HEDG Annual One-Day Workshop, SOAS, May 10th, 2000

Panel: Methodology of Comparative Economic History.

Papers by:

Janet Hunter,'Culture, Confucianism and Comparison'

Barbara Ingham, 'Pattern Modelling in Development Economics'

William Gervase Clarence-Smith, 'Commodity chains and comparative economic history: the case of cocoa in the liberal era, 1765-1914'

Bishnupriya Gupta, 'Economic Growth and Well Being: Quantitative Methods for Comparison'

 Tom Tomlinson, 'External Shocks - the possibilities and limitations of comparative economic history'.



Culture, Confucianism and Comparison

Janet Hunter


This brief paper raises issues but does not claim to offer answers. It hopes merely to stimulate discussion of an area that too many economic historians have been reluctant to come to grips with.

  1. Culture
  2. We cannot pretend that any picture of economic change is complete without incorporating non-quantifiable elements, and that applies to any comparative exercise as well. The recent emphasis on institutions is one reflection of this recognition, but this approach rarely tackles what is in some ways an even more fundamental question, namely why certain socio-cultural (not just political and economic) environments give rise to (in)appropriate institutions and what a socio-cultural context may mean for the operation of institutions. Those concerned with comparative economic history/development thus come up against two specific obstacles. One is that we don’t have an accepted theoretical framework for dealing with the relationship between culture and the economy, even within a single economy, let alone on a comparative basis, and few are willing to accept any kind of determinist paradigm for this relationship. The other is that since socio-cultural factors tend to be regarded as the enemy of systematic social scientific comparison, when we think in terms of comparative economic history we find ourselves compelled us to take a stance somewhere between the universalist paradigms of development that assume common features or at least convergence towards them, and the view of historical development as lots of unique and separate stories,

    Economic historians’ reluctance to tackle this issue is compounded by two additional factors. One is the problem of trying to define what we are talking about when we use the word culture. However, disagreement over definition is found in the case of many other concepts, and is not in itself sufficient reason not to talk about it. One potentially useful definition is that of the anthropologist Robert Smith, who argues that culture is ‘formal and informal rules, learned and shared information pool, thus impacting on both the production function and the utility function, and establishing the range of choices for action’. This permits inclusion in the term of belief and ideology as well as social modus operandi. The second factor is the persistent perception of culture as a fixed and unchanging tradition, when it is usually quite the opposite, subject to a constant process of change and reinvention. We are thus talking about a moving target.

  3. Confucianism

We can all cite examples of literature and debates that have attempted in some way to deal with this issue, but one prominent example of the problems inherent in approaching the relationship between culture/thought/belief and economic development is the ‘Confucian’ economic model often applied to Asian economies from Japan through to Singapore. Certainly all these economies lie within what is conventionally thought of as the Chinese cultural area, but history has produced enormous diversity in the forms of ‘Confucianism’ that prevail. Among the issues this application raises are:

  1. Is this use of a ‘Confucian’ cultural-geographical area as a yardstick for comparison really legitimate, or is it merely a lazy way of trying to account for the growth of Asian economies very diverse in size, endowments and historical experience, and perhaps largely a response to Eurocentric interpretations of development?
  2. If it is a valid concept for comparing economies, it must be linked to more apparent features of these economies. For example, the link between government intervention and Confucian values is often taken as pivotal to an ‘East Asian model’, and the prevailing value systems may provide a more conducive environment within which states can operate effectively according to the Gerschenkronian paradigm. If that is so, we need to think about whether such a supportive value system is unique to Confucianism or shared with some other cultures. Moreover, if this is the key attribute of Confucianism as far as economic development is concerned, is what we should really be concentrating on for the purposes of comparative explanation the imperatives of late development that face all ‘relatively backward’ economies. This might make the Confucian value system merely a red herring.
  1. Comparison

So where does this leave us? There are many questions and few answers. The following observations may help to act as a basis for discussion.

