occupants on the stream of housing services received from renter and owner occupied subsidized housing. Subsidized housing is a significant type of "in-kind" income for which the recipient's valuation is needed as an important input to the determination of an optimal mix of cash and in-kind aid. At the present time, little is known about the recipient's valuation of assisted housing. In order to provide data in this area, HUD should work closely with DHEW's developmental effort related to the Survey of Income and Program Participation.

Data on the inventory of mobile homes used or intended for use as residences is another area in which available data are deficient for assessing the adequacy of the housing stock relative to need. Improved data on the housing role played by mobile homes is of great importance because mobile homes are an essential source of new private housing for families in the lowest third of the income distribution. HUD should play a major role in developing a followon survey to the 1980 census to obtain detailed characteristics of mobile homes and their occupants.

In order to monitor changes in the housing stock relative to need, policymakers need to go beyond a purely descriptive approach to current data about the process through which change takes place. Change in the housing stock takes place through the addition, loss, division or combination of housing units. It is also frequently associated with changes in the population living in units basically identifiable as the same units over time. Data concerning the process of change is provided on the national level by the "components of inventory change" of the Annual Housing Survey. The "components" fill what was previously a serious gap in current data by making it possible to follow annual trends in the location, magnitude, and characteristics of housing losses and additions, and to evaluate requirements for new units in relation to the overall housing stock. The process of change in the population associated with the same units can be analyzed using data based on reinterviews of a fixed panel of the Annual Housing Survey, thus providing data relevant to the analysis of factors associated with the maintenance or deterioration of housing over time.

Allocation of Block Grant Funds

Title I of the Housing and Community Development Act of 1974 provides for the distribution of significant funds to local areas including metropolitan cities and urban counties according to a three-factor formula involving population, extent of housing overcrowding, and extent of poverty. In order to maintain the comparability of

the data among localities, it is necessary to employ data obtained from a common source. Current population estimates are used by HUD to update the population factor; however, the most recent source of comparable data for the remaining factors is the 1970 decennial census. As the age of the census data increases, the data are becoming increasingly less descriptive of the actual distribution of these factors.

Title II of the 1974 act imposes two additional data requirements as a basis for determining the allocation of housing assistance to local governments. Housing vacancies and substandard housing are specifically mentioned in addition to population, overcrowding and poverty, which are used for title I. For small areas, the additional information is available only from the 1970 decennial census. Moreover, the usefulness of vacancy data is relatively shortlived and subject to rapid change due to changing market and seasonal conditions. In order to minimize the inequities resulting from the allocation of funds on the basis of decennial censuses, the mid-decade census program should be used to provide benchmark data for updating the data employed in formula grants.

Housing Market Analysis

Concerns with both the supply and demand for housing are included in this topic. HUD and other agencies have a need to monitor changes in both sides of the market in order to develop policies for the national as well as the local housing market that will help assure a reasonable balance. On the supply side of the market, data are needed on interest rates, the volume and terms of transactions in existing units, costs of housing production and construction activity. On the demand side, data requirements include demographic estimates and projections, detailed estimates of the economic status of families, characteristics of market transactions and current intentions with respect to moving. Since much of the data of importance for analyzing the balance between the supply and demand for housing is covered in other chapters, only those aspects not covered elsewhere will be reviewed here.

At the broadest level of analysis, currently available data generally provide a moderately adequate overview with respect to major trends in housing supply and demand. Demographic estimates and projections indicate the trend of family formation and increases in the population; series on family income and factor costs indicate short- and long-term trends in the ability to pay for housing; and data series on construction starts and market absorption pinpoint current construction activity. In addition, the Annual Housing Survey provides

information on the major components of change in the housing stock, vacancy rates, and trends in the balance between the housing stock and the population to be housed.

While data are generally useful for an overall assessment of trends, important specific weaknesses exist in a number of areas. The housing situation of particular population groups of concern such as the elderly and welfare recipients is known only in the most general terms due to the difficulty of sampling groups that form a fraction of the overall population. Little is known concerning the market for housing of special types, such as housing limited to Occupancy by the elderly, or seasonal housing. Information concerning mobile homes and their occupants remains fragmentary. Another area of data weakness relates to the charateristics of transactions involving the sales of existing homes, which is the way many families accommodate their need for a change in housing.

