On the Record

Housing Policies in the New Millennium

Anthony Downs

My daunting assignment is to present an overview of what our housing policies ought to be for the New Millennium. I cannot cover all policies, but only those I feel are most crucial.

I will begin with certain key background factors that underlie my subsequent analysis.

1. First, the American housing production and urban growth processes already provide excellent shelter opportunities for most households with middle- and upper-incomes, except in a few very high-cost areas like Northern California. Our housing markets are working well for most households with money.

2. Second, the most widespread and serious U.S. housing problem lies in the discrepancy between the low incomes of many poor households and minimal costs of “decent” housing as judged by middle-class standards. Household incomes go right down to almost zero, but the minimal costs of “decent” shelter level off at some amount needed to cover basic space, plumbing, kitchen, utility, and heating needs.

  • In 1997, 42.7% of all renter households had incomes below 50% of their area medians, 27% had incomes below 30% of those medians about equivalent to the poverty level and 15% had incomes below 20% of those medians.
  • The 1997 fair market rent for a household of 4 was $720 per month (derived from 15 large MSAs). That is equivalent to 90 cents per square foot per month if such an apartment contained 800 square feet. If that is equivalent to 30% of income, then the minimum income needed to avoid “housing poverty” was $28,800. That is much higher than the 1997 food-based poverty level of $16,400. In 1997, an estimated 57.3% of all renter households had incomes below $28,800, so they were “housing poor.”
  • If the 27% of all renter households with incomes below $12,000 spent 30% of their incomes on rent, and if rent was 90 cents per square foot per month as derived from fair market rents, then such households could only afford to occupy 333 square feet. The 15% of all renters with incomes below 20% of the area medians could afford to occupy only 222 square feet. These unit sizes are far below the sizes considered “decent” and required by nearly all building codes.
  • This shows that millions of American families cannot afford ANY newly-built dwellings without spending over 30% of their incomes for housing. Hence they are excluded from living in all new-growth areas unless subsidized, and we do not subsidize anywhere near all households that are eligible. The only way you can occupy only 222 square feet is by having 3 households in one unit!
  • Clearly, low incomes comprise by far our largest “housing problem.” The best way to attack that problem is by raising incomes of the poor. That means through programs beyond HUD’s jurisdiction, except for vouchers, which in essence increase the money incomes of families who receive them.

3. A third critical factor is that the population of the U.S. will rise by 48 million from 2000 to 2020, and housing should be created that those people can afford to occupy. Over one-third of these added people will be immigrants from abroad or their children. Many will be very poor; so they will be unable to afford new units built to our high-quality standards without subsidies. Yet poor immigrants are not deterred from entering our cities by high housing prices that force them to double and triple up, because overcrowded housing is superior to what they experience in their home areas.

  • Recent economic prosperity, combined with some negative conditions produced by growth in high-growth areas, have intensified anti-growth feelings in many communities, especially new suburbs. Although individual localities can slow growth within their own borders, there is no way for individual regions or the nation as a whole to stop or greatly slow their future growth. Yet concern with how to cope with this growth is arising all across the nation and will greatly effect the environments in which future housing must be created.
  • As a result, effective housing policy must be concerned with the entire growth process, not just the building of new housing as though it were separate from the growth process. Housing policies must be integrated with growth policies.

4. A fourth crucial factor in housing policy is that most middle- and upper-income households of all ethnic groups do not want to live in neighborhoods containing any sizable number or percentage of poor people. This is especially true of households with school-aged children. Therefore, non-poor households withdraw from areas where many poor live and erect barriers to the subsequent entry of the poor into their own neighborhoods. This behavior is the foundation for the socio-economic hierarchy of neighborhoods found in every U.S. metropolitan area.

