The Conditions of Agricultural Growth: The Economics of Agrarian Change under Population Pressure

Project 2001: Significant Works in Economic History

Ester Boserup, The Conditions of Agricultural Growth: The Economics of Agrarian Change under Population Pressure. London, G. Allen and Unwin, 1965; Chicago: Aldine, 1965. 124 pp.

Review Essay by Giovanni Federico, Department of Modern History, University of Pisa.

Population, Agricultural Growth and Institutions: The Real Long-Run View

This may be an unusual review for the series. In fact, Ester Boserup was not a professional economic historian and this is not properly speaking a work of history. Boserup was part of the staff at the United Nations and she wrote the book out of her experience as a consultant in developing countries. The book does not discuss in depth any specific historical event, and quotations of historical works are rather rare. It nevertheless is one of the most widely quoted works in economic history. Usually, it is labeled as “anti-Malthusian” and encapsulated with a sentence such as “population growth causes agricultural growth.” This is undoubtedly an implication of her model and comes in handy to scholars who do not believe that the (human) carrying capacity of a given area is set, and cannot be exceeded. From this point of view, one can draw a parallel between The Conditions of Agricultural Growth and another highly influential book, Amartya Sen’s Poverty and Famines: An Essay on Entitlement and Deprivation (Oxford 1981), which dismantled another tenet of Malthusian theory — i.e. that famines were always (or mainly) caused by absolute deficiency of food.

However, Boserup’s book is much more than a simple rejection of Malthus. It aims at explaining all the characteristics of agriculture in any specific area and time according to the resource endowment — the land/labor ratio. The more dense population is, the more intensive cultivation becomes. Agrarian economists in the 1950s focused on the Western world, and thus they could appreciate only a relatively narrow range of techniques. Looking at less developed countries, Boserup could list five different agricultural systems, according to the length of fallow between periods of cultivation (pp.15-16): 1) forest-fallow or slash and burn (15-20 years of fallow), 2) bush-fallow (6-10 years); 3) short-fallow (1-2 years); 4) annual cropping (a few months); 5) multi-cropping (no fallow). Even if the original evidence comes from the observation of primitive societies in the 1940s, the leap from changes in space to changes in time is short. Thus the rest of the book explores the consequences of intensification — i.e. of the move from one stage to another caused by population growth. Each of them entails more labor per unit of (total) land, and thus the intensification increases the productivity of land and reduces that of labor. A household has to work more to keep the same level of income. The intensification brings about an improvement in tools (from the digging stick, to the hoe, to the plough) and in the long run also brings some investments in land improvement (e.g. irrigation schemes). With pre-industrial technology, land improvements had to be done manually by peasants. Thus, they are typical of the last stages of the process, when there is enough work-force and enough demand for food to justify them. Total factor productivity may increase in the long run, but surely most of the increase in total output is achieved with a massive growth of work effort by the agricultural population. Finally, the intensification also shapes institutions, and this is the most innovative aspect of Boserup’s model. The forest-fallow system is inconsistent with household property of any given plot of land. The land belongs to (or more precisely is exploited by) the tribe as a whole. Property rights have to be created only when the cultivation cycle is shorter, and the quality of each single piece of land begins to matter. In the later stages of development some people could cease to work, and be entitled to rights to a part of the product (a “two-tier” society). However, Boserup is not nostalgic about primitive societies. She makes it crystal clear that the “two-tier” societies are better, even if in these latter some men did not work as hard as others.

Some years later, Boserup extended her model from agriculture to the whole of society (Population and Technological Change: A Study of Long-term Trends, Chicago, 1981). She added the concept of economies of scale. Many technologies can be properly exploited only if the population is dense enough. Population growth makes urban civilization possible. The second book is highly interesting, and has many insightful passages. Yet it fails to reach the simple elegance of The Conditions of Agricultural Growth — that quality which makes this book really deserving of being added to this list of masterpieces.

