The economic history of agriculture since 1800

GIOVANNI FEDERICO

Introduction: two centuries of successes

Today agriculture does not enjoy a great reputation. In advanced countries, it is blamed for being inefficient and polluting.

Mapping the growth of total production in the first seventy years of the nineteenth century is not easy. There are few, and often only tentative, series of data, which refer almost exclusively to countries in Europe and North America. In none of these countries, however, other than Portugal, did total output decline, and in the majority of cases it increased faster than popula­tion. We cannot rule out that these gains were offset by a decline in output per capita elsewhere in the world, but this hypothesis is not terribly plausible. The world's growing population was mostly employed in agriculture and, as we will detail in the next section, land was abundant. The quantitative evidence is more solid after 1870. It is possible to estimate an index of “world” output from national production series for twenty-five countries, ^[54]

accounting for about 50-55 percent of world population.^[55] The list includes all European countries (except the Balkans), the United States, Canada, the main Asian countries (India, China, Indonesia, and Japan), and three countries in South America (Argentina, Chile, and Uruguay).

If the output per capita of the missing countries had been growing as much as their population (a reasonable hypothesis), total world production increased by 90 percent and per capita output by about 10 percent. Figure 3.1 then links the twenty-five country series to the series by the Food and Agriculture Organization of the United Nations (FAO), which covers all countries. From 1938 to 2010, world production increased by almost five times, and output per capita by 60 percent. This total includes some dubious series for LDCs and for socialist ones (China, USSR, and satellite countries until the 1990s), but no plausible bias in country series can question the outstanding performance of agriculture. The output is more than sufficient to feed the world population. Undernourishment, which, according to the latest data by the FAO, still affects about ι billion people, is a consequence of inefficiency in distribution and wastage.

Production can be augmented by using more inputs - capital, labor, and land - and/or by using them more efficiently. Section two of this chapter outlines the growth in inputs, and shows that it can account only for a part of the increase in output. Most of the growth was the result of an increase in productivity, and the remaining sections in this chapter deal with its causes. Growth in productivity is usually identified with technical progress, and section three describes the main innovations and the pattern of adoption. However, efficiency depends also on how factors are allocated and techniques are used - and thus ultimately also on institutions (section four) and policies (section five). The final section speculates about the prospects for the future.

The growth of inputs

The evidence on the growth of inputs is fairly abundant, but incomplete. Before 1913, sources refer mostly to Europe and North America, and during the interwar years the country coverage, although increasing, is still only partial.

Table 3.1 proxies the amount of labor with the number of workers, of both genders. Thus for every 100 workers in agriculture in Africa in the year 2000, for example, there were 51 workers in i960; and for every 100 in Europe in

Table 3.1 Workforce

Source: Federico, Feeding the World, tables 4.16 and 4.17

* including Mexico and Central America

2000, there were 309 workers in i960.

The table highlights a key distinction between the number of agricultural workers and their share of the total workforce. This latter is bound to decline as a consequence of modern economic growth, but the relative decline does not necessarily cause the absolute number of workers, the labor input, to decrease. Actually, the number has declined only in a handful of advanced countries and only long after the start of their modern economic growth. The agricultural workforce did peak in the United Kingdom around 1850, but in other advanced countries of Europe and North America it peaked sometime in the first half of the twentieth century, and collapsed only after the Second World War. Worldwide, the agricultural workforce grew because of global population growth until 2010, although at a decreasing rate.

The number of workers might not be an accurate measure of labor input. In fact, it assumes a constant number of hours per worker and a constant intensity of work. It omits part-time work by non-agricultural workers (e.g. at harvest time) and does not take into account changes in the human capital of workers. The evidence and some reasoning suggest that these biases work in different directions. For instance, the human capital of agricultural workers has increased spectacularly in advanced countries after 1950, but the trend has been compensated by a decline in the number of hours worked and possibly in the contribution from non-agricultural workers. The net effect is difficult to measure but probably not very large - so that the total increase of the input of labor should not differ much from the headcount.

Table 3.2 reports comparable indexes for the extension of cropland and tree crops, the best available proxy for land input. Since 1880, acreage has increased all over the world.

Table 3.2 Acreage

Source: Federico, Feeding the World, tables 4.1, 4.3, and 4.5

* including Mexico and Central America

but there is very little doubt that acreage had been growing parallel to population in all areas, except perhaps Western Europe.

There are two reasons to suspect that figures in Table 3.2 overstate the actual increase in land input. First, they assume that all additional cropland was not productive at all. This inference is plainly wrong for pastures. Unfortunately, the data on acreage under pasture before i960 are scarce and plagued by inconsistent definitions. Since 1950, pastures have increased about as much as cropland - i.e. this latter is not a biased measure of total acreage in use. Second, the figures would overstate the increase if the best land were settled first. This seems a common-sense rule, but it does not hold true for large-scale colonization, which depends on the available infrastruc­tures. For instance, the pampa humida in southern Argentina is more suited for agriculture than the area around Buenos Aires, but it was settled later.

