N 156-169. New housing units started, by ownership, type of structure, location, and construction cost, 1889-1970.

Source: U.S. Bureau of the Census, 1889-1962, Housing Construction Statistics, 1889 to 1964, tables A-1 and A-5; 1963-1970, Construction Reports, Housing Starts, series C 20-73-7, July 1973.

The data for 1889-1919 are from David M. Blank, The Volume of Residential Construction, 1889-1950, Technical Paper No. 9, National Bureau of Economic Research, 1954; data for 1920-1929 are from David L. Wickens and Ray R. Foster, Nonfarm Residential Construction, 1920-1936, National Bureau of Economic Research, Bulletin No. 65, 1937. The data for 1930-1944 are from U.S. Department of Labor, Construction, 1948 in Review, Bulletin No. 983, 1950; data for 1930-1936 represent a revision by the Bureau of Labor Statistics of the Wickens-Foster data. Data for 1945-1970 are Census Bureau estimates derived from its monthly estimates based on building permits and supplemented by sample surveys of housing starts in nonpermit-issuing areas. Data for 1945-1958 are revisions of data from the Bureau of Labor Statistics; detail data, such as number of units by type of structure, are not available for these years.

For methods used by Blank and Wickens-Foster, see the sources. Blank's data are based on a comprehensive tabulation of historical building permit data collected by the Works Progress Administration and made available by the Bureau of Labor Statistics (BLS). For methods used by BLS, see BLS Bulletin No. 1168, Techniques of Preparing Major BLS Statistical Series, December 1954, chapter 2. Basically, compilation of the housing starts series depends on four steps. First, an estimate is made of the number of housing units for which building permits have been issued in all permit-issuing places each month. For the country as a whole, about 85 percent of the private housing units were constructed in permit-issuing places in recent years. Since 1967, the series have pertained to approximately 13,000 places identified in 1967 as having local building permit systems. Coverage from 1963 to 1967 was based on a permit-issuing universe of 12,000 places, identified as permit-issuing in 1962. Back to 1959, the series relate to the 10,000 places identified as permitissuing in 1959. Prior to 1959, 6,600 places was the universe. Second, a survey is made each month in a sample of permit places. In each place, a sample of building permits is selected each month and an inquiry is made of the owner or the builder to find out whether and when the units covered by the permits have been started. From this sample of permits, ratios are calculated, by type of structure, of the number of units started to the number of units covered by permits. These ratios are then applied to the total number of units authorized by permits in the corresponding months to provide estimates of the total number of units started each month with permit authorization. Third, the estimates of the number of one-family units started in each month with permit authorization are adjusted upward by 3.3 percent to take care of the units started within permit-issuing areas but without permit authorization. The fourth step in estimating total housing starts is to estimate the number of units started in areas where building permit systems do not exist. In a sample of 100 areas, visits are made to a select group of persons who are presumed most likely to know about local housing activities. A list is obtained from them of all residential buildings they know to have been started within the nonpermit portions of these areas during the preceding month. Within those portions of the sample area, a subsample of areas is canvassed for all units started since the previous month, identifying those not reported by the sources as well as those reported by them. This canvass provides a basis for estimating the number of units not reported by the local sources. The number of

units not reported is then added to the number of units reported, to provide an estimate of total housing starts in areas not covered by building permit systems.

The housing units covered in these series are permanent housekeeping units in new residential structures. Excluded are temporary units; accommodations without housekeeping facilities such as transient hotels, dormitories, and clubhouses; mobile homes, trailers, houseboats, sheds, and shacks used for housing purposes; units provided by conversion of existing structures; and housing units in nonresidential structures such as factories, warehouses, or public buildings.

For regional estimates of the number of new private nonfarm housing units started, 1920-1950, see Leo Grebler, David M. Blank, and Louis Winnick, Capital Formation in Residential Real Estate: Trends and Prospects, Princeton University Press, 1956, table H-1. Regional estimates for later years appear in Bureau of the Census, Construction Reports, series C 20.

