The current system of funding and performing biological and medical academic research in the United States has produced excellent science. The field has been built over the last 75 years as emerging technologies have made molecular biology possible, and it is rapidly growing. However, a federal funding plateau over the past decade coupled with personnel saturation at all levels of the research totem pole has led to inertia in hiring practices at universities. A cascade of consequences now renders the practice of academic research unsustainable, as long as it remains trapped in a funding and hiring framework designed for rapid growth rather than for a steady state. In the long term, this calls for serious reform; in the short term, graduate students and postdocs must embrace careers beyond the heavily trodden academic path.
First, after more than sixty consecutive years of significant increases, the National Institute of Health (NIH) budget – the primary funding source of biological and medical academic research – abruptly stagnated in 2003 and has remained roughly constant since. Second, universities are incentivized to grow their science programs by adding faculty and buildings to their campuses, because both are indirectly reimbursed by federal funding. Additionally, as tenured principal investigators (PIs) remain in their posts through old age, they maintain fiefdoms of graduate students and postdocs, whose turnover rates dwarf those of professors. This unhappy confluence breeds reliance on graduate students and postdocs as cheap labor before their release into a bloated job market while professors enjoy lifelong posts in precious few job positions.
Problematic consequences plague students, postdocs, professors, and the academic research system in general:
The average time to PhD completion for graduate students in biological and medical research is currently seven years, which is a long time to be confined to a notoriously low stipend that ranges from about $11,000 to $32,000 annually, depending on local cost of living. (Graduate students in the sciences are typically prohibited by their universities and PIs from holding other jobs, with the occasional exception of tutoring for the university.) For this and other reasons, including general stress and getting “scooped” by a competitor who publishes similar work, the dropout rate in these programs in the US is as high as 44%.
Even so, only about a third of those who begin PhD schooling graduate and find a next-level research position in academia, the private sector, or the government, which remain saturated despite the low PhD graduation rate. Only about 10% of PhD students will ultimately secure a tenure-track professorship, if current conditions persist. Clearly, too many people begin PhD programs, which then lose many of their entrants, and then churn out the remainder into a crowded job market. While one might sardonically point out that the high dropout rate at least culls the glut of unemployed PhDs, that does nothing to help the fact that young scientists struggle to apply their skills and passion to a career.
Fortunately, several other occupational choices for graduate-level scientists do exist – those will be mentioned later. However, the requisite skills for anything other than academic research are rarely taught in graduate school. Non-academic job postings and recruiting are relatively rare. Most frustratingly, declaring interest in and discussing “alternative” career paths with one’s PI is widely considered taboo – so much so that it’s likened to coming out of the closet. PIs are frequently unsupportive of a student’s decision to abandon a perceived academic legacy; they may withhold mentorship or letters of recommendation to potential employers.
Performing more research in a second lab after earning a PhD is almost always an essential step to becoming a PI or even to securing any science-related position. This level of the totem pole is most tightly choked by the hiring bottleneck in academic research. The average length of a postdoc is undocumented, but as of 2010, 30% of erstwhile postdocs had made a lateral move to a second postdoc rather than acquiring a professorship, and 10% of recent postdocs were unemployed. Postdocs are also poorly paid, typically earning less than $45,000 per year. That modest hike in salary relative to the PhD student’s helps little during a phase of life in which many people hope to marry and raise families, and it disregards their extensive education and experience.
In 1980, newly minted PIs waited one year on average for federal funding; in 2009, the average wait was four to five years. A separate survey revealed that, regardless of wait time, the average age at which an American professor in biology or biomedical sciences will receive his or her first federal grant is 42. Even for established, successful professors, sufficient funding is never guaranteed.
The effects of a lean budget on a crowded industry are pervasive. Only the grant submissions that are deemed to be in the top decile commonly receive NIH funding; “top” typically translates to short-term, low-risk proposals, which are most likely to generate publishable results. Creative approaches, long-term cohort studies or evolution experiments, and high-risk/high-reward science are often passed over, to the detriment of scientific discovery. Moreover, frequent grant rejection means that professors, postdocs, and students spend a lot of time applying for grants and fellowships rather than actually practicing research.
When a funded, successful project’s results approach publication, researchers who think ahead to grant renewal often sensationalize their results, vocally link them (however tenuously) to medical significance, and ignore irreproducibility in their experiments. Squabbles over authorship rank are the norm, as are attempts to exclude minor contributors from authorship to avoid the appearance of diluting the major authors’ efforts. On the other hand, projects that yield negative results – “A does not correlate with B”, or “Drug X does not influence tumor size” – are considered low-impact and often aren’t submitted for publication at all, which leaves other scientists in the field to waste time and resources in order to arrive at the same conclusion.
