Are you considering leaving academia, but you’re afraid of not being able to go back if you change your mind? Believe me, I understand your fear.

I don’t have any tattoos, I backup my files regularly, and I appreciate having an “undo” button above these lines as I write them. In other words: I always try to avoid irreversible changes. So, when I started pondering leaving academia to become a data scientist, that same fear was holding me back. But is leaving academia really an irreversible career change?

The short answer

Of course, a definitive answer to this generic question would require gathering data and thoroughly analysing it. But I’m in a lazy Sunday afternoon mood (plus I’m not an academic anymore!). Instead, I’ll simply consider the chances of going back to academia for someone in my shoes: an ex-astrophysicist who has been working in industry for over 3 years. Please note that I’m just discussing whether I would be able to go back to academia, not if I would like to (that’s another interesting question I may cover in a future post).

One does not simply walk into academia

Academic jobs and grants often explicitly require the “ability to obtain funds” and a “proven record of academic publications”. Therefore, I now rank lower than many other researchers who have been publishing papers and applying for grants while I was busy solving “real world” problems in industry (speaking of which, please check out this other post). Of course, I could submit a hundred applications around the world, and probably get at least one positive answer. But the chances of having an offer that would suit my personal and professional expectations are pretty low.

So, as time passes outside of academia, I see that “Prof.” before my name drifting further and further away… Unless there was a way to keep publishing papers on the side, right?

Is it possible to write papers in your free time?

While publishing papers on the work done for a company is possible, it is a rare luxury (I’m not considering private research for this post). So any publication must come out of your free time.

If you are reading this, you probably know what publishing a paper involves. Among other things, submitting papers to a journal is not a simple “upload pdf” button. You have to deal with the responses from referees (ranging from insightful all the way to arbitrary) and editors. Also, unless you are affiliated with a research institution, your access to those journals is limited, or otherwise expensive. On top of all that you have to keep up with the literature, do the actual research and write the paper itself! All the while bearing in mind that publications are basically irrelevant in industry.

Therefore, publishing papers is a big investment of time with very little return. So you may as well free yourself from the burden of publishing papers, which is one of the perks of leaving academia!

In my case, when I left academia I had a few first author papers in the making, and several follow-up ideas. How many of those papers did I publish? None. I have a pile of drafts, messy code and data that never went public, and probably never will. However, I have happily participated as a co-author in several publications while working in industry. But the more I disconnect from research over time, the lower I see my chances of further contributions.

So, given that I haven’t published enough papers during the past few years, my professional value in academia has decreased… Which automatically makes me a worse researcher, doesn’t it?

The talent diode

Paradoxically, I think my value as a researcher has actually increased in industry. I’d be more proficient now at doing research than before. Sure, I feel a bit disconnected from the latest findings, and I would need some time to catch up in my field. But that extra time would easily be outweighed by the new skills I’ve acquired in industry.

Academics who move to industry bring great value. Why can’t talent flow in the opposite direction?

In a similar fashion, I have met many non-academics who would excel at research. Industry workers (in particular data scientists) often require as much analytical thinking as researchers. But they usually have a stronger need to extract incremental value from ongoing projects, communicate results to people with different professional backgrounds, prioritise which problems to solve first, write efficient and robust code… All of which would be invaluable assets in academia.

And when I think of some of my most prolific astrophysics colleagues, they were also savvy in technical skills like programming and data analysis. If such academics moved to industry, they would not only be extremely valuable, but they could also develop those skills more effectively. However, after some year working in industry, they would not be able to bring those new skills back to academia.

One-way sign.

Clearly, academics who move to industry bring great value. Why can’t talent flow in the opposite direction? The recruitment process in academia creates a one-way road (or a diode, if you also fancy electronics), allowing talent to flow from academia to industry, but not the other way around.

Conclusions

I think it’s fair to say that going back to academia after some years of industry work is unlikely. The main reason being that if you can’t keep up with publishing papers, your value as an academic decreases. And publishing papers in your free time is just not a realistic option. That being said, I have met a handful of people who did go back to academia after some years of industry work. But they are definitely the exception. Put simply, going back to academia (or even entering academia from a purely industry background!) is very uncommon.

The way academia recruits creates a “talent diode” that lets skilled academics move to industry, but impedes talent flow back to academia. And in my opinion, this not only restricts the career choices of non-academics, but ultimately also slows down scientific progress.

So, if you are thinking of moving to industry, don’t be afraid, there are plenty of good reasons to do so. But just like getting a tattoo, make sure you think it through thoroughly: There’s no way to hit “undo”!