In spite of the areas mentioned above in which the lack of data hampers analyses of situations with broad national policy implication, the data on a national basis are incomparably better than for local areas. Federal agencies require reasonably current housing related data for local areas in order to assist in reviewing the viability of local government and private plans for building or rehabilitating housing with direct Federal subsidies, Federal mortgage insurance or guarantees, or with loans from federally regulated financial institutions.

Every 10 years the decennial census provides a rich source of comparable data for all localities. Beginning in 1985 much local data should become available as part of the quinquennial census program. As mentioned earlier, in 60 Standard Metropolitan Statistical Areas (SMSA's) data are available every 3 years from the Annual Housing Survey. For markets other than the 60 SMSA's housing permits constitute the principal data series developed on a comparable basis for all active housing markets. For current data in the balance of the metropolitan areas, only local data of variable quality are available. The middecade census provides an opportunity to replace the present 3-year cycle of data for 60 SMSA's with a 5year cycle for all SMSA's.

The need for improved current data for market analysis at both the national and local level is particularly evident with respect to (1) vacancy data, (2) existing home prices, and (3) the volume of transactions and characteristics of buyers and renters of existing units.

Data giving the characteristics of vacant units available for rent or sale are among the most im

portant indicators of the operation of local housing markets. In determining the ability of local markets to absorb new units, Federal agency field staffs are required to approximate vacancy rates from sources such as utility meter hookups and postal vacancy

surveys.

Information on existing home transactions, including selling prices, and rental transactions in existing units is fragmentary. The only available current monthly series is derived from a sample of real estate brokers and provides data for the United States and four regions on the number of transactions and sales prices. Owner and renter transactions in existing housing are several times more numerous than new home purchases and thus should provide a sensitive indicator of the extent of choice for all segments of the housing market.

The need for improved data for housing market analysis should be met through more frequent updating of data, which can be used to benchmark local data series through selective improvements in current data series, and through additions to survey programs focusing on subgroups of the population. The requirement for improved benchmark data for local areas should be met by the inclusion in a middecade census of basic housing data such as occupancy status, value, rent, and housing expense. To provide current vacancy data on a local basis consideration should be given to the use of existing commercial directory services and encouragement of the standardization of local utility data and local occupancy permits. In special situations of unusually active local markets consideration should be given to one-time surveys conducted by the Bureau of the Census.

The requirement for separate tabulations for SMSA's as part of the Annual Housing Survey should be reevaluated with respect to the mid-decade census program and the benchmarking of local data series.

The need for data on existing housing transactions should be met through the initiation of a HUD sponsored survey conducted by the Bureau of the Census and providing data for selected individual active market areas as well as for the Nation. Comparable data should be provided for existing mobile home transactions. Special-purpose surveys such as the periodic surveys of the aged and disabled should be more fully utilized as possible sources on the housing situation of special groups.

Community Development

Federal requirements for data concerning community development are considerably less well

developed than for housing. At this time one area of statistics relating to community development has been identified for investigation. The inclusion of a requirement in the International Investment Survey Act of 1976 that the "the President shall conduct a study of the feasibility of establishing a system to monitor foreign direct investment...including the feasibility of establishing a nationwide multipurpose land data system..." provides a mandate for the examination of the extent of Federal needs for land use data. The need for urban and rural land use on a comparable basis derives from concerns with environmental quality, energy conservation, transportation, the provision of community services, and the preservation of farm land. Currently available land use data on a comparable basis across jurisdictions were developed for limited purposes such as tax assessment or single-time transportation studies and are not comparable from area to area. Sources of land use data include local governments, special transportation studies, studies of the Corps of Engineers, maps of the National Geodetic Survey and the Geological Survey, studies conducted by the Department of Agriculture, data collected in the housing census and the shelter survey conducted by the Defense Civil Preparedness Agency.

