Last month I began a three-part series with A Bayesian Analysis of Susannah Rees’s Ishtar-in-the-Manosphere Thesis and A Bayesian Analysis of Kate Loveman’s Pepys Diary Thesis. Today I will conclude with the third random selection from my test set: LaDale C. Winling and Todd M. Michney, “The Roots of Redlining: Academic, Governmental, and Professional Networks in the Making of the New Deal Lending Regime,” The Journal of American History 108.1 (June 2021): 42–69. Yes, another random selection on a social justice focus, this time not from feminist studies but the field of Critical Race Theory. Yes. The most terrifying bugbear for conservatives today, the Notorious C.R.T.! And yet, another fantastic article.

Indeed, if you want to see what a serious and solid academic paper on CRT looks like, this is a paradigmatic example. On what my project is this month, what I’m aiming for here and how I settled on these selections, see my first article. But the gist of it is: I’m describing some real, peer-reviewed papers in history and showing how “under the hood” as-it-were their entire argument is Bayesian, and then illuminating what this can tell us about how to evaluate (or even challenge) claims about history, and why that’s important. My selections were entirely random. So that they all landed in social justice topics is interesting, but expected for reasons I outlined in that first entry: this holds the vanguard of popularity in history now, due to centuries of neglect and rising pressure from (usually conservative) naysayers to prove the claims underlying social justice claims. In effect, we need to know with rigorous certainty what really happened in history, if we are to ground our politics in factual reality and better understand how to dismantle the norms and institutions our national past has saddled us with. And finally, I can make the observation now that by random selection, producing a total number of authors of four, half were women, and half men. None were evidently people of color, although by random selection that’s not so improbable as to surprise (roughly a third of historians are non-White; so the probability of selecting four at random and their being all White is therefore 1 in 5, which is better than rolling a 1 on a single six-sided die, a not-uncommon outcome). But now to the present test case.

The Journal of American History is a major journal in its field, with a decent impact factor for a history journal (and a proper ISSN), and though it isn’t indexed by the DOAJ, it is published by Oxford University Press (which is again about as prestigious as you can get) and it ranks in the top quartile of all journals worldwide. LaDale C. Winling is a professor of history at Virginia Tech with an extensive peer reviewed publication record, particularly in the subject of redlining (which is the subject of this paper and which, if you don’t already know what that is, I’ll explain shortly), and his book, Building the Ivory Tower: Universities and Metropolitan Development in the Twentieth Century (University of Pennsylvania Press 2017), is on the directly-related subject of real-estate history. You really can’t find a more qualified expert here. Todd M. Michney is a professor of history and sociology at the Georgia Institute of Technology who also has an excellent publication record, particularly in urban history, racism studies, and redlining in particular. So we have two top-notch authors collaborating on this article.

The Thesis

As with the other two articles in my series, this is another treasure trove of diverse assertions and revelations, whose footnotes and bibliography alone are priceless. But I will here focus on the central assertion of their paper: that “the system of racial segregation and housing discrimination” that economist Richard T. Ely “and his allies set in motion” in the 1930s “did indeed proliferate over the succeeding decades and even to our present day” and have still yet to be fully dismantled. What they are referring to is commonly called redlining, a historical practice in the United States so significant it has its own Wikipedia page, which is well-sourced and thorough and well worth reading—especially its coverage of how it’s still happening. Even as certain racist practices become outlawed, institutions find loopholes to work around the law (e.g., the creation of food deserts, moving businesses out of minority-majority neighborhoods, predatory interest rates), or even simply just don’t follow the law. And even past practices that were ceased are still continuing their racist effects. There are countless scientific and historical studies on this, so thoroughly and conclusively proving these points that one really has to be devoted to ignoring evidence to deny systemic racism exists in the real estate sector (on the general distinction between personal and systemic racism, see my article Actually, Fryer Proved Systemic Racism in American Policing).

