Racial Disparity in K99 Awards and R00 Transitions

Oh, what a shocker.

In the wake of the 2011 Ginther finding [see archives on Ginther if you have been living under a rock] that there was a significant racial bias in NIH grant review, the concrete response of the NIH was to blame the pipeline. Their only real dollar, funded initiatives were to attempt to get more African-American trainees into the science pipeline. The obvious subtext here was that the current PIs, against whom the grant review bias was defined, must be the problem, not the victim. Right? If you spend all your time insisting that since there were not red-fanged, white-hooded peer reviewers overtly proclaiming their hate for black people that peer review can't be the problem, and you put your tepid money initiatives into scraping up more trainees of color, you are saying the current black PIs deserve their fate. Current example: NIGMS trying to transition more underrepresented individuals into faculty ranks, rather than funding the ones that already exist.

Well, we have some news. The Rescuing Biomedical Research blog has a new post up on Examining the distribution of K99/R00 awards by race authored by Chris Pickett.

It reviews success rates of K99 applicants from 2007-2017. Application PI demographics broke down to nearly 2/3 White, ~1/3 Asian, 2% multiracial and 2% black. Success rates: White, 31%, Multiracial, 30.7%, Asian, 26.7%, Black, 16.2%. Conversion to R00 phase rates: White, 80%, Multiracial, 77%, Asian, 76%, Black, 60%.

In terms of Hispanic ethnicity, 26.9% success for K99 and 77% conversion rate, neither significantly different from the nonHispanic rates.

Of course, seeing as how the RBR people are the VerySeriousPeople considering the future of biomedical careers (sorry Jeremy Berg but you hang with these people), the Discussion is the usual throwing up of hands and excuse making.

"The source of this bias is not clear...". " an analysis ...could address". "There are several potential explanations for these data".

and of course
"put the onus on universities"

No. Heeeeeeyyyyyuuullll no. The onus is on the NIH. They are the ones with the problem.

And, as per usual, the fix is extraordinarily simple. As I repeatedly observe in the context of the Ginther finding, the NIH responded to a perception of a disparity in the funding of new investigators with immediate heavy handed top-down quota based affirmative action for many applications from ESI investigators. And now we have Round2 where they are inventing up new quota based affirmative action policies for the second round of funding for these self-same applicants. Note well: the statistical beneficiaries of ESI affirmative action polices are white investigators.

The number of K99 applications from black candidates was 154 over 10 years. 25 of these were funded. To bring this up to the success rate enjoyed by white applicants, the NIH need only have funded 23 more K99s. Across 28 Institutes and Centers. Across 10 years, aka 30 funding cycles. One more per IC per decade to fix the disparity. Fixing the Asian bias would be a little steeper, they'd need to fund another 97, let's round that to 10 per year. Across all 28 ICs.

Now that they know about this, just as with Ginther, the fix is duck soup. The Director pulls each IC Director aside in quiet moment and says 'fix this'. That's it. That's all that would be required. And the Directors just commit to pick up one more Asian application every year or so and one more black application every, checks notes, decade and this is fixed.

This is what makes the NIH response to all of this so damn disturbing. It's rounding error. They pick up grants all the time for reasons way more biased and disturbing than this. Saving a BSD lab that allegedly ran out of funding. Handing out under the table Administrative Supplements for gawd knows what random purpose. Prioritizing the F32 applications from some labs over others. Ditto the K99 apps.

They just need to apply their usual set of glad handing biases to redress this systematic problem with the review and funding of people of color.

And they steadfastly refuse to do so.

For this one specific area of declared Programmatic interest.

When they pick up many, many more grants out of order of review for all their other varied Programmatic interests.

You* have to wonder why.
__
h/t @biochembelle

*and those people you are trying to lure into the pipeline, NIH? They are also wondering why they should join a rigged game like this one.

13 responses so far

  • JL says:

    DM,
    You make a good point that this should be easy to fix with a few pickups.

    Maybe I can make an additional point: if it would only take a few pickups to balance the success rates, isn't this evidence that a larger issue is too few applications from some groups? Balancing the success rate of the tiny number of applications from black PIs will help, and should be done, but it will barely touch on the overall proportion of PIs from each group.

  • slipodza says:

    You know why NIH won't do anything to directly address racial disparities. This thing called "congress", which has fully revealed itself as a state-sanctioned white supremacist racket at this point. It doesn't make it right, but all these government agencies know exactly on which side their bread is buttered

  • Rheophile says:

    @JL - Balancing success rates would not solve underrepresentation. But it would solve biased refereeing, which is no small thing! Having seen really convincing evidence of bias at the R-scale seven years ago, and now this - as DM says, we are asking people to join a system rigged against them. It doesn't fix the pipeline, but at least we'd feel better about recruiting young URM folks if we are giving them a fair shot.

    Because the actual $ cost to the added pickups is so low, I can only assume that the NIH is afraid that affirmative action for underrepresented groups would have a big political cost, and make them a target. Either that or NIH has so internalized the "white moderate" disease that they reflexively go "pipeline" when there's a simpler answer.

  • drugmonkey says:

    Balancing the success rate of the tiny number of applications from black PIs will help, and should be done, but it will barely touch on the overall proportion of PIs from each group.

