High wages & induced innovation

Thesis

Inventions that succeed are inventions that pay. The reason machines that replace labor with capital and energy were invented and adopted in 18th-century Britain — and not in equally talented France, Germany, India, or China — is that Britain had a peculiar factor-price structure: unusually high wages (in silver and grain terms) and unusually cheap energy (coal). Mechanization is profitable when the wage you save exceeds the capital and fuel cost of the machine. Almost everywhere outside Britain, hand-spinning and hand-weaving were cheaper than the early water frame and Arkwright’s machinery. In Britain, they were not. The same calculation extended to the steam engine, the puddling furnace, and the railway.

In this view, the IR is not a story about who invented things but where it paid to use them. Inventors were everywhere; the British market was the only one that could absorb their output and fund their refinement.

The high-wage logic is the demand-side counterpart to the upper-tail human-capital supply-side story, and the two are the central rivalry in the post-2010 literature: was British labour expensive (Allen) or merely productive (Kelly–Mokyr–Ó Gráda)?

Lead proponents

Robert Allen — The British Industrial Revolution in Global Perspective (2009) is the canonical statement, building on a decade of prior papers reconstructing real wages across European and Asian cities. Allen’s two prior moves underwrite the book: his “great divergence in wages and prices” work (e.g. Explorations in Economic History, 2001) built the comparable silver- and welfare-ratio wage series, and his theory of induced innovation formalized the claim that the direction of technical change — labour-saving vs. capital- or land-saving — responds to relative factor prices, reviving a Hicks–Habakkuk tradition with hard price data.
John Habakkuk (antecedent) — American and British Technology in the Nineteenth Century (1962) is the intellectual ancestor: Habakkuk argued that American labour scarcity relative to land drove the US toward labour-saving mechanization faster than Britain. Allen inverts and globalizes the logic — Britain was the high-wage economy relative to Asia and most of the Continent — but the induced-innovation skeleton is Habakkuk’s.
Joel Mokyr, Jan Luiten van Zanden, Jaime Reis (qualified fellow-travellers) — none endorses the strong price-primacy claim, but each accepts that the British factor-price configuration was real and shaped adoption. Van Zanden’s work on the European “little divergence” in wages, and the broader European real-wage reconstructions, supply the comparative baseline against which Allen’s London-is-expensive claim is read.
Critically engaged contributors — much of the position’s current shape is set by its critics, whose work is now part of the literature on it: Judy Stephenson on the wage-data foundations, Jane Humphries on the spinning sector, and Morgan Kelly & Cormac Ó Gráda on efficiency-unit wages and the within-England geography (see Major critiques below).

Key arguments

Britain had high wages. Allen’s reconstructed real-wage series (Allen wage series) shows that in 18th-century London (and to a lesser degree Amsterdam), an unskilled male laborer’s day wage bought roughly 2–3× the basket of subsistence goods that the same wage bought in Paris, Vienna, Beijing, or Delhi. This was not a 19th-century artifact; it preceded the IR and helps explain it.
Britain had cheap energy. Per the coal & resource geography story: London coal was a fraction of the price (per unit of energy) of charcoal or peat in continental cities.
The wage/energy ratio is what matters for mechanization. A spinning jenny, a Watt engine, or a coke-iron furnace is worth installing when (wages saved) > (capital cost + energy cost). In high-wage / cheap-energy Britain, these inequalities held early. In low-wage France, India, or China, they held much later, if at all.
The pattern of British innovation matches. The IR’s signature inventions — water frame, mule, steam engine, coke iron, puddling — were almost all labor-saving and energy-using. They are exactly what the relative price story predicts.
Adoption lags trace the price ratio. Allen tracks the diffusion of British innovations to the Continent and shows that uptake correlates with the local wage/energy ratio — French manufacturers adopted British machinery only as French wages rose and energy costs fell, not when the machinery itself became “available.”

