« Memo Question for March 5 | Main | The Industrialization and Urbanization of England »

March 05, 2008


F. Gerard

Francois GERARD
Memo 7: Technology, Investment, and the Industrial Revolution

“… so small relative to the whole economy that it no longer deserves the title of Industrial Revolution”. Do I agree with such a statement? No, I still consider the second half of the eighteenth and the first half of the nineteenth centuries as a key period in the general process that lead to the creation of our modern economies and the shape of the economic relations worldwide.
Growth accounting exercises failing to reveal a serious change in those years are not sufficient to neglect the importance of the period. First, those exercises are known as perilous given the number of assumptions required (as clearly apparent in the paper by Crafts himself) and, when applied to such a remote period, the weakness of the available data. But even without those technical considerations, such a growth argument is not appealing to me. Indeed, as Williamson interestingly shows in his paper, external causes like wars can totally explain the low rates of capital accumulation and growth prevailing around 1800. Temin and his Ricardian model of international trade also reveals that applying stylized economic models (with their apparent weaknesses) to the data not always come to support the idea of the obsolescence of the classical view. But more crucially, a revolution, even if industrial, is much more than a change in some indicators. On this point, Berg and Hudson are particularly convincing. Even today, new inventions, new organizational techniques need some time to be diffused broadly within our economies and seriously affect them. Two centuries ago, without neither the same communication networks nor the same entrepreneurial history, it doesn’t seem unrealistic that a change took more time. The new machineries as used in the cotton industries were heavy structure requiring huge investment, such kind of capital investment were generally risky (the machines were far from unbreakable and rapidly obsolete) and new for the entrepreneurs as well as the workers. A technical revolution even if occurring first in limited industries (supposing that Crafts may be right) but reaching slowly the various sector of the economy, changing the classical working relations, the pattern of urbanizations, some demographic behaviors … does not deserve to lose his name.
I would also come with an idea that is maybe more a question. While Williamson seems to be convinced that the French wars explain most of the low capital accumulation rates during the Industrial Revolution, can we imagine that the same events are partly responsible of the Industrial Revolution itself. To be more precise, a war when no battle takes place on your soil can have positive effects on the shape of the industries and on the speed of innovations: tons of usual weapons, dozens of ships must be produced, new weapons must be produced to have a technological advance on the enemy and everything must occur in a really short time. May that explain part of the industrialization story?
Nonetheless as underlined in the first paragraph of this memo, it is important to acknowledge that the Industrial Revolution was part of a general process that started earlier and that was much more than some crucial technical inventions. The last weeks were clearly dedicated to convince us of the existence of such a process concerning institutional reforms, financial innovations, long distance trade and new consumerism and industriousness. Then, the point of view of the new school of thought must maybe be perceived not as a coup to withdraw the King from his throne but as an attempt to show to his zealous partisans that the King is not a God.

Tijl Vanneste

Berg and Hudson argue in their article for a reevaluation of the Industrial Revolution, its characteristics as a period of profound change being revised by literature relying on a national accounts approach. This literature posed the question why British economic growth was so slow and pushed this as far as questioning whether the First Industrial Revolution truly deserved that name. According to Berg and Hudson, the methodology with which the revisionists come to this question is not flawless. The numbers could be an underestimation, and macro-economic indicators fail to show certain dynamic important aspects of the period. Berg and Hudson stress four points: technological and organizational innovation outside the factory sector, deployment of female and child labor, regional specialization and demographic development. In questioning the revisionist analysis and in coming up with a new agenda, the authors manage to build a convincing case in favor of the First Industrial Revolution as a period of change as hailed in the traditionalist view. Temin, using international trade as an indicator, seems to agree with them, although his article should not be seen as a firm reestablishment of the I.R., but rather as a methodology that showed the revisionist’ division of cotton and iron as modern industries and other manufactures as relatively backwards as problematic, a division that should show that growth and innovation was not widespread but limited to these two industries, playing down the general importance of the I.R. Temin’s article takes the ground away for this view, but it does not allow the historian to discard completely the revisionist view.
The argument of traditionalists versus revisionists seems reminiscent to the debate regarding, what professor de Vries called, soft and hard globalization (in the early modern period). Hard globalization, price convergence, did not exist, while a soft globalization, of growing interaction, exchange and interconnectedness, might have taken place. It’s a quantitative approach that doesn’t yield the right numbers that leads to a questioning of the concept, a questioning partially countered by a more qualitative approach, similar to Berg and Hudson, in stressing that even if the numbers would be right, there is a need to focus on other aspects to fully comprehend the change that the I.R. has brought forward. Their article was written in 1992, and what was true then is not necessarily true now. I would not be sure whether the literature has moved too much in the direction of the revisionist view. It is at least my impression that in recent years, history writing is popular in which the I.R. still plays its role. It might be true that some of this literature focuses on the long run, on modernity, on processes of globalization. In this literature, the I.R. is not regarded as an almost insignificant event. It is true that there is, I think, a tendency to embed the I.R. in a longer time frame, and hence to stress at least some notion of continuity or the idea of change that was taking place before the British First Industrial Revolution (for instance the modernity mentioned by Neal of capital markets, de Vries’ concept of industrious revolution).
A part from the revisionist/traditionalist or continuity/discontinuity schemes, there seem to be at least two other dividing lines. Williamson claimed that effects war (crowding out because of war debt, mobilization and war-distorted prices) counterplayed effects of the I.R., not necessarily questioning the concept of I.R. per se. And secondly, there is the debate of timing. Some suggest, fundamental change occurred only after the 1820s. Be that as it may, but the productivity paradox that effects of technological innovation might only be visible later, and might even have a negative effect on investment and growth at first, seems a very important element that could be fundamental in the debate on timing. It could link different views, in stating that rapid growth worth the name Industrial Revolution might be numerically visible at a time when the most fundamental changes of such revolution had already taken place.