It is impossible to sum up trends in agricultural capital in a simple table, because it consists in a number of widely different items - tree crops, buildings, livestock, irrigation works, tools and machinery, and so on - which must be added together in monetary terms to get a meaningful total. The FAO provides an estimate of “world” capital stock, which shows a 25 percent increase from 1975 to 2000 - i.e. much less than output. For the period before 1975, there are few series, which refer only to advanced countries and often cover only a subset of items. Most of them show an increase, which is, unsurprisingly, faster in the USA, Canada, and Russia than in Western Europe. It is thus necessary to use the information on specific components of capital, such as the number of tractors or the extension of irrigated land. Such an analysis highlights four different patterns:

(ι) In 1800, the capital stock in the countries of Western Settlement was negligible, except in small parts of old colonization areas, such as the American East Coast. Colonization entailed massive investments, and the rate of accumulation dropped after the end of the process. The adoption of modern, capital-intensive technologies caused a boom in investments from the 1930s onwards.

(2) The “advanced,” long-settled countries ofWestern Europe had tradition­ally a very substantial capital stock. Thus it grew until the Second World War decidedly slower than in the Western Settlement countries. The postwar spurt was similar.

(3) In “backward,” long-settled countries, most notably China, the capital stock around 1800 was quite large, possibly even greater than in Europe because of the extensive irrigation works needed for rice-growing. It grew very slowly or did not grow at all until quite recently and then boomed, with the intensive use of fertilizers and also mechanization.

(4) In “backward,” less densely populated countries, such as many of those in Africa, the capital stock was initially minimal, and it grew only as much as the population did, probably until the Second World War. Since 1950 the per capita stock of capital has increased, but much less than in Asia.

Summing up, all inputs have been growing throughout the whole period, but there is evidence of a slowdown after 1950, at least for labor and land. It seems highly unlikely that the acceleration in the growth of capital stock was large enough to compensate for this. Therefore, the sharp acceleration in the growth of output since 1950 must reflect an increase in the efficiency in using these inputs - or to use the economists' jargon, an increase in Total Factor Productivity (henceforth TFP). This inference has not escaped economic historians and economists, who have produced literally hundreds of estimates of TFP. Table 3.3 shows the estimates available for the period to 1938 as continent-wide averages, though these are often derived from data on only one country. (“Number” on the table refers to the number of countries in the data set.)

The geographical coverage is limited and the estimates are often fairly crude, but two facts stand out. First and foremost, TFP grew almost every­where, and this by itself is a major change from notions of the (alleged) stagnation of traditional agriculture. A 0.5 percent yearly growth may seem slow, but, when cumulated over forty years, corresponds to a 25 percent increase, and this is far from trivial. The cases of falling TFP are very rare: the two negative signs in the table refer to Argentina and Egypt, and one must

Table 3.3 Growth in total factor productivity to 1938

Sources: Van Zanden, “The first green revolution,” Federico, Feeding the World, Statistical Appendix table ιv, Lains and Pinilla, Agriculture and Economic Development

add the Philippines and the Soviet Union in the interwar years. Second, in most countries, productivity growth has accelerated over time. For instance, the growth rate of TFP in the United States was 0.4 percent in 1840-1870, increased by a third in 1870-1910 and then, even according to the most conservative estimates, by a further half in the next thirty years.

The available estimates for the postwar period cover almost all indepen­dent countries with a variety of methods and for different periods. Whenever a direct comparison is possible (i.e. for eighteen country cases) the data show a further acceleration of productivity growth: the postwar rates are on average almost three times higher than the prewar ones. About 70 percent of the estimates are positive, and the negative rates concentrate in socialist countries and in Sub-Saharan Africa. The rates for some African countries are so low as to be somewhat suspicious. These countries drag down the unweighted mean of country rates to “only” 0.7 percent per annum. Indeed computing a “world” rate over the period 1960-2000, as if the whole world were a single country, yields significantly higher figures. They vary, accord­ing to the method of computation, between ι percent and 1.25 percent per annum. TFP grew decidedly faster in OECD countries, where rates range between 1.5 percent and i.8 percent, than in the rest of the world (rates from 0.8 percent to ι percent). The excellent performance of agriculture in the postwar years emerges by a comparison with the rate of TFP growth in manufacturing for a sample of thirty-six countries in 1967-1992: agriculture outperformed manufacturing in twenty-two countries.^[56]

Technical progress in agriculture

In over two centuries, farmers have introduced thousands of innovations, which for the purpose of illustration can be grouped into four categories - new practices of cultivation, new plants and animals (biological innovations), chemical products, and machinery (industrial innovations).

