Stabilizing Funding Cycles

Top

00:00:07.29 Hello. I'm Michael Teitelbaum from the Labor and Worklife Program
00:00:10.28 at Harvard Law School.
00:00:13.06 I'm not a lawyer. I'm a demographer.
00:00:15.10 The Labor and Worklife Program
00:00:17.15 is very interested in the careers of scientists and engineers,
00:00:21.01 and that's what I have been writing about recently.
00:00:24.17 The first thing I want to say is that
00:00:27.18 the research enterprise is a critical component of the US economy,
00:00:31.04 a very important leading edge of the economy
00:00:33.16 and of all developed countries, really.
00:00:37.09 And in my humble opinion,
00:00:39.14 biomedical research is the most exciting area of basic scientific research
00:00:43.29 out there right now.
00:00:46.19 But there are puzzles
00:00:49.10 about the biomedical research enterprise.
00:00:51.28 It's one that is very heavily funded...
00:00:54.04 massive budgets,
00:00:55.23 and yet it sees itself as experiencing funding crises
00:01:00.19 that seem to continue.
00:01:02.10 And while it has the highest potential
00:01:06.03 in terms of research outcomes,
00:01:09.13 it is presenting rather unattractive career prospects
00:01:13.00 for recent entrants into the field.
00:01:15.15 Well, where does US biom...
00:01:19.13 basic biomedical research fit in the world?
00:01:21.15 It's very productive.
00:01:23.05 It's been the world leader from World War II.
00:01:25.13 And it is strengthening, not weakening.
00:01:28.29 Part of it was that the US came out of World War II
00:01:32.24 in much better shape than most other developed countries,
00:01:35.20 but part of it was very wise US government policies
00:01:38.22 which facilitated the rise of basic research
00:01:41.18 to a global leader.
00:01:44.27 And this was particularly due to decisions by the government
00:01:50.01 to increase the amount of funding provided for scientific research
00:01:53.03 but to direct the funding, in the form of grants,
00:01:55.29 to research universities
00:01:58.20 rather than to government research labs
00:02:01.10 that would be built and would become freestanding.
00:02:05.07 And US funding is very large.
00:02:07.28 It is the highest level of funding of all large countries in the world,
00:02:14.01 both in absolute terms, or dollar terms,
00:02:15.27 and as a percentage of the gross domestic product.
00:02:18.14 And it represents a half of the total funding
00:02:21.24 of all developed countries.
00:02:23.07 So, it's large and import... imposing.
00:02:26.01 Unfortunately, there's also a long history of cycles in this funding,
00:02:30.18 which I describe as a cycle of alarm, boom, and bust,
00:02:35.20 in which the alarm is first sounded by somebody
00:02:38.10 that there's a shortage of funding or the US
00:02:42.28 is falling behind its competitors, followed -- with a lag sometimes --
00:02:47.13 by government action to increase the support
00:02:50.05 or the numbers of people in these fields,
00:02:51.28 and then a loss of interest after a few years,
00:02:54.22 and a bust in this system.
00:02:58.11 And there have been five rounds... surprisingly...
00:03:00.13 five rounds of alarm, boom, bust
00:03:02.15 in scientific research and funding
00:03:05.22 since World War II.
00:03:07.22 From World War II to 1957... first... that first round...
00:03:11.14 that was mostly in physics.
00:03:13.13 From '57, when Sputnik 1 was first launched,
00:03:16.27 to the 1970s.
00:03:18.27 From 1981 to the early 1990s,
00:03:22.05 which was heavily related to the defense build-up
00:03:25.21 under the Reagan administration.
00:03:27.28 And then those were...
00:03:31.02 as you can see, they were Cold War kinds of cycles.
00:03:33.18 Followed by two cycles that were not Cold War related...
00:03:36.13 after the Cold War,
00:03:38.26 the first having to do with high-tech industries
00:03:40.26 booming all together in the late 90s,
00:03:45.02 and then all busting all together around 2001 and subsequently.
00:03:49.27 And then an unrelated boom and bust in the biomedical research field,
00:03:53.25 which had to do with funding for National Institutes of Health
00:03:58.27 from 1998...
00:04:01.15 under the perception that there was a funding crisis,
00:04:04.11 the political system agreed to double the budget of NIH
00:04:08.19 in only a five-year period.
00:04:10.29 That doubling was completed in 2003.
00:04:13.20 The budget increases went from 14 or 15% a year