  1. The problems of incorporating culture/belief/ideology into models are well known. The ideas of eg. Hoselitz have been little incorporated into mainstream thought on economic development, and the issue continues to be dealt with largely at an intuitive level. NIE has suggested that cultural endowments make institutional change more or less costly, and there is the concept of ‘mental models’, but I am not persuaded that these attempts to incorporate such things into the neoclassical approach leave us much the wiser. And yet, without a satisfactory model (in the sense of an interpretative rather than predictive model), comparison will continue to be plagued by the problem of balancing the ‘cultural peculiarities of each economy’ with what have been called the ‘universal concomitants of the industrialisation process’.
  2. In the absence of any such interpretative model at the macroeconomic level, perhaps the most practical way forward on this issue is comparison at the micro-level, comparing aspects that are more easily pinned down: commodities, institutions, growth rates, levels of competition etc. We must ensure, however, that any interpretation of our findings does not relegate the socio-cultural/ideological context to an error term, or ignore it altogether, but instead views it as a key factor that may constrain or determine the production and utility functions in those economies being compared.
  3. At the level of a single economic unit, whether the nation state, the family, the village, the locality or even the region (eg. W. Europe, E. Asia), culture and ideology are an important part of any overall explanation of economic change, helping to illuminate why and how certain outcomes are achieved. However, comparison between different economic units at the same level, both within economies and across economies, must start from the level of results. Culture/belief are essentially integral to the process whereby outcomes are achieved. They are key elements in the ‘why/how’ questions which are our ultimate concern. These questions are inherently much more difficult to answer, and more interesting, than the ‘what’ questions, but it is understanding ‘what’ happened that is at present the only way into understanding why and how.




Barbara Ingham

Pattern Modelling in Development Economics 

"The historical dimension [needs strengthening] so that we understand how things came to be what they are, so that we may know the limits of, and opportunities for, desirable change" [Paul Streeten, 1984]

Most development economists agree with Streeten that "history" plays a vital role in development economics in explaining "how things came to be". More controversial, however, is analytical economic history, specifically pattern modelling in development economics. Analytical economic history seeks to identify an intelligible pattern in the development experience. Usually it employs a comparative approach, pairing countries which have significant similarities and / or differences. The method of enquiry is also distinctive, i.e. a combination of deduction with extensive qualitative and quantitative induction.

Development economics is the only area of mainstream economics that continues to employ historical inductive methods of inquiry. Lal and Myint (1996) who employed them in a World Bank sponsored study of 21 developing countries between 1950 and 1985, were at pains to stress their controversial nature. Others, such as Snooks (1993) express regret that economics does not make greater use of inductive methodologies,

"the discipline has been transformed from the practical empirical art of 'political economy' to the abstract deductive science of economics. While this methodological transformation was necessary for the development of the discipline, it brought with it significant costs, particularly the loss of applied and historical skills…".

How useful are inductive methodologies and the associated pattern approach? They are the methods of inquiry that give comparative approaches their distinctive quality. In particular:





Bishnupriya Gupta

London school of Economics

Economic performance and outcomes have diverged widely across countries. As the industrial revolution changed the wealth of nations, one measure of prosperity has been the level of industrial development. Comparisons between Britain and the late industrializers in Europe focussed on the share of the industrial sector in national output and the rate of growth in manufacturing output. The "catching up" by the United States and the subsequent "surging ahead" was measured in terms of GDP per worker. Comparisons of economic growth in the post war period have been made in terms of growth in productivity measured by output per worker. More sophisticated measures include "total factor productivity". In all these measures, the ultimate comparisons of success and failure tend to be measured by rates of economic growth. Economic historians writing on comparative economic development in the developing world have also focussed on growth of GDP or GDP per capita. Reynolds described the change from extensive to intensive growth as the "turning point". In many developing countries this was a 19th century phenomenon as these economies had rapid economic growth as they integrated into the international division of labour. In others, this came with political independence in the second half of the 20th century. Comparisons of GDP per capita placed Argentina among the top economies of the world in the 19th century, Her decline in the course of the 20th century as she came to belong in the group of developing countries has been explained by lower economic growth.

Economists and economic historians have used quantitative methods to measure the effects of structural change in low-income economies. The relative achievements of developing countries in response to the Great Depression have been measured in terms of relative rates of growth in industrial output. as these countries adopted policies restricting trade.