Most new practices aimed at reducing the length of periods of rest for the soil, which was the traditional way to restore soil fertility depleted by cultivation. In traditional agricultural systems, the number of crops per year (or cropping ratio) ranged from 0.05 in the so-called slash and burn systems (i.e. two or three years of crops followed by twenty to thirty years of rest) to figures above one in the most intensive irrigated rice cultivation in Asia, where fertility was restored by water and intense manuring. In most of Europe, land was left idle (fallow) one year out of three - i.e. the ratio was around 0.6. The ratio was increased by substituting fallow with the cultiva­tion of fertility restoring plants, such as grass, roots (e.g. turnips and pota­toes), and maize. This practice had been used in some areas of Europe since the Middle Ages, but in the late eighteenth and early nineteenth centuries it spread widely in England, earning the name of the “Agricultural Revolution,” and with some delay all over the Continent. Towards the end of the nine­teenth century the succession of crops (rotation) became very sophisticated. The increasing use of rotations and of fertilizers improved the cropping ratios. On the eve of the Second World War, they were slightly below one in Europe and the United States, and over 1.3 in China, Korea, and Taiwan, with a country-wide world record of 1.6 in Egypt. Since then, the ratio has increased further, approaching one worldwide at the end of the 1990s, with maxima around three in some areas of Asia.

Any variety of plant (or animal) is new from the point of view of the individual farmer who cultivates or raises it for the first time, but from the point of view of agriculture it is important to distinguish its source - the casual discovery, the transfer from other areas, or the artificial creation via the hybridization of existing varieties. Casual discoveries have always been rare and in the most recent period their role has further diminished. The latest discovery of a new plant, the sugar-beet, dates back to the late eighteenth century, and the latest discovery of a new wheat variety to 1862. The contribu­tion of long-range transfers has been similarly declining. It had been very important during the age of discoveries, and had had a sort of revival in the nineteenth century because of the combined effects of the efforts of govern­ments, which wanted to foster the economic potential of agriculture, and immigrants, who often brought with them the seeds and plants of their native countries. There were some notable successes - such as the introduction of early ripening varieties of Russian wheat in the Canadian prairies or of the rubber tree, originally from Brazil, in Southeast Asia. Hybrid animals had always been known (mules are hybrid of horses and asses), and the theoretical possibility of hybridizing plants was first suggested in the eighteenth century, with experiments starting in the mid-nineteenth century. However, the initial results were disappointing, as techniques were still primitive and not yet supported by the knowledge of mechanisms of genetic transmission, which were (re-)discovered at the end of the century. The first great success was the production of hybrid corn (that is, maize) in the 1930s. In the 1940s researchers started to work on wheat and maize varieties fit for the Mexican environment, and in the late 1950s on rice in the Philippines. The effort paid off handsomely: some varieties, aptly named high-yield varieties (HYV), in the right conditions, produced up to eight times that of the traditional ones. The results of their adoption were so stunning that this transformation earned the nickname of the “Green Revolution.” Since the 1980s, the potential for the production of new varieties has been boosted by genetic engineering. The first genetically mod­ified variety of tomato was commercialized in 1994.

The major contribution of the chemical industry to agricultural progress was the solution, apparently for good, of the problem of restoring soil fertility. In 1840 the German scientist Liebig discovered that soils needed (different combinations of) phosphates, potash, and nitrogen. The commer­cial production of phosphates started the following year and that of potash in 1856. Nitrogen was first supplied by natural sources, such as Chilean nitrates and Peruvian guano, and then by the production of coke. However, output boomed and prices collapsed after the discovery of the Haber-Bosch method of producing ammonium sulfate in 1909. The first chemical fertilizers were marketed in Germany in the 1920s. Chemical products were used since the late nineteenth century to fight diseases and parasites, but results were rather poor until the 1940s and the discovery of DDT.

Water and, later, steam-powered machines had always been used for processing agricultural products (e.g. for milling), but mechanization of fieldwork started quite late. Arguably, the first machine was the wheat harvester, or reaper, which Hussey and McCormick patented independently in 1833-1834. In the next decades, investors focused on increasing labor productivity in harvesting, introducing joint machines for reaping and thresh­ing (the combine) in the 1880s, the corn picker in 1900, the cotton picker in 1907-1912, and so on. However, mechanization was delayed by the lack of a suitable source of inanimate power. Neither water nor steam was, for different reasons, really suitable. Mechanization really took off only after the introduction of tractors powered by internal combustion engines since the early 1900s, and was boosted by the introduction of the power-take-off shaft (or PTO), which transformed the pulling power of the engine into a rotatory movement.

All (successful) innovations cut production costs by reducing the amount of factors per unit of product, but their effects on the demand for factors differ. Some innovations save all factors in the same proportion (neutral), others save mostly one factor (say land), and some others need more of a factor to save others. With the possible exception of rotations, agricultural innovations were capital-intensive - i.e they needed additional investments to purchase seeds, machinery, fertilizers, and so on. Machinery saved labor, by definition, but also land, as tractors reduced the number of horses and other animals and thus the need for feed. Chemical products and new varieties increased production per unit of land and thus saved land.

Basic economic theory suggests that the benefits of substituting for a factor are, all things being equal, greater the more it costs, and that the cost depends on its scarcity. Thus, one would expect that the transfer of technologies from advanced countries to LDCs was hampered by absolute scarcity of capital and by the poor development of institutions for agricultural credit. One would also expect that land-scarce Europe was on the forefront of the adoption of fertilizers and new varieties, while labor-scarce Western Settlement countries pioneered mechanization. Indeed, on the eve of the Second World War, tractors already accounted for about two-thirds of total power used in the United States and for a third in the United Kingdom, the most mechanized country in Western Europe. In contrast, the United States consumed on average 9 kg of fertilizers per hectare, versus 26 kg in Italy and 300 kg in the Netherlands. These differences have narrowed since then, but they still persist: in 1998-2000 Italy consumed 60 percent more fertilizers than the United States, and the Netherlands five times more, while the United States had about double the number of tractors per worker than did Western Europe. Hayami and Ruttan go a step further along this line of reasoning.^[57] They argue that factor endowment affects not only the adoption of innova­tions but also their production. Land-scarce countries invest more in research on land-saving innovations. Their view is controversial, however.

