Lewis (1965): Building Cycles and Britain's Growth

Summary

Building Cycles and Britain's Growth (Macmillan, 1965, 384 pp.) is a study by British economic historian J. Parry Lewis tracing the history of building activity and its relationship to broader economic growth in Britain from roughly 1700 to 1950.[1] Like Homer Hoyt's One Hundred Years of Land Values in Chicago for the United States, it functions in the Georgist land-cycle literature as a long-run, empirical, national data source documenting that British construction and property activity did not grow smoothly but moved through recurring booms and slumps over more than two centuries.[1][2]

Fred Harrison's Boom Bust draws directly on Lewis's data, citing Table 5.1 of the book (p.74) for an average cycle duration of 17.4 years across the building cycles Lewis identifies — figures Harrison uses as UK corroboration for the same roughly-18-year periodicity that Hoyt's Chicago data documents for the US.[2] Lewis's book is cited as this wiki's discovery source for the claim, and secondary summaries describe the cycle as tracing back at least to the early 1700s, with reported cycle lengths in the literature ranging narrowly around 17–21 years.[3]

This page is drafted from secondary and discovery-source summaries, a book-listing page, and the peer-reviewed reception record (Cairncross's 1967 Economic Journal review and Barras's 2009 synthesis, cited below), not from a direct read of Lewis's primary text, which this wiki has not obtained. The exact peak/trough dates, the precise definition of "building cycle" Lewis uses, and his own stated conclusions about periodicity should accordingly be treated as second-hand — the 17.4-year average in particular is Harrison's reading of Lewis's Table 5.1, not a figure independently confirmed against the original — and could be checked against the primary source or a full academic review should a future editor gain access. Checked again 2026-07-14: the book's one Internet Archive copy (item buildingcyclesbr0000unse) is lending-restricted, and its search-inside API returns an access error without a library loan; Google Books' search-within-book index confirms the digit string "17.4" occurs exactly once in the scanned text (a weak sign the figure is not fabricated) but the record has no preview tier, so no snippet or page number is visible. This narrows but does not close the gap: the figure's existence in the book is now corroborated by an independent index, but its context, table, and exact page still are not.

Checked again 2026-07-17 — page now pinned, context still unread. Two independently digitized HathiTrust library scans of the book (University of Michigan copies, search-only tier: items mdp.39015007186649 and mdp.39015030457421) each report exactly one occurrence of "17.4" in the full text, and both place it at printed page 314 of Lewis's book. That triples the independent-index corroboration (Google Books plus two HathiTrust scans) and, for the first time, gives the figure a location in the primary text. What it does not give is context: the search-only tier shows no snippet, so whether p. 314 states the 17.4-year average as Lewis's own summary figure (p. 314 falls late in the book, where a recap table is plausible) or in some other role remains unread. Note the page number is not in tension with Harrison — his "Table 5.1, p. 74" citation locates his own table in Boom Bust, not a page of Lewis. Two further probes this pass: Barras's 2009 synthesis (searchable preview) contains no "17.4" anywhere and quotes Lewis only for methodology (its preface quotes Lewis 1965, p. 1, on the building cycle presenting "a problem of method as well as of theory"); and a second Internet Archive copy surfaced (buildingcyclesbr0000unse_b5o0, University of Victoria), under the same controlled-digital-lending restriction. The cheapest remaining unblock is a free archive.org account plus a one-hour lending loan of either copy, turning to p. 314 directly.

Relation to the Georgist Case

Lewis's book, like Hoyt's, is a historical-empirical source for the premise that construction and land markets exhibit a recurring cyclical rhythm — it does not itself argue for land value taxation or test an LVT counterfactual. Its evidentiary role on this wiki is to extend the land-cycle claim beyond the single American city (Chicago) that Hoyt studied to a second, independent national dataset (Britain, 1700–1950), which Harrison then uses to argue the pattern is not a Chicago-specific artifact.[2] As with Hoyt, readers should not treat Lewis's data as evidence that an LVT would dampen the cycle it documents — only as evidence that the cycle itself is a real, recurring, long-run historical pattern in more than one country.

Nuances and Limits

• Pre-econometric methodology. As with Hoyt's contemporaneous Chicago work, Lewis's 1965 study reflects mid-twentieth-century economic-history practice — descriptive statistics and narrative periodization — rather than modern time-series econometrics.
• Secondary sourcing. The specific figures cited here (17.4-year average, Table 5.1) come from Harrison's Boom Bust, not from an independent read of Lewis's original tables; this wiki has not yet independently verified them against the primary text.
• No LVT variation to test. Britain in this period had no land value tax regime distinct from ordinary property taxation, so — exactly as with Hoyt — the book cannot speak directly to how a taxed-land counterfactual would have altered cycle amplitude or timing.

See Also

• One Hundred Years of Land Values in Chicago — the US counterpart Lewis's UK data parallels
• 18-Year Land Cycle — the pattern this study's data feeds into
• Fred Harrison — cites Lewis's Table 5.1 directly in Boom Bust
• Boom Bust (Harrison, book page) — the wiki's page for the book that cites Lewis
• Land Speculation Causes Boom and Bust

Sources

1. J. Parry Lewis, Building Cycles and Britain's Growth (Macmillan, 1965), 384 pp. — used for the book's title, author, publisher, year, and scope (1700–1950 British building-cycle history, with forecasts to 2000) (search-indexed listing; direct fetch of a full-text or table-of-contents source was not completed in this research pass). Google Books listing
2. Fred Harrison, Boom Bust: House Prices, Banking and the Depression of 2010 (Shepheard-Walwyn, 2005), Ch. 5 §1, p.74, Table 5.1, as cited in this wiki's book page for Boom Bust and its underlying discovery report — used for the 17.4-year average cycle duration figure Harrison attributes to Lewis (B-claim; empirical, secondhand via Harrison's citation, not independently confirmed against Lewis's primary tables).
3. Alec Cairncross (1967), review of J. P. Lewis, Building Cycles and Britain's Growth (London: Macmillan, 1965), The Economic Journal 77(306), pp. 361–364. DOI: 10.2307/2229317 — a peer-reviewed academic review of Lewis's book by the economic historian Alec Cairncross; used as the citable scholarly anchor (replacing the earlier search-indexed summaries) for the book's standing and reception in the building-cycle literature. The review's bibliographic record was fetched this session; its full text sits behind the EJ paywall, so it anchors the book's scholarly standing rather than the specific 17.4-year figure, which remains sourced to Harrison.
4. M. C. Fleming (1966), review of J. P. Lewis, Building Cycles and Britain's Growth, The Economic History Review, New Series 19(2), pp. 435–436. DOI: 10.2307/2592282 — a second contemporary peer-reviewed notice of the book, located 2026-07-17; used (alongside Cairncross) to anchor the book's scholarly reception. Full text behind the JSTOR paywall; not used for the 17.4-year figure.
5. Richard Barras, Building Cycles: Growth and Instability (Wiley-Blackwell, 2009), esp. Ch. 3 ("The Nature of Building Cycles," including its survey of historical building-cycle research) and Ch. 5's UK building-cycle chronology (1785–2005). DOI: 10.1002/9781444310009 — the standard modern scholarly synthesis of the building-cycle literature, cited as contemporary context for the long-run British building-cycle research tradition of which Lewis's study is part.