Chapter 5 ECD.pdf

141 5 THE ROLE OF AGRICULTURE AND SURPLUS LABOUR FOR INDUSTRIALIZATION Introduction The role of agriculture in development Barriers to agricultural development Land reform The su...

141 5 THE ROLE OF AGRICULTURE AND SURPLUS LABOUR FOR INDUSTRIALIZATION Introduction The role of agriculture in development Barriers to agricultural development Land reform The supply response of agriculture Transforming traditional agriculture The growth of the money economy Finance for traditional agriculture The interdependence of agriculture and industry Economic development with unlimited supplies of labour A model of the complementarity between agriculture and industry Rural–urban migration and urban unemployment Disguised unemployment: types and measurement Incentives and the costs of labour transfer Summary Appendix: the functioning of markets in agrarian societies Discussion questions Notes Websites on agriculture 142 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S Introduction The task of a true theory of economic growth and development must be to explain why some societies developed sooner than others, why some societies have experienced such rapid increases in living standards while others have lagged behind, and why development has not spread more evenly across the globe. The answer must be that at different stages of development, different constraints on progress operate. While some of these factors are likely to be sociological and political, the major con- straints are likely to be economic. One of the most critical factors in the early stages of develop- ment is the health of the agricultural sector, because without a surplus of food production over subsistence needs, little else can be done. There would be no surplus labour, no saving, no invest- ment and no food to feed labour working in alternative activities. It is no coincidence that the material progress of mankind started 8,000 years ago in the region of Mesopotamia (the cradle of civilization, now Iraq), where, for the first time, agriculture became settled. Unless agriculture is settled, there is no prospect of agricultural productivity increasing to provide the basis for the development of non-agricultural activities, the building of cities and the enjoyment of leisure. Where shifting agriculture is practised, as by nomadic tribes in the Kalahari Desert of Botswana and Namibia, for example, there is no basis for an agricultural surplus. As the World Bank says in its World Development Report 2008: Agriculture for Development (World Bank, 2007): agricultural growth was the precursor to the industrial revolutions that spread across the tem- perate world from England in the mid-18th century to Japan in the late-19th century. More recently, rapid agricultural growth in China, India and Vietnam was the precursor to the rise of industry... the special powers of agriculture as the basis for early growth are well established. In many developing countries today, agriculture is still extremely backward. Low productivity is a major cause of poverty and retards development of the whole economy. Over 3 billion people live in rural areas, and most of them live in households engaged in agriculture earning just a few dollars a day. The World Bank (2007) recognizes ‘that agriculture must be a prominent part of the development agenda whether for delivering growth in the agricultural-based countries or for reducing rural poverty’. It also recognizes that the state has a role to play in providing core public goods and incentives for investment in the agricultural sector. In this chapter, we consider some of the reasons for agricultural backwardness, and why prod­ uctivity is so low. We look at the process of transforming traditional agriculture and the growth of the money economy, and model the interrelationship between the growth of agriculture and industry. Then, we use Arthur Lewis’s (1954) well-known model of economic development with unlimited supplies of labour to illustrate the important role that surplus labour in agriculture (and other sectors) plays in the development process and in fuelling industrial growth. The precise meaning of surplus labour is addressed, as well as the rural–urban migration process. An appen- dix describes the various markets in agrarian societies – land, labour and credit – and how they interlock. The role of agriculture in development Agriculture makes four major contributions to the process of economic development: a product, a factor, a market and a foreign exchange contribution.  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 143 Product contribution The product contribution of agriculture refers to the fact that agriculture must supply food above subsistence needs in order to feed labour working in alternative occupations. If other sectors of the economy are to be developed, labour needs to be fed, and this cannot be done by imports until export activities have been developed to provide foreign exchange to pay for the imports. It will be remembered from Chapter 3 that in Rostow’s model of economic growth, the take- off stage of development must be preceded by an agricultural revolution. Indeed, as mentioned above in the quote from the World Bank, one of the major reasons why Britain was the first coun- try to industrialize was that it was the first to experience a significant agricultural revolution based on the abolition of serfdom and on the enclosure movement, which raised agricultural productiv- ity and provided surplus labour and food to support industrial expansion. The difference between total agricultural output and subsistence needs is called the mar- ketable surplus. Economic progress in the early stages of development requires an increase in the marketable surplus, which, in turn, requires an increase in labour productivity. If prod­ uctivity does not increase naturally or ‘voluntarily’, a marketable surplus can be forcibly extracted, as it was in Japan at the time of the Meiji Restoration in 1868, when landowners were compulsorily taxed, and more dramatically in the Soviet Union in the 1920s, when there was mass genocide of the kulaks (small prosperous landowners) during Stalin’s collectiviza- tion programme. ‘Marketable surplus’ is an important concept in the neoclassical model of the development process, because unless the marketable surplus rises as the demand for food increases, the price of food will tend to rise. This will turn the terms of trade against industry, higher wages will have to be paid to workers in industry, which will eat into profits and capital accumulation. The market- able surplus therefore becomes the major constraint on industrial growth. Factor contribution The factor contribution of agriculture consists of two parts: a labour contribution and a capi- tal contribution. Labour for industry and other activities must come from agriculture, but can be released only if productivity in agriculture rises. The existence of surplus labour (or disguised unemployment) plays a major role in the development process, as we shall see when we consider the Lewis model of economic development with unlimited supplies of labour. The lower the cost of industrial labour, the faster the rate of industrial expansion is likely to be, but this depends on the rate at which the agricultural sector is releasing labour. Industrial development today in many of the rapidly growing countries of Southeast Asia is being fuelled by cheap labour drawn from agriculture. In this respect, China’s industrial potential is enormous. Agriculture is a source of saving and capital accumulation for industrial development. The sav- ing can be voluntary or involuntary. Examples of voluntary saving are rich landlords voluntarily investing in industrial activities (the Industrial Revolution in Britain was partly financed in this way), and peasant farmers investing small savings in rural banks. Involuntary saving could take the form of the government taxing the agricultural sector and using the proceeds for investment, or, more drastically, the forced extraction of the agricultural surplus through expropriation or col- lectivization (as in Stalinist Russia). Another traditional way in which governments have taxed the agricultural sector is through the pricing policies of marketing boards, established to market agricultural produce. The prices paid to farmers are lower than the prices at which the goods are sold on the market – the differ- ence providing net revenue to the government. 