Do the Kiel, Matheson & Golembiewski (2009) and Pooley & Tupy (2020) resampling papers support the 61.2% and 54.2% win-rate figures attributed to them?

population-policyforecastingsimon-ehrlich-wagerstatisticsverificationresamplingunverified-quant-resolved

Both underlying papers were located and read directly this session — the Kiel, Matheson & Golembiewski working paper via its own institutional PDF host (hcapps.holycross.edu, TLS-verified), and the Pooley & Tupy piece via Human Progress, the project both authors run and publish through directly (cross-checked against an identical Gale Pooley Substack repost). This resolves question-verify-simon-ehrlich-wager-resampling-primaries — with a correction, not a simple confirmation. The 61.2% figure checks out exactly as attributed. The 54.2% figure is real but has been circulating in this vault (and in the HumanProgress secondary-summary chain feeding claim-three-resamplings-estimate-ehrlich-would-have-won-54-to-63-percent-of-intervals) attached to the wrong side of the bet.

Claim: Kiel, Matheson & Golembiewski (2009) directly supports "Ehrlich would have won 61.2% of ten-year intervals, 1900-2007"

Claim type: quantitative. Floor: Tier 1-2 required — met (Tier 1, author's own institutional working-paper server, PDF read directly). verifies: question-verify-simon-ehrlich-wager-resampling-primaries

The paper's Table 1 ("Results of the Ehrlich-Simon Bet - 1900-2007") reports, for 98 ten-year intervals: "Percentage of bets won by Ehrlich 61.2%" with an "Average return on bet for Ehrlich" of 10.5%. (The same table also reports a 25-year-interval version: 59.0% of 83 intervals, average return 13.8% — not the figure in question, but from the identical dataset and method.) The paper's own methodology section states the data source as U.S. Geological Survey nominal prices for the five metals in the original bet (chrome, copper, nickel, tin, tungsten), deflated to real prices using the CPI (with the pre-1913 span deflated using McCusker 2001 estimates), tracked over rolling 10-year and 25-year windows starting from 1900. The paper's text glosses the finding as: "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." The 61.2% figure attributed to this paper elsewhere in the vault is exactly what the primary source states — fully confirmed.

Claim: Pooley & Tupy (2020) do not support "Ehrlich would have won 54.2% of the time" — the 54.2% in their paper is Simon's win rate, under a different price metric

Claim type: quantitative / technical-mechanism (which side of the bet a figure describes is load-bearing here). Floor: Tier 1-2 required — met (Tier 1, authors' own venue, read directly, identical text cross-confirmed on a second author-controlled host). verifies: question-verify-simon-ehrlich-wager-resampling-primaries

Pooley & Tupy's own text is unambiguous: "When analysed with time prices, Simon wins the bet 54.2 per cent of the time." This is Simon's win rate, not Ehrlich's, and it rests on a different price construction than Kiel et al.'s figure: "time prices," defined as "nominal prices divided by nominal hourly compensation" (using blue-collar hourly wage data from measuringworth.com), rather than CPI-deflated real prices. The period is also different — "110 ten-year intervals" spanning 1900-2019, versus Kiel et al.'s 98 intervals spanning 1900-2007. Under this alternate metric, Ehrlich's implied win rate is 45.8% (100% − 54.2%), not 54.2%. The vault's existing claim-note (claim-three-resamplings-estimate-ehrlich-would-have-won-54-to-63-percent-of-intervals) and the question it fed both treated 54.2% as belonging to the same "Ehrlich would have won X%" family as the Sunstein 63% and Kiel 61.2% figures; that framing does not survive contact with the primary text. Pooley & Tupy's own paper, using their own preferred metric, finds Simon — not Ehrlich — the more frequent winner across their study period.

