“If you would understand anything, observe its beginning and its development.” Aristotle1 The tireless effort of mankind to examine its origins with the hope of gaining insight on its
1
present is ultimately a universal struggle. Even economists are not immune to the cancers of repetitive error, whose occurrence is inevitable should one disregard the extensive examples authored by those who existed before us. More intrinsic however, is the opportunity to justify present institutions and framework, achieved by the development of a clearer understanding on the derivation of existing economies with respect to time. In the discourse of economic history, the Industrial Revolution is a topic that has consumed the interests of historians and economists alike. In the realm of economics – and specifically in the stream of growth accounting – the once unanimous opinion of the magnitude and eminence of this time period has been profoundly debated in recent years, essentially forming a disparate set of perceptions. There is on one side the opinion that the original works of Rostow and others alike are more appropriate figures of the aggregate growth rates, embodied by Maxine Berg and Pat Hudson (Berg and Hudson); otherwise known as the ‘qualitative’ group of thinkers, which is qualified later on in the study. Their conceived opposition is the opinion that the original calculations were plagued (due to crucial index errors)2 and misleadingly overstate these macro growth rates implying misconstrued generalizations, embodied by NFR Crafts and CK Harley (Crafts and Harley); otherwise known as the ‘quantitative’ or ‘revisionist’ thinkers, also qualified below. Consequently, the format of this piece will work to firstly summarize each outlook, then to elucidate the differences – which is derived from the former. Finally, the purported conflictive nature of these views is challenged, and it will effectively be shown that they are in fact not mutually exclusive, but are co‐dependent and equally correct.
1 2
Szasz, Ferenc M, ‘The Many Meanings of History, Parts I‐IV’, p208 Crafts & Harley, ‘Output Growth and the British Industrial Revolution’, p6
We begin with the qualitative aspect of Berg and Hudson. The basis of this term is drawn from
the article itself, as the underlining tone in their contention is that quantitative measures such as growth accounting is not a sufficient method to encapsulate the truly revolutionary nature of the Industrial Revolution: “we stress that growth rates on their own are inadequate to the task of identifying and comprehending the industrial revolution.” 3 Berg and Hudson select to highlight four areas whereby this time frame justly receives its revolutionary name: technical and organizational innovation outside the factory sector, the deployment of female and child labour, regional specialization, and demographic development. 4 Each concept is analysed in turn.
Firstly, the authors challenge the notion that innovation in any form was confined to a
minuscule sector in the industry of factories, based on several grounds. They assert that the assumptions of Crafts on the cotton sector are based falsely on the concept that the innovativeness of the factory sector functioned independently of and owed little to the change in the rest of the manufacturing and service sector at large. And secondly, that it is based on the idea that innovation only existed in the introduction of capital‐intensive plants and equipment and nothing else (i.e., in processes or ideologies). This sets the platform for an in depth criticism on the legitimacy of growth accounting and total factor productivity (TFP) models. Berg and Hudson then consult their concerns with the revisionist’s estimates of output and national income, which are calculated from the weighted averages of their components. The approach neglects the difference in opinion on the calculations of the weights, the changes of the weights with respect to time, and any fluctuations in price levels. Additionally, any errors in computing weights or component levels will have a multiplier effect. The mathematics of the model thereby magnifies these errors when it comes to any residual calculations,
3 4
Berg & Hudson, ‘Rehabilitating the Industrial Revolution’, p4 Ibid., p5
such as that of productivity. 5 Furthermore, the information needed to conduct quantitative analyses such as growth accounting is dependent on statistics that are in their own respect questionable. Lindert, who gathered and produced some of these statistics, also professes that the margins of error in some cases are close to 60%, and that some of the figures are little more than guesses. Moreover, Lindert relies on the very few records available, including the burial records of adult males in order to estimate the size of the labour force thus ignoring any participation by that of females or children. Macro‐data’s reliance on multipliers becomes questionable when they are derived from only a few examples of a small sample of industries. This is apparent with respect to estimation of the service sector, which becomes near impossible to achieve. Large areas of the economy at the time left no quantitative source or independently observable trace. This lack of credible statistics argument also shows up when constructing price‐level estimates, fundamental to the value‐added methodology highly used by the revisionists. 6 Next, the modern approach to interpreting the Industrial Revolution relies on an analytical divide between large and relatively unproductive traditional and service sectors, and the small yet highly productive factory industry. This mathematical necessity is a complex tactic due to the difficulty in separating the two industries as they were in many cases present within the same sector or even present within firms themselves. For example, many firms had institutions set up that produced with innovative machinery and processes as well as traditional factories that stuck to conventional ways of production. This was done as a form of hedging risk and in attempts of establishing a balance between new technology and cheap labour. 7 Moreover, the plastering of ‘stagnation’ as a trait of the traditional sector is also problematic. Berg and Hudson reference examples of the textile and metal trades to highlight the triumph of transforming products and processes, as well as the overlaps in labour that were created through the traditional sector’s inventiveness. Essentially industrial labourers were
