there is a curious result in behavioural economics, which shows that paying people to do what they like to do, sometimes provides a disincentive for them to do it, and people correspondingly lower their effort to do the task.
The example I recently came across, from the 7 Rules of Behavioural Economics, or some such, was that if you pay people to have their friends for dinner, they entertain less.
But, enough about the decline of intradepartmental socialization...
The reason I thought of this, is that yesterday we congratulated a colleague on a nice result, and another colleague asked some question about "which agency the grant was from", which lead to a quick chortle, and "no, this was unfunded research".
Which is true, quite often. In fact one of my mentors, many years ago, told me that he found the grant system quite irritating, because by the time you got the proposal through and had a grant, it wasn't interesting any more...
There is quite a lot of truth to that.
Now, it is not universal - there are many important large projects which take long term funded efforts, and are fun, mostly, through the years.
But, it is also true, that a lot of the really nice results, especially the little and medium results, come from unfunded research - stuff people do because they can, and it is fun.
Effectively, a lot of this research is subsidised by the more ponderous, bread'n'butter, grant funded research. The people and facilities are still paid for, somehow. And, mostly, the agencies know this and tolerate it.
So, why don't we do more of it?
Why the pretense?
Well, apart from the need to do large systematic multi-year projects which actually employ most of us most of the time?
Why don't we just accept that researchers do random fun stuff a lot of the time, as the mood strikes them, and really that is where a lot of the good results come from?
Here I'd bemoan the loss of the Golden Era, when it was ever such, but it wasn't really, unless you were an independently wealthy English aristocrat in the Enlightenment, or a precocious upper middle class scion at an Ivy during the Gilded Age.
Research ought to be fun.
It works better when the people doing it enjoy it.
The current system is structured to make what ought to be progressive joy into dreary drudgery (ok, I exaggerate).
Worse than that, it, may be producing disincentives for good research and lowering net productivity.
There is a story I heard years ago from Leon Lederman about an unnamed physicist who had a remarkable track record of discovering what he proposed to discover. The system worked smoothly until one of his proposals was rejected. He went to talk to his program officer about that, and the PO asked him why he think he could do what he had proposed. He pulled a bunch of papers out of his coat and said something along the lines of, "Because, you fool, it's already been done!" His game all along had been to propose doing what he had just done and use the funding to support his follow-up project.
The above tale is probably apocryphal, but some days I feel that a proposal to do work that isn't already substantially underway has a much lower probability of success than a proposal to do what you have already done. Especially during proposal season, which is right now for the NASA programs I usually propose to.
"Why don’t we just accept that researchers do random fun stuff a lot of the time, as the mood strikes them, and really that is where a lot of the good results come from?"
Because that is not the researcher's call to make. it's his patron's. If his patron accepts that he'll spend a lot of the patron's money just playing, then he can go and play; iif not, then he can't. When the patron is the public purse, which is to say the taxpayer, it seems unconscionable to simply assume that he doesn't mind the researcher playing.
As a data point, I made my money in industry, and then retired and did some self-funded research into acoustics and audio synthesis. So it's still eminently possible to be an independently wealthy English gentleman - in fact, I'd recommend it! Afterwards, I found that I became a lot less approving than I had been of a lot of the taxpayer-funded research that's out there. And especially of the kind where the researchers consider "obtaining grants" to be their primary job function!.
As a supplementary data point, when I was being paid to produce the best software in my market (by being paid royalties and not just a salary), I was really motivated to do just that. When I was working for my own enjoyment, the results were good, certainly, but by no means great. So I'd suggest that the most important word in the first paragraph of your article is "sometimes".
Thanks, Steinn! This is a topic near and dear to my heart. I gave up long ago on writing proposals. My MO is to find a project which someone else is already doing which sounds really cool, and ask if I can join to do some little piece of the big puzzle -- for free. It's a win for me, because I get to do something interesting. It's a win for the project, because they get a small bit of effort without having to pay.
It's certainly true that one can't do a big project in this way -- or, more properly, one can't lead a big project this way -- but as the astronomical community builds larger and larger collaborations, there will be more and more opportunities for people to work in this fashion.
And some tiny little fun projects can be done without any funding. When SN 2011fe went off last August, I realized that it was a rare chance for me to get back to my roots in supernova photometry with the 12-inch telescope at our campus observatory. It took a lot of work -- getting up at 2 AM in the middle of January and spending half an hour setting things up in the dome was hard -- but watching that light curve grow each morning was terrific. It made me feel like a grad student again :-)
On a related note, I am always amused by the process of submitting revised proposals. In my field (and I expect most others), you rarely get the first grant submission funded. But, if you got a reasonable score and make specific changes in response to the reviewers, you can get funding on your second attempt.
However, I rarely see people actually implement the revisions - they are just word service to appease the reviewers. In reality the research that gets done is often only related to what was proposed not precisely what was proposed. Fortunately there is little oversight other than to be able to tie some publications to the grant in some way and show that you spent the money more or less how you had it budgeted.
