White Paper

RSS

Census data, most recently Census 2000 data, traditionally offers a reasonably accurate decennial count of the U.S. population. Every corner of the land, from each city block to each block-like equivalent in remote rural areas, receives the once-every-ten-year scrutiny of the census takers. The census numbers feed the constitutional requirement to reapportion the 435 seats of the U.S. House of Representatives among the 50 states. They also provide grist for the redistricting mills, those state legislative committees and their consultants, that carve up congressional and state legislative districts within each state. Oh yes, and the money formulas…government dollars for numerous programs are frequently divvied up based on population counts and current estimates. These uses of demographic data drive a long-standing quest for accuracy on the part of official data collectors as well as politicians and program beneficiaries. With so much at stake, count accuracy has always been the primary goal of census enumerations.

Complicating this situation is the inherent messiness of demographic reality. Take household decline. Household decline can result from temporary housing vacancies, typically in units "vacant for rent" or "vacant for sale." These vacancies can be short-term, long-term, or seasonal. Likewise, household growth can occur where new homes are built or where housing unit vacancies are filled.

Thankfully, the U.S. Census Bureau has provided increasing accuracy of enumeration counts with each successive decennial census. Census 2000 was by most measures the most accurate census ever. Nevertheless, business users of demographics are acutely aware of the fading accuracy of aging census data. Current year estimates have long been a staple of data vendor offerings, along with five year projections, because the future trend is as important as the current magnitude (size) of population.