Home / UCLA Housing Voice Podcast / Episode 15: The Legacy of Redlining with Jacob Faber

Episode 15: The Legacy of Redlining with Jacob Faber

Faber, We Built This: Consequences of New Deal Era Intervention in America’s Racial Geography

More

Episode Summary: In the 1930s, in the midst of the Great Depression, the Home Owners Loan Corporation (HOLC) was created to protect households from foreclosure and in some cases repurchase homes they’d already lost. As a part of its efforts, HOLC created “residential security maps” to categorize neighborhoods by lending risk, with low-risk neighborhoods shaded in green and blue, and high-risk neighborhoods colored in yellow and red. These infamous maps are where we get the familiar term, “redlining,” and they helped institutionalize America’s racialized housing market. Jacob Faber, associate professor at New York University's Robert F. Wagner School of Public Service, joins us to discuss his fascinating new research into HOLC’s influence on racial segregation in the cities where it operated, and the persistence of its effects nearly 100 years after the agency was created.

  • “Despite evidence of moderating racial attitudes and declining housing discrimination, black/white segregation remains stubbornly high … Many scholars place blame on the federal government for creating this sociospatial dichotomy through segregationist housing policies initially developed during the New Deal and expanded in subsequent decades. The Home Owners Loan Corporation (HOLC), Federal Housing Administration (FHA), and GI Bill created the contemporary U.S. homeownership society but largely excluded people and communities of color from affordable mortgage credit through explicit and implicit means, thereby reifying racialized neighborhood boundaries. By giving federal backing to the idea that proximity to people of color necessarily leads to property value decline, these policies created a powerful financial incentive for white communities to segregate themselves. Discouraging white home-buying inside black neighborhoods and constraining black home-buying outside of black neighborhoods strengthened stereotypes about quality of life and structural strength of black neighborhoods, creating a “vicious housing cycle” of self-reinforcing narratives that deepened racial stereotypes, the wealth divide, and racial isolation.”

 

  • “This is the first article to estimate the effects of early-twentieth-century housing policies on subsequent segregation patterns. Specifically, I leverage a full century of census data to show that the cities and towns appraised by HOLC became far more segregated than cities and towns ignored by HOLC. These findings are consistent across a range of specifications, including controlling for racial and socioeconomic characteristics prior to HOLC as well as change over time in those characteristics.”

 

  • “Appraised and unappraised places were similarly segregated in 1930 (i.e., before HOLC), and both groups experienced increases in segregation between 1940 and 1960, but the rise was significantly steeper for cities and towns HOLC evaluated. HOLC-graded places continued to be more segregated than their ungraded counterparts through the “decline” of the American Ghetto during the latter half of the twentieth century. In 2010, the black-white dissimilarity, black isolation, and white-black information theory indices were 12, 16, and 8 points higher, respectively, in places touched by HOLC than in other cities.”

 

  • “If America’s stark racial geography was intentionally constructed, work can be done to intentionally deconstruct it—although only through redirecting the momentum carried by public policy.”

 

  • “The Great Depression was not only an employment crisis, but a housing crisis—in 1933, approximately half of the nation’s mortgage debt was in default. During that pivotal year, to stem the tide of foreclosures and spur economic activity, the Roosevelt administration created the Home Owners Loan Corporation (HOLC), a new federal agency, as part of the New Deal. HOLC granted loans to homeowners to help them avoid foreclosure and reacquire homes already lost to foreclosure. HOLC also institutionalized the long-term, fixed payment mortgage, which largely replaced the ad hoc, patchwork system for purchasing homes that existed previously. In doing so, HOLC created the primary tool for wealth generation in the United States.”

 

  • “As the federal government took an increasingly involved role in the provision of mortgage credit, its interest in assessing the default risk of potential borrowers also grew, leading HOLC to map the distribution of perceived risk within cities across the country. HOLC appraisers, in consultation with local lenders, realtors, and other industry representatives, gave neighborhoods one of four grades: “A” being the most desirable (displayed in HOLC’s “Residential Security Maps” in green), “B” being slightly less desirable (blue), “C” being declining neighborhoods (yellow), and “D” being undesirable (red)—also the origin of the term “redlining.” Adopting the practices of the real estate industry, appraisers based these grades on housing characteristics, proximity to industry, and the sociodemographic characteristics of a neighborhood’s residents … Grades were geographically distributed in starkly segregationist fashion. The presence of low-income immigrants, or—most importantly—black families, effectively guaranteed a D grade. For example, in 1930, not a single black person in St. Louis—home to some 94,030 African Americans at that time—resided outside of D areas. By conflating race with mortgage default risk and home equity growth, these policies not only justified racial discrimination, but they also created a marketplace whose metrics of risk made discrimination necessary.

 

  • “...HOLC was designed to be a stopgap intervention. Had HOLC only served the short-term purpose of its intent, it may not have had a meaningful, long-term impact on America’s racial geography net of endogenous racial preferences held by citizens and real estate interests within the cities HOLC evaluated. In other words, HOLC may have been a relatively “weak” institutional intervention … However, the historical record documents that HOLC’s practices of racial exclusion were adopted by subsequent federal programs, which were larger and more durable. Most notable among these were the Federal Housing Administration (FHA) and the GI Bill.”

 

  • “My analytic strategy is designed to explore differences in segregation over time between cities and towns (i.e., census places) that were appraised versus not appraised by HOLC. I focus on cities and towns because HOLC appraisals were generally organized by place, and this unit of geography provides a large sample of both appraised and unappraised observations … My primary concern is the extent to which black individuals were segregated from white individuals [several reasons provided].”

