Technology
Overcoverage and Undercoverage in Census Enumeration and Its Social Consequences
Quick fact
In the United States 2020 Census, an estimated 0.24% of the population was overcounted, but the undercount was far more severe for specific groups: Black people were undercounted by about 3.3%, Hispanic people by about 5%, and American Indians living on reservations by about 5.6%—even as other groups were overcounted.
Why this is interesting
Imagine a government trying to count every single person in a country—everyone. Now imagine it misses 2% of them. That might sound small, but in a country of 300 million, that's 6 million people, more than the population of many small nations. What happens when those 6 million are invisible?
Read the full explanation
Understanding Overcoverage and Undercoverage in Census Enumeration and Its Social Consequences
A census is a complete count of a population, but it is rarely perfect. Two types of errors can occur: overcoverage and undercoverage. Overcoverage happens when people are counted more than once, or when someone who should not be counted (like a visitor from another country) is included. Undercoverage is the opposite—people who should be counted are missed entirely. These are not random; they stem from practical challenges. For example, a college student might be counted at home and at school, or a family might fill out two forms listing the same child. On the other hand, a homeless person sleeping in a shelter might be easy to miss, or a person who fears government might avoid an enumerator. The result is a census number that is an estimate, not a perfect count, and the errors are not spread evenly. Some groups are overcounted, while others are undercounted.
A deeper explanation
The mechanism behind coverage errors lies in the design and execution of the enumeration process. Modern censuses use a combination of self-response forms, door-to-door visits, and statistical adjustment. Overcoverage often occurs because people are recorded in multiple locations—like a second home, a college dormitory, or a temporary job site—even though they live in only one place. Undercoverage happens when individuals are not captured at all, often because they are mobile, homeless, living in unconventional housing, or belong to populations that distrust the government. To measure these errors, statistical agencies conduct post-enumeration surveys that independently count a sample of areas and compare results to the census, estimating both over- and undercounts for various demographic groups. This matters because census counts have direct social consequences. They determine political representation, as the distribution of seats in national parliaments and electoral districts depends on population figures. They also drive public spending, as billions of dollars in federal funds are allocated to regions based on census-derived data. When certain groups are undercounted, they lose political power and funding, which can perpetuate inequality. Overcounting can also skew representation, benefiting some groups at the expense of others. Thus, coverage errors are not merely technical glitches; they are a form of institutional neglect that can affect the most vulnerable members of society.