    With African-Americans representing 12% of the US population, yeah, it'd be nice to budge the percentage of K99 applicants away from 2.2% (of the applications where race was reported). But bringing this up in the context of outcome bias for the existing pool of applicants just serves the distracting and minimizing agenda of not doing anything. We can work on these two things independently, especially since the one has a really easy fix. And btw, the easy fix is to equalize success rates. It really wouldn't be that hard to boost African-American success rates even higher than those for good-old straight-whitey-male-Americans, if we really were committed to diversifying the long term successful faculty pool.

    You know why NIH won't do anything to directly address racial disparities. This thing called "congress",
    Yes, I am in fact aware of the fear that NIH has of right wing congressional inquiries. This is why I consistently say that Francis Collins should take the Directors aside, in private, and tell them to fix this funding bias shit on the DL.

    Either that or NIH has so internalized the "white moderate" disease that they reflexively go "pipeline" when there's a simpler answer.
    NIH is a very large place and the usual cultural baggage comes with that. Academic issues of diversity are clearly caused in substantial part by overt or implicit bias on the part of the people who are in power. So yeah, it isn't only fear of Congress that is the problem here. There are in fact anti-diversity opinions and I would warrant that just as with anywhere else, these are correlated with advancing age, whiteness, maleness and the delusion of meritocracy of science. In short, the closer you get to people who can do anything about the disparity, the less likely they are to really believe that it is a problem that needs to be addressed. This is why I beat the drum on the intertoobs whenever I can.

  • A Salty Scientist says:

    I can only assume that the NIH is afraid that affirmative action for underrepresented groups would have a big political cost, and make them a target.

    My educated guess as well, considering how the NSF--and their notable use of "Broader Impacts" as a merit criterion--is targeted more by certain politicians than the NIH.

  • A Salty Scientist says:

    This is why I consistently say that Francis Collins should take the Directors aside, in private, and tell them to fix this funding bias shit on the DL.

    My experience is with NSF, where both geographic and PI diversity is part of "programmatic interest," and POs have a lot of latitude. Why NIH apparently sticks their head in the sand on these issues is beyond me.

  • drugmonkey says:

    Why NIH apparently sticks their head in the sand on these issues is beyond me.

    Real talk: My experience is that even seemingly well intentioned majoritarians of academia do not really get it. They are literally unable to comprehend that disparity really means that there is a bias that needs to be made up for. They cannot grasp this. They are convinced at their very deep center that the people who are in front of them right now really are inferior. Lipservice is made to the idea of diversity but they do not really believe it.

    This is why if solutions are sought they are always displaced (pipeline) or extra (Dean's hire, special competition"diversity K99" ploy, etc). This latter means that we cannot ever redress the bias against one group (women, racial or ethnic groups) by decreasing the bias in favor of the other group, i.e., white men, by even a micrometer.

    Notice how NIH "equalized" male and female R grant success rates in 2003 but for fifteen straight years female rates never actually exceed male rates?
    Things that make you go hmmmm.

  • wally says:

    Another issue currently is that some people from marginalized groups (like people of color and LGBT people) want to do research on marginalized groups. Some POs and review committees get nervous about that. I know of a researcher who was told to change the title, abstract, etc. to remove mention of the marginalized group (this happened under Bush jr as well to my postdoc mentor) and another researcher who was told by the review committee that the population to be studied wasn't a priority one for the NIH (which was bizarre in the context of the health issue and who is disproportionately affected - but not bizarre given the political climate). I know another researcher who has historically gotten funding for research on a marginalized group from a specific institute, got a super low and extremely fundable score on the new R01 and was told it wasn't a priority, wouldn't be funded, and basically to never apply to that institute again. So, I know some ECP from marginalized groups think they shouldn't even bother applying as a result.

  • […] to the bias against black PIs when they try to get research funding (Ginther et al., 2011), Asian-American and African-American K99 applicants are also at a disadvantage. These issues trigger my usual remarks about how NIH has handled observed disparities in the past. […]

  • Almost tenured PI says:

    I'm surprised that 40% of black K99ers didn't transition to the R00. The most obvious possible reason is that they didn't get faculty positions, yet this seems hard to believe. I'm at an R1 university (that is admittedly not on one of the coasts). We do not get black faculty applicants. Literally never. I wish it was more widely known that credible minority applicants are almost certain to get an interview at my uni because of the possibility of hiring using a separate pot of money for "diversity hires", thus allowing us to ultimately hire two people. So I'm having a hard time imagining why these faculty applicants that have good enough records to get a K99 would not be able to land a faculty job. Are they opting out of academia at a higher rate? As my anecdata suggests, are they applying too narrowly? Whatever the answer, it seems like this needs to be figured out. What's the point of giving out more K99s to this group if they're not becoming faculty?

  • drugmonkey says:

    If your Uni is already unattractive to POC why would they fact they have a K99 make it attractive?

  • Almost tenured PI says:

    My comment above was suggesting that perhaps POC with K99s recognize that they are unicorns and in high demand. Yet, Harvard can only make so many hires.

  • Dnaman says:

    There were only 25 black K99 (over 10 years). 40% didn't transition, that's 10 over ten years. So, that's about one per year.

    Just perusing the NIH Report and googling old K99's. It looks like the most common reason for not transitioning is that they took a research job outside the US.

    Is it possible the black K99'ers are more likely to leave the US? A significant fraction of the black PhD students I've known were international students.

Leave a Reply