With only a handful of readings, it is hard to say whether the broader literature has moved too far in the direction of de-emphasizing the changes wrought by the Industrial Revolution. However, we can say something about whether or not it might be appropriate to still regard the Industrial Revolution as an important turning point.

The crux of the debate is how much change the Industrial Revolution itself provoked. Growth accounting, according to Berg and Hudson, has shown that gross domestic product; industrial output; productivity; fixed capital proportions, savings, and investment; and workers’ standard of living all grew relatively slowly through this period (1992). Thus, if the Industrial Revolution is to be any kind of turning point, let alone a unique one, something else must be changing.

To Berg and Hudson, some of this change was in “technical and organizational innovation outside the factory sector, the deployment of female and child labour, regional specialization, and demographic development” (1992, 27). We can also think of other possibilities. As transportation systems were improved, information could flow more quickly, increasing the chances of innovations. Perhaps the possibility of work in textiles was important even to those who did not pursue that work. The machines and factories certainly caught the imagination, and perhaps even contributed to a change in consumerism parallel to that discussed by Austen and Smith (2002); increased people working side by side in cities meant that they could more readily observe each other and adopt new tastes.

One can make up countless stories and it is easy to say in hindsight that something significant must have changed. After all, we see the start of technologies being adopted that remind us more of the modern world, and we know that things like GDP/capita did eventually take off; given that everything follows from what came before in some form or other, one would naturally like to say that it laid the groundwork for subsequent growth, but this alone would not give us a unique turning point. What we really need is hard evidence of change without which we would not have subsequently obtained great growth.

The mere fact that “English population increased by 187% between 1770-9 and 1860-9 while a wide group of other European countries saw population increase by only 79%” is fairly convincing evidence by itself (Clark, 2002). Here was something different happening in England, both different from other European countries and different from England’s past, as it subsequently kept growing in population and GDP/capita beyond all what came before (Clark, 2002). Population breeds innovation, and when the two start to symbiotically grow there is something distinctly new which one could call a turning point. Even if growth accounting tells us that productivity didn’t increase all that much over all the sectors combined, the startlingly different population trends tell us that something had changed. Interestingly, however, when Kremer ran regressions of the population growth rate on population, he found no break at the time of the Industrial Revolution (1993). This might suggest that while we’d like to take the Industrial Revolution as a turning point for England, if not the world, it was indeed bound to happen.

V Paredes

Berg and Hudson point out that even though radical change during the industrial revolution was obvious to contemporaries, current consensus now favors continuity and gradualism. They argue that “the industrial revolution should be rehabilitated”. Their hypothesis is that growth and productivity change have been underestimated. Moreover, they argue that growth rates on their own are inadequate to identify and comprehend this process.

One important argument they give is that if productivity growth was not important in this period, then it couldn’t have sustained population growth. In a Malthusian economy, population growth would have lowered income, mortality would have raised and population would decline to its new equilibrium level.

Their other principal argument is that data used in previous studies had measurement errors. There are several measurement errors when using aggregate data estimates. First, this data suffers from an index problem. As productivity growth is calculated as a residual, these errors are magnified. Second, occupational information is generally constructed for adult males, and there is very little information of females and children. Yet, women and children played an important role in this period. Third, there is no quantitative data at all of large areas of economic activity. Finally, price data is sparse and partial, which creates a problem when trying to calculate the aggregate value of a sector.

