"In my world there is an important middle ground: "uncertain" - as in I personally do not yet have enough information to draw a conclusion. That's where I am in regard to this assertion."
Yeah, maybe this is where we're tripping up. I hold almost all specific claims in the "uncertain" category. Certainly all claims in the social sciences.…
"In my world there is an important middle ground: "uncertain" - as in I personally do not yet have enough information to draw a conclusion. That's where I am in regard to this assertion."
Yeah, maybe this is where we're tripping up. I hold almost all specific claims in the "uncertain" category. Certainly all claims in the social sciences. Because almost all information could be "disproved" or cast into doubt by new evidence or more careful analysis.
But once there's *enough* evidence to satisfy me (multiple sources, logical conclusions, zero claims of the inverse), they move to the "more likely than not to be true" end of that spectrum. As I said earlier, it's about how much time one is willing to spend digging into every piece of data on a single assertion in order to "prove" it.
If a single piece of research claims the opposite, I note it, but that single data point probably won't move me very much unless it presents something compelling. Especially if it appears to me to have fairly serious flaws (to be clear, this conversation has moved my confidence in the assertion back toward "uncertain", just not very much). Again, COVID is an excellent example of how futile it is to try to prove something beyond any doubt. And that's far harder science than this.
If I'm writing an article, I'll never base my argument on a single piece of research or a single data point. Precisely because if that research is cast into doubt, the entire thesis falls apart. So I'll cite numerous independent pieces of evidence, all pointing to the same conclusion, and say, "All of this together is why I believe X to be true."
My attachment to the "factuality" of a single piece of data therefore, is usually pretty loose. I'm convinced enough by "the assertion" that I'm happy to use it in an aside to a reader in a comment section. But I'd never use it as the cornerstone of an article. I'm not even using it as the cornerstone in my argument that Charles Murray is racially biased.
So yes, while I think it's wise (and sadly rare), that people keep information in the "uncertain" category, it's paralysing if we don't accept some information as "likely enough" to be true for a particular purpose. If everything you said had to be true beyond any possible shadow of disagreement, you'd never say anything. I mean, look at how many people (including many fellow experts in his field) claim the findings in Charles Murray's book aren't true.
So yeah, if there were definitive evidence that black people were the primary beneficiaries of affirmative action (or that white women and black people benefited more or less equally), it wouldn't change anything for me, because a) lots of people already assume that to be the case (all without *any* evidence) and b) none of my ideas rest on it being false. But some of Charles Murray's arguments about affirmative action *do* seem to rest on it being true.
And the wider point is that grievances about racism don't rest on things like this single assertion being true. Or, for example, on the Atlantic slave trade being the most brutal form of slavery. Neo-progressives aren't wrong to feel aggrieved, they're wrong in what they suggest is done about those grievances. And in implicating all white people as the source of those grievances. But there is soooo much indisputable evidence of the cruel, deliberate, and state-sanctioned oppression of black people in America, that these assertions being proven wrong beyond a shadow of a doubt wouldn't move the needle.
Good points, Steve. In particular, I agree with your final paragraph, and many of your points about evidence and "proof".
(Well, I might question some of the specific points or degrees of grievance used by neo-progressives, but I do agree that many of the grievances are valid and that their prescriptions are the bigger issue).
I'm glad that we agree that absolute proof is not available, and that neither of us expects that. That seemed like a red herring, so it's good to take it off the table.
As I've said, for years I had (without discomfort) put "the assertion" into the "likely true" category. But I had done zero real research (since it didn't seem to have big operational implications), I was just lazily assuming it was probably true, because I had heard it many times, and had never heard anybody dispute it. I didn't seek out information about it.
Then *relatively recently*, I heard somebody dispute it, so I did a cursory search and after filtering, the first peer reviewed paper I found came to a contrary conclusion, and also described the flaws in superficial analysis. I explicitly did not take that as definitive, but as enough to raise some questions of factuality, and as raising the right questions (unlike the popular press).
So for me, the currents status my knowledge about "the assertion":
* mainstream popular sources: near universal belief, cited by activists as another grievance; I find most activists to be deeply into confirmation bias and cherry picking, so this consensus is not convincing and I try to look for better sources
* scientific sources: 0 sources supporting it, 1 source questioning it, not nearly enough evidence to make any conclusion, but enough to consider it "needs more investigation". I do not know if this paper is part of a consensus, or a rare exception to a consensus, or if there is n scientific consensus.
If I relied entirely upon the sources which also consider it to be established "fact" that the George Floyd murder was obviously due to racism, then "the assertion" would be a no-brainer; we know that the popular press is very much in favor of the assertion, when it comes up. But is that because they neutrally assessed the evidence, or (once again) because it fits the same narratives which support their George Floyd "factual" conclusions?
I figured that if your research into the matter had established that there was indeed a scientific consensus supporting the assertion, then you could easily pass on a few links to the best of them, opportunistically bootstrapping my low-priority research. As I've said, I would have asked the same from somebody certain of the opposite conclusion. I don't have a dog in this race, other than seeking truth.
Science is not immune to ideological or other biases, but it typically does at least consider major confounding factors and avoid the worst abuses of process that the popular press often commits. Like assuming that 100% of the increases in female and POC employment is based entirely on AA, and then citing statistics which have meaning only if you accept that assumption as fact.
