Disparate Impact
by Randal O’Toole, The Antiplanner, March 27, 2020
In 2000, Dartmouth University economist William Fischel coined the term homevoter to describe the fact that, in most of the United States, a majority of voters were also homeowners and that they would tend to support policies that boosted the value of their homes. In particular, Fischel argued that homevoter preferences were responsible for the rise of no-growth and slow-growth laws that became popular on the West Coast and in the Northeast after Earth Day, 1970.
This suggested that efforts to abolish urban-growth boundaries were doomed to failure. Why would Portland or San Jose homeowners ever vote to get rid of boundaries when doing so would reduce the value of their homes by 50 percent (in Portland) to 75 percent (in San Jose)? Even in places where a majority of people were renters, renters are less likely to vote than homeowners, so getting rid of the boundaries appeared to be impossible.
However, a new rule published by the Department of Housing and Urban Development in 2013 offered an alternative solution, one that by-passed voters and their elected representatives. This solution was called disparate impact, and it said that land-use regulations that made it more difficult for low-income minorities to live in an area were effectively the same as putting up a sign saying “No Blacks Allowed.”
The Fourteenth Amendment to the Constitution forbade the federal and state governments from doing anything that would “deny to any person within its jurisdiction the equal protection of the laws.” In 1968, Congress passed the Fair Housing Act, which effectively applied to Fourteenth Amendment to housing. Under this law, no one could refuse to rent or sell a home to someone else based on their race, religion, or national origin. This was a follow-up to the 1964 Civil Rights Act that said that employers were not allowed to refuse to hire someone on the basis of race. Refusing to sell or rent a home or to hire someone on the basis of their race was known as disparate treatment.
Duke Power in North Carolina had a long history of discriminating against blacks in its hiring practices. On the same day that Congress passed the 1964 law, Duke Power changed its policies to require people to pass certain tests to be eligible to be hired in certain jobs. Whites were ten times more likely to pass these tests than blacks. When the company was challenged in court, Duke Power attorneys argued that there was no proof that the company had imposed these new requirements in order to discriminate against blacks.
When the case reached the Supreme Court, however, the court unanimously agreed that it didn’t need proof that the company was deliberately discriminating against blacks. Instead, just the fact that the tests resulted in a different impact on blacks than on whites was evidence enough to “create a strong inference of discriminatory intent.” “Federal judges have decades of experience sniffing out pretext,” the decision observed. Thus, policies that have a disparate impact on minorities were effectively equated with those that disparately treat minorities.
HUD’s 2013 rules noted that the language in the 1964 and 1968 laws were similar so what applied to jobs discrimination also applied to housing discrimination. HUD specifically objected to two types of housing policies that produced a disparate impact: bank lending policies that refused to give mortgages on homes in minority neighborhoods and government land-use rules that made housing expensive.
The HUD rules didn’t automatically exclude any policy that made housing more expensive. Instead, they included a two-part test to review such rules and policies. First, does the justification for such rules outweigh the cost of potential discrimination against minorities? Second, is there another way to achieve the goal of such rules without discriminating against minorities? For example, rules requiring that people hook up their homes to city sewer systems may make housing more expensive but can be justified by their public health benefits, especially if it can be shown that alternatives (such as septic tanks) are not as effective.
In 2015, disparate impact for housing reached the Supreme Court. It was a curious case because the side that won at the circuit court level was the side that appealed it to the Supreme Court. In the case, the Texas Department of Housing had allocated low-income tax credits to developers in various parts of the state. In doing so, it followed HUD guidelines that encouraged states to use the tax credits as effectively as possible, in other words, where the most housing could be provided for a fixed amount of tax credits. A group called the Inclusive Communities Project challenged the state’s allocation, noting that most of the tax credits went to minority neighborhoods and thus reinforced segregated housing patterns.
At the circuit court level, the state argued that its allocation wasn’t subject to legal review. The court disagreed, but agreed with the state that its allocation of tax credits weren’t discriminatory because the Inclusive Communities Project had failed to identify any other method of allocating tax credits that would have a less discriminatory effect.
