How committed is NIH to addressing its race problem? Hint: kinda sorta

Jan 16 2013 Published by under Uncategorized

"Now we know, and now we have a chance to do something about it. The leadership here is absolutely committed to making that happen"
-Francis Collins (NIH Director)

Remember that  NIH race thing?

ginter11-r01byrace1

No? I quick reminder, in the summer of 2011 a paper published in Science reported that NIH grants showed a racial disparity. When controlling for a host of factors, African-American scientists still tended to have less of a chance to be awarded a grant from NIH.  There was a minor storm of commentary and coverage, including Science and the New York Times. NIH officials stated that, 1) the disparity was a big concern and 2) they were going to do something about it (NIH reactions here and here). Initially that something was, of course, to form a committee. Not exactly a dramatic action, but I suppose understandable given the giant bureaucraticness of NIH. There needed to be a report written and recommendations made. That was well over a year ago and what have the results been? (a reaction here).

Basically four things have been recommended,

1) Increase undergraduate research opportunities
2) Create a mentoring network
3) Hire a Chief Diversity Officer
4) Test some possible bias interventions, e.g., "diversity training".

Even if these recommendations were enacted tomorrow, and worked exactly as hoped, the  gains would be slow and marginal. #1 seem to more address the problem of under representation. The report is not about under representation it is about an outcome disparity. #2 I'm not sure how this will work. Is this network going to be the Facebook of science mentoring? Pick-a-Mentor? Mentoring is great but it's not as simple as being assigned a mentor who then provides you with all knowledge needed for professional development. In my own experience it is a constant active slog to find people and pick their brains. #3 Seems good enough. Of course this person will have to take actions and have the power to take actions other than issue more reports. #4 The standard diversity training response. How well this will work may depend on the source(s) of the disparity.

I want to contrast this with NIH actions regarding other issues. In that same blog post I linked there is also discussion of the ongoing early career investigator issues. Here is a selection of some of the actions directed towards that problem.

NIH plans to increase the funding of awards that encourage independence like the K99/R00 and early independence awards, and increase the initial postdoctoral researcher stipend.

In the past NIH has also taken actions in modifying how grants are awarded. The whole Early Stage Investigator designation is part of that. Grant pickups, etc.

A commenter on the Rock Talk NIH blog put the contrast this way:

A grant outcome disparity for younger investigators leads to a massive and immediate initiative to fund proposals out of order.
A grant outcome disparity for African-Americans leads to….”initiatives” and “study”. In other words, nothing.

It's tough for me to disagree with this. I don't want to get all Kanye ("NIH doesn't care about black researchers"), but priorities, be they individual or institutional, really come though not in talk but actions. Now, I don't have any special knowledge about the source or solution to the racial disparity. But the NIH response here seems more along the lines of adequate than overwhelming.

If it is the case that NIH leadership believes the report flawed, or the effect due to something beyond their control, then they should say so. Thus far I have not seen any indication. The optimistic view of NIH's plans is that a multifaceted approach is more likely to show some effect. Another view is that NIH is throwing a few okay plans at the wall and seeing what sticks.

Two interesting notes:

  1. The NSF numbers don't seem to show nearly as much a disparity.
  2. Interesting wrinkle in the data: black scientists had less co-authors per paper.

21 responses so far

  • drugmonkey says:

    The comment you quote from RockTalk blog is exactly right. The NIH took defined action on the ESI and NI apps by immediately starting to fund them out of order of the review rankings. They didn't table it for more study or even, so far as I am aware, appoint a Chief Youngsters Officer. They just started picking up grants.

    The contrast with the fate of African-American PIs could not be more striking. It shows very clearly that the NIH as a whole does not in fact agree that these data represent a bias. That is, the NIH is not agreeing that there are applications that should have been funded save for the bias. Thus it is no wonder that they are trying as hard as they can to find some other reason for why the disparity has occurred.

