More Science, Same Low Price!

Let's take a quick look at the economy of fundamental science in the United States. To greatly simplify, here's about how it works:

1. Congress gives a limited amount of money to the funding agencies.
2. Funding agencies want to make a big impact with their limited funds, so they solicit grant proposals.
3. Scientists compete against each other to produce the best grant proposals.
4. A few lucky winners receive grants. They use the grant money to conduct scientific experiments.
5. Universities put pressure on the scientists to get more grants.

Can you spot the problem in this system? Here, I'll give you a hint:

Perhaps what you measure is what you get. More likely, what you measure is all you’ll get. What you don’t (or can’t) measure is lost. H. Thomas Johnson

What do you imagine scientists really spend most of their time doing? Walking around their laboratories helping their graduate students? Sha, as if! When they're not teaching or helping undergraduates, most of their time is spent in their offices writing grant proposals, of course! Grant proposals are what lead to grants. The universities skim money from those grants to maintain their facilities. And the results that come from those grants bring prestige, which means wealthy students will be willing to pay more tuition to attend that university.

Everything in science hinges on grants. The funding agencies determine what leads to grants. And the funding agencies have decided that grant proposals lead to grants. So it's really all about writing grant proposals.

What do we want scientists to spend most of their time doing? Producing knowledge, right? Isn't that what scientists are supposed to be good at?

So how can we fix this broken situation? We just need to make one important little change. Instead of rewarding promises, we need to actually reward contributions to knowledge. Here's what that would look like:

First, let's observe what this proposed change will not fix:

• Congress won't give more funding for fundamental research.
• Universities won't reduce the pressure they put on scientists.
• Doing science won't get any cheaper. (Notice the number of green light bulbs is still the same.) This won't buy any more super-colliders or robots or giant telescopes for science. It will only incentivize low-cost preliminary work and theoretical investigations. But there is plenty of that to be done in every domain.

The only change I am proposing is:

• Let's reward doing science instead of rewarding promises to do science.

The consequence of this change is that universities will pressure scientists to do science instead of pressuring them to write more grant proposals. So instead of competing against each other to write more grant proposals, scientists will compete against each other to publish contributions of knowledge. For the same low price, the funding agencies get a whole lot more science. The broader impacts will be tremendous!

The catch

Of course there is a catch. There is always a catch. The funding agencies already know very well that they could buy a whole lot more science with the same amount of money. Why don't they do it? Because they are scared of liability for corruption. Consider the following scenario:

Some scientist publishes a break-through advance that fundamentally shifts his domain of science. So the National Science Foundation gives him a substantial grant to produce more work like that. Then he uses the money to study the behavioral patterns of young girls washing cars in wet T-shirts. And of course, to conduct this research properly, he needs to buy an outrageously expensive sports car for himself. Well, the Flat Earth Society makes a stink on social media about how socialism is corrupting science. An outraged right-wing congressman summons the director of the NSF to account for how he is using money from his tax-paying constituents, and he points at a scathing article in the Daily Flat Earth Chronicle about this particular scientist's "work".

If that scientist had written a 15-page grant proposal about how he planned to use the money, then the director of the NSF could simply use that proposal to shift the wrath onto the university of that scientist. But if the NSF was just giving money to that scientist because they wanted him to keep producing more of what he had already done, then it would be entirely up to his discretion to determine what counts as important science. Funding agencies just don't trust scientsts enough to do that.

...or do they? Well, who do you think really picks the winning proposals? Yep, other scientists who volunteer to serve as reviewers. So scientists ultiately determine who the grant winners are anyway. Yet, the very same system turns them against each other to compete for limited money by not doing science! This whole situation is so messed up! Isn't there a way we can limit liability for corruption, keep the process fair, and still encourage scientists to spend their time doing science instead of just writing proposals? I think there is. Here is the solution I propose:

Implementation details

1. The funding agency surveys scientists to identify the perceived impact of various publication venues in relevant domains of study.
2. The funding agency randomly selects a number of recent publications from those venues, giving more weight to the venues determined to publish more impactful works.
3. Volunteer reviewers are asked to rank the intellectual merit and broader impacts of a few of these publications.
4. The rankings are combined to produce an overall ranking for all of the randomly-selected publications.
5. Funds are allocated for the authors of the top-ranking publications to receive grants.
6. Only the winners have to write proposals, and only program managers have to review them. Scientists who are not eligible (perhaps due to citizenship or retirement plans or who already have excessive funding), or who are unwilling to produce a reasonable plan, might still be denied. But otherwise, nearly all proposals in this step are funded.

Significantly, this plan is nearly identical to the way it already works. The major differences are:

1. Only the winners have to write proposals.
2. Writing proposals is no longer competitive.
3. Scientists mostly do science.

