Comment by epistasis | Hacker News Reader

Comment by epistasis | Hacker News Reader

Parent Full thread

>the first is not. If you are not doing research efficiently, then you should not be doing research.

My point, and perhaps I could have been clearer, is that funding is shrinking, while the quality of proposals is increasing or getting better. Funding rates for grants are now in the 5-10% rate, which means that even excellent, efficient, outstanding labs that are leaders in their fields can starve due to random chance.

It ha nothing to do with being entitled to research money. It has everything to do with the US spending research money efficiently and not capriciously. A research lab is not like a startup in that the intial time investment is quite long, and cutting that off inappropriately is wasteful. Additionally, it is incredibly inefficient to spend most of a researchers time churn through good proposal after good proposal, and no let them do research.

US science is shrinking dramatically right now, and rebuilding it will be a slow and long process, an if we lose the crown to China in the interim, the US may never be the bastion it once was.

I disagree. The quality of proposals is getting worse. There's a horrifying cycle of cronyism, lack of managerial skill that is going into science. People are promoted into research positions based on the impact factor of the journals they publish in, not the quality of their work. High impact factor journals publish things that are more likely to be scandalous than solid. The most important aspects of scientific success (versus excellence) is "were you your PI's golden child", and "just how lucky did you get"?... etc. etc. etc.

These trends are reflected in the proposals that I see getting pitched.

1) Who cares if "we" "lose the crown" to china? 2) I'm not worried about china, anyways, the scientific culture there is miserable and dastardly, and deeply entangled with the cultural value of "face" (I'm east asian, I think this applies to Korea and Japan, and troubling - increasingly in the US - see Hellinga, for example) and values Feynman's "extra integrity" less.

If you claim it has nothing to do with being entitled to money, then perhaps you should rephrase. Cutting and pasting what you said,

"First, the researcher has run out of funding ... the first is a travesty"

Which implies that a researcher running out of funding is a travesty. In the general case, as implied by your use of the indefinite article, that is not true.

Do you claim that 90% of proposals are not worthy of funding? Because it it was not an extremely high number, we could significantly increase spending and still only fund the best.

You stated: " It can be debated if grant reviewers are actually doing a good job (versus being a whitewashed version of cronyism)"

If they are already evaluating research and researchers the way the rest of profession does, I don't see how cutting funds would help. It would not drive better proposal writing. And if the cronies are already getting their money anyway, the added batch might have a higher percent of high quality.

I understand the desire to punish the undeserving of your profession, but when weighed against the advancement of science (even if artificially inefficient)...I don't care. And cutting funding won't make a significant dent in the problems you describe.

I'd claim that a lion's share of the top 10% of proposals are not worthy of funding. The best proposals are probably usually rated in the 70-80% range, depending on the funding agency (NSF, NIH, DOE), but most of it is utter embarassment.

"when weighed against the advancement of science (even if artificially inefficient)...I don't care." Well, no, we're talking about funding here. What if the structure of funding was causing a net regression in the advancement of science? Is it a safe assumption that funding directly correlates to advancement?

"And cutting funding won't make a significant dent in the problems you describe." If I had my druthers, I'd cut science funding to zero (there would be other ways to raise money for science - endowments, 501(c)(3)s, foundations, bake sales, etc. Keep in mind Peter Mitchell validated the chemiosmotic effect, a cornerstone of basic biochemistry and cell biology, by launching his own completely independent science research foundation) That would make a significant dent in the problems I describe.

What about the medical funding for stuff like cancer, where the vast bulk is in the public sector?

Also, while I agree with Theodore Sturgeon that 90% of everything is crap, in science it is especially hard to tell what that 90% is going to be in advance, so I think it is better to overfund and then concentrate on the improvement of quality in the review process to filter out the crap.

I guess I'm uncomfortable using taxes to pay for stuff (especially overfunding) that will be mostly crap. Taxes, because if you don't pay for them, you go to jail, should be subject to a particularly strong level of accountability[0], and "we don't know what is going to be useful" is just really a recipe for spending abuse.