144 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S The general policy in developing countries of keeping agricultural prices low used to be justified on two main grounds: that low prices benefit the industrial sector, and that peasant farmers have limited horizons and do not respond to incentives, so if prices are higher, they may actually produce less if all they are interested in is a fixed money income. This is the notion of a backward-bending supply curve of effort. It can be said without hesitation that the deliberate policy of keeping agri- cultural prices low has done enormous damage to the agricultural sector in developing countries. As we shall see later, there is ample evidence that peasant farmers do respond to price incentives. They not only increase supply in response to price rises, but also switch crops as relative prices change. Market contribution The market contribution of agriculture refers to the fact that the demand from agriculture must be the major source of autonomous demand for industrial goods. If industry is to grow and pros- per, it must be able to sell its goods. In the early stages of development, the agricultural sector is likely to provide the largest market for industrial goods. There is a complementarity between agri- cultural and industrial growth. This is well documented in the historical experience of developed countries, and in the contemporary world economy. In his classic study of Japanese economic development, Lockwood (1954) wrote: The growth of primary production was interrelated with industrialization and urbanization at every point... As industry developed, it offered a widening market for the food and raw material surpluses of the countryside... On the other hand, the increasing productivity of the primary industries created a growing home market for manufactures and services. The World Development Report 1979 (World Bank, 1979) noted that ‘a stagnant rural econ- omy with low purchasing power holds back industrial growth in many developing countries’. The World Development Report 1982 (World Bank, 1982) documented the close correspondence across countries between agricultural development and industrial growth: ‘fast growth of indus- try and sluggish agriculture were evident only in countries with oil or mineral-based economies, such as Algeria, Ecuador, Mexico, Morocco and Nigeria... These were exceptions but they prove the rule.’ In other words, a precondition for rapid industrial growth is a rapidly expanding agricul- tural sector, at least in terms of purchasing power. This has implications for the pricing of agricultural goods relative to industrial goods, or what is called the agricultural (or industrial) terms of trade. Low farm prices are good for industry from the point of view of supply potential, because this means that industry can obtain cheaper raw material inputs and wage goods, which increases profitability. On the other hand, low farm prices are bad for industry from the demand side, because this means low farm purchasing power and therefore a lower demand for industrial goods. There needs to be an equilibrium terms of trade between the two sectors to achieve balanced growth between the two sectors, so that indus- trial growth is not constrained from the supply side by agricultural prices being too high or con- strained from the demand side by agricultural prices being too low. Later in the chapter, we bring the two sectors together in an equilibrium framework and derive the equilibrium terms of trade that maximizes the growth rate of the economy as a whole. Foreign exchange contribution In the early stages of development, the only source of foreign exchange is likely to be primary commodity exports. Agriculture therefore makes an important foreign exchange contribution. Foreign exchange is a resource, just like savings. It provides access to goods that either cannot be  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 145 produced domestically or can be produced only at higher cost in an opportunity cost sense. Either way, the imports made possible by exporting agricultural products will be very ­productive – the more so if they are investment-type goods necessary for the development process. There are not many countries in the world that could not grow faster, given the greater availability of for- eign exchange. The link between trade, the balance of payments and growth is explored fully in Chapters 15 and 16. Barriers to agricultural development For the agricultural sector to supply food, release labour, provide savings, contribute to the mar- ket for industrial goods, and earn foreign exchange, it must generate a steadily rising surplus of production in excess of subsistence needs. Since land is relatively fixed in supply, this requires ris- ing agricultural productivity. The ‘grassroots’ school of economic development, which came into fashion as a reaction against the emphasis on industrialization at any cost, lays stress on policies to raise the level of productivity in agriculture as the most crucial development priority and an indispensable element of a long-run development strategy. Overall, agricultural productivity in developing countries is less than one-twentieth of the level in developed countries, and there are even bigger differences between countries. Table 5.1 gives figures on agricultural productivity in selected regions of the world in 2013 measured in US dollars at 2005 prices. Notice the huge disparities that exist, which go some way to explaining divisions in world income. In low-income countries, in which over half of the popu- lation are engaged in agriculture, value-added per head is only $310, which is less than a dollar a day. In sub-Saharan Africa and South Asia (which includes India, Bangladesh and Pakistan), prod­ uctivity is less than $2 a day. In East Asia and the Pacific region (which includes China), value- added is just over $2 day. By contrast, in the high-income countries, productivity is over $24,000. Some progress has been made in recent years with particular crops in particular countries, but Table 5.1 Agricultural productivity, 2013 Value-added per worker, US$ World 1,406 Low income 310 Middle income 1,053 Lower middle income 939 Upper middle income 1,156 Low and middle income 956 East Asia and Pacific 803 Europe and Central Asia 5,247 Latin America and Caribbean 4,125 Middle East and North Africa 3,264 South Asia 711 Sub-Saharan Africa 706 High income 24,509 Source: World Bank, 2015. 146 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S the performance of the agricultural sector is still disappointing, and the lack of marketable sur- plus still holds back development on a wide front. So what impedes agricultural productivity? There are several factors, particularly related to geography and land–labour ratios, the existence of urban bias in the treatment of agriculture and the allocation of resources, and unfair competi- tion in world markets, but the most important factors of all are the structure of rural societies, the organization of agriculture and the land tenure system that operates. As far as geographical factors are concerned, climate and terrain determine, to a large degree, what goods a country can produce, the amount of cultivatable land available per inhabitant and the land’s fertility. To some extent, the application of capital to land can compensate for un­favourable natural forces, but there are obvious limits. Mountains cannot easily be flattened or deserts readily watered. This is the concept of geographic determinism, which can be advanced as a hypothesis of underdevelopment in its own right. Having said this, however, differences in natural conditions and the fertility of the soil can be no more than a partial explanation of low productivity. Poor people are to be found along the highly productive alluvial banks of the Nile, as well as on the barren plateaus of Asia and South America. Productivity is also affected by land–labour ratios. Low labour productivity may be associated, for example, with a high population density and a high ratio of labour to land. In this case, product­ ivity might be increased