Olmstead and Rhode argue that even the quintessential labor-scarce country, the United States, invested more in biological innovations than in mechanical ones until the 1930s.^[58]

Factor endowment can explain a lot, but it is not sufficient to account for all differences in the rate of adoption of innovations in agriculture across countries, for two reasons. First, most agricultural innovations are environment-specific. A new variety might perform wonders where devel­oped and prove unsuitable in another location, with different soil, water, and so on. Second, many innovations are complementary or interrelated, i.e., they can develop their potential only if adopted jointly. The high-yielding varieties need more fertilizers and more water than the traditional ones. The Green Revolution was not just a change in varieties, but a comprehensive package that transformed agriculture altogether. The environmental suitability of a specific innovation and the right package of innovations can be discovered only with systematic testing in specialized research facilities.

The need for additional testing increases the overall cost of R&D (research and development) but this is not the only problem for the production of agricultural innovations. The main problem is the appropriability of “bio­logical” innovations - i.e., the possibility to recover the costs of successful R&D. In fact, new practices of cultivation or natural varieties are very easy to imitate and hybrid seeds can be produced by any firm with the necessary skills. The inventors risk losing part of the potential returns from their investments and thus the expenditures in R&D may be less than those socially optimal. The gap should be filled by non-profit expenditures. In the late eighteenth and early nineteenth centuries, some money was disbursed by enlightened landlords. Some very enlightened landlords, such as John Bennet Lawes, ceded their estates to set up research facilities (Rothamsted in the United Kingdom). Less generous ones gathered in learned societies, such as the Royal Agricultural Society of England (established in 1838), to test innovations and spread knowledge about new techniques. Some money was supplied by non-profit organizations, such as the American Ford and Rockefeller foundations, which funded the early research on high-yielding varieties in Mexico. However, the resources of private foundations were too limited relative to the needs, while the willingness to commit by landlords was beset by a free-riding problem: why should they organize field trials if they could get the information from trials organized by someone else? Thus, by far the major source of funding of R&D in not appropriable agricultural technologies has been the public purse.

Some research was performed in universities and government-established experimental stations. The idea was pioneered by the government of Saxony (1851), which was imitated by the United States and by most European countries (with the notable exception of the United Kingdom). At the turn of the century, colonial powers funded R&D in tropical cash crops for exports (cocoa, rubber), while research in food crops took off after the Second World War, with the creation of specialized institutions (e.g. CIMMYT in Mexico for wheat, the IRRI in the Philippines for rice), which were co-ordinated in 1971 by the Consultative Group on International Agricultural Research (now CGIAR). The total investment was substantial. In the United States, total expenditures increased from 2 million (1993) dollars in 1889, equivalent to 0.03 percent of gross agricultural output, to 50 million on the eve of the Second World War (0.7 percent), and exceeded 500 million (over 2 percent of output) in the late 1990s. According to the best estimates, worldwide public expen­diture increased by 150 percent in the 1960s, by 50 percent in the 1970s, by 30 percent in the 1980s and only by 15 percent in the 1990s.^[59] By 2000, total expenditure was equivalent to 2.4 percent of gross output in advanced countries, but only to 0.53 percent in developing ones (and to 0.8 percent worldwide). Expenditures for diffusion of best practices among farmers (so- called extension) doubled from 1959 to 1971 and increased by 25 percent in the next decade. The clear slow-down in public spending in R&D was compen­sated, at least in the advanced countries, by a surge in private investments, which was related to a major institutional change, the extension of patenting rights to living species. This measure had been advocated by firms selling seeds and plants since the beginning of the twentieth century and it was granted for the first time, but for trees only, in the United States in the 1930s. In 1960, European Union countries extended the right to all plants and in 1961 they signed an inter-country agreement for the mutual recognition of patents, the International Union for the Protection of New Varieties of Plant (UPOV). The United States followed suit in 1970. As of 2011, UPOV has sixty- nine member countries. Private expenditure in agriculture-specific R&D (i.e. excluding mechanical or chemical research) overtook public expenditure in the United States already in the 1980s, and by 2000 private expenditure

The economic history of agriculture since 1800 accounted for about a third of world total and for over half of the expenditure in advanced countries.

Institutions and agricultural performance in the long run

In the economists' jargon, institutions can be defined as the set of formal or informal rules that determine the ownership of the goods and factors (property rights) and regulate the interactions among individual agents or households (contracts, markets, and other forms of distribution). There is a wide consensus among historians, economists, agricultural experts, and policymakers on the relevance of institutions as a whole for agricultural growth, but also wide differences of opinion about the importance of each type of institution and its effects.

