I recently decided to quit my PhD after taken a year of leave!

My decision was mainly motivated by the fact that I felt much more happy outside of academia and that I don't care so much about the PhD title (I had about ~1 1/2 year left).

There are of course many other parameters that influenced my decision.

One thing I was surprised about is the messages I've received from other PhD/Master students asking for advice on whether they should also drop out. One of them particularly resonated with me:

I would like your advice on something and I'll try not to take too much of your time.

I recently read a post where you mention quitting your PhD. I am in my first year PhD in Comp. sci.

I'm considering quitting PhD and going for industry directly instead.

My question is: based on your experience, do you consider having a PhD an asset when going for that type of career ?

Thank you for taking the time to read this, I really appreciate it

I'll formalize in this post my opinion about the dropping out subject, which is very subjective to my personal experience.

TLDR;

yes, quit. It's not an asset.

Overview

It's not as black-and-white as I made it out to be in the TLDR, but if you lay out the good and bad sides of the PhD experience, everything points towards quitting.

I still cherish the time I've spent working on my PhD. I've learned a lot. However, if I was to remove all the part that were absolutely useless to my growth I would have learned a lot more.

Actually, I think I would have learned way more just volunteering in a lab for free and working part-time.

Let's start by the big overarching problem with doing a PhD! 🙂

Problem with doing a PhD

This list of problem I had with doing a PhD is of course not exhaustive and not even necessarily accurate.

The amount of admin work that needs to be done is mind-numbing.

This is the part that I disliked the most. It's not so much that there was a lot of it to do, just the fact that it was there was an issue for me.

By far the admin work that was the most dreadful was the grant writing. I can't think of an exercise I hated more than spending 2+ month at the beginning of the year trying to convince the government to give me money for my PhD.

If we were talking about millions of dollar in funding that would have been a whole different experience for sure. But here we are talking about highly competitive grant with the intellectual cream of the crop of the whole country... for like [20K, 50K] CAD a year.

The main issue I have is that the whole exercise is kind of pointless. In life-science in Canada all graduate position are paid by the supervisor anyway. Which means that you already have the base of 20K a year for your work at the lab (thankfully... some PhD in other field are just not paid).

Usually what happens is that if you apply for these grant and you get them, it's your supervisor that gets the money and then you get some part of it (or none). That means that if you get the 21K grant from NSERC, most PhD student will effectively gets nothing (my PI gave me a small fraction of the grant as a thank you).

This kind of makes the whole exercises feels like a drag because it's so tedious. You need to understand what is the angle to write the grant, do multiple version of it, learn about best practice for writing good grant, etc.

This issue is tightly tied to the second major problem for me in a PhD.

The pay is atrocious for the hours put in.

The pay for a PhD student is ridiculously low, 20K CAD a year in Montreal, Canada.

This creates a very negative environment where your whole life is kind of on hold since you won't be able to financially grow. You can't realistically take other jobs without working yourself into a burn-out since a PhD is a full time job.

In some case, it's more than full time. I had PhD friends in molecular biology routinely working on the weekends and 50h+/week was the norm.

Imagine going into a job for [4,6] years for ~20K a year with limited potential of making more. That amounts to 120K gross for your whole PhD. By the way out of that you have to pay tuition for some reason 👍!

If you are in computer science, this is the worst financially move that you can make. A starting position in some institution is 120K CAD/year with near certainty of growth every year. By the time you would be done with your PhD you would have made near 1M$ more and have a very stable career.

This is the part I had the most problem with because the people who didn't mind the pay where mostly students whose family were financially well-off. Having had to bootstrap my way through undergraduate and working through every summer since I was 18, the pay for me was a huge deal.

It further crystalized the feeling that I did not belong in a PhD since I was struggling a lot while other felt like it was completely normal. Yes, you can take debt in order to pay for a more comfortable life while you are doing your PhD, but the prospect of finishing a PhD with a mountain of debt made no sense to me.

This isn't even the worst case. International student needs to pay way higher tuition in order to be in the same program and finding a part-time job is sometime not even an option.