substantially with small applications of capital in the form of drainage schemes, fertilizers and so on. On the other hand, low productivity may be associated with the opposite situation of a high ratio of land to labour, in which case the solution to low productivity is likely to involve much larger doses of capital for labour to work with. Most countries in Asia have high ratios of labour to land, while in Latin America and Africa, the reverse is true, as was the case in many of today’s richest countries at an equivalent stage in their economic history, for example the USA, Canada and Australia. Urban bias against agriculture takes many forms: The holding down of agricultural prices to favour the industrial/urban sector. The concentration of investment in industry. Tax incentives and subsidies to industry. Overvalued exchange rates, which keep the price of industrial inputs low, and the domestic price of agricultural exports low. Tariff and quota protection for industry, which raises the price of fertilizers, seeds and equipment. Greater spending in urban areas on education, training, housing, nutrition and medical provi- sion, which all affect productivity and the quality of life. Unfair competition consists of the subsidies that developed countries give to their farm- ers, and the tariffs that developed countries impose on imported agricultural products from developing countries. The USA and the European Union (EU) alone spend nearly $150 billion a year on farm subsidies. This has two major consequences. First, it leads to overproduction, and the surpluses are then frequently dumped on the markets of developing countries, impoverish- ing domestic farmers. Second, farmers in developing countries are not able to compete in their own markets, let alone overseas markets. The situation is made worse by developing countries being forced by international agreements to lower their tariffs against imported agricultural produce, while developed countries continue to protect their own agricultural sectors. The average global tariff on agricultural commodities is 62%. The maize growers of Mexico cannot compete with cheap maize from the USA; nor can the cotton growers of West Africa compete against subsidies of $4 billion a year given to the 20,000 cotton growers in the southern states of America. Unfair competition between developed and developing countries in the markets for  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 147 agricultural goods is one of the central issues in ongoing world trade talks under the auspices of the World Trade Organization. Geographic factors, the land–labour ratio, urban bias and competition from developed coun- tries can only explain a small part of the low productivity of agriculture in most developing coun- tries. There are more fundamental forces at work concerned with the structure of rural society, the organization of agriculture, the incentives to produce and the supply of inputs (see Binswanger and Deininger, 1997). In a typical developing country, rural society consists of rich landowners, peasants, share- croppers, tenants and labourers. Apart from the landowners, most others in the rural sector are extremely poor. Because they live on the margin of subsistence, they tend to be risk averse. In all developing countries, peasant subsistence farming is a traditional way of life, and attempts to raise productivity will alter that way of life and necessarily involve risk. As Theodore Schultz (1980) perceptively remarked in his Nobel Prize-winning lecture: Most of the people in the world are poor, so if we knew the economics of being poor we would know much of the economics that really matters. Most of the world’s poor people earn their living from agriculture so if we knew the economics of agriculture we would know much of the economics of being poor. People who are rich find it hard to understand the behaviour of poor people. Economists are no exception, for they, too, find it difficult to comprehend the preferences and scarcity constraints that determine the choices that poor people make. We all know that most of the world’s people are poor, that they earn a pittance for their labour, that half and more of their meagre income is spent on food, that they reside predominantly in low-income countries and that most of them are earning their livelihood in agriculture. What many economists fail to understand is that poor people are no less concerned about improv- ing their lot and that of their children than rich people are. Poor people on the margin of subsistence may be reluctant to make the changes necessary to improve productivity because if things go wrong it will spell disaster. But even if poor people wanted to change the traditional ways of doing things, there is the serious constraint of lack of access to credit to finance the purchase of new inputs such as seeds, fertilizers, pesticides, drain- age schemes and so on. Then there is the question of the incentive to change. Where there are tenant farmers, there is little or no security of tenure, and therefore no incentive to invest in improved methods of production. Where there is sharecropping, a certain proportion of output must be relinquished to the landowner, which also reduces the incentive to invest. Any serious programme of agrarian reform must provide greater security of tenure for farmers and give incentives to raise agricultural production, coupled with access to credit, water, fertilizers and extension services for advice. The appendix to this chapter gives a detailed description of the markets for land, labour and credit in rural societies, how they are interlocked, and the inefficiencies that arise as a result of the structure of the agricultural sector of developing countries. Land reform The system by which land is held and farmed is a serious impediment to increased productivity in many developing countries. The structure of peasant agriculture differs between countries, largely for historical reasons, but the structures have many common characteristics that keep productiv- ity low. In many countries, landholding tends to be highly concentrated. The average Gini ratio for 148 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S the concentration of landholdings in Latin America is 0.8, and in Asia, 0.4. In Latin America, 1% of landowners own roughly 70% of the land. In Brazil, 15% of landowners own 90% of the land. In many parts of Latin America, agriculture is based on a combination of large estates (latifundios), owned by a wealthy few, and small farms (minifundios), which are often so small that they cannot support a single family. When land is held and worked in the form of large estates, it is frequently underutilized and farmed inefficiently by peasants, who may have no security of tenure and may have to relinquish to the landowner a large fraction of their output. In these circumstances, there is little incentive to increase efficiency and improve productivity. In Asia, the organization of peasant agriculture is also an important determinant of productiv- ity. Because of the high population density, the major problem is that too many small farms are operated by sharecroppers and tenant farmers, the land being owned by absentee landlords. As families multiply and debts rise, land is continually sold and subdivided, leading to a very inef- ficient structure. Land reform has two aspects: first, the redistribution of land in favour of landless or near- landless households, and second, tenancy reform in favour of sharecroppers and other forms of tenant farming. Such reforms involving land rights, and security for tenants, can contribute both to an increased intensity of land use and to improved efficiency and initiative on the part of the tenant farmers, particularly if they are allowed to reap fully the rewards of their own labour. There is impressive evidence that where a change in the tenure system has permitted the pro- ducers themselves to reap the rewards of new techniques, peasant farmers have been ready to break with custom and tradition. The task of persuading producers to adopt more modern meth- ods of production and to purchase improved seeds and fertilizers has been much easier. In a study of China from 1978, Lin (1992) finds that the shift from collective to household farming led to big