On the other side of publishing responsibility lies peer review. Funded PIs are periodically assigned submitted papers relevant to their field to check for adequacy. However, professors whose schedules are strained by grant writing often delegate this responsibility to postdocs or students, who may lack the breadth of knowledge to evaluate a manuscript’s contribution to their fields.
Tight competition also brings the journal hierarchy – the self-reinforcing idea that certain journals publish “better,” more important research than do others – into sharp focus. In a rough job market, a strong publication record is crucial for an aspiring postdoc’s or professor’s employability; it’s tantalizingly simple for universities and labs to select a candidate whose CV boasts publications in Cell, Nature, or Science. As we will discuss in an upcoming PhDISH piece, though, this hierarchy is unhelpful to subscribe to and dangerous to perpetuate. Fundamentally, it tends to ascribe value to publications based on the reputation of the lab publishing that information and not only on the merit of the research.
A handful of respected professors in the biomedical sciences have considered the progressively worsening state of academic research and have offered plans to mend it. Two cogent reviews, “Rescuing US biomedical research from its systemic flaws” by Bruce Alberts, et al. and “A fair deal for PhD students and postdocs” by Henry Bourne, itemize causes of and potential solutions to financially strained, hypercompetitive research. The recommendations below are theirs unless stated otherwise.
Young scientists should be encouraged to complete a PhD and a postdoc only when absolutely necessary for subsequent work. To this end, PhD programs can uniformly decrease their acceptance rates and award Master of Science (MS) degrees about two years into the program, thereby creating a formal branch point. Armed with a master’s degree, students who are better suited for or interested in careers other than research should accept ‘honorable discharge’ from the program to pursue said careers. Alberts, et al. recommend that these programs offer courses related to consulting, public policy, and other potentially science-related jobs in order to expose students to and prepare them for such options, but Bourne disagrees; with such low graduation rates to contend with, he posits, MS/PhD programs should first ensure that at least research skills are well-taught.
Second, the NIH should separate research grants for labs’ projects from training grants for individual PhD students. Notably, this would require extending eligibility for NIH funding to foreign students. This move would give students greater liberty to choose a lab based on research interest rather than on that lab’s financial situation. (Speaking from personal experience, a small pay increase wouldn’t hurt, either). Those training grants, furthermore, should be limited to five years, to remove PIs’ motivation to retain graduate students as cheap labor for as long as possible. The NIH might even consider preferentially funding research institutions whose biological and biomedical PhDs are shorter on average.
Unquestionably, a gradual increase in the average postdoc’s starting salary is in order. Limiting postdocs’ pay to five years would likely confer the same benefit as it would for students. After five years, postdocs should earn the salary and title of staff scientist (see below). Additionally, NIH administrators would be wise to catalog the number of former and current postdocs in the US and which career paths they choose, for reference.
Shortening and clearing the path to professorship will reduce the average age at which PIs receive their first federal grant.
A much-needed redistribution of laboratory responsibilities could occur with the widespread adoption of staff scientists. Staff scientists are semi-permanent or permanent fixtures in a lab who hold an MS or a PhD in science. They may perform their own experiments or assist with others’ or maintain a resource like a database or a mouse colony, and they can train students and postdocs. Such positions exist somewhat rarely in academia, but significantly increasing the ratio of staff scientists to postdocs and students will stabilize the revolving door of the latter high-turnover employees, particularly by making training more uniform and efficient. Increasing the number of these positions also introduces a financially stable career choice for those who wish to remain in academic science without directing an entire lab.
With respect to the NIH’s belt-tightening, no clear solutions beckon. However, even if the federal government cannot promise greater largesse in the foreseeable future, it would be very helpful for researchers and their institutions to have access to a federal prediction of the next five to seven years’ budget, updated annually. The capacity to anticipate financial means is still an improvement over a laboratory budget that’s both stretched and uncertain.
The peer review process deserves changes as well. PIs stereotypically shirk the responsibility of serving on a review board more than once during their careers. Thorough critical review of submitted manuscripts is essential for quality control, and more frequent participation in the review process should be enforced, punishing noncompliance with revocation of publishing privileges or of funding. Review boards must be carefully constructed of PIs from various fields so that a given area – cell biology, say, or development – isn’t overrepresented on any board.