The Winling-Michney thesis argued in this paper contributes to that body of knowledge by addressing one outstanding dispute: which is not whether redlining happened, still happens, and still is having racist effects even when ceased (facts no qualified experts today dispute), but who is primarily responsible for it, whether private industry or the government. They find that at its start these two centers of power were so thoroughly intertwined in cooperation toward this common end that no such distinction can even be made: both were fully culpable. And central to this having happened, they find, is the personage of Richard T. Ely, an economics professor with influential roles in several think-tanks and lobbying groups, who developed the ideology that would be used to justify systemic racism in real estate policy, and helped the government and private sector implement it. The details are extensive and complex, outlined and referenced in the paper (another example of what I have remarked on before as the tremendous advantage modern history has over ancient history in the preservation of records and evidence).

The Case

It’s already well known that, as Winling and Michney put it:

A particularly notorious tool for promoting segregation has been redlining—the targeted denial of home mortgages and other financial supports according to presumed risk. Institutionalized by the Home Owners’ Loan Corporation (HOLC) and adopted by the Federal Housing Administration (FHA), historical redlining defined the riskiness of investing in urban neighborhoods based on demographics, infrastructure, housing quality, neighborhood stability, and proximity to amenities or hazards such as manufacturing. HOLC’s assessment of urban neighborhoods in the 1930s was racially and ethnically discriminatory. It used the power of the federal government to formalize patterns of segregation and discrimination.

They cite the previous scholarship already establishing these facts (in which the evidence is overwhelming, with likelihood ratios indisputably above millions to one favoring the conclusion). This does mean “discriminatory mind-sets translated into unequal access to state subsidies underlying suburbanization.” In other words, the Federal system literally funded white supremacism, giving White families enormous advantages in generational wealth that still produce observable disparities today. They responsibly note exceptions and nuances to this picture, but any reader can tell those qualifiers don’t affect the fundamentals of the conclusion. At any rate, this isn’t what Winling and Michney set out to prove; they instead cite a vast body of prior research already establishing it (pp. 42-44), and are merely summarizing the State of the Field, as historians would put it.

The authors then explain that their “article reframes the debate by examining the intellectual origins of real estate economics, how the resulting theories were channeled into collaborations with real estate
professionals, and how the collaborators implemented them as federal policy, thereby restructuring the private real estate sector” (emphasis mine). These are the three points Winling and Michney aim to prove and thus add to our database of knowledge on this story. And central to their narrative is the claim that “in the work of the Institute for Research in Land Economics, founded by Richard T. Ely,” the rest of whose backstory they document (pp. 44-46), they can “identify robust networks among academia, private industry, and government that developed, implemented, and spread these ideas,” such that although “Federal policy makers considered more than just race, yet racism factored into the equation from the very beginning of the Great Migration.” And if you don’t know what the “Great Migration” is, the National Archives will get you up to speed. Which is important to do. Since Republican opposition to teaching CRT probably will, and in many cases already has, nixed most Black history from your education. One key way to support white supremacism is to erase the history of it and then deny it existed, a denial an uneducated public might then fall far, thanks to their education being censored.

Crucial to testing all this will be weighing alternative explanations of the evidence against the Winling-Michney thesis: where does background knowledge settle the priors, and where do the likelihoods lean, especially compared to the most plausible alternatives, whatever they may be. All the undergirding points (like Ely’s career and thoughts, or the status of real estate economics before he transformed it, and after) they document thoroughly with citations of high quality evidence (which means beyond millions to one likelihoods even by a fortiori estimates; see my book Proving History, index, “canon of probabilities” and “a fortiori, method of”). And the prior probability of their thesis is well bolstered by their survey of the context of race relations and segregation efforts in which Ely developed and realized his ideas, which demonstrate that the latter exactly fit the trends of actions and thinking of the time, both in the private and the public sector (e.g. pp. 46-47).

So the question comes down to their evidence for Ely’s ideas becoming the key influence merging private and government efforts to segregate real estate by race. They start by showing how the National Association of Real Estate Boards (NAREB, the central clearinghouse for private-sector real estate policy and culture) courted Ely and worked with him in building out ways to implement his ideas for valuing and developing real estate (pp. 47-48). For this they have access to bounteous evidence, from conference minutes and agendas to correspondence and contracts; as well as books, articles, and leading textbooks composed under the direction of Ely and his students. The probability these things would exist and contain the statements they do on any other explanation is well beyond millions to one. The authors likewise document the influential status gained by many of his students, among them even the American President-to-be Woodrow Wilson (p. 44).