increases in agricultural productivity related to the acquisition of property and land rights. Likewise, in a study of India, Besley and Burgess (2000) find that rural poverty was reduced by land reform, particularly reforms that strengthened property rights over land. In Vietnam, efficiency has increased and poverty has been reduced since the end of collectivization in the 1990s. As was first discovered by Amartya Sen (1964), using Indian farm data, small farms are more productive (per hectare) than large farms. This has been shown in many other studies subse- quently. The reason is that land tends to be more fertile on small farms, and family labour tends to work the land more intensively. In other words, small farms tend to employ more labour per unit of land than large farms. Thus, land redistribution from owners of large estates to smaller family farms can raise agricultural output and employment simultaneously, helping to reduce poverty. Land reform may be a necessary condition for increased productivity, but it is clearly not a sufficient condition. It needs to be accompanied by other measures of agrarian reform. New landowners must be given access to credit, water, fertilizers and extension services for advice. Farmers need to be brought within the organized money market to improve access to credit and to reduce the role of village moneylenders, who charge exorbitant interest rates. Improved farm implements, irrigation and new social infrastructure are likely to be important. There needs to be improved dissemination of agricultural research. Too often, the agricultural extension services available are perfunctory and ineffective because the personnel are ill-trained and ill-equipped. Conditions vary from country to country, but in theory at least, agrarian reform, coupled with the application of complementary inputs, offers substantial scope for increased agricultural product­ ivity.1 See Case example 5.1 on the attempt to raise productivity in Africa, and Case example 5.2 on the approach of the World Bank.  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 149 Case example 5.1 Raising agricultural productivity in Africa African agriculture was often neglected by most governments and donors in the 1980s and 1990s. Now, however, there is a renewed commitment to agriculture centred on the Comprehensive Africa Agricultural Development Programme. There is broad agreement that there needs to be more investment in agriculture, particularly in public goods, such as rural roads, agricultural research and extension services, rural schooling, clean water and healthcare. But, often in rural Africa, there are market failures in that farmers cannot get access to credit, insurance and necessary inputs. These failures can be severe and leave small farmers in a poverty trap from which they struggle to escape even when the technology exists for them to produce more. Subsidies can help overcome poor farmers’ inability to obtain credit or take risks and to allow farmers to learn about new agricultural inputs such as new seeds and fertilizers. Subsidies can also be justified, on the grounds of equity, to overcome soil degradation and improve soil quality in the case of fertilizers, and to stimulate produc- tion to reduce the cost of food and raise the real incomes of the poor. But subsidies can be costly, with costs rising over time if not targeted properly. Where subsidies are used, they need to be ‘smart’: targeted to those who need them most, limited in time, and designed to enhance commercial distribution rather than supplant it. Complementary investment in transport and input dealer training can reinforce these programmes and make it easier to reduce or remove subsidies in the future. There are alternatives to subsidies, as Keyna’s experience of liberalized fertilizer distribution shows. Source: Africa Progress Panel, 2010. Case example 5.2 The World Bank tackles low productivity in agriculture The World Bank is committed to boosting agricultural productivity and agricultural- related investment. The world needs to produce at least 50% more food by the year 2050 to feed a projected 9 billion people. To help meet this goal, the bank is working with countries to boost the productivity of farms, livestock and fisheries. To raise yields sustainably, the bank supports ‘climate-smart’ approaches that have the potential to increase productivity, enhance resilience, promote agricultural innov­ ation through research and education, and facilitate responsible agricultural invest- ment. Gender-specific interventions are important because women account for the majority of smallholder farms – up to 70% in Africa. Improved land governance can help smallholder farmers increase the productivity of their land and improve their livelihoods. The bank supports government policies that implement systematic land surveying and titling programmes that recognize all forms of land tenure. The bank is committed to agricultural innovations that boost productivity, as well as better land and water management. It also promotes the use of new livestock breeds, bet- ter animal nutrition, improved veterinary services, vaccinations and improved husbandry to sustainably increase livestock productivity for about 1 billion farmers who depend on livestock for their livelihoods. The bank supports work on breeding and animal nutrition in India, grassland management in China, and sustainable aquaculture at coastal fisheries in Tanzania and Vietnam. These are just a few examples of the World Bank’s work. Source: World Bank, 2013. 150 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S The supply response of agriculture What may also be required is a rise in the price of agricultural products relative to industrial products in order to induce extra supply. Traditionally, attempts have been made to ‘tax’ the agricultural sector by keeping prices low in order to maintain the terms of trade in favour of the industrial sector. This policy was justified by the widespread belief that peasant producers in traditional societies would not respond to price incentives, but this assumption has proved to be wrong. Depressing the agricultural terms of trade has depressed agricultural output and caused problems for the feeding of a growing urban population. Many countries have had to introduce a positive price policy to act as a stimulus to agricultural output in general and to alter the composition of agricultural output as circumstances warrant. There is, in fact, considerable evidence that producers, especially those in close proximity to large markets with good transport facilities, respond positively to price changes, as economic theory would predict. Schultz (1964) gave early warning that ‘the doctrine that farmers in poor countries either are indifferent or respond perversely to changes in prices... is patently false and harmful. Price policies based on it always impair the efficiency of agriculture.’ When discussing the supply response of agricultural output to price, however, a distinction needs to be made between three types of response: A change in the composition of agricultural output to a change in the relative price of indi- vidual agricultural commodities. An increase in total agricultural output with respect to an improvement in the relative price of agricultural commodities compared with industrial goods. An increase in the marketed surplus in response to an increase in the price of agricultural ­commodities. Most of the studies on the supply response in peasant agriculture in developing countries relate to how producers respond to changes in the relative price of different agricultural commodities. But, of course, it would be quite possible for the supply of any individual commodity to be quite elastic with respect to price, yet the total supply of agricultural output and the marketed surplus to be quite inelastic, or even to fall, in response to a change in prices. Having said this, there are reasons for believing that the other two elasticities are likely to be positive if the supply of individual commodities is positive, especially when crops are grown not just for subsistence purposes. For example, for any crop grown commercially, the elasticity of marketed supply will be virtually equal to the output elasticity, and unless inputs are withdrawn from the production of other commodities, the elasticity of total agricultural supply will also be positive. Only in cases where peasants are content with a fixed money income, or all increased production of a commodity is consumed within the subsistence sector, will the elasticity of mar- keted supply be zero or negative at the same time as the price elasticity of supply is positive. These conditions are not likely to prevail. Empirical research on the supply response of agriculture can be divided into four main categories: 1. Cross-country studies that look at output differences in relation to price differences across countries. 