What if Scientists Didn’t Compete?

By John Tierney

What if scientists, instead of rushing to publish or perish, chose to cooperate? Sean Cutler decided to do “a little experiment,” as he calls it, and you can see the results in the forthcoming issue of Science.

The journal carries an article by Dr. Cutler and 20 other researchers in the United States, Canada and Spain reporting a long-sought technique for helping plants to grow with less water by activating the natural defenses that enable plants to survive during droughts. (Here’s the Science article; here’s a summary of the research.) Dr. Cutler, an assistant professor of plant cell biology at the University of California, Riverside, knew that the rush to be first in this area had previously led to some dubious publications (including papers that were subsequently retracted). So he took the unusual approach of identifying his rivals (by determining which researchers had ordered the same genetic strains from a public source) and then contacting them. He told me:

Instead of competing with my competitors, I invited them to contribute data to my paper so that no one got scooped. I figured out who might have data relating to my work (and who could get scooped) using public resources and then sent them an email. Now that I have done this, I am thinking: Why the hell isn’t everyone doing this? Why do we waste taxpayer money on ego battles between rival scientists? Usually in science you get first place or you get nothing, but that is a really inefficient model when you think about it, especially in terms of the consequences for people’s careers and training, which the public pays for.

One counter-argument would be that competition makes everyone work harder and more quickly, thereby benefiting society. But Dr. Cutler argues that this competition can amount to an expensive arms race that doesn’t leave anyone better off. His experiment in cooperation with four other laboratories, he said, yielded “a very compelling body of data validated by many labs,” and has inspired the researchers to go on freely sharing with one another.” Here’s Dr. Cutler’s conclusion:

My bottom line is that scientists are publicly funded individuals and we have to make decisions that are good for the bigger picture and scientific progress. Obviously there is a balance between self and community interests, but as it stands there are very few metrics of scientific “niceness” and few ways to reward community-minded scientists (some grants consider “broader impact,” but that is not the same thing). What is even worse, is there are even fewer mechanisms for punishing selfish (sometimes horribly so) scientists. If it were their own money or private money they were spending on their research — fine, they can be as selfish as they want and hold others up. But 99 times out of 100, it’s not their money- it’s the public’s money and it drives me absolutely crazy that there is no meaningful oversight of behavior.

I welcome reactions — and accounts of similar experiments.

Comments are no longer being accepted.

Bryan April 30, 2009 · 4:13 pm

Open Source Science

P April 30, 2009 · 4:23 pm

I have a simple, yet inspired solution. Take all of the public money away. All of it.

If anything truly that important was being funded, then surely the public will be willing to fund it by choice rather than at the point of a gun.

caveat bettor April 30, 2009 · 4:53 pm

I’m thinking where would we be with Einstein vs. Newton, Schrodinger vs. Einstein, Tesla vs. Edison (just off the top of my not so scientific head).

Chris April 30, 2009 · 4:54 pm

Competition is good for science–very good. Why else work so hard at it? Sure, the public good is important, but tenure is nice, too, and respect, and the excitement of being the one who cracks the code, and salary, and so on.

Philip Kitcher’s paper “The cognitive organization of labor” which appears in his “The Advancement of Science” (Oxford, 1993) makes a very strong technical and rhetorical case for the good of competition. One major point–it encourages some scientists to seek out non-standard, lower probability routes, which ensures that more ideas are exploited. And they sometimes pay off, but wouldn’t be pursued without competition.

Paul April 30, 2009 · 4:57 pm

As a layperson, so, admittedly, not a scientist, I’m enclined to agree with P. It’s bad enough that we give our tax dollars to egomaniacal elected officials, let alone to unelected scientists only looking for “first place!”

anonymous April 30, 2009 · 4:58 pm

“I have a simple, yet inspired solution. Take all of the public money away. All of it.

If anything truly that important was being funded, then surely the public will be willing to fund it by choice rather than at the point of a gun.”

I don’t know if you noticed this, but the government is made of people, and these people were chosen by the rest of the people. So if there is public funding for something, it’s because the people wanted it.

If you’re curious to know more about this phenomenon, google “democracy”.

anonymous April 30, 2009 · 5:03 pm

Cooperation and competition each have different roles. Competition increases the incentive for each individual or group to produce quality output, while cooperation reduces the total cost to society by avoiding replication of effort. Both have their place, and that is why scientists today cooperate sometimes and compete at other times.

CSGradStudent April 30, 2009 · 5:05 pm

In computer science research, cooperation often does have “selfish” benefits.

Consider two computer scientists with competing approaches, A and B, that both solve some problem. Both publish their approaches, but A also makes his algorithm or data freely available (for example, as a downloadable tool or open-source code). Then other scientists who wish to do further work upon this problem are more likely to download, use, and cite A’s work rather than B’s in their own publications.