I caught only one minor defect. They point out that before Ely’s involvement, NAREB’s “Code of Ethics” already contained Article 34: “A Realtor should never be instrumental in introducing into a neighborhood a character of property or occupancy, members of any race or nationality, or any individuals whose presence will clearly be detrimental to property values in that neighborhood.” They do show that when Ely and his student Ernest M. Fisher started advising and writing NAREB textbooks and policy, this rule was never altered or challenged, but in fact codified; and Fisher openly endorsed the NAREB “Code of Ethics” in general. So even if they never explicitly mentioned Rule 34, we still can only conclude they supported it. But when this paper asserts that “along with the profession as a whole, Fisher assumed that the selling of real estate to differing races and nationalities was just as detrimental as introducing industry or commerce that would bring pollution, traffic, and crime to residential districts,” they have no specific evidence to cite (pp. 48-49).

I thought this might be poor wording. If they had written “Fisher probably assumed what the profession as a whole did,” their documentation (of this being indeed the common sentiment in that industry) more than proved the point. But the words “Fisher assumed” is a declaration of fact that one has to establish with more evidence than just this (like actual statements from Fisher himself), and I did not see it. Still, the corrected assertion (of what Fisher probably assumed, given his endorsement of a code of ethics enshrining it and what they well document to have been commonly assumed within his profession) would remain sufficient to sustain the paper’s core thesis, when placed in conjunction with the other evidence they go on to present. But I thought this an unlikely error for a paper of this quality, so I researched their citation of Fisher’s textbook and confirmed that in it he did say “the purchase of property by certain racial types is likely to diminish the value of other property in the section,” which is the evidence they needed to cite (and could have done better simply quoting).

So I think they simply forgot to add the page number to their citation in a corresponding footnote, thereby indicating where we would find in that work the evidence for their assertion. This is not something a reader of a peer-reviewed article should have to do, but any reader (even an amateur) could have done what I did here and thus benefitted from testing and verifying their claim. And I encourage all people to engage with sources this way, and not leave this to “the experts.” The oft-mocked claim that someone “did their own research” is only a joke because that doesn’t actually mean what it sounds like—rather than actual, productively critical fact-checking, too often what people who say that mean is more like “I went and found people who made assertions that agree with what I want to be true,” rather than actually checking if those assertions are credible, or even based on any evidence at all. But you actually could do that—do your own actual research, the real kind (on how that works, see my article A Vital Primer on Media Literacy).

In this case, all I did was some targeted Googling until I found an article on the matter at Platform, which linked to a digital copy of Fisher’s textbook, where I could verify the quote directly. One might object that Fisher was simply stating the reality (that all homebuyers are racists); but he did not object to this reality, rather treating it as an unalterable fact as plain as any other, which is precisely Winling and Michney’s point. This is how systemic racism works: Fisher does not have to be a racist himself; merely going along with the racist regime, and simply “accepting” racist pricing of the market, is sufficient to ensure the system itself remains racist. Because “Black houses are worth less” is racist: it improperly mistreats persons based solely on their perceived race (such as, in this case, by property valuation). And this has even wider racist effects, e.g. neighbors will seek to prevent Black people living near them for fear of its effect on their own property’s value, which they might do even if they are not themselves racists, but simply because the system (in this case, how our society chooses to assess the value of property) forces them to (e.g. if a drop in their own property’s value will cause an economic hardship they cannot afford).

This minor omission, of simply a page number, was easily fixed. All their other key premises that are themselves conclusions from accumulated evidence are well supported and cited. For example, the fact that by 1925 “Ely had a prominent university position for his research institute in a major metropolitan area, corporate and philanthropic funding, supportive students and colleagues, an alliance with the nation’s leading real estate organization, an emerging network of interested and sympathetic scholars across the country, and a journal to showcase new research,” Winling and Michney so thoroughly document that the likelihood ratio in its support is off the charts (pp. 48-51). There simply is no other plausible explanation of the evidence they present than that.

