«by Robert C. Allen Professor of Economic History Department of Economics and Nuffield College Oxford University Email: ...»
And yet the French not only failed to invent mechanized spinning, they did not adopt it even when it was freely available. This was not for lack of knowledge. John Holker was an English Jacobite, who fled to France in 1750 where he established himself as a cotton manufacturer. In 1754, he succeeded in being appointed Inspector General of Foreign Manufactures charged with importing successful foreign technology. In 1771 he sent his son to Lancashire to report on the new machines, and his son brought back a jenny. This was copied and made available to French producers; indeed, the state subsidized its use. It was installed in some large scale factories but was otherwise ignored by the cotton trade. In 1790, there were about 900 jennies in France compared to 20,000 in England (Aspin and Chapman 1964, p. 49). The disproportion was at least as great with water frames. About 150 large scale mills were in operation in Britain in the late 1780s. In France, there were only four and several of these were extremely small and not representative of British practice. (Wadsworth and Mann 1931, pp. 193-208, 503-6, Chapman and Butt 1988, pp. 106-11).
Why did the French ignore the new spinning machines? Cost calculations for France are not robust, but the available figures indicate that jennies achieved consistent savings only at high count work, which was not the typical application (Ballot 1923, pp. 48-9). In France, a 60 spindle jenny cost 280 livre tournois in 1790 (Chassagne 1991, p. 191), while a labourer in the provinces earned about three quarters of a livre tournois per day, so the jenny cost 373 days labour. In England, a jenny cost 140 shillings and a labourer earned about one shilling per day, so the jenny was worth 140 days labour (Chapman and Butt 1988, p. 107). In France, the value of the labour saved with the jenny was not worth the extra capital cost, while in England it was. French cost comparisons show that Arkwright’s water frame, a much more capital intensive technique, was no more economical than the jenny. The reverse was true in England where water frames were rapidly overtaking jennies. The French lag in mechanization was the result of the low French wage.
Global competition was the impetus to invent mechanical spinning. The result was a biased technical improvement that benefited Britain with its high wage economy much more than continental producers like France.
Why the British rather than the French invented mechanical spinning
As we have indicated, both the jenny and the water frame required considerable expenditures in R&D to make them work. The same would have been true in France. Would these expenditures have been worthwhile in France? No–mechanized spinning brought no economic benefit there in view of the low wage. We need look no further to understand why the spinning jenny and the water frame were invented in England rather than France or, indeed, most other parts of the world.
Steam engine An idea from science The steam engine presents a variation on the theme. Big Ideas did not have much to do with coke smelting or mechanized spinning, but the low pressure steam engine, developed by Newcomen and improved by Watt, was the best example of a scientific spin-off in the eighteenth century. It was based on the idea that the atmosphere had weight, which was a seventeenth century discovery and a hot topic in experimental physics. Even in this case, however, economic incentives were a key to the application of this new knowledge. Without the British coal industry, the steam engine would not have been developed.
The link from science to the steam engine was direct. The science began with Galileo, who discovered that a suction pump could not raise water more than about 34 feet–despite a vacuum existing above the column of water that had been drawn up to that height. Aristotle had said that nature abhorred a vacuum but only, it seemed, for 34 feet! Galileo suggested to Evangelista Torricelli, his secretary, that he investigate this problem. In 1644 Torricelli inverted a glass tube full of mercury and placed its bottom in a bowl of mercury. The mercury stabilized in the tube forming a column 76 centimeters high with a vacuum above it.
This was the world’s first barometer, and Toricelli concluded that the atmosphere had weight and pushed the mercury up the column. This was confirmed in 1648 by placing the barometer in a larger container and pumping the air out of it–the column of mercury collapsed and then reappeared as air was readmitted into the larger container.
A particularly important set of experiments was performed in Magdeburg by Otto von Guericke. In 1655, he put two hemispheres together and pumped the air out of the space they enclosed. It took sixteen horses to pull them apart. In another portentous experiment in 1672, von Guericke Figure 12 found that if the air was pumped out of cylinder A (Figure 12), the weights D rose as the atmosphere pushed the piston down into the cylinder. Evidently, the weight of the air could perform work.