  • Such behavior is similar to the unwillingness of most whites to live in areas where more than 25 to 33% of the residents are African Americans no matter what their income levels. This underlies continuing racial segregation. Similar but less intensive feelings are held by whites about other minorities too.
  • The non-poor exclude the poor not through purely market forces, but through local zoning and other regulations that prevent construction of affordable units. This behavior is rooted in the strong desire of homeowning households to protect and increase their housing values, since housing is their greatest financial asset. Hence suburban governments, which are almost always dominated by homeowning voters, tend to adopt parochial policies that aim at benefiting their own residents, without regard to the impacts upon anyone else.
  • Exclusionary zoning is reinforced by the desire of local officials and citizens to minimize taxes by blocking land uses that generate more local expenses than revenues. This means housing, especially housing for people with children. So truly affordable housing is regarded as a fiscal anathema by suburban governments, and they all try to shunt it off onto other areas. Yet our economy cannot run without many low-wage workers who cannot afford “decent” units.
  • These exclusionary motives are emphasized during periods of prosperity when residents have the luxury of not worrying about their jobs or incomes. Then they can worry about elements like congestion and life style. Another result of prosperity has been rising prices and rents that hurt the poor. From 1990-1999, median home prices in the largest 21 metro areas rose an average of 33.9%.

5. A central, if unintended, consequence of such exclusionary behavior is significant concentration of the poorest households together in high-poverty neighborhoods in central cities and older suburbs – especially poor minority households, who are doubly excluded. Yet concentrating many very poor people together produces adverse neighborhood environments that reduce the life-chances of people there, compared to environments with much more economically diverse populations. I am convinced, and so are most big-city mayors and many HUD officials, that we cannot improve the quality of life for the very poor without reducing big poverty clusters.

  • Such poverty concentrations have been worsened by federal and other government policies that focused most housing assistance and incentives on the very poorest households, in the poorest areas. This was done in the name of humanely aiding those who needed it most. But this policy has failed because it increased concentration of the poorest households together, as in high-rise public housing projects – thereby creating socially destructive environments.
  • It is time to re-align incentives created by federal policies so they encourage greater income diversity, even if that means giving public aid to people who are not the very poorest. For example, more points for low income housing tax credit projects should be given to those with moderate percentages of very poor residents, rather than to those with the highest percentages, as is done now.

6. The preceding points show the dominant American housing development process causes a progressive abandonment of parts of many large older cities through growth at the suburban fringe, aggravating urban decline. Many observers blame such urban decline on suburban sprawl, but that conclusion is not accurate.

  • Major suburban growth in U.S. metropolitan areas was, and still is, inevitable because of population increases in those areas, plus rising real incomes that generate desires for low-density living. “Sprawl” is just one possible form of suburban growth marked by very low densities, leap-frog new development, unlimited outward extension, and automobile dominance. In the U.S., sprawl has been so dominant that most people wrongly think it is the same as growth.
  • I have conducted extensive regression analyses that show almost no connection between the basic traits of sprawl I mentioned and urban decline – measured as either city population change from 1980-1990 or by an index of decline indicators like high crime rates, high poverty rates, etc. This surprised me.
  • Further analysis convinced me that it is not the low-density aspects of suburban growth – those which comprise sprawl – that lead to urban decline. Rather, it is the operation of six other basic characteristics of our development process.

    (1) We require all new housing to meet high quality standards that poor people cannot afford without subsidies, (2) we do not provide most of the poor with such subsidies, so few can live in new-growth areas, (3) we encourage exclusionary suburban zoning, (4) we engage in widespread racial segregation in almost all housing markets, (5) we maintain major obstacles to redevelopment of older core areas, and (6) we have social values that encourage households to move to higher-status neighborhoods when their incomes rise. These factors – not low density – are the reasons why our growth causes poverty concentrations.

  • This is an important conclusion because it implies that continued growth would still lead to concentrated poverty even if we shifted from sprawl to much more compact forms of growth – unless we altered those six basic characteristics. And concentrated poverty would still generate withdrawal of the middle class to the suburbs, which aggravates urban decline. In fact, the poor themselves often move out of high-poverty areas once they get higher incomes.