N 157-158, new housing units started, by ownership. For bases of estimates for privately financed housing units, see text for series N 156-169. Publicly owned housing includes housing units in buildings for which construction contracts were awarded by Federal, State, or local governments. Information on public housing is obtained, for the most part, from the agencies involved, e.g., Department of Housing and Urban Development, Department of Defense, New York City Housing Authority, and others. The criterion for classifying housing units as public is ownership of the facilities rather than the source of funds. Thus, low-rent housing projects owned by local housing authorities are classified as public even though they may be financed by local bonds issued to private investors, and military housing units owned by the Department of Defense are also classified as public even though they may be financed by mortgages held by private lending institutions. Figures exclude temporary dwellings built during the defense period and World War II (1940's), veterans temporary re-use housing (see text for series N 186-191), and temporary structures on military posts and similar installations. Units in structures built by private developers for sale upon completion to local public housing authorities under the Department of Housing and Urban Development "Turnkey" program are classified as private housing.

N 162-163, new housing units started, by location. The distribution of housing starts between units inside and outside standard metropolitan statistical areas (SMSA's) is based on the definitions published by the Office of Management and Budget in Standard Metropolitan Statistical Areas. Data for 1959-1960 are based on 1959 definitions; for 1961-1963 on 1961 definitions; for January 1964March 1968 on 1964 definitions; and for April 1968-1970 on 1967 definitions. The term, "urban" was applied to all incorporated places with a population of 2,500 or more and to a relatively small number of areas urban under special rule. "Rural-nonfarm" housing included all housing (except farm housing) in unincorporated areas and in incorporated places of less than 2,500 inhabitants. This classification for 1920-1929 was based on the 1930 Census of Population and for 1930-1944 on the 1940 census. This classification system was abandoned in 1954 because of the difficulties of resolving differences between the geographic areas used for building permit systems and the urban areas as newly defined in the 1950 census. Beginning in 1950, housing starts have been classified by those inside and outside the standard metropolitan statistical areas.

N 164-169, construction cost. The construction cost data for the privately owned units are not reported directly but are based on permit valuations adjusted for understatement of construction cost

and relationship between costs in permit places and nonpermit areas. They cover the cost of labor, material, and subcontracted work, and that part of the builder's overhead and profit chargeable directly to the building of the housing units started. Included is the cost of equipment which becomes an integral part of the housing unit and is essential to its general use. Excluded are the costs of land, site improvement, architectural fees, and sales profit.

The data for the publicly owned units are based on contract award values or estimated construction costs for individual projects, as reported by the several agencies administering the various public housing programs.

The source, Housing Construction Statistics, 1889 to 1964, was designed as an historical supplement to the current data issued by the Bureau of the Census in three publication series of its Construction Reports program-Housing Starts, series C 20; Building Permits, Housing Authorized in Individual Permit-Issuing Places, series C 40; and Building Permits, Housing Authorized in Permit-Issuing Places, Summary Statistics, series C 42.

N 170. Mobile home shipments, 1947-1970.

Source: U.S. Department of Housing and Urban Development, HUD Statistical Yearbook, annual issues.

Statistics on manufacturers' shipments of mobile homes are provided by the Mobile Home Manufacturers' Association, and include estimates for firms not associated with the MHMA. Mass production of 10-foot wide homes began in 1955; 12-foot wide homes in 1962.

Manufacturers' shipments of mobile homes are included in this volume because an addition to the housing supply is made by mobile homes as well as by the construction of new housing units. Some of the mobile homes, however, are used as seasonal homes and second homes and do not add to the supply of housing units occupied as usual places of residence. Furthermore, some are used for nonresidential purposes. The number of mobile homes used in these ways is not now known.

N 171-179. New publicly owned housing starts, by ownership and program, 1949-1970.

Source: U.S. Bureau of Domestic Commerce, Construction Review, monthly issues.

These series, which are compiled by the U.S. Bureau of the Census, represent an actual count of publicly owned housing starts as reported by the Public Housing Administration (PHA), the Defense Department, the New York City Housing Authority, and other State and local housing authorities. Publicly owned housing units were not reported separately until 1935. It is considered that the volume of permanent publicly owned housing units prior to 1935 is insignificant. Housing provided under the Federal emergency programs, including those of World War II, consisted largely of units in temporary or converted structures and, therefore, are not included in the permanent units shown. Type of program data, i.e., Federal, State, or local, are not available for publicly owned housing prior to 1949.

