In Washington today, major political efforts are under way to eliminate, or at least drastically shrink, the U.S. Department of Housing and Urban Development (HUD). Since HUD spends around $26 billion per year on a great many different housing and other programs, and since most of the money flows into real estate markets around the nation, HUD's elimination would have notable effects upon money flows into U.S. real estate markets. For this and other even more cogent reasons, I believe it is in the public interest to maintain a separate federal department concerned with housing and urban development - even if the way it performs its functions is radically changed. This article presents my arguments supporting that conclusion.

HUD was created in the late 1960s during the Johnson administration through the consolidation of several previously separate agencies dealing with housing, housing finance and urban mass transit. HUD has since evolved through the addition of many more housing programs to its purview, and the removal of urban mass transit to the Department of Transportation. A recent HUD document officially defines HUD's basic missions as follows: (1) assisting homeless persons and families, (2) reducing the number of distressed public housing units, (3) developing affordable housing and making homeownership a reality for more Americans, (4) reducing racial barriers in housing markets, (5) empowering communities - especially poor ones and (6) creating an environment (within HUD itself) that supports teamwork and organizational excellence. The first four of these missions all deal directly with housing, and the last two do so indirectly. It is therefore clear from these missions that one of HUD's central purposes is improving the housing situation of the American people, especially poor people who cannot afford "decent" quality housing without aid from others.

Why does the nation need a separate federal department to pursue the housing missions described above? For one thing, living in adequate housing is a central element in the well-being of every person and household. Yet housing is the single costliest element in people's basic living standard. Consequently, millions of households with low incomes cannot afford to occupy "decent quality" housing if they must rely solely upon their own resources. Therefore, insofar as the federal government is concerned with its citizens' well-being, it must be concerned with improving the quality and quantity of their housing.

But no other agency within the federal government focuses on housing as part of its central mission. All other federal agencies are concerned with housing only peripherally, if at all. So no others are likely to pay enough attention to housing to improve the chance that most citizens will be well-housed. Only a separate department with that mission as one of its central purposes will have such a focus.

Why not leave this element to state and local governments? A key reason is that providing housing aid to poorly housed citizens amounts to redistributing resources from others to those citizens. But all resource redistribution schemes are more effectively financed (though not necessarily best administered) at the national level. Local governments cannot tax their better-off citizens highly in order to aid their poorer citizens without causing many of the former to move away to nearby communities that have no such redistributive schemes. Even state governments suffer from that competitive disadvantage. This is shown by the tax-induced flight of many firms from high-tax New York to neighboring Connecticut and New Jersey. Only the federal government can impose higher taxes on its wealthier citizens to aid its poor ones without causing many of the former to move elsewhere. (Even that is hard to do politically in times of relatively slow household income growth, such as the past 20 years, because taxpayers then resent such redistribution.)

Furthermore, the cost of a household's financing a home purchase is heavily influenced by mortgage interest rates and mortgage fund availability. Those conditions are determined by national financial conditions and markets, not state or local ones. Therefore, it is necessary to have an agency with nationwide scope and authority to deal with housing financial conditions and availability, rather than having multitudes of parochial state or local agencies try to do so.

Similarly, the racial and ethnic discrimination and segregation that permeate American housing markets cannot be dealt with effectively by purely state or local agencies. Those narrower-jurisdiction agencies need the legal empowerment of national anti-discrimination and anti-segregation laws and policies to support local enforcement efforts. For the same reason that the civil rights movement ultimately required federal action to overcome powerful state and local resistance, policies attacking racial barriers in housing markets also need a strong national agency dedicated to that mission. It must be equipped with powers that transcend locally dominant prejudices.

For all these housing-related reasons, the nation needs a separate federal agency dedicated to the housing missions described above.

