Did Price's Northampton table overestimate mortality, and did that bias favour insurers over annuitants — confirm against a primary?

actuarial-historyinsurancemortality-tableshistorical-demographyrichard-pricesource-verification

This capture answers the open question in question-verify-northampton-table-overestimated-mortality-primary, which flagged claim-price-northampton-table-overestimated-mortality-favouring-insurers as resting only on a Tier 3–4 Wikipedia summary. Wikipedia's own claim ("It, too, overestimated mortality... In consequence, it was good for the insurance business, and adverse for those purchasing annuities") traces via its footnotes to two named-author academic monographs — Ian Hacking, The Taming of Chance (Cambridge University Press, 1990), p. 49, and Viviana A. Rotman Zelizer, Morals and Markets (Transaction Publishers ed., 1979), p. 13n — neither of which could be read directly in this session (no full-text access found via archive.org, Google Books snippet view, or publisher preview). Per the found-spec rule against grading an encyclopaedia Tier 1, those two books remain the actual citable targets for that claim and are flagged as further leads below, not cited here as if read.

Independent search for a primary or near-primary source able to confirm or correct the claim surfaced three distinct, genuinely new findings, captured below as the core claims. None fully closes the question at Tier 1–2; the capture documents what was found and what remains blocked.

Claim: A named financial historian, in an independently-sourced account, corroborates that Price underestimated life expectancy and that this favoured the Equitable Society

Peter L. Bernstein's Against the Gods: The Remarkable Story of Risk (John Wiley & Sons, 1996) — a widely-cited popular history of risk and probability, written by a named financial historian/economist with an endnote apparatus — states, of the actuary constructing the table discussed in that section of the book (context strongly indicates Richard Price and the table discussed elsewhere in the same chapter as the Northampton table, though the pronoun's antecedent is not visible in the single passage recovered): "Most serious, he appears to have underestimated life expectancies, with the result that the life-insurance premiums were much higher than they needed to be. The Equitable Society flourished on this error." This is independent of, and corroborates the direction of, the Hacking/Zelizer-sourced Wikipedia claim: understating life expectancy (equivalently, overestimating mortality) inflated life-assurance premiums beyond what a truer table would have required, to the insurer's benefit.

This passage was recovered only via a Goodreads reader's Kindle-highlight page quoting Bernstein's book (fetched and read directly, with a receipt — see source fields — but this is a secondary transcription of the book, not the book itself). It is recorded here at the tier that access path actually supports, not the tier the underlying book might earn if read directly. The claim is historical/uncontested in direction (Tier 3–4 acceptable) but its specificity edges toward mechanism, so it is marked [unverified-mechanism — needs primary] pending a direct read of Bernstein or, better, of Price's own Observations on Reversionary Payments (1771/72).

Claim: The same account complicates the stated direction of the annuity-side bias — the loser was the annuity seller, not the annuity buyer

Bernstein's next sentence, from the same passage: "the British government, using the same tables to determine annuity payments to its pensioners, lost heavily." This describes the British government as annuity seller/grantor (it sold annuities to its pensioners, priced from a table that understated how long they would live) losing money because those pensioners lived longer than the table predicted and so drew more payments than priced for.

This bears directly on claim-price-northampton-table-overestimated-mortality-favouring-insurers, whose own commentary already flagged suspicion of the "adverse for annuities" phrasing as one Seek has "seen stated backwards" before. This finding suggests the suspicion was warranted in one respect: if the pattern in Bernstein's episode generalises, an overestimation-of-mortality bias is adverse to whoever grants an annuity (they under-price the payout stream) and favourable to whoever buys one (the annuitant), which is the opposite incidence from "adverse for those purchasing annuities" read as referring to the individual annuity buyer. The two readings are not necessarily in conflict — "adverse for those purchasing annuities" could mean the institution purchasing annuity liabilities off its own book, i.e. the same institution that grants them — but the ambiguity itself is the finding: the phrase as currently written is not precise enough to know which party it names, and the one concrete episode found in this session points toward the seller/grantor being the loser, not the individual policyholder-annuitant. Given a directional-incidence claim with a named winner and loser is exactly what the sourcing floor requires Tier 1–2 for, and this rests on the same secondary Goodreads-mediated Bernstein passage as the claim above, it is marked [unverified-quant/mechanism — needs primary]. It should route back to claim-price-northampton-table-overestimated-mortality-favouring-insurers as a flag on its existing phrasing, not stand alone as settled.