 

  • “Gaps between HOLC-appraised cities and cities HOLC ignored widened dramatically in the post-World War II period. Figure 1 shows predicted values from linear regressions of each segregation measure on an interaction between indicators for HOLC appraisal and year (i.e., the average of each index for each year by appraisal status with 95 percent confidence intervals). Both groups of cities experienced increases in segregation, although the rise was sharper among the appraised cities—especially along the dimension of black isolation. By 1950, black families in appraised cities were more isolated than black families in unappraised cities, although again, this difference was not significant. This gap gained significance in 1960 and remained significant through subsequent decades. By 2010, ungraded cities had a similar average black isolation index as they did in 1930 (.202), whereas black households in graded areas are far more isolated today (.366) than in 1930.”

 

  • “The gaps between appraised and ignored cities that emerged in the 1940s have not closed in the intervening six decades. In 2010, appraised cities had .117 higher dissimilarity, .084 higher information theory, and .164 higher isolation—all of which are statistically significant (p < .001) and substantively important.  The isolation gap indicates that, compared to black residents of ungraded cities, black people in graded cities live in neighborhoods with a 16.4 percent higher black population share, on average. The difference in dissimilarity suggests that 11.7 percent more white households in appraised cities would need to move in order for there to be an equal distribution of white vis-à-vis black households in those cities. Along the measure of information theory, appraised cities were approximately as segregated in 2010 as unappraised cities were two decades earlier.”

Shane Phillips 0:04
Hello, this is the UCLA Housing Voice podcast, and I'm your host, Shane Phillips. Every two weeks we discuss a different housing research paper, translating it into non-expert, non-academic language to better understand why we face the challenges we do and what can be done about them. My co-host today is Paavo Monankenn. And our interview is with Professor Jacob Faber of NYU. This time, we're talking about the legacy of redlining and specifically the association between the Homeowners Loan Corporation and segregation. The Homeowners Loan Corporation or HOLC is responsible for creating the infamous racist redlining maps that many of us are familiar with, and which Richard Rothstein further popularized in his 2017 book 'The Color of Law'. Professor Faber finds that the places where HOLC created these maps saw a sharp rise in segregation between black and white households compared to places where HOLC didn't create its maps. And that effect has persisted all the way to the present day, more than 80 years after HOLC was established. That association is really interesting and really important, but it also does raise questions. HOLC was a short-lived organization, and as work by Amy Hillier, and others have shown, even though it created maps based on racist assumptions, it actually lended pretty broadly to black households. Other agencies, including the massively influential Federal Housing Administration, seem to have developed their own racially discriminatory practices independent from HOLC. So why are the places mapped by Hulk still so segregated compared to the places that didn't map even though redlining and similar discriminatory practices spread nationwide, and through so many other avenues such a long time ago? We get into that debate in this conversation, and in echoes of our interview with Kathy O'Regan, we talk about some of the human consequences of this racialized geography, and what responsibilities we're left with as the inheritors of federally institutionalized segregation. This is an interview that raises as many questions as it answers, but they're questions we should all be grappling with a lot more so it's absolutely worth your time. The Housing Voice Podcast is a production of the UCLA Lewis Center for Regional Policy Studies. And you can contact me with questions or research paper ideas at Shanephillips@ucla.edu or on Twitter @ShaneDPhillips. Now let's talk to Professor Faber. Joining us this week is Dr. Jacob Faber, Associate Professor at New York University's Robert F. Wagner School of Public Service with a joint appointment at the NYU Sociology Department because we love our sociologists on this show, you're the third or fourth on here. So far, I think the paper we're discussing, published in American Sociological Review, is We Built This: Consequences of New Deal era Intervention in America's Racial Geography. And this paper looks at how practices institutionalized in the 1930s by the federal agency, the Homeowners Loan Corporation, continue to be associated with segregated cities and neighborhoods in present-day America almost 100 years later. So Professor Faber, welcome to the show.

Jacob Faber 3:24
Thank you for having me excited to talk about this.

Shane Phillips 3:26
And I I feel obligated to note that we are holding this interview under protest from Mike Lens.

Paavo Monkkonen 3:36
Well, my presence is the protested part. He was mad that I got a firmer neuron without him. or we could just say he says hi.

Shane Phillips 3:45
Yes, that too. It's not a sincere protest. Yeah, he regrets that he cannot be here.

Jacob Faber 3:51
Hi, Mike.

Shane Phillips 3:52
So we'll start off how we always start off, tell us about where you're from where you live. If you were given us a tour, whether it's New York City where you're at or somewhere else you consider home, what would you want to show us? .

Jacob Faber 4:04
Yeah, I love this question. I, I live in New York City. And I think that's where I would say I'm from. I was born in the city, but grew up in northern jersey, but I've been back in New York now for about 15 years. And I am a lover of food and no food is the thing that I am most excited to show people. New York, to be honest, has really have ruined my ability to enjoy travel to a lot of places just because of how good the food is here. And it's you know, it's so diverse also that you can find just incredible variety in the type of food.

So were you taking us for hand-pulled noodles? Where are we going?

Yeah. Yeah, so one of my absolute favorite restaurants is a Thai place in Queens in Woodside called Tripper Pie. And it's at the top of every, like, best Asian food or best Thai food in New York City lists. And it is just transcended. It does not taste like any other food that I've had before. And we went there after my kid got her first COVID shot a couple weeks ago, and it felt like it was the first time we had been there since the start of the pandemic. And it's yet outside, but it was still just incredible. So that's a must see, or must eat. I'll put that on the list.

Paavo Monkkonen 5:50
And wait. So is there a better hand-pulled noodles than Xi'an Famous? I mean, that's the....

Jacob Faber 5:54
Yeah, it's a good question. I love that place too, and there's one just a couple blocks from where we live.