In addition, I believe HUD has another central set of missions related more to urban development patterns than to housing in itself. Almost 80% of Americans live in metropolitan areas, each containing one or two central cities and their surrounding suburbs and "exurbs." The basic economic, social and other functions of American society are therefore mainly carried out within these over 330 metropolitan areas. If American society is to function well enough to provide its citizens with "decent" living standards and a satisfactory quality of life, its activities must be located in space in ways that promote the mutual efficiency and effectiveness of those activities. Therefore, the spatial efficiency of how activities function within metropolitan areas is of vital importance to our entire society. If those activities are located in ways that greatly reduce economic efficiency, increase social conflict, or isolate whole groups of citizens from receiving proper schooling or job opportunities, the results can be very harmful to our whole society.

Yet no existing state or local governments, and almost no regional ones, have any direct responsibility to help insure the appropriate spatial efficiency of activities within our metropolitan areas. True, many metropolitan areas lie entirely within a single state. Therefore, in theory, that state's government could take upon itself some responsibility for improving the spatial efficiency of activities within its metropolitan areas. But few states have assumed that responsibility. And many of our largest and most important metropolitan areas lie within two or more states. Examples include New York, Boston, Washington, Chicago, Philadelphia, St. Louis, Memphis, Cincinnati, Charlotte, Louisville, Kansas City and Portland, Ore. These metropolitan areas contain 22.4% of the nation's total population. No one state can have jurisdiction over their overall efficiency.

Why do we need to worry about the spatial efficiency of our metropolitan areas? Because current trends are seriously undermining that efficiency in many of our largest metropolitan areas in ways that local governments cannot halt. The fragmentalized government structure in nearly all U.S. metropolitan areas makes it impossible to develop coherent, consistent and effective policies toward several key growth-related problems. These include air pollution and other environmental degradation, traffic congestion, provision of sufficient new infrastructures to accommodate growth, selection of sites for regionally necessary facilities like airports that have negative impacts upon their immediate surroundings, and the absorption of excessive open space.

All of these problems are essentially regional or metropolitan-areawide in nature. But they are all related to the patterns of land uses emerging from ongoing urban development controlled by fragmentalized local governments. That is why HUD's spatial efficiency mission - which focuses on land use patterns - is relevant to problems like air pollution that technically fall under the jurisdiction of other agencies, such as the Environmental Protection Administration. At present, power over land uses and other regulations that would need to be coordinated to cope with such regional problems is divided among hundreds of local governments. Moreover, each such government is competing with all the others to capture activities that generate net local tax revenues, and to reject activities that generate net local fiscal losses. This ferocious fiscal competition makes development of a single coordinated land-use policy among local governments almost impossible.

The result is an immense spreading of people and activities across the landscape that increases the need - and cost - of travel within each metropolitan area. In the 20 years from 1970 to 1990, the Chicago metropolitan area gained only 4% in population, but increased over 40% in urbanized area. Similar "de-densification" occurred in dozens of other major metropolitan areas. It is no accident that total vehicle miles traveled in the United States rose 37% from 1983 to 1990, compared to a rise of only 4% in population. More and more time and resources are now being consumed by travel, often waiting in creeping lines of congested traffic. This is a sign of decreased overall spatial efficiency.

Much more important in the long run is the deteriorating fiscal and social condition of many major central cities and older close-in suburbs. As rapid growth on the fringes of each metropolitan area draws the most prosperous and viable households and firms out of central cities and older suburbs, those communities are left with declining per-capita tax bases and increasing concentrations of poor households with high public service needs. In 1990, the nation's central cities contained only 31% of the entire U.S. population, but 42% of all poor persons. Within the central cities combined, about 19% of all residents were poor, compared to about 8% in suburbs as a whole, and 14% in the entire nation. This concentration of poverty is greatly influenced by the exclusionary policies of many suburban local governments. They erect zoning, building-code and other regulatory barriers to the creation of any added low-cost housing within their boundaries in order to keep "the wrong kind of people" out of their neighborhoods. That forces high fractions of all poor residents to live in central cities and older suburbs that cannot engage in such exclusionary behavior. This situation loads local governments in older central cities and suburbs with increasingly unbearable fiscal burdens and extensive areas of deteriorating or abandoned neighborhoods.