Developing countries moved further away from an international division of labour after the Second World War. The success of import substitution was measured by the relative rates of industrial growth. The subsequent criticisms of these policies were made in terms of slow growth in productivity and the tapering off of industrial growth.

Economic growth has been explained in relation to rate of investment, level of human capital, degree of openness and also factors like income distribution. etc. Econometrics provides us with an important tool to test how important each of these variables are in explaining economic growth and if generalisations can be made. Econometrics also allows us to estimate of long-term growth and therefore control for biases arising in point-to- point estimates. Long term growth is a useful measure for making comparisons and has raised the question of convergence in long term growth across countries: Are all countries converging to a common level of per capita income, as predicted by the simplest versions of the Solow model? It has been found that this is not the case --- there seem to be two groups of countries, with convergence within but not across groups. Bringing human capital as an explanatory factor is growth makes convergence conditional on the stock of human capital.

The East Asian economic miracle has focussed on comparisons of growth and productivity. Alwyn Young’s paper entitled "The Tyranny of Numbers" has convincingly demolished the argument that the Asian tigers have grown due to massive increases in productivity often explained by the openness of these economies. His statistical analysis shows that most of GDP growth can be explained by factor accumulation, that is increase in investment and labour participation rather than increases in productivity. Rodrik finds the predominant the role of human capital and initial conditions such as the level of education and income distribution, using quantitative methods. There has also been a large literature (e.g. Pertotti) on the relation between inequality and growth, with some evidence that additional equality promotes growth.

It is pertinent to argue that economic growth does not explain the well being of the majority in a society as the economic gains may be enjoyed by a few. Do quantitative methods provide us with tools to measure well being? Once again, it may be argued that quantitative methods are useful for comparative analysis, Comparisons across countries are meaningful when we look at quantitative indices that seek to measure well being- for example indices constricted using measures on GDP/capita, literacy, life expectancy and infant mortality. These indices may also include political and civil rights enjoyed by citizens. The interesting comparisons have been made between countries with relatively low per capita income and high social indicators, e.g. China and Sri Lanka, on the one hand, and countries with higher incomes but lower social indicators, for example Brazil and Jordan. A negative relationship between per capita income and under 5-mortality rate for 120 countries also finds outliers like China and Oman. Quantitative techniques also allow us to answer the following question: are there common relationships between different economic variables, for example fertility rate and per capita income.

Generalisations have an important role in understanding development experiences of countries across a wide spectrum. It is an important tool to understand why some countries grow faster than others do and can growth improve the well being of the majority of people in a society. It is not merely an academic exercise, but is also important for policy makers in changing conditions of underdevelopment.


Commodity chains and comparative economic history: the case of cocoa in the liberal era, 1765-1914

William Gervase Clarence-Smith (wc2@soas.ac.uk)

Studying a commodity chain is like cutting a cross-section through history. The focus on a single product engenders a laser beam, illuminating hidden corners of the past, revealing unexpected connections between apparently discreet phenomena, and testing general theories about the wealth and poverty of nations. Looking at the same commodity in different geographical settings allows an assessment of variables. For example, if a crop is grown on estates here and by smallholders there, the nature of the commodity cannot, of itself, dictate the mode of cultivation. Similarly, tea and silk powered Japan into development, but failed to work the same magic for China.

Extending Appadurai's metaphor of the 'social life' of commodities, cocoa had a life cycle that went 'from plantation to cup', and involved two major 'life crises'. In its youth, cultivation depended both on the skills of farmers and the social relations governing access to land, labour and capital. In 'middle age', sacks of fermented and dried beans were not transacted in entirely impersonal markets, but were caught up in a network of social relations and political institutions. Each agent in the chain of trade and taxation attempted to maximise rents and profits, as sacks were nudged along from field to factory. In its 'old age', further transformed into a multitude of products by the beverages, confectionery and baking trades, and passing through another network of wholesalers and retailers, cocoa 'died' in the act of consumption.