Other PhD position are not paid. You need to pay in order to get in the program with no guarantee that you will get that title.

The worst possible case is the unpaid position where you are obligated to work as a teacher assistant for your supervisor.

This kind of stressful financial environment just plain sucks if you are poor.

Speaking of stress...

The pressure to published is demoralizing.

I published 7 papers during my short time in academia. I wasn't too worried about papers since I had great colleagues working with me to get our research out there.

However, I've seen plenty of other researchers struggling routinely to get their research published in order to meet their quota for graduating.

Not only that, but when the pressure came from their supervisors to just publish something it made the whole experience very stressful for them.

Sometime, experiment leads to nowhere or the result are too ambiguous to do anything about it. Other time data collection might be stalling for lots of reasons (i.e. difficulty recruiting, data quality being poor, limited access to equipment, ethics taking a long time to clear out, etc.).

When you have that big Damocles sword above your head to publish, it makes getting a PhD a pretty bleak journey.

What is also adding on that stress is the review process...

The review process for publishing is broken.

Oh boi where to start.

Having to fight against the review process is a Kafkaesque endeavor.

You never know what will happen. You send the papers to Journal X, but it gets desk-rejected. You tweak the title and the abstract, it can now get into Journal Y desk.

Now the review process start, but one of the reviewers think your project suck and don't understand why you had the audacity of sending it for publication.

Some other reviewers give helpful pointers about what they want you to add to the experiment (after taking 3 months to write a review).

Now you go into a political session of politely telling them that you don't care about some of the comments or actually doing some of the analysis requested.

You iterate on the experiment, tweaking it so that there is the least amount of work on your side (i.e. no way we are going back to ethic, it will take 3 more months for sure).

Some of the reviewers start agreeing with what you bring back to the table (after 2 months to read the new version).

One of them is adamant that this thing plain suck. Even after the improvement they reject the whole paper.

The editor send a nice thank you letter telling you that he is sorry it didn't work out.

Now you take your beefed up paper and send it to Journal Z. Thanks to the review process you now have a whole bunch of more analysis added to your paper, some of them were just there because previous reviewers asked for it. You don't think that it adds much value to the narrative, but you keep them just in case.

This process repeat from journal to journal or conference to conference. It is so slow, cumbersome and stochastic that it suck out the life of you. At the very end, right before it gets publish you start to question why you started this experiments in the first place.

Then your paper get accepted and your whole lab celebrate! 🎉

You will invariably start to question why you need to send to a journal in the first place, then you learn about what is up under the hood of academia...

The research publishing industry relationship to academia is weird.

You need to publish your experiments. Yet, you can't just post them out there on Reddit and have other researchers gives you all sort of feedback.

You have to go through the process of working with big journals. The bigger the journal (in term of impact factor) the more valuable your research is, because this is what supervisor are graded on to get grants.

The twist is that the journals use volunteer researchers in order to get those reviews going. Volunteers do so because it is a positive signal for grant agency to see that a researcher is a reviewer for the prestigious Journal X.

However, all of these people reviewing have 100s things to do, including pumping more paper into the system in different journals. Meaning that reviewing papers isn't their top priority.

At the end of the day, the grants could come straight from a consortium of executives in the research publishing industry and it wouldn't make a difference.

As a researchers you are implicitly working for the publishing industry, which is very weird.

Speaking of working for someone else...

The career opportunity are not clear at all.

It truly is very confusing. There are very little job in Academia and depending on your research field it might be the only path in terms of career opportunity.

If getting a grant is competitive, imagine all of the ones that got multiple of them competing for even fewer seats in an academia track.

Researchers doing post-docs over post-docs to get just another chance of being accepted at their faculty of choice is mind blowing.

It never was my plan though so I don't have much insight here, but the fact that I knew there was just no way I could get a job in this current setting after I was done made the pull from the industry that much stronger.

So if career prospect are not great in academia, you are surely at least helping advance your field and help people! (right?)...

The research you are doing will most likely not help the population you are studying.

