« Memo Question for February 20: Institutions | Main | Memo Question for February 27: Colonialism »

February 13, 2008

Comments

Tijl Vanneste

In his article, Voth lays out a path based on empirical data that backens his hypothesis about the importance of abstention from leisure with regard to economic growth. The fact that he uses data from a new type of source in this regard, is remarkable and is a welcome contribution to a large debate. The idea is convincing, but the article alone is not enough to accept the hypothesis and ‘close the discussion’ (if such a thing would ever be possible). First of all, he used data from a limited region (London) and a limited time-period (1749-1763 and 1799-1803). Taking the particularity of the source material into account as well, his measurements can and shoud serve as an indicator and an encouragement to go further, as Voth himself suggests towards the end of his article. Other regions should be included, as well as a longer time-frame. Adding more places could counteract the possible bias of the specificness of London crime records, and could as well shed more light whether the hypothesis that is to be tested should serve as the only explanation for a phenomenon or one out of different explanations. Did people work longer in the same period in regions that did not experience an Industrial Revolution just yet? Is the same true for, for instance, the south of Belgium or did other factors apply? These are questions that cannot be answered by Voth. Further, although he goes quite a long way in testing his model against possible biases, it seems to simple to declare that witnesses related freely about their activities, going ‘awhoring’ and derive from comfortably that they did not sketch an ideal image of their use of time to stay within a socially acceptable image. This, and other issues related to the source material, such as memory issues, should be handled in my view with more attention. Also, the sleeping pattern does pose a problem. It is to me not entirely clear how a perfect distinction can be made between work, leisure and sleep.
The increased input of labor, and the industrious revolution, is also a more complex phenomenon than people working longer. There are differences about who works longer, there’s a change in household economy, gender issues need to be adressed etcetera. It is also something that should be embedded in a larger time-frame. One of the reasons the concept of the Industrial Revolution is under attack is exactly that it was not just a post-1750 radical change but something that fits in a more expanded time-frame, with changes in productivity already occurring long before. Voth or others should work on time allocation before 1750. In my view, Voth needs more than his data. He needs a more elaborated conceptual framework, taking into account more aspects of economic change. Also, downplaying the role of innovation with regard to labor means that he should do more than just show that people worked longer. He should also show that this factor was more important than other possible explanations. In general, Voth lacks a comparative horizon, in time as well as in space, in methodology as well as in analysis.
In general however, his model deserves following and it could offer an important manner in which the Industrial Revolution can be reexamined. Clark agrees with the trend of working longer, a fact that he used to value his idea that standard of living was not better around 1800 than it was in hunter-gatherer societies. He finds Voth’s model ‘interesting’, but refrains to analyse his method deeper. Clark accepts Voth’s numbers as ‘quite consistent’ with his own research. So, is his evidence convincing to Clark? I would think not entirely, although Clark does not give a reason why not.

Mark Borgschulte

Voth uses an innovative method of building a dataset, by surveying published records of witness testimony in court cases, to test the hypothesis that increased labor supply drove the industrial revolution. While he finds little to contradict his claim, his method relies upon several untestable assumptions that call into question his inference. Specifically, Voth assumes that the witnesses provide something like a random sample of the activities of the average citizen in 18th century London. Plainly, this is a strong assumption, as we must worry about bias in the sample selection and distortions in the witness accounts, to name a few prominent issues. Voth's attempts to address these issues rely upon comparisons of his sample to other survey data regarding occupation, as well as comparisons to a dataset of workers in 1800.

The counterfactual question we are left with: is it possible that the hardest-working people across occupations tend to give testimony more often in 1800 than in 1750? I do not believe we can completely reject this explanation. In a society with relatively little social mobility, this seems entirely plausible, as occupation would not necessarily reflect ability or work/leisure preferences. Presumably the trend in the data would be accounted for by changes in the selection of witnesses. No evidence is presented that can answer this.

Voth's most compelling test compares the work patterns of his sample to that of workers on the Cheshire Canal. More tests of this sort would be useful, as it is, once again, a stretch to consider the canal workers to be a representative sample. Unfortunately, trend data is not available for the canal workers, ie the period surrounding 1750 is not included. This leaves open the possiblity of bias in the 1750 witness sample.

Ultimately, Voth's evidence is suggestive, but hardly conclusive, to both this reader, and Clark. The problem seems not so much in constructing a cross-section of workers at a given point in time, but in a steady measure of hours worked across many decades. This task is further complicated by changes in the workday and patterns of seasonal employment. Overcoming these issues seem crucial in providing a clear picture of changes in labor supply and leisure demand in late 18th century England.