Despite having won, the state decided to take the circuit court’s ruling that the allocation was subject to legal review to the Supreme Court. This was probably a dumb idea, and the result was that, in a five-to-four decision, the court not only affirmed the decision that the state’s allocation was subject to legal review but overturned the circuit court’s decision regarding discrimination, saying that disparate impact rules applied to housing as well as jobs. The court did place the burden of proof that policies were discriminatory on those who were challenging those policies.
I eagerly read the Supreme Court’s rulings, especially the minority opinions written by justices Alito and Thomas. Alito argued that applying disparate impact to housing would lead to endless litigation. “No one wants to live in a rat’s nest,” he noted, but cities would be reluctant to pass ordinances requiring landlords to remove rat infestations for fear of litigation. I didn’t buy this because it would be easy to argue that the public health benefits of removing rats outweighed the potential discriminatory effects of such an ordinance. More important, he argued that, since Congress “did not specifically authorize” the use of disparate impact to ban certain policies, it wasn’t appropriate.
Thomas went further and argued that the case against Duke Power was wrongly decided in the first place. It was not enough to show that a policy would harm minorities, he said; opponents had to prove that the intention of the policy was to discriminate against minorities. This was a stronger argument that Alito’s claim about rats. But I also thought it was too easy for people to cloak their discriminatory views with statements about environmental protection and other warm and fuzzy ideas.
In any case, I was elated to find both a federal regulation and a Supreme Court ruling that offered a way to challenge urban-growth boundaries. Using disparate impact, there was no need to persuade a majority of voters to abolish policies that made their homes more valuable. Instead, we would only have to persuade a few judges.
While doing research on this subject, I called the Inclusive Communities Project to interview them about their role in the case. The person I talked to was supposedly the director of the project, yet she didn’t seem that knowledgeable about it, was reluctant to talk about, and it wasn’t even clear that the organization was still active.
At first glance, the project’s web site looks very appealing, but a close look reveals that most of the menu items on the site — “resources,” “mobility assistance program,” “fair housing,” etc. — merely provide links to other organizations. A link to “our new booklet, Inclusive Communities Project: 10 Years and Counting,” is dead. The group’s address listed on the web site just happens to also be the address for the law firm that represented it in the case. All I could conclude was that the organization was created for the sole purpose of acting as a front group for this litigation, which tends to support those who believe that disparate impact is an invitation to lawsuits. While that was somewhat disturbing, I didn’t let it dissuade me from arguing at the November 2015 American Dream Coalition conference that growth-management opponents should use disparate impact to challenge growth boundaries and similar policies.
However, I soon ran into another roadblock. It turned out that the Cato Institute had signed onto an amicus brief written by another think tank arguing to the Supreme Court that disparate impact should not be a test of the legality of various regulations. That was a surprise since you would think that Cato would be interested in less regulation, not more, and disparate impact threatened to dismantle a lot of regulations. In any case, Cato decided it would be hypocritical for them to publish a paper I had written about using disparate impact to abolish growth management.
The great thing about working for Cato is that, while it may not publish everything I write, it doesn’t censor me. It had no problem with me submitting the paper to another think tank so long as Cato’s name was not on the paper. However, it turned out that the other national think tanks I sent the paper to, including the Competitive Enterprise Institute and Reason Foundation, had signed on to the same amicus brief.
As a result, I asked think tanks in Colorado, Hawaii, and Oregon to publish versions of the report customized to their states. I also wrote op-eds on the subject for various publications.
Meanwhile, states and cities with growth-management laws and plans were considering all kinds of policies to make housing affordable except to abolish growth boundaries. Most of the policies actually made housing less affordable.
The first idea was inclusionary zoning, a requirement that builders dedicate a fixed percentage of the homes they build to “affordable housing.” This usually meant renting or selling homes at less than cost. Builders responded by building less and charging more for the market-rate homes they built, thus making housing less affordable for everyone except the lucky few who got an affordable unit. To make sure that people didn’t simply flip the homes, cities also passed rules saying that anyone buying an affordable dwelling would not be allowed to sell it for more than they paid for it plus inflation, effectively making them second-class homeowners.
Cities followed with more crackpot schemes. How about a tax on every new home to pay for the construction of a handful of “affordable” homes? How about a property tax on all existing homes to pay for affordable housing? How about rent control? All of these ideas are guaranteed to make housing less affordable, yet cities in California, Oregon, and Washington were approving them with glee.