N 180-185. Privately owned housing units in major Federal programs, 1935-1970.

Source: All series except N 181 and N 185, U.S. Housing and Home Finance Agency, 1935–1956, Annual Report, 1956, tables A-6, A-37, A-42, and A-54; 1957, Annual Report, 1957, tables A-7, A-48, A-53, and A-68; 1958–1963, Annual Report, 1964, table III-3; 1964-1970, U.S. Department of Housing and Urban Development, 1970 HUD Statistical Yearbook, tables 146, 160, 162, and 164. Series N 181 and N 185, U.S. Veterans Administration, Loan Guaranty Service, unpublished data.

Figures are based on reports of the agencies administering the programs. The Federal Housing Administration (FHA) and the Veterans Administration (VA) are agencies which insure or guarantee loans made by private lenders.

N 180-181, new privately owned units started under FHA and VA. Data are based on monthly reports of these agencies. These reports are based on the first of several inspections of newly started units required by the agencies, the timing of which coincides roughly with the definition of housing starts by the Bureau of Labor Statistics and the Bureau of the Census.

The figures may be used roughly to derive the units started under FHA and VA as a percentage of all private housing starts. However, an unknown number of units started under FHA or VA inspection is sold later for cash or with conventional (uninsured or unguaranteed) mortgage loans. On the other hand, the number of units started under the FHA program understates the role of FHA inasmuch as previously unoccupied (new) houses, for which the builder did not apply for FHA insurance before construction, are classified by FHA as "existing construction" when the houses are sold later with FHAinsured loans. In 1956, about one-fifth of the FHA units classified as "existing construction" were previously unoccupied (new). However, by 1970, such previously unoccupied (new) units amounted to less than 3 percent of those classified under “existing construction." For problems of coverage and comparability, see Department of Commerce and Department of Labor, Construction Review, "FHA and VA Housing Statistics and the Housing Market," June 1957. N 182-185, new and existing privately owned units covered by loans under FHA and VA. Data are based on monthly reports of these agencies and refer to loans on both new and existing construction at the time such loans were closed or actually insured. FHA "homes" include 1- to 4-family dwellings; FHA "rental projects" include structures having 5 or more dwelling units. Practically all VA loans are on single-family dwellings. The VA program was authorized in 1944 and the small 1944 activity is included in 1945.

N 186-191. Low-rent public housing units, by progress stage, and war and defense housing and veterans housing units available for occupancy, 1941-1970.

Source: U.S. Department of Housing and Urban Development, Housing and Mortgage Credit-FHA, HUD Statistical Yearbook, various issues.

These data comprise low-rent, publicly financed housing units occupied or available for occupancy, units to be constructed, and units that were to go directly under management since they needed no rehabilitation. The data are not comparable with series N 156-159 which relate to new construction starts and include all publicly financed units (Federal, State, and local).

N 192-195. Nonfarm dwelling units standing and selected components of change, 1890-1950.

Source: Leo Grebler, David M. Blank, and Louis Winnick, Capital Formation in Residential Real Estate: Trends and Prospects, Princeton University Press, 1956, tables 15 and A-1 (copyright, National Bureau of Economic Research, New York).

Estimates do not represent all components of change in the nonfarm housing inventory; that is, the units added through new construction or conversion minus the units destroyed through demolition or disaster losses during a certain period do not equal the net change in the inventory during the same period. This difference is due mainly to the following factors: (1) The net change in the nonfarm housing inventory reflects the reclassification of farm dwellings, as farmland is absorbed in suburban and urban development; (2) the inventory includes temporary dwelling units, shacks, trailers, and dwellings in nonresidential buildings such as factories or warehouses, which are not included in the estimates of housing starts; (3) the periods of the inventory estimates are not fully reconcilable with the calendar-year estimates of new or converted units and of demolitions; (4) minor changes in census definitions; and (5) deficiencies of estimates, particularly for conversions and demolitions. For a reconciliation of the net change in inventory and the various components