| Abstract |
Computerized U.S. Census data has been most widely used for (1)employment, fertility, demographic and stratification researchinvolving Public Use Sample (PUS) microdata on the national level,and (2) applied research (for planning, administration, marketing,and other applications) with summary (aggregated) data forlocalized (i.e., block, tract, community, etc.) geographic units. Athird, highly productive avenue of research, involving Census PUSmicro-data for localized urban units (i.e., SMSAs, counties andespecially selected large-city neighborhoods), has not received theattention it merits, either among sophisticated public data usersor among novice users.Three forms of current or future small area,Census microdata constitute resources for urban research. First,conventional 1970 Census PUS data sets are available for countiesand/or SMSAs (with minimum populations of 250,000). Second, specialtabulations for the two largest U.S. cities permit analysis of 1970household and person records group by (sub-county) urbanneighborhoods (27 in New York City; 12 in Chicago). Third, 1980Census microdata, by allowing identification of geographic areas ofsmaller population size (100,000 population), will vastly expandthe applications of localized research with the conventional PUS orspecial tabulations. In addition, the 1980 PUS microdata will, forthe first time, allow comparative time-series analyses of county(or SMSA) area populations, over the 1970-1980 decade.In contrastto national PUS microdata research, local level analyses have theadvantages of (1) smaller data set size and processing costs, (2)more immediate integration of computerized research hypotheses withadditional sources of (qualitative) information and questions(stemming from direct knowledge of the communities studied), and(3) increased ability to zero in on specialized ethnic,occupational-industrial, migrant, age, etc. urban population groupswhich are disproportionately represented in particular localenvironments. Our own research projects (at various stages ofdevelopment) which attempt to exploit these advantages includecomputerized analysis of:1. Patterns of household composition, andsource and structure of family income, among Upper East Side andUpper West Side Manhattan residents with family incomes of \$50,000. or more (as reported in the 1970 Census)2. Employmentpatterns of married women of Cuban immigrant background, inrelation to family class position and period of immigration, forHudson County, New Jersey3. Contrasts in the occupational positionsand household patterns of first-generation and second-generationhusbands and wives of Italian background in a New York City workingclass community (Astoria-Long Island City, Queens)4. Wives'employment patterns in relation to ethnic background and husbands'occupations and income levels in a working class community locatedin a manufacturing center (South Side, Chicago)5. Change in thesocial and demographic characteristics of succeeding groups ofmigrants to an expanding "sunbelt" metropolitan area (Albuquerque,New Mexico)6. Contrasts in local housing markets and housingavailability, involving analysis of the number and characteristicsof vacant housing units for New Jersey counties.These, as well asother projects we have assisted, have been undertaken with variedsoftware resources, including packages (such as CENTS-AID) withunique hierarchical file processing capabilities, as well as moreversatile (non-hierarchical), general purpose packages (such asSPSS). The advantages and research applications of small area,micro-databases can be realized with a range of software techniquesand user-formulated research strategies. |