2. Time-series studies that examine output movements in relation to price movements within countries over time. 3. Cross-section studies that look at output differences in relation to price differences across farms within a country.  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 151 4. Intersectoral general equilibrium models that examine how the output of agriculture varies in response to changes in the prices of agricultural goods relative to the price of other goods in the economy. The evidence shows that aggregate supply elasticities of agricultural output range from 0.3 to 0.9 (Chhibber, 1988).2 Long-run elasticity is obviously higher than short-run elasticity, and elasticity tends to be higher in the more advanced and land-abundant developing countries. The supply response of farmers to price changes depends crucially on the ability of farmers to respond to price signals, which, in turn, depends on transport, infrastructure and access to agricultural inputs. In poorer countries with inadequate infrastructure, supply elasticity is low (0.2–0.5). In fact, the sup- ply elasticity of agriculture with respect to non-price factors (e.g., the provision of public goods and services) is much higher than it is with respect to price, especially in poorer developing countries with inadequate infrastructure and marketing facilities. In a study of farm households in Ethiopia, Abrar et al. (2004) find a high supply response of different crops to changes in relative prices, but non-price factors such as access to fertilizers, land, infrastructure and marketing are often more important than prices in determining how much of which crops is produced for market. The International Monetary Fund (IMF) and the World Bank are naturally concerned with the performance of the agricultural sector in countries to which they lend under various adjustment programmes (see Chapters 14 and 16). Three interrelated issues are typically addressed: The terms of trade between agriculture and the rest of the economy. The efficiency of the agricultural sector. The supply response of agriculture to price changes. With regard to the agricultural terms of trade, the IMF normally insists that the prices paid by state marketing boards to producers be increased. Traditionally, governments have ‘taxed’ the agricultural sector through agricultural marketing boards, driving a large wedge between the prices paid to producers and the market prices of the commodities concerned. One implication, therefore, of raising producer prices is that government revenue may fall. This has implications for government expenditure if there is a budget constraint. Only if the elasticity of the supply of output with respect to producer prices is greater than unity will government revenue not fall; but as we saw above, supply elasticity is typically less than unity. To achieve efficiency within agriculture, the IMF concentrates on factors such as improving storage and transport facilities, increasing the availability of agricultural inputs, improving exten- sion services, insisting on the economic pricing of output and inputs, and privatizing marketing and extension services. We saw earlier that the supply response of farmers to price changes depends a great deal on the ability to respond, which, in turn, depends on infrastructure, transport, access to inputs and so on. Governments may be in a dilemma here because raising producer prices and reducing their own revenue may impair their ability to spend on infrastructure and other facilities. Given that the elasticity of supply with respect to non-price factors is higher than with respect to price, it would seem unwise to cut public expenditure as far as it affects the agricultural sector. Transforming traditional agriculture The task of transforming traditional agriculture is not simply a question of land reform or price policy, however. The transformation of traditional agriculture is also dependent on new inputs. The policy issue is to determine the form that the new inputs should take if agriculture is to 152 FAC TO R S I N T H E D E V E LO P M E NT PRO CE SS attract an adequate share of investment resources. New seeds are especially important to raise agricultural productivity (see Lipton and Longhurst, 1989). The way to transform traditional agriculture into a dynamic source of growth is by investment to produce a supply of new agricultural inputs that will be profitable for farmers to adopt. What is lacking is not so much an unwillingness on the part of the agricultural sector to accept new ideas, but public expenditure and the organization of particular public activities to serve the agricultural sector. Agricultural research, and investment in people to improve human capabilities in agricul- ture, has been neglected. The state of agriculture in Africa is particularly dire. Agricultural yields are low and food short- ages and undernourishment are rife. Much of the support in place for agriculture in Africa was dismantled in the 1970s by World Bank structural adjustment programmes (see Chapter 14); for example, subsidies for fertilizers and seeds, guaranteed prices for crops, and research and devel- opment – all the policies that supported Asia’s so-called Green Revolution in the 1960s, which tripled and quadrupled yields of crops such as wheat, rice and maize. ʾF
GBUIFS
PG
UIF
T
(SFFO
3FWPMVUJPO
XIJDI
CZQBTTFE
"GSJDB
XBT
Norman Borlaug, an American biologist working in Mexico, who crossed Japanese dwarf wheat with a disease-resistant local strain to produce a high-yielding hybrid, which came to be known as ‘Mexican dwarf wheat’. In countries such as India and Pakistan, yields increased threefold. Borlaug is credited with saving more lives than anyone in history – at least 1 billion in Asia alone. His motto was ‘it is impossible to build a peaceful world on empty stomachs’. He was awarded the Nobel Peace Prize in 1970 and died in 2009, aged 95. Borlaug’s breakthrough – the result of 30 years’ research – is a prime example of the impact that technology can have on the productivity of agriculture. "
TJNJMBS
CSFBLUISPVHI
PS
ALJDLTUBSU
JT
SFRVJSFE
JO
"GSJDB
5P
UIJT
FOE
BO
"MMJBODF
GPS
B
(SFFO
3FWPMVUJPO
JO
"GSJDB
"(3"
XBT
GPVOEFE
JO

XJUI
B

NJMMJPO
HSBOU
GSPN
UIF
3PDLFGFMMFS
'PVOEBUJPO
BOE
UIF
#JMM

.FMJOEB
(BUFT
'PVOEBUJPO
UP
IFMQ
SBJTF
ZJFMET
UISPVHI
JNQSPWFE
farming methods, new seeds and fertilizers, working with the African Agricultural Technology Foundation based in Nairobi, Kenya. One of the major projects is to develop ‘water-efficient’ maize to cope with long periods of drought now being experienced in southern Africa. *O
HFOFSBM
UIFSF
JT
B
OFFE
GPS
B
TFDPOE
(SFFO
3FWPMVUJPO
JO
BHSJDVMUVSF
UP
GPMMPX
UIF
mSTU
JO
the 1960s, which has now run its course. Modern science can help. Biotechnology, including genetically modified (GM) technology, has the potential to raise productivity substantially and UP
SFEVDF
UIF
JODJEFODF
PG
GBNJOF
BOE
NBMOVUSJUJPO
"
(.
DSPQ
JT
BOZ
DSPQ
WBSJFUZ
UIBU
IBT
IBE
a gene or genes from a different species or variety inserted into its genetic material using genetic engineering techniques. Currently, about 15% of the world’s farmland (approximately 180 million IFDUBSFT
JT
EFWPUFE
UP
(.
DSPQT
3FTFBSDI
JT
CFJOH
EPOF
PO
NBOZ
DSPQT
CVU
WJSUVBMMZ
BMM
QMBOUJOH
covers just five crops: soya beans, maize, cotton, rice and oilseed rape (canola). There is, however, TUSPOH
PQQPTJUJPO
UP
(.
DSPQT
GSPN
DPOTVNFST
BOE
FOWJSPONFOUBM
HSPVQT
PO
UIF
HSPVOET
PG
SJTL
UP
IVNBO
IFBMUI
CVU
(.
DSPQT
BSF
BMSFBEZ
JO
UIF
GPPE
DIBJO
CFDBVTF
UIFZ
BSF
XJEFMZ
VTFE
as processed food ingredients and for animal feed. So far, there is no scientific evidence that they are harmful. ʾF
CFOFmU
PG
(.