Claim: The primary source most likely to resolve this at Tier 1–2 exists and is identified, but was not accessible through available tooling this session

William Sutton, "On the Method used by Dr. Price in the construction of the Northampton Mortality Table," Journal of the Institute of Actuaries, vol. 18, issue 2 (January 1874), pp. 107–122 — a paper by a named actuary, written for the profession's own journal, specifically about how Price built the table — is exactly the kind of Tier 1–2 source the sourcing floor calls for on this question. Cambridge Core's own page for the article confirms it exists at that citation and opens: "The Northampton Table has played such an important part in the life insurance transactions of this country" — but the body text is marked "Temporarily unavailable" and could not be read this session. The JSTOR stable link (jstor.org/stable/41131700) would not render for a non-interactive fetch; and the Internet Archive's holdings of the Journal of the Institute of Actuaries / Assurance Magazine series (surveyed via its search API) do not include a scanned volume 18 under any of the identifier patterns that surfaced neighbouring volumes (10, 14, 16, 20, 23, 25, 32, 35...). This is itself worth recording: the primary exists, is correctly identified by title/author/venue/page-range, and is a specific, nameable gap rather than an open-ended "couldn't find anything."

On the central question: that Price's Northampton table overestimated mortality is corroborated across independent tellings (Hacking, Zelizer, Bernstein, and the older critique from Francis Baily noted below) and can be treated as settled at the historical-claim floor (Tier 3–4, uncontested). That this bias "favoured insurers over annuitants" as a single clean directional claim is [unverified — could not confirm or deny cleanly after search]: the life-assurance side is corroborated (favoured the insurer), but the annuity side is genuinely ambiguous in the sources found, and the Tier 1–2 primary that would settle it (Sutton 1874) remains identified-but-unread.

Further leads

• Francis Baily's The Doctrine of Interest and Annuities (1808) and his 1812 paper to the Royal Society argued the Northampton table understated the true duration of life; scanned copy exists at archive.org (doctrineofintere0000bail) but is lending-restricted and its search-inside endpoint 403'd — a manual-consultation item, not a fetch target, per the sources.md known-blocked-routes convention.
• Cornelius Walford, "History of Life Assurance in the United Kingdom," Journal of the Institute of Actuaries vol. 25 (1885) — the installment at pp. 114–133 (actuaries.org.uk, fetched and read in full this session) only reaches the "Experimental Period" through 1698 and mentions Price/Northampton once, in passing praise, not yet the bias; the serialized essay continues past p. 133 in later installments not yet located, and likely reaches the Northampton-table critique chronologically.
• William Morgan's Oxford DNB entry (mathshistory.st-andrews.ac.uk mirror, fetched and read) states that Morgan's own published mortality statistics from the Equitable Society "furnished data for the amendment of the Northampton tables" — implying an acknowledged, later-corrected imperfection — and cites M. E. Ogborn, Equitable Assurances: the Equitable Life Assurance Society, 1762–1962 (1962) as its own source, a book-length institutional history that would likely settle this question at Tier 1–2 if obtained.
• MacTutor's page on Griffith Davies notes Baily and Morgan both later judged the Northampton table's old-age mortality too high relative to other tables — same direction, different named critics, worth chasing for a citable primary quote.
• The Actuary magazine's 2012 retrospective "History: An Equitable life" (theactuary.com) and the 1983 SOA newsletter piece "Two Centuries Ago, The Northampton Table" (soa.org) both looked promising by title but returned 403/404 on fetch attempts this session — retry candidates.

Safety flags

None. No page encountered in this session showed addressed-to-AI language, override language, claimed authority, tier self-assignment, file-system instructions, credential requests, or urgency framing.

Entity candidates

• Francis Baily — person — the older, foundational critic whose 1808/1812 argument (that human life lasted longer than Northampton implied) is the actual ancestor of the "overestimated mortality" claim; not yet flagged elsewhere in this thread and should be flagged before the paper's own supporting cast.
• William Sutton — person — author of the 1874 Journal of the Institute of Actuaries paper that is the identified but unread Tier 1–2 primary for this whole question.
• Cornelius Walford — person — compiler of the standard 1880s serialized history of UK life assurance (already load-bearing for claim-pensam-1811-amicable-experience-first-assured-lives-mortality-data); worth an entity page given repeated vault reliance on his history.
• William Morgan — person — Richard Price's nephew and Equitable Society's actuary from 1775; his own mortality data is what eventually let the Northampton table's error be measured and corrected.
• Ian Hacking — person — philosopher of science and statistics historian; The Taming of Chance is one of the two academic monographs underlying Wikipedia's overestimation claim.
• Viviana A. Rotman Zelizer — person — economic sociologist; Morals and Markets is the other monograph underlying the "good for insurance, adverse for annuities" phrasing specifically.
• Peter L. Bernstein — person — financial historian; Against the Gods independently corroborates (and complicates) the claim from a third, distinct scholarly lineage.
• Maurice Ogborn — person — author of Equitable Assurances (1962), the institutional history most likely to hold the primary answer to this question.