The spinning-jenny calculation

Allen’s most-cited single exhibit is a microeconomic profitability calculation for the spinning jenny, because it converts the abstract factor-price story into a concrete adopt/don’t-adopt decision. The jenny was a cheap wooden machine (a few pounds to build), so its adoption turned almost entirely on the wage it saved relative to hand-spinning. Allen’s reconstruction (Explorations in Economic History, 2009, “The Industrial Revolution in Miniature”) computes the rate of return to installing a jenny at local wage and capital-cost levels around 1780. In Britain the return was high — perhaps 38% — because British spinning wages were high enough that the labour saved paid for the machine quickly. In France the return was around 2–4%, far below the threshold that would induce a workshop to switch; in India, where spinning labour was a fraction of the British wage, the calculation was deeply negative. The same logic was meant to extend up the chain to the water frame, the mule, Cort’s puddling-and-rolling process, and ultimately the Watt engine: each is profitable to install only where the wage saved exceeds capital plus fuel, and Britain’s high-wage/cheap-coal configuration crossed that line first.

The jenny calculation is also where the critics concentrate their fire, because the wage figures that drive the return — the spinning wage in particular — are exactly the numbers Humphries & Schneider dispute (below). If British spinning was low-paid female and child work rather than the relatively high wage Allen’s basket implies, the 38% return shrinks and the central exhibit weakens.

The welfare-ratio method and the Asian comparison

The empirical claim “Britain was a high-wage economy” depends entirely on how wages are made comparable across centuries and continents, and Allen’s method is itself part of the argument. He converts nominal day-wages into silver wages (grams of silver per day, to compare across currencies) and into welfare ratios — the number of subsistence “baskets” a year’s full-time earnings would buy, where a basket is a fixed bundle of bread or grain, a little meat and oil, fuel, cloth, and lighting calibrated to a bare-bones annual consumption of ~1,940 calories per adult-equivalent. On the welfare-ratio metric, London and Amsterdam labourers around 1700–1750 could afford several baskets per year (a comfortable margin above bare subsistence), while labourers in Leipzig, Florence, Beijing, and Delhi hovered near one — at or barely above the subsistence floor. The London-versus-Leipzig silver-wage gap, visible in the Allen wage series, is roughly 3:1 across the 17th–18th centuries.

The Asian comparison is doing heavy lifting and is therefore heavily scrutinized. Against the California school’s claim of rough living-standard parity between the most advanced regions of England and China c.1750, Allen’s reconstructed welfare ratios place Yangzi-delta and Beijing labourers well below their London counterparts. The dispute turns on the content of the basket — whether a rice-and-vegetable Chinese consumption norm is being unfairly priced against a bread-meat-beer European one — and on the quality of the underlying Chinese and Indian wage data, which are far thinner than the European series. Allen’s defence is that the silver-wage comparison (currency-neutral, basket-independent) shows the same ordering, and that the ordering is what the inducement argument needs: machinery was profitable to install where labour was dear in whatever the locally relevant terms, and British labour was dear in all of them.

The diffusion test

The position’s cleanest comparative prediction is about timing: British innovations should diffuse to the Continent as continental wage/energy ratios rise toward British levels, not when the machinery becomes physically available. The observed adoption lags broadly fit — French cotton mechanization scaled only from the 1810s–20s, German from the 1830s–50s, and these lags track the slow convergence of continental wages and the fall of continental coal and machine costs rather than mere knowledge transfer (British machinery and British mechanics were available abroad far earlier; see the upper-tail position’s “British workmen abroad” pattern). The diffusion evidence is among the more robust planks because it does not depend on the contested level of any single British wage series, only on the cross-country ordering.

Key evidence

The Allen real-wage series (dataset) — silver and grain wages for unskilled workers in ~10 European cities and several Asian cities, 1500–1900. The single most-cited dataset in modern IR debates.
City-level wage ratios — e.g., the silver wage of a London laborer relative to a Beijing or Delhi laborer, ~1700: roughly 4:1.
Profitability calculations for individual machines — Allen’s worked examples for the spinning jenny, the Newcomen engine, and the coke-blast furnace, showing the British internal rate of return positive at British prices and negative at French prices through the 1770s.
Diffusion lags — French cotton mechanization waited until ~1815; German until ~1850. The lag tracks wage convergence.