Ernie Tedeschi

Voth makes the argument that the economic growth of the Industrial
Revolution was more driven by increases in hours-worked rather than gains
in efficiency from new technology. As evidence, he posits that a hike in hours
worked is the only way to explain observations of both increased per-capita
consumption between the mid-18th century and the early-19th century and
declining real wages. Voth concludes that “abstention from leisure” was the
more important factor in escaping the Malthusian stagnation of the previous
centuries.
Several reasons exist for questioning his hypothesis given the evidence he
cites. First, in reconstructing a typical labor schedule for Londoners for both
c. 1750 and c. 1800, Voth concludes that the number of holidays enjoyed by
English workers – political, religious, and unofficial such as “Saint Monday” –
decreased during this period while daily labor obligations remained the same.
This loss of leisure to a large measure return by the 1850s, however, and yet
economic growth continued unabated. Since it is unlikely that marginal
returns to labor are increasing at the point of an 11-hour work day, Voth
has only answered one puzzle by uncovering another: if the role of declining
leisure has been the untold story of the Industrial Revolution, then what do
we make of the leisure recovery of the mid- to late-19th century? This could
simply be a rational response to the lower opportunity cost (lower real wages)
of leisure, or it could be the result of government intervention limiting labor
hours. Yet without technological change, either of these explanation would
imply lower economic growth, and, in fact, the higher productivity from
technology would itself imply higher real wages. At best, if we understand one
of defining characteristics of the Industrial Revolution to be an extraordinary
hike in per-capita GDP growth, then all Voth has done is move its bookends
1slightly further ahead into the 19th century.
Another reason for doubt based on Voth’s evidence – and what I think
would be Clark’s main objection to his argument – is that throughout me-
dieval times, Malthus observed per-capita income that was strongly tied to
population growth (or, to think of it differently, the labor supply). In other
words, up until 1800, if there were any increasing returns to labor, they were
not enough to grow income beyond the birth rate, and so larger populations
in history generally enjoyed lower personal incomes. Simply working longer
hours during the Industrial Revolution – essentially expanding the labor sup-
ply if we measure ”labor” on an hours- rather than persons-willing-to-work
basis) – should not have changed this dynamic, and we would have predicted
personal consumption to have decreased. Yet between 1750 and 1800, it
did not. So technology must have played some role; indeed, Voth himself
concedes this when he cites the Solow growth model.

F. Gerard

Memo 3: Industrial and Industrious Revolution

What drove the Industrial Revolution? A sudden (let say that a time period of 50 years is more or less a sudden event in the history of humanity) rises in ingenuity leading to technological progress and more productive workers or a sudden rise in industriousness of mankind leading to more hours of labor provided. This is the debate between McCloskey and Voth enlightened in the question for this memo.
The concept of industrious revolution is new for me. Generally, in the classical history classes one may get, the industrial revolution is invariably explained through an innovation channel. But as we saw in the first weeks of this class, the steam engine was discovered in Egypt centuries before… basically, it is more the practical use of inventions rather than inventions per se that was key for the industrial revolution (at least at the very beginning). Therefore the concept of industrious revolution seems a very appealing complementary explanation for the start of the industrial revolution or even the start of what is generally perceived as the main cause of this new historical period: technological progress and its extended practical application.
But behind the concept, what are the evidences of an industrious revolution in the second half of the eighteenth century? That is the aim of Voth’s paper. By studying witnesses’ testimonies (provided by court reports) in London around 1750 and around 1800, Voth tries to infer the daily schedule of the witnesses and the number of days a year an average worker was effectively exercising his labor activities. The originality and ingenuity of his method is undeniable and the results he gets seem to come to support the new concept considered here above: the daily schedule was about the same in both periods but the number of working days a year was “significantly” higher around 1800. It is difficult to be completely convinced by Voth’s evidences, several severe bias clearly (possible lies of the witnesses in front of an official court to appear as a “good” citizen, …) exist and the London population that is likely to testimony in front of a court is hardly a representative agent of the English economy. A bunches of studies of the same sauce are required to strengthen the argument. But even if some historians may be tempted to add tomatoes in the Bolognese, I think that will not be the most efficient use of their time. Indeed, the debate opposing McCluskey and Voth is maybe not that relevant if it appears that the two possible driving forces of the industrial revolution are just two different manifestations of a unique phenomenon. In short, what is (are) the driving force(s) of the driving forces? Why were people suddenly ready to work more and likely to invent more (or to find applications of previous inventions)? I strongly believe that this is the question that should crystallize the attention. A possible explanation could come from the point I was underlining in last week memo: the eighteen century is also century of the Enlightment and some crucial political revolutions. Modifications in the immobility of the social classes (or weakening of the justifications for this immobility), in the conception that the religion is the only way to realize yourself as a human being (and that working has only a subsistence goal) can encourage people’s industriousness or ingenuity. Then new goods appear on the market, you can now reduce your household production and increase your provision of labor on the market to buy those product and the voice that was telling you that those “material” goods were of no intrinsic values (in itself or for the social class you come from) is now weakened, so you want to do so!