The Lynchpin of the Thesis

And this is where we get to the “Implementation” part of their thesis. And here Winling and Michney begin with how Ely and his school of thought (advanced by his students and colleagues) answered the question, “What is the nature of value and on what basis is it determined?” And how they got the government to agree with them. Which gets us back to my point about systemic racism: we usually don’t stop to realize that we could question the very underlying basis of real estate valuation itself. So why we ended up with a system that simply takes racist pricing as a given is something we need still to explain. Ironically, this is how “race-blind” policy ends up generating a racist system: by simply calculating supply and demand pricing blind to race, the racism of buyers will cause you to produce a racist pricing structure. No amount of “but I don’t see color” or “I never included color in my analysis” will save you from this charge: if all you look at is what people will pay, you are implicitly endorsing all the unsavory and unjust ways they are deciding that. You are thus helping to sustain a racist system, even if you yourself are not racist, or indeed never even think about race.

Winling and Michney show that until the Ely school’s influence on the industry, real estate had several different systems of valuation (emphasis mine):

First was the capitalization method, focused on investment, in which the value of a property represented the current price of future income from the property. Second was the cost replacement approach, oriented toward organizations and institutions that seldom sold their real estate. In this method, value was calculated as the cost of replacement or reconstruction—for example, if disaster befell the property. Third was the market approach, oriented toward consumers, in which the real estate in question was evaluated against comparable properties that had recently sold in a proximate sales area—value as the price a home buyer was willing to pay.

These aren’t even the only ways we could do it. They are just the three then used. But key to understand here is that the “natural assumption” that free market pricing should prevail is itself incorrect, because what really drives all of this is the mortgage economy. Few people buy homes (or other real estate) outright; and even most of those who do are speculators who depend on being able to turn their buy around and sell it to someone who is relying on a mortgage. Which in turn means that when the government subsidizes mortgages (through insurance, tax breaks, and other guarantees), the government itself is then inextricably affecting those decisions. The government’s decisions as to which mortgages to back, the conditions it sets, are then actually in substantial part dictating pricing. For example, the FHA requires backed homes to meet certain safety and livability requirements. But if the FHA is already doing that, it could as well mandate 14th Amendment equity requirements. If it doesn’t, it can only be because it simply chooses not to. Otherwise, that it can set any conditions it wants for which mortgages to back is already proved by the conditions it has already set.

In case you don’t follow the point, let me expand. Because most property is bought on mortgage, it is not in fact “the buyer” who is dictating prices, as if in some fictional Adam Smith utopia. If the buyer can’t get a mortgage, their price offer is going to be far lower than anything you can imagine, and sellers will either have to treat property as worth a great deal less than we now treat it, or just never sell (or only sell to the mega-rich, which will inevitably result in all land only being owned by what will then be in effect a landed aristocracy). So it’s really the banks deciding what property is worth, and it’s mortgages that drive prices to the lofty heights we now see. The buyer has some impact, insofar as what mortgage terms they can accept limit what prices they can offer to a seller, but even those terms are set by the banks. And banks do have some limits in what terms they can offer and still make a net profit on the market, so buyer limits on what terms they can accept still play a role in setting market prices. But the banks play an even larger role. And when the government is involved in backing mortgages to expand home ownership (which it has an interest to do to secure economic stability, and thus national prosperity, and thus even national security), really, the government is setting the actual pricing regime.

Thus, for example, the government, and banks, could simply valuate property in Black neighborhoods based on comparables in White neighborhoods—thereby accounting for all factors except race. That would require conscious admission that racism exists, and a conscious accounting of race—the opposite of “color-blind” decision making. The color-blind can’t see racism, and thus ironically sustain and perpetuate it; whereas the color-conscious can control for the variable of race and thus eliminate racism in pricing, returning it to just an evaluation of actual objective facts that matter. This would not eliminate all racism in the system; e.g., because of other forms of systemic racism, poverty or crime rates can be higher in Black neighborhoods, and poverty and crime rates are objective factors justifiably affecting property value, so by proxy some Black neighborhoods will get under-valued, but this will be for legitimate reasons (like demand for greater harm reduction), where the problem is not pricing, but the systemic causes of racial disparities in poverty and crime, which are their own separate problems to solve. But simply removing “color” as a factor has been proven even still today to radically alter home pricing even accounting for those secondary effects. As a Black family will learn if they hire White actors to pretend they are living in the house they are selling, or to pretend to be the ones buying the home (and yes, these are things actually being done); likewise if the homeowners on a street all hide when buyers come to check out a house. After all, if the buyer can’t see what race anyone is, then their racism won’t affect the price they are willing to pay.