Paavo Monkkonen 6:02
... and I was thinking the other thing you said it ruined travel for you. But for me, it's like then you can go to the most special place for a certain dish. And like it gives you a new appreciation, right? Because even in New York, maybe the hand-pulled noodles are not as good as they are in Xi'an.

Shane Phillips 6:18
So I don't want to bury the lead here. So I'll kick us off by sharing the major finding of your article, which is that you find that cities where the Homeowners Loan Corporation assessed and graded neighborhoods using their pretty infamous redlining maps. These neighborhoods and cities are more segregated today, almost 100 years after the Agency was created compared to similar places where HOLC did not create these maps. And we'll get into what all that means in more detail. But since the history of HOLC and these other New Deal housing market interventions is really integral to understanding these findings. I want to start there, in the introduction you write "the Homeowners Loan Corporation, Federal Housing Administration, and GI Bill created the contemporary US Homeownership Society, but largely excluded people and communities of color from affordable mortgage credit through explicit and implicit means thereby reifying racialized neighborhood boundaries. So what problems were these programs trying to address? And how did they actually work? And I think we can probably focus on HOLC and the FHA since these came earlier, starting in the 1930s. Whereas the GI Bill came a little later, starting after World War Two.

Jacob Faber 7:36
Yeah, so I think most of the time when we talk about the Great Depression, we talk about it as an employment crisis. And it certainly was, you know, the worst employment crisis in American history, probably. But it was also a housing crisis. In 1933, about half of mortgage debt was in default, there were hundreds of 1000s, if not millions of people who were losing their homes, due to foreclosures, and evictions, and really just enormous amounts of Americans who were homeless. And so, these policies, especially HOLC, were designed to kind of stabilize the housing market, which was and remains a big part of the economy, and also as an employment program to encourage housing construction.

Shane Phillips 8:33
Right.

Jacob Faber 8:34
So HOLC was part of the 1933 Homeowners Loan Act, and the program bought distressed residential mortgages from private lenders and then refinance them with the original borrowers on easier terms. So in helping people at risk of foreclosure or already in the foreclosure process, to stabilize their credit obligations. And then FHA (Federal Housing Administration) came to be as part of the National Housing Act just a year later in 1934. And this was a mortgage insurance program run by the government to replace the private mortgage guarantee industry that had collapsed with the fall of the housing market in the early 1930s. And it required mortgage lenders to offer mortgage terms with longer time periods, lower interest rates, and lower down payment requirements that were previously available on the market. And because of these changes, these are these requirements, owning your home or holding a mortgage became much cheaper than renting in many parts of the country. And so, you know, we really have these programs to thank for the contemporary homeownership society, they are responsible for institutionalizing and making quite common the long term fixed payment mortgage. And this financial instrument lifted tens of millions of Americans into homeownership. And this became the primary tool for wealth accumulation for most Americans.

Shane Phillips 10:32
Right, right. Yeah, have it to thank or to blame.

Jacob Faber 10:35
Sure. Yes, right to blame, yeah.

Shane Phillips 10:39
So I realized I didn't really write about this in my notes, but I kind of took it for granted that people would would know what redlining was, but I don't think we should assume that here. So can you just talk a little bit about what redlining, where that concept comes from, and what role can FHA played in that?

Jacob Faber 10:58
Yeah, yeah. So now that the government is concerned about mortgage lending, they are concerned about lending risk. And so they hire appraisers, to go to hundreds of cities across the country and assess the lending risk in neighborhoods within those cities, and may base those assessments on a number of different characteristics of the neighborhood, including the existing housing stock, proximity to environmental hazards, or parks, as well as the characteristics of the people who live there. So well, you know, income people there, migration status as well as race, these appraisers were obsessed with the precedence of black people, a single black person in your neighborhood could guarantee you the lowest grade, which was a D, these grades went from A to D with A being the highest grades, and D being the lowest grades. And appraisers use these grades, excuse me to make maps of the cities, these maps were then called residential security maps. And it became harder to get mortgage credit in D neighborhoods, red neighborhoods over the subsequent generations, and this is where the term "redlining" came from, because they were literally drawing red lines around poor neighborhoods, black neighborhoods, and the like.

Paavo Monkkonen 12:35
And I had a follow up question about, this is also off-script, but I'm very curious to know about the appraisers themselves. I wonder if you in your research you looked into that at all, there's been a lot of good work historical work on realtors, like you know, Page Glotzer's book and Laura Redford. And I wonder if I mean, probably in some cases, they were the same people, or were the appraisers kind of coming from DC to these cities or were they already kind of people active locally in the real estate market?

Jacob Faber 13:06
Yeah, it's a really good question. So we have some historical record of the business of appraising for HOLC for FHA, but a lot of the pieces are missing. We know that appraisers coordinated with local real estate concerns to assess risk. But we also know that appraisers, you know, received extensive training from HOLC and we're supposed to be able to just be dropped in any city and successfully able to appraise neighborhoods within that city. Of course, that's not how anything works. And I'm sure there was quite a bit of variation in the stringency of neighborhood appraisals across appraisers. But, you know, I think that, this is still a question to be answered.

Shane Phillips 14:04
Maybe we should walk through a little bit how this redlining actually could cause segregation, and then we can talk about what the consequences of that segregation would be. So we have these red line districts, these D grades and these C grades, how is that actually preventing, you know, black households from moving into white neighborhoods, discouraging white households moving into black neighborhoods and so on.

