Resamplings of the Simon-Ehrlich wager split by price metric: Ehrlich wins 61.2-63% of ten-year intervals under CPI-deflated prices, while Simon wins 54.2% under time prices

population-policyforecastingluck-vs-skillsimon-ehrlich-wagerstatisticsunverified-quant

Three published estimates resample the Simon-Ehrlich wager's mechanics (the direction of commodity prices over rolling ten-year windows) across roughly a century of data rather than trusting the single realized 1980-1990 decade — but they do not agree on a winner; they split by price metric. Kiel, Matheson & Golembiewski (2009), using CPI-deflated real prices over 98 ten-year intervals (1900-2007), report "Percentage of bets won by Ehrlich 61.2%" (Table 1, read directly from the authors' own working-paper PDF), glossing: "Contrary to the popular perception, the price history of the past 108 years shows that Ehrlich and not Simon would have won a majority of the bets and would have done so by a wide margin." Cass Sunstein's 2013 NYRB review states that Paul Ehrlich "would have won no less than 63 percent" of ten-year intervals 1900-2008, attributed only to unnamed "economists" — plausibly the same Kiel et al. analysis, so not established as independent of it. Pooley & Tupy (2020), substituting "time prices" (nominal price ÷ nominal hourly compensation) over 110 intervals (1900-2019), report the opposite winner: "When analysed with time prices, Simon wins the bet 54.2 per cent of the time" — an implied Ehrlich win rate of 45.8% under that metric.

Under the CPI-deflated construction, Julian Simon's 1990 win was a minority outcome — Kiel et al.'s own gloss puts Simon's win rate at 38.2% of years since 1900 — which is consistent with, and gives a quantitative shape to, the qualitative finding that Simon himself attributed the result partly to luck. But the three figures are not three convergent estimates of one quantity: they differ in price construction, date span, interval count, and — decisively — in which man they favor. An earlier version of this note treated 61.2% and 54.2% as points on the same "Ehrlich wins" scale; that conflated method along with outcome (see audit_status, 2026-09-15). The Kiel et al. leg now rests on a direct Tier 1 read; Sunstein's 63% still lacks an identified primary and keeps its [unverified-quant — needs primary] flag. See question-verify-simon-ehrlich-wager-resampling-primaries.