That may be so, if you are already familiar with the arguments made and their rebuttals. I speak as someone who is not, and who is on the outside looking in as it were. From my point of view, the responses both by the college and by posters on this forum seemed very dismissive without ever reaching the merits of the letter. WN is the only one who addressed the points of the author with counterpoints, which then helped me weigh up both sides of the debate for myself.

Some of the reaction is just because of the political context, which I think may get lost a bit for non-Americans?

I do think you have a point that universities as teaching institutions can and should look for opportunities to teach, although again a letter to faculty about internal politics isn't exactly the classroom.

Also, fwiw, if there are specific points you want me to try to address in more detail, I can try.

While I may not be attuned to the political context experientally, I would say that, having visited these boards daily over the last couple of months, I am probably better informed as to the current state of affairs in the US than the average non-American. However, that is not a particularly high bar to clear.

Your offer to respond to the points made in the letter is appreciated. There is one point that did strike me - below - any counterpoints are very welcome.

- Is the prevailing narrative that the justice system is white supremacist, or anti-black American? The author does seem, to me, to make some compelling arguments about other nonwhites who have suffered discrimination historically, yet who are not over-represented (or are, even, under-represented) in their interactions with the criminal justice system.

In the interests of full disclosure, I am a white guy who did a bit over a year in jail in Boston, where the ratio was about 90% nonwhite, so I do also have some personal experience (and perhaps, bias) on this subject.

Race is a social construct, it's not a binary, and it's not immutable. Irish immigrants faced horrible discrimination, but now they're white. Chinese immigrants faced horrible discrimination, but now, they're not white, but they're definitely on a rung above black people.

Fine, respond as though race is a social rather than a physical/biological/evolutionary construct, then. The question made no such distinction.

Edit: or was that the answer? Because I don't understand it, if it is.

Instead of trying to assess what is the prevailing narrative, I'm mostly just going to tell you what I think, based on the research I'm familiar with. Because everything is complicated enough without introducing another layer on top.

I'm not sure exactly what you mean by white supremacist here. If you are asking whether or not the CJS is consciously and intentionally white supremacist, a la Jim Crow, then I would say no, and I think most researchers would distinguish between the Jim Crow era and the present. There is a book called The New Jim Crow: Mass Incarceration in the Age of Colorblindness, which I didn't recommend because I haven't read it. The slightly more nuanced argument is that the outcomes the system produces create and reproduce racial inequality unjustly, just like Jim Crow did -- regardless of intent. This review is useful in making that distinction, e.g. (quoting the book):

"The Age of Colorblindness" is a reference to Bonilla-Silva's book, which I mentioned before, which further develops this idea that more subtle prejudices, and ideological commitments he calls "color-blindness," contribute to the reproduction of racial inequality even while most people now reject explicitly racist beliefs and find them repellent.

I think the most important part of this explanation, central to explaining the second part of your question, is the role of economic inequality. TeflonDawg stated this in a rather pithy way the other day, so I'm going to borrow it:

The difficulty is that the two are intertwined. For example, it's not just that blacks in the US have less wealth and income than whites, it's that they are far more likely to live in neighborhoods of concentrated poverty, because of residential segregation. Various research documents the harms that compound from growing up under those circumstances. Raj Chetty's recent work (here and here) is useful.

Segregation was brought about during the Jim Crow era, by law. Though we eventually made de jure segregation illegal, we have not made that much progress at desegregation, because it isn't legally required. The Color of Law, which I recommended earlier, documents this history very well. It also discusses the way that explicitly racist housing law, redlining policies, the GI bill, and other legal discriminatory practices around housing contributed over time to the wealth gap, because a lot of the wealth accumulation of middle class America post WWII is tied to appreciation of real estate. For a brief summary, see here.

Several consequences of this are important:

1) With regard to your questions about other non-white groups, I think most of the differences can be explained by aggregate differences in wealth and income. The author of the Berkeley letter mentions Asian Americans. There is no doubt that they have (and do) face discrimination. The internment camps of WWII are a particularly horrific example. But it didn't go on long enough to create the kind of economic inequality that exists between blacks and whites. Nor is there the same kind of residential segregation. The number of Asian Americans has grown from immigration, and Asian immigrants tend to be already better educated, and earn incomes higher than the median for the US:

2) Differences in economic status, and particularly differences in exposure to neighborhoods of concentrated poverty, explain a lot about differences in exposure to the criminal justice system. The relationship between poverty and crime isn't that simple, but again race is relevant, illustrated by problems caused by tension between police and community residents, which contribute to problems with crime.