At the beginning of the chain, the chief question is why economic 'take-off' did not occur in cocoa producing territories. Embroidering on Bulmer-Thomas' notion of a 'commodity lottery', the ideal agricultural staple yielded forward and backward linkages, transfers of technology and skills, foreign exchange, state revenues, and income for entrepreneurs. If locally consumed, it lowered the cost of living, and hence the wage rate. Finally, it was sustainably produced, without damaging the environment.

Cocoa engendered few linkages, but cheap inputs and low imports advantaged poor countries chronically short of capital and foreign exchange. Simple hand tools were most effective for cultivation and primary processing, and new transport links were of little significance. There were more chocolate factories in tropical countries before 1914 than is usually appreciated, but fixed capital formation was low, and they never exported on any scale.

Bad governance was more of a problem. Independent states squandered funds from the 'golden bean', and colonial authorities lacked vision for investment. States failed to protect the forest, favoured estates and labour coercion, discouraged savings, restricted the immigration of entrepreneurs, fostered cartels, and interfered in marketing. There were no economies of scale, so that estates merely pushed up the import bill and leaked profits abroad. Above all, states taxed heavily without providing essential public goods. Western governments did not intervene to reduce the price of cocoa, but fiscal and protectionist policies at times severely depressed final demand for chocolate.

Cocoa also suffered from price volatility, due to sharp variations in supplies, a problem common to all tree crops. Vulnerable ageing cocoa trees fell victim to natural disasters, causing prices to shoot up. This incited producers to plant new trees in a rush, but the lag of several years between planting and full production provoked a surfeit of new planting, leading to gluts and price falls. Producers compensated by ceasing to plant, and the cycle moved in reverse, pushing prices up again. This cycle might have levelled out over time, as producers learned to anticipate its effects, had it not been for planters based in the tropics forcing up prices through withholding schemes from the 1890s.

A further liability of cocoa was its voracious appetite for virgin forest. Pioneer cocoa cultivators benefited from fertile soils, enriched with ashes, and they offset the cost of clearing forest by selling timber and firewood. There were few weeds, pests or diseases in such soils, and rainfall may have benefited from remaining forest stands. Producers seeking to replace decrepit old trees faced the opposite scenario, and costly additional inputs of labour and capital made their product uncompetitive. This 'forest rent' made cocoa cultivation a wasting asset, and drove the cocoa frontier restlessly along, exacerbating uncertainties in supply and price volatility, and threatening the very existence of the tropical forest.

The intermediary stages in commodity chains are usually the most obscure, pointing to a dramatic gap in the research strategies of economic historians. Too little is known about peddlers, storekeepers, muleteers, boatmen, cask-rollers and porters. Even exporting merchants and shippers have been poorly studied in relation to particular commodities. The least is known about merchants, dealers and brokers in importing countries. They are routinely accused of conspiracy, and yet barriers to entry were low and competition was intense. Moreover, they could easily be cut out by manufacturers. The fact that vertical integration was rarely attempted and never successful suggests that such accusations were groundless.

Intermediaries formed tight knit 'communities of trust'. Agency was crucial, given high risks and delayed returns, and ethnic or religious groups could threaten deviant members with 'social death'. However, competition was not eliminated. Entrepreneurial communities encouraged fierce rivalry, not only with other groups, but also within their own ranks. What remains obscure is why particular peoples were attracted to specific commodities. In the case of cocoa, the 19th century transition from Basques and Sephardic Jews to Hanseatic Germans, Corsicans and Genoese remains to be fully explained.

Accusations of cartelisation have also been frequent in manufacturing, as large Western chocolate companies began to emerge from the 1850s, exploiting new economies of scale in roasting and grinding. The only properly studied national chocolate industry is that of Britain, where the three largest firms reached agreements on pricing and advertising in the late 19th century, assisted by close personal ties between inter-related Quaker families. However, these agreements were limited in scope and largely ineffective prior to 1914. Moreover, such links were not replicated in the chocolate industries of the Netherlands, Germany, Switzerland or the United States. The manufacturers of France and Spain remain to be studied by historians, but there are no obvious signs of a British situation.

Chocolate was subjected to unpredictable whims of fashion, and much affected by competing products. The Western temperance movement was a powerful ally, but it also favoured other substances, notably tea and coffee, but also guaraná and maté in the Americas, salep in the Middle East, kola in Africa, and betel in Southeast Asia. Technical change probably played a lesser part in competition for market share before the 1880s than product diversification and advertising, a situation that surfaced again after 1914.

