Many of the comments on my post yesterday, “Bernie Sanders, Minimum Wage, and Systemic Racism,” March 1, were particularly good. They have convinced me that I need to walk back some of the things I said. Like most people, I hate to admit that I’m wrong. Unlike most people, I’m typically quite quick to admit that I’m wrong. Back in the 1970s or 1980s, my friend Michael Walker, who founded Canada’s Fraser Institute, asked Friedrich Hayek why people don’t seem to be convinced by evidence or logic. Hayek replied (I can just picture him doing this with his characteristic half wince/half grin) that people’s ideas are one of their most treasured forms of private property and when you convince them that they’re wrong, they suffer a capital loss. That fit my experience. But my next thought on hearing it was that you already had the capital loss; you just, to use tax lingo, realized it. And to push the metaphor maybe too far, you get a tax break on the loss. Translation: you’re better off admitting you’re wrong because then you won’t make that mistake again and you’ll get clarity for the future.

Now to the issue at hand.

I’ll focus on 8 comments.

BC writes:

Then, is “systemic racism” used synonymously with “unequal outcomes”?  If not, then what would be an example of unequal outcomes that are unfavorable for African Americans that does not reflect systemic racism?

Good question, BC, to which I don’t have a good answer. It’s the first comment that got me wondering whether there is such a thing as systematic racism, at least as defined by the NAACP’s President Derrick Johnson.

John Hall, after raising the issue of SAT scores, writes:

I think David needs to think on this issue a little harder to distinguish between this case and the minimum wage case.

I think John is right.

Vivian Darkbloom, as is Vivian’s wont, has one of the best comments, writing:

I’m opposed to a minimum wage increase to $15 for all the reasons likely discussed here previously.

Nevertheless, I’m wondering how someone in favor of such a hike would respond to this definition (and the conclusion therefrom that the hike constitutes “systemic racism”):

“[NAACP President Derrick] Johnson defined systemic racism, also called structural racism or institutional racism, as ““systems and structures that have procedures or processes that disadvantages [sic] African Americans.””

It’s highly likely that a disproportionate number of African Americans would either lose their jobs or fail to get one (or work fewer hours) as a result of such a minimum wage hike.  However, is it possible that a disproportionate number of African Americans (those that keep their jobs or manage to get one) would benefit from a wage raise and enjoy a slightly higher standard of living?

What I’m getting at here is whether, per the proposed definition, one should look  solely at those disadvantaged, or  should one weigh the *net* effect on those all those affected by the change?  Could Bernie Sanders reply “I acknowledge that there will be some loss of employment and some reduced employment as a result of this hike.  But, those negative effects are exceeded by the benefits of those who will get a pay raise”.

The same thought goes for those “disproportionately arrested” for whatever reason.  Is it possible that African Americans “disproportionately benefit” from more policing in those communities?

I don’t have a ready answer, but on the side of those opposed, one tends to refer only to those who tend to lose, and those in favor only to those who tend to gain, with neither side seriously attempting a net benefit analysis.  I think it is rare with respect to policy changes that there are only winners *or* losers rather than winners *and* losers.

I agree with Vivian that one should look at all the effects and that, in doing so, one might find that the gains to the black gainers outweigh the losses to the black teenage losers. I doubt it, because a high percent of low-wage people are employed in industries that produce goods that are sold to other low-wage people. Think of McDonald’s in Alabama, for example. But Vivian’s point remains.

Also, Vivian’s point about the disproportionately arrested black people is also relevant. Heather Mac Donald often makes the point that cops in low-income neighborhoods are valued highly by residents, including black residents, because they protect them from crime by other people and that the criminals are disproportionately black. So one would want to look at the total effect.

That last point is interesting because it was precisely the “disproportionately arrested” point that my friend made who convinced me that there is systemic racism. I called my friend to talk it out and actually got him to doubt his own claim, based on Vivian’s reasoning.

zeke5123 writes:

I think this definition [of systemic racism] is dangerous. I am reminded of an example by Kendi that lowering capital gains tax is racist, because it disproportionately favors non-blacks since blacks are less likely to have wealth.

The counter to this is if lower capital gains rates lead to net efficiency, it is a sound policy regardless of the impact on certain cross sections of the populace. Indeed, if you implement enough net efficiency policies, it is very likely that you end up lifting up all cross sections. But this is basically just saying: implement policies that are Kaldor-Hicks improvements.

