US Political Violence Database, 1780–2010 (Turchin)

The event-level database behind every instability figure in THE COUNTED CROWD: 1,855 raw rows, 1,602 after excluding rows coded D or X (the file carries no legend for these codes; 1,602 is this studio’s count of the remainder), each event dated, typed (riot, lynching, rampage, assassination, terrorism, insurrection), subtyped, and sourced. The companion paper (Journal of Peace Research, 2012, https://web.archive.org/web/20260829173309/https://peterturchin.com/wp-content/uploads/2012/06/Turchin_JPR2012.pdf) supplies the published shape of the record: an instability wave across the second half of the nineteenth century, bracketed by two peaceful periods, with a fifty-year oscillation peaking around 1870, 1920, and 1970.

Corrected 2026-10-01. This record said the paper shows “the calm stretches between” the peaks. The paper’s calm is two bracketing periods; “From 1860 to 1920, the level of violence fluctuated around a very high level.” The 1,602 count is this studio’s: it excludes rows coded D or X and keeps rows coded D? and XD. The paper itself says “The USPV database includes 1,590 unique instability events.” The paper and the dataset now 404 at the author’s site and are read through the Internet Archive; the paper link above is the capture of 29 August 2026, re-read 3 October 2026, and both quoted sentences are in it.

Scope of this record, stated plainly. The essay’s decomposition figures (41% of events race-coded, 42% of fatalities, 438 lynching episodes, the 1860–1879 fatality concentration) were computed directly from the downloaded file by this studio on 2026-08-06; the computation lives in this studio’s predictive-history repository, which is private, (06_TESTS/e9_psi/who_is_the_crowd.md) and can be re-run from the file this url names. The codings are Turchin’s team’s judgments, and the essay’s claims inherit that: this record supports “the record counts,” never “history was.” The “age of concord” phrasing for the 1940–1960 trough follows Turchin’s own characterization as relayed in Richerson’s review of Ages of Discord (Cliodynamics 8, 2017).