So by sharing his work with the community, A has actually increased his influence in the field. The whole field also benefits because future researchers can directly build upon A’s work, rather than having to start from scratch every time.

More fields in science and engineering could benefit from this kind of community cooperation.

JD April 30, 2009 · 5:05 pm

There is some value to different, independent groups being able to validate each others work. If everyone in the field is the same group, who independently tests the claims? Yes, in this case it was internally validated, in a sense, but that’s not really independent, is it?

anonymous graduate student April 30, 2009 · 5:06 pm

(1) Scientists already cooperate like this — lots of *new* collaborations actually start this way. Looking for this kind of external collaborator is standard operating practice in some fields. How prevalent this kind of cooperation is probably a function of how capital-intensive that field is … higher start up and operating costs probably mean higher levels of collaboration formed in this way, as opposed to forming out of pre-existing relationships between investigators.

(2) This “experiment” ignores the downsides of larger collaborations — larger groups working together, especially when in separated geographically, mean the project is more likely to get bogged down or even fail altogether. Collecting the data, sharing it, and ultimately writing the paper all become harder when there are more cooks in the kitchen.

(3) Some research projects (beyond just collaborative data collection) can be pursued by different means or methods — there are often genuine differences of opinion between investigators about the best way to attack a research question, and so forging these kinds of collaboration in those cases would be useless (or counter-productive).

(4) Even if it does make sense to join forces on a particular research project — the more people there are involved in a project, the more incentive there is for someone to “cheat,” to break off from the group and publish on their own.

Anyone who’s paid attention to the rise of large sequencing centers and genomics research institutes in the wake of the human genome project would have realized that forming institutions which foster this kind of cross-group collaboration has been an important dynamic in some fields for quite a number of years — and it’s hardly unique to biology. Other fields like astronomy and physics have even longer histories of long-term, large-scale collaboration and cooperation between researchers. Do you honestly think that “cooperation” is a foreign concept throughout science?

At the same time, anyone who’s actually participated in a scientific research project realizes just how motivating the competition really is — a lot of basic research really does move forward at a much faster pace than it otherwise would, thanks to the fear of being scooped. We all benefit from the rapid advance of basic and applied research as a result…

Kai April 30, 2009 · 5:06 pm

P, you really want to end university research? It provides some of our biggest technological and productive advancements.

Lisa April 30, 2009 · 5:07 pm

As a scientist I agree with much of what Dr. Cutler has said. There is no control on obnoxious behavior that in fact impedes the flow of science but enhances the careers of a few well-placed scientists. What most people don’t realize is how few resources are really available for scientists. There are lots and lots of smart and hardworking scientist but only a very small percentage get NSF grants. The last NSF panel I was on (in early 2009) funded less than 6% of proposals. A big improvement would be to increase the number of National Science Foundation grants that are awarded but to limit the amount of money any one lab can get. What we have now are a few very large very powerful labs that can push the smaller labs run mostly by younger scientists out. The problem with our current model is as Cutlter points out the losers here are not just the scientists that get no funding but also the tax payers who support the research and paid to train all these great people who now lack funding.

Kent April 30, 2009 · 5:07 pm

Re: P

Scared of a little science, are we?

When it comes to experimentation, topics can be so obscure or specialized as to completely elude the public’s grasp, though the findings may eventually contribute to something much larger–and in the public’s interest.

Mike April 30, 2009 · 5:09 pm

Fantastic topic, and probably one that should get wider discussion in the policy world, not to mention the scientific community.

The funding mechanisms are at the heart of this question. NSF / NIH / NOAA / DOE / DOD etc. — many of the important gvt. funding agencies — do not put any sort of stress on this point of collaboration. Certainly, it is increasingly typical for funded grants to have multiple PI’s and to be multi-university. Nobody else will be able to compel this change — not journals, who simply judge the work sent to them, not universities, who compete more fiercely than PIs.

I am often inspired by the true — and groundbreaking — collaborations that happen internationally, especially the ocean drilling program (ODP). Recent projects in Antarctica have shown the enormous value a tad more civility, equanimity and collaboration can lend to science. This collaborative work can also speed up the results hugely. It is not uncommon for full-fledged research analysis from ODP trips to be published in the same year as the conclusion of the field work (atypical for smaller research groups).

I am a graduate student at another University of California campus and have recently decided not to complete a PhD given the internecine political backstabbing, gypsy lifestyle and ego-driven atmosphere of the academic world. I am leaving with an MS in a science funded by the government and taxpayers, wishing that I could contribute more to society. I think I can, but it’s not through the myopic and self-centered world of academia.