00:04:17.05 to 0% a year,
00:04:19.05 and in some cases declined,
00:04:21.25 and this led to a second crisis in funding,
00:04:25.13 even though the budget was by then almost twice as large
00:04:28.29 as it had been only five or six years earlier.
00:04:32.13 So, this is a very successful system in research terms,
00:04:35.07 but it's hardly a stable system,
00:04:37.28 as you can see.
00:04:39.15 And the core of the source of instability, in my view,
00:04:42.14 is positive feedback that's built into the system.
00:04:45.08 Nobody planned it that way, but it's there.
00:04:48.10 If you increase research funding,
00:04:51.02 automatically you will increase the number of slots
00:04:53.25 for PhD students and postdocs
00:04:56.02 who can be funded by the increased research funding.
00:04:59.16 And that will happen even if there's no demand...
00:05:03.06 increase in...
00:05:05.07 for PhDs after the PhD and postdoc have been completed.
00:05:10.04 At the same time, research budgets have grown,
00:05:12.21 grant success rates have declined.
00:05:14.15 And let me just show you...
00:05:16.16 I'll show you in a moment what that means.
00:05:18.07 So, let's first talk about the positive feedback loop
00:05:21.18 for PhDs and postdocs.
00:05:24.03 Most of these PhD students and postdocs
00:05:27.24 are funded as research assistants
00:05:30.27 on research grants.
00:05:32.25 And you see some numbers there from NSF and from NIH
00:05:35.26 which I think make that point.
00:05:38.13 And then, as I said earlier,
00:05:40.16 if you increase the research budget,
00:05:42.23 you'll increase the number of PhDs and postdocs,
00:05:45.05 even if there's no demand increasing...
00:05:48.09 no perceived demand increase for PhDs and postdocs,
00:05:51.22 who will then face weak career prospects.
00:05:55.22 Now, aren't there negative feedback loops
00:05:58.01 that are stabilizing as well?
00:05:59.21 The answer is yes, there are.
00:06:01.09 After all, talented US students in these fields
00:06:04.07 really do have alternatives.
00:06:06.04 They can opt for other careers and other degrees:
00:06:09.08 in law, in business, in medicine.
00:06:13.25 And those who have completed their PhDs and postdocs
00:06:17.15 and have started their research careers
00:06:19.27 can shift out if they find the funding situation
00:06:22.09 too unpleasant.
00:06:25.07 But those are pretty destructive and counterproductive
00:06:28.19 negative feedbacks.
00:06:30.08 They are... they are destructive to the system,
00:06:34.00 not helpful to the system.
00:06:35.14 And even if they take place
00:06:38.00 -- and they probably are taking place --
00:06:39.20 the overall system still does not have to adjust
00:06:43.27 to the new reality,
00:06:46.05 because there is a large and increasing
00:06:50.29 international supply of PhD students and of postdocs
00:06:55.21 who can be recruited and who can be supported financially
00:06:59.16 under the US research grant system.
00:07:02.09 There are no limits on the number of such postdocs, PhD students,
00:07:08.17 and funding under research grants.
00:07:12.15 So, here's a graph of the NIH research budget
00:07:16.05 from 1960 to 2012.
00:07:17.23 I think you can see...
00:07:20.09 the dark bottom line is the absolute... or...
00:07:24.00 the current dollars, unadjusted for inflation.
00:07:26.15 The other two lines are different ways of adjusting for inflation.
00:07:29.09 I think you can see what the trajectory has been.
00:07:32.05 And the rectangle, the red rectangle,
00:07:34.17 shows you the period of the budget doubling for NIH,
00:07:38.11 when those lines went very sharply upward.
00:07:42.29 And then it's not been so great since.
00:07:46.03 And one of the...
00:07:48.01 this is a graph showing what's happened to the success rate
00:07:50.27 with the increased funding curves you've just seen,
00:07:53.29 again going back to the early 1960s.
00:07:57.19 The blue line at the bottom is the number of research grants
00:08:01.25 that have been supported financially.
00:08:03.14 The red line is the number of proposals
00:08:06.10 that have been submitted.
00:08:07.29 And if you compare those,
00:08:10.14 you'll see what the green line is,
00:08:12.04 which is the success rate of submitted proposals to NIH,
00:08:16.01 which has been in a downward trajectory, pretty much,
00:08:19.21 since the 1960s, with one major exception:
00:08:23.27 a slight increase during the period of doubling
