PhD?” is a top-level decision, but there are a number of lower-level questions that might be equally important if I want to have as much impact as possible. I think sometimes these lower-level questions can get neglected, since thinking only about the “big” career decisions, once we’ve made them we think we’re done. Part of the reason I bit the bullet and forced myself to make a decision on the PhD vs. 80,000 Hours, despite remaining uncertainties, was because I thought it was actually pretty important to start thinking about some of these lower-level issues, which include:
What research area/question should I focus my PhD on?
How can I make the best use of my time during my PhD?
How can I ensure I have time to work on other projects?
What kinds of other projects should I work on? Should I volunteer for 80,000 Hours/similar organisations, or try and do my own projects (e.g. a blog, a startup, other projects)
Should I try and do internships or work experience during my PhD, and where?
(As above, if you have any insights that might help me answer these questions, or are interested in discussing them, I’d love to hear from you!)
I’m planning on writing up a post soon to discuss some of the more generalisable considerations for when it’s valuable to do a PhD that I’ve learned from this process. But some very brief takeaways that I’ve found useful:
PhDs may not be as useful for skill building as they seem initially: it’s worth thinking about other ways you might build the same skills you’re hoping to get from your PhD with a lesser time cost
Doing a PhD carries a pretty big opportunity cost if your aims are altruistic, especially if there’s a particularly unusual opportunity to do good in the short-term. Because of this, I think it’s only worth doing a PhD if a) there are future career options that require you to have a PhD which you’re not prepared to rule out, or b) you think there’s a good chance you’d be able to work on other projects with more immediate impact during the programme.
This said, I am uncertain about whether this is the correct weighting of factors. There are plausible arguments that in some cases immediate impact might dwarf future impact which concern me, and I want to think about more when I make more specific decisions of what to do in the next few years.↩
An additional point worth mentioning is that I have some concerns that 80,000 Hours and related organisations, whilst full of smart and inspiring people, may be at risk of suffering from lack of experience and diversity. As I’m fairly young and don’t have a lot of real-world/work experience myself, I’m afraid that not interacting enough with other groups of people - especially older and more experienced ones - may lead to blind spots in my own views.↩