Jacob Faber 14:30
So as you know, the top line impact was that it became harder to be a homeowner in redlined areas, especially if you're a person of color, but also for white people became easier to be a homeowner in neighborhoods that got the best grades, the green grades, and so this kind of pulled people in both directions to segregated neighborhoods and cities and towns,

Shane Phillips 15:01
And you've done a lot of work on the consequences of racial discrimination and segregation in the housing market, including how communities and households are, you know, ultimately impacted by it. Could you give us just a quick primer on some of those consequences? Like, what do our neighborhoods look like today, at least, you know, in part because of these government interventions, and how have people been actually harmed? And how do they continue to be harmed by the racial geography that you described in this paper?

Jacob Faber 15:29
Yes, this is a great question. I mean, these programs were a long time ago. And there's been a lot of change since then. And quite a bit of variation in the trajectories of neighborhoods. But on average, the places that were redlined, almost a century ago, are still marked by disadvantage today. So they tend to be places of higher levels of racial isolation, higher poverty, lower home values, and homeownership rate, lower rates of intergenerational economic mobility, contemporary mortgage credit is harder to get in historically redlined places, and it's more expensive when it is given. And there's even some research not done by me but connecting the legacy of redlining to vulnerability to climate change that historically redlined neighborhoods are much hotter than neighborhoods that got better grades.

Shane Phillips 16:31
Yeah, we'll have to link to that paper in our show notes. Another line from your paper that stood out to me is this, "new deal housing programs certainly did not invent segregationist mortgage provision but they institutionalize the practice and implemented it at an unprecedented scale". This is a point that Richard Rothstein also drives home in his book, The Color of Law, and in that book, he also quotes Sherilyn Eiffel, who says something to the effect of as a society, we enjoy privileges and freedoms that we did not earn that were secured by our predecessors and ancestors. And that means we also carry responsibility for the harms that were imposed or not addressed by our predecessors and ancestors. I'm curious if you have any of your own reflections on what it means for us to institutionalize these practices through government programs and policy, and what kinds of obligations that might impose on us that a less institutionalized system might not.

Jacob Faber 17:28
Yeah, we too often equate responsibility with blame, which makes people defensive, you know, especially in your conversations around race and the history of race.

Shane Phillips 17:40
Yeah.

Jacob Faber 17:41
But let's say I stole $1,000 from you and gave it to my kid who took that $1,000 and invested it in a mutual fund, you know, 50 years later that $1,000 has appreciated quite a bit. And, you know, my child has more wealth than your child's, you know, assuming you have a kid in this example, of course, and, you know, let's go another generation forward, let's say my child use the wealth accumulated to invest in their kids' education, their kids' entrepreneurship and homeownership. You know, how should we be thinking about the difference in success between my grandkid who benefited from my thefts, and your grandkid? You know what if we go another generation forward? How much has that initial $1,000 grown not just in the form of financial capital, but also human capital, like education, investments in health cetera. I think most people would agree that my descendants should compensate yours in some fashion in this example. And this example is just within, you know, my family and your family. What happens when a government enacts a policy that has intergenerational consequences for inequalities? What if the federal government took that $1,000 from you and gave it to me? And there are plenty of government decisions that were made well, before I was born, and most well before most of your listeners were probably born that we are still paying for today; the Korean and Vietnam wars are fought, decades before I was born, but my tax dollars still support the pensions of people who fought in those wars. So to me, there's a moral responsibility to right the wrongs of the past. And this is research that really exploded, documenting the long-term consequences of this federal redlining and really underscores the importance of reparations. And there is also a powerful cultural impact of these policies, which is even harder to measure these policies through the appraisal process, conflated race with lending risk, and property value, thereby linking whiteness to wealth accumulation through this really powerful market mechanism that encouraged segregation. And I don't even know how we go about measuring the impact of these policies on racial animus, you know, these policies that legitimated and institutionalize racism, wrecking havoc on virtually all aspects of American life.

Paavo Monkkonen 20:32
Yeah, I think that's a great point. I mean, I think these are the things that it's much harder to measure, right? It's kind of how this conflation between race and property values like plays out in so many different ways in society. And I think that point about, you know, imagine this one generation forward, 2, 3, 4. Not only does it illustrate just how, how big the effects can be, but as you go forward in time, and the space between when things started and where we are today grows, teasing out, like who's benefited, who's been harmed, how you, you know, shift resources just gets more and more difficult as well. And I think we're in that place now, for sure.

Jacob Faber 21:11
Definitely.

Paavo Monkkonen 21:12
But let's talk about some things we can measure, I guess, I would compliment this paper, especially I'm teaching a class on zoning for equity that has a lot of history in it. And so I thought this paper was really useful for that class to connect kind of 100 years of history. And I complement the painstaking data work that went into doing this research. So you know, the basic idea is looking at places that were appraised by HOLC and places that weren't, and kind of how did that affect their segregation trajectories and levels of segregation today? So can you maybe talk about how you came up with that idea? And how you went about making it happen?

Jacob Faber 21:50
Yeah, so you know, ideally, I would have used longitudinal datasets with the same measures of race and the same number of measures of municipal boundaries and neighborhood boundaries, from before HOLC was implemented all the way through today. And unfortunately, those data don't exist. So I pulled together data from almost a century's worth of decennial censuses, and measured segregation on the city level for, as you said, places that were appraised in places that were not in have compared the trends over time. And many of your listeners might be thinking, "okay, well, there maybe was something inherently different about places that were appraised compared to places that weren't". And there were differences, appraisal was not random; the federal guidelines requested residential security maps for cities with at least 40,000 residents. But there was quite imperfect implementation of that guideline. You know, we know we have maps, for example, of entire counties that include a number of towns and cities, we have maps that were drawn for sections of metropolitan areas that include, again, multiple cities. And so there are a lot of cities that overlap with these residential security maps, redlining maps, that had populations below 40,000. And there are also a number of cities that we don't have maps for that were above 40,000. So what I also do as a robustness check, is to compare the impact of HOLC over time, within a number of subsets, sub-samples of cities and towns, that were around this 40,000 cut-off of just below it, and just above it, which includes a mix of small cities that were appraised and large cities that weren't. And, you know, admittedly, this rests on a non-trivial assumption, which is that all cities near that cut-off were at risk of being appraised, which is admittedly a large assumption but, you know, I think that the the results are pretty compelling and consistent, so I believe them.