UFDIOPMPHZ
JT
UIBU
JU
DBO
QSPEVDF
DSPQT
UIBU
DBO
SFTJTU
QFTUT
DBO
HSPX
JO
salty soil, are drought resistant, use nitrogen more efficiently, can be stored for longer, and are more nutritious. Rice is an important staple food for poor people. Over 3 billion people in the world get at least 20% of their calorie intake from rice. The International Rice Research Institute in Manila, Philippines is at the forefront of research to produce different varieties of rice to suit particular environments where the problem may be drought, flooding, heat or salty water. With  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 153 regard to nutrition, Ingo Potrykus, working in Zurich, teamed up with Peter Beyer to genetic­ ally engineer a type of rice (‘golden rice’) to contain beta-carotene, which is the pigment that produces vitamin A. This is an important breakthrough since vitamin A deficiency kills 2 million children a year and blinds many more. This research has been funded not by biotechnology com- panies, concerned with maximizing returns by patenting and the exercise of intellectual property rights, but by the Swiss government and the Rockefeller Foundation. The plan is for growers to be given the new rice free by national research centres supervised by the International Rice Research Institute. Agricultural innovation cannot flourish without well-resourced agricultural extension services within countries. Research is now under way to cross ‘golden rice’ with a grain implanted with three genes to boost iron content to combat anaemia, which many people suffer from in developing countries. A quality protein maize has also been developed by Norman Borlaug, con- taining many important amino acids that could dramatically reduce the number of children who die of malnutrition. GM cotton has increased yields by nearly 100% in India by being more disease resistant. Some people argue that GM is the only technology that can prevent future world food crises and rising prices of basic foodstuffs. Malnutrition remains a major scourge in developing coun- tries, and by 2030, there will also be 2 billion more mouths to feed. The application of new tech- nology is urgently required. What matters most are the incentives and associated opportunities that farm people have to augment production by means of investments that include the contribution of agricul- tural research and the improvement of human skills. We emphasize again that subsistence agriculture is an uncertain activity and therefore risky, particularly when survival is at stake, and this is another factor that breeds conservatism and makes change difficult, even in the face of opportunities. Poor people prefer to be safe than sorry; they tend to prefer an inferior outcome that is relatively certain to the prospect of a higher average return with a greater degree of risk attached. They are risk averse. This is clearly not irrational behaviour for poor people living on the margin of subsistence, even if the greater risk is imagined rather than real. To overcome inertia on this score, an integral element of agrarian reform must be policies designed to minimize risk and uncertainty through the provision of various types of insurance (as discussed in Chapter 2). The growth of the money economy The question of the willingness to change customs and traditions leads naturally to a consid- eration of how peasant subsistence economies, producing goods for consumption only, typically transform themselves into money economies with an export and industrial sector. From historical experience (see Chang, 2009), two factors would appear to be crucial for the expansion of the agricultural sector and the eventual production of goods for exchange at home and abroad: 1. The expansion of communications to create outlets and markets for surplus production – and to encourage the production of the surplus itself. 2. The emergence of a class of middlemen or export–import merchants acting as agents between world markets and the domestic agricultural sectors. If these conditions prevail, purely subsistence farming can develop first into mixed agriculture, where part of the crop is retained for subsistence and part is sold in the market, and then into modern agriculture with production entirely for the market, very often based on one crop. In 154 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S the transition from subsistence agriculture, cash crops can utilize slack labour and land when the ­subsistence crops are finished; but the transition into mixed farming is possible only if the farmer has the inputs to raise productivity and the credit to purchase those inputs, as well as the market- ing facilities. Modern agriculture, run on strictly commercial lines for profit and based on one crop, must rely on exports since the size of the domestic market will generally be too small. The system of modern commercialized agriculture, upon which many developing countries depend for their export earnings, is often termed agribusiness. This is a catch-all phrase referring not only to the production of the commodity in question, but also to the backward and forward linkages associated with the production process: the provision of finance, machinery, fertilizers, seed and so on at the input end, and the processing, manufacturing and marketing of the product at the output end. Today, multinational corporations have a powerful position and a strong hold over the pro- duction and export of major agricultural commodities produced in developing countries. To give just a few examples: three US firms control over one-half of the global banana trade, five European companies control 90% of the tea sold in developed countries, and the two largest coffee com­ panies control 20% of the world market. The ability to export and the ability to market internally imply surplus production over sub- sistence needs, and it is the size of this surplus that will largely determine the speed with which the subsistence sector can be drawn into the money economy. Again, we come to the fact that unless productivity in agriculture increases, the expansion of the monetized sector will tend to decelerate as the land for cultivation dries up. When land has been exploited to the full, it acts as a constraint on development unless agricultural productivity increases or non-agricultural activ­ ities can be established. The emergence of an export sector provides a powerful stimulus to the development and extension of the money economy. Exports create the capacity to import, and the very purchase of foreign products can encourage further export specialization. A population that acquires a taste for imported goods provides the impetus to producers to export more. In the case of new goods, as well as new techniques, there is strong evidence that peasant producers respond to incentives, and are not as different from ‘Western economic man’ as is sometimes claimed. Imports also provide a stimulus to industrialization. If a market for a foreign manufactured good becomes established, it becomes easier and less risky, with the aid of tariff protection, for a domestic manu­facturer to set up in business because the market is assured. Imports can also substitute for domestic capital and raise the growth rate directly. When farmers start to specialize in goods for export, and rely on other producers for goods they previously produced themselves, the money economy will spread from the foreign trade sector to the rest of the domestic economy. This is nothing more than the international division of labour giving rise to the need for a means of exchange within a country as well as between countries. The emergence of an export sector, the spread of the money economy and the establishment of industries typically occur concurrently. What form industrialization takes will depend, in the first instance, on the initial impetus. One stimulus to industrialization that we have already mentioned is imports creating a market for goods that can be produced domestically without much difficulty. A more obvious factor leading naturally to industrialization is the availability of resources from the land, forming an indigenous industrial base. In this case, industrialization takes the form of the processing of raw materials. There are few countries that do not possess  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 155 some natural resource or other, and every country will have a comparative advantage in the production of one or other raw material that can be processed. These are the agribusinesses mentioned earlier. In many of the present developing countries, formerly under colonial rule, the initiating force behind industrialization was the foreign exploitation of resources. Industrial activity took the form of mining operations and plantation agriculture. The establishment of foreign enclave activities undoubtedly exerted a