Now let us turn to conclusions about housing policy from these background factors.

1. My first conclusion is that by far the most important housing policies are being set by local governments – not the federal government. True, the federal government influences the financial climate that affects housing affordability and production. But local governments set the rules of housing quality and density that really determine the amount of housing built and where different income groups will live. And those local policies are mainly determined by the parochial and exclusionary perspectives of suburban homeowners, as described earlier. This perpetuates poverty concentrations.

  • Therefore, the most important thing the federal government can do to improve housing opportunities for the poor is to exert influence on local governments to be less parochial in deciding what types of housing can be built where. This would require HUD and Congress to create incentives for local and state governments to modify their currently exclusionary behavior towards housing.
  • Examples have already been created by EPA and DOT – the latter in the form of Metropolitan Planning Organizations. Both require each metropolitan area to establish a regional planning agency that considers area-wide plans before the federal government will provide any funds to any governments in the region. HUD could make its financial aids similar contingent on regional planning.
  • Possible goals would be (1) to require each metropolitan region to establish “fair share” allocations of low-cost housing among its communities, (2) to allow owners of single-family homes of a certain size to create accessory apartments in their homes even if the local government does not permit it, (3) to require every community to zone some land for multi-family housing, and (4) to expand use of vouchers to encourage the “moving to opportunity” program. At the very least, HUD could make planning grants to local governments within regions where all such governments agreed to develop a voluntary regional plan.
  • Because suburban homeowners form a majority of voters in the U.S., Congress will be reluctant to permit such a policy. Yet HUD should press to get one because HUD cannot reduce inner-city decline without affecting the growth process as a whole, which means affecting local suburban growth policies.

2. A second key goal of housing policy should be to deconcentrate existing high-poverty enclaves in two ways. One is by using vouchers to give households voluntary opportunities to move to middle-income neighborhoods. The other is to encourage greater income diversity in poor neighborhoods, including in public housing. These are already present goals of HUD policy that should be promoted more strongly.

  • This approach contradicts the desire of middle-and-upper-income households to isolate themselves from the poor. Since the non-poor are much more numerous and more powerful than the poor, society has resisted nearly all attempts to deconcentrate the poor by integrating them into more affluent communities.
  • I have long believed in my heart that we in the middle-and-upper-income groups cannot ultimately upgrade our poorer brothers and sisters at arm’s length, but must somehow share our neighborhoods with them if we want to create truly equal opportunities in our society. But this is distinctly a minority viewpoint strongly resisted by the vast majority of non-poor Americans, and by many poor.

3. My third key goal is striking a better balance between aid to homeowners “mainly as tax benefits” and aid to poor renters by greatly increasing the latter. In the 1990s, HUD placed much more emphasis on promoting home ownership than aiding low-income renters, even though the latter have more serious housing problems. Homeownership rates have risen notably, and that is good. But today builders are putting up many houses costing $2 to $3 million. It is ludicrously unjust to give those buyers tax benefits of as much as $58,500 per year while not providing any assistance to most eligible low-income renters. We do not need to encourage more mansions.

  • Changing interest and property tax deductions to tax credits would be much fairer without eliminating big tax benefits to homeownership. This would either shift more benefits to less affluent homeowners or save money. The money thus saved could be allocated to greater assistance to low-income renters.
  • More federal money should be spent on aiding low-income renters – mainly as vouchers and as aid to the rehabilitation of older units. HUD should also consider setting higher fair market rents in suburban markets with high housing prices to enable Section 8 households to live there; the present single-rent policy in all parts of a metropolitan area further encourages concentration of the poor.

4. Because it is politically impossible to focus all federal aids on deconcentrating existing poverty enclaves, we should also invest notable resources in improving conditions there. We will be unable to deconcentrate even a majority of existing high-poverty areas within any short time, so we should not ignore those still there.

