The Census Series, as Used

Decennial volumes, 1850–1940 · PL 94-171 and DP1, 2020 · ACS five-year files — page images wherever the machine text interleaves

What it is: the statistical spine of the Population file, taken from the volumes themselves. The historical volumes on census.gov divide by their machinery: the 1850, 1880, 1900, 1910, 1930-state and 1950 parts carry OCR text layers, scrambled exactly where it matters — in the two-column minor-civil-division tables — so the township figures in this build were settled from page images wherever the OCR interleaved columns (1850, 1880, 1900, 1910, 1930, 1940 all read as images); the 1910 volume I general chapters, the 1930 volume I, the 1960 A-parts, the 1840 compendium and the 1860/1870 parts carry no text layer at all. From the 1850 subdivision table comes the frame the Population and Tribes files print: Clay and Harris returned no free colored residents; the county's twenty-nine lived in Centre, Penn, Portage and Union townships and South Bend — categories printed as each census printed them. The modern line runs on the redistricting file and the demographic profile: PL 94-171 and DP1 give the CDP's 2020 count, 30,337; the delineated series runs 1990: 20,241 (Clay part 8,764, Harris part 11,477) · 2000: 28,284 · 2010: 30,465 · 2020: 30,337.

Its standing, from this build's experience: the census is this build's most trusted counter and its most carefully bounded one. The ACS speaks for a CDP only in five-year estimates — the 2020–2024 profile, 30,321 (±575), reached through the Missouri Census Data Center's profile for GEOID 16000US1828800 — and there is no annual estimates series for a place of this class: QuickFacts prints the line not-applicable. The Bureau's API, tested first-hand this build, is now key-gated (a 302 "Missing Key" returned), and Census Reporter serves only the current release; the vintage-by-vintage ACS series stands as a named gap with three settling archives (GAP-POP-6). Where the decennial and the survey disagree — median age 41.3 (DP1) against 40.4 (ACS) — both print as their own instruments' readings.

Where it fails: silently, at its own edges. No census of agriculture prints township figures for Clay or Harris in any volume reached; no township influenza count exists in the federal tables; no CDP-level COVID series exists anywhere, and none was manufactured from the ZIP-code dataset, because ZIP 46530 is not the CDP. The census proves counts at its own geography and definitions, in its own years — and this build asks it for nothing else.

Sources:

2. U.S. Census Bureau — decennial volumes as above; 2020 PL 94-171 file and DP1; 2020 Gazetteer files; QuickFacts, FIPS 1828800.