Chocolate was a prey to fiscal pressure and protectionist regulations. Customs dues on cocoa beans and sales taxes on prepared chocolate were crushing in the mercantilist era, as on other 'luxuries'. Liberalisation gathered pace from 1765, but suffered a reversal in the wars around 1800. Furthermore, its geographic and chronological tempo was highly diverse, and the temptation to tax this discretionary item in people's consumption remained, only tempered by fear of political unpopularity.

Topik and Wells consider that exporting primary commodities was a 'dead end', but that is too harsh a judgement. The removal of mercantilist shackles from the world economy from the 1760s brought real benefits to producers and consumers alike. To be sure, liberalisation was partial, hesitant, and strongly opposed by vested interests. Moreover, there was no take-off into self-sustained economic growth. However, liberalism was no mere trick on the part of dominant countries to sustain the exploitation of weaker ones at lower cost. As long as liberal policies were hegemonic, the economies of most cocoa producing countries were moving in the direction of a sustained rise in incomes per head, the closest there is to a working definition of development. As liberalism collapsed in the holocaust of 1914-18, that evolution stopped or went into reverse.

Policy makers of today could do worse than to ponder the successes and failures of the liberal era. Glib talk about globalisation conceals the fact that the international environment remains less free today than in 1914. If coerced labour is now marginal, both workers and entrepreneurs face huge obstacles in crossing frontiers. There have been ill-judged attempts to resuscitate cocoa estates, even to set up state farms, and government meddling in marketing lingers on. The world payments system has not returned to gold standard conditions, and protection and cartels persist, from 'fortress Europe' to shipping rings. Worst of all, Third World governments continue to gang up to drive up the cocoa price, harming the industry over the longer term.



External Shocks - the possibilities and limitations of comparative economic history.

Tom Tomlinson (University of Strathclyde)

Preliminary statement for meeting of round-table discussion of ‘Methodology of Comparative Economic History’ at History and Development Group Annual Workshop, SOAS, 10th May 2000.

These are introductory remarks - they are not to be taken too seriously, and should not be used or cited without permission. If you wish to discuss them further, and cannot attend the Workshop, contact me here.

Before we can discuss the use of comparative methods there is one crucial definitional problem that must be addressed - the well-known paradox that there is no simple collective phrase in English to describe those parts of the world that we all study which does not define it in terms of what it is not. Terms like 'the non-European world', or 'the developing world' contain obvious traps for the unwary; the term 'the Third World' seems, in retrospect, to be a special rather than a general description, and to be limited in validity to the historical period from the mid-1950s to the late-1980s. More imaginative coinages, such as 'Gondwanaland' (a reference to the fact that all of present-day Africa, Latin America, South Asia and Oceania was once part of a distinct super-continent sited well away from Laurasia - the geographical ancestor of the rest of Eurasia and North America) does not quite work, and ignores the extent to which similar environments and ecologies, and similar histories of human colonisation, have been created since the Earth's tectonic plates stabilised 50 million years ago. My current preference is for 'the Tropics' (in a loose geographical sense), which encapsulates the simple truth that the regions south of 20 degrees of latitude north and north of 20 degrees of latitude south were dominated by a physical environment that did not respond well to European agricultural technologies, that did not encourage European settlement (or that gave indigenous peoples valuable weapons - especially endemic diseases - to resist European incursions), and that required its inhabitants to adopt structures of social organisation that were distinctively different from those found in temperate regions. But strictly speaking this term excludes the northern half of South Asia, and the whole of West Asia, plus China and Japan, and works well only for the period of European industrial dominance in the world (c1820 to c1960?) when the 'tools of Empire' devised in the northern temperate regions of the world acted as the chief driving force for development throughout the globe.

