areas. The pool of vacancies which existed in 1940 in many cities was rapidly dried up during 1941 and 1942. By 1943, most of the unoccupied units which existed were relatively undesirable because of their condition or because of their location. In 1945, however, as a result of cutbacks in war production, a slight easing in the vacancy situation occurred. But by late 1945 the huge demobilization of persons from the armed forces resulted in an even tighter situation. in the number of habitable vacancies for rent or sale in most cities than existed during the war. For example, Philadelphia in April 1940, had a gross habitable vacancy rate of 4.9 percent. The

habitable vacancy rate declined sharply during 1941 and 1942 until by late 1943 only 1.0 percent of total dwelling units were habitable vacancies. In early 1945 as a result of cutbacks in defense industries, the total habitable vacancy situation had loosened slightly. However, the percentage of habitable vacancies for rent or sale remained at about the same level as in 1943. Finally, a survey of Philadelphia made in the last half of 1946 showed that the total habitable vacancy rate had dropped to 0.9 percent.

It should be pointed out that the term "habitable" does not mean that the vacant dwellings were necessarily of standard quality, since units needing major repairs were classified as habitable if located in a neighborhood where similar units were occupied.

Total dwelling units

Texas:

Area 2

April 1940 $

1946

98,000 115,000 6.2 4.3-4.8 2.7-2.9 2.7-2.9 60,000 5.3 2.0-2.3 1.9-2.1 62,000 5.2

Houston city.

113, 326

124,000

5.1 4.0-4.6 3.3-3.7 2.7-3.3 5. 2

1.0

San Antonio city.

69, 731

Greater San Antonio

The survey was made by the Bureau of Labor Statistics.

The 1940 data include public housing dwelling units while estimates for subsequent years are based on the results of sample surveys among privately financed dwelling units only. However, the number of public housing units in any locality in 1940 was too small to have any significant effect on the figures.

For places designated city, county, or metropolitan district, the 1940 to 1946 figures cover identical geographical areas. For those places designated "area," the 1940 to 1946 figures generally included the central city and nearby communitics which comprised the housing market area for that city. The geographical boundaries of the housing market area occasionally changed from one survey to the next but in only a few cases did the coverage vary by more than 10 percent. A detailed description of each survey "area" is contained in the individual report for the survey.

The 1940 dwelling unit figures are for an area geographically identical to the area covered by the latest survey for all areas except those carrying footnote 7, 8, or 9 in this column.

For places where no survey was made in 1946, the estimate is for the latest survey in that area.

The 1940 vacancy rates include both seasonal and nonseasonal units. For surveys made between 1941 and September 1945, the vacancy rates may include some seasonal units while the vacancy rates for areas surveyed in September 1945 and later are for nonseasonal vacancies only. The effect of such units on the vacancy rates needs to be taken into account only in the few places which included a resort area in the survey.

The definition of "habitable" changed several times between 1940 and 1946. (Refer to the section on Vacancy and Occupancy in the text.) The

enumeration procedure for surveys made from 1941 to August 1942 combined vacant units in need of major repairs with those unfit for use. In order to show a comparable set of figures for all surveys, two vacancy rates are shown for the surveys made during this period-the first representing vacancies in good condition or in need of minor repairs, and the second representing vacancies in good condition or in need of minor repairs plus those in need of major repairs or unfit for use. Consequently, the proportion of vacant units which compares with the April 1940 rate and those after August 1942 lies somewhere between the two figures.

7 Since the latest survey in this area included a few small unincorporated places for which 1940 dwelling unit figures are not available, the 1940 and 1946 areas are not geographically identical.

The 1940 figure is for the metropolitan district while the 1946 estimate is for an area slightly smaller geographically than the 1940 metropolitan district, so that the 1940 and 1946 areas are not geographically identical.

The 1940 figure is for the central city and all parts of nearby towns while the 1946 estimate includes only the built-up portions of the nearby towns which were part of the survey area, so that the 1940 and 1946 areas are not geographically identical.

Sources: April 1940-Census of Housing, 1941 to 1946-Dwelling Unit Vacancy and Occupancy Surveys made for the Defense Housing Coordinator and the National Housing Agency by the Work Projects Administration until August 1942, and after August 1942 by the Bureau of the Census and the Bureau of Labor Statistics. Those made by the Bureau of Labor Statistics are indicated by an asterisk.

This chapter is concerned with the characteristics of occupied dwelling units,' particularly those qualities which are pertinent to the determination of current and future housing requirements. All of the national and regional data in this chapter are taken from the census of 1940 or the sample survey of November 1945. Data for smaller areas, States, counties, cities, and minor civil divisions may be found in the 1940 housing census. Data for the Nation as a whole, for 4 regions and 35 metropolitan districts will also be available from the April 1947 sample survey. The data concerned with the characteristics of housing in specific localities as of 1946 were obtained from surveys conducted for the National Housing Agency by the Bureau of the Census or the Bureau of Labor Statistics. The comparable locality data for 1940 are from the Census of Housing.

1 The Census Bureau defines a dwelling unit in such a way that the number of occupied dwelling units is approximately equivalent to the number of private families or private households. Because of some technical points a discussion of which appears in ch. V, the number of private households in the United States in 1940 actually was slightly greater than the number of occupied dwelling units. The difference is so slight that for all practical purposes the statistics for occupied dwelling units and families can be used interchangeably.

results of the sample survey the median corresponding in November 1945 was 2.97. It should be remembered, however, that at the time of the sample survey demobilization was far from complete, a fact that tended to keep down the average size of households. In the compilation of these medians and the distribution of households by number of persons appearing in table 60, all persons in the household have been considered. Inasmuch as more than 9,300,000 persons in 1940 were living in private nonfarm households as boarders, lodgers, servants, etc., and therefore were not members of the private families therein, the medians and distributions on a private family basis would have been somewhat different.

One-person families accounted for 3,158,000 occupied dwelling units, or approximately 10 percent of all nonfarm dwelling units occupied in 1945. Even in 1940, however, there were 2,316,000 such families, representing more than 8 percent of all occupied nonfarm units. Some of the 1945 total represents servicemen's wives still

awaiting the return of their husbands. A large number of the others are widows, widowers, and elderly persons.

The median number of persons per owneroccupied urban dwelling unit in 1940 was 3.33 as compared with 3.06 for tenant-occupied units. In urban areas of all geographic divisions, the median number of persons per household was larger in the case of owner-occupied as compared with tenant-occupied units. The reverse situation was true in rural nonfarm areas where the median number of persons per tenant-occupied unit, 3.32, exceeded the 3.11 median for owneroccupied units. The geographic division medians followed the national pattern in this respect except for the Pacific and Mountain States where the median numbers of persons per owner-occupied rural nonfarm unit were larger than the medians for tenant-occupied units.

The distributions of occupied nonfarm dwelling units by number of persons per unit in 1940 are presented in table 61 separately for urban and

NEW

PERCENT

20

occupied rural farm dwelling units are shown in table 62 as well.

In nonfarm areas the five-room unit was the type of unit encountered most frequently in the November 1945 survey and in the census of 1940. Twenty-four percent of the occupied nonfarm units in 1945 were of this size while 19 percent had four rooms and the same percentage had six rooms. There were 2,569,000 units, or 8 percent of the nonfarm total, which had one or two rooms, Table 63.-Number of persons per room: Occupied nonfarm dwelling units, urban and rural nonfarm, 1940 and 19451

[Dwelling units in thousands]

20

10

SOURCE: Bureau of the Census

MOUNTAIN