Finally, four years after the start of the 2020 census, you can see how your state or county fared in the 2020 census count of young children.
This April, the Census Bureau released experimental data measuring the net count of young children ages 0 to 4 in the 2020 census in every state and in counties with at least 1000 children. The net count takes into account both the number of children missed and those counted twice or improperly counted–for example, because they were born after the census date. This data was calculated using the Demographic Analysis technique, or DA, where they compared the births, deaths, and migration in each state or county with the Census data. You can find your state and county here, and also see for each county how that compares to four factors that have some correlation with the undercount.
We thank the Bureau for its increased attention to the count of young children, and for this really excellent analysis.
Their findings confirm the accuracy of the research we have released from Dr. O’Hare using a somewhat different approach, using the vintage population estimates.
So, what can you tell about the count of young children in your state or community?
Under both approaches, the undercount of young children is pervasive; it also varies a lot from state to state. The Bureau’s data indicates that young children were undercounted in every state; Dr O’Hare’s research suggests there was a slight overcount in Vermont but an undercount in all other states. Using the Bureau’s approach, Florida had the largest undercount with -9.87%, while Vermont had the smallest undercount with .02%.
The Census Bureau analyzed the correlation between the county undercount and four factors:
- The share of children living in poverty
- The share of adults without a high school degree
- The share of households headed by women without a spouse
- The share of renting households.
Prior research has shown us that each of those factors has a correlation with a higher undercount, and the Bureau’s analysis confirms this moderate correlation. As you assess the count of young children in your state by county, you may want to look at these four factors; if a county has a large undercount even thought the county’s share of these factors was low, what else might explain the poor coverage of young children? Conversely, what happened in counties with high levels of these four factors but pretty good coverage of the count of young children—did they try an innovative strategy to count young children that might explain their success?
The District of Columbia is a special case, and understanding why DC has such a high undercount may help you understand results in your state. The DA approach suggests that the District had an undercount of -15.85%. Some of this very large undercount is undoubtedly due to DC having very high populations of children of color who are often missed in the census. But some of it may be because the Demographic Analysis approach looks at where babies are born. Although the birth record should reflect the baby’s home county, apparently often it reflects the county of their birth, so the extremely high DC rate may be partly due to many women who live in Maryland and Virginia giving birth in DC. This is worth considering as you look at the counties in your state. If a county has an unusually large or small population of color, that might affect the results. If a county with maternity hospitals or wards has an unusually high undercount, and the counties that do not have maternity services have a smaller undercount than you might expect, consider whether the location of maternity hospitals is affecting the count.
We still await further analyses that may also help explain the outcomes in your state. In particular, the Bureau plans to analyze the correlation between the young child undercount and race and ethnicity.