Economists believe that modern property rights are necessary to exploit the full potential of an economy and thus they assume that their diffusion improves agricultural performance. Property rights are not antithetical to personal freedom, yet a sizeable share of the world population was denied this as late as the mid-nineteenth century. The slave trade had been formally banned in 1807, but slavery was outlawed in many countries only much later (in the United States in 1865, after the Civil War, in Brazil in 1888) and survived in some areas, such as Nigeria, well into the twentieth century. Serfdom, which tied workers to the estate rather than to an individual master, was abolished in 1861 in Russia and three years later in Romania.

Economists strongly believe in the benefits of modern property rights to land - including the right to sell and bequeath it.^[60] They prevent the excessive exploitation of land for short-term gains (the so-called tragedy of the commons), stimulate location-specific investments and experiment­ing with new techniques, and allow the use of land as collateral for borrow­ing. By 1800, full ownership prevailed only in Western Europe, in the already settled areas of countries of Western Settlement, such as the East Coast of the United States, and in parts of Asia, including most of China.^[61]

Elsewhere, the rights to land were jointly held. In some areas, feudal lords or other powerful individuals had the right to claim a part of the product and/or of the time of workers. In most areas, however, the rights to land were owned collectively by all users. Tribes of hunter-gatherers used it collec­tively, while in more settled areas, such as Russia, land was allocated to households for cultivation for a predetermined period and under the control of the village.

In the last two centuries, these traditional property rights have been disappearing. The process has been slow, featuring massive reversals, such as the collectivization of previously private land in the Soviet Union in the 1930s and in China in the 1950s, and is not yet fully over. The feudal systems of Eastern Europe were the first to disappear, in the first half of the nine­teenth century. The land was divided between peasants (former serfs) and former lords, who often also got financial compensation. In theory, com­mon ownership could be abolished by transforming the temporary alloca­tion of land into full permanent ownership. In most cases, the process was very gradual, with intermediate stages: in Turkey it lasted from 1858 to the 1940s, in Indonesia from 1870 to i960, and so on. Such a direct transfer was not feasible in Africa, Oceania, and the Americas, where, under traditional systems of swidden agriculture or hunting-gathering, only a tiny fraction of the total land was in use. Anyway, European colonial powers ignored the rights of the native peoples and seized all the land European immigrants wanted. The temperate areas of the New World attracted large numbers of European immigrants and so the governments expropriated most of the land and distributed it to settlers, with different procedures (sales, block concession to railway companies, homesteads, etc.). Indigenous peoples got a somewhat better deal in the tropical countries of Asia and Sub­Saharan Africa. There the demand for land by Europeans was compara­tively modest and most of the land remained under common ownership of Africans. Colonial administrations started to register the ownership of individual farmers (“titling”) in the 1940s and the process has continued since then, with strong support by international organizations such as the World Bank. In 1990, tribal land was down to 0.34 percent of the worldwide total, but it still accounted for 14 percent of land in Africa. The moderniza­tion of property rights must have enhanced efficiency and accumulation of capital, but benefits may have been smaller than hoped for. In some cases, as in Mexico, titling was unfair or blatantly rigged and peasants lost their rights. Furthermore, anecdotal evidence and some quantitative analyses suggest that in many cases, farmers found arrangements to circumvent the shortcomings of traditional property rights.^[62]

Historians and agricultural experts tend to pay a lot of attention to the patterns of size and ownership of farms. They deem the concentration of land in the hands of absentee landlords (latifundia) a major hindrance to technical progress. Economists acknowledge that land concentration may negatively affect economic growth to the extent that it entrusts political power to an elite who are not interested in growth-enhancing expenditures, such as investment in education.^[63] In contrast, the evidence about the negative effect of concentration on technical progress is very weak, if existing at all. Allegedly absentee landlords did introduce innovations whenever profitable. On the other hand, it is fairly clear that large “capitalist” farms do not enjoy any advantage in cultivation (processing is an altogether different case, and it has become an industrial activity). The economies of scale are small, if any, and large farms manned by hired workers are less efficient than family farms. Family farmers have strong incentives to work hard, while preventing wage workers from shirking is much more difficult in agriculture than in manufacturing. It is possible to monitor agricultural workers effectively only while they perform simple tasks, such as harvesting.

The superiority of family-owned farms is revealed by their growing share of total land. In the nineteenth century, they prevailed in Western Settlement countries, in many countries of Europe, especially in the north, and possibly in China. In the late 1930s, according to the so-called agricultural censuses by the FAO, “farms managed by owners” (admittedly including the very few capitalist farms) accounted for about 55 percent of acreage and this figure rose to almost 80 percent at the end of the twentieth century. The coverage by country differs between censuses, but results from a pairwise comparison between identical samples are similar. Furthermore, land is, if anything, more productive in small than in large farms, and thus the share of acreage is likely to underestimate the percentage of output.