What I took away from the paper I cited was (1) they defined the issue more carefully, (2) they acknowledged the difficulty of determining factual truth, (3) they noted the major confounding factors which needed to be taken into account in any serious study, and (4) within their limited scope, their analysis actually disconfirmed "the assertion" with much more care than the popular press. Because it's only one paper of limited scope, #4 has little weight in isolation - but note that points 1-3, which are still very valid in understanding the deep conceptual flaws in many journalistic or activist "analyses". We should not think that #4 was the only thing of value in the paper.
TL;DR: The paper did not establish any "scientific consensus" by itself, but it did identify why non-scientific analyses may be deeply flawed, by noting things which absolutely *must* be accounted for in a serious study.
So far, you are asserting that science has come to consensus supporting the conclusion so that the only peer reviewed study I have seen *must* obviously be an outlier, which thus can be summarily dismissed as having little weight compared to that established broad scientific consensus.
However, you haven't provided even one citation of a (science) journal article you has used in forming your conclusion, much less any indication that your previous research had established that there actually *is* such a scientific consensus - based on your having reviewed many such scientific papers and found a consensus among them (with the paper I cited being the first and assumed only exception to come to your attention).
> "I'm convinced enough by "the assertion" that I'm happy to use it in an aside to a reader in a comment section. But I'd never use it as the cornerstone of an article."
Fair enough. Good practice.
> " if there were definitive evidence that black people were the primary beneficiaries of affirmative action (or that white women and black people benefited more or less equally), it wouldn't change anything for me, because a) lots of people already assume that to be the case "
Steve, are you saying that you both that (1) you know that lots of people assume that black people rather than white women are the primary beneficiaries of affirmative action, and (2) you know of zero claims of the inverse of the assertion? ("there's *enough* evidence to satisfy me (multiple sources, logical conclusions, zero claims of the inverse)"?
Oddly, I had never run into anybody asserting that POC were the primary beneficiaries of affirmative action until fairly recently, as I've described. I had ONLY run into "the assertion", which I have seen many, many times. And that asymmetry makes sense; "the assertion" is part of the narrative of ongoing pervasive oppression, but the opposite (that POC have benefitted more) does not lend itself to any common narratives. We won't find angry far right folks asserting that it's a known (and implicitly outrageous) fact that white women have received fewer benefits from AA than POC. This is not ideologically symmetric, the incentives for confirmation bias are not equal in both directions.
Obviously, we read different places and our experience will differ, but I do think that most of the mainstream media is on the side of assuming "the assertion" to be a known fact, as you believe, and while you may believe that "lots of people" assume otherwise, I suspect that lots *more* people assume the assertion true than false.
Let's be careful here not to mis-frame the question. That is, one cannot assume that the assertion must be factually true, UNTIL and UNLESS somebody can provide "definitive evidence" that the opposite is true; there has to be room for "the evidence known to me is mixed, not yet sufficient to form a factual conclusion". (Which happens to be my current estimate, pending more research).
One can question whether an assertion has accumulated enough solid evidence to be called a known "fact", without needing to counter-assert that enough evidence has accumulated to establish it's opposite as a known "fact". Sometimes your framing of this does not appear to acknowledge this.
And before I close, I will note that there is a lot of sloppiness in the popular press (including the one you cited) between using statistics about "women" versus statistics about just "white women". I noticed that immediately; did you? Reread the article and see how often they refer to each. In general, women of color can get grouped in on either side of the statistics, depending on what the journalist wants to "prove".
And I'm still wondering if you think there is anything wrong if white women (or women in general) actually had benefitted more from AA than POC (assuming that both needed it to overcome previous stereotypes). Since you are not generally trying to reinforce a grievance narrative, does it even matter to you, beyond curiosity and respect for truth? Do you have any preference for assertion to be shown true or false?
Good discussing this with you, Steve. If we can discuss things well, we can help refine each other's views, on the relatively few points where we disagree.
"If I relied entirely upon the sources which also consider it to be established "fact" that the George Floyd murder was obviously due to racism, then "the assertion" would be a no-brainer;"
Yeah, as demonstrated by my recent conversation about the Atlanta, an assertion of racist motivations in the popular press is worth next to nothing. But that's quite different to a statistical analysis of employment and income advancement during affirmative action. It's not really fair to compare the two.
Also, there is scientific "evidence" insofar as the assertion is based on scientific research. We could debate the quality of that research, but it's untrue that there is 0 evidence. There is research claiming that the assertion is true, there is a paper saying the assertion might not be true, but there are 0 papers claiming that the inverse of the assertion is true. Sorry, I thought there were links to the research in the TIME article I originally linked. This one has some (https://www.vox.com/2016/5/25/11682950/fisher-supreme-court-white-women-affirmative-action).
Interesting that you say you've never heard of anybody who questioned the assertion. I hadn't even heard it until about 5 years ago. Until then, all I'd ever heard is the stereotypes that are spun from the narrative that black people are the primary (or even in many people's minds the *only*) beneficiaries of AA. In his interview with Coleman, Charles Murray states this extremely clearly (https://youtu.be/wCJFr6zB2NM?t=457):
"The reality of aggressive affirmative action, which nobody wants to say, is (but everybody kinda knows), whites widely consider that any time there is a new black face in the office, that's an affirmative action hire. That's the default assumption."