Paavo Monkkonen 24:26
Yeah I know, and even if they're, you know, thinking about the reasons that one city would have been appraised while a similar city wouldn't have been appraised, I think no matter what those reasons are the differences 100 years later approved are pretty striking right? But and so on that have you, I wonder what you think about kind of why some cities that were bigger, weren't appraised and I mean, I understand more why some smaller cities were if they're kind of caught in a county or in a metro area that had a larger kind of appraisal catchment area but kind of these larger cities. weren't appraised or, you know, the role of local actors in seeking appraisal, because that was kind of the first thing I was thinking of is, you know, if you have these segregatory places, asking for this redlining maps to be drawn, maybe that could be one explanation.

Jacob Faber 25:15
Yeah, we know, thanks to work by historian Nathan Connolly, that there were instances where local real estate concerns, tried to manipulate the appraisals for the purpose of real estate speculation. So you know, if you were a real estate investor, and you could convince the home appraiser to give a neighborhood a lower grade, that would make accessing these federal programs more difficult, which would increase foreclosures, and therefore reduce property prices, making it less expensive for you to buy land in those places. But to circle back to your question, you know, we really don't know, this could just be a complete accident of history. It's possible my data have errors in them. But for example, you know, we don't have a map for Washington, DC, which was definitely bigger than 40,000 people in 1930. But we don't have a map, and because of the secrecy, that HOLC and later FHA, exhibit around their practices, which we can talk about later, you know, it's possible that there was a map made for DC, and it was just lost to history, and we just don't have it anymore. So it's, you know, it's possible that there are a number of cities in my data set that I have coded as not graded that, in fact, were. And, you know, whether or not the historical data exists to figure that out is an open question. Maybe they're sitting, you know, maybe these maps are sitting at the bottom of a dusty box in some basement somewhere, you just don't know, right?

Paavo Monkkonen 27:13
Or the archival retention strategy was to destroy them at some point.

Jacob Faber 27:17
Yes. Yeah.

Paavo Monkkonen 27:18
Which would be very unfortunate.

Shane Phillips 27:19
Can you give me a sense of in 1930, there were 5000 municipalities, incorporated cities in the country are and how many, I don't know if that's true, I'm just guessing, and for how many did you have maps, and kind of how many are in your study sample?

Jacob Faber 27:37
Yeah, so there were about 200 cities that we have maps for, or that kind of overlap with, with maps, right. And the number of cities in my sample varies a little bit from year to year based on census data availability, you know, census data is was good in the early 1900s, because we have access to the complete counts on the individual level, and then the data get bad in 1950, and 60, and then they get better again in 1970, 80, 90, and 2010. And so I have a lot of cities in the sample, several 100 and fewer, and then, and then many more.

Paavo Monkkonen 28:26
Interesting, yeah, so I guess moving on to the kind of outcome of interest and your paper, which is the kind of level of segregation I'm used three different... I'm a big information theory index proponent, but I like also using multiple indicators for a phenomenon. So maybe talk about the pros and cons of the three different indexes you use and why you chose to use three.

Jacob Faber 28:49
Yeah, so I use these three different measures, which are among the most commonly used measures of segregation, the isolation Index, which measures that typical neighborhood, percent black for a black resident within a city, the dissimilarity Index, which measures the percent of either blacks or whites that would have to change neighborhoods have to move for there to be an even distribution of that group within the city, and then the information theory Index, which measures unevenness by comparing the diversity of neighborhoods, in my case tracks, within a city to the overall diversity of that city. And, you know, ideally, in social science, you'd like to have a number of different ways of measuring that outcome you care about, because every measurement has limitations. And these measures of segregation have a number of limitations like they're somewhat sensitive to the racial makeup of places as well as the choice of geographic unit. And perhaps most importantly, none of these measures are spatial. They're all aspatial measures of segregation. So, you know, thankfully, for me professionally, all the results all poin in the same direction. And there are no real differences across any of these measures.

Paavo Monkkonen 30:29
And maybe before we get to the results, I'm curious how you think about kind of what I see as kind of a challenging complicating factor, which is the role of municipal boundaries in metropolitan level segregation, and kind of the you the choice you made to measure segregation within a municipality, rather than kind of metro areas where more of the places were HOLC mapped, compared to other metro areas where fewer of the places were mapped?

Jacob Faber 30:58
Yeah, this is a really excellent question and something that I continue to struggle with, empirically, and theoretically, as well, you know, a lot of the segregation literature analyzes metropolitan areas, so like a city, and its suburbs, because those geographic units are usually drawn to capture, like an entire employment market or housing market. And when we think about the way that kind of form that segregation typically took in the middle to end of the 20th century was a very stark-like city suburb split, and so, if you measure segregation on the metro level, metropolitan area level, you will get higher estimates of segregation than I did, that's why I measured on the city level or place level, were in places and cities tend to be more homogenous than metropolitan areas, which is why their levels of segregation tend to be lower. And again, I have really struggled with trying to figure out what was the right geography for measuring this, because again, of this city suburb dynamic, especially, you know, moving past HOLC, and into the Federal Housing Administration, which really ramped up suburbanization. I feel like my paper really misses that whole story. But the choice I was, I was really kind of forced into this choice because of identification strategy. That, you know, trying to identify a causal effect on on the Metropolitan level felt like it was much more difficult, given the nature of HOLC maps. You know, I'm trying to challenge that assumption of mine, a bit in more recent work, but I think I would love to figure out a way to understand the kind of total effect that these policies had on segregation, which, again, my estimates, probably understate.