In a number of countries, the diffusion of family farms was helped by the intervention of governments, which deemed the concentration of land­ownership unjust. The first land reforms were enacted in the early 1920s by the new countries of Eastern Europe and they became a flood after the Second World War.¹¹ King lists twenty-three measures to 1975 all over the world.^{[64] [65]} However, the share of owner-operated farms rose also in countries or areas where no land reform was enacted. Of course, family farms in advanced countries in the early twenty-first century are very different from traditional ones. According to the latest American agricultural census, in 2007 the 100,000 “very large family farms” (with sales in excess of 1 million dollars and on average 1,366 acres of cropland) accounted for 50 percent of total sales. Yet, even the largest of them was tiny relative to the whole market for agricultural products.

Whenever landowners are not willing (or are not forced) to sell, the advantage of small-scale self-monitoring cultivation can be captured by hiring a tenant and his or her family, for a predetermined sum (fixed-rent tenancy) or for a share of the product (sharecropping). This latter is a very contentious institution. Historians argue that it hinders technical progress, while econo­mists, following Alfred Marshall, suspect it to be inefficient as both tenants and landlords get only half of the returns from the additional amount of factors they provide. Marshall's argument spawned a huge literature to defend the rationality of sharecropping. This literature offers a lot of good points, but no encompassing theory of the choice of contracts. The scarce quantitative evidence shows that sharecropping was less diffused than fixed rent in the nineteenth century and that its share on total tenanted land, and thus, a fortiori, on total acreage, has been declining in the twentieth century.^[66] However, without a good theory of contract choice it is difficult to explain this change. The few available tests have uncovered no evidence that share­cropping is less efficient or less innovation-fostering than fixed-rent contracts.

Agriculture is plagued by serious problems of asymmetric information between farmers and buyers of their products or lenders. The former know much more than the latter about the quality of the goods and about their own creditworthiness. Buyers can react to uncertainty by refraining from purchas­ing, while informational asymmetries in the market for credit can produce a dual market. Banks and other “formal” institutions would lend only to land­owners, who can pledge real assets, while all other farmers have to resort to “informal” sources (moneylenders, traders, etc.) and pay very high rates. Even if economically rational, this behavior would reduce output and invest­ments below their potential. Governments have tried to help - e.g. by setting up specialized credit banks for farmers - with mixed success. They invested huge sums, but only part of their money reached the farmers and the overall rate of repayment was low. However, the most popular solution to the asymmetric information problem has been the self-organization of farmers into co-operatives. The members of a production co-operative have a collec­tive interest in the quality of the product and thus they will monitor each other to avoid cheating. Similarly, a farmer can assess the credit risk of another farmer in the same village much better than can a bank clerk from the city. On the other hand, local credit co-operatives are highly vulnerable to shocks. A drought can cause the crop of all its members to fail at the same time, and no bank can stand this, unless supported from outside. This latter was provided either by the state or by regional or national associations of co-operatives.

The first co-operatives for agricultural credit were organized in Germany in the 1850s, and the first producer co-operatives in Denmark in the 1880s. In spite of some setbacks, co-operatives have been growing since then. In the mid-1990s, when world agriculture employed about 1.3 billion people, agricultural co-operatives had 180 million members - i.e., one worker out of six if membership was individual and more if the figure refers to house­holds. About 80 percent of these people were in LDCs, but advanced countries, including Japan, account for 80 percent of the total turnover. In Europe co-operatives have up to 80 to 90 percent of markets for perishable or non-homogeneous products, such as fruit and vegetables, wine, and, above all, dairy products. Some of them have built highly successful consumer brands, becoming, in some extreme cases, a threat to competition.

Markets for factors and products are essential to elicit all benefits from specialization and modern technology. Of course, markets had existed well before 1800, but in most cases their beneficial effect was reduced by poorly defined property rights and by high costs of transport and barriers to trade. The evidence on the historical development of factor markets is too sketchy to be of much use. At most, one can surmise that the diffusion of family farming reduced the percentage of full-time laborers in the total agricultural workforce, and thus possibly the depth of the market for their services. Modernization of agriculture, increasing the need for fertilizers and other capital goods, should have increased the demand for capital. The available data refer to “formal” credit and thus they are likely to overstate the growth of total lending to the extent that “formal” credit substituted for “informal” credit.

We know more about the market for commodities. In the long run, the share of marketed products in total output has undoubtedly increased a great deal, and the share of long-distance trade has increased even more, in spite of growing barriers to trade. The most compelling evidence for this hypothesis is the growth in the gross output per worker, well beyond the consumption of farming households, and in the urban population, who have to be fed via markets. In fact, the data on the share of marketed production show a steady increase in all countries. In the 1960s, it exceeded 95 percent of output in the United States and 80 percent in all advanced countries, and it hovered around 40 percent in the least developed African countries. It is likely that since then the share has increased further.

This analysis, although very sketchy, suggests that agricultural institutions were quite flexible although not perfect. In some cases, they may have hampered technical progress or reduced efficiency, and thus output relative to its potential, but on balance these negative effects seem to have been minor.

