My Last Word on TED

A Response to Chris Anderson and Adam Grant

Last week I wrote a piece for the Free Press (now reposted on my substack) about how TED sandbagged my talk on color-blindness. Chris Anderson, the head of TED, wrote a response on X, which I responded to promptly. But Chris wrote a follow-up, and Adam Grant has also written a public response, neither of which I have publicly responded to. 

A number of people have asked me if I plan to respond, so here I go.

Disclaimer: This is the last thing I will be writing about the TED debacle!

Chris:

From my perspective, it’s pointless to continue a public conversation with Chris without him acknowledging the core allegation of my piece and apologizing for it. 

Simply put, Chris pleaded with me to adopt a strange release strategy in order to “amplify” my talk––which I suspected was a bit of corporate spin-doctoring at the time. Then, TED did the opposite of amplify my talk: they deliberately under-promoted and sandbagged it on their website.

While Chris has called my version of events “reasonably accurate,” he has been careful to neither admit nor deny this core accusation. If he thinks I’m wrong, he should say so. But if he acknowledges that I’m right––which is all but conceded by his silence on the issue––then decency requires owning up to it and apologizing. Instead, he has chosen a middle path where he says a bunch of stuff without addressing the elephant in the room. 

Besides that, I have little to say about Chris’s public statements that has not already been said. See Jesse Singal’s substack post, for instance. I will just echo some of Singal’s points here. 

Chris seems to view this situation as a dispute between two equally reasonable parties––me on one end and his staff on the other. That is the wrong way to think about this situation.

Here are two key differences between me and his staff: 

(1) I believe that there should be all kinds of TED talks: woke ones, anti-woke ones, and apolitical ones. Free speech and viewpoint diversity should reign supreme! But TED’s staff appear to believe that there should be tons of woke talks and zero anti-woke talks. That’s a big difference. I want a bigger tent of allowable ideas, they want a smaller tent.

(2) I believe MLK’s prescription of race-blind, classed-based social policy––as he advocated for in his book Why We Can’t Wait (see point #5 in this post)––is both wise and within the bounds of acceptable opinion. The people on the other side of this appear to believe that anyone who advocates for MLK’s position should be de-platformed. Equally reasonable? 

Finally, Chris ought to reframe his view of his staff’s feelings. 

For instance, when Chris writes, “Some commenters below just don’t understand how anyone could be upset by a talk arguing for color blindness,” he is straw-manning in spectacular fashion. Sure, somewhere there is probably a person with Asperger’s who truly doesn’t get why someone could get touchy about race. But the vast majority of TED’s critics understand why people get emotional about race––we just don’t think that those emotions should have de facto veto power over what can be argued in a public forum. 

Last year, TED had a pro-communism talk (“socialism” is in the title, but “communism” is advocated for in the talk). I bet there was at least one member of the audience who grew up in the Soviet Union or Cuba and was fiercely triggered by this––given the ghastly toll of communism in those nations. 

But would their offense ever be taken as a reason to censor the talk? Of course not. And that’s the way it should be. The adults in the room should say “Sorry, but your feelings, while totally understandable, do not give you veto power.” End of story.

Life is triggering. Part of being an adult is learning to take responsibility for your feelings instead of insisting that it’s the world’s responsibility not to trigger you. We all understand this 99% of the time. Why woke institutions suddenly forget this when confronted with black anger (or Indigenous anger or  LGBTQ anger, etc.) is what TED’s critics, myself included, “just don’t get”.

Adam:

There has been much talk about a meta-analysis that Adam Grant sent to Chris, and Chris sent to me. Because of the way that this meta-analysis has been framed, I imagine that it is very easy to lose the forest for the trees here.

Big picture: the meta-analysis is really bad and in no way refutes my talk.  

Let’s return to square one.

In my talk, I argued that color-blindness is the right way to go, both in our personal lives and in our public policy. In other words, race should not be a “but-for” cause of your decisions about who to accept into your college, who to hire, who to befriend, who to marry, who to distribute aid to in a pandemic, and so forth. And too often, it is. That was the message of my talk.

My argument didn’t rely on social science––purposely so. For starters, there is a replication crisis in social science. By one estimate almost half of social science findings turn out to be BS, and those BS findings are far more likely to be cited than the real ones. What’s more, combining a bunch of BS studies into a “meta-analysis” doesn’t make them any better. (Kind of like CDOs, as explained by Anthony Bourdain in The Big Short). 