Shane Phillips 33:22
Alright, so we shared the very high level findings at the very beginning. But I do want to get into those in more detail now. So, you know, it's one thing to find that the cities that were appraised or mapped by HOLC are more segregated than unappraised cities. But really the scale of the difference between these two groups, and the statistical significance is really impressive. And you know, you, you were kind of humble and saying, maybe the data's just not there, the maps are hidden away somewhere. But I do think just based on the statistical significance here, the amount of maps that would have to turn up that have the opposite results of the ones you've already looked at, it would have to be very large to kind of overturn your findings here. So for each of these measures of segregation that you looked at, what were the results here?

Jacob Faber 34:15
So just to use the numbers in the paper, the average, white black, the similarity, isolation and information theory indices were 12,16 and 8 points higher, respectively, in places that were appraised compared to places that weren't an on average across all of the post HOLC years. And, you know, these numbers, they don't mean anything in the abstract, but they are in the, you know, the world of segregation scholarship, really quite large differences. The isolation gap indicates that compared to black residents of ungraded cities, black people in graded cities live in neighborhoods with a 12% higher, excuse me 16% higher black population share on average, which is a pretty big difference.

Shane Phillips 35:13
And that's percent not percentage point, right?

Jacob Faber 35:16
Yes. percent. Yes. Okay. Right. And, you know, in terms of the dissimilarity and information theory indices, you know, one another way of thinking about the magnitude of these effects, cities, segregation along these two indices has been declining since about, you know, since 1960, or 70. And cities that were appraised were or remain, effectively two decades behind cities that were not appraised on both of these indices. So that's a lot of difference.

Shane Phillips 35:52
What do you think the mechanism or mechanisms are, by which these appraisals are contributing to this increase segregation, and not just, you know, this sort of almost immediate increase in the 40s and 50s. But then this persistent increase that has lasted in many cases to the present day, you know, the association between these things - between the appraisal, and the segregation is very clear. And the magnitude of the difference in the statistical significance speak for themselves. But what's the story for how HOLC's involvement may have actually caused or contributed to this increased segregation? And I'll kind of combine a few questions here, because I think what Paavo and I have in mind here is some other papers that have come out over the past 20 years really, starting, or at least, including Amy Hiller's paper, looking at Philadelphia, and HOLC there, and more recently worked by Fishback, Snowden, Rows and Stores, which found that HOLC itself actually loaned pretty broadly in black neighborhoods. So how does that all fit in together? What's the mechanism? Just curious to hear your collections on this picture?

Jacob Faber 37:06
Right, so you know, first, there is a ton more work needed here, as you know, especially on the relationship between these HOLC maps and HOLC inheritors, like the FHA and the GI Bill. And so in some way, it's, you know, I kind of do this in the paper kind of hand wave mechanisms. But, you know, there are likely or were likely a wide range of mechanisms, including kind of limiting access to neighborhoods and towns, for primarily people of color, so kind of forcing people to live separately. Also penalizing whites who lived in mixed-race neighborhoods with lower grades. And then you can think about the second-order consequences of this limited access to homeownership, and its implications for racially differentiated wealth accumulation. So, you know, if you can't buy a home, or can't buy a home cheaply, then you were unable to accumulate wealth, and then pass it on to your children and to later generations. And, you know, that wealth kind of accrues over time, and now we have an incredibly large wealth gap that limits people's mobility, and where they can live even more. And there's also these self-reinforcing consequences of segregation, like the segregation of information, and you know, what neighborhoods people even know about. But it's, you know, I'm really happy that there is a rapidly growing body of work in this field. And I've been in conversation with, with Thomas Stores about the, you know, kind of differences in our findings, and hopefully, we'll have an answer for you in maybe a year from now. But, you know, it seems incredibly clear that HOLC maps did something; my work is a small part of this growing body of research that shows really profound impacts of these maps. But at the same time, the specific mechanisms which social scientists really have stressed over are kind of unclear. I'm working with a team of graduate students here at NYU, Riley Sandow, and Kate Thomas to find more data on lending specifically, but I'm also really interested in how HOLC fits into other historical processes. We've had massive social and economic, cultural and geographic changes, since these maps were made in the late 1930s including, of course, the banning of housing discrimination and redlining, you know, lightly enforced, of course, but, you know, one of the things that I really strongly believe, and this is not an original thought, is that HOLC helped set up subsequent processes, just going to set them in motion, or help direct them in discriminatory ways. So for example, maybe mid-century zoning policies were implemented differently in places that had maps, and in places that didn't. It's possible that these maps, and the effects that I'm finding, you know, that the effects are less about housing than it is about something else, like schools, maybe, or congressional districts or commercial in- development. So this really excites me, there's so much more work that needs to be done. And I'm really happy to be part of this conversation.

Shane Phillips 41:00
Got your career laid out for you.

Paavo Monkkonen 41:03
That is exciting. I'm curious. I mean, back to this, like the idea of selection bias in terms of which places get get mapped, in which places don't, I wonder if there would be other ways or whether you've thought about how to kind of get at that question, just looking at, you know, other kinds of data on racial animosity, or kind of like how racist local leaders were in different cities at the time and kind of trying to correlate that with, with getting mapped or not.

