December 28, 2018

Scientists who leave academia aren’t dropouts. They’re scientists.

Posted by AGU Career Center

I’m a researcher turned policy wonk turned scicommer turned communications trainer and storyteller. Even though I’ve changed trajectories many times throughout my (so far) short career, one thing has remained constant: I am a scientist. However, this hasn’t always been how I felt. It’s taken me a long time to accept that I made the right decision. But sometimes, something happens to make me feel guilt, uncertainty, other-ness… Recently, this exact situation happened.

Me as a researcher, communicator, and both as a scientist.

I came across a PNAS (Proceedings of the National Academy of Sciences) manuscript entitled, Changing demographics of scientific careers: The rise of the temporary workforce. Intrigued by “temporary”, I dug in. I didn’t even get through the abstract before my blood started to boil. Before I get into the reasons for my shock and disgust, let’s talk about what the manuscript was actually (I think) about.

The authors set out to examine the relationships between number of publications by (primarily academic) research scientists, whether those scientists were lead or supporting authors, and how long they stayed in their research career. The authors found some trends between how productive (measured by number of publications) researchers were early in their career and how long they remained researchers. Additionally, they found that a growing number of researchers never transition from supporting authors to lead authors, though no relationship exists between being a supporting author and time spent as a researcher. The authors argue that this change in authorship structure (from lead to supporting, from single to teams), is a good thing and “critical to the production of contemporary science.” They also make the point that while universities and institutions don’t really reward supporting authors as much as lead, they should.

Lead tweet of the thread after ~24hrs live.

This is good stuff. Unfortunately, it took me multiple reads to realize that these were the main points/findings of the manuscript. Why? Well…the language.

The problem starts with the title: Changing demographics of scientific careers: The rise of the temporary workforce. That “temporary” word intrigued me, and rightly so. The authors define “temporary” as basically anyone who leaves academia/research. (NOTE: I’ll use academia/research(er) interchangeably for active-research scientists as most researchers are academics. I do this because the authors’ language reflects similar connotations.) Worse, per the language in the manuscript, “temporary workforce” actually means “temporary scientists.” Other problematic language exists throughout, starting in the abstract by referring to “careers of scientists” as only researchers, “scientific career survivability” as only careers in research, and scientists “full careers” as those who have authored multiple papers over a 20-year period.

Per that final point of career length, the authors placed scientists into one of five categories to measure “survival status” (i.e. how long they stay in academia):

  1. Transient: authors with a single publication
  2. Junior dropouts: authors who leave [research] <10 years after their first pub
  3. Early-career dropouts: authors who leave [research] 11-15 years after their first pub
  4. Mid-career dropouts: authors who leave [research] 16-20 years after their first pub
  5. Full-career scientists: authors who have careers >20 years after their first pub

“Dropout” hurts me personally. I did not drop out of science. I changed how I contribute to science, as have many of my peers and colleagues who are not active researchers. To say that all of use who pursued science, no matter if we have Bachelors, Masters, Ph.D.s, or more, have dropped out of science, is plain wrong. “Transient” might be OK when paired with “authors”, but even then, non-research scientists still publish, especially those in science communication and journalism. In some cases, they publish more than researchers.

A follow-up tweet after ~24hrs live.

The attitudes of some research scientists to their fellow non-research scientists is an area of ongoing conflict. As I was writing this piece, I tweeted out a small thread about the findings, language, and my attitude towards this study. The response was overwhelming. Research and non-research scientists alike came out of the Twitter woodwork to offer their support of a system where we as the larger scientific community don’t treat each other differently depending on our title or career path.

To anyone reading this piece, wondering why exactly I was so incensed when I read this manuscript, I have some suggestions of phrases to avoid/think about when talking about non-research scientists:

  • Avoid referring to folks who leave academic science as a “temporary workforce.”
  • A “scientific career” can be (and for many is) more than a career in research or working for a university.
  • Don’t use “survivability” as a term to refer to ability (or desire) to remain in academia.
    • Also, I know that it’s the name of the analysis, but literally using a survival analysis to study the “survivability” of scientists when referring to their research career is incredibly insensitive.
  • All of the scientific categories: transients, junior dropouts, early-career dropouts, mid-career dropouts, and full-career scientists.
  • Who’s measuring “premature” in “leaving the field prematurely”?
  • Leaving academia/research does not equate to “abandoning” a scientific career.

I get what the authors were going for. I often hear (and say) that there are too many Ph.D.s and not enough tenure-track jobs, but seldom do I see any citations to back that up. This manuscript was an exhaustive attempt to put numbers behind that statement and to track the factors that determine how long a scientist remains a researcher. And in some ways, it did. It showed that fewer scientists are staying in research. And it showed that those who do stay have fewer first-author publications. My primary issue is that the manuscript frames all of this as a negative, using terms like “dropout” and “survival” (and literally using survival analysis to analyze how long folks stay in academia).

Let’s stop framing non-academic/research scientists in a negative way. When thinking about non-academic career options, let’s switch “other” to “another.”


Shane M Hanlon is Program Manager of AGU’s Sharing Science Program. His views expressed here do not represent those of his program or employer.