5
Ibid., p6 Ibid., p6 7 Ibid., p5 6
also a part of the service and agricultural sectors’ labour force, thus creating external economies. This emphasizes the economy’s resemblance to an intricate web as opposed to a sum of distinguishable components. The productivity growth rate exhibited by the cotton textiles sector is by no means applicable to other industries, but it was clearly not disconnected to the rest of the economy and depended on the innovations and transformations of these other sectors. Clearly then, the attempts of separating the economy into two distinct sectors or components runs counterintuitive to the reality of a highly interdependent and interrelated nation, and any quantitative modelling that approaches with this basis will result in erroneousness. 8 The authors then confront what is, in their opinion, a more alarming issue of productivity change assessments by the quantitative crowd. TFP, a residual calculation, is a common measure favoured by the revisionists, and can be attributed for the low levels of productivity growth they computed. Aside from the issues of error magnification mentioned earlier, Berg and Hudson find further dilemma with the methodology. The approach rests on a number of assumptions which arguably do not apply to the Industrial Revolution: perfect factor mobility, full employment of labour, perfect competition, neutral technical progress, constant returns to scale, and parametric prices. 9 Evidence exists to negate these conditions with respect to this time period and therefore to the use of such an application. Additionally, TFP takes no account for innovation of outputs or change in the quality of inputs; a benchmark accomplishment of the Revolution times. The improved quality of inputs can be characterized by the change in labour experienced by firms, as well as the expansion of raw materials usage (for example, cheap alloys in place of precious metals). Similarly, altered products were developed by many industries, consequently revolutionizing consumption patterns and habits. Also, the TFP process emphasizes on saving and capital formation at the expense of science and economic organization; major feats of Britain and are objectively revolutionary traits. Finally, the concept of a ‘productivity paradox’ is used to underscore the underestimation of growth accounting results. To
8 9
Ibid., pp 8‐9 Ibid., p11
briefly summarize, the paradox explains how changes in technology yields higher unemployment and downward pressures on wage‐rate in the short run and thus the model does not express these key advancements in the aggregate calculations for the time period. When we add the rapid transformation of society and culture, also present during this period, the longer the time‐lag before increases in productivity are reflected numerically. Berg and Hudson then move onto the second major concept of deployment of female and youth labour. The restricted definition of the workforce – due to prior mentioned issues of burial records – disregards the uniqueness of this period and the calculations thus do not encompass the innovativeness of the labour force. Female and child workers were not new to this time; however their dependency for the spread of manufacturing and factory expansion is vital. On the supply side of the issue, there is a large portion of household income attributable to the earnings of these workers, which is also compounded by the increase in population. On the demand side, there is a high need for hand‐skills which is coupled with the favourability of employers towards long‐hours and low wages, as well as with the increasing presence of these workers in commercial production. This led to new work disciplines, subcontracting, and a restructuring of the labour framework.10 Clearly then, innovativeness of the labour force influenced the rest of the economy, and equally the economy’s innovative nature influenced the labour force. These details are too specific to be observed at the aggregate level, and quantitative measures would not be able to communicate these occurrences accurately, if at all. With respect to regional specialization, the authors mention the period’s great disparity in regional rates of change and economic fortunes. The slow moving aggregates intuitively conceal the volatility of change and transformation taking place in industry. Developments in transportation such as the canal construction as well as both the availability and dispersion of information was the cause of the specialization of geographic regions in specific forms or types of production; in essence creating
10
Ibid., p14