This was one of the most demoralizing aspect for me as I was doing consciousness research.

It hit me hard when I was working for a new analysis coming from a review we got. As I was looking at my external hard drive, I realized that 50% of the people I've collected electroencephalographic data from were dead.

Meaning that what I had in my hard drive was the last remnant of what their brain was doing deep into their coma. One of them I'm pretty sure still had a brain capable to support conscious before being scheduled for organ-donation.

Yet, the source of that data is long gone and whatever I was doing at the moment to get it published wasn't for them.

I understood that in the grand scheme of thing the field I was studying will improve, but I 100% felt like a sham sitting there in front of my computer trying to please reviewers #3.

Benefit of doing a PhD

I've focused mainly on what I considered the negative side of the PhD. Yet, there was plenty of cool benefit from doing a PhD! It just didn't weight as much as the negative in the end.

Get paid to learn awesome stuff.

If you don't mind the monetary aspect, this is great. You get to learn the cutting edge aspect of whatever you are studying and really immerse yourself in that learning!

On top of that you get a small revenue stream!

Take part in solving important problem.

At least in the grant proposal, it's always cutting-edge-human-race-defining research problems that you are solving. In the practice it's a bit different, but being part of something larger than yourself is a fantastic experience.

This is well on display in conferences where you can see all of these bright mind etching away at a problem you are also part of. I loved having all of these great discussion in front of posters or in-between talk.

Get to play with big toys!

Some of the tools you are using are impossible to get your hands on outside of a research lab. For instance there is very few MEG available anywhere in Canada, but at a few research facility.

Being able to geek around these science thing for your experiments is amazing!

Same with molecular biology reagent. When I've learned how much some of the liquids I was using in my first ever experiment cost I freaked out.

Lots of useful translatable skills are learned.

The main one is to be able to write properly and convey your ideas clearly (which as you can see I haven't mastered at all).

Data science skills also comes to mind as a very transferable and desirable outside academia.

The one skill though that has been the most useful for me is the ability to figure stuff out on my own when no one can possibly help me. The problem you are solving are usually too niche for someone to take the time and tell you "go this way and you will find what you are looking for".

When asking people for help you might get contradictory guidance, there might also be just wrong because they don't understand the problem as deeply as you do.

Same with the technical aspect of your studies.

Some math that I needed were  so cryptically written that they could have been as well an incantation from the Necronomicon.

Transitioning towards industry research or even entrepreneurship is a natural path out of academia!

Alternative way of Getting the Benefits

So the benefits are great, but they aren't exclusive to a PhD/Master degree. You can actually get them outside of academia simply by being driven.

Volunteer in a lab :)

Before starting my master and PhD I've been volunteering at labs for 3 years. My experience was much different because I didn't care much about either the grant proposal aspect or paper publishing.

I was just some kid running experiments and learning in the process. This is highly valuable for a lab, even if your experience is limited. Ironically I got 3 of my 7 papers published from past work I was involved in without much effort.

Contribute to open source endeavor.

If you like coding, there are plenty of way you can help out advance any field by just contributing to open source. Most field in academia rely on a core backbone of open source codebase which are maintained by volunteers (or not maintained at all).

Helping out making these open source project better is a sure way to make a big impact on the life of many researchers and getting result out faster.

Literally work on whatever problem you want to solve outside of academia.

This was an eye-opener to me. I could just do whatever I want and work on issues I think was important to solve. I got that 'ah-ah' moment as I was sitting next to an entrepreneur starting a business in the neuroscience field on a topic adjacent to me.

Yes, you are left on your own, but at that point the freedom to chip away at a problem you deem relevant is exhilarating.

Conclusion

In my view, the conclusion is pretty clear. There is so much reason why a PhD shouldn't be done and only one that support it:

because I want to do a PhD with this Professor in this particular lab

If this is the case, go right ahead and have some fun while you are at it. Don't take what I'm saying too seriously. My perception is biased by my experience and the learning I've done along the way.

The most important thing you should take out of this is to do a retrospection about why you are doing a PhD and then take a decision that you are happy with!