Funding Fixes pt. 2

If you’ll bear with me, I’ll wrap up my rant about changing the grant system in this post.  Picking up where we left off, here are two more changes that could improve government research funding:

a)    Fund the researcher/lab, not the project

This change has the same intention as a) and b) from my previous post [link]. Loosening the restrictions on the intellectual focus of funding recipients will open the doors of innovation instead of locking them into dead-end or myopic projects dictated from the top down. Funded researchers would have the freedom to switch tracks away from a project once they know it won’t work, rather than wasting months of research simply because their pay is tied to the project.

b)   Insulate the funding system from public opinion and politicians

All of my suggestions are based on the well-established concept of how innovation works: many people have to try many things, of which a few will succeed, but most WILL FAIL – this is especially true for the sort of low-immediate-value, boundary-pushing research that should be publicly funded.

Obviously we want success, but even the failures are valuable because if information is disseminated as it should be (which is sadly a big if) they act like flashing traffic cones blocking off a road that ends in a cliff.

Sadly, most people don’t seem to recognize that discovery is essentially a stochastic process where you can’t know what paths are going to work a priori. This lack of understanding can in part be blamed on the human tendency to assign stories to successes that tell a narrative ignoring all the points of pure dumb luck.

This misconception is just one of a slew of other mental fallacies that distort most people’s perception of research. Instead of listing them, I will just point you to “Thinking, Fast and Slow” by Daniel Kahneman. Worse still, many politicians and large parts of the public see government-supported research as a business investment that needs to yield clear, immediate value. That sort of research should be done by, well, business.

The outcome of having funding tightly controlled by groups with a distorted idea of how the research process should work – especially the sort of research that should be funded by public, rather than private money – is well summarized by Aubrey De Grey:

 “The situation is the worst in the more mathematically extreme cases. These are the situations that can be summarised as “high risk, high gain”—low perceived probability of success, but huge benefits in the event of success. As any academic will aver, the mainstream mechanisms for supplying public funding to research have gravitated to a disastrous level of antipathy towards high-risk high-gain work, to the point where it is genuinely normal for senior scientists to stay one step ahead of the system by essentially applying for money to do work that is already largely complete, and thus bears no risk of failing to be delivered on time.”

A synthesis of these ideas into a vision for a new system really wants an essay, or at least its own post.  However, my sketch of the alternative funding system looks something like this:

Rather than putting out calls for grant applications and spending time reading them, the gatekeepers of funding money spend that time actually paying close attention to their own fields. They meet to discuss the work of individual researchers and labs based on a number of useful metrics and award long duration funding based on perceived merit.

 Obviously, this new funding scheme does not solve all the problems of the grant system and leaves several lingering questions. However, the discussion of how to fix our broken system needs to start somewhere if the innovation rate we have come to enjoy is going to continue.


Funding Fixes pt. 1

This post is a follow-up to points I made about the absurd inefficiency of the grant system that funds most public research. While there is some value in just shedding light on a problem, it is far more valuable to point out the problem and propose some solutions.  Most of my proposed fixes stem from a conversation I had with Duke Economist Mike Munger, so he deserves the lion’s share of the credit.

These ideas apply primarily to government money aimed at research with very limited or nonexistent near-term applications. The government shouldn’t be funding research that has obvious near term benefits.  I argue that these projects should be left to the private sector for a number of reasons including efficiency disparity and the simple fact that there is only so much the government should spend on research, so it shouldn’t fund things that would be funded through other means.

However, plenty of research without obvious near term value will benefit humanity down the line in completely unexpected ways. Take pure mathematics for example – it doesn’t seem to benefit anybody, but given a few dozen to hundreds of years, theories that mathematicians are coming up with now will percolate into physics, chemistry and biology, generating new discoveries there that will be taken by engineers and used to benefit millions of people.

Here is a rough sketch of some solutions:

a)    Decrease focus on deliverables

Research is inherently a process of discovery. Think of the classic (though apocryphal) story of that paragon of scientists, Issac Newton. He clearly didn’t sit down under the apple tree thinking “I need to discover a new natural law today.” And in fact, if he was funded by a grant to turn lead into gold (people forget about it, but Newton was also a practitioner of alchemy) he wouldn’t have sat down under the tree in the first place – he would have been in his lab, too busy frantically trying to tune proportions of sulfur and mercury so that he could demonstrate progress and get his grant renewed.  Grants constrain the focus of recipients, suppressing serendipitous innovation.  Thus, a revamped system should give money based on quality of past performance, rather than the promise of future deliverables.