Edson R Severnini

In his article, Voth presents evidence of increases in annual work hours in eighteenth-century London to argue that the abstention of leisure was the main force driving the Industrial Revolution (“Perspiration rather than abstention governed the Industrial Revolution”). His conclusion contradicts the McCloskey's argument that the technological change was more important than a abstention of consumption in driving economic growth during the Industrial Revolution (“Ingenuity rather than abstention governed the Industrial Revolution”). Even assuming that the Voth's estimates of working hours are completely accurate (there are some important caveats presented in his article, for example), I think that he did not provide sufficient evidence to support his argument. Neither Clark think that that evidence is convincing. I will present my worries and the Clark's argument below.
First of all, I think that change in the annual-work-hour pattern of individuals does not mean change in such pattern for the whole corresponding household. In other words, the labor participation status of each member of a family matters. Suppose, hypothetically, that in 1750 every member in a family worked to guarantee the subsistence, and in 1800 only the spouses of that family worked and still guarantee the same standard of living as before, since each of the spouses was working more hours now. Assuming that the hours worked by children were perfectly substituted by hours worked by adults, in this example there is an increase in the working time of the adults of the family and the leisure time of the whole family is unchanged. Since the Voth's estimates of annual work hours are probably of adults (in general, witnesses are adults), they do not account for the working pattern of the entire household and thus his evidence does not support his argument that the abstention of leisure was at the core of economic growth during the Industrial Revolution.
Besides this abstention issue, I think that the Voth's interpretation of the TFP was precipitated. The TFP does not involves only “efficiency”: the technological advance is also a component of it. So, the falling in the TFP that Voth refers can be a balance of the negative pressures of efficiency due to the entrance of a great contingent of people in the labor force (remember that the population growth is rising in that moment), and the positive pressure of the increasing innovation rate, the main force driving economic growth, as we have learnt from the Solow model. In fact, he himself mentions that the technology may have played an important role. He said: “It is likely that, in the absence of technological advances, declining marginal returns would have rapidly acted to depress the living standard of the population”. Therefore, I still believe that the “ingenuity” was the driving force in the Industrial Revolution.
Clark, in turn, gives another argument to do not support the Voth's conclusion. He does a hypothetically experiment in which a Malthusian economy where workers work 2,000 hours per year experiences an increase in labor inputs to the 3,000 hours per year (typical of English workers in the Industrial Revolution period), and he argues that the Malthusian era actually covers this situation. According to him, the higher labor inputs would correspond to higher annual material output, and thus a short run situation where births exceeded deaths, and hence population grew. But with this population growth the economy would again attain equilibrium, with the same annual real income as before, but workers now laboring 3,000 hours per year as opposed to the previous 2,000 hours.
To sum up, although the Voth's new method for reconstructing the annual-work-hour pattern of Londoners in the eighteenth century was (almost) effective, I think that his estimates are not enough to support his conclusion that abstention of leisure was the main force driving the Industrial Revolution. Neither Clark does. As I have argued, I still think that the technological advance (“ingenuity”) was at the core of the economic growth of that time.

Mitchell Hoffman

Voth seems to provide convincing evidence that individual labor supply, that is, the number of hours each individual worked per week or year, was higher in 1801-1831 than in 1760-1801. His use of the Old Bailey data is intuitively clever, and also careful, as he takes steps to ensure that those giving testimony at Old Bailey have characteristics that are representative of the broader English population. There are possible steps where the data could have been fiddled with to make his point, like when Voth imputes hours slept, though again, he seems careful to make his results appear honest and robust to different imputations.
The case that increased labor supply was the cause of Industrial Revolution growth is still interesting, though a bit less clear. Voth argues that this fact about the large increase in labor supply links several important facts, one, moderate output growth in Industrial Revolution England, with minimal TFP growth, and two, the estimates of capital accumulation that have been revised downwards recently. I am largely ignorant of the literature on macro measurement issues, but they seem extremely important here. With all the industrialization in early 19th century England, must technology, or at least use of technology like the growth of water and machine-power driven factories, have increased substantially, thus meaning that TFP growth was more than minimal? If TFP growth and/or capital growth was substantial, Voth’s argued large increases in labor supply are no longer a unifying explanatory tool, but rather an anomaly requiring explanation.
Still, I think that Voth’s arguments on the existence of labor supply and its macro implications are substantial and important. The question I would like to see answered in future work is why did people start working more. This question is particularly prescient if TFP did not change very much. Why did people start working on Mondays and Holy Days if they hadn’t before? Was it required as part of their jobs in factories and other employment with employer-mandated hours, or was it mostly farmers and self-employed shopkeepers who made must of the changes in hours? How is this issue related to observed constant or decreasing wages, and what is the relation of this question to central labor economics questions on intertemporal labor supply and the relative importance of income versus substitution effects?
The Greg Clark chapter was interesting, but I don’t think it relates to most of the substantive issues in the Voth article. His descriptions of foraging people working less and have decent living standards, or discussion of calories per working hour in Malthusian societies, is not direct related to the economic impact of English people working more in the early 19th century, or the empirical question of whether people started working more.