Steve April 30, 2009 · 5:09 pm

They often do not share their data and even when funded by the public, they act like its their personal property. Most that I’ve been around want to spend their time in their own private sandbox, like some artists, they want the money and reject any attempts to suggest what they can do with it. Pear review I’ve seen is a whitewash.

I agree, get the public out of funding most of this stuff, then see what it looks like when the tide goes out. I often refer to science funding as welfare for the anointed.

Justin Thyme April 30, 2009 · 5:13 pm

hooray.

I think competition is driven by habit and fear.

About overcoming a similar divide: a quick story regarding student collaboration at the Harvard Business and Law schools where I was a student in the early 90s. Both schools use what’s known as the “case method,” in which instructors guide a class discussion that generates the concepts to be learned. Students are encouraged to study together and are graded on collaborative participation. At the same time, grading is individual, and a fairly consistent number of underperforming students are asked to leave both programs each year. So you’re motivated to collaborate and outshine.

Law School students were all about outshining. Classroom collaboration was much more guarded, careful debate than dialogue. B-School students, in contrast, were much more willing to share information (even surreptitiously slip notes to students being grilled by the instructor), build on others’ ideas, take risks and be creative.

Although the B-School students are generally older and perhaps more comfortable solving problems collectively, I think the difference was more structural than about the students: it was all about the teacher.

I invited several HLS professors to visit HBS classes to see if they noticed what I was seeing, and if it would have an impact on their teaching style. I was surprised by the results. The less-senior profs were intrigued but unchanged, while the most senior guy, Professor Abe Chayes, changed his teaching style dramatically after the experience. He said he’d gotten into a rut, basically. From then on his classrooms were much more lively.

David Annis April 30, 2009 · 5:13 pm

P, (#2)

Sure, and while we’re at it why not take away all money for roads, sewers, defense, police, and fire? Whoever wants to donate for those things can.

Oh wait, that might be hard to get passed, so you and I need another way to get to a no tax paradise. I know – Let’s move to Somalia! You go first and I’ll meet you there. I just need to add your brilliant idea to my new blog //moderatelyliberal.blogspot.com first.

David

MD April 30, 2009 · 5:14 pm

That is just a stupid republican argument to take public funding away. I think we should increase public funding to find more cures and preventive measrures and away from defense

BSE April 30, 2009 · 5:14 pm

This is an interesting and compelling approach to scientific effort. Other examples of cooperation among large groups of scientists usually involve limited or very expensive resources for research such as particle accelerators in experimental physics. Hopefully, greater public awareness of this problem will lead to encouragement of greater collaboration among rival researchers.

Cook April 30, 2009 · 5:14 pm

P, clearly you and/or your child aren’t suffering from a potentially deadly or debilitating disease on the verge of a cure. We are at the point of a gun but it’s not your metaphorical one.

j April 30, 2009 · 5:15 pm

I would suggest to the previous poster that the next time he has an opportunity to require medical services, quite possibly in his case due to gunshot wounds, that he follow his principles and refuse all treatments developed directly or indirectly through the expenditure of government funds. Aspirin and Band-Aids would be his main treatment, though I suppose he could choose the plain or Scooby-Do varieties at his discretion.

Robert April 30, 2009 · 5:16 pm

I’m afraid Mr. Tierney is completely wrong.

Look at a random issue of Science Magazine, note how most of the research reports have contributions from scientists working at more than one institution.

The correct question should be: “What if scientists didn’t cooperate?”

Sean Cutler’s report of his “little experiment” is junk science and the gullible reporter believed it without performing a simple fact check.

GDT / Buffalo April 30, 2009 · 5:16 pm

We scientists compete and collaborate – nothing new about that.

John Kuehne April 30, 2009 · 5:17 pm

Sometimes scientific competition is used to ensure that results are correct, by setting independent teams on the same problem. The competition between GSFC, CSR, and JPL produced better results for earth’s gravity map than could be expected from any one outfit Some kinds of collaboration run the danger of everybody making the same mistake and jumping on the same bandwagon. Scientists, mathematicians, and engineers are self-organizing. The system works without top-down control deciding on how collaboration and competition should work. Inviting your peers to work with you a big project is usually just called a Research Institute.

The comment by “P” is just sad. There are already countries that practice this, but they didn’t invent GPS, MRI, space telescopes, genetically engineered insulin, interstate highways . . . well the list goes on. Maybe P should emigrate.

Nancy April 30, 2009 · 5:18 pm

P –

There is such a thing as the common welfare, and government has a role to play in supporting it. We benefit as a society from having fewer sick people, faster computers, better communication systems, strong transportation systems, etc. We benefit as a society from the advances that come through research. Paying for these things through taxes is the most cost-efficient method possible, especially in the case of peer-reviewed research projects. No guns are involved.