This absolutely matches my experience in talking about the issue or hearing people talk about it. But I know of no *evidence* of the claim. Just the culturally accepted assumption that AA is a "black" thing.
No, I don't think there's anything particularly wrong if AA benefitted white women more than POC. No meaningful preference for the assertion to be true or false (I'd find it funnier if the racists who use AA as a cudgel for POC were proven wrong I suppose). but yes, the grievance aspect is non-existent as far as I'm concerned. I wouldn't feel aggrieved if the assertion were proven false. Women *do* deserve more opportunities in the workplace. But if a new female face pops up in the workplace, especially a new white female face, the assumption is not that she's an AA hire. This the issue.
> "But that's quite different to a statistical analysis of employment and income advancement during affirmative action. It's not really fair to compare the two."
In terms of the *possibility* of a relatively neutral scientific analysis, I agree that they are quite different. In terms of popular sources potentially having a bias which could inspire cherry picking of only that which supports a pre-set narrative and not considering off-narrative hypotheses, they can be very similar. It depend on what facet one is referring to.
I did not assert that there was 0 scientific support for "the assertion"; I strongly suspect there is some (and it's quite possible that you are correct that there is a "consensus" in favor of it, for all I know). I was just summarizing that, at my current level of research (including what had been referenced here), I have not yet encountered a scientific paper supporting it - reflecting my own lack of knowledge of the field.
Prior to a few years ago, I had not heard anything much about which groups were benefitting from AA, period. I had of course heard that Black folks, Latinos, American Indians, and women all benefitted, but I do not recall anybody making a point in any direction about "who gets more benefit". It just wasn't an issue. So I never heard "AA is benefitting Black people more than white women" in those days, nor the reverse - I just didn't hear any comparisons at all in general discourse.
Then a few years ago I began hearing frequently that white women benefitted more; that idea spread rapidly and persisted, and I passively assumed it was likely true (because I heard it frequently, it seems plausible, and I did not hear any counterpoint) until recently. I do not see that emergence as countering a then prevailing narrative that "Blacks benefit more than women"; it's more like the beginning of comparing benefits between competing identity groups.
---
I take your point about assumptions regarding AA hiring of women and Blacks (and Latinos, etc).
However, I think it depends on the field. You may be right about office workers, but among firefighters, police, combat soldiers, welders, masons, etc it's going to be the reverse - women are going to be more suspected of being an AA hire than are Black men. In the US, Black police officers are approximately proportionately represented, but female officers are not even close. So the degree of wariness has a lot of it has to do with how common vs unusual the new face is. (And some may be reality based, as it's not unreasonable to think that males more often can carry people out of burning building than females, statistically).
In my experience with workplace dynamics, the main concern I have observed is hiring of somebody who is less competent and who will not hold up their share of the job, making other's work harder (or less successful or more dangerous, etc). This would be true for the boss's nephew, or the manager's squeeze, but today AA is one of the possible causes for such wariness. But if over time the new person shows themselves as competent and easy to work with, they will be welcomed; if not, there is also the fear that they cannot be terminated without a mess. It's important to distinguish these basically rational fears, from irrational hatred or antipathy.
The ideal concept of AA is that when two candidates are equally competent, we should choose the one from the more marginalized group. If it had always worked out that way, AA would likely not have nearly as much pushback. Alas, in the real world this is not always the case; sometimes the difference being bridged to meet a goal is not insignificant. I have personally encountered frankly incompetent people hired under AA, so I know it exists. Alongside those are many AA hires who turn out to be competent and good co-workers, so I'm not saying that's typically the case, I'm not saying that the concept of AA is wrong, and I definitely abhor pre-judging anybody - but it does happen and pretending it doesn't, just alienated people.
All that acknowledged, it's unfair that somebody who is or might be an AA hire may face more caution and scrutiny than other newcomers, due to no fault of their own. For any AA category.
Some people buck up and show their mettle under that kind of pressure of feeling watched, but it can cause others to fail ("stereotype threat"). Humans vary.
I'm not sure how to intervene and reduce that dynamic. Well, eliminating AA would do it, but other than that. Heavy handed interventions, I believe, are likely to make the problem worse - increasing distrust and suspicion (of both management and the new hire), even if overt expression thereof is successfully suppressed. My tendency would be to emphasize the concept of judging people on the content of their character rather than their skin color (literally, or metaphorically in the case of non-race based AA). That is - judge each new coworker as an individual, give everybody the fair chance that you would want, treat them as you would wish to be treated if you were a new hire, allow them a fair chance to show their stuff. Perhaps urging a bit more slack, as they may feel under more pressure as described above.
Alas, this "treat everybody the same, judge their competence not their superficial traits" concept has been demonized among neo-progressives. The latter tends to strongly promote stereotyping and differential treatment - but only politically approved stereotypes and treatment. I do not think that produces the outcomes they claim to want (back to the criticism that neo-progressives often prescribe bad strategies even for real problems).