Using the comparative method has always held great attractions for economic historians - especially, perhaps, those working on Asia, Africa and Latin America. There are a number of reasons for this, some more respectable than others.
1) The first problem is that most economic historians seeking to explain the historical process of development in Asia, Africa, Latin America face is the inapplicability of the general laws that dominate the core of the discipline. We are dealing with places where the usual assumptions of neo-classical or Marxist economics, on which so much orthodox economic history is based, do not seem to appropriate. Supply curves are, apparently, backward-bending; individuals are apparently, not profit-maximising; class formations are apparently, fragmented or imperfect. The general laws that have been devised to explain the history of these places - based on concepts of underdevelopment or colonialism - seek to impose simple models that operate only at a very high level of generalisation, and that have proved increasingly unsatisfactory. Looming behind these issues is the problem of monoeconomics - are economic laws the same for all times and places, or should the underlying behavioural assumptions of economists be adapted for different circumstances - and if so, how?
2) The second problem is that of exceptionalism. Given the difficulties of operating in the field, and the problems of extracting economic information from the overlapping complexities of social, cultural and anthropological analysis, much of the work that is done - especially by those with the energy of youth - is based on small units. This makes it difficult to generalise. There is the associated problem that so many of the categories used by economic historians (in the past, at least) were built around the categories of colonial administration which glorified in the particular; or around cultural categories, which again explained the exceptionalism of each case. Thus in South Asia, for example, much effort was expended in teasing out the impact of different forms of land revenue collection - ryotwari and zaminadari - rather than looking beyond the categories in the records to structural divisions between 'wet' and 'dry' agricultural regions etc.
3) There is also the problem of 'presentism'. Much of the literature on individual areas has been concerned with 'writing history backwards' - with proving the inevitability of current conditions and the differences between them, especially differences in recent economic growth. We need to find a way of asking whether historical circumstances are in fact so distinct in areas which currently show variations in economic development - to write history sideways rather than backwards.
4) Because it is hard to apply general laws to the areas we study, it is hard to justify a purely economic analysis of them. One result has been to glorify the particular, with economic phenomena being explained in terms of cultural, political or social structures. We need some way of suppressing the extremes of particularism, while retaining a sceptical view of the general explanations that are currently on offer.

A comparative methodology helps to provide a solution to these problems. But what you pick for your comparison will have significant effect on your findings - reading many existing studies leads to the suspicion that scholars select the comparisons they want to prove a case they know in advance. Compare, for example, the literature on comparative industrialisation among Gersenkronian 'late-comers' (Germany, Russia, Japan) with comparisons based on impact of imperialism (China and Japan), or on different political structures (China and India). In general, comparative studies follow one of two approaches - either to show what is different about apparently similar places, or to show what is the same about apparently dissimilar places. In the first case historians are using the differences between places to identify a crucial variable (role of state, impact of external forces, ideology of government etc.). The alternative perspective uses a comparative method to show what is the same, rather than what is different. This is much more difficult and more open-ended, but much more stimulating.

What is the difference between the comparative method and the construction of general explanations? Comparative methodology allows for the details of each case, but tries to establish similarities and differences, and tends to argue by analogy and coincidence, rather than by conceptual rigour. The essential problem with the comparative method is that problem is that of assigning meaning to particular pieces of data outside a particularistic context. This problem is clearest for statistical data - what do comparative data on, say, agricultural productivity mean? There have been a number of classic attempts to provide this for Asian economies (India, China, Japan) - but do these tell us anything, except about the different as in which such data are gathered? Qualitative data raises the same problems. What does a comparative description of landlord/tenant relations tell us? We need some conceptualisations of the 'essential' nature of society to understand this. Contrast 'paternalist' landlordism in Japan, with feudal landlordism in India, for example. These problems refocus our attention on the basic question for economic historians of non-Western world - are essentially western concepts of economic behaviour and motivation appropriate to understand the mechanisms of change and development - are they universal or not?