But another problem is that the criminal justice system creates feedback loops which make it even more difficult to break out of poverty. Chetty's work documents this effect, via measures of intergenerational mobility. Qualitatively, the federal complaint against the Ferguson PD, from 2014, provides a useful illustration of how this can play out. I think this note from the introduction, in particular, is important:

Basically, and this isn't unique to Ferguson (though Ferguson was particularly egregious), the police treated these kinds of neighborhoods, which are largely inhabited by blacks, as a source of revenue, engaging in all sorts of discriminatory practices:

I don't think those practices were mostly motivated by racial animus. They were mostly motivated by institutional self-interest, as noted above. But the net effect is like a kind of regressive tax. People who are already poor end up unfairly paying lots of fines. People end up with criminal records which make finding employment more difficult. All of this compounds with just normal issues caused by poverty, lack of education, and other social problems. All of this is inextricably tied to segregation, and acts to compound racial inequalities which were originally created by racism. When that's not enough, there are also still some issues of lingering discrimination, some of which I documented earlier. I believe The New Jim Crow focuses around racism in drug policy, as another example.

This is a long post, but hopefully it's useful at least to begin to answer your question. I think the essential conclusion is that you can't understand these issues without understanding how the status quo was created by racism. You can't really understand why we have never made much progress at fixing them without understanding how more contemporary racial ideology ("colorblindness") explains our unwillingness to examine that fact. Which is basically the reason why people like the letter writer insist on looking for other explanations, no matter how weak.

I love this post.

I have one point of contention:


The problem with that book is, it ignores a lot of what you pointed out. It ignores the cultural/social reasons people choose the drugs they choose, and it also ignores what happens when the use of one of these drugs reach epidemic proportions.

It's a fact the war on drug severely hurt the black community, during the crack epidemic, as it's hurting the white community during the opioid/meth epidemic.

I think at least part of the story with the some of those other minority groups not being as targeted by the cops(thinking long-term/historically as well) is that they are 'so foreign'(Asian/Hasidic etc) as to be nearly impenetrable by the cops It's not that there's zero crime--there are language barriers, some are very insular etc. As well as very obvious barriers--a white cop even undercover is going to stick out like a sore thumb in their communities. How many Asian cops have you seen in your life? We've added more of those kinds of minorities to police forces over the years but there still aren't very many. Aside from certain parts of the country those groups are also very tiny comparatively. Christ, look how long it took to really crack into the Italian mob. The other minority org crime groups from those eras(Irish/Jewish etc) were able to assimilate somewhat easier--and move away from 'street' crime by owning legit bizs to shield them somewhat so they own a bar or whatever and are a bookie on the side and largely get left alone--but there are still plenty of outlaws from those groups today. Now there are Russians and Armenians etc that are similarly tough to penetrate for some of those same reasons.

The entire story for black people in our country has been very different than those groups. Mexicans/Latino people are kind of a separate category as well story-wise...but a trip to your local courthouse will surely show they're getting plenty of attention from the cops.

This is something Jason Whitlock talks about a lot, which is also why I think so much focus on police on the streets is not going to change very much, and the cynical part of me wonders if that is the actual goal of Neo liberal progressivism (to appear to be doing something, but not really).

I understand officers on the streets are visible and easy targets, as they are the ones enforcing the (allegedly) discriminatory policies. But at the end of the day they are just following the orders their employers are giving them. As a matter of policy they are constantly putting themselves in high leverage dangerous situations (for themselves and the public), for the benefit of the state. It is true that some officers probably enjoy this a little more than they should, but that isn't the main problem IMO.

I find it extremely cynical how generally city halls are getting woke and now turning on their employees (and deflecting as much blame as possible off of themselves) who have been acting as defacto collection agencies on their orders.

Lumping Asians all together borders on ridiculous. Seems crazy to group Filipinos, Chinese, and Vietnamese, for example. It's basically racist thinking given how divergent these cultures are.

The internment camps of wwii had lsmall scale long term economic ramifications because the Japanese American population was small in absolute numbers.

The US black population is very special because it is not an immigrant population. It is a subculture with deeply American roots. Immigrant groups have a lot sociological advantages relative to what amounts to a historically oppressed indigenous (on the relevant time scales) group. The most comparable population I can think of elsewhere are roma in Europe (who are, however, a far smaller subpopulation). The only comparable US populations are Indian tribes and Puerto Ricans.

This is also true, it's not a great oversimplification. The source I linked does go into more granular detail. It just happens to be the case that the point I was trying to illustrate generalizes well enough and I was trying to simplify where I could.