I strongly believe in not stereotyping (which for my purpose I define as assuming that some group trait - whether said trait has a high or low correlation with the group - can be meaningfully applied to every individual in the group). Some group stereotypes are plain out false, like if the trait is actually no more common in the group than average. Others may have some degree of truth to them, like the trait could actually be significantly more common in the group. But in either case, each individual can be typical or atypical of that population group, so I advocate starting with a blank slate of no pre-suppositions, and learning about that person as a unique individual rather than as an interchangeable unit of the group identity. By population group I mean Latinos, males, engineers, Catholics, whatever group may be associated with stereotypes, false or with some statistical level validity.
Do you have better ideas on dealing with wariness of new hires based on AA? Whether due to race or to sex or whatever.
"Some people buck up and show their mettle under that kind of pressure of feeling watched, but it can cause others to fail ("stereotype threat"). Humans vary."
Yeah, I think this is a key source of irritation amongst minorities and women. I absolutely resent the implication (not saying this is your implication, but it certainly exists in society generally) that we need to "buck up" when people make assumptions about us based on some immutable characteristic. That doesn't mean we won't or can't, it means we shouldn't need to. The frustration is about having to do this thing we shouldn't have to do, and which an entire class of people *never* has to do, in addition to the ordinary frustrations and mistreatments of life that we all deal with. It's like running a race where there's an extra hurdle only in your lane.
I won't fail because somebody believes that I'm less capable because I'm black. But that in no way diminishes the irritation that I should need to "prove myself" to the satisfaction of my white colleagues as Murray put it. And for the people who *will* fail under that pressure, it's absolutely wrong that they should face it for no reason other than some facet of how they were born.
And yeah, while eliminating AA would reduce this dynamic, it would also leave in place the dynamic that AA was intended to solve. All those people who would assume that POC and women are less capable would continue to make that assumption and so would hire/promote them less often. This, in turn, would leave the employment and opportunity disparities in place.
After all, the purpose of AA wasn't to hire POC who were less capable, instead, as you say, it was to give POC who *were* capable a fairer chance in the employment market. Maybe AA wasn't the best way to do this. Or wasn't implemented in the best way. There are smart people on both sides of that debate. But other than some kind of conscious intervention, I'm not sure how it would be possible to break the hegemony that was already in place. I don't think we'd see the degree of diversity we see today in the employment market without AA and diversity programs, for example.
I think the reason the assumption around "affirmative action hires" arose is that too many people really couldn't wrap their heads around the fact that pretty much every single president, CEO, movie star, etc. being a white man wasn't just a coincidence or a simple question of merit. They saw the challenge to the monopoly as preferential treatment for minorities and women, while failing to realise that people with their immutable characteristics had enjoyed preferential treatment since forever.
So I guess the best way to reduce the dynamic, though sadly by no means easy (and almost endlessly frustrating), is to encourage the people who hold this mindset to really examine the world from a perspective other than their own. And this, in turn, means learning to talk about racial and sex-based bias in ways that are honest and clear without being divisive or accusatory. There's a lot of talk about how white people get defensive in discussion about race. And it's true. But I think a fair amount of that defensiveness has been generated by the demonising way we talk about it.
And it's hard to wisely intervene without "undesired side effects".
I don't have the answer, either.
And yes, as I noted - in the ideal case, somebody chosen via AA is just as competent (or very close) as the person who was passed over. In the real world, that doesn't always happen tho. I was reading somebody whose boss was convinced there was an unlimited supply of qualified Black doctors, so they could focusing on hiring almost exclusively those doctors, until decades of past practice had been balanced out and the total workforce (of this institution) was proportionate to the population. The correspondent was saying that in practice, there just were not enough such physicians, so they either had to understaff (unacceptable), hire more Asians and whites than they wanted, or reduce the quality they expect. No magic pills to avoid those tradeoffs (other than waiting many years).
I don't like that situation, I'd much prefer that the boss's illusions were true. Sometimes maybe it is, and that's good. Other times it's a tight market for employers, and they can't afford to pass up too many good and available employees to find enough people who fulfill the diversity goals. Hopefully that will change over time.
One of the purest meritocracy examples I've heard of is the practice of doing blind auditions for major orchestras - the candidates play behind a cloth, so the reviewers have zero knowledge of their sex, race, ethnicity, sexual orientation, appearance, etc. This was demanded of them back in the 60's and 70's. However this process has consistently resulted in greatly disproportionate hiring of Asians, slightly disproportionate hiring of whites, and too few Latin and Black hirings - compared to the overall population. Of course, talented musicians in different groups may be attracted more or less to classical music for their expression. But to fix that, the major orchestras in the US have abandoned blind auditions, so they can deliberately choose certain groups more often and be socially accountable. The assertion is that top musicians are so close that they are functionally indistinguishable - but they have been being distinguished for many years without any prejudice involved at all, so the new paradigm is to deliberately distinguish so as to meet diversity goals. I am not that much into US orchestras, so it doesn't matter a lot to me, but it's objective evidence that Kendi is wrong in asserting that the ONLY explanation for different outcomes is racial discrimination. This practice may still be justified in the big picture, but not on the basis that it's *reducing* discrimination and bias, but rather on the basis that it is practicing "positive discrimination" (as the AA concept is sometimes referred to in England). That is, a major improvement in social justice (represented by the proxy of proportionate hiring) may arguably be more important than the minor hit to merit. However, I wonder if there is any dynamic here which will ensure that the hit to merit really is only minor.