Agricultural policies

A comprehensive list of state policies which may affect agriculture directly would include:

(ι) Measures affecting ownership of factors of production (creation of property rights, land reform);

(2) Provision of public goods to farmers (R&D, infrastructures, marketing support, well-enforced property rights, etc.);

(3) Provision of public goods to the population (health regulations, repres­sion of food frauds, etc.);

(4) Transfers to farmers (subsidies, low-cost credit, etc.) or from farmers (taxation);

(5) Interventions in the domestic market of agricultural products (purchases by marketing boards, etc.) or of agricultural factors (provision of low cost credit, regulation of the agricultural labor market, etc.);

(6) Interventions in international trade of agricultural products (tariffs, taxes, quotas, etc.).

This section will concentrate on the last three items, which affect directly and immediately the farmer's income. We have already touched briefly on the two first items in this list, and can add that the earliest measures to protect consumers and also producers of premium products such as wine were adopted in the late nineteenth century, and multiplied in the second half of the twentieth century.^{[67] [68]}

Most pre-industrial polities traditionally did not intervene in agricultural markets, but there were notable exceptions. European states and cities regulated the market for staple foods in order to supply the urban population (and prevent it from revolting), while Qing China set up a network of state granaries. Between the end of the eighteenth century and the beginning of the nineteenth, the Chinese granaries were progressively shut down because of a shortage of funds, and the European regulations were phased out or abolished. At roughly the same time, many European governments, includ­ing the United Kingdom, started to protect wheat growers against foreign competition for the first time. This first wave of protection lasted for a few decades, from the late 1810s to the 1850s. After a brief spell of free trade in the 1880s, most European countries, excluding the United Kingdom, returned to protecting wheat growing against the alleged invasion of overseas grains. This is often considered an epoch-making change, but actually duties were not very high and other products were affected much less than cereals, if at all. Indeed, aggregate protection, as measured by the so-called Nominal Rate of Assistance (NRA), remained very low or even negative until the First World War.1⁵

The real epoch-making change was the outbreak of the Great Depression, which caused a fall in relative prices of agricultural goods. The European countries protected farmers by increasing duties and by adding quantitative restrictions and regulations of the markets, with the exception of the United Kingdom, which let in imports from the Empire free and compensated farmers with subsidies. On the Continent, the NRAs shot well above 50 percent, with a peak of 160 percent in Germany in 1934. Overseas producers tried to prop up their export by setting up marketing boards or subsidized their farmers for the losses (e.g. the United States with the Agricultural Adjustment Act, one of the first measures of the New Deal). State intervention was not phased out after the war. Japan main­tained the state monopoly in the rice trade (established in 1942) and in Europe the Common Agricultural Policy (formally enshrined in the Treaty of Rome of 1958 and implemented since 1962) resumed the key principle of French and German prewar policies. It set prices of agricultural products, which had to be equal (and very high) in all countries. As a result, the NRAs in the 1950s and 1960s come close to 100 percent in Japan, exceeded 50 percent in Europe, while they remained fairly low in exporting countries. The Latin American countries and the former colonies in Africa and Asia chased the dream of industrialization and used agriculture as a cash cow to finance it. They imposed heavy taxes on agriculture, set up marketing boards to control exports, and subsidized the urban consumption of food. The NRA was negative (i.e. domestic prices were lower than world market ones) in at least two-thirds of LDCs, and the average hovered around 10 percent, in spite of the presence of a small group of protectionist outliers, such as South Korea. Since the 1990s, state intervention has been slowly phased out, although not entirely. Poor countries liberalized their domestic markets and dismantled state-owned marketing boards, while advanced OECD ones switched from price setting and market intervention to directly subsidizing farmers. As a result, in the former, the NRAs declined sharply in the early years of the new millennium, although the OECD average remains high, as smaller European countries, such as Switzerland and Iceland, did not share the liberalizing zeal. In LDCs the average NRA, while still below zero in African countries, became positive in South America and especially Asia.

Economists reckon that state intervention is justified only when it can foster competition (not an issue in agriculture) or when it can redress some market failure. A case in point was arguably the low level of investment in R&D, as noted above. It seems impossible to defend policies aimed at augmenting prices of agricultural products above their world market level. They transferred income from consumers to producers, with a net loss for the consumers, which has been estimated to amount in the 1980s to about one-quarter of total transfers. Consumers did not complain too much, as food accounted for a very small share of their consumption. For the same reason, they did not rejoice that much at the liberalization of the 1990s, which reduced their losses by about two-thirds. The effects of policies on LDCs were the opposite: producers lost and urban consumers gained. The total effect for the world economy is very difficult to measure. There are some estimates of the effect of trade restrictions only, and they are quite impressive.^[69] A complete liberalization of trade in agricultural products in the early 2000s, without modifying other support policies, would have increased world GDP by about half a point, and the GDP of poor countries by about i to 1.5 percent.

