Why do a PhD?
Answers to FAQs and misconceptions about doing a PhD
I often talk to students who are thinking about higher studies and over the years I’ve found that many of them have misconceptions about why you should or should not do a PhD. In this article, I’ve tried to capture the most common advice I give: What are the benefits (if any) of doing a PhD, busting the common misconceptions about doing a PhD, and answers to other questions I frequently get asked.
Skip this article if you’re not at a stage in your life (early in your career, or midlife crisis) where a PhD makes sense. (However, you probably know others who are in that stage, so please share this post with them.)
Warning: This article applies to doing a PhD in Computer Science from a highly-ranked college. Much of it probably applies to other engineering branches. Most of it probably does not apply to PhD in non-technical fields. Also, Much of it probably does not apply to if the college is an average one.
Before getting into a detailed discussion of PhDs I think it is important to realize that after completing a Bachelor’s degree, you can pick one of three major directions:
MS/PhD: higher technical studies for those who want to stay in their technical field for a long time,
MBA: for those who are less interested in being technical and more interested in a large company career,
Job: for those unsure or who want to do a startup or need the money right away.
I’ve written a more detailed article on Job vs MBA vs MS that you should read if you aren’t sure which of these applies to you. I have also written articles related to choosing undergraduate studies. You can check those out here, here, here, here, here, here, here, and here.
For those interested in a career in technical roles for at least the next 10-15 years I recommend not only doing a Masters in a good university but also considering a PhD seriously. This article is for them.
Benefits of a PhD
The Brand: Having a PhD on your resume, especially from a well-known college, gives you a brand name that stays with you for the rest of your career. 30 years later, after having various other (more important) achievements on your resume, the PhD causes people to look at you a bit differently and opens doors more easily. Joining a large company, joining a startup, asking angels and VCs for funding, making the first few sales of your startup—in all these contexts, the fact that you have a PhD will be considered, discussed, and given weightage.
Communication Skills: The ability to communicate with people from different backgrounds is one of the most important skills in almost any career. During your PhD you repeatedly have to explain different pieces of your thesis to people with varying levels of understanding of your chosen topic and you’re forced to get good at quickly judging the audience and customizing your explanation for them. This includes your grandmother and the famous Mr. Gupta who lives next door—the one whose son has a lucrative job in Bangalore. It’s a tough audience. You get good at it and this pays rich dividends for the rest of your career.
Experiencing The Dip: Doing anything non-trivial in life—something that is going to take a lot of time and effort—has a familiar and frustrating trajectory. Initially, you make a lot of quick progress, but soon it either gets frustrating or boring or both. And one of the keys to long-term success is knowing how to get past this stage instead of giving up. Doing a PhD is one of the few things where you can learn to deal with The Dip in a conducive environment where everyone around you is either going through or has gone through the same dip. It is far more difficult to learn this once you join the industry because of the much shorter-term focus of the goals and assessments. I’ve written a more detailed article on The Dip.
Note: Gaining expertise in the topic/area of your PhD is missing from the list of benefits. Why? Because the area doesn’t matter. Are you doing a startup for online reputation management? A PhD from Stanford will give you instant credibility. The fact that the PhD was in VLSI will not matter to anyone. Most of my friends who did their PhDs with me are not working in the same area now.
Misconceptions
A number of students stay away from doing a PhD because the word “PhD” brings up some common associations which are not necessarily true.
Myth: After a PhD you have to become a professor and/or work in academia
Just because you do a PhD doesn’t mean that you have to become a professor in a college. A majority of CS PhDs don’t work in academia. They work in the industry.
Myth: After a PhD you have to do research
Just because you do a PhD doesn’t mean that you must do research for the rest of your life. A majority of CS PhDs who join the industry work on applications and implementations of technology. A number of them end up moving to non-tech business-oriented roles too.
Myth: A PhD results in loss of earnings
Many students feel that doing a PhD means losing 5 years of earnings and that the starting salary of a PhD isn’t really that much higher than classmates with a bachelor’s degree + 5 years of experience. This is incorrect thinking for two reasons.
First, your career will span roughly 40 years, with income increases compounding. The earnings of the first 5 years are trivial compared to the total career earnings. What matters isn’t the amount of money you make in the first year or the 5th year, but rather the slope of the graph. And doing a PhD (in CS, from a good college) (arguably) puts you on a different curve with a higher slope.
However, if you really want to or are in circumstances that need you to optimize for earnings right from the word go, then probably doing a PhD is not the right choice for you. There are many other ways of making lots of money. Doing a PhD is really about optimizing for interests—you’ll make good money and you’ll get to work on more interesting things than the others.
Myth: Doing a PhD is difficult, painful, and frustrating and/or PhD advisors are slave drivers who exploit their students
Reading social media, many students get the impression that doing a PhD is a horrible experience where your advisor has full control over your life and tortures you for his own benefit with little or no upside for you. You are paid peanuts and in the end, it is impossible to get a decent job.
I don’t know if this description is true for PhDs in non-technical fields or for PhDs from average/low-ranked colleges. But this is generally not true of a PhD in CS from a good college. I know a lot of PhDs who fit this description and almost all of them had a good time doing a PhD. Most CS PhDs in good colleges will get tuition-fee waivers and enough of a stipend to live comfortably (assuming you’re single at least for the first 3-4 years). Most advisors in CS in good colleges are not assholes. They want their students to succeed and find satisfaction in their students’ success.
Other Questions and Comments
How to pick a branch/area of CS that will have “scope” in the long term?
You can’t and you shouldn’t. Most PhDs don’t end up working in the area of their PhD. And, as I mentioned in the benefits section, the most important benefits of a PhD have little to do with the topic/thesis or even the area. Computer Science just moves too fast for that. My advisor, who was one of the top experts in the world in databases, had done his PhD in VLSI! Of my 5 colleagues who did their PhD in databases around the same time as I did, only one is still in databases proper.
Don’t focus too much on the specifics of the area. Instead, focus on learning the right general principles and learning them well, and on making sure you have a good advisor who will guide you through your ups and downs.
Doing a mid-career PhD
I often run into people with 20+ years of experience who feel that they should do a PhD. Often, they feel they’ve done shallow things all their career and would now like to go deep in some subject. The increase in status associated with a PhD is also sometimes a draw.
I am generally not in favour of this. A much better way to (force yourself to) gain expertise (and status) in some area is to write a book. It will give similar benefits as doing a PhD without all the bullshit requirements that the university bureaucracy imposes on the PhD process. (In some cases, people want to do a PhD because it is a requirement for something they want to do, like becoming a professor in some college. In that case, go ahead and do the PhD.)
For much more details on the actual nuts-and-bolts of doing a PhD, check out Shriram Krishnamurthi’s Getting a Computer Science PhD in the USA FAQ (Shriram is a Professor in the CS Department at Brown University) and Neeldhara Misra’s Twitter Thread on Advice to incoming PhD Students (Neeldhara is an Associate Professor, Computer Science and Engineering, IIT Gandhinagar.
Other questions?
Do you have questions about doing a PhD that I haven’t covered? If yes, please leave a comment and I’ll try to answer it.
Thanks to Meeta Kabra for reading drafts of this article and editing it.
Navin, have you ever written around research itself, as in how one does it?
Hi Navin, thanks for the article. It's so good to read it and clarifies many lingering doubts. Meanwhile could you please advise if its okay to do PhD while working or should one do it full time? My goal is to work in academic environment eventually.