WN - your thorough response is very much appreciated, thank you. A lot to take in there, going through it all now.

From the Ferguson case WN linked above (https://www.justice.gov/opa/file/823486/download [emphasis mine]):

WTF? They fined on average every person over once a year? Am I reading these numbers right?

ETA - this **** gets better and better:

I don't see a mistake in your reading.

FWIW, I haven't read most of them, but you can find links to documents related to other DOJ Office of Civil Rights interventions with the police here.

WN - I do find persuasive the argument that economic inequality plays a large part in explaining disproportionate representation of black Americans within the criminal justice system. Is it possible to control for this in the analysis of the statistics? I envision something along the lines of the following:

- Define some quantifiable measure, let's call it poverty index (PI). Each person in our data set is assigned a PI of e.g. 1-10, based on some agreed upon criteria.
- Assume the following hypothesis is true: there is a direct relationship between PI and "involvement in the CJS", however we currently quantify that.
- Assign a probability score to each PI for a random person with that PI being involved with the CJS, regardless of race.
- Disaggregate the data set by race to see if there are still racial disparities within each PI value.

Statistics is not my strong point, so this suggested methodology might be bunk, but something similar would seem like a sensible way to verify or refute the hypothesis. Are you aware of any studies which have done something along these lines? Apologies if you've already posted about this before, just direct me to your previous posts if this is the case.

There are some problems with data collection to begin with, I think. A lot these kinds of analyses rely on data collected by governments or other large public surveys, and my impression is that the easily available data is not as rich as one would like. There are a lot of methodological problems, including difficulties inferring causal direction from correlations.

But the reason I included the note that poverty and crime are not very simply associated is that I believe that's the conclusion of the available research. I don't know an R² value off the top of my head, but I think it is fairly low using your methodology. If all you know about a person is that they are poor, that's not by itself a very useful predictor that the person will be arrested, charged with a crime, spend time in jail, or etc. You can intuit this just by thinking about differences in crime rates across different countries with very different levels of absolute poverty. There's no very strong association.

I only have a pretty limited understanding of macro criminology theory, but here's one meta-analysis, from 2005: Assessing Macro-Level Predictors and Theories of Crime: A Meta-Analysis (full text):

Starting on page 389 (17 of the PDF; "Predictor Domains") you can find summaries of all the macro-level theories investigated by those studies. Some of their are pretty complex, involving much more nuanced ideas that are also more complex to measure, e.g.

Here's a summary of the results:

So poverty, by itself, is better than a lot of other predictors, but it's not the most important.

To be clear, there's a lot to unpack in this chart. For example, how do you measure "strength of noneconomic institutions?" What is "collective efficacy?" It should also be apparent that these are not all independent of each other. It's this interdependence that is often summarized as "concentrated poverty". This meta-analysis talks about "concentrated disadvantage," by which they basically mean the same thing:

You will notice of course the entries for "percent nonwhite" and "percent black." They also show up (for example) in Chetty's work. The point about the interrelation of factors and "concentrated disadvantage" is basically this: what the data are revealing is that the non-race factors which contribute the most to crime tend to be concentrated in areas where black Americans live. That is why race becomes a meaningful predictor. It is statistically associated with unemployment, incarceration, family disruption, and weaker institutions. This is also why history is so important, and why I think neighborhood-level analyses are often more useful than the individual-level ones. How can you begin to understand those associations without taking into account segregation and discrimination?

I also thought this chart would be a useful way to explain Wookie's comment about the social construction of race. The thing about statistical association is that it's difficult to infer causality, right? The crux of the Berkeley letter's argument is that the association between race, crime, and police killings is grounded in differences in behavior between people in different racial groups. This is couched in the language of individual agency, but the implication of all the cross-group comparisons is supposed to be: "there must be something wrong in black communities, or with black people, specifically." It's the latter implication that tends to be inescapably racist. While there's no doubt that problems exist within black communities which are not solely the result of direct discrimination, the persistent effect of racial prejudices in the US is that people view those problems in a way that essentializes race: that somehow these are not just problems faced disproportionately by black Americans, but that they are caused by racial differences.

And the difficulty with statistical associations is that they don't easily resolve that. White supremacists are happy to cite the associations noted in this study. And they are happy to infer racist conclusions from them. The point that Wookie is making is that those conclusions rest on conceptual assumptions about the "essential" character of race that are false. They require the assumption that our racial categories reflect underlying biological differences in a way that is meaningful to the question. That is not true. The enormous problems with racial categories invalidates the inference. This is another reason why history matters, alongside quantitative analysis. The categories we bring to our analyses are presuppositions. History shows why they are problematic and culturally contingent. These are the kinds of considerations that should lead, abductively, away from racist interpretations of the data.