While of unknown extent, existing duplication of data gathering efforts for direct Federal purposes or local purposes partly financed through Federal funds is probably considerable. The outlays of the Defense Civil Preparedness Agency are illustrative. It is estimated that in the mid 1970's $1 1/2-$2 million was spent annually in updating the inventory of shelters. Since the buildings in which shelters are found are located on parcels of land, it is conceivable that an inventory of shelters could be accommodated as part of a general land use inventory. Although quantitative data are not available, it is safe to say that an even more important use in terms of dollar outlays is the employment of land use data in land use planning, especially in transportation planning.

A program which would provide a significant portion of the urban and rural land use data needed for the multiplicity of uses would be a large undertaking involving extensive Federal and local cooperation. Because of this, a developmental program is needed which will experiment with alternative frameworks for what could become a multipurpose vehicle having both local and national uses. Consideration should also be given to a national survey of land use to provide data relevant to national energy, environment, housing and community development policy formulation.

The Department of Housing and Urban Development should establish a developmental effort to

determine the optimal means of meeting requirements for urban land use data including the need for a multipurpose land data system, the use of special purpose studies and the possible establishment of a Federal/State/local cooperative system. The Department of Agriculture should play a similar role with respect to rural land use data as a natural outgrowth of its role as the lead agency for conducting a feasibility study under the provisions of section 4(d) of the International Investment Survey Act of 1976 (P.L. 94-472). The Bureau of the Census, as the most likely data collection agency, should play an active role in the developmental effort. Consideration should also be given to the initiation of a national survey of land use to provide data relevant to national energy, environment, housing and community development policy formulation and monitoring foreign investment in land during the period in which a cooperative statistical program is being established.

Coordination

Two areas stand out as being in need of more coordination to ensure that the statistical needs outlined in this chapter are addressed. The first need is for the initiation of a sustained effort to improve the comparability of locally generated data on housing and community development. A second need is for an agency to exert broad leadership in the development of housing and community development statistics needed for broad analytical purposes. Each of these areas is discussed in turn.

In the discussion of Federal requirements for local area data, a recommendation was made that local area data be relied on as much as possible to provide updates of data produced through the decennial and mid-decade programs. Implicit in this recommendation is the notion that systematic exploitation of data for statistical purposes would be possible for data that are produced as a part of the operation of local governments. As indicated in the chapter on Federal/State/local Cooperative Systems of Data Collection, a significant degree of such cooperation exists. However, relatively little in the way of Federal/local cooperative statistical efforts is to be found in the area of housing and community development statistics. Differences in definitions, methods of data collection, the extent of quality control and differences in the geographic boundaries of administrative subdivisions generally preclude meaningful comparisons between communities, or even within the same community over time.

Difficult as the problem may be of introducing a minimal degree of standardization into selected local statistics of housing and community development,

the potential return to serious efforts should be significant. At this stage what is needed is the designation of a lead agency to assess the potential for a cooperative Federal/local statistical program. The lead agency would need to identify potential users and their data requirements, facilitate the determination of priorities and propose approaches to continuing coordination.

Currently, no one agency attempts to oversee Federal and local cooperation in this subject area. Private, local, and State outlays dwarf Federal outlays for housing and community development. At the same time, HUD has been assigned the role of reporting to the Congress on progress in meeting national housing goals, including its environment. HUD is therefore the most appropriate agency to take the lead in exploring the benefits and costs of one or more Federal/local cooperative statistical programs and appropriate mechanisms for organizing the effort.

The office of the Assistant Secretary for Policy Development and Research (HUD) should be assigned the lead responsibility for exploring the feasibility of initiating Federal/local cooperation in the standardization of local area data on housing and community development. The Bureau of the Census should play a major role in this effort.

Assuming the preliminary assessment to be favorable, experience with the development of cooperative systems in the fields of education and health indicates that any continuing systems would evolve slowly. Important first steps in the process would involve the identification of Federal requirements for housing and community development statistics for local areas and discussing joint interests in local area data with appropriate government and professional groups.