Beyond textiles: the macro-invention claim

Allen extends the inducement logic beyond spinning to the IR’s other signature technologies, and this is where the thesis becomes a general theory rather than a sectoral observation. The Newcomen and Watt steam engines were coal-hungry and capital-intensive — viable first at the pithead, where coal was nearly free, and only later economic where coal had to be carted; the wage saved by replacing horse-gins and hand-pumps had to clear the engine’s capital and fuel cost, which it did first in high-wage, cheap-coal Britain. Cort’s puddling-and-rolling process (1783–84) substituted coke and capital for the labour and charcoal of older iron-making. Allen’s narrative is that all three of the IR’s “macro-inventions” share the same factor-bias — labour-saving, energy- and capital-using — because they were selected for by the same relative-price gradient. The contrast with John Habakkuk’s America sharpens the point: nineteenth-century America, even more labour-scarce relative to land than Britain, pushed further and faster toward labour-saving mechanization (interchangeable parts, the “American system of manufactures”) — exactly as an induced-innovation theory predicts when the factor-price signal is stronger still. The strong version of Allen’s claim is thus that the direction of the whole technological cluster, not merely its adoption, was set by British factor prices.

Major critiques

— STEPHENSON: the wage data overstate worker earnings. Stephenson (2018) reconstructed the actual disbursement records behind Allen’s London building-trades wages and found that the day-rates he used were contractor fees paid to master builders, not the take-home earnings of the labourers themselves — the masters skimmed a margin, so the figures overstate what workers actually received by perhaps 20–30%. Because the London building wage is the spine of Allen’s “Britain is the high-wage economy” claim, a downward correction there propagates through the welfare ratios and the profitability calculations. The exchange is the most consequential live empirical front around this position.

— HUMPHRIES & SCHNEIDER: the lead sector was low-wage. Humphries & Schneider (2019) took the dispute to the place it matters most — hand spinning, the activity the first machines actually replaced and overwhelmingly a female and child occupation. Reconstructing spinners’ earnings across 1500–1800, they find spinning was a persistently low-paid activity with no rising-wage trend in the decades before the jenny and water frame. Their conclusion inverts the thesis: “the route to mechanization and factory production was a response to low not high wages.” Allen has responded defending the broad framework while conceding particulars and disputing the spinning-wage reconstruction; the disagreement turns on fragmentary piece-rate evidence and is unresolved.

— KELLY, MOKYR & Ó GRÁDA: wages were high because labour was productive, not expensive. The canonical Precocious Albion (2014) argues that British wages were not unusually high per efficiency unit once you adjust for skill and human capital. British workers around 1750 were physically larger, better-nourished, more literate, and more mechanically skilled than continental counterparts — products of the Poor Law and an unusually open apprenticeship system. What Allen reads as “expensive labour” they read as highly productive labour, which reroutes causation through the supply of skilled labour rather than the price of labour. If they are right, the inducement mechanism loses its exogenous shock.

— KELLY, MOKYR & Ó GRÁDA (within-England): the geography runs backwards. Their 2023 JPE paper tests a sharp implication: if high wages induced mechanization, the counties that textile-industrialized first should have had the highest pre-IR wages. Regressing the 1831 share of male textile employment across England’s 41 counties on pre-IR wages and pre-IR mechanical-skill supply yields elasticities of roughly −6 on wages and +2 on skills (joint R² ≈ 0.7) — high pre-IR wages were a disincentive to textile mechanization at the county level. Real wages then rose in the industrializing (low-wage) north and collapsed in the previously prosperous south, reversing the pre-IR geography. Allen’s defence is that the within-England test does not bear on the between-country story (the paper concedes English wages were high relative to Asia and most of the Continent) — defensible, but it concedes that the within-country test runs against the intuitive form of the thesis.

— From the Industrial Enlightenment school: the supply-of-inventors gap. Even granting the demand-side story, Allen is famously thin on where the inventions came from. The high-wage thesis explains why Britain adopted labour-saving machinery; it does not explain why Britain was the place that generated it. Mokyr’s useful-knowledge account is essentially a critique of this silence.

— Counter-examples: several non-British high-wage cities (Antwerp, Amsterdam) did not industrialize first despite favourable-looking wage/energy ratios. Defenders insist the British coal-plus-wages combination was unique; critics read the accumulation of necessary conditions (high wages and cheap coal and skill and institutions) as a sign that no single factor, the price ratio included, is doing the explanatory work the strong thesis assigns it.

Status

Mainstream, with caveats. The basic claim that factor prices shaped the adoption of mechanization is broadly accepted. The strong form — that prices were the primary cause of the IR — depends on the contested wage-data work and on dismissing the supply-side story. Most modern syntheses (Mokyr, Broadberry, Crafts) treat Allen’s framework as one of two or three required ingredients, not the whole story.