I am either not tracking our argument very well, or I don't agree with it. I am not sure. It is very reasonable for someone to look at racial statistics and make inferences and presuppositions, knowing full well that race is a social construct.

Wookie (and every other progressive) himself does it all the time when it is convenient. The very idea that having more black CEOs (just to give one example) would be a benefit to the "black" group as a whole rests upon such premises. I don't even disagree with this presupposition, but it clearly is essentializing race.

Maybe it is correct analysis and maybe it isn't, but I think it is entirely reasonable to make an evidence based argument, "black people are disproportionately affected by the criminal justice system because they commit more crime, and Asians are not because they don't" while at the same time realizing race is a social construct, there is no biological determinism, and not having "racism" as a primary motivation factor in your analysis.

Also, I think you give white supremacists WAY too much agency assuming they are driving the conversation, or motivating the conversation. Whether one should refrain from making an argument in good faith because they know others will distort their argument in bad faith towards their own ends is one matter. But making the blanket statement that you can't make an argument because it is something white supremacists agree with and hence racist, I don't agree with.

The point of the paragraphs you quoted is about the relevance of the social construction of race to "scientific racism" style arguments. That is, arguments where the inference is that the differences are explained by by supposedly inherent, biological factors. I don't mean that no inferences at all are possible. Quite obviously, my posts have drawn a number of inferences from the data.

I would say that it is just the nature of social science that causal inference is really hard, and methodological problems with categories is not a small part of that. But I mentioned abduction because I think that's the best method for social science, and we certainly can apply it despite the tentativeness of our causal inferences. This is something I was trying to get at back in the playground.

I don't even think they need to drive the conversation. I think, if you're at all familiar with the history, that you can't help but notice the pedigree of arguments like "what about the Japanese, or Jewish Americans." I think you also have to distinguish between the conversation among elites and popular understanding. Clearly scientific racism doesn't get much as much traction in elite circles (though Charles Murray published another book this year). But my experience is that essentialist ideas about race are still very, very common. How many times has someone tried to argue for biological notions of race, relevant to socio-political questions, based in observations of athletics? I've probably observed this hundreds of times, just on my own, just across a few pretty narrow corners of the internet. So, sure, avowed white supremacists can't get much air time, but 19th century ideas about race itself are still pretty ubiquitous.

I am guessing I shouldn't be expecting any book reviews of this work from you anytime soon?

I was planning on reading it when it came out but I never quite got around to it.

Ok but then how much weight do you put into systemic factors that play into those numbers?

This is why use of "systemic" is 100% bullshit. There is no legitimate answer to that question. People claim "systemic" issues because of disproporation results. But how do you get those results? It's nothing more than a collection of individual events, for instance criminal sentencing. If there were systemic issues, every single black person convicted of a crime, would get a harsher sentences than a white in similar circumstances. That does not happen.


How you tackle this problem of "systemic" is to isolate and correct the individual instances. One thing you can do is monitor individuals judges to see if their sentencing is harsher for one race, over another. However, this really does not occur much, because the entire focus is on changing the "system". You can change every law to a progressives wet dream, and if there are prejudiced people within the system, you will still get racial bias in the results.

People are prejudiced, not systems. People commit crimes, not systems. Lastly, it's reasonable to have a culture racism, but it's racist to have a culture of violence? That's the main flaw in Well Named argument. All these cultural factors that make white people "racist' or prejudices" are reasonable. All the cultural factors that lead to more crime in the black community is considered racist.

You could have a conservative like me who agrees there is racial bias among white people (which is a cultural issues), generally, but you will hardly get a progressive to acknowledge cultural issues within the black community, and they will attack that claim as racist, almost 100% of the time.

This take is on FIRE. If a system of 100 judges has 99 racist judges who give black people worse sentences and 1 fair judge, the fact that any black person (you said it's only systemic if every black person is treated unfairly) gets a fair sentence disproves systemic injustice? There's no systemic problem in that system that allowed 99 racist judges to get seated in the first place?

If anyone is interested in the study WN linked above (Assessing Macro-Level Predictors and Theories of Crime: A Meta-Analysis), it is behind a paywall in his link, but I found a copy you can read with a free account here: https://www.jstor.org/stable/3488363?seq=1

There was a link to the text in my post: https://wellnamed.s3.amazonaws.com/pratt2005.pdf