Q & A

• Even theoretical research requires money.
  Yep. You know what else requires money? Writing grant proposals. Universities find a way to pay their professors to write proposals (usually by drawing more tuition from undergraduates) because they must. If making knowledge contributions was required to obtain a grant, universities would find a way to pay for that instead.
• Won't this be bad for junior faculty and faculty in under-represented minorities?
  No. It is no different than before. The funding agency can still put their thumbs on the scale when they rank papers, just like they do now when they rank proposals. If they really need to gather demographics information to do their affirmative action thing, funding agencies could require scientists to register in order to be eligible to receive a grant. The registration form would provide them with any data they need.
• Won't this give too much power to journal editors?
  Perhaps. This is my biggest concern about the proposed change. However, there are two reasons not to be excessively concerned about this:
  
  (1) It's really not very much of a change. Publishing in top scientific venues is already a soft prerequisite to obtaining grant money. The proposed change would just formalize that.
  
  (2) The market will probably do a better job anyway. Being centralized or authoritarian has never really been the way of science. And since there are not many barriers to entry for a new publication venue, competition can be expected to do a better job of resolving issues than benevolent-but-oblivious authorities (like the panels that operate under the direction of funding agencies).
  
  Some fast-moving scientific domains, such as computer science, have started shifting away from long journal articles and toward emphasizing the more rapid publication of short conference papers. The process of getting published in these domains more closely resembles how panels of reviewers pick the grants at funding agencies, and gives little-to-no authority to any particular editor. Although change is often painful, perhaps the evolution of science would be accelerated by shift away from those antiquated methods.
• How can funding agencies encourage projects that include under-represented minorities?
  Let's be real for a moment: In all of my proposals I promised to hire all the qualified black and hispanic females I could find to conduct my scientific experiments. That's called lying under the immense pressure that the University heaped on me to say whatever the funding agency expected me to say. Sure, I really was planning to do what I said I would do. And the truth is I really wanted to do my part to help women and under-represented minorities have a greater role in science. And I did make deliberate efforts to do that.
  
  So why was it lying? Because I'm a computer scientist, not a social scientist! I don't know how to promote social change. And worst of all, I know that I don't know how to do that. I have absolutely no business claiming I'm going to make broad societal impacts in the domain of social equity! And all the volunteer reviewers who read my proposals knew that too. We all roll our eyes when we read those silly boiler-plate paragraphs. All of us knew the exact number of black and hispanic females qualified to conduct scientific experiments in my lab that I would probably be able to find. Let's call it what it really is: We're all just conspiring to say what the funding agencies expect to hear. That's not promoting real change. That's just called lying.
  
  How do we fix this? Well, for starters, let's all stop lying to ourselves and the world about it! Pasted boiler-plate paragraphs in a Broader Impacts section aren't going to get girls excited about science. Maybe give grants to social scientists if you want something real to happen. But if you really want to press scientists who know nothing about the domain of social science to pretend they do, that could still be a required part of the proposals, just as before--except maybe scientists will take their promises more seriously if they are really a condition for receiving money, instead of merely one criterion by which our grant proposals are evaluated. So, the proposed change isn't going to weaken social justice efforts. If anything, it will make them much stronger. But mostly, it's just not going to do anything. This is not the problem I am trying to solve.
• But we want to know what scientists are going do with the tax-payer money we give them!
  Oh, please! You don't even know what scientists do with your money right now! I once commented to a highly-successful senior faculty member how anxious I was to get funded so I could spend more time doing research and less time begging for money. He just laughed and told me how naive I was. He told me when I get a grant it doesn't really change anything. He said that's just grease for the wheels of the great machine. It keeps the graduate students in your lab doing what they are already doing, and it keeps you in your office writing more grant proposals, just like you are already doing.
  
  I was a little puzzled by this response, so I asked him what was the point of all the care and effort I was putting into describing my research plans in my grant proposals. He laughed again and assured me no one was really ever going to hold me to any of the promises I made. He then explained how the system really worked. Consider two professors: One does exactly what he said he was going to do in his grant proposals. The other uses the money to buy out of his classes so he can spend more time writing proposals. Which of these two professors is going to have a successful career?
  
  As the implications of this started sinking in, he laughed at me yet again. He told me all junior professors come in full of ideals about the glory and purity of science. He said the ones who figure out how the system really works get tenure, and the ones who cling to their ideals about science just aren't cut out for the job.
  
  Was he right? Well, he's still a full professor. And to be clear, he wasn't the only full professor who gave me a version of that lecture. And here I am, still clinging to my ideals about science, refusing to be a cog in a corrupted system that milks money from funding agencies, and pretends that ignoring graduate students who struggle alone in the lab to fulfill our grant proposal promises is "doing science". (Don't cry for me. I chose the one I wanted more.)
  
  The unfortunately reality is that our universities are packed with sharp tenured individuals who are experts at working the system, and who have to varying extents let go of their intial enthusiasm for changing the world. But this situation will not be fixed by rooting out individual cases of corruption. It is the system itself that is doing the corrupting. The only way this problem can be fixed is by adjusting the incentives of the system.
  
  Paying for promises is just getting us promises. Let's start paying for science instead.