Of course, one could quibble endlessly about the “details” of Boserup’s model such the number and the exact features of the “stages.” The overall view provides a short, but powerful, history of the world, from prehistory to the nineteenth century arranged around one of the basic principles of economic theory — that techniques (and much else) depend on resource endowments. As you would expect from a seminal work, The Conditions of Agricultural Growth launched and refocused many modern debates. Let me give two examples. When Boserup was writing, the British agricultural revolution (i.e. the change in rotations with the substitution of fodder crops for fallow) was considered an epochal change with far-reaching implications for the entirety of world history. This view is still diffused, if no longer dominant. In Boserup’s model, the change is only part of the long-run process of world-wide intensification, and Europe was trailing behind the two other major civilizations, India and China. In fact, the most advanced areas of Europe reached Stage 4 while China was already at Stage 5. Another, and perhaps less obvious, example may be Greg Clark’s thesis on the differences in work intensity between Eastern Europe and the West (including the US). He argues that in the early nineteenth century Eastern Europeans were less productive than Westerners, because they worked less hard, and that they worked less hard because “they were different” (Clark, “Productivity Growth without Technical Change in European Agriculture before 1850,” Journal of Economic History, Vol. 47, 1987, p. 431). The thesis is very controversial (see the subsequent debate with John Komlos in the Journal of Economic History, in 1988 and 1989), but let’s assume it is true. Is it not possible that the “different” work ethic had been shaped over the centuries by different land/labor ratios? Other examples could follow, but the main point is clear: Boserup’s book is a treasure-trove of ideas. Unfortunately, it is more often quoted than used in actual research. As far as I know, there are very few really “Boserupian” works — i.e., long-term analyses of agricultural change as driven by changes in factor endowments. The most ambitious is Kang Chao’s book on Man and Land in Chinese Economic History: An Economic Analysis (Stanford 1986).

Why this (relative) neglect in spite of the so frequent quotations? One can put forward three causes, which are not mutually exclusive. The first is academic specialization. Intensification lasted for centuries, even for millennia, and few scholars would feel at ease in discussing both pre-historical agriculture and nineteenth century techniques. This fate is common to all interpretations of long-term change (cf. J. L. Anderson, Explaining Long-term Economic Change, Basingstoke, 1991). Second, the evidence on early-stage societies is very scarce, and by its nature it is often unfamiliar to historians. “Real” historical sources exist for Western Europe, China and India in the last three stages.

Last, but not least, the model has its own weaknesses. It is surely convincing as an account of long-term growth. It is less convincing as an explanation of short-term trends, and in this case the “short” term can last for decades. Boserup speaks as if all the techniques were known since the beginning, so that the population had only to choose the one best suited to its resource endowment and adjust its institutions if necessary. On the contrary, new techniques had to be learned, and sometimes discovered or re-discovered. In backward economies, information travels very slowly or not at all, and thus a people may not know that another one, maybe hundreds or thousands of miles away, has successful managed to overcome a specific problem. And, even if it gets to know the right technique, plant, or implement, the population still may need time and effort to master it and to adapt it to its own environment. Thus a success in the long run may conceal several short-term crises. Outright failure cannot be ruled out entirely.

Second, Boserup assumes that population growth is exogenous, following a standard practice among economists in pre-Beckerian time. Today, however, most consider population growth to be endogenous, and largely affected by economic calculations. People could reduce population increase by delaying marriages, controlling births, migrating and the like. Slower population growth would, ceteris paribus, reduce the drive to agricultural intensification. This is, of course, an empirical issue.

Finally, Boserup seems to neglect the different nature of modern technology or, if you want, the new role of capital. Her world is a two-factor world — labor and land. As said, capital does exist either as simple tools or as labor-intensive investment projects — but not as labor-saving machinery and above all land-saving fertilizers. In her world, intensification is possible up to a point, but sooner or later it has to reach a limit. It is unclear whether in real history this limit had ever been reached, even if China in the eighteenth and nineteenth centuries may be a good candidate. Aside from China, even in, say, 1800 there was a lot of “free” land on the Earth and thus a “Malthusian” crisis was still far away for the world as a whole. But sooner or later, a limit had to be reached, and further population increase beyond it was bound to cause a Malthusian crisis (even if smart people may prevent it with birth control). As everyone knows, the solution was technical progress, which has increased the productivity of both land and labor. (One wonders whether there are ecological or maybe ethical limits to technical progress). Boserup should have added a Stage 6 to her intensification model. Of course, she was very well aware of the technical progress, but she did not. One may speculate that she was more interested in less developed countries than in advanced countries, or simply she did not want to add a stage which could not fit easily in a model based on the length of fallow.

It is too easy to criticize ex post with the hindsight of decades of research. In spite of all its shortcomings, The Conditions of Agricultural Growth remains a small masterpiece which economic historians should read — and not simply quote.

Giovanni Federico is the author of An Economic History of the Silk Industry, 1830-1930 (Cambridge University Press, 1997) and (with Jon Cohen) The Economic Development of Italy, 1820-1930 (Cambridge University Press, forthcoming for the Economic History Society).

?