  • However, we should recognize that efforts to up-grade such areas are not likely to be effective, unless many non-poor residents can be attracted to live there. Billions of dollars have been spent to encourage Community Development Corporations and empowerment zones to upgrade their entire neighborhoods. But few have succeeded, though they have built many new housing units. It is time to stop wasting such aid by focusing more of it on encouraging diversity.
  • But that means devoting public funds to providing incentives to persons other than the very poorest. This is a politically controversial but necessary strategy that most city mayors are now promoting because they realize it is necessary.

5. Another goal of federal housing policy should be closer integration of land-use planning, transportation planning, and environmental planning since each type of planning is heavily influenced by the others in a process of mutual causation. In theory, the Departments of Housing and Urban Development and Transportation should be merged into one federal agency. It could be called either THUD because of its heavily influence, or maybe HUDAT just for fun. That department should then require each metropolitan area to develop some type of coordinated affordable housing and ground transportation plan as a prerequisite to receiving federal funding.

  • Transportation planning is generally not done with full recognition of its impacts upon land use, and vice versa. The U.S. and state DOTs do not realize they are not just building roads but creating the future skeletons of entire metro areas.
  • However, it would not be possible to combine these departments without major changes in the committee structures of Congress. Yet Congress is far more resistant to reform than any other institution in America. And trying to closely integrate the actions of separate federal agencies is usually a vain exercise.
  • As a starter, HUD should clean up its own act a bit. For example, HUD’s Economic Development Initiative financed a hotel in Huntington Beach without inquiring whether the low-wage workers to be hired could afford to live there, or would have to drive miles to work. No such grants should be approved without examining their impacts upon local housing and transportation requirements.

6. A final policy recommendation is that the federal government should continue to promote economic policies that keep interest rates low and labor markets tight. Low rates make it easier to build more housing, and tight labor markets raise the incomes of many low-income workers. The long run of nine years of prosperity has done more for the economies of our cities than any other federal policies of any types.

As we consider these policy recommendations, one enormous obstacle to achieving them emerges. It is the fact that present institutional arrangements in housing markets and growth processes favor the middle-and-upper-income majority at the expense of the low-income minority – especially low-income ethnic minorities. For example, concentration of poverty in older core-areas permits more affluent households to live in neighborhoods mostly free from problems associated with poverty. That is what they like and want.

1. In short, democracy is working, because the majority has created and sustained arrangements which benefit it and which its members therefore do not want to change. For decades, it has proven extremely difficult to persuade that majority to alter those institutional arrangements which benefit it by imposing high costs upon the poor. Appealing to their sympathy for the poor has had only modest impacts.

2. But now the suburban majority itself is beginning to complain about some of the costs of the growth processes it has generated, especially traffic congestion. The challenge to those of us trying to implement the policies I have described is to use these feelings of dissatisfaction to achieve institutional changes – such as at least some regional planning mechanisms – that might help remove some of the unfairness and ineffectiveness of our present housing and development processes.

3. Such appeals to the self-interest of the majority should include pointing out two consequences of the failure to up-grade the skills and incomes of people now living in concentrated poverty areas or to permit more of them to move to the suburbs.

  • First, such failure will weaken the ability of those people to buy the suburban homes of the current white baby-boomers when they or their children want to sell. The potential market will then consist heavily of minority households now living in cities. But those possible buyers cannot maintain future home values of the present residents if they have little money.
  • Second, if low-wage workers essential to both business and residential areas have to live far from where the jobs are located because there is no affordable housing nearby, then both traffic congestion and air pollution will rise, reducing the quality of life for the affluent households who can afford to live there.

In conclusion, I hope we will recognize that future housing policies and future metropolitan growth policies are inextricably intertwined, and cannot be treated separately. If so, we can perhaps use the growing awareness among even middle-and-upper-income households that our growth policies must be changed to achieve major improvements in our housing policies in the new millennium.

HUD Conference on Housing Policies for the Millennium, October 3, 2000
Washington D.C. – U.S. Department of Housing and Urban Development