    I will note that, interestingly, when it comes to younger, newer investigators, the NIH has always (R29/FIRST award, checkbox, ESI designation...there was even the R23 prior to the R29!) tried to ameliorate the bias against the younger investigators while essentially admitting their apps might in fact be inferior. Right in the review criteria it suggests that reviewers should feel free to take their relative handicaps into account. Think we're going to see anything like this for the differences observed in Ginther? "Reviewers should account and not penalize African-Americans for having fewer co-authors"....hahhhhahah, oh, I kill myself. hahahaha. right. /wipes tears.

  • I agree with DM: the one policy that could rectify this problem is one that can never be (and will never be) mentioned in this political climate.

  • drugmonkey says:

    They don't have to "mention" jack squatte. FC could sit down his Directors (as a group or better, individually) and say "this disparity shit makes us look bad. make it go away. And no, we won't be discussing how, it is going to eventually look like the attention was sufficient to fix the problem. Get it? ok, good. "

  • Hermitage says:

    #2 honestly pisses me off a little, because the implication is that despite equivalent education, training, awards, papers, institutions, prior NIH funding, and NIH panel service black people somehow don't know how to write a grant and need someone to come tell them how the game works.

    From the major case of bigotry-itis that flared up from "scientists" when Ginther et al. first came out, I honestly think the NIH is afraid to do anything other than some mealy mouthed bullshit because of the barely-concealed racist outcry that would rage across NIH-land. It might even be big enough to get picked up outside of Academic Blogland, and then it'd really be a PR shitshow.

    • Bashir says:

      I think NIH would rather have the racial disparity than the ire of those that may think they are doing too much. As bad as the press may be for having the disparity, I don't think it's nearly as bad as a bunch of "NIH uses Affirmative Action to give grants to sub-par researchers" articles from FOX et al.

      The best strategy for NIH may be to eat around the edges of the problem and hope no one pays that much attention to it.

      • drugmonkey says:

        Agreed Bashir....and they can probably do it if they want to. Picking up one or two more grants per IC per FY would barely make a ripple. Yet the number of AA PIs is low enough this would make a dent on the disparity

    • becca says:

      Isn't the implication that you can't succeed in anything in life without the right connections, which black scientists might disproportionately lack?
      It's only insulting if you think the role of a mentor is to tell you how to do your job. I'm surprised anyone in grad school feels that way, honestly. The role of the mentor is to clear the way for you doing the things that you cannot do for yourself. You can find someone to polish a grant for you, even (especially?) as a raw unknown n00bster. You can't make sure a grant you write is received favorably until you've got a toehold in the field. THAT is a role for a mentor.
      (Ok- one could argue that PIs should know, and probably do, that their job at the beginning IS to become known well enough so their grant will be received favorably, and that one does NOT need a mentor to do this and that it's still a form of telling black scientists how to do their jobs. But I personally tend to think that there is a real 'activation energy' in getting established in a field; a mentor can serve as a catalyst for some of the crucial first steps)

      Which isn't to say that a magic NIH hand wavey program to assign mentors is going to be a reasonable approach to solving the disparity problem. Just that there *could* be a pathway toward ameliorating the disparity there.

  • A. Postdoc says:

    Getting a grant is inverse to grantsmanship - i.e. the sport of lying and getting away with it. Grantsmanship is correlated with low morals, perhaps the problem is that African Americans just have better morals.

    Don't believe my first sentence, here it is from a white guy at a top school, grantsmanship is lying and not getting caught. Too bad for him (and his mentees) he got caught:

    http://retractionwatch.wordpress.com/2013/01/09/paul-muchowski-sanctioned-last-month-by-the-ori-for-falsely-reporting-results-offers-an-apology/#more-11638

  • drugmonkey says:

    The lower rate of submitting revised grants didn't impress you Hermitage?

  • Anonymous Consultant says:

    One question I have relates to the places where the bias has impact. It would be really interesting to tease out the data to look for the correlations. Is there a difference between success rates of African Americans applying from State U, Ivy League or HBU? Is there a difference between two African American applicants from State U, where one has an HBCU on his/her biosketch?