Mitchell Hoffman

Voth seems to provide convincing evidence that individual labor supply, that is, the number of hours each individual worked per week or year, was higher in 1801-1831 than in 1760-1801. His use of the Old Bailey data is intuitively clever, and also careful, as he takes steps to ensure that those giving testimony at Old Bailey have characteristics that are representative of the broader English population. There are possible steps where the data could have been fiddled with to make his point, like when Voth imputes hours slept, though again, he seems careful to make his results appear honest and robust to different imputations.
The case that increased labor supply was the cause of Industrial Revolution growth is still interesting, though a bit less clear. Voth argues that this fact about the large increase in labor supply links several important facts, one, moderate output growth in Industrial Revolution England, with minimal TFP growth, and two, the estimates of capital accumulation that have been revised downwards recently. I am largely ignorant of the literature on macro measurement issues, but they seem extremely important here. With all the industrialization in early 19th century England, must technology, or at least use of technology like the growth of water and machine-power driven factories, have increased substantially, thus meaning that TFP growth was more than minimal? If TFP growth and/or capital growth was substantial, Voth’s argued large increases in labor supply are no longer a unifying explanatory tool, but rather an anomaly requiring explanation.
Still, I think that Voth’s arguments on the existence of labor supply and its macro implications are substantial and important. The question I would like to see answered in future work is why did people start working more. This question is particularly prescient if TFP did not change very much. Why did people start working on Mondays and Holy Days if they hadn’t before? Was it required as part of their jobs in factories and other employment with employer-mandated hours, or was it mostly farmers and self-employed shopkeepers who made must of the changes in hours? How is this issue related to observed constant or decreasing wages, and what is the relation of this question to central labor economics questions on intertemporal labor supply and the relative importance of income versus substitution effects?
The Greg Clark chapter was interesting, but I don’t think it relates to most of the substantive issues in the Voth article. His descriptions of foraging people working less and have decent living standards, or discussion of calories per working hour in Malthusian societies, is not direct related to the economic impact of English people working more in the early 19th century, or the empirical question of whether people started working more.

Gabriel Chodorow-Reich

The Industr-? Revolution


Hans-Joachim Voth uses daily schedules compiled from court testimony to estimate the change in hours worked in England from 1750 to 1800. He finds that a decrease in holy days and the addition of Monday to the normal workweek increased average labor input by roughly one-quarter over the course of a year. Combined with data on output and capital, the labor input estimate implies negative total factor productivity growth from 1760 to 1800. Thus workers’ increased industriousness brought about the industrial revolution, rather than a surge in technology.

In the standard Malthusian model increased labor input cannot lead to permanently higher living standards. As Clark (Chapter 2) explains, raising the number of hours worked per year has an equivalent effect to a shift in the technology curve. In the short run incomes rise and the death rate falls, but the subsequent increase in population size eventually leads the economy back to its original equilibrium. The industrious revolution should have led to a transitory increase in income but a permanent increase in population.

So what happened? According to the welfare ratios shown in class, English living standards did not begin their acceleration until the middle of the nineteenth century, after the repeal of the Corn Laws and fifty years after the conclusion of Voth’s study. Likewise, Brown and Hopkins (1956) study of real wages of construction workers in southern England show little movement (except a downward spike coinciding with the Napoleanic Wars) from 1750 to 1850, and then a series break. The stagnation of factor prices supports Voth’s claim of low TFP over the period.

