Intro

Any decision that costs six years of your life should be considered carefully, so you should weigh the pros and cons before deciding to do a PhD. I think doing a PhD has been a good decision for me (as you might have inferred from the line on my “about” page that says “PhD candidate”), but it’s not necessarily a good decision for everyone. To help aspiring grad students decide whether a PhD is right for them, I’m writing two posts giving my best arguments for and against doing a PhD. This post will take the position that you should not do a PhD. As always, questions, comments, concerns, angry retorts, and any relevant sea shanties should be sent to me via Twitter.

This post is not comprehensive. Everyone has a different background, and the factors I consider important may not be important for everyone. There are certainly valid reasons to not apply to grad schools that vary from person to person. For example, uprooting your life and moving to a new city may not be feasible for you. I selected the topics I did because I expect they will apply to most people. Most people will make less money while doing a PhD, many find grad school emotionally difficult, and all PhD students will have to interact with the academic system.

It’s also worth mentioning my background as it informs my perspective. My undergrad degree is in bioinformatics from a program based in the computer science department. The alternatives to a PhD that I researched were primarily software engineer positions, so I may be more optimistic about what other choices look like than someone with a pure biology background. I’m also from the United States, so I had the privilege of being a domestic applicant to most of the schools I applied to. If you’re an international applicant, every step of the process becomes more challenging.

With the preliminaries out of the way, here are three reasons you shouldn’t apply to graduate school:

Money

Financially, you shouldn’t expect to make more money in the long run by doing a PhD, and in the short run you’ll make less. If you have a CS background, you can go into industry as a software engineer or data scientist and make more money than you will as a PhD student. If your skills are more biological, you could work at a pharma company instead.

That’s abstract though, let’s talk about numbers. As a conservative estimate, I looked up Pfizer associate scientist (an entry-level bio position) and software engineer salaries on Glassdoor. The mean salaries for the positions were $60k and $100k per year, respectively¹. On the other hand, a lucky grad student will make somewhere around $34k per year. Over the course of six year PhD, even if you had remained at your industry job’s starting salary, the difference is still between $150,000 and $400,000.

Imagine someone offering you a large check, to be paid in six years, and all you have to do is not go to grad school. Sure money doesn’t buy happiness, but it certainly can prevent unhappiness. A six figure decrease in student loans/increase in your net worth is a big change, especially if you’re just coming out of college.

Emotional Difficulty

One in four people doing a PhD program show signs of depression and one in five show signs of anxiety. Maintaining a healthy mental state in grad school is hard, and there are several factors working against you. Here’s a sample of those challenges:

Isolation

The goal of a PhD is to become the world’s leading expert in some narrow slice of your field. As you voyage out to the frontier of knowledge, the number of people you can talk to about your work decreases. You begin to have strong feelings on whether the manifold hypothesis applies to gene expression data, but few people to talk to about it. Eventually you end up in unmarked territory, seeing things daily that no one has ever seen. Like a rocket getting further from earth, the communication gap grows and grows. You find that to explain what you’re actually researching to someone, they’d have to already know what a variational autoencoder is and the difference in expression distributions between microarray and RNA-seq data. Instead you come up with simplifications to tell your family like “I teach computers to see whether people are sick” and nod sagely at the inevitable reply, “Ooh, that sounds hard.”

Ill-defined Progress

The purpose of academic research is to test hypotheses in order to learn more about how the world works. Because you don’t already know whether the hypotheses are true, you end up getting a lot of negative results of the flavor, “The interesting thing I thought might be true was not, in fact, true.” These experiments are still useful, because you gain knowledge about your research question, but aren’t publishable results². Research doesn’t come with a progress bar, you have no way of knowing “if I just run five more experiments I’ll finally understand what’s happening.” You just keep doing experiments until you realize how to solve a problem or decide you need to work on something else.

This lack of feedback is baked into the structure of the academic system to some degree. Instead of receiving frequent feedback in the form of grades, you write papers which are published months later. By the time you receive the good news that your paper will be published, you’ve already been working on a different project for weeks.

In high school I never understood how someone could do something as boring and painful as become a cross-country runner. In grad school, I found the “work harder -> run farther” feedback loop was a welcome reprieve from the “work harder -> ???” feedback loop as a PhD student.