    I think about these questions for a number of reasons, one of which is that I've had the opportunity to work on proposals with people in a lot of different contexts. IIRC, NIH pulled out the data supposedly to normalize for pedigree, etc. I more consistently see a failure of mentoring for African-American scientists, regardless of the institutions in which they trained and now work. As an outside consultant, I find myself telling them things their post-doctoral or even graduate advisors should have told them. (I've talked about this on RockTalk.) Point being, bias enters the picture in training (not even worrying about undergrad research opportunities), because the mentoring seems uneven. There are so many things that play into this, some of them well meaning ("I don't want to offer correction because I don't want to be seen as too critical to a person of color."), some of them fear-based ("I don't know how to talk to a Black person. I will look stupid and let my unconscious prejudices show, and I don't want to look prejudiced, so I won't say anything."), and some just stemming from un-admitted prejudice (no example quote needed here). Then when it comes to the applications themselves, you add the often unconscious bias from reviewers that correlates with 'markers' of being African American. We have systematic problems. Yes, this is a microcosm of the larger cultural picture.

    Some of the bias is in your face in ways I don't think NIH recognizes. I'm not sure what NIH meant to address with the SCORE program, the SC1, 2, and 3 grant mechanisms available to researchers at majority-minority institutions. If you read the solicitations, they seem meant to help researchers at these institutions transition to standard NIH research funding, but I see two things that are wrong--one with the PAs themselves and one with the way reviewers react to investigators who compete for R funding when they are eligible for SCORE grants. First, the SCORE PAs themselves seem pretty paternalistic, because of the requirement for a mentor for applicants who are TT, not post-docs or K applicants. "Just because you're at a SCORE-eligible institution, you probably need help" is the message. Second, the reviewers also appear to resent any SCORE-elegible applicant from going for a 'real R03' or, heaven forfend, an R01. I regret the hear-say of what follows, but clients have told me that reviewers have essentially told them to go back in the SCORE box--to use the money set aside for them, and stop trying to compete in the general pool. I wonder if that doesn't also affect the overall success rate at the R level.

    So I wonder if there isn't some hesitancy at NIH that the "just fix it" approach they took for younger faculty wouldn't apply--that the assumption, some conscious or unconscious consideration might be that it isn't the applications that are inferior, as DM notes above for young faculty, but that the science might somehow be. But if the mentoring isn't as good, the standards aren't set just as high early on... Maybe that's the thinking behind the mentoring requirements for SC grants, but it would be late at that point, and paternal at the least. I'm in Bashir's blog, and he knows better than I would if I'm just being a different kind of paternalist here.

  • drugmonkey says:

    Regarding the "SC Box", this issue was a major reason the R29/FIRST was a failure. It became the "expected" starter grant. And the budget was crippled.

    • Anonymous Consultant says:

      Right, and SC budgets are pretty much insufficient, too, so that the intended 'ladder' to the R01 wouldn't be adequate, especially since it would be (assuming the progression of 2-year grants) six years of insufficiently funded work leading to the R01 application. It's a rare person, who probably has additional resources, who could produce enough quickly enough under those circumstances to be competitive. And then reviewers come in questioning productivity without thinking about the actual resources that were available...

      • drugmonkey says:

        exactly the problem with the FIRST award. Not to mention the problem with PO policies that used to, when picking up a junior person's grant out of order, cripple them with the old "let's see how you do sonny" paternalism.

  • [...] How committed is NIH to addressing its race problem? Hint: kinda sorta [...]

  • [...] What’s up with the CDC Influenza Vaccine Effectiveness Calculations? Is it really healthier to be a few pounds overweight? That’s not what the study says. Flu studies suggest vaccine ‘match’ not super predictor of effectiveness Averting the Approaching Apocalypse How committed is NIH to addressing its race problem? Hint: kinda sorta [...]

  • [...] noted recently by Bashir, the NIH response to the Ginther report contrasts with their response to certain other issues of [...]

  • […]  Black scientists less likely to get NIH grants. […]

  • […] Remember her? Wonder how NIH is coming along on the R01 funding disparity issue? HAAHA, I crack myself […]

Leave a Reply