And, in a sense this all works - science gets done and researchers get to do what interests them. But, it would be nice to get the funding system to match the actual research mechanism. That is, fund those who produce in the fields that are of greatest importance - general aims rather than specific aims.
"But, it is also true, that a lot of the really nice results, especially the little and medium results, come from unfunded research – stuff people do because they can, and it is fun."
Interesting point, but do you have any specific examples? Are you talking about the kind of unfunded projects that ultimately support a funded research project (or answer questions tangent to a funded project)? Are you talking about unfunded research projects that are intriguing enough to receive later funding? Or are you talking about unfunded projects that stand alone yet have a significant impact on a particular field?
It seems to me that the small projects alluded to in this blog post are of the type that reaffirm a scientist's love of science, although have little impact. Even on tiny budgets, I think we will always find ways to enrich our lives.
I would have to agree with BKsea - fund general aims rather than specific aims. I would add that even though you may state a specific aim in a grant, carrying out a specific aim is quite a different matter. Troubleshooting the techniques in specific aims and trying new approaches to achieve specific aims are a large part of the daily creative process of science. These types of issues leave a lot of flexibility, and can produce interesting, unforeseen results. They can also be engaging and exciting.
I also suppose that the ease of doing unfunded projects will depend on the field of research. I work in a biology lab. People in my research group (myself included) have been known to do small side projects for fun and interest. However, few of these results get published and, therefore, have little impact. This is mostly because 1.) it is very time consuming and expensive to do the kinds of experiments that will get nice results and 2.) small side projects usually don't tell a very complete story.
@Ian - clearly economic incentives are often motivating. The interesting thing is that not only are they sometimes not an incentive, but can be disincentives. In fact I'd argue that the way economic incentives are tied into the grant system is what makes researcher "obtaining grants" become the end rather than the means. I think it can lead to research focused on long term projects that make incremental progress - a phenomenon also seen in the software industry.
"Becoming rich, then doing fun shit" is an excellent business plan, some of my best friends have done it, but it is not very practical as a way of providing a sustained large scale research effort supporting long term growth of a tech heavy economy.
Pure research doesn't work the way applied development does.
@CFP - so, my example was the paper that just came out.
More generally, none of my Nature or Science papers were based on funded research when they came out.
All lead to or were part of long term research projects, some of which was fundable for part of the time, some of which were even fundable for me to do the research, as opposed to others being funded to follow through on the work. All are well cited and continue being cited and all initiated lines of research I continue pursuing.
I find funding panels to be annoyingly conservative. Even when I sit on the panels - the guidelines for funding and need to show proven incremental progress forces that dynamic.
From a European perspective, doing 'unfunded' research is effectively what many people in faculty positions do, since we don't need to apply for funds to pay our salary when we are not teaching (in fact, in most cases such funds don't even exist, unless you land a very large grant which you can use to buy yourself out of teaching). There is still pressure to write grant proposals though, if you want funding for travel, postdocs, grad students etc. and also simply to help you with career progression. Effectively I can do what I like during my research time... but it doesn't usually work that way due to commitments of collaborations, supervision etc. I would feel bad about giving a student or postdoc a project which didn't have a strong guarantee of good results at the end of it.
Doesn't the obvious solution deserve mention? More tenured faculty positions with their associated academic freedom.
Well, my jaundiced perspective is that fellowship postdocs are the people with most freedom to do blue sky and radical research. Tenure track faculty are driven to conservative incremental research, because that is what maximises the chance for grants and reduces the risk of having insufficient grants. Once they have tenure, there is a trap - there is now an infrastructure which relies upon the flow of funding, you lose it other people get hurt. So the tenured faculty are driven, to some extent, to protect their sunk costs and preserve the research infrastructure they built; which continues the motivation for conservative, long term, incremental research. Much of which is still good of course. Some of which must be done that way. But I fear it sometime squeezed out more interesting research.
@SS: Agreed about the reality that tenured faculty are under great pressure to bring in grants. Isn't that one of the features of the current system that you are critiquing? Also the "grass is always greener" applies to your comment about fellowship postdocs, who have similar pressures to conservatively produce results in 1-2 years to prepare for the next job cycle.
I really like the NSERC Discovery Grant system (the mainstay of grad student funding in Canada)--5-year grants that support a researcher to do a broad category of research. Pretty high success rates, too (I think something like 60% or so these days, used to be rather higher), so obviously they're not terribly large; but it's a reliable source of funding for grad students.
So Steinn, this sounds like you'd be more in favor of a block-grant or rolling-grant model (like many European countries). However, that model has its problems too: Big-names tend to score the cash since they've produced in the past so talented newbies may not get a look in.
Maybe the solution is to have say 10% of a funding pool devoted to high-risk projects or people who haven't gotten funding before or are at institutions without a big soft money population?
From the other side, too often I have seen grants to get basic data needed for just about everything turned down for being number grubbing.