That background isn’t in Winling and Michney’s paper because it’s understood by anyone reading real estate history. I provide it here only so you can better understand the context Winling and Michney are writing in. So when they get to documenting the Ely school’s effect on real estate pricing, you’ll better understand the importance of this. Accordingly, when they document that “the Roosevelt administration populated HOLC and the FHA with experienced professionals from Ely’s network,” and other links and connections whereby Ely’s ideas become policy (such as through his protege H. Morton Bodfish, another central character, who had his hands in almost everything relating), they have secured all three components of their thesis: these ideas originated and disseminated from Ely, and he and his network of ideological supporters played major roles in dictating the entire direction of the private real estate industry, even running the major government bureaucracies relating to that industry (as well as influencing the key legislation underlying all of that). The evidence they present for all of this is solid (pp. 52-60), entailing unassailable likelihood ratios.

The Role of Racism

This leaves the question of racism. While the creation of a system that had the effect of racist devaluation and segregation is undeniably in large part the work of Richard Ely and his network of students, colleagues, supporters, and proteges, Winling and Michney do not claim that Ely himself was a racist—whether one could argue that (and you could; just see his remarks on pp. 61ff. on “The Negro Problem” in Outlines of Economics in 1908), it is not relevant to their thesis and thus not a question they take up. This is the entire reason CRT exists as an academic paradigm: it doesn’t matter whether “Ely himself” was a racist; he still nevertheless created the resulting system that happened in result to be racist. By promoting certain changes in the way property is valued, mortgages awarded, and government involves itself in both, he established the system that would embody and magnify the fixed and openly-admitted assumption that White people don’t want to live near Black people and thus had to actively restrict or hinder Black access to White neighborhoods (and ghettoize Black people into segregated neighborhoods subject to poor investment, and government and industry abuse and neglect), hence those in power could use the levers of power he created to do this. Ely was not ignorant of this effect. He just didn’t care about it (as Winling and Michney outright quote him saying, “One thing that I long ago resolved was that I would not take up the Negro question”); neither did anyone else in his school of influence (who were all, perhaps needless to say, well-to-do White men).

Hence, the authors note, the federal government’s “secretive Mortgagee Rehabilitation Division
(MRD),” whose records later came to be disclosed, employed Ely’s system to advance White supremacism:

While scholars have debated whether race was the foremost consideration or but one important factor among several, it is beyond dispute that virtually all African American neighborhoods were rated hazardous for mortgage lending, receiving a corresponding “D” in the four-grade scheme and colored red on the maps (hence “redlining”). Not all “D”-rated neighborhoods were Black, but all Black neighborhoods were marked red, with only six known exceptions

And this was an open secret at the time (pp. 61-68). They document many members of the Ely school were involved, and expressed openly racist justifications for it in private correspondence. The most ironic of which is documentation that “Particularly confounding” to senior government bureaucrats involved “was the finding that in some cities the best record of repayment by mortgagors was actually in ‘D’ neighborhoods,” essentially refuting and contradicting the entire rationale for their mortgage-risk grading system. Nevertheless, the authors document how the racist principles for demarcating neighborhoods for mortgage appraisals used by government mortgage backing were disseminated throughout various industries through training, workshops, conferences, and guides.