Jacob Faber 41:31
Yeah, no, so I have been trying to think about this. And I know that there are many other people who are in more advanced stages of this, which I'm happy about is understanding, related to your question earlier, a variation in appraisers' decision making. So for example, you know, if you could find a measure, you can estimate a measure of the racial animus of appraisers on the city level, you know, that measure is likely to be predictive of future outcomes.

Paavo Monkkonen 42:08
Right.

Jacob Faber 42:08
And, you know, that could really give... you know, provide a lot of insight into this mechanism question, you know, it's possible that there are archival records, you know, correspondence between appraisers or between HOLC, and local real estate concerns that we don't know about. And, you know, there are, I think it's possible, just to kind of pinpoint the critique of all this work, that there were plenty of vectors of discrimination that were well in place and well underway before HOLC...

Paavo Monkkonen 42:47
Right,

Jacob Faber 42:48
....but you know, the counterfactual is that billions of dollars of federal investment in segregation, and reorganizing an entire financial market had no impact. And, you know, I find that just, like, impossible to believe.

Shane Phillips 43:06
Yeah.

Jacob Faber 43:06
And, you know, I, I also tried really hard to make the findings go away, and put it so, you know,

Paavo Monkkonen 43:15
There's clearly good social science there. What people have said, like, so this is my history of this correct that like no one had written about HOLC maps until Ken Jackson found them in a basement somewhere and kind of like going to this like, seek, like, " Did people see these at the time, and how prevalent was their kind of visibility?" I wonder if you know more about that, I assume you know more about that than I do.

Jacob Faber 43:39
I, you know, know maybe a little bit more, and I'm still in the process of learning and going through some of the archival data myself that, that's the data that's available online, this is the way we do research and during pandemic. But my understanding of the history, and I found this again, in the archival records, as well as in like newspaper and magazine reporting, at the time, there was quite a bit of concern among civil rights groups that these programs were discriminatory. And so, you know, people knew at the time, what they were doing, the FHA was sued, and you know, it's likely that the fact that we don't have any FHA maps today is because they destroyed evidence during a discrimination lawsuit.

Paavo Monkkonen 44:34
I see a raid

Jacob Faber 44:35
Right, right. You know, we have one map left, Chicago, that's the only one. And lending records also don't seem to exist anymore, which also seems hard to believe that they never did. But I think that Ken Jackson is probably the first published social scientist to write about the impact of these maps but certainly people at the time were organizing around them.

Shane Phillips 45:04
Good reminder that academics that researchers are not always the most on top of these things.

Paavo Monkkonen 45:10
And I think it's not surprising that the white Academy ignored this for Yeah, middle of the whole middle of the 20th century. Right.

Jacob Faber 45:19
Right. Right.

Shane Phillips 45:20
Okay. So, you know, you go to great lengths here to show this association between Hulk appraisals and racial segregation and do so really effectively. And you've been, again, very humble about kind of the limitations and how much more there is to learn here. But I'm curious to hear kind of some of the four things you haven't addressed already. What are some of the criticisms you've you've received? And if you're trying to kind of steal man, your own position here, or at least, you know, even if you're not fully standing behind, like everything is paper is definitely true, no matter what, if you are trying to make that case? What are some of the criticisms you've heard? And how do you respond to those?

Jacob Faber 46:00
Yeah. So I think that there are two broad criticisms that overlap quite a bit. The first is, you know, the endogeneity concern that there is something inherently different about cities that were appraised and cities that weren't that is the reason that they became more segregated over time. And I hope that my taking advantage of this imperfect implementation of the mapping guidelines, allay some of that concern. And the other, again, kind of largely overlapping critique is that these policies didn't do anything new. That, you know, banks were, they were redlining, before that term was invented. You know, there was plenty of mortgage discrimination that existed beforehand, and that's certainly true. But, you know, again, it seems quite clear, I'm not the only person who's saying this now, that these maps had a causal effects. And, you know, again, the counterfactuals, that these huge programs that created the American homeownership society and explicitly discriminated against people of color, had no effect on racial inequality. And that just doesn't doesn't make any sense.

Shane Phillips 47:26
Can you talk a little bit about the difference you observe between impacts in the South and the rest of the country?

Jacob Faber 47:34
Yeah, so I find a much stronger effects among cities in the South than in the rest of the country. You know, the effect is statistically significant for what that's worth. But it's, you know, I feel less confident about it than I do about the overall effects. I mean, the historical narrative of it makes sense. You know, there was more explicit resistance to civil rights and housing discrimination efforts in the South than there was in the North, but there was plenty of that in the North, as well. And so, you know, once we kind of get into this regional differences piece, I become a little bit more concerned about the endogeneity. But, you know, the effect is there, and it's quite big, the impact in the South was like, twice the size of impact in the rest of the country, which is really quite big.

Shane Phillips 48:45
So near the end of the paper, you consider some counterfactuals, one of which assumed that HOLC had just never existed and the other assume sort of the opposite, that we adopted an anti-racist integrationist housing policy agenda, instead of the discriminatory segregationist policies we actually had. And, you know, I realize, anytime I start asking on this podcast about forecasts, projection, speculation, all researchers are hesitant to do this. But it's in your paper so I feel like it's fair game. You know, what would you want to have seen from that anti-racist counterfactual? You know, I think we can take it as a given that FHA would not have engaged in redlining and the promotion of racial covenants as they did that the GI Bill would have been available to veterans of all races and ethnicities, but what else would you be looking for out of this?

Jacob Faber 49:42
So full disclosure, this section of the paper was not in the original draft, and I added it on the insistence of one of the anonymous reviewers. And I'm really glad that I did because It's the part of the paper that I get the most questions about from journalists, especially, because it kind of is quite straightforward. And it's something that I plan on incorporating into future papers as well, because...

