The decennial census is described as a headcount, and its consequences are structural. Two separate systems, political representation and program funding, both take their inputs from the same numbers.
Apportionment is the constitutional purpose
The census exists because seats in the House of Representatives are distributed among the states by population, and that distribution must be recalculated as people move.
The total number of seats is fixed, so apportionment is a zero-sum exercise. A state gaining residents faster than the national rate gains seats from states growing more slowly.
Electoral College votes follow directly, since each state's total is its House seats plus its two senators.
District lines are drawn from the detailed counts
Apportionment uses state totals, but redistricting requires population down to the block level, and the census is the only source of that detail.
Courts have long required districts within a state to be close to equal in population, which makes the block-level counts legally operative rather than merely informative.
State legislatures, commissions and courts all work from the same released tables when boundaries are redrawn.
Funding formulas consume the same data
A large number of federal programs distribute money using population-based formulas. Highway funds, health programs, education support, nutrition assistance and emergency preparedness all draw on census-derived figures.
Many of these formulas use not just headcount but characteristics such as age distribution, income levels and housing status, which come from the census and its associated annual surveys.
A jurisdiction counted short receives less on every one of those lines simultaneously.
Undercounts persist and are not evenly distributed
Certain groups are consistently harder to count, including renters, young children, people in irregular housing and households that move frequently.
Areas with those characteristics tend to be undercounted, which is why local governments spend money on outreach that has no purpose other than improving response.
The bureau applies statistical methods to fill gaps, but the estimates are only as good as the underlying response.
The ten-year lock is what makes it consequential
Annual surveys track change between counts, and the decennial figures remain the anchor for apportionment and many formulas throughout the decade.
A community miscounted in a census year carries that number through ten budget cycles before the next opportunity to correct it. The delay, rather than the count itself, is what turns an administrative error into a durable one.