By the time we get to this paper’s assessment, every sentence of it has been well established on a strong basis of evidence:

Ely’s venture into land economics is a particularly poignant example of academic outreach that had broad effects on public policy and the landscapes of everyday American life. The Federal Home Loan Bank, the Home Owners’ Loan Corporation, the Federal Housing Administration, and their restructuring of the real estate sector were outgrowths of the network of academics, private sector professionals, trade organizations, and lobbyists that Ely assembled in the 1920s. His research institute generated, encapsulated, and promoted ideas that worked their way through academic and policy conversations into law, regulation, and professional practice. Ely’s colleagues and collaborators created and supported new professions with agendas that powerfully shaped federal policy. In the process they structurally institutionalized racial discrimination as a fundamental feature of real estate value.

Conclusion

Sadly the controversial claims of Winling and Michney (that CRT is an epistemically productive, indeed necessary, way to study and understand American history; and that structural racism has plagued the real estate industry for over a hundred years, and openly racist decisions made then are still having substantive racist impacts now) are not even the ones they set out to prove, because they have been so widely and throughly established across thousands of academic papers and reports there is no functional need for them to be further proved. The only reason they are controversial is due to ongoing racist denials of scientific and historical facts. There is no such controversy among the actual experts studying this. Rather, Winling and Michney are merely filling in one unanswered piece of this story: “whose” ideas and influence launched the creation of this system; and to what extent did it move from the private sector to the state or from the state to the private sector. They find the answer to be Richard Ely, and that the state and industry were so throughly intertwined in their efforts, and so simultaneously influenced by Ely’s vision, that there simply is no way to assign blame to one over the other. The whole system was racist, top to bottom, public and private.

Of course they are not claiming Ely made the system racist. Racism pervaded and defined the conduct of private industry and the public sector long before (a fact Winling and Michney also document). Their thesis is that by not caring about its racist molding and abuse, and instead conceiving and enacting a system of valuing and mortgaging property tailor-made for it, Ely is the person most responsible for creating the system that became, within the real estate sector, the most effective tool of public and private white supremacism that could have been contrived by any nation ostensibly governed by the 14th Amendment. He didn’t do it alone, but through a network of sympathizers, students, and proteges. But that still would not have happened without him. Though possibly someone else might have stepped up and filled his social role had he been absent, that would only shift the blame to them. What remains clear is that had someone come forward with a different vision for molding the real estate industry, we would not have the problems we still suffer today, nor the even worse problems we suffered as a nation over the last century, owing to this blatantly racist machine “managing” citizen access to real estate.

The Winling-Michney thesis has a strong prior probability. They demonstrate this with extensive summaries and citations to background knowledge establishing how it fits the context of American society at the time, in society and government. And it has a strong likelihood ratio. Their accumulated evidence has no other plausible explanation (because it’s vastly less probable on any alternative to theirs). But this in turn tells us what we would need to “refute” their thesis. There are two possible avenues. First, one could try to show that they have misrepresented the evidence in crucial ways, thus undermining their case for a strong prior probability or likelihood ratio. But their documentation is extensive and by my own sampling solid and accurate, so this does not seem likely. Second, one could try to show that they have left out evidence that, when reinserted, would flip these probabilities the other way around. For example, maybe some other figure was more important and influential than Richard Ely in producing the outcomes documented in the Winling-Michney thesis. The evidence would have to be more probable on that alternative explanation, which means something would have to be improbable about Winling and Michney’s evidence on their thesis. And from what they present, that also seems very unlikely. In fact, it’s hard to find an example of anything not fully expected on their thesis; which means it cannot be outdone by any greater probability. Which brings us back to the “misrepresentation” concern, already deemed unlikely.

Nevertheless, if one wants to refute a thesis like this, they have to do it empirically, and that means they have to do either of those things: show that the evidence they present is not as expected on the Ely Nexus causal model as Winling and Michney claim, but more expected on some alternative causal model; or show that some evidence has been left out that, when put back in, tanks the Ely Nexus causal model’s prior probability or likelihood ratio. That is, you need evidence that is more probable on another causal model than theirs. If you can’t find it, despite a diligent search, then you simply have to concede that epistemically, their explanation of the evidence is by far the most likely. So much so, in fact, that it is itself a historical fact. And indeed, that is what we mean by “historical fact,” a proposition whose posterior epistemic probability is so high that it would be a logical absurdity to regard it as false.