Furthermore, central cities and older suburbs contain a majority of members of two large minority groups that will play ever-more-important roles in the nation's future. These are African-Americans and Hispanics. Together, the 58.3 million persons in these two groups comprised 22.2% of the nation's total population in 1995. But they will make up 56.2% of the projected 58.3 million increase in population between 1995 and 2020. So in 2020, just 25 years from today, these two groups will comprise 28.5% of our total population. At present, the average household incomes of these groups are between 30% and 40% below those of non-Hispanic whites, and the educational attainment of the former two groups is similarly below that of the last. Unless this condition is changed in the future, most of the nation's future population growth will occur among its poorest and least skilled major minority groups. To improve their incomes and skills, American society needs to improve the quality of education and the job opportunities available in the central cities and older suburbs where most of these groups live. But that will be extremely difficult if continued future peripheral growth of our metropolitan areas keeps on draining economically strong households and businesses out of those central cities and older suburbs.

Right now, no government or private groups anywhere within our society - except HUD - are officially entrusted with the mission of thinking about, or doing something about, these adverse trends. Even HUD has not focused much of its past attention on developing policy responses to such trends. Under the Republican administrations in power from 1981 through 1993, HUD was deliberately prevented from paying much attention to these trends because those administrations drew most of their political support from suburban voters. So they considered central cities as mainly "enemy territory" politically. They were not terribly concerned about whether rapid peripheral growth weakened central cities and older suburbs.

Since the Clinton administration took office, HUD secretary Henry Cisneros has talked a lot about responding to these metropolitan-area trends. But in reality, HUD has not refocused its missions to include strong emphasis on improving the spatial efficiency of metropolitan areas, as its above-quoted mission statements show.

Nevertheless, American society needs some highly visible, politically active public focal point for thinking about and improving the spatial efficiency of our metropolitan areas. An important aspect of such improvement would be creating incentives for the development of some type of regional decision-making within each metropolitan area. That does not mean metropolitan government. Rather, it means at least some type of planning for functional activities (such as housing, health care and transportation) that takes account of the needs of each region considered as a whole, not just each locality considered apart from all the others. An example is the metropolitan planning requirement concerning highways and mass transit adopted by Congress in the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA, often referred to as "Iced Tea"). A similar regional planning requirement might eventually be connected to the housing and other programs under HUD's jurisdiction, as a means of improving the spatial efficiency of our metropolitan areas.

HUD is the logical body to take on the challenges associated with such improvement because its official area of jurisdiction includes both housing and urban development. HUD needs to provide leadership in focusing the nation's attention on the need for regionwide approaches and policies aimed at coping with the regionwide problems described above. If HUD does not take on this mission in a powerful way, the adverse trends described above will further weaken many older central cities and suburbs and cause calamitous decay in large parts of our inner-city neighborhoods.

This conclusion does not imply that HUD alone can solve all the problems associated with declining cities and deteriorating inner-city neighborhoods. Even at best, it could not do so without help from many other parts of society, both public and private. But if HUD does not exist to at least focus public attention on these issues, few - if any - remedial policies will be directed at them.

American society needs to retain the U.S. Department of Housing and Urban Development as a separate federal agency because it needs someone to perform two crucial missions now entrusted to HUD. One is upgrading the housing of the population; the other is improving the spatial efficiency of our metropolitan areas. Both these missions are vital to the future quality of life in our society. But no other agencies - public or private, federal, state or local - are likely to focus on those missions in the future.

This conclusion does not mean that HUD should remain in its present form, or do business "in the same old way." Many of the radical changes in HUD and its operations now being proposed by HUD's leaders themselves and by Congress are certainly worthy of serious consideration. But I believe completely eliminating this agency, or weakening its ability to pursue the two central purposes discussed above, would be a disservice to the nation.

Anthony H. Downs is senior fellow at the Brookings Institution, Washington, D.C. The views expressed here are those of the author and not necessarily those of officers, trustees or other staff members of the Brookings Institution.