This idea had been anticipated by Christian Huygens in 1666 who used exploding gun powder to drive a piston up a cylinder. When it reached the top, the gases from the explosion were released creating a vacuum. Air pushed the piston down and raised the load. This design was not effective. However, his assistant, Denis Papin, realized that filling the cylinder with steam and then condensing it accomplished the same purpose. In 1675, Papin built the first, very crude steam engine.
The first practical application of steam technology Figure 13 was Savery’s steam vacuum pump patented in 1698. It created a vacuum by condensing steam in a reservoir; the vacuum then sucked up water. The purpose of Savery’s devise was draining mines, but it was not widely used, and it was not a steam engine.
But still an R&D project The first successful steam engine was invented by Thomas Newcomen.9 Like Savery’s device, it was intended to drain mines. Newcomen’s engine applied the discovery that the atmosphere has weight. That application required a major R&D project, and that project meant that the invention was an economic commitment as well as a scientific spin-off.
Newcomen’s design (Figure 13) was suggested by von Gierecke’s apparatus: First, replace the weights with a pump (I). Second, construct the ‘balance beam’ so it is slightly out of balance and rests naturally with the pump-side down (H). Then, if a way were contrived to create a vacuum in the cylinder (B), air pressure would depress the piston (E) and raise the pump. Next, if air were reintroduced into the cylinder, the vacuum would be eliminated and the pump would drop since the beam is slightly out of balance. Finally, recreating the vaccum would raise the pump again since the pressure of the atmosphere would again depress the piston. Thus, creating a vacuum and relieving it raises and lowers the pump. This apparatus becomes a ‘steam engine’ when steam is made by boiling water (A) and drawing it into the cylinder when the piston is raised, and the vacuum is created when cold water is injected into the cylinder (B) to condense the steam. This is a low pressure engine since it is not steam pressure that pushes the piston up: the point of the steam is simply to provide a gas that fills the cylinder and which is condense to create the vacuum. At the heart of the Newcomen engine was seventeenth century science.
While the Newcomen engine differed from other eighteenth century inventions in its scientific basis, it was similar in the engineering challenges it posed. Twentieth century engineers who have built Newcomen engines have found it to be tricky and difficult to make them actually work (Hills 1989, pp. 20-30). That Newcomen could resolve the engineering problems was a remarkable achievement. He began experimenting around 1700 and apparently built an engine in Cornwall in 1710, two years before his famous engine at Dudley.
In this decade of R&D, Newcomen learned many things. He discovered by accident that the steam could be condensed rapidly if cold water was injected into the cylinder (B). He found that the water supply tank (L) for the injector worked best if it was placed at the top of the engine house, so the injection water entered the cylinder at high pressure and volume.
The pipe (R) that drained the condensed water from the cylinder had to run far enough down into a hot well (S), so that atmospheric pressure could not force condensed water back into Recent work on the development of the steam engine includes Hills (1989), Nuvolari (2004a), von Tunzelman (1978).
the engine. The top of the cylinder had to be sealed with a layer of water–nothing else worked. The dimensions of the balance and the weights of the engine’s piston and the pump (K) had to be coordinated for smooth operation. Linkages between the beam and the valves had to be designed so that they would open and shut automatically at the correct moments in the cycle. No wonder it took Newcomen ten years to create an operating engine. It was a time consuming and expensive undertaking.
Like many practitioners of R&D, Newcomen hoped for a pay-off through patenting his creation. In this he was frustrated because the Savery patent was extended 21 years to 1733 and construed to cover his very different engine! Newcomen was forced to do a deal with the Savery patentees to realize any income at all.