More broadly, NIH research grant disbursement must be modified. Allocating equal subsets of funding to early-, mid-, and late-career professors, rather than favoring the longest-standing labs, is necessary. In science, because any given line of interrogation may not yield interesting results, previous success is a poor indicator of future success – particularly when prior ‘failure’ is all but guaranteed on a slim budget. And although it would be difficult to incentivize, it would be beneficial in the long term to increase the proportion of funding for creative and long-term proposals and their potentially deeply rewarding findings. Ideally, also, grant applications should be considered based on their qualitative recent and potential significance, not on the impact factors of the journals in which applicants have recently published.
With respect to prevailing attitudes in academic science, current notions of feudal hierarchy in the lab are the most harmful, because they romanticize top-down authority and ladder-climbing at the expense of collaboration. I have personally witnessed postdocs and graduate students getting barked at by each other and their supervisors, being forced to compete with one another for a project assignment, and generally expected to act as inexpensive labor rather than as well-educated teammates with agency and creativity. It would be refreshing to embrace a more communal structure akin to those in some western European institutions (and in the engineering field in general), in which multiple PIs may lead a team of other professors, students, and postdocs; staff scientists abound; and collaboration with other groups is extensive. On the publishing front, perhaps biomedical scientists could adjust to the concept of submitting manuscripts bearing numerous – even dozens of – authors, as often occurs in physics, engineering, and even in genome sequencing. On the way to publication, keeping publicly accessible lab notes would be a brave step forward in the spirit collaboration and of practicality (see Andrew Lee’s post The Sweet Taste of Open Science).
REAL CHANGE: Fund people, not projects
In a heartening acknowledgment of the academic research struggle, the NIH is developing a new funding option: the Maximizing Investigators’ Research Award, or MIRA. The MIRA is designed to fund all of a lab’s research that pertains to the NIH’s NIGMS, or National Institute of General Medical Sciences, for five years at a time, provided that recipient labs devote at least half of their research efforts to the MIRA-funded work. Awards will be distributed in amounts of $150,000 to $750,000 per year and may be modulated annually. Additionally, a lab that holds a MIRA will not be eligible for a second MIRA or an R01 grant, which is currently the NIH’s highest-paying research grant.
The idea behind the MIRA is to prevent small numbers of labs from acquiring disproportionately high funding[, to stabilize funding in the midterm, and to afford researchers some much-needed freedom to explore different hypotheses as they arise in their general line of interrogation, rather than remaining beholden to the rigid, prescient ‘specific aims’ demanded in a traditional grant application. MIRAs are meant to be distributed evenly between conventional low-risk and more creative high-risk, high-reward proposals.
Like any funding scheme, this one has its risks – MIRAs might end up dropping in the laps of perceived academic elites, or the award’s criteria, once established, might be otherwise biased. However, a recent request by the NIGMS for feedback from PIs about the concept of the MIRA generated an “overwhelmingly positive” response. At a recent Open Session of the National Advisory General Medical Sciences Council meeting, the MIRA was greenlit for development, with the first grants scheduled to be offered in the 2016 fiscal year.
Of course, this discussion of the research pipeline is not exhaustive, and solutions to its shortcomings are not entirely obvious. Many of the problems above are rooted in “ivory tower culture,” in which, broadly, the top prioritization of information and discovery can be supplanted by ego. Pervasive attitudes like these are difficult to dismantle. The most important thing for participants in the academic research system is to be aware of its current constraints and their effects on career options. PhD students and postdocs must understand that despite their PIs’ and their own expectations of becoming a PI, landing a tenured academic research position is less likely than, collectively, any other job options, as illustrated by a popular infographic. However, those who hold an MS or PhD in the biological and medical sciences are free to consider a wealth of fascinating career choices, including:
· science journalism
· public policy
· research in industry
· research for the FBI or the CDC
· forensic science
· biology- or medicine-related startups
· research in other areas of science, including geology, paleontology, and biostatistics
· teaching high school or college students
· research and development for large firms such as Unilever or Proctor & Gamble
· editing at a scientific journal
Scientists at all levels of experience can effect change by contacting their school dean, government representative, or submitting an editorial to a newspaper or scientific journal of their choice. Pursuing a job in policy is a long-term means for passionate advocates for change to alter the way that the government funds science.
Of course, attaining tenured professorship in academia is a privilege. Not all scientists deserve to become a PI simply by virtue of completing a PhD and a postdoc. However, especially in the current financial climate, there are so many professional outlets for scientific excellence that narrowly defining success as academic ladder-climbing is counterproductive. A healthier body of American researchers begins with embracing this concept.