Source of data. -- The estimates presented in this report are based on data obtained from two surveys conducted by the Bureau of the Census. Data concerning vacancy rates and tenure of occupied housing units are from the monthly sample of the Current Population Survey (CPS). Characteristics of occupied housing units are from the Quarterly Household Survey (QHS).

The Current Population Survey sample is spread over 461 areas comprising 923 counties and independent cities with coverage in each of the 50 States and the District of Columbia. Approximately 47,000 occupied households are designated for interview each month; of this number 2,000 occupied units, on the average, are visited but interviews are not obtained because the occupants are not found at home after repeated calls or are unavailable for some other reason. In addition to the 47,000, there are also about 8,000 sample units in an average month which are visited but are found to be vacant or otherwise not to be interviewed.1 Data on vacancy rates and tenure of occupied units for the fourth quarter 1973 are averages over the 3 months, October, November, and December.

The data concerning characteristics of occupied housing units are from the Quarterly Household Survey (QHS). The figures for the fourth. quarter 1973 are based on the average of four quarterly estimates produced from interviews conducted in January, April, July, and October, 1973. Over this period of time, January to October 1973, the QHS was redesigned; two sample designs were involved. For January and April 1973, the QHS sample was in 266 areas, comprising 578 counties and independent cities. For July and October, the areas were reduced to 103, comprising 296 counties and independent cities.

The estimates of occupied housing units for this report were based on approximately one-third of the units designated for the full QHS samples within each of the areas. Estimates for a full QHS sample were based on interviews conducted in a subsample of households selected from a larger sample within the areas (the subsample was

1A more complete description of the Current Population Survey sample is given in Current Population Reports, Series P-23, No. 22, June 1967.

derived by eliminating a random subset of owneroccupied and low income households identified in the larger sample). The samples varied in size over the four surveys reflecting the changes in the sample designs. For the January, and April interview periods, the larger samples amounted to about 16,000 designated units, and about 7,000 units were designated for July and October. Subsamples of about 13,500, and 6,000 interviewed households were selected for the respective full QHS samples. The occupied housing unit data used for this report are based on about one-third of these units. Units with family incomes below $4,000 were treated as low income households for the purpose of identifying the smaller samples.

In addition to the interviewed households, for each of the quarters a set of units, amounting to about 5 percent of the larger sample of occupied units, were visited but interviews were not obtained because the occupants were not found at home after repeated calls or were unavailable for some other reason. There were also additional units, amounting to about 15 percent of the larger sample, which were visited but found to be vacant or otherwise not to be interviewed. Data for these units were not included in the figures for this report.

The estimating procedures used for the QHS involve the inflation of the weighted sample results to independent estimates of the number of owner-occupied and renter-occupied housing units by urban and rural residence based on data from the Current Population Survey (CPS). These independent estimates have a lower sampling error due to the larger sample size and because other current independent estimates of the population of the United States are employed in the CPS estimation procedure.

Reliability of the estimates.--Since the estimates are based on a sample, they may differ somewhat from figures obtained if a complete census had been taken using the same schedules, instructions, and enumerators. As in any survey work, the results are subject to errors of response and of reporting as well as being subject to sampling variability.

The sampling error is primarily a measure of sampling variability, that is, of the variations that occur by chance because a sample rather than the