"In my world there is an important middle ground: "uncertain" - as in I personally do not yet have enough information to draw a conclusion. That's where I am in regard to this assertion."
Yeah, maybe this is where we're tripping up. I hold almost all specific claims in the "uncertain" category. Certainly all claims in the social sciences. Because almost all information could be "disproved" or cast into doubt by new evidence or more careful analysis.
But once there's *enough* evidence to satisfy me (multiple sources, logical conclusions, zero claims of the inverse), they move to the "more likely than not to be true" end of that spectrum. As I said earlier, it's about how much time one is willing to spend digging into every piece of data on a single assertion in order to "prove" it.
If a single piece of research claims the opposite, I note it, but that single data point probably won't move me very much unless it presents something compelling. Especially if it appears to me to have fairly serious flaws (to be clear, this conversation has moved my confidence in the assertion back toward "uncertain", just not very much). Again, COVID is an excellent example of how futile it is to try to prove something beyond any doubt. And that's far harder science than this.
If I'm writing an article, I'll never base my argument on a single piece of research or a single data point. Precisely because if that research is cast into doubt, the entire thesis falls apart. So I'll cite numerous independent pieces of evidence, all pointing to the same conclusion, and say, "All of this together is why I believe X to be true."
My attachment to the "factuality" of a single piece of data therefore, is usually pretty loose. I'm convinced enough by "the assertion" that I'm happy to use it in an aside to a reader in a comment section. But I'd never use it as the cornerstone of an article. I'm not even using it as the cornerstone in my argument that Charles Murray is racially biased.
So yes, while I think it's wise (and sadly rare), that people keep information in the "uncertain" category, it's paralysing if we don't accept some information as "likely enough" to be true for a particular purpose. If everything you said had to be true beyond any possible shadow of disagreement, you'd never say anything. I mean, look at how many people (including many fellow experts in his field) claim the findings in Charles Murray's book aren't true.
So yeah, if there were definitive evidence that black people were the primary beneficiaries of affirmative action (or that white women and black people benefited more or less equally), it wouldn't change anything for me, because a) lots of people already assume that to be the case (all without *any* evidence) and b) none of my ideas rest on it being false. But some of Charles Murray's arguments about affirmative action *do* seem to rest on it being true.
And the wider point is that grievances about racism don't rest on things like this single assertion being true. Or, for example, on the Atlantic slave trade being the most brutal form of slavery. Neo-progressives aren't wrong to feel aggrieved, they're wrong in what they suggest is done about those grievances. And in implicating all white people as the source of those grievances. But there is soooo much indisputable evidence of the cruel, deliberate, and state-sanctioned oppression of black people in America, that these assertions being proven wrong beyond a shadow of a doubt wouldn't move the needle.
Good points, Steve. In particular, I agree with your final paragraph, and many of your points about evidence and "proof".
(Well, I might question some of the specific points or degrees of grievance used by neo-progressives, but I do agree that many of the grievances are valid and that their prescriptions are the bigger issue).
I'm glad that we agree that absolute proof is not available, and that neither of us expects that. That seemed like a red herring, so it's good to take it off the table.
As I've said, for years I had (without discomfort) put "the assertion" into the "likely true" category. But I had done zero real research (since it didn't seem to have big operational implications), I was just lazily assuming it was probably true, because I had heard it many times, and had never heard anybody dispute it. I didn't seek out information about it.
Then *relatively recently*, I heard somebody dispute it, so I did a cursory search and after filtering, the first peer reviewed paper I found came to a contrary conclusion, and also described the flaws in superficial analysis. I explicitly did not take that as definitive, but as enough to raise some questions of factuality, and as raising the right questions (unlike the popular press).
So for me, the currents status my knowledge about "the assertion":
* mainstream popular sources: near universal belief, cited by activists as another grievance; I find most activists to be deeply into confirmation bias and cherry picking, so this consensus is not convincing and I try to look for better sources
* scientific sources: 0 sources supporting it, 1 source questioning it, not nearly enough evidence to make any conclusion, but enough to consider it "needs more investigation". I do not know if this paper is part of a consensus, or a rare exception to a consensus, or if there is n scientific consensus.
If I relied entirely upon the sources which also consider it to be established "fact" that the George Floyd murder was obviously due to racism, then "the assertion" would be a no-brainer; we know that the popular press is very much in favor of the assertion, when it comes up. But is that because they neutrally assessed the evidence, or (once again) because it fits the same narratives which support their George Floyd "factual" conclusions?
I figured that if your research into the matter had established that there was indeed a scientific consensus supporting the assertion, then you could easily pass on a few links to the best of them, opportunistically bootstrapping my low-priority research. As I've said, I would have asked the same from somebody certain of the opposite conclusion. I don't have a dog in this race, other than seeking truth.
Science is not immune to ideological or other biases, but it typically does at least consider major confounding factors and avoid the worst abuses of process that the popular press often commits. Like assuming that 100% of the increases in female and POC employment is based entirely on AA, and then citing statistics which have meaning only if you accept that assumption as fact.