00:08:26.25 that is outlined there;
00:08:28.17 that's that rectangle.
00:08:30.16 Now, these booms and busts are not benign.
00:08:32.27 They cause harm for PhDs and postdocs
00:08:36.03 who are in a long period of study.
00:08:39.07 They may enter at a time
00:08:43.04 when career prospects are very promising,
00:08:45.12 and by the time they finish,
00:08:49.03 their career prospects may have waned
00:08:51.11 through no fault of their own.
00:08:53.04 For research faculty, they may begin their research careers,
00:08:55.27 and be very productive in their research output,
00:09:00.00 and yet find their careers disrupted
00:09:02.07 because funding suddenly becomes less possible to obtain.
00:09:06.02 And universities are given incentives
00:09:08.25 to leverage up -- to take on more risk --
00:09:11.19 by the current structure.
00:09:13.28 They can pay more salaries of their faculty
00:09:17.08 from research grant funding.
00:09:20.06 And they can take on more debt,
00:09:22.13 and they're encouraged to take on more debt,
00:09:24.07 to build new facilities or remodel labs
00:09:26.09 because they can claim the costs of those labs
00:09:30.18 from their indirect grant support.
00:09:34.14 And they are certainly incentivized
00:09:37.23 to hire or recruit more PhDs and postdocs
00:09:40.08 as research assistants as the lab...
00:09:45.06 the bench labworkforce for these grant-supported activities.
00:09:47.05 If their assumptions about future budget growth, though,
00:09:50.09 turn out not to be right,
00:09:52.02 they will find themselves exposed to greater risk
00:09:54.14 and even financial crises,
00:09:56.13 and that is happening now at many US universities.
00:10:00.29 This is an article I published in 2008 about the system.
00:10:04.23 I said I hoped I would be wrong.
00:10:07.06 Unfortunately, I was not wrong,
00:10:10.05 and things have gotten worse since.
00:10:13.22 Okay.
00:10:15.15 So, how does all this fit in with the claims
00:10:17.15 I'm sure you've all heard about there being
00:10:19.20 serious general shortages of science, technology, engineering,
00:10:22.14 and mathematics professionals in the US workforce?
00:10:26.05 Absolutely standard fare in American political discussions,
00:10:29.03 in the mass media,
00:10:31.02 though not in the specialist media, which knows better.
00:10:34.16 Researchers who have looked at this question
00:10:37.03 have been able to find
00:10:40.13 no credible evidence of these general shortages
00:10:42.21 that are standard fare.
00:10:44.20 Instead, what they find is,
00:10:46.03 while there are no general shortages,
00:10:48.03 you can find huge variation by field, by time,
00:10:52.15 by degree level, and by geography.
00:10:55.11 And most of the claims of shortages
00:10:58.19 come out of the IT sector
00:11:00.14 and especially from Silicon Valley,
00:11:02.04 and these are just very atypical,
00:11:04.25 very erratic kinds of industries
00:11:07.01 with their own massive booms and busts,
00:11:10.00 and very high housing costs
00:11:12.17 that have to be supported in Silicon Valley.
00:11:15.22 So, generalizing from Silicon Valley experience
00:11:18.24 is a really bad idea.
00:11:21.01 However, the shortage claims
00:11:22.25 are supported and continue
00:11:26.09 because those involved know that they work at the political level,
00:11:30.07 and there's very heavy financing of lobbying and campaign finance
00:11:34.11 from those industries to make the claim
00:11:38.01 that there are shortages of STEM workers.
00:11:40.25 Meanwhile, on the biomedical research side,
00:11:43.19 the leaders of biomedical research have been...
00:11:47.08 in recent years, have been... leading.
00:11:50.13 The Director of the National Institutes of Health, Francis Collins,
00:11:56.08 commissioned a report on the biomedical research workforce
00:12:00.00 by a task force of his Advisory Committee to the Director,
00:12:03.25 and I commend that report to you.
00:12:06.08 It's on the NIH website,
00:12:07.28 and you should read it because it's actually a very good piece of work.
00:12:11.11 And then, more recently,
00:12:13.06 a group of distinguished biomedical researchers,
00:12:15.23 who have come to be called The Quartet,
00:12:18.28 have taken a leadership role by publishing
00:12:22.03 these two articles at the bottom, here,
00:12:24.19 in the Proceedings of the National Academy of Sciences.