Challenging Work-Life Balance

Grad students are paid a salary, not a wage. That is to say that if you work an eighty hour week, you will make no more money than if you went on vacation. However, the win condition for a PhD is to produce enough research in your field that you are acknowledged as an expert by your professor and thesis committee. As a result, you end up calculating whether the breaks you take are getting you well-needed rest or just delaying your graduation that much longer. The logical conclusion of these pressures is that many grad students end up working long hours.

The Reality of Academia

Academia isn’t a bad place (says the person with a vested interest in graduate degrees being considered valuable). However, there are bad reasons to go into academia. This section shouldn’t be read as “academia is the worst, go elsewhere”, but instead “here are some realities of the academic world, be warned.”

Not Just More Undergrad

There is a sort of person³ who enjoys the structure of undergraduate life. Learning new things within a defined scope, being evaluated on your knowledge in predictable intervals, and having well-defined time constraints is something they enjoy. To such a person the idea of leaving their undergrad program to live an entirely different lifestyle in the workforce can be terrifying. They might conclude, incorrectly, that the only way to avoid change is to pursue even more education.

While the trappings of PhD and undergraduate programs are similar, the actual experiences are very different. You do take classes for the first year or two, but the classes themselves aren’t like the ones in undergrad. In grad school, classes focus on reading papers, designing experiments, and communicating ideas rather than learning facts and taking tests. Where the objective of undergrad is to do well in classes, PhD coursework is just a mechanism for gaining information and learning skills.

Grad school is also less structured than undergrad. In undergrad you know a year in advance when the final exams that determine your performance are going to be. In grad school (as in life) you rarely know when the events with disproportionately large impact will be. You will look daily at experimental results, and the difference between discovering penicillin and throwing out a contaminated plate will be dependent on your skills and your luck rather than your preparation for that event.

The social life in grad school is different as well. What social life looks like is very program/person dependent, but in general your peers will be from a wider age range and will have more concerns than just what is happening at school.

The Academic Job Market

Even if the classic CS/bio jobs aren’t something you want to do, you should be aware that the academic job market is hypercompetitive. If your goal is to become a tenured professor, you will have to repeatedly move cross-country and take postdoc positions just for a ~15% chance at a tenured position. An average timeline is around six years of doing a PhD, 4.5 years of being a postdoc, and six years to get tenure. That is to say that even if everything goes well, you’re still not getting a tenured position until you’re 40+ years old.

Having a PhD can also work against you in the industry job market. While many technical jobs like to hire PhDs, the further away from the very specific thing you have been training for you get, the less relevant skills you will have.

Academia is Sales

A common reason to become an academic is to have the freedom to choose what to research, but often your freedom to research what you want is a function of your skill as a salesperson. Your success at every step of your job advancement (writing papers, giving job talks, presenting at conferences) is at least partially dependent on your ability to sell the idea that your work is important.

For example, as a grad student what you research is heavily dependent on your lab’s focus. If you want to do something else, you have to convince your boss that it’s a good idea⁴. If that fails, you’ll have to convince a panel of scientists to fund your ideas via a fellowship instead. If you’re fortunate enough to become a professor, you will ostensibly have control over your lab’s research direction. You will, of course, have to convince a granting agency to fund that research direction though.

As an academic, you’ll have an unparalleled ability to choose what you work on. However, that freedom and the future of your career are circumscribed by your ability to persuade others that your work matters.

Inability to Switch Workplaces

A PhD lasts around six years, and you’ll spend five of those years in the same lab(s). If you decide to leave your thesis lab, you will often have to change projects, leading to a delay in your graduation time. Since you’re locked into your work arrangement, and lab heads aren’t trained in management, you can end up in toxic lab situations that are difficult to escape.

Conclusion

PhD life isn’t all bad. In fact, I have an entire post about the good parts going up next week. If I have failed to dissuade you from doing a PhD, maybe you should read it.

Further Reading:

• “Everything I wanted to know about CS graduate school at the beginning but didn’t learn until later.” Some of the specifics about e.g. funding agencies don’t really apply to computational biology, but the broad strokes are good.

• Another take on whether or not to do grad school

• This analysis about postdoc positions

• “…although few PhD holders end up unemployed, it is not clear that spending years securing this high-level qualification is worth it for a job as, for example, a high-school teacher.”.

• shouldigetaphd.com - I haven’t actually read the book, but having listened to to the author talk about PhD programs before on podcasts I’m fairly confident the answer will be “No.”

Acknowledgements

Thanks to Rachel Ungar for suggestions and proofreading.

Footnotes