What I took away from the paper I cited was (1) they defined the issue more carefully, (2) they acknowledged the difficulty of determining factual truth, (3) they noted the major confounding factors which needed to be taken into account in any serious study, and (4) within their limited scope, their analysis actually disconfirmed "the assertion" with much more care than the popular press. Because it's only one paper of limited scope, #4 has little weight in isolation - but note that points 1-3, which are still very valid in understanding the deep conceptual flaws in many journalistic or activist "analyses". We should not think that #4 was the only thing of value in the paper.
TL;DR: The paper did not establish any "scientific consensus" by itself, but it did identify why non-scientific analyses may be deeply flawed, by noting things which absolutely *must* be accounted for in a serious study.
So far, you are asserting that science has come to consensus supporting the conclusion so that the only peer reviewed study I have seen *must* obviously be an outlier, which thus can be summarily dismissed as having little weight compared to that established broad scientific consensus.
However, you haven't provided even one citation of a (science) journal article you has used in forming your conclusion, much less any indication that your previous research had established that there actually *is* such a scientific consensus - based on your having reviewed many such scientific papers and found a consensus among them (with the paper I cited being the first and assumed only exception to come to your attention).
> "I'm convinced enough by "the assertion" that I'm happy to use it in an aside to a reader in a comment section. But I'd never use it as the cornerstone of an article."
Fair enough. Good practice.
> " if there were definitive evidence that black people were the primary beneficiaries of affirmative action (or that white women and black people benefited more or less equally), it wouldn't change anything for me, because a) lots of people already assume that to be the case "
Steve, are you saying that you both that (1) you know that lots of people assume that black people rather than white women are the primary beneficiaries of affirmative action, and (2) you know of zero claims of the inverse of the assertion? ("there's *enough* evidence to satisfy me (multiple sources, logical conclusions, zero claims of the inverse)"?
Oddly, I had never run into anybody asserting that POC were the primary beneficiaries of affirmative action until fairly recently, as I've described. I had ONLY run into "the assertion", which I have seen many, many times. And that asymmetry makes sense; "the assertion" is part of the narrative of ongoing pervasive oppression, but the opposite (that POC have benefitted more) does not lend itself to any common narratives. We won't find angry far right folks asserting that it's a known (and implicitly outrageous) fact that white women have received fewer benefits from AA than POC. This is not ideologically symmetric, the incentives for confirmation bias are not equal in both directions.
Obviously, we read different places and our experience will differ, but I do think that most of the mainstream media is on the side of assuming "the assertion" to be a known fact, as you believe, and while you may believe that "lots of people" assume otherwise, I suspect that lots *more* people assume the assertion true than false.
Let's be careful here not to mis-frame the question. That is, one cannot assume that the assertion must be factually true, UNTIL and UNLESS somebody can provide "definitive evidence" that the opposite is true; there has to be room for "the evidence known to me is mixed, not yet sufficient to form a factual conclusion". (Which happens to be my current estimate, pending more research).
One can question whether an assertion has accumulated enough solid evidence to be called a known "fact", without needing to counter-assert that enough evidence has accumulated to establish it's opposite as a known "fact". Sometimes your framing of this does not appear to acknowledge this.
And before I close, I will note that there is a lot of sloppiness in the popular press (including the one you cited) between using statistics about "women" versus statistics about just "white women". I noticed that immediately; did you? Reread the article and see how often they refer to each. In general, women of color can get grouped in on either side of the statistics, depending on what the journalist wants to "prove".
And I'm still wondering if you think there is anything wrong if white women (or women in general) actually had benefitted more from AA than POC (assuming that both needed it to overcome previous stereotypes). Since you are not generally trying to reinforce a grievance narrative, does it even matter to you, beyond curiosity and respect for truth? Do you have any preference for assertion to be shown true or false?
Good discussing this with you, Steve. If we can discuss things well, we can help refine each other's views, on the relatively few points where we disagree.
"If I relied entirely upon the sources which also consider it to be established "fact" that the George Floyd murder was obviously due to racism, then "the assertion" would be a no-brainer;"
Yeah, as demonstrated by my recent conversation about the Atlanta, an assertion of racist motivations in the popular press is worth next to nothing. But that's quite different to a statistical analysis of employment and income advancement during affirmative action. It's not really fair to compare the two.
Also, there is scientific "evidence" insofar as the assertion is based on scientific research. We could debate the quality of that research, but it's untrue that there is 0 evidence. There is research claiming that the assertion is true, there is a paper saying the assertion might not be true, but there are 0 papers claiming that the inverse of the assertion is true. Sorry, I thought there were links to the research in the TIME article I originally linked. This one has some (https://www.vox.com/2016/5/25/11682950/fisher-supreme-court-white-women-affirmative-action).
Interesting that you say you've never heard of anybody who questioned the assertion. I hadn't even heard it until about 5 years ago. Until then, all I'd ever heard is the stereotypes that are spun from the narrative that black people are the primary (or even in many people's minds the *only*) beneficiaries of AA. In his interview with Coleman, Charles Murray states this extremely clearly (https://youtu.be/wCJFr6zB2NM?t=457):
"The reality of aggressive affirmative action, which nobody wants to say, is (but everybody kinda knows), whites widely consider that any time there is a new black face in the office, that's an affirmative action hire. That's the default assumption."