The study of external shocks provides a good test of the strengths and weaknesses of the comparative method, since such shocks are assumed to be universal in their impact. If we find similarities or differences in their local effects, we can begin to hypothesise about the nature of economic structures and institutions in different parts of the world, and to differential particular from general causes.
One problem is that the number of such external shocks is limited, but we can illustrate the issues by considering one at least, the impact of the Great Depression of 1929-33 (henceforth GD) on economic activity in Africa and Asia. This is a good example to pick because there is a substantial literature on it, and at least one well-known volume that can provide sophisticated comparative data, to which several of the members of this Group have contributed - Ian Brown (ed.), The Economies of Africa and Asia in the Inter-War Depression (London, 1989). This book shows both the potentialities and limitations of comparative methodology. It consists of 12 essays on aspects of economic change in the major regions of Asia and Africa in the 1920 and 1930s. Most of the essays are based on the analysis of coherent political units - states, colonial territories, or provincial or local units of administration, although two - by Susan Martin and Gervase Clarence-Smith - are based on economic regions, rather than administrative units. Perhaps the most significant feature of the volume for our purposes is that the text consists of 279 pages, but the attempt to produce a comparative analysis in Ian Brown's 'Introduction' takes up only seven of them. The 'Introduction' states explicitly that it does 'not attempt to bring together the conclusions of the main contributions into a rounded, concluding statement. The diversity and complexity of the impact of the depression in these continents and the still preliminary state of scholarship in this field mean that it is premature to establish a new orthodoxy'. The aim instead is to 'sketch in a number of general themes [and] fix a number of points of reference, which may guide the reader through the detailed explorations which follow'. [page 2]
The purpose of the book, according to the 'Introduction', is to challenge the general laws set out by an 'old orthodoxy' which stated that the GD was a traumatic event for all colonial, peripheral economies since they were based around the export of primary produce. The new research which is showcased in the volume demonstrates three important modifications to this view:
1) the 'great contraction' of 1929-33 was not a single, traumatic event. There were separate crises in the early 1920s and late 1930s, and also an underlying 'long depression' in tropical agriculture brought about by overproduction and changed consumer preference, which led to continued falls in the terms of trade.
2) the economic events of the inter-war period had a differential impact in different economies, and in sectors within economies. There were significant variations in how much they affected by external factors, and, especially, there was a significant stimulation for industry that off-set the contraction for agriculture. The most important effect of increased industrialisation was 'export substitution' (Clarence-Smith's phrase, page 196), which had more important backward linkages than the more commonly-identified phenomenon of import-substitution.
3) individual economies varied widely, and so made different responses. Different groups within particular economies were able to respond in different ways - often finding new opportunities even within patterns of existing economic activity.

The overall conclusions of this volume succeed in denting the old orthodoxy about the GD considerably, and leaves us forced to consider whether this classic external shock was, in fact, either external or a shock. It is clear that the rapid fall in international prices for primary produce in 1928-9 and, especially, the problems of liquidity associated with this phenomenon, did have an impact on African and Asian economies, but that this impact was much more complex and variable than had been thought hitherto. Many of the essays in the volume make the point that economic actors in Asia and Africa were not passive puppets whose actions were determined by forces beyond their control, but that their active participation in local economic institutions provided new opportunities within the internal economy to balance the falling profitability of exports. Discovering that local participation was robust and sophisticated, and that similar reactions and innovative strategies were found in many distinct regions, with different cultures and political structures, provides a way to restore economic explanations of development issues to the centre of the Use of uninitialized value in concatenation (.) or string at E:\listplex\SYSTEM\SCRIPTS\filearea.cgi line 455, line 447. stage, provided that such explanations are sensitive to the different, and often imperfect, institutional and market structures that existed in different parts of Asia and Africa in the 1930s. A comparative study of the GD allows us to investigate the relationship of economic institutional structures to those of informal and formal structures of political and social power. The GD was a case of market failure - from a comparative study of its effects we are able to deduce how the market for export-based cash crops worked, and what alternative market systems underlay this, and could replace it. Thus the ability of African and Asian economies to respond to the events of the GD was a symptom of the strength of market systems there, not an indication of their weakness.

Perhaps the best that can be offered by comparative studies of this nature is that they provide a series of typologies - a range of the different things that can happen, with local explanations. If we find the same things happening elsewhere, where local contexts are different, then we can begin to expand the literature significantly, and to introduce general rather than particular explanations. By doing so we can then close the gap between general and particular explanations, and begin to refine the general laws of economics that can be applied across the 'seams of Pangea' (to use Crosby's phrase), in both tropical and temperate economies.