So instead of relying on a field of research that’s notoriously unreliable, I relied on other arguments: intellectual history, moral philosophy, and recent real-world case studies. Nobody at TED had a problem with this at the time.

Weeks after my talk, I received a social science meta-analysis, along with a snippet of a message from Adam Grant saying that this meta-analysis was an example of “rigorous research” that  “directly contradicted” my “inaccurate” talk: 

“Really glad to see TED offering viewpoint diversity—we need more conservative voices—but as a social scientist, was dismayed to see Coleman Hughes deliver an inaccurate message.

His case for color blindness is directly contradicted by an extensive body of rigorous research; for the state of the science, see Leslie, Bono, Kim & Beaver (2020, Journal of Applied Psychology). In a meta-analysis of 296 studies, they found that whereas color-conscious models reduce prejudice and discrimination, color-blind approaches often fail to help and sometimes backfire.”

At this point, Adam would certainly want me to remind you that this was not his full message to Chris. So this is that reminder. His full comments are here.

But frankly, I don’t see how that’s relevant since nothing else in his message contradicts this excerpted part. Even in its full context, Adam appears to believe that this meta-analysis is the “state of the science“, “directly contradicts” my talk, and shows my talk to be “inaccurate”. And Chris appears to believe that Adam’s “overall stance is indeed backed up by the paper.” As you will see shortly, these are both very unreasonable, retraction-worthy takes.

There are two levels at which this meta-analysis doesn’t refute my talk. First, as mentioned, you should rate meta-analyses of social psychology studies fairly low on the believability scale to begin with. “A grain of salt” does not begin to approach the attitude we should take towards such studies, given the replication crisis. It is the opposite of “rigorous research”. 

But let’s, for the sake of argument, assume that the field of social psychology were super-rigorous. What does this meta-analysis actually claim and do those claims challenge the ones I made in my talk?

The overall structure of the meta-analysis is fairly simple: it examines how four different ideologies–– “color-blindness”, “meritocracy,” “assimilation,” and “multiculturalism”––affect four different outcomes––”prejudice,” “discrimination,” “stereotyping,” and… “support for diversity policies” (like “affirmative action and permissive immigration”, the authors clarify). 

Yes, you read that right. In this “rigorous” meta-analysis, one of the ways that success is measured is whether the ideology leads to support for policies like affirmative action. Affirmative Action, in other words, is assumed to be good and treated like an effect rather than a cause. That strikes me as backwards. Shouldn’t the point be to study whether policies like racial affirmative action cause good or bad outcomes? That, certainly, was the point of my talk (or one point of it). 

The interesting question is not whether color-blindness leads to opposing race-based affirmative action (it does, almost by definition). The interesting question is whether policies like race-based affirmative action lead to bad outcomes like more discrimination. This meta-analysis does not even ask that question, much less answer it. Because of the strange way its terms are defined, it’s not designed to weigh in on the claims that I make in my talk.

Similarly, Grant hangs his argument on the paper’s claim that “multiculturalism is more consistently associated with improved intergroup relations than any identity-blind ideology.” In order for me to be right, he argues, I’d have to show the opposite: that the color-blind approach “has greater efficacy than a multicultural approach that acknowledges differences”. And I didn’t do that, so…game, set, match. Right?

Not exactly. It would depend on what this paper means by “multiculturalism”. They define “multiculturalism” as “acknowledging [cultural] differences by learning about, maintaining, or valuing them.” The problem for Grant is that I support this kind of multiculturalism and not a word of my talk suggests otherwise. I’m not against learning about and celebrating cultural differences. I’m against race essentialism and policies that discriminate on the basis of race. I’m not coming for your school assembly on Diwali or Ramadan or Chinese New Year––that stuff is great. I’m coming for your race-based affirmative action, race-based emergency aid, and so forth. (And again, this “rigorous” meta-analysis does not even ask what the effects of those kinds of policies are. A bit strange, no?)

Besides this conceptually-confused and irrelevant meta-analysis, Adam cites some other studies which are paywalled. I don’t want to spend more time on this, so I will not be reading or addressing those here. If Adam wants to hang his argument on those other studies instead, then perhaps I will look at them in the future. (You, dear reader, are welcome to dig into them if you want). But I hope I’ve shown that the efforts to debunk my talk with silly social science are, in general, deranged. This is what we would call working backwards from a conclusion.

In sum, let me say that I do not think either Chris or Adam has issued the kinds of retractions and/or apologies that are truly merited by this situation. But at this point I do not expect those to be forthcoming.