The 2010 Census had a net undercount of 4.6 percent for the population age 0 to 4 compared to a net overcount of 0.1 percent for the total population (Hogan et al. 2013). This 4.6 percent translates into a net undercount of almost 1 million young children. The coverage of young children in the decennial census is a persistent problem (West and Robinson 1999). Moreover, there is evidence that the undercount for this population has been increasing while there has been an improvement in coverage for other age groups (O’Hare 2015). The growth in the net undercount of young children is concerning, and in response, the Census Bureau formed the Task Force on the Undercount of Young Children and the Children Undercount Research Team to address this issue (U.S. Census Bureau 2014).

The Research Team has used data from multiple sources including vital statistics, Census operations, Census evaluations, and household surveys to investigate this problem. A common finding across this research is that variation in household structure is closely related to the undercount of young children (U.S. Census Bureau 2016, U.S. Census Bureau 2017a, b, & c). In recent decades, shifts in demographic, social, and economic patterns in the United States have led to changes in household and family structure. As a result, today there is less overlap between families and households than in the past as families are often spread over multiple households or multiple families may be living in one household (Cherlin 2010). Family diversity and complexity may cause ambiguity for census respondents about whom to include on the household roster, which may increase the likelihood that some household members are not counted.

Research on the undercount of young children has included indicators of household structure, mainly by focusing on the relationship of the young child to the householder. Relying only on the direct relationship between individuals and the householder may mask additional intricacies in living situations. In this report, we use the complex household typology, developed by Schwede and Terry (2013), to measure household structure. The typology uses information on the relationship to the householder for all household members to capture the diversity and complexity in structure that may not be reflected in the individual relationships. For instance, the living situation of a young child who lives with just his or her biological parents may be qualitatively different from that of a young child living with his or her parents and extended family or nonrelatives.

In this report, we use the complex household typology to analyze household structure and the undercount of young children in the 2010 Census. First, we classify households in the 2010 Census using the complex household typology, which includes a wider variety of household types than has previously been analyzed in the census. Next, we provide a descriptive profile of children in the 2010 Census by the complex household typology. In addition to summarizing the distribution of enumerated young children across these household types, this report expands recent Coverage Followup (CFU) and Census Coverage Measurement (CCM) analyses to see if specific types of complex households had a greater likelihood of errors involving young children. We conclude with a discussion of the importance of including complex household types in analyses of coverage.