There is currently no single organizational unit which has been given the assignment and technical skills necessary to enable it to take a broad role in fostering improvements in Federal housing and community development statistics. In the absence of such a focal unit, advances in statistics have been spotty, and there have been significant delays in addressing emerging needs. For example, there are no data on the characteristics of mortgage transactions underwritten by private mortgage insurers (PMI's) despite the fact that PMI's have replaced HUD as the principal source of residential mortgage insurance. Creation of a Federal statistical center for housing and community development would place these fields par with the fields of health and education which have established centers that are in a position to

on

extend considerable leadership within their respective areas. The Department of Housing and Urban Development, as the agency with the most immediate Federal involvement, would be the logical site for a statistical center for housing and community development statistics.

A Federal statistical center should be established within HUD to exert leadership in the area of housing and community development. A study should be conducted to recommend specific functions, lead agency responsibilities, and organizational location.

Programs To Be Discontinued

Potential savings in statistical programs might be made in the outlays made by State and local governments for the collection of data, much of which is financed out of Federal grants. For example, a significant proportion of the approximately $60 million HUD annual outlays for the support of local planning is expended for the purpose of providing updated data on land use, housing conditions and the physical plant of local governments. While the cost of the supported data programs is not available, it is estimated that some 5% of the annual Federal outlays in this program area is for data collection and processing. To the extent comparable data are made available for local use through programs such as a mid-decade census or through estimation procedures, some portion of the current outlays could be applied to other activities.

Summary of Recommendations

Recommendations for new programs which apply to several headings shown below are listed only under the first heading where they appear.

Housing Need

1. The HUD Office of the Assistant Secretary for Policy Development and Research should coordinate a continuing interagency effort to develop measures of housing and neighborhood adequacy which have a significant relationship to Federal policy issues. This effort should include investigation of the relationships between housing and neighborhood characteristics and a variety of social and economic phenomena which are included in the HEW Survey of Income and Program Participation.

2. Improved estimating procedures for data on income and wealth which are developed in the DHEW Survey of Income and Program Participation should be introduced into the

Annual Housing Survey as soon as possible after development.

3. HUD should play a major role in developing a follow-on survey to the 1980 Census to obtain detailed characteristics of mobile homes and their occupants.

Housing Market Analysis

4. The requirement for separate tabulations for SMSA's as part of the Annual Housing Survey should be reevaluated with respect to the middecade census program and the benchmarking of local data series.

5. The need for data on existing homes. transactions should be met through the initiation of a HUD sponsored survey conducted by the Bureau of the Census and providing data for selected active market areas as well as for the Nation. Comparable data should be provided separately for existing mobile home transactions. Special-purpose surveys such as the periodic surveys of the aged and disabled should be more fully utilized as possible sources on the housing situation of special groups.

Community Development

6. The Department of Housing and Urban Development should establish a developmental effort to determine the optimal means of meeting requirements for urban land use data including the need for a multipurpose land data system, the use of special purpose studies and the possible establishment of a

Federal/State/local cooperative system. The Department of Agriculture should play a similar role with respect to rural land use data as a natural outgrowth of its role as the lead agency for conducting a feasibiliy study under the provisions of section 4(d) of the International Investment Survey Act of 1976 (P.L. 94-472). The Bureau of the Census, as the most likely data collection agency, should play an active role in the effort. Consideration should also be given to the initiation of a national survey of land use to provide data relevant to national energy, environment, housing and community development policy formulation and monitoring foreign investment in land during the period in which a cooperative statistical program is being established.

Coordination

7. The Office of the Assistant Secretary for Policy Development and Research (HUD) should be assigned the lead responsibility for exploring the feasibility of initiating Federal/local cooperation in the standardization of local area data on housing and community development. The Bureau of the Census should play a major role in this effort.

8. A Federal statistical center should be established within HUD to exert leadership in the area of housing and community development. A study should be conducted to develop recommendations for specific functions, lead agency responsibilities and organizational location.