Scott Alexander, curated
← Back to curation

Bush Did North Dakota

Quality
74
Strong
Claude Shift
58
Moderate
RWI
3
of 10

Summary

Introduces the 'North Dakota constant.' The 2016 Chapman Survey of American Fears slipped in a PLACEBO conspiracy — a cover-up of 'the North Dakota crash' that never happened — and 33% endorsed it. Scott argues this is reasonable-ish given a forced agree/disagree with no 'don't know' option and a guilty-until-proven-innocent prior on government, and posits a North Dakota constant of ~33%: the share who'll endorse any conspiracy for no reason at all (vs the 4% Lizardman constant). So the alarming 'Guess et al: 46% of Trump voters believe Pizzagate' (vs 38% overall, 45% of Hispanics) isn't outrageous — only ~10-15% above the floor, balanced by ~10-15% fewer Clinton supporters. He probes why North Dakota (33%) far exceeds Lizardman (4%) — the missing 'don't know,' low-information voters reading the question charitably, conspiracy-priors — and uses birtherism (28% of Black respondents) to warn against over-reading 'X% of GROUP believes CONSPIRACY.' Conclusion: don't draw strong conclusions from such stats when X is near 33%, and don't treat Pizzagate's numbers as strong evidence that fake news is very convincing.

Why this score

Quality 74 · Strong. Strong, high (74). A concise, frequently-cited piece introducing a genuinely useful, durable concept (the North Dakota constant, companion to Lizardman) plus an important methodological lesson about reading conspiracy/poll statistics. Held just below Excellent by its brevity and single-concept focus.

Claude’s paradigm shift 58 · Moderate. Moderate, high (58). The 'North Dakota constant' coinage — extending the Lizardman constant to a conspiracy-belief floor with a placebo-option proof — was a fresh, useful contribution.

Real-world impact 3 · Moderate. Moderate (3). The North Dakota constant became used vocabulary for interpreting polls and debunking alarmist 'X% believe Y' reporting; its reach is intellectual.