This absolutely matches my experience in talking about the issue or hearing people talk about it. But I know of no *evidence* of the claim. Just the culturally accepted assumption that AA is a "black" thing.
No, I don't think there's anything particularly wrong if AA benefitted white women more than POC. No meaningful preference for the assertion to be true or false (I'd find it funnier if the racists who use AA as a cudgel for POC were proven wrong I suppose). but yes, the grievance aspect is non-existent as far as I'm concerned. I wouldn't feel aggrieved if the assertion were proven false. Women *do* deserve more opportunities in the workplace. But if a new female face pops up in the workplace, especially a new white female face, the assumption is not that she's an AA hire. This the issue.
> "But that's quite different to a statistical analysis of employment and income advancement during affirmative action. It's not really fair to compare the two."
In terms of the *possibility* of a relatively neutral scientific analysis, I agree that they are quite different. In terms of popular sources potentially having a bias which could inspire cherry picking of only that which supports a pre-set narrative and not considering off-narrative hypotheses, they can be very similar. It depend on what facet one is referring to.
I did not assert that there was 0 scientific support for "the assertion"; I strongly suspect there is some (and it's quite possible that you are correct that there is a "consensus" in favor of it, for all I know). I was just summarizing that, at my current level of research (including what had been referenced here), I have not yet encountered a scientific paper supporting it - reflecting my own lack of knowledge of the field.
Prior to a few years ago, I had not heard anything much about which groups were benefitting from AA, period. I had of course heard that Black folks, Latinos, American Indians, and women all benefitted, but I do not recall anybody making a point in any direction about "who gets more benefit". It just wasn't an issue. So I never heard "AA is benefitting Black people more than white women" in those days, nor the reverse - I just didn't hear any comparisons at all in general discourse.
Then a few years ago I began hearing frequently that white women benefitted more; that idea spread rapidly and persisted, and I passively assumed it was likely true (because I heard it frequently, it seems plausible, and I did not hear any counterpoint) until recently. I do not see that emergence as countering a then prevailing narrative that "Blacks benefit more than women"; it's more like the beginning of comparing benefits between competing identity groups.
---
I take your point about assumptions regarding AA hiring of women and Blacks (and Latinos, etc).
However, I think it depends on the field. You may be right about office workers, but among firefighters, police, combat soldiers, welders, masons, etc it's going to be the reverse - women are going to be more suspected of being an AA hire than are Black men. In the US, Black police officers are approximately proportionately represented, but female officers are not even close. So the degree of wariness has a lot of it has to do with how common vs unusual the new face is. (And some may be reality based, as it's not unreasonable to think that males more often can carry people out of burning building than females, statistically).
In my experience with workplace dynamics, the main concern I have observed is hiring of somebody who is less competent and who will not hold up their share of the job, making other's work harder (or less successful or more dangerous, etc). This would be true for the boss's nephew, or the manager's squeeze, but today AA is one of the possible causes for such wariness. But if over time the new person shows themselves as competent and easy to work with, they will be welcomed; if not, there is also the fear that they cannot be terminated without a mess. It's important to distinguish these basically rational fears, from irrational hatred or antipathy.
The ideal concept of AA is that when two candidates are equally competent, we should choose the one from the more marginalized group. If it had always worked out that way, AA would likely not have nearly as much pushback. Alas, in the real world this is not always the case; sometimes the difference being bridged to meet a goal is not insignificant. I have personally encountered frankly incompetent people hired under AA, so I know it exists. Alongside those are many AA hires who turn out to be competent and good co-workers, so I'm not saying that's typically the case, I'm not saying that the concept of AA is wrong, and I definitely abhor pre-judging anybody - but it does happen and pretending it doesn't, just alienated people.
All that acknowledged, it's unfair that somebody who is or might be an AA hire may face more caution and scrutiny than other newcomers, due to no fault of their own. For any AA category.
Some people buck up and show their mettle under that kind of pressure of feeling watched, but it can cause others to fail ("stereotype threat"). Humans vary.
I'm not sure how to intervene and reduce that dynamic. Well, eliminating AA would do it, but other than that. Heavy handed interventions, I believe, are likely to make the problem worse - increasing distrust and suspicion (of both management and the new hire), even if overt expression thereof is successfully suppressed. My tendency would be to emphasize the concept of judging people on the content of their character rather than their skin color (literally, or metaphorically in the case of non-race based AA). That is - judge each new coworker as an individual, give everybody the fair chance that you would want, treat them as you would wish to be treated if you were a new hire, allow them a fair chance to show their stuff. Perhaps urging a bit more slack, as they may feel under more pressure as described above.
Alas, this "treat everybody the same, judge their competence not their superficial traits" concept has been demonized among neo-progressives. The latter tends to strongly promote stereotyping and differential treatment - but only politically approved stereotypes and treatment. I do not think that produces the outcomes they claim to want (back to the criticism that neo-progressives often prescribe bad strategies even for real problems).