A biased technical improvement that favoured the British
R&D costs mean that the link between Galileo and Newcomen was mediated by economics. Scientific curiosity and court patronage may have been reason enough for Torricelli, Boyle, Huygens and other scientists to devote their time and money to studying air pressure (David 1998), but Newcomen was motivated by prospective commercial gain. What was that gain? The object of the engine was to drain mines, so the demand for the technology was determined by the size of the mining industry. In 1700, England’s lead was immense: It produced 81% of the tonnage in Europe and 58% of the value. Germany, which had been Europe’s mining centre in the late middle ages, produced only 4% of the tonnage and 9% of the value in 1700. The change was all down to coal. Servicing the drainage needs of England’s coal industry is one reason why steam engine research was carried out in England.
Coal mattered for a second reason as well. There were alternative ways of powering pumps–water wheels or horse gigs–so there was effective demand for steam power only if it was cost-effective. The early steam engines were profligate in their consumption of fuel, so they were cheap sources of power only if fuel was remarkably cheap. Desaguliers (1744, II,
pp. 464-5), an early enthusiast of steam power, put the matter succinctly:
But where there is no water [for power] to be had, and coals are cheap, the Engine, now call’d the Fire-Engine, or the Engine to raise Water by Fire, is the best and most effectual. But it is especially of immense Service (so as to be now of general use) in the Coal-Works, where the Power of the Fire is made from the Refuse of the Coals, which would not otherwise be sold.
The Newcomen engine was a biased technological improvement that shifted input demand away from animal feed and towards combustible fuel.
Free fuel overcame high fuel consumption, but, by the same token, the energyintensity of the Newcomen engine restricted its use to the coal fuels. Since most of the coal mines were in Britain, so were most of the engines. At the expiry of the Savery-Newcomen patent in 1733, there were about 100 atmospheric engines in operation in England. By 1800, the total had grown to 2500 in Britain of which 60 - 70% were Newcomen engines.10 In Kanefsky and Robey (1980, p. 171). The uncertainty depends on how one classifies the engines of unknown type. As the production of Watt engines is reasonably well established, the unknown engines were probably Newcomen, and that choice yields the higher contrast, Belgium, with the largest coal mining industry on the continent, was second with perhaps 100 engines in 1800.11 France followed with about 70 engines of which 45 were probably Newcomen (installed mainly at coal mines) and 25 were Watt. The first steam engine was installed in the Netherlands in 1774, in Russia in 1775-7, and in Germany at about the same time. None seem to have been installed in Portugal or Italy (Redlich 1944, p.
122, Tann 1978-9, p. 548, 558). The Newcomen engine “was adopted in numbers only in the coal fields...The machines were, until well into the 19th century, so symbolically linked to the coal-fuel matrix in which they had come to maturity that they could not readily pass beyond its limits” (Hollister-Short 1976-7, p. 22). The diffusion pattern of the Newcomen engine was determined by the location of coal mines, and Britain’s lead reflected the size of her coal industry–not superior rationality.
Why the steam engine was invented in Britain rather than France or China
Moreover, the diffusion pattern of the Newcomen engine indicates that it would not have been invented outside of Britain during the eighteenth century. Non-adoption was not due to ignorance: The Newcomen engine was well known as the wonder technology of its day. It was not difficult to acquire components, nor was it difficult to lure English mechanics abroad to install them (Hollister-Short 1976). Despite that, it was little used. A small market for engines implied little potential income for a developer to set against the R&D costs. The benefit-cost ratio was much higher for Newcomen than for any would-be emulator on the continent. Newcomen had to know about the weight of the atmosphere in order to make his engine work, but he also needed a market for the invention in order to make its development a paying proposition. The condition was realized only in Britain, and that is why the steam engine was developed there rather than in France, Germany, or even Belgium.
Why did the industrial revolution lead to modern economic growth?
I have argued that the famous inventions of the British industrial revolution were responses to Britain’s unique economic environment and would not have been developed anywhere else. This is one reason that the Industrial Revolution was British. But why did those inventions matter? The French were certainly active inventors, and the scientific revolution was a pan-European phenomenon. Wouldn’t the French, or the Germans, or the Italians, have produced an industrial revolution by another route? Weren’t there alternative paths to the twentieth century?