Part of the problem is that too few people understand the difference between affordable housing and housing affordability. The first is subsidized homes built for a few very low-income people who otherwise couldn’t afford housing. The second is a measure of the general level of housing prices compared with incomes for everyone in an area. You can’t make housing more affordable for everyone by building a handful of subsidized units, especially if you pay for those subsidized units by increasing the cost of all the other housing in the area.
The next problem I encountered was that some people who claimed to be free-market advocates were buying in to New Urbanist arguments for density. In 2015, the Mercatus Center, a free-market think tank associated with George Mason University, published a paper titled “How Land Use Regulation Undermines Housing Affordability.” The paper claim that single-family zoning and other land-use rules prevented people from living as densely as they would like. That was a surprise to me since all of the research shows that most people want to live in lower densities.
The paper was written by Sanford Ikeda, an economist with the State University of New York, and Emily Washington (more recently Emily Hamilton), with Mercatus. I was invited to debate Washington at a conference that took place, I think, in Atlanta. When she presented her argument in the debate, she claimed that urban areas like San Francisco-Oakland had to grow denser because they were out of land. After all, she noted, the region was bordered by the Pacific Ocean.
I couldn’t believe she could be so ignorant. Yes, the Pacific Ocean is to the west of San Francisco, but to the north is Marin County, which is 84 percent rural open space. To the east are Alameda and Contra Cost counties, which are about 60 percent rural open space. To the south are San Mateo and Santa Clara counties, which are 70 to 75 percent rural open space. At least 70 percent of the rural open space in these counties was private and many of the owners would be happy to develop but couldn’t because the land was outside of urban-growth boundaries.
At the end of the debate, someone asked why not get rid of both growth boundaries and single-family zoning. I would be happy to support that if single-family zoning could be easily replaced by protective covenants, which historically preceded zoning. Nearly all new home construction today that is built in areas that aren’t zoned come with such covenants because developers know that single-family homebuyers want the security of knowing their neighborhood is not going to be flooded with commercial, industrial, or high-density residential development. Effectively, people give up the right to develop their land to higher densities on the condition that their neighbors give up the same right.
In Houston, if you live in a neighborhood that has no covenants, you can petition your neighbors and if 75 percent of them agree you can write such covenants. I would support an end to single-family zoning only if it came with a similar mechanism. But I don’t think it will make any difference for housing affordability either way because single-family zoning doesn’t make housing expensive; growth boundaries do.
When growth boundaries are in place, land prices climb to many times the price of land without boundaries. Abolishing single-family zoning won’t fix that. Multi-level, multifamily housing also costs a lot more to build, per square foot, than one- or two-story single-family homes. You can’t make housing more affordable without reducing land costs and you can’t make it more affordable by building housing that is even more expensive than what you already have.
I also met Sandy Ikeda at the conference where Washington and I debated and was stunned to realize that he had bought into all of the Jane Jacobs hype that had shaped the New Urbanist movement. Jane Jacobs successfully defended her neighborhood from being taken by eminent domain and cleared in the urban renewal era of the 1950s. Since New York City was only allowed to do such urban renewal in slums, she had to prove that her neighborhood wasn’t a slum.
She went further than that, however. She endeared herself to libertarians by arguing that urban planners didn’t know how cities work. But she inspired a new generation of urban planners by claiming that she did know how cities work.
She didn’t seem to understand that she lived in a vestigial remnant of a nineteenth-century city, built at a time when the only methods of urban travel were by horse, streetcar, or on foot. Only the upper classes could afford to stable horses and streetcars were used mainly by the middle class, so working-class people went mainly on foot. That meant they lived in Jacobs’ dense neighborhood not because they wanted to but because they had to in order to be within walking distance of their work.
That all changed when Ford developed the moving assembly line that made cars affordable to the working class. As soon as they got cars, most of the residents of neighborhoods like Jacobs’ moved out, and the people left behind were either poor blacks or recent immigrants who hadn’t yet earned enough to buy a car. The mix of immigrants gave the neighborhood a worldly atmosphere that attracted Jacobs, but the downsides — tiny apartments, lack of privacy, noise, high crime, and congested streets — turned away most people.
