Undercount

Undercount refers to a situation in which the number identified by a counting process is smaller than the true quantity. In statistics and demography, it occurs when some members of the population are not counted or counted less than once. It is distinct from miscounts, where recording errors occur, and from underreporting, which involves incomplete disclosure of information rather than missing units.

Common contexts include national censuses, large surveys, and administrative data. Causes include nonresponse, inaccessible populations (such as the homeless or undocumented migrants), language barriers, concerns about privacy, sampling biases, and data processing errors. Political and administrative factors can also play a role, such as deliberate exclusion or fear-driven nonparticipation.

Consequences can be wide-ranging: biased demographic estimates, distorted resource allocation, and impacts on political representation and policy planning for health, education, housing, and infrastructure.

Mitigation and estimation methods include post-enumeration surveys and follow-up operations to locate missed households, weighting adjustments, imputation, and statistical techniques such as capture-recapture or dual-system estimation. Transparency about methods and ongoing coverage evaluation help reduce undercounts.

Undercounts are a core concern in census-taking and related data collection, and are often discussed alongside overcounts and miscounts.