Nonetheless, eventually wages and productivity rose, and dramatically so. This fact points to a crucial step in the theory of the industrious revolution, namely why increased labor input should induce technological innovations of the scale observed. Indeed, without such inducement the higher labor hours could not have moved England out of the Malthusian world. Standard neoclassical theory presents a story for why productivity-enhancing technological improvements and competitive markets should lead to higher real wages, causing labor supply to increase to meet the new demand. Why should causation also run from labor supply to productivity improvements?

Without an answer to this question Voth’s claim is strikingly incomplete. One possibility is that industriousness and innovation were complements. The advent of the division of labor and the creation of the “pin factory” required increased labor input relative to the old technologies. If product and factor markets remained competitive, the pin factory could have resulted in little observed TFP growth. Both the technology underlying the pin factory and workers’ willingness to meet the higher labor demand may have constituted necessary conditions for the revolution.


Wayne Feng

In the analysis of which was of more importance, invention or abstention, Voth argued for the abstention of consumption and an increase in labor input. This increase in labor input is touted as what enabled the industrial revolution versus technological development which would increase labor output (in terms of significance). The evidence of changing work hours does seem to be convincing in showing that there is a noticeable change in labor inputs. While very innovative, the limited nature of the data does beg some questions about whether this labor change was all prevalent and not counter balanced by any other phenomenon. For Clark there are ample examples across time and space which offer evidence that living standards are independent of technology as well as time across different groups. Thus for Clark it is the utilization of labor and classification of wealth which matter. People that may be materially not as well off may be wealthy in leisure. For Clark this was the cause of varying living standards. While this may be true, the question of household expenditures and choice of production/goods should also be addressed to make a complete argument (such as the “industrious revolution” proposed by professor deVries). For this analysis we would want to look at household consumption and choice of consumption of goods. In the data provided from Voth the specificity of the people involved leads us to ask for further analysis of the role of the family (individually analyzed) in the changing labor marketplace. From this we should also look at the expenditures associated with family members, such as leisure. Clark would argue against the idea of more work leading to a more elaborate standard of living as with less work you also have more leisure time. However with the story that is told by de Vries, Clark, and Voth we can argue that in terms of material wealth abstention of consumption does play a role in the industrial revolution. This is in some sense contrary to what we expect due to Solow growth model expectations. At the same time it could be that in the short term we follow TFP analysis with increased labor inputs which raised output and provided incentives for technological advancements to follow. This is in fact what Professor de Vries argues with the “industrious revolution” preceding the Industrial Revolution. So while Voth makes a very interesting point with creative analysis we see that the analysis is not altogether complete. Only with more data can we make any comprehensive conclusion as to what role changing labor inputs played in the industrial revolution.

Alexandre Poirier

The common perception about the Industrial Revolution is that it was a period of high growth caused by a sudden apparition of many productivity enhancing inventions. For example, the invention of the steam engine and sewing machine are often thought of as greatly improving productivity. The concept of Industrious Revolution introduced by De Vries is completely opposite: the growth during the period extending from roughly 1750 to 1850 was mainly caused by a great increase in labor supply. Voth tries to quantify the increase in hours worked ingeniously by looking into the Old Bailey Sessions Papers to find witnesses accounts and recalls about their whereabouts. He uses these testimonies to infer the length of the working day, and the number of days worked.

There are a lot of potential shortcomings to such an approach. One may think of selection bias, memory failure and exaggeration from the witness. Voth seems to be very careful about these, and addresses these concerns by comparing the sample of witnesses to London's population, and by performing other comparisons with the population at large, to ensure the sample is representative. Overall, with all these checks, I think Voth's evidence is as convincing as can be, given the rarity of data. One shortcoming is that he uses only data on the city of London. It would be interesting to see if these changes were really widespread in England, and then, in most of Western Europe. One may think that the workers of smaller towns might have a lighter work schedule, as is often the case in modern times.

There are also other interesting questions that have not been answered. For example, the main reason working hours increased was by the reduction of holidays and the end of Saint Monday, the practice of taking Monday off to recover from the weekend. What is the reason that this practice has stopped? Did the other European countries also have that quantity of religious holidays? If so, did their number also drastically go down? All these questions would certainly help complete Voth's argument.

Finally, in the introduction to his paper, Voth mentions that the quantity of labor inputs increased in response to better nutrition. The rest of the paper is that this increase in labor input provided the majority of the growth during the industrial revolution, not technological progress. But my question is what is it that increased the quality of nutrition? If it is some sort of agricultural technology, we could think of that technology as having caused the Industrious Revolution, because such an increase in working hours does not come out of nothing. The better nutrition just before the Industrial Revolution does not seem to correspond to Clark's account of calories per day across different periods.

The comments to this entry are closed.

Search Brad DeLong's Website

  •