b)   Increase the duration of funding

Significant, fundamental research of the type the government should be funding can’t be rushed. Grants with short time limits encourage a number of counterproductive behaviors: rushed and shoddy work or a project that might have gone somewhere given more time, but is terminated due to lack of funds because it couldn’t be completed in the set time limit.  Uncertainty about year-to-year funding also incentivizes researchers to take on projects with low probabilities of failure, limiting the advances of research.  To minimize these problems, the funding system should increase the duration of awards to provide researchers with the flexibility to truly explore.

c)    Base awards on real-life metrics, rather than proposals

It turns out that researchers don’t keep all their ideas and work in vaults and can only be convinced to bring them forth when tempted by grants. Instead, we love to show off our work and talk about our ideas whenever possible.  The most pedigreed method is of course the peer-reviewed paper. However, thanks to the Internet, researchers can now share completed work and future plans with the world through an ever-expanding list including articles, blogs, educational videos and recorded talks.

Basing funding on communication with actual value rather than grant proposals has a number of upsides:

  • Those controlling the funding could spend the time they would have wasted reading grants paying attention to the work and ideas being presented in their field, which is something they would have done regardless.
  • Researchers are incentivized to share their work with the world in as clear and compelling a way as possible. They will have time to share and perform actual research because grant proposals won’t continuously eat their time.
  • This increased information flow alone will benefit everybody because more and better information about up-to-date research will not only make more people informed, but will spur on innovation at a greater rate.

If we could implement just those three changes to the government funding system, I believe the benefits would be outstanding. However, this list is far from complete. I don’t wish to cause any eye-glazing, as too-long posts are wont to do, so a future post will take up where this one leaves off – finishing the list and synthesizing the changes into an architecture for an improved (but still imperfect) funding system.

Grant Me Some Change

This week, the journal Nature published a short article with the headline: “Funding: Australia’s grant system wastes time.” Here is a link for those with access. Scientific journals aren’t known for their sense of humor, but the understatement of the title is so extreme that it must be an attempt at a joke. From the article:

 “We found that scientists in Australia spent more than five centuries’ worth of time preparing research-grant proposals for consideration by the largest funding scheme of 2012. Because just 20.5% of these applications were successful, the equivalent of some four centuries of effort returned no immediate benefit to researchers and wasted valuable research time.” (emphasis mine)

Having directly experienced the grant process here in America, I can attest that the Australian system is not an anomaly in its ability to waste time. You should also consider the unseen waste. In addition to the time spent the scientists writing the grants, other highly trained scientists then need to spend their time reading and evaluating the proposals. This adds even more scientist-time to the gruesome casualty list of the grant system.

One could argue that writing grants helps scientists to spread their ideas, and refine their communication abilities. However, in terms of these benefits, grants are far inferior peer-reviewed papers, blog posts, verbal presentations, or other ways that researchers share their ideas. Grant proposals are only seen by a few individuals and don’t become part of a larger body of knowledge. Additionally, the feedback to the writer in the grant system is far inferior to other communication methods.

Explanations written for a journal paper can be reused in a talk or subsequent papers. However, grants often require many components that cannot even be reused for subsequent grants, let alone useful applications. Specific budget proposals (that are just meticulous bits of hand waving which probably will change anyway) are one example of these non-reusable components.

This list of inefficiencies is far from complete, but the point is that the side benefits generated by writing grant proposals are tiny compared to the amount of time wasted.

Coincidentally, I recently had an excellent email exchange with Duke economist Mike Munger on this very problem. The same inefficiency issue caused by grants occurs in almost any case involving bidding for government money. He suggested some solutions that I found very intriguing and will discuss in a future post.

I believe that some of the few things that truly improve the human situation are scientific and technological innovations. If you agree with that sentiment, it is clear that the grant system wastes hundreds of thousands of arguably some of the most important man-hours in the world. If we could change the system, think of the time that could be spent actually producing results and changing lives (or just as productively, getting some sleep) rather than grinding out funding spreadsheets destined for a government filing cabinet.


Coolest. Inverted Pendulum. Ever.

SpaceX has been testing a system called the Grasshopper that will allow reusable rockets to perform powered landings. Currently, everything that comes back from space either parachutes in and has to be picked up from the middle of nowhere or glides.

The link contains a very cool video of a test flight and more information. 

It’s innovations like this that actually change our ability interact with the world that make me proud to be a human. When I first saw the animated video of the concept, it triggered the normal warning sirens that should go off whenever you see a space project pitched with an awesome animation and little else. My initial skepticism makes it even even more exciting that SpaceX seems to be pulling it off.

This is a good example of a project that just straddles the tinfoil hat line.