Shane Phillips 50:17
You don't get a lot of questions about research design on the forum?

Jacob Faber 50:22
Not always, but, you know, just to quick complain, tangent about academic publishing, you know, we are so often told to be so conservative and narrow with our conclusions, which oftentimes makes the research kind of context lists. You know, I do this research because I care about policy. So if I'm not able to say something about policy, and what good policy could have looked like, or the impact that good policy could have had, and, you know, I feel like the work that I do isn't as useful. But to your question, there there are many things that a policy could have done to advance an anti-racist agenda; HOLC could have, you know, encouraged financing of affordable housing in wealthier neighborhoods, for example, through rewarding rather than penalizing neighborhoods that had both single-family and multifamily homes. Similarly, it could have rewarded neighborhoods for being integrated rather than referring to it as it was referred to on the appraisers documents as "infiltration". They could have specifically encouraged lending on affordable terms to African Americans, and interestingly, so I've been reading through the Federal Housing Administration's archives recently, and at the start of these archives, the first ones in 1935, these annual reports, and they are strongly encouraging restrictive covenants and exclusionary zoning. And then they stop doing that once the Supreme Court said that restrictive covenants were illegal in, in Shelley versus Kramer in 1948. And then, after that decision, all of the FHA annual reports start claiming that they care about minority homeownership and that they're encouraging minority homeownership with no data. Of course, they have no data to back up this claim. But anyway, you know, these policies FHA and HOLC could have fought discrimination in lending through hiring black employees, or audits or data collection, you know, we have the Home Mortgage Disclosure Act, now that requires lenders to provide data on mortgage applications, and we could have implemented something like that. But again, going back to something that we talked about earlier, the original sin of all of this to, you know to borrow the famous phrase, is the conflation of race with capital accumulation. And right now, I think a lot of this damage would have been avoided if HOLC simply didn't do that.

Shane Phillips 53:34
I mentioned you, you know, if you want it, you've got a lifetime's worth of research to do on all of this. And I know that this paper, even that we're talking about is not even your most recent one on the subject. So just curious if you could kind of talk a little bit about your paper with Daniel Aronson, Daniel Hartley, Bhashkar Mazumder, and Patrick Sharkey, that compared socio-economic outcomes for people living in HOLC and non-HOLC cities, so not just how segregated are they, but actually, you know, how are people's lives going in these places? What do you find there, and you know, what does the future hold for your research?

Jacob Faber 54:13
Yeah, so the, that group of economists, Dan Aronson and Dan Hartley and Bhashkar Mazumder wrote a really good paper on the neighborhood-level impacts of redlining on racial segregation and home values and credit scores, etc. And so I had Sharky as a collaborator of mine and I reached out to him, reach out to them, to see if there were other outcomes that were worth looking at. And so we kind of applied the method that they innovated in their paper, looking at buffers around neighborhood grade boundaries, so like boundaries between C graded neighborhoods, and B grade neighborhoods, and then C and D neighborhoods, and we looked at a large range of outcomes, including the likelihood of upward mobility across generations, incarceration, poverty, having a single parent, or having a kid as a teenager, etc. And we found a pretty consistent pattern across all of these outcomes, that the likelihood of having a positive outcome was significantly lower in redlined neighborhoods.

Shane Phillips 55:37
You know, I like this kind of paper just as a compliment, because it it sort of illustrates what you know, the Kerner Commission report 50 years ago stated that, you know, segregation is not this harmless thing. There are actual negative consequences for people's lives beyond the isolation itself. So I think that's a really important follow up. Anything else you want to add before we let you go?

Jacob Faber 55:59
Yeah. So I think that this work makes two broad arguments. The first is the importance or centrality of public policy in shaping opportunity, right? These policies, made homeownership possible by making it cheaper per 10s of millions of Americans and homeownership, skyrocketed from the low 40% to the mid 60 percents because of these policies. And flowing from that is a real importance of acknowledging the intentionality behind these programs is where we built this part of the title of this paper comes from that, you know, the racial inequalities that we see today, whether that's segregation or incarceration or mobility, this was all intentionally constructed through public policy. And I think that this sheds light on the kind of the scope of the problem and the bearer of responsibility, which is , you know, collective responsibility, and therefore, a real moral need for remedy.

Shane Phillips 57:19
Professor Jacob Faber, thanks for being on the show.

Jacob Faber 57:22
Thank you. This is great.

Paavo Monkkonen 57:23
Yeah, I want to thank you for your service. I know these data intensive projects are unrecognized in terms of how much painful work go into it. So I think it's extremely-the research community is grateful to you, you know,(for) highlighting these things. We're here to recognize.

Jacob Faber 57:40
Thank you.

Shane Phillips 57:46
You can read more about Professor Faber's research and find our show notes and a transcript of the interview at our website lewis.ucla.edu. If you haven't done so already, please be sure to subscribe to the show on your favorite podcast platform and tell your friends and colleagues where to find us. We release a new episode every two weeks. The UCLA Lewis Center is on Facebook and Twitter and I'm on Twitter at Shane D Phillips. You can find Paavo there @elpavo Thanks again for listening to the UCLA housing voice podcast. See you next time.

Transcribed by https://otter.ai

About the Guest Speaker(s)

Jacob Faber

Jacob William Faber is an Associate Professor at New York University's Robert F. Wagner School of Public Service and holds a joint appointment in NYU's Sociology Department. His research and teaching focuses on spatial inequality. He leverages observational and experimental methods to study the mechanisms responsible for sorting individuals across space and how the distribution of people by race and class interacts with political, social, and ecological systems to create and sustain economic disparities.
Learn More

Suggested Episodes