A New Focus: Studying Linkages Among Household Structure, Race/Ethnicity, and Geographical Levels,
with Implications for Census Coverage
KEY WORDS: Households, families, race, Hispanic origin, household structure, household types, complex
households, Census 2000, census coverage, African American, white, Hispanic, Asian, American Indian,
This paper compares and contrasts household structure indicators for six of the main U.S. race/ethnic
groups at two geographical levels: at the national level and in custom-tailored local areas surrounding
research sites where small-scale qualitative studies of complex households for each race/ethnic group
were done. Four research questions are explored. To what extent do measures of household structure and
complexity in the overall U.S. population reflect similar patterns for different race/ethnic groups, or mask
variations among those populations? How do local measures of household structure and complexity add
to our understanding of population and household trends? What factors interact with and influence
household structure patterns? What are the implications of these findings for the perennial goal of
improving census coverage across race/ethnic groups in future censuses?
Nine household structure indicators are explored at local and national levels for six race/ethnic groups:
African Americans (coastal Virginia); non-Hispanic whites (upstate New York), Hispanic immigrants
(Central Virginia), Korean immigrants (Queens, New York), Navajos (northern Arizona), and Eskimos
(northern Alaska). For the latter groups, national statistics are presented for Asians, American Indians,
and Alaska Natives, respectively. Data come from Census 2000 tables in the American FactFinder
system at www.census.gov. The decennial census is the only dataset large enough to permit reliable
statistics on very small race/ethnic subpopulations at local geographical levels, allowing these finegrained
comparisons to be done.
Selected findings include the following. Measures of household structure and complexity in the overall
U.S. population do not reflect similar patterns for different race/ethnic groups; they mask notable
variations among the subpopulations and geographical levels. This is because non-Hispanic white
households differ markedly from those of the other groups, with the lowest proportions of households in
eight of the nine household structure indicators. Yet, as the largest group comprising 75% of all U.S.
households, they heavily influence (and skew downward) statistics on the overall national household
structure indicators examined here. Non-Hispanic whites also have one of the lowest growth rates, with
the result that as the proportion of these households declines over coming decades, their influence on
overall household structure statistics is likely to wane. If trends remain constant, household structure and
complexity indicator rates are likely to rise to reflect patterns of the minority subpopulations.
The census coverage tie-in is that race/ethnic minorities, complex households, and nonrelatives are all
subpopulations that have been associated in past research with census coverage errors: as they increase
over time, coverage errors may increase in tandem. These findings suggest that further ongoing research
and monitoring of changing household structure trends by race/ethnicity and geographical levels may
suggest adjustments that might be made to enumeration methods and operations to improve census
CITATION: Schwede, Laurie, 2007. “A New Focus: Studying Linkages Among Household Structure,
Race/Ethnicity, and Geographical Levels, with Implications for Census Coverage.” SRD Research
Report Series (Survey Methodology - #2007-38).