Although I was convinced by Berg and Hudson that previous quantitative work on the Industrial Revolution have measurement errors and so cannot provide precise answers, I am still not convinced that the process was not gradual. I think the readings don’t give enough evidence to be convinced about one hypothesis or the other.

Mark Borgschulte

I would agree that there is certainly room to argue that a fundamental shift in human history occurred in Great Britain, starting just before the turn of the 19th century, and proceding through its first half.

The primary argument Berg and Hudson attempt to counter is that growth accounting has debunked the notion that rapid technological change occurred in this period. Although they present compelling theoretical justifications for being suspicious of growth accounting as the sole source of historical evidence upon which to build one's opinion, as far as I can tell, they do not present any solid evidence demonstrating the failure of growth accounting to document rapid industrialization that other quantitative sources establish. The most useful theoretical argument they advance (mixed in with the laundry list of complaints about aggregate studies) is that growth accounting depends on factors being paid their marginal product.

Into this void steps two empirical explorations of the causes and consequences of rapid (or slower than rapid) industrialization in the early stages of Briatain's Industrial Revolution. The first paper, by Williamson, argues that industrialization was slower than contemporaneous accounts suggested because Britain's war debt crowded out investment. This argument fits somewhere in between supporting and refuting Berg and Hudson's thesis. Rapid and dramatic changes DID take place, but were not fully catalyzed, due to the lack of capital investment- an argument which supports their thesis. However, this evidence contradicts the specific assertions upon which Berg and Hudson base their case- growth WAS slow, though this did not reflect societal changes occurring during this period.

The other paper that supports their hypothesis, by Temin, uses evidence from trade records and a simple Ricardean model of trade to test whether productivity growth occurred in sectors outside of cotton and iron manufactures. I'm not familiar enough with the Ricardean model to know its limitations, but the evidence they present seems to rather conclusively rule out a restricted story of technical change.

All in all, it seems the quantitative evidence supports the contemporaneous accounts- the early19th was a period of rapid productivity growth, which was mitigated by low capital investment. The result of these competing forces is a dilution of TFP growth when calculated using growth accounting methods- probably because of the failure of one of the key assumptions of the model, namely that factors are paid their full share

Ernie Tedeschi

The Industrial Revolution is such a fixture of the discipline of economic history that it serves as the fulcrum around which we analyze later and even earlier growth phenomena in a diverse array of nations. It occupies a singular place in the collective imaginations of economists and historians alike; yet, its primacy is not so universally accepted that it stands unchallenged. The Industrial Revolution has --- especially in recent years --- been subjected to heightened scrutiny and revisionism. Crafts and Harley, for example, argue that the wide sweeping economic upheaval of lore was, in actuality, a narrow event that only touched a few English industries. On this week's reading list, Williamson concedes the technological changes of the Industrial Revolution but points out that economic growth was hardly extraordinary, and places the blame on British war debts. It is fair to say, then, that the Industrial Revolution is not the sacred cow once assumed.

Or is it? Against Crafts and Harley (and indeed the whole gang of Industrial Revolution detractors), a counter-backlash has arisen attempting to push consensus back to the traditional view of the revolution. Berg \& Hudson are the broadest and most systematic, aiming their academic salvos at a wide range of targets, including critiques based on demographic behavior and revised macroeconomic estimates of the British economy. Peter Temin is more surgical, focusing solely on the Crafts/Harley rebuttal and testing both it and the traditional hypothesis using the Ricardian model of international trade, ultimately siding with the traditionalists. To use a historical analogy, here the Counter-Reformation seems to have been at least as fierce, and maybe as prolific, as its impetus.

So if the question of dissention is simply descriptive --- a question of volume --- then the answer has to be that the literature of Industrial Revolution revisionism pales in comparison to the traditionalist view, and is indeed chronologically bookended by clusters of it. If instead the question is more normative: whether it is unfair or even harmful to the discipline that Industrial Revolution history turned so incrementalist in recent times, again I have to say ``no,'' for the uncontroversial reason that the academy is well-equipped to handle healthy debate, and I imagine both schools of thought would agree that if any subject in Economic History were due for a healthy debate, it was the Industrial Revolution.

These are the easy questions. The hardest one is whether the incrementalist side of the debate in fact has the better argument. I agree that the lack of extraordinary growth during the Industrial Revolution in Great Britain ought to give economic historians pause when weaving superlatives into their narratives about the period. The Industrial Revolution was not a growth explosion defined by sustained and significant augmentation of national income. Berg and Hudson, however, rightly point out that contemporaries knew something extraordinary was occurring with the technology and productivity methods being employed in Britain. Perhaps, then, the Industrial Revolution was a more fundamental change that wouldn't yield income dividends until later.

The comments to this entry are closed.

Search Brad DeLong's Website