Peer review is one thing (and it looks like it's going to be getting better)... But reviews for funding - that's really tricky. How do you select whom to review science? In our current system, you wind up with program administrators (people who advocate and shepherd funding projects) who are mostly[1] people who 'couldn't make it in science'. So you get effectively second-rate-intelligence fanboys and fangirls who get paid six figures to fawn over the scientists with the biggest mouths and most outlandish claims. To actually conduct the reviews, typically PIs are hired (in a secreted process[2]), so there's of course a level of cronyism involved - not to mention how PIs are generally clueless for other structural reasons. It is hard to believe that this system is efficient or effective at being a good steward of the public trust.

[0] Science is not by far the worst area where government excels at failing to be accountable, but I also see no reason to increase our investment in a system that isn't working.

[1] I know one program administrator who isn't incomptent, she's my ex-boss; the reasons why she wound up as a program administrator are rather complicated and everyone hates her because she isn't interested in doing things the usual way. My neighbor growing up - who introduced me to science - was a physics administrator was probably also not incompetent, but I don't really know.

[2] by which I mean the process is not secret, but the actual goings on are, i.e. you don't know who reviewed you etc.

I agree that using taxes to pay for stuff that isn't going to be used is wasteful, however there is waste in all activities and at least science funding has a good economic return on the whole, even when there is a lot of waste. Besides, I don't see stuff like CERN happening privately.

Also, to avoid confusion, I was meaning tightening up the peer review process and loosening the funding review. The biggest problem I see in poor science is not in the fact it was funded in the first place, but that it is difficult to filter out later. Like you said, funding decisions are tricky, but I think the trick would be to fund even more things and concentrate on excellence in peer review. This would, I think, increase the overall benefit to society, which is the only thing that justifies paying for it with tax in the first place.

I guess we'll just have to disagree. In my opinion, "benefit to society" is a necessary but insufficient condition to morally justify paying for something via taxation. The other necessary condition being "requires force to administer", for example, police (if someone is murdering another person you are going to have to detain them with force).

The reason, why it is insufficient, is the question, "if it were so beneficial to society why didn't society pay for it voluntarily?" Now, the usual argument is "because they were too stupid or selfish to" - but then, why would we expect the politicians to be any better? Day by day, it seems to me that the political class is more capricious, narcissistic, and self-centered than most humans they rule over. I do not see this as a particularly effective cadre of individuals to be overseeing spending money on science research, however indirect. Might as well just get money from the people directly with outstretched hands.

To your original point, "What about the medical funding for stuff like cancer, where the vast bulk is in the public sector?" This is a particularly bad example because there is a large BINGO (although I would not call it a shining example of an efficient charity) that funds quite a bit of cancer research - the ACS. The early decades of the NCI were a disaster. To be sure, the NCI has gotten way better over time, and there are some indispensible services they provide (which nonetheless in theory could be spun off to nonprofit private orgs) like the NCI-60 cell line screening service; but there are also plenty of stories about the siloing/turfwarring that goes on between the NCI and the NIH - which has led to severe inefficiencies and plenty of lost opportunities. The ex-boss program manager manages projects for the NIH that OFTEN cross over into NCI territory and the political machinations she's complained about make my head spin.

I'm not sure that "requires force to administer" is that good a qualifier for deciding on what should be funded by taxes, assuming that you mean force in the political sense of force against people. By that standard you wouldn't fund fire departments.

Also, government paying for things is society paying for things. Government is part of society, not an appendage.

On the subject of cancer research, I don't know that much of the situation in the US, but I did look into this a while back and found that the amount spent by governments globally dwarfs the money spent by industry and charity combined. I apologise for not having the figures to hand though and if you think I am wrong on this, let me know.

edit - just had a look at your profile... I was not aware of what you are involved in. Good luck with it all, I hope you do well and I will stop trying to tell you about cancer funding as you are bound to know a hell of a lot more about it than I am.

Reply on news.ycombinator.com