Claim: the 61.2% and 54.2% figures are not measurements of the same question — they differ in price metric, time span, and interval count

Claim type: technical-mechanism (methodology comparison). Floor: Tier 1-2 required — met (both figures drawn from the primary texts above). verifies: question-verify-simon-ehrlich-wager-resampling-primaries

Kiel, Matheson & Golembiewski (2009): CPI-deflated real prices, 1900-2007, 98 ten-year intervals, reports Ehrlich's win rate (61.2%). Pooley & Tupy (2020): "time prices" (nominal price ÷ nominal hourly wage), 1900-2019, 110 ten-year intervals, reports Simon's win rate (54.2%) — explicitly framed in the paper as "two modifications to the Kiel et al. methodology, namely using time prices and the war clause." Pooley & Tupy also report a third figure from the same dataset with war years excluded: Simon wins 69.9% of 73 remaining intervals. None of these three numbers (61.2%, 54.2%, 69.9%) is answering an identical question; they are three different resamplings of a related but non-identical construct (real-price win rate vs. time-price win rate vs. time-price win rate excluding war years), over two different date ranges. Treating 61.2% and 54.2% as two data points on the same scale, as a prior vault claim-note did, conflates method along with outcome.

Further leads

• Pooley & Tupy's war-clause-excluded figure: Simon wins 69.9% of 73 non-war ten-year intervals, 1900-2019, average return 18.0% (same source as above; not independently notable enough to be a core claim here, but a further sensitivity result worth its own note if the war-clause methodology itself becomes a topic).
• Kiel et al. explicitly position their paper as contradicting two earlier, less rigorous analyses: McClintick & Emmett (2005, PERC Reports 23(3):16-17), which found "the price history of the twentieth century provides evidence that [Simon] would have won five of the ten decades by large margins, and he would have won a bet over the entire century," and Perry (2008, Carpe Diem blog), which concluded Simon would have won a hypothetical 1990-2000 rematch. Neither earlier source was read this session — worth a direct read to characterize exactly how limited their "decade-by-decade" method was relative to Kiel et al.'s full rolling-window resampling.
• Lawn (2010), "On the Ehrlich-Simon bet: Both were unskilled and Simon was lucky," Ecological Economics 69(11):2045-2046 — a published critical response to Kiel et al. in the same journal; not read this session.
• Later citers of this literature already noted in RePEc's citation list: Stan Becker (2013, Demography) and Vincent Geloso (2022, Review of Austrian Economics) — not read this session, potential further corroboration or complication.
• Cass Sunstein's NYRB "63 percent" figure (see claim-three-resamplings-estimate-ehrlich-would-have-won-54-to-63-percent-of-intervals) still has no identified specific academic source distinct from the two papers checked here — Sunstein's review does not appear to cite either paper by name in the portion previously captured. Unresolved; a further lead, not answered by this capture.
• Both papers' true journal-of-record venues (Ecological Economics via ScienceDirect/Elsevier, and Economic Affairs via Wiley) remain paywalled and were not independently accessed; this capture relies on the authors' own freely-hosted versions (Holy Cross's own PDF server for Kiel et al.; Human Progress, which Pooley and Tupy run, for Pooley & Tupy), which match the published citations' titles, authors, and venue metadata exactly.

Entity candidates

• Mark J. Perry — person — economist (AEI/Carpe Diem blog); his 2008 hypothetical-rematch analysis is one of the two earlier, narrower studies that Kiel, Matheson & Golembiewski explicitly say they "extend" and reach "a distinctly different conclusion" from — the foundational prior work this capture's core claim is measured against.
• D. McClintick & Ross B. Emmett — person/concept — authors of the 2005 PERC Reports piece concluding Simon would have won "a bet over the entire century"; the other prior analysis Kiel et al. explicitly contradict.
• Katherine A. Kiel — person — co-author of the 61.2% resampling (College of the Holy Cross economist).
• Victor A. Matheson — person — co-author of the 61.2% resampling (College of the Holy Cross economist, sports/public economics).
• Kevin Golembiewski — person — co-author of the 61.2% resampling.
• Gale L. Pooley — person — co-author of the 54.2%/time-price resampling; economist at Utah Tech University, board member of HumanProgress.org.
• Marian L. Tupy — person — co-author of the 54.2%/time-price resampling; Cato Institute senior fellow, editor of HumanProgress.org.
• Time price (economic concept) — concept — the nominal-price-over-nominal-wage metric Pooley & Tupy substitute for CPI-deflated real prices; central to why their 54.2% figure isn't directly comparable to Kiel et al.'s 61.2%, and recurs across Pooley/Tupy's broader body of work (e.g. Superabundance).