physically separated markets of goods and factories. This locational restructuring led to increased competition as well as a drastic increase in regional intelligence. The stimulation of informational networks then led to advancement and overall economic organization, which was met by specialized financial and mercantile services. The collective outcome was an increase in external economies and a significant drop in transaction costs. 11 This regionalization generated a prolific social and economic interaction otherwise unachievable without the spatial adjustments of industries. Broadened, this dynamism cannot be found by the monitoring or study of macro‐data movements and levels. Demographic development is the final point of the article, and Berg and Hudson discuss the radical structural shifts in the composition and location of population. An increased population tends to shadow the mutual existence of decreasing mortality rates; especially in the exposition of aggregate data. Social classes as they were traditionally defined were facing evolutionary changes, and the reactions to the surrounding economic changes were in many cases incongruent between these classes. Finally, it was all facets of society and the economy that faced innovations in the organization and use of labour – not only technologically speaking. The demography of Britain over this time was visibly impacted by these aforementioned changes, yet growth accounting procedures will necessarily oversight these unquantifiable aspects. 12 We can now proceed to the quantitative perspective of the Industrial Revolution. The revisionists’ view is summarized in the article of Crafts and Harley, whose work is published in the same volume of The Economic History Review. The following is a summary of said article. Crafts and Harley primarily cover three areas of concern in their work: slower growth based on estimates, a review of the construction of industrial productivity indices and an examination of possible amendments, and finally the wider implications of revisions to overall growth, productivity, and the Industrial Revolution at large. Presumptuously, the revisionists’ view of Britain’s growth during this period depends heavily on the
11 12
Ibid., p17 Ibid., p17
assessment of the rate of industrial growth. An aggregate grows at the weighted average of its components, and even if the weights are not discernable, the aggregate growth lies somewhere between the lower and upper limit of these components. The heavy censure on the quantitative view has brought about many legitimate concerns, though they are in many cases overstated. However the authors feel that their work has been largely misconceived. Many critics of Crafts and Harley are charged with having erroneous beliefs of standard indices (such as Laspeyres and Paasche indices) as well as a general fallacy of growth accounting. 13 The claim of available data being a poor reflection of entire industrial activity is indeed a correct one. In fact, this leads to the notion that current quantitative measures are an overstatement of growth – and not the reverse as alleged by many. This result holds under the assumption that the average growth rate of omitted sectors is no higher than those that records are available for; an unlikely case as research indicates. Adding more sectors to the model reduces the value‐added shares of included sectors, which is also why current work may be overstated. Furthermore, even incorporating possible revisions will have a negligible effect on the industrial growth figure. This is ascribed to the fact that the missing sectors would have to be very large and grow at much higher rates; again, highly unlikely given the spectrum of possibility based on research.14 Furthermore, many critics in fact justify the low rates of productivity growth in their own work, specifically with the fact that innovations in technology have no observable impact on economic growth for the foreseeable medium run (productivity paradox). Their immediate effects of downward pressures on wages and increased unemployment further serve to the overestimation of national income and growth. 15 Criticisms of the TFP approach are also exaggerated, and claims that the restrictive assumptions on which the model functions is not applicable for this time period is also refuted. It is shown how similar growth rates in capital and labour negate the relevance of weighting 13
Crafts & Harley, ‘Output Growth and the British Industrial Revolution’, p6 Ibid., pp 13‐15 discusses these points at length 15 Ibid., p16 14
issues. Further, the economy as a whole experienced returns quite close to constant, and thus the margin of error in the approach is conceivably narrow compared to the wide scope of error that critics allude to. Finally, it is imperative to note that there is yet to be any quantitative challenge to their work. The figures of growth rates are indeed best guesses, and there is plausibly a degree of error and uncertainty however one cannot outlandishly claim that this range is infinite. The article also qualifies recent revisions to both Crafts and Harley estimates made by various suggestions and re‐examinations; nevertheless the outcome is largely unchanged. 16 Also, Crafts and Harley discuss the fact that productive criticism of a model suggests areas of improvement and well‐thought methods for quantification; they do not simply rehearse the known difficulties. The authors found the study in itself rewarding, it has reinforced the notion of comparative advantage, and in no way was it intended to ignore or usurp focus from the other focal aspects of the Industrial Revolution. Rather it looks exclusively through the lens of growth accounting and macroeconomic data. We can now gauge the differences in the views, and though there are several they can be summed up quite simply. It is imperative to note however, that the reasoning of differences is highly correlated with the subjective notion that both articles are co‐dependent. Subsequently, for the sake of clarity, the two ideas are combined. Clearly, the first cause of dissimilarity is due to the nature of each approach: the first considers the Revolution in a qualitative sense, whereas the second takes a quantitative approach. There are many factual sources for a qualitative analysis relative to a quantitative one; especially in the case of the Industrial Revolution with respect to available data. Secondly, though both approaches rely on the ability of judgment, a quantitative model can only form judgments based on components that are by necessity quantifiable. Even with perfect data availability and accuracy, a quantitative analysis will always encounter issues with the inclusion of relevant factors that have no numeric value or presence. This is the overlap that bridges into the final aspect of this