00:12:27.15 You can see in that picture Shirley Tilghman, the former President of Princeton University;
00:12:31.10 Bruce Albert's, former President of the National Academies;
00:12:35.01 Harold Varmus, the former Director of NIH and of NCI;
00:12:39.15 and Marc Kirschner, a distinguished researcher at Harvard Medical School.
00:12:43.17 And they've taken a leadership role which is very positive.
00:12:46.19 It's healthy because this kind of leadership is the best kind of leadership.
00:12:51.14 People like this can...
00:12:54.27 can most effectively identify
00:12:58.04 what stabilizers might be considered
00:13:00.18 and which would minimize any damage to what is,
00:13:03.04 as I said earlier,
00:13:05.00 a very productive research system.
00:13:06.25 And it also shows that senior scientists in biomedical research
00:13:09.23 are really concerned about early career scientists.
00:13:12.12 They're doing fine themselves,
00:13:14.10 but they are worried about the early career scientists.
00:13:17.13 Here, I'm including PIs and university leaders,
00:13:21.08 the National Academies,
00:13:23.06 NIH leadership, etc.
00:13:25.12 The young investigators, the early career investigators,
00:13:27.28 are their intellectual progeny,
00:13:31.14 and they know they're the future leaders of biomedical research.
00:13:34.09 So, let me close with three possible stabilizers
00:13:37.05 that ought to be discussed and considered, in my view.
00:13:40.12 First, find ways to reduce that positive feedback
00:13:43.08 that I discussed earlier.
00:13:45.17 That would mean really allocating more support
00:13:50.01 to PhD students and postdocs
00:13:53.03 as trainees rather than as research assistants.
00:13:56.02 Not more altogether
00:13:58.00 -- there's no need to grow the number --
00:13:59.25 but shift the proportion from what is now
00:14:01.23 a very high percentage as research assistants
00:14:04.14 back down to, say, half and half,
00:14:07.11 for trainees and research assistants.
00:14:10.03 Second, to seek to grow the budgets for research and development,
00:14:14.26 but to do so in a steady way...
00:14:18.03 in a steadier way, in a more predictable way.
00:14:20.29 It would require biomedical researchers and their leaders
00:14:26.02 to actually oppose any proposals
00:14:29.13 for booming increases in funding,
00:14:31.24 for additional doublings that might be proposed.
00:14:35.02 These have turned out to be destabilizing,
00:14:37.29 and the way I would put it is,
00:14:40.15 be careful what you wish for.
00:14:42.22 But also to support longer term budget plans
00:14:45.06 that do have sustained and sustainable budget growth,
00:14:48.10 such as a five-year trajectory,
00:14:50.19 as proposed by The Quartet.
00:14:53.11 And finally, we really need to understand this system better.
00:14:56.20 It's a critical system for the future.
00:14:58.17 We have very weak data.
00:15:00.15 For example, we don't even know to a factor of two
00:15:02.20 how many postdocs there are in the United States right now.
00:15:06.12 It's sort of shocking when you think about it.
00:15:08.14 And we need to have better research on the critical system
00:15:11.01 that we're talking about
00:15:13.09 and the way it's working,
00:15:15.05 and what problems it faces and how those might be mitigated.
00:15:18.21 Thank you for your attention.
00:15:20.25 That's the end of what I have to say.

This Talk

Speaker: Michael S. Teitelbaum

Audience:

Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: May 2015

Talk Overview

Although the US research enterprise is successful and productive, Michael Teitelbaum explains it is not very stable. This instability is primarily due to a history of boom & bust funding cycles. The most recent cycle occurred from 1998 to 2008, when the NIH budget doubled and then abruptly stalled. These boom & bust funding cycles are harmful to universities, faculty, and, most especially, early-career scientists and trainees. Teitelbaum offers three solutions for how to bring stability to the US research enterprise and make it sustainably productive.

Speaker Bio

Michael S. Teitelbaum

Michael Teitelbaum is Senior Research Associate at the Labor and Worklife Program, Harvard Law School, and former Vice President of the Alfred P. Sloan Foundation. He is a demographer who studies labor and worklife issues of scientists and engineers. Teitelbaum has authored many books over the course of his career, most recently Falling Behind? Boom,… Continue Reading