period in Series P-20, No. 305, Geographical Mobility: March 1975 to March 1976; for the 1970-75 period in Series P-20, No. 285, Mobility of the Population of the United States: March 1970 to March 1975; for the 1970-74 period in Series P-20, No. 273; and for the 1970-73 period in Series P-20, No. 262. Data for the 1970-71 period were issued in Series P-20, No. 235, and similar statistics were published in this series each year beginning with the report for 1947-48.

Statistics on geographical mobility of the population for cities, counties, SMSA's, urbanized areas, State economic areas, States, divisions, regions, and the United States appear in Volume I of the 1970 Census of Population (based on State of birth or residence 5 years before the census). Detailed statistics on mobility status by race and sex for these areas and the United States appear in Volume II, Subject Reports: PC(2)-2A, State of Birth; PC(2)-2B, Mobility for States and the Nation; PC(2)-2C, Mobility for Metropolitan Areas; PC(2)-2D, Lifetime and Recent Migration; PC (2)-2E, Migration Between State Economic Areas; and PC(2)-7E, Occupation and Residence in 1965. Some other subject reports of the 1970 census present statistics on mobility status in relation to the main subject of the report.

COMPARABILITY OF METROPOLITAN AND NONMETROPOLITAN DATA FROM THE 1980 CPS WITH DATA FOR PREVIOUS YEARS

Changes in CPS design and procedures over the last several years have made the annual series of sample population data for metropolitan and nonmetropolitan areas inconsistent. Analytic comparisons of year-to-year changes in these figures should be avoided. Trends in metropolitan

and nonmetropolitan population growth over the 1970-80 and 1975-80 periods should not be appreciably affected by the procedural changes.

The major revisions to the CPS sample design and estimation methods have involved the expansion of the number of sample units from 55,000 housing units to 65,500 housing units. This incorporation of approximately 10,000 supplemental housing units into the March CPS sample in 1977 was accompanied by new procedures for inflating the sample results to reflect national estimates. It was determined subsequent to the introduction of the additional sample that the new inflating (weighting) procedures used for processing both the March 1977 and March 1978 CPS supplement data had resulted in an apparent overestimate of the nonmetropolitan population and corresponding underestimate of the metropolitan population for those years. For March 1979, another revision of the weighting process was introduced to correct the problem discovered in the earlier procedures. The result of this change was a spurious large increase in the metropolitan population and decrease in the nonmetropolitan population relative to March 1978 levels.

Beginning with the March 1979 CPS, metropolitan and nonmetropolitan population estimates also reflect other operational changes including the introduction of a coverage improvement sample designed to provide greater accuracy in survey estimation. The net effect of all changes in procedure was to increase the metropolitan area estimates. Research and detailed analysis of the impact of each procedural change on the population estimates is underway and the results will be issued in a forthcoming technical report.

period in Series P-20, No. 305, Geographical Mobility: March 1975 to March 1976; for the 1970-75 period in Series P-20, No. 285, Mobility of the Population of the United States: March 1970 to March 1975; for the 1970-74 period in Series P-20, No. 273; and for the 1970-73 period in Series P-20, No. 262. Data for the 1970-71 period were issued in Series P-20, No. 235, and similar statistics were published in this series each year beginning with the report for 1947-48. Statistics on geographical mobility of the population for cities, counties, SMSA's, urbanized areas, State economic areas, States, divisions, regions, and the United States appear in Volume I of the 1970 Census of Population (based on State of birth or residence 5 years before the census). Detailed statistics on mobility status by race and sex for these areas and the United States appear in Volume II, Subject Reports: PC(2)-2A, State of Birth; PC(2)-2B, Mobility for States and the Nation; PC(2)-2C, Mobility for Metropolitan Areas; PC(2)-2D, Lifetime and Recent Migration; PC (2)-2E, Migration Between State Economic Areas; and PC(2)-7E, Occupation and Residence in 1965. Some other subject reports of the 1970 census present statistics on mobility status in relation to the main subject of the report.

COMPARABILITY OF METROPOLITAN AND NONMETROPOLITAN DATA FROM THE 1980 CPS WITH DATA FOR PREVIOUS YEARS

Changes in CPS design and procedures over the last several years have made the annual series of sample population data for metropolitan and nonmetropolitan areas inconsistent. Analytic comparisons of year-to-year changes in these figures should be avoided. Trends in metropolitan

and nonmetropolitan population growth over the 1970-80 and 1975-80 periods should not be appreciably affected by the procedural changes.

The major revisions to the CPS sample design and estimation methods have involved the expansion of the number of sample units from 55,000 housing units to 65,500 housing units. This incorporation of approximately 10,000 supplemental housing units into the March CPS sample in 1977 was accompanied by new procedures for inflating the sample results to reflect national estimates. It was determined subsequent to the introduction of the additional sample that the new inflating (weighting) procedures used for processing both the March 1977 and March 1978 CPS supplement data had resulted in an apparent overestimate of the nonmetropolitan population and corresponding underestimate of the metropolitan population for those years. For March 1979, another revision of the weighting process was introduced to correct the problem discovered in the earlier procedures. The result of this change was a spurious large increase in the metropolitan population and decrease in the nonmetropolitan population relative to March 1978 levels.

Beginning with the March 1979 CPS, metropolitan and nonmetropolitan population estimates also reflect other operational changes including the introduction of a coverage improvement sample designed to provide greater accuracy in survey estimation. The net effect of all changes in procedure was to increase the metropolitan area estimates. Research and detailed analysis of the impact of each procedural change on the population estimates is underway and the results will be issued in a forthcoming technical report.