How should private foundations support science?

It is amazing that there exist quite some private foundations in the United States who care about science and are enthusiastic about funding scientific projects. However, a lot of them, if not all, don’t seem to know how they should support science in a complementary way when compared to government funding agencies like NSF and DoE.

Many of these private foundations claim in their mission statement that they intend to support high-risk yet high-reward scientific projects. The goal sounds very good and it looks like exactly what is needed to complement agencies such as NSF and DoE as we all know that government funding agencies tend to support low-risk (often low-reward) projects.

Unfortunately, the practices of these private foundations do not seem to help on their goals according to my own experience. There seems to be a severe disconnection between their reviewing process and intended goals.

Since early last year when I proposed the new mirror matter theory and feasible laboratory experiments for unique tests of the new n-n’ oscillation model, I have been trying to get support from private foundations. Because of the novelty and associated risks with the new theory, I thought that the best chance of getting support for this kind of projects would be through private foundations, in particular, after I read several of their mission statements. After all, government funding agencies are notorious for their risk-averse nature. Fortunately, Notre Dame has been pretty supportive for my applications of such proposals. I thought I got an ideal case for winning such kind of support.

However, what puzzled me was that all of my proposals were rejected by private foundations. Typically they don’t give any feedback at all when they reject one’s project. Their operation is like a black box and it is often hard to know how they actually select projects to support. Luckily, the stuff at the office of Notre Dame foundation relations made diligent effort to get me some feedback when my proposal got rejected by one of the private foundations.

Here shows their feedback regrouped in four points by the ND stuff as an example:

1. started with a statement that they receive many applications in this field, and it is very important to make your idea stand out above the rest.
2. They wanted to see that there was interest from the community in your idea – by the way of citations and publications – and this is the area that your proposal scored lower on.
3. They really liked your methodology, especially the point that your signal was expected to be high during your experiment. This is something that many proposals are lacking or unclear on.
4. Also a note, even though you spoke to NSF Program Officers, they were hoping to see feedback from actual applications to them and other agencies.

Point #1 just states that the competition is very intensive and therefore it is not easy to get supported. Point #3 should be one of the most important criteria for such support – the high-reward factor. They obviously liked the strength of my proposal on this aspect. Another criterion should be stressed on the novelty of a proposal, or, the high-risk factor. Maybe point #1 is related but it is effectively abandoned by the requirements of points #2 and #4.

The ND stuff did not make point #4 clear. According to my previous communications with the foundation, what they wanted to see in point #4 is that one’s proposal has been rejected by the national funding agencies. This might be a valid point if their emphasis is on the novelty of the proposal. However, the intention is distorted due to point #2 (strong community support). In fact, point #2 is what national funding agencies really care about. That is exactly why they tend to be risk-averse. Under point #2, the proposal should be well received in the community and hence tend to get good peer-review reports. Emphasis on point #2 will definitely lead to a low-risk factor.

For private foundations, points #2 and #4 effectively make them support the 2nd-rate proposals that have just been rejected by national agencies. In some sense, they become just a secondary low-risk agency. On the contrary, they could have covered the other end (high-risk) of the risk spectrum of the proposals and supported truly high-risk high-reward projects had they removed the requirement of point #2.

In my opinion, they should have required the opposite of point #2. That is, any proposals with high citation number and good community support should seek funding from government instead of private foundations. Truly novel ideas have never been well received in the beginning.

As such, the best criteria should be: 1. true novelty (high risk/high claims), 2. if it is scientifically reasonable or not, and 3. what unique testable predictions it provides. These new three points should provide better guidelines for private foundations. To be considered novel or high risk, the proposal must have not received enough community support.

However, they can not just simply support the proposals with the craziest ideas or biggest claims, which are often just crackpots. There has to be some baseline. The idea has to be at least scientifically reasonable and self-consistent. At least, it does not contradict with well established data and observations. This criterion should be simply one of pass or not and the baseline should be something close to what the most open-minded scientists require. But where exactly this baseline should be might be still debatable as human innovation can often go beyond our imagination. In reality, this may be the most difficult point to follow.

The last point (unique testable predictions) may not exist for many proposals. If they do, such proposals should be rather supported first than others without such predictions. This distinctive merit should be weighted more, in particular, when too many proposals are in competition.

I wish these foundations would adopt better review criteria as suggested above to truly fulfill their mission of supporting high-risk / high-reward scientific projects.