Conclusions: the challenges ahead

In spite of its past achievements, world agriculture faces a difficult task. The United Nations forecasts that by 2050 world population will range between a minimum of 7.4 billion and a maximum of 10.4 billion - that is, it will be 20 to 65 percent larger than at the beginning of the twenty-first century. On top of this, the increase in income is bound to shift demand towards fruits, vegetables, and, above all, dairy and meat, which require much more land than cereals per unit of calories produced. But, unfortu­nately, available land is scarce, and it is constantly reduced by urbaniza­tion. Some pastures can be transformed into cropland, but irrigating deserts would be prohibitively expensive and a massive deforestation would cause huge environmental and social problems. Scarcity of land is not the only problem. The modern varieties of seeds tend to lose their beneficial properties after a few years, and thus they have to be replaced with new ones, with a continuous investment in research. Last, but not least, peasants and other rural residents are moving to cities in greater numbers all over the world, so that agricultural manpower is bound to decrease in the future. Thus, the only solution to the problem of production increase seems to be a further increase in capital intensity and further technical progress. Unfortunately, modern techniques, although extremely efficient, damage the environment. Irrigation causes salinization or waterlogging (an accumulation of salts or water in the soil), which might make it unsuitable for cultivation, although the real extent of the problem is still controversial. Chemical products are harm­ful to farmers and to the whole population, as they pollute the environ­ment and aquifers. The massive adoption of selected seeds and improved breeds threatens biodiversity and thus the stock of potentially useful varieties. The impact of genetically modified organisms is highly controversial.

To some extent, the needs of increasing production and of preserving the environment are in conflict. This conflict cannot be solved with a return to traditional agriculture, which, although environmentally more sustainable, would be unable to feed the current and projected world population. Developing efficient and environmentally sustainable techniques is the great challenge for the future of agriculture and of humankind.

Further reading

Anderson, Kym, Johanna Croser, Damiano Sandri, and Ernesto Valenzuela. “Agricultural distortion patterns since the 1950s: what needs explaining?” Agricultural Distortions Working Paper 90 (May 2009).

Berry, R. Albert, and William R. Cline. Agrarian Structure and Productivity in Developing Countries. Baltimore, md, and London: Johns Hopkins University Press, 1979.

Cline, W. R. Trade Policy and Global Poverty. Washington: Institute for International Economics, 2004.

De Soto, H. The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else. London and New York: Bantam Press, 2000.

Engerman, Stanley, and Ken Sokoloff. “Factor endowment, institutions and differential paths of growth among New World economies.” In Stephen Haber, ed., How Latin America Fell Behind: Essays on the Economic Histories of Brazil and Mexico. Stanford University Press, 1997, pp. 259-304.

Federico, Giovanni. Feeding the World: An Economic History of Agriculture, 1800-2000. Princeton University Press, 2005.

“The growth of world agricultural production, 1800-1938.” Research in Economic History 22 (2004), 125-181.

“The ‘real' puzzle of share-cropping: why is it disappearing?” Continuity and Change 21 (2006), 261-285.

Gardner, Bruce. American Agriculture in the Twentieth Century: How It Flourished and What It Cost. Cambridge, ma: Harvard University Press, 2002.

Hayami, Yujiro, and Vernon Ruttan. Agricultural Development: An International Perspective, 2nd edn. Baltimore, md, and London: Johns Hopkins University Press, 1985.

Jorgensen, Hans. “The interwar land reforms in Estonia, Finland and Bulgaria: a comparative study.” Scandinavian Economic History Review 54 (2006), 64-97.

King, Russell. Land Reform. London: Bell and Sons, 1977.

Kishimoto, Mio. “Property rights, land, and law in Imperial China.” In Debin Ma and Jan Luiten van Zanden, eds., Law and Long-term Economic Change: A Eurasian Perspective. Stanford University Press, 2011, pp. 68-90.

Lains, Pedro, and Vincente Pinilla. Agriculture and Economic Development in Europe since 1870. London and New York: Routledge, 2009.

Larson, Donald, Rita Butzer, Yair Mundlak, and Al Crego. “A cross-country database for sector investment and capital.” World Bank Economic Review 14 (2000), 371-391.

Marshall, Alfred. Principles of Economics, 8th edn. London: Macmillan, 1920.

Martin, Will, and Devashish Mitra. “Productivity growth and convergence in agriculture versus manufacturing.” Economic Development and Cultural Change 49 (2001), 403-423.

Nafziger, Steven. “Peasant communes and factor markets in late nineteenth-century Russia.” Explorations in Economic History 47 (2010), 381-402.

Olmstead, Alan, and Paul Rhode. Creating Abundance: Biological Innovation and American Agricultural Development. New York: Cambridge University Press, 2008.

Pardey, Philip, Julian Alston, and Roland Piggot. Agricultural R&D in the Developing World: Too Little, Too Late? New York: IFPRI, 2006.

Pomeranz, Kenneth. “Land markets in late imperial and republican China.” Continuity and Change 23 (2008), 101-150.

Simpson, James. Creating Wine: The Emergency of a World Industry, 1840-1914. Princeton University Press, 2012.

Swinnen, Johan. “The growth of agricultural protection in Europe in the 19th and 20th centuries.” World Economy 32:11 (2009), 1499-1537.

Van der Eng, Pierre. Agricultural Growth in Indonesia: Productivity Change and Policy Impact Since 1880. London and Basingstoke: Macmillan, 1996.

Van Zanden, J. L. “The first Green Revolution: the growth of production and productivity in European agriculture, 1870-1914.” Economic History Review 44 (1991), 215-239.