16
Ibid., pp 17, 20‐21, 26 contain these issues at length
paper. Qualitative and quantitative analyses in essence will always highlight distinct points, yet to immediately suggest that these points thereby oppose one another is incorrect. The view of both articles is akin to debates of positive versus normative economics; they can both be equally correct. They take a different stance on the analysis, and the difference between the two is based on the perspective undertaken. Yet we find that largely, each article does not in actuality refute the other, rather it simply highlights certain factors that the other does not or cannot analyze. Thus we see that they are not in contest with each other, but fulfill one another. Co‐dependency can also be seen by the fact that if you take either article on its own, the resulting view of the Industrial Revolution is incomplete or even partially misconstrued. For example, to look at the quantitative analysis solely would leave the reader ignorant of other revolutionary feats that took place in Britain. Conversely, the qualitative analysis in isolation would leave the reader with overstating views of aggregates over this time period, or even with the incorrect notion that these same revolutionary feats translated to immediate and substantial impacts on macro‐data. The likelihood of readers drawing such conclusions is actually quite high, and therefore in order to encompass the true nature of Britain’s Industrial Revolution, one should consult the pair of articles for an accurate depiction. The tendency of aligning oneself with a single perspective is a crucial error and is largely counterproductive to the primary goal of attempting to understand the importance of this unique period in history; any gains in comprehension will only be achieved through the acceptance that both approaches are correct. The infamous quote by Aristotle headlining this paper exhibits a double entendre: it is both appropriate in a denotative or literal sense, as well as in a connotative one. To highlight further, in the attempts of gaining understanding, we find ourselves striving to comprehend the beginning and developing stages of the Industrial Revolution. Equally, the undying quest of qualifying our present is only achievable if we delve into our points of origin and move forward. Through the summaries and examination of the articles by Berg and Hudson and Crafts and Harley, we have discerned the necessity
of considering multiple viewpoints when attempting to analyze any subject matter. Moreover, we have seen how certain overtones of articles can incorrectly lead one to label them as opposing forces whereas the truth is that indeed both can coexist amicably. Extending this we observe how such a coexistence is necessary for one to form an advanced level of comprehension of the Industrial Revolution. A fair conclusion to draw is the minimal variance of economic aggregates in the face of drastic structural and ideological change. The natural question then is whether the ingenious results of such analyses can be developed in the absence of debate and the hostile arena in which it is carried out? In this respect, it is safe simply to state that these competitive stages produce a fruitful tension.
Works Cited 1) Szasz, Ferenc M., ‘The Many Meanings of History, Parts I‐IV’, The History Teacher (1974) 2) Berg, M. and Hudson, P. ‘Rehabilitating the Industrial Revolution’, Econ. Hist. Rev., XLV (1992) 3) Crafts, N.F.R. and Harley, N.K. ‘Output Growth and the Industrial Revolution’, Econ. Hist. Rev., XLV (1992)
ABSTRACT:
The Industrial Revolution A Comparative Essay
KhalilHussein Najafi EC227 Dr. Peter Sinclair 041161140
In the debate between numbers versus words, can one be more correct than the other? Clearly this seems to be an unlikely possibility. Conceivably then, we can extend the question towards quantitative and qualitative research; can one outweigh the other? With the assumption that both analyses are grounded in truth there is no discernable way to dismiss one based on the other. This piece deals with articles of separate stances with respect to the Industrial Revolution, and effectively aims to summarize both views and then highlight why they differ as well as why they are both correct. It is in the reader’s benefit to acquaint themselves with both articles, and thus with the attempts to uphold clarity and avoid redundancy, many of the specifics are not included in the essay. Nonetheless, the major points and relevant examples are mentioned in order to encapsulate what it is the authors are mentioning, however much of the original tone is removed in order to highlight simply the facts and themes of the two views. The importance of the British Industrial Revolution is increasingly apparent in any study or attempts to understand modern Western economic institutions, and therefore this work is essentially an attempt to clarify the initial period of economic evolution relevant to us all. Both articles can be found in The Economic History Review.