Once when, with a single policy, you bring into the discussion how it impacts certain cross-sections of the populace (i.e., in this case race), you are more likely not to implement Kaldor-Hicks efficiency polices because racism is bad. No — that way lies madness. Focus on Kaldor-Hicks efficiency generally* and the rest will follow.

I think in your examples you believe changing those policies is Kaldor-Hicks efficient — that is, you think on net minimum wage is bad. But if you argue we should eliminate minimum wage (or not raise it) to reduce structural racism, then why can’t Kendi argue we should raise capital gains to reduce structural racism? I get that people can walk and chew bubblegum at the same time, but ultimately I think your case against the minimum wage is that it is bad economics; not that it harms black people because that would be your argument against raising capital gains tax rate (assuming you agree doing so would be bad).

*I would worry about situations where a policy causes massive losses to Group A to benefit Group B by the losses plus epsilon. There are mismeasurement errors, political risks, etc. So, I would limit Kaldor-Hicks efficiency where there is not (i) massive losses to a particular group and (ii) the gains clearly outweigh the losses. For me, a good rule of thumb is a strong presumption of liberty.

I pretty much agree with what zeke said, including that it’s a dangerous definition and that there’s a strong presumption of liberty. My guess is that my presumption of liberty is even stronger than his, but that’s just a guess.

Knut P. Heen writes of Derrick Johnson’s definition of racism, which I adopted but am now furiously backpedaling on:

A definition of systemic racism which excludes all races but one. Great. I thought this was the definition of racism.


Andrew_FL writes:

Minimum wage laws have disparate impacts on many groups, but no one would say the minimum wage is part of “systemic ageism” or “systemic credentialism” even in the rare case of someone who would both correctly understand the consequences of such laws and be inclined to agree with this definition of “systemic racism”

The whole notion is more about proving that we are all guilty collectively of a crime none of us can be said to be guilty of individually. Of course, if committing the category error of attributing racialist ideology and malice to non-human things like “systems” leads to a vast deregulatory push in housing and labor markets, that’s great. If it leads to a mass redistribution of wealth from the supposed perpetrators to their supposed victims, not so much.

The first paragraph is particularly on target. Andrew’s second paragraph points out the perils of using bad arguments for good policies: they can also be used for bad policies. Better just not to use bad arguments, period, even when they help you, which is getting to where I’ll get to soon.

Tom Nagle writes:

While I concur with all your facts, your concluding logic seems flawed when you assert that “Bernie Sanders, therefore, advocates systemic racism”.

Think in terms of a Venn diagram. Circle A is defined to encompass all policies that embody systemic racism as defined in your post. Overlapping a small part of Circle A is another Circle, label it B, that encompasses all policies designed or intended to increase the share of national income earned by low wage workers. Bernie Sanders is a socialist who likely supports all the policies in Circle B. But is it fair to claim that Sanders “advocates systemic racism”–that is, the policies in circle A–even if the overlap is small, the consequences unintended, and he clearly does not support an overwhelming share of the policies that could be classified as systemically racist. Would it be fair to state that “David Henderson advocates the suppression of women’s rights in Afghanistan” because that is an unintended consequence of pulling the US military out of there, a policy that you support?

Good point. Tom, by the way, is the person who convinced me, with the police example, that there is systemic racism. As I noted above, in response to Vivian Darkbloom, he is starting to wonder about his own belief.

Finally, BW writes:

Here is a relevant Slate Star Codex post.

I read it. As is often true, that post is very good. Damn the New York Times.

So where am I? I’m back to doubting that we have systemic racism. And this might answer the first questioner, Dylan. He asked what I had thought systemic racism was before Tom Nagle convinced me that it existed. I had in mind things like slavery, compulsory segregation of restaurants, government requirements that people on buses and streetcars be segregated, etc.

Moreover, I doubt that the concept is useful. Let’s say that you convince me that Bernie Sanders does not believe in systemic racism. And, by the way, the commenters have convinced me. Does that make me dislike his proposal for a $15 minimum wage any less? No. Does it mean that the $15 minimum wage would not cause as much harm as I think it would? No.

I hereby resign as an expert on systemic racism.