From about 1910 to 1990, almost no neighborhoods like Jacobs’ — five-story walk-up apartments with shops on the ground floor — were built anywhere in the United States. When cities started building them again, at the instigation of the New Urbanists, they had to subsidize them because not enough people wanted to live in them if they could find any alternatives. Despite this, the New Urbanists persist in arguing that a nineteenth-century style of living for poor working-class families should be the ideal for everyone in the United States today.
Somehow Ikeda had bought into this and refused to admit that anything Jacobs ever wrote might be wrong. Ikeda, Washington, and others who call themselves “market urbanists” seem to be millennials who have bought into the New Urbanist claim that there is a huge pent-up demand, particularly among millennials, for high-density housing. The reality is that such housing is attractive to young, single people — a group that no longer includes millennials — because it gives them more opportunities to meet other young, single people. But most cities already have sufficient housing for this small group of people, and we don’t need to build more of it.
The claim that there is a pent-up demand for dense housing was led by Arthur Nelson, an Oregon urban planner who currently teaches at the University of Arizona. In 2006, Nelson had a paper in the Journal of the American Planning Association claiming that most people wanted to live in denser neighborhood and that urban planners needed to take the lead in making sure that almost all new development would be dense. The only source he cited for the critical data in his paper was “author’s analysis,” and even the Journal didn’t completely buy it because it followed his article with critical reviews that argued he was misinterpreting the numbers. Yet it led publications such as the Atlantic to claim that the nation would have “22 million surplus large-lot” homes by 2025. We’re only five years away, and I don’t see any signs of a surplus yet.
Nelson’s problem is that he doesn’t understand the fundamental concepts of supply and demand. In 2014, a Minnesota think tank called the Center of the American Experiment asked me to review a report Nelson wrote for the Twin Cities Metropolitan Council, which has the job of projecting future housing demand to make sure it allocated enough land to housing. Nelson claimed that the “demand” for almost all new housing by 2040 would be for multifamily housing. Once again, the sole source he cited for most of his tables and figures was himself.
Demand, as I explained in my review of his report, is not a single number but a relationship between price and quantity. If you ask people, as planners like Nelson do, if they would rather live in a neighborhood where they can walk to grocery stores and coffee shops or live in one where they have to drive everywhere, many will say the walkable neighborhood. But if you add prices, asking if they would rather spend $200,000 on a 2,000-square-foot home with a large private yard within an easy driving distance of several supermarkets eagerly competing for business based on price and selection vs. spending $400,000 for a 1,000-square-foot condo with no private yard within walking distance of a single, high-priced, limited selection grocery store, most people will choose the former.
Bolstered by Nelson’s misconceptions and supported by Market Urbanists, by 2019 a movement had grown among urban planners to abolish single-family zoning in order to allow the construction of supposedly more affordable housing. Minneapolis got rid of single-family zoning in its ordinance. The Oregon legislature banned single-family zoning in major cities. The California legislature attempted to pass several laws that would have overruled local zoning in order to build denser developments, particularly in transit corridors.
As stated above, these steps won’t make housing affordable because they don’t address the underlying artificial land shortages created by growth management and because dense housing a la Jane Jacobs costs so much more per square foot than single-family housing. Portland and Seattle typically spend $300 to $500 a square foot on housing in transit-oriented developments. Such housing is even more expensive in California. Meanwhile, single-family homes in areas with minimal regulation, such as Buckeye, Arizona, typically cost around $100 a square foot.
Unfortunately, the willingness of supposedly free marketeers to support central planners’ density goals killed my hope of ever using disparate impact to abolish growth management. It would be hard enough to prove to the court’s satisfaction that a growth boundary made housing expensive when urban planners piously claimed that high housing prices weren’t their fault. It would be practically impossible when economists who were supposedly on my side were claiming that the real problem was single-family zoning.
The question may be moot as the Trump administration has since revised the rule, making it more difficult to challenge land-use rules that make housing more expensive. I’m sorry to say that I no longer have much hope of winning this battle.
---30---
Background:
2016: Hawaii Land Use Laws Violate Fair Housing Act
2016: Grassroot Report: How to Fix Hawaii's Housing Crisis
2018: How Much Does Over-Regulation Increase Hawaii Housing Costs?