I strongly believe in not stereotyping (which for my purpose I define as assuming that some group trait - whether said trait has a high or low correlation with the group - can be meaningfully applied to every individual in the group). Some group stereotypes are plain out false, like if the trait is actually no more common in the group than average. Others may have some degree of truth to them, like the trait could actually be significantly more common in the group. But in either case, each individual can be typical or atypical of that population group, so I advocate starting with a blank slate of no pre-suppositions, and learning about that person as a unique individual rather than as an interchangeable unit of the group identity. By population group I mean Latinos, males, engineers, Catholics, whatever group may be associated with stereotypes, false or with some statistical level validity.
Do you have better ideas on dealing with wariness of new hires based on AA? Whether due to race or to sex or whatever.
"Some people buck up and show their mettle under that kind of pressure of feeling watched, but it can cause others to fail ("stereotype threat"). Humans vary."
Yeah, I think this is a key source of irritation amongst minorities and women. I absolutely resent the implication (not saying this is your implication, but it certainly exists in society generally) that we need to "buck up" when people make assumptions about us based on some immutable characteristic. That doesn't mean we won't or can't, it means we shouldn't need to. The frustration is about having to do this thing we shouldn't have to do, and which an entire class of people *never* has to do, in addition to the ordinary frustrations and mistreatments of life that we all deal with. It's like running a race where there's an extra hurdle only in your lane.
I won't fail because somebody believes that I'm less capable because I'm black. But that in no way diminishes the irritation that I should need to "prove myself" to the satisfaction of my white colleagues as Murray put it. And for the people who *will* fail under that pressure, it's absolutely wrong that they should face it for no reason other than some facet of how they were born.
And yeah, while eliminating AA would reduce this dynamic, it would also leave in place the dynamic that AA was intended to solve. All those people who would assume that POC and women are less capable would continue to make that assumption and so would hire/promote them less often. This, in turn, would leave the employment and opportunity disparities in place.
After all, the purpose of AA wasn't to hire POC who were less capable, instead, as you say, it was to give POC who *were* capable a fairer chance in the employment market. Maybe AA wasn't the best way to do this. Or wasn't implemented in the best way. There are smart people on both sides of that debate. But other than some kind of conscious intervention, I'm not sure how it would be possible to break the hegemony that was already in place. I don't think we'd see the degree of diversity we see today in the employment market without AA and diversity programs, for example.
I think the reason the assumption around "affirmative action hires" arose is that too many people really couldn't wrap their heads around the fact that pretty much every single president, CEO, movie star, etc. being a white man wasn't just a coincidence or a simple question of merit. They saw the challenge to the monopoly as preferential treatment for minorities and women, while failing to realise that people with their immutable characteristics had enjoyed preferential treatment since forever.
So I guess the best way to reduce the dynamic, though sadly by no means easy (and almost endlessly frustrating), is to encourage the people who hold this mindset to really examine the world from a perspective other than their own. And this, in turn, means learning to talk about racial and sex-based bias in ways that are honest and clear without being divisive or accusatory. There's a lot of talk about how white people get defensive in discussion about race. And it's true. But I think a fair amount of that defensiveness has been generated by the demonising way we talk about it.
Agreed. It's unfair.
And it's hard to wisely intervene without "undesired side effects".
I don't have the answer, either.
And yes, as I noted - in the ideal case, somebody chosen via AA is just as competent (or very close) as the person who was passed over. In the real world, that doesn't always happen tho. I was reading somebody whose boss was convinced there was an unlimited supply of qualified Black doctors, so they could focusing on hiring almost exclusively those doctors, until decades of past practice had been balanced out and the total workforce (of this institution) was proportionate to the population. The correspondent was saying that in practice, there just were not enough such physicians, so they either had to understaff (unacceptable), hire more Asians and whites than they wanted, or reduce the quality they expect. No magic pills to avoid those tradeoffs (other than waiting many years).
I don't like that situation, I'd much prefer that the boss's illusions were true. Sometimes maybe it is, and that's good. Other times it's a tight market for employers, and they can't afford to pass up too many good and available employees to find enough people who fulfill the diversity goals. Hopefully that will change over time.
One of the purest meritocracy examples I've heard of is the practice of doing blind auditions for major orchestras - the candidates play behind a cloth, so the reviewers have zero knowledge of their sex, race, ethnicity, sexual orientation, appearance, etc. This was demanded of them back in the 60's and 70's. However this process has consistently resulted in greatly disproportionate hiring of Asians, slightly disproportionate hiring of whites, and too few Latin and Black hirings - compared to the overall population. Of course, talented musicians in different groups may be attracted more or less to classical music for their expression. But to fix that, the major orchestras in the US have abandoned blind auditions, so they can deliberately choose certain groups more often and be socially accountable. The assertion is that top musicians are so close that they are functionally indistinguishable - but they have been being distinguished for many years without any prejudice involved at all, so the new paradigm is to deliberately distinguish so as to meet diversity goals. I am not that much into US orchestras, so it doesn't matter a lot to me, but it's objective evidence that Kendi is wrong in asserting that the ONLY explanation for different outcomes is racial discrimination. This practice may still be justified in the big picture, but not on the basis that it's *reducing* discrimination and bias, but rather on the basis that it is practicing "positive discrimination" (as the AA concept is sometimes referred to in England). That is, a major improvement in social justice (represented by the proxy of proportionate hiring) may arguably be more important than the minor hit to merit. However, I wonder if there is any dynamic here which will ensure that the hit to merit really is only minor.