A population census is a complete enumeration of all persons in a defined territory at a specific point in time. Censuses provide the denominator for virtually all demographic rates and the baseline for population estimates and projections between census years. Key methodological choices include de facto (counting people where they are on census night) versus de jure (counting people at their usual residence), individual versus household-based enumeration, and the trade-off between census frequency and cost. Census errors include under-coverage (missing people, particularly marginalized populations), over-coverage (counting people twice), content error (inaccurate responses), and processing error. Post-enumeration surveys estimate coverage rates, and census data are adjusted for known biases before use. Modern census alternatives include register-based censuses (using administrative records) and rolling surveys, adopted by some countries to reduce cost.
Examine the questionnaire and methodology documentation for a real census (e.g., US Census 2020 or a country with published methodology). Identify the decisions made about enumeration type, question design, and coverage evaluation. Then compare coverage estimates across demographic groups to see which populations are hardest to count.
Every demographic rate you have studied — crude birth rates, crude death rates, age-specific rates, dependency ratios — requires a denominator: the population at risk. That denominator ultimately comes from the population census, the most fundamental data collection operation in demography. Without census counts, the entire architecture of demographic measurement collapses into numerators without denominators.
A census aims to enumerate every person in a defined territory at a specific point in time. This distinguishes it from a survey, which draws a sample. The complete enumeration principle means a census can provide data for small geographic areas and small population subgroups that surveys cannot — a property essential for electoral apportionment, service planning, and resource allocation.
Two fundamental approaches exist. In a de facto census, enumerators count people wherever they are found on census night. This is operationally simple — you count the person in the dwelling where they sleep — but it places visitors and travelers in the wrong geographic unit. In a de jure census, people are assigned to their usual place of residence, regardless of where they are on census night. This produces better denominators for local demographic rates but requires more complex enumeration (determining each person's usual residence, handling people with multiple residences, tracing people who are temporarily away).
No census achieves perfect coverage. Under-coverage — missing people entirely — is the most serious error because it systematically affects certain groups: young men (who may be transient or avoiding enumeration), racial and ethnic minorities, homeless populations, undocumented immigrants, and people in remote or conflict-affected areas. Over-coverage — counting people more than once — can occur when people with multiple residences are enumerated at each one. The post-enumeration survey (PES) estimates these errors through an independent re-enumeration of a sample of areas, using capture-recapture logic: if the census found 95% of the people that the PES independently found, then the census missed approximately 5%. These estimates allow statistical adjustment of census totals.
Modern alternatives to the traditional census are emerging. Several Nordic countries now conduct register-based censuses using administrative records (population registers, tax records, property databases) instead of direct enumeration, dramatically reducing cost and respondent burden. France replaced its traditional decennial census with a rolling survey that covers the entire country over five-year cycles. These innovations reflect the tension between the census ideal (complete enumeration) and its practical limitations (enormous cost, declining response rates, and the inherent difficulty of counting mobile, complex populations).
Topics in reflective domains aren't scored by quiz answers. Read, reflect, and mark when you've thought it through.