I did some preliminary research on biomedical research as a career. The case for becoming a biomedical researcher looks to be weak for most candidates for the career. Are there important points in favor of pursuing a career in biomedical research that I'm missing?

Summary

• Some people find biomedical research very rewarding, but the job involves a lot of grant writing, not only research.
• Job security for biomedical researchers in academia is extremely poor before tenure. We still have to research exit options for those who leave academia.
• Biomedical researchers make substantially less money over a life time than they could in other fields.
• The job involves ~60 hours of work per week
• While biomedical research has historically produced a great deal of value, the situation today is more ambiguous, and it appears that the average biomedical researcher does little to advance the field.

The nature of the work

According to How to succeed in science: a concise guide for young biomedical scientists. Part I: taking the plunge by Yewdell (2009)

for individuals with a hunger for knowledge and an insatiable curiosity about how things work, science offers a constant challenge and, best of all, the intense thrill of discovery.  What can match being the first person who has ever lived to know something new about nature? And not just the big, infrequent, paradigm-making (or breaking) discoveries, but the small, incremental discoveries that occur on a daily or weekly basis too. If this doesn’t give you goosebumps and if you are not in a rush to get to the laboratory in the morning to find the results of yesterday’s experiment, then you should seriously consider a non-laboratory career.

However, research is not the only part of the job: Yewdell writes 

For your entire career as a PI, you will put inordinate efforts into writing grants

This is in consonance with GiveWell's post Exploring Life Science Funding which says

The existing system focuses on time-consuming, paperwork-heavy grant applications for individual investigators.

GiveWell's post also hints at researchers being constrained with respect to the research that they're able to get funding for:

The existing system favors a particular brand of research – generally incremental testing of particular hypotheses – and is less suited to supporting research that doesn’t fit into this mold. Research that doesn’t fit into this mold may include: (i) Very high-risk research representing a small chance of a big breakthrough. (ii) Research that focuses on developing improved tools and techniques (for example, better microscopy or better genome sequencing), rather than on directly investigating particular hypotheses. (iii) “Translational research” aiming to improve the transition between basic scientific discoveries and clinical applications, and not focused on traditionally “academic” topics (for example, research focusing on predicting drug toxicity).

Job security

Our writeup on job security in academia gives some general considerations.

Concerning biomedical research specifically, The Scientific Workforce Policy Debate: Do We Produce too Many Biomedical Trainees? reports that

During the period from 1993-2003, the probability that a postdoc in the U.S. was in a tenure-track PI position 5-6 years after obtaining their PhD ranged from 15-23% (Garrison and McGuire, 2007).

This graphic says that after finishing graduate school / postdoc, of biomedical research PhDs, 18% go into non-research science jobs, 6% go into government research, 43% go into academia or teaching, 18% go into industrial research, 13% do work outside of science and 2% are unemployed. Roughly 50% of those who complete a postdoc and go into academia get tenure, and the career outcomes for those who don't get tenure are unreported.

Some of the jobs that biomedical researchers get outside of academia are jobs that they could have gotten without doing a PhD or postdoc.

An important question is that of how correlated research ability is with job security. If luck plays a sufficiently large role then high ability doesn't guarantee a job, whereas if skill can overcome luck, then those who are skilled can be confident that they'll be able to get jobs. An interview with Prof. Andrew McMichael at the 80K blog seems to suggest that sufficiently high quality researchers can get jobs and funding. However, going into graduate school, one's ability level may not be clear.

It's unclear how job security is changing over time. In 2010, the Bureau of Labor Statistics reported that the number of jobs was expected to grow 36% over 10 years (much faster than average). But in 2012, the Bureau of Labor Statistics reported that the number of jobs is expected to grow 13% over 10 years, and in the intervening time the number of jobs had grown only 3%. So there appears to have been a substantial change in outlook in only two years. The job growth rate forecasts have to be viewed in juxtaposition with the expected change in number of new PhDs. According to one source, the National Institutes of Health found that the number of new PhDs increased by 50% between 2002 and 2009. If this rate were to be sustained, the ratio of jobs to job candidates would decrease even more.

I plan on researching exit options 

Work-life balance

According to Yewdell (2009) 

As a graduate student, you should be spending a minimum of 40 hours per week actually designing, performing or interpreting experiments. As there are many other necessary things to do during the day (for example, reading the literature, attending seminars and journal club, talking to colleagues both formally and informally, and common laboratory jobs), this means you will be spending 60 or more hours per week in science-associated activities.

This is corroborated by career coach Marty Nemko, who wrote 

You spend most of your 60-to-70-hour workweek alone in a lab or at your desk, with little people contact.

Biomedical researchers who stay in academia are often constrained with respect to the geographic location where they can get jobs. See our writeup on job location options for academics.

Earnings

Getting a PhD in a biomedical research field takes 6 to 7 years, during which one makes substantially less money than one could otherwise make. It's been reported that the average biology PhD had $45k in debt as of 2004.

Salaries rise afterward, but not rapidly: as of 2009, the starting salary for a postdoc was ~$37k/year (pg. 141), and postdoctoral appointments last 4 years.

According to the Bureau of Labor Statistics

Colleges, Universities, and Professional Schools are next in employment, and pay a mean wage of $61,320 per year. Completing the five areas with the most employment are Pharmaceutical and Medicine Manufacturing ($92,130), General Medical and Surgical Hospitals ($80,090) and Drugs and Druggists' Sundries Merchant Wholesalers ($93,090).

The "Colleges, Universities, and Professional Schools" category includes postdocs: if one considers professors only, the figure will be more like $80k/year.

According to Yewdell (2009)

If you do achieve the ‘Holy Grail’ of full professorship then you will not be poor, but you will be far worse off financially than nearly all of your peers who have similar levels of talent, energy and dedication, but who chose other careers.

Career coach Marty Nemko wrote

"According to MIT faculty member Philip Greenspun, Adjusted for IQ, quantitative skills, and working hours, jobs in science are the lowest paid in the United States...."

A small number of biomedical researchers command high salaries: for example, one source reports that there are 20 in the country with earnings at the $240k+ level.

Some sources report that biomedical researchers can become very wealthy if as early employees of successful biotech startups, but this is very rare.

Social Value

Society desperately needs your talents [...] For rationally thinking people with an altruistic bent, life can be no more rewarding than when practising the scientific method for the benefit of all of the denizens of this fragile planet.

Some points to keep in mind in assessing the social value of biomedical research are

• Diminishing returns — Much of the increase in lifespan between 1950 and now was due to cardiovascular disease research, with the gains mostly halting by 1990. There have been significant advances in recent years, such as AIDS treatment drugs, statins, psychiatric drugs. But one should expect the increase in quality of life and lifespan per researcher to go down over time, because of low hanging fruit being plucked, barring radical advances coming from anti-aging research and unexpected sources.
• Low replication rates — The fact that large fraction of studies don't replicate suggesting that much research doesn't move science forward.
• Power law distribution of research contributions A small fraction of researchers produce 100x+ as much value as the average researcher. To the extent that success is driven by skill rather than luck, prospects for impact depend heavily on your ability.

80,000 Hours plans to publish an overview of biomedical research that will address the social value of going into biomedical research in more detail.

See also

Biomedical Research Workforce Working Group Report (2012) by the National Institutes of Health.

How to succeed in science: a concise guide for young biomedical scientists. Part I: taking the plunge (2009) by Jonathan Yewdell.

Cross-posted from the Cognito Mentoring blog