development impact, but it is sometimes argued that devel- opment would have been more rapid if countries had been left to their own devices. Some claim that the long-run development of these countries was impaired because the availabil- ity of cheap labour from the subsistence sector discouraged the installation of more mod- ern productive machinery, and also that the foreign ownership and exploitation of countries’ resources considerably reduced the potential level of investment through the remittance of profits to the host country. This is the argument of dependency theorists, which is discussed more fully in Chapter 10. Finance for traditional agriculture For many years, traditional agriculture has been starved of investment resources. While it accounts for approximately 30% of output and 50% of total employment, it attracts little more than 10% of total investment resources. Private capital has no doubt been deterred by the risks involved and the low returns in traditional agriculture. But institutional investment has also been meagre. For example, in the early years of the World Bank, 1947–59, only $124 million was spent on agriculture out of total loans of $4 billion. Official development assistance (ODA) to agriculture from multilateral and bilateral sources rose sharply in the 1970s, but since 1979, the share of ODA going to agriculture decreased from 18% to 3.5% in 2004. In absolute terms, it reached a peak of $8 billion (measured at 2004 US$) in 1984, falling to only £3.4 billion in 2004 (World Bank, 2007). Within agriculture-based developing economies, the share of public expenditure spent on agriculture has also decreased from 7% in 1980 to 4% in 2004 (World Bank, 2007). This gives some measure of the neglect of agriculture, which was partly responsible for the world food crisis and food price rises in 2007–08. The public sectors of developing countries, and multilateral institutions such as the World Bank, have a responsibility to invest in agriculture to raise productivity and combat poverty. Some projects will involve increasing the output of traditional crops through the more effective use of seeds, fertilizers and water. Other projects will involve changing the product mix from subsistence crops to the production of high-value crops. At present, the largest single component of lending to agriculture is irrigation, which permits the expansion of cultivation and makes more intensive cultivation possible by permitting dou- ble cropping. Bank-financed irrigation schemes have had a major impact on rice yields and pro- duction in Asia. The World Bank has also become the most important source of financial and technical assistance for the construction of fertilizer plants in developing countries, and these have played an important role in increasing yields and output. The World Bank gives credit for rural infrastructure projects, such as roads to reduce marketing and supply bottlenecks, and rural electrification schemes. Agricultural extension is another important aspect of the World Bank’s assistance to the rural sector. In India, where ‘contact’ farmers disseminate knowledge to their neighbours of improved techniques learnt from field agents, over 10 million farm families have 156 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S been helped. The rural poor now have more extensive and easier access to credit financed by the World Bank. In India, much of the credit has been used by small farmers to provide supplemen- tary irrigation. Finally, the World Bank operates various multipurpose projects that combine a wide range of activities, normally in conjunction with a regional development programme. In Mexico, some 75,000 low-income families have benefited from such a project in about 30 localities through investments in irrigation, soil conservation, electrification, schools, healthcare, water supplies and marketing services. Each dollar the bank invests in rural development is supplemented by local investment, and the bank rightly stresses that its contribution to the total flow of resources can be effective only if appropriate national policies are pursued on pricing, taxation, land reform and so on. The major part of the World Bank’s programme to reach the rural poor is still in the process of implementation, and is therefore difficult to assess reliably, but indications suggest that a com- bination of additional resources, institutional reforms and national government commitment to improvement in the rural sector can have a major impact. Apart from the World Bank, other multilateral institutions exist to help traditional agriculture, notably the UN’s International Fund for Agricultural Development (IFAD), which seeks to inte- grate small farmers and landless people into the development process. IFAD states that its priority is for ‘projects which will have a significant impact on improving food production in developing countries, particularly for the benefit of the poorest sections of the rural population’. Between 1978 and 2014, $15 billion had been dispersed. In the absence of external institutional investment, the sources of capital for the expansion of agriculture and industry are relatively limited in the early stages of development. In a truly subsistence economy, in the sense of an economy producing only what it needs for itself and no more, everyone is a Robinson Crusoe, supplying their own capital by refraining from present consumption. With specialization in the production of goods for export, and the producer’s need for capital to expand productive capacity, mechanisms grow up spontaneously to meet the need for credit. It is a good market maxim that demand will create a supplier at a price. The suppliers are generally village moneylenders, shopkeepers, landlords and, not infrequently, the Church – ­especially in South America – charging rates of interest that often exceed 50%. The interdependence of agriculture and industry Once agriculture emerges from its subsistence state and starts to specialize and produce goods for export, and industry develops under the impact of growth in the agricultural sector, the two sectors of agriculture and industry become interdependent. The industrial sector adds to the demand for goods produced by agriculture and absorbs surplus labour, which may raise product­ ivity in agriculture. In turn, the agricultural sector provides a market for industrial goods out of ris- ing real income, and makes a factor contribution to development through the release of resources if productivity rises faster than the demand for commodities. Adam Smith, in Wealth of Nations (1776), clearly recognized this interdependence (Kim, 2015). Demand coming from agriculture can be a major stimulus to industrialization. Adelman (1984) has described the process as ‘agricultural demand-led industrialization’. Taking 27 social account- ing matrixes for low- and high-income economies, Vogel (1994) has shown that the impact of agriculture on industry is much higher than the impact of industry on agriculture, and it increases with the level of income. At low levels of income, a $1 expenditure in agriculture generates a $2.75 increase in induced demand for non-agricultural inputs and services, and a $10 increase  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 157 in high-income countries. It is rural household demand that contributes most to the backward multiplier, which leads Vogel (1994) to conclude: ‘the early development theorists failed to articu- late a place for rural household demand for consumer goods. Not recognizing the centrality of these institutional feedbacks in agriculture’s production linkages in developing economies has been one of the great failures of theories of economic development.’ It is true that a stagnant rural sector has held back industrial development in several developing countries. (See section below, A model of the complementarity between agriculture and industry.) The transfer of resources from agriculture to industry may be in the form of capital or labour or both. Since labour is in abundant supply in most low-income countries, there is generally no difficulty in releasing labour for industry, except during harvest time. In any case, labour will tend to migrate naturally in response to seemingly better opportunities in the industrial sector and higher real incomes. The real earnings of labour in the industrial sector may be more than twice as much as the agricultural wage. If the industrial sector is to be guaranteed an adequate supply of labour, some wage differential is inevitably required to offset the higher real living costs in an urban environment, to compensate for the forfeit of non-monetary benefits of rural life, and to compensate for greater job uncertainty in the industrial sector. Real earnings may also be higher because of genuinely higher productivity in the industrial sector, where labour has more factors of production to work with. Most models of rural–urban migration make migration a positive function of the expected urban–rural wage differential, which is the difference between the urban wage, adjusted for the proportion of the total urban labour force employed (as a proxy for the probability of finding work), and the agricultural real wage (see the section below, Rural–urban migration and urban unemployment, for an outline of the model). Capital may be less ‘mobile’ than labour, and if there is considered to be insufficient lending from the agricultural sector on a voluntary basis, it may become necessary for a government to extract savings compulsorily from the agricultural sector by taxation. As mentioned already, this method was resorted to in a harsh manner by Japan at the time of the Meiji Restoration and Soviet Russia after the communist revolution. In Japan between 1880 and 1900, the land tax provided approximately 80% of central government tax revenue, and in Russia forced extrac- tion of the agricultural surplus took the form of expropriation of land and the extermination of labour. Industrialization in Western Europe, and particularly in England, was also financed to a large extent by surpluses generated on the land, but transference of these surpluses was, on the whole, voluntary through a rapidly expanding banking system. Today, developing countries, despite their access to foreign sources of capital, must also rely heavily on extracting the sur- plus from agriculture to finance industrialization. The difficulty is to decide on the best means of extraction without impairing the incentive to produce, or damaging the growth of productivity, on which a growing agricultural surplus depends. The financing of economic development will be discussed more fully in Part IV. Economic development with unlimited supplies of labour The process of the emergence of a money economy from a subsistence state was formalized by Sir Arthur Lewis in his classic paper ‘Economic Development with Unlimited Supplies of Labour’ (1954).3 There, he presented a ‘classical’ model of a dual economy with the purpose, as he described it, of seeing what can be made of the classical framework for understanding the issues of distribution, capital accumulation and growth in developing countries. His ultimate aim was to emphasize the crucial role of the capitalist surplus in the development process. 158 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S Arthur Lewis Born 1915, St Lucia, West Indies. Died 1991. Professor of Economics, Manchester University, University of the West Indies and Princeton University. Vice-Chancellor, University of the West Indies; Director of the Caribbean Development Bank. Wrote the first textbook on development economics, The Theory of Economic Growth (1955), but most famous for his 1954 paper, ‘Economic Development with Unlimited Supplies of Labour’, one of the most influential papers in development economics, still widely consulted today. One of the ‘fathers’ of development econom- ics; awarded the Nobel Prize for Economics, 1979. The Lewis model therefore starts with the assumption of a dual economy with a modern exchange (capitalist) sector and an indigenous (non-capitalist) subsistence sector, and assumes that there are unlimited supplies of labour in the subsistence sector, in the sense that the supply of labour exceeds the demand for labour at the subsistence wage; that is, the marginal product of workers in the subsistence sector is equal to, or less than, the subsistence or institutional wage. It has even been argued that the marginal product of labour may be zero or negative in an economy that is still at a fairly low level of development and experiencing a rapid growth of popu- lation. Indeed, Lewis (1954) said: ‘there are large sectors of [a developing] economy where the marginal productivity of labour is negligible, zero or even negative’. One of the distinguishing features of agriculture is that it is an activity that is subject to dimin- ishing returns owing to the fixity of the supply of land. If there is rapid population growth and labour has little employment opportunity other than on the land, a stage may be reached where the land cannot provide further workers with a living unless the existing workers drastically reduce their hours of work. These propositions are illustrated in Figure 5.1. The curve drawn represents the marginal product of successive units of labour added to the land. After the employment of X units of labour, the marginal product of labour begins to fall owing to diminishing returns; after X1 units of labour, labour’s marginal contribution to output falls below the subsistence wage; and Figure 5.1 Marginal product of successive units of labour added to the land Marginal product of labour W Subsistence wage 0 Units of labour X X1 X2 added to land  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 159 after X2 units of labour, labour’s contribution to output becomes negative and total product will decline with successive additions of labour beyond X2. In Lewis’s model, labour in excess of X1 in Figure 5.1 is in completely elastic supply to the industrial sector at whatever the industrial wage.4 The industrial or capitalist sector is repre- sented in Figure 5.2. The curve NR represents the marginal product of labour in the capitalist sector, W is the industrial wage and, on the profit-maximizing assumption, labour is employed in the capitalist sector up to the point where the marginal product is equal to the wage rate. That is, M will be employed. Workers in excess of M earn what they can in the subsistence sec- tor. The industrial wage is assumed to be determined in some relation to the wage that workers can earn in the subsistence sector. The differential (WW1) between the industrial wage and the subsistence wage will be a function of many factors, some of which were mentioned earlier, for example higher real living costs in the capitalist sector and greater job uncertainty. Given that the industrial wage is based on earnings in the subsistence sector, capitalists have a direct interest in holding down productivity in the subsistence sector, and Lewis commented that the record of every imperial power in Africa in modern times was one of impoverishing the subsistence economy. In Figure 5.2, the total product of labour, 0NPM, is split between the payment to labour in the form of wages, 0WPM, and the capitalist surplus, WNP. The expansion of the capitalist sector and the rate of absorption of labour from the subsistence sector depends on the use made of the capitalist surplus. If the surplus is reinvested, leading to greater capital formation, this will increase the total product of labour. The marginal product curve will shift upwards to the right, say N1R1, which means that if wages remain constant, the capitalist sector can now afford to employ more labour and will do so by drawing on labour from the subsistence sector to the extent of MM1 workers. The size of the capitalist surplus will increase from WNP to WN1P1, which is available for further reinvestment and so the process goes on. For Lewis, this is the essence of the development process. The stimulus to investment in the capitalist sector comes from the rate of profit, which must rise over time because all the benefits of increased productivity accrue to capital if the real wage is constant.5 According to Lewis (1954), the share of profits in the national income (P/0) will also rise. First, the share of profits in the capitalist sector (P/C) will increase, and second, the capitalist sector relative to the national income (C/0) will tend to expand; that is, if P/0 5 P/C 3 C/0, then P/0 will rise as P/C and C/0 increase. For Lewis, the latter is the more important. Lewis (1954) said: ‘if we ask why the less developed countries save so little, the answer is not because they are so poor but because their capitalist sector is so small’. Figure 5.2 Industrial/capitalist sector N1 N Marginal product P P1 W Industrial wage W1 Subsistence wage 0 Units of labour M M1 R R1 160 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S The process outlined by Lewis comes to an end when capital accumulation has caught up with population, so that there is no surplus labour in the subsistence sector left to absorb. When all surplus labour is absorbed, the supply of labour to the industrial sector becomes less than perfectly elastic. It is now in the interests of producers in the subsistence sector to compete for labour, since the marginal product of labour is no longer below the institutional wage. This is the so-called ‘Lewis turning point’. This change in producer behaviour in the subsistence sector has also been defined as the end of the take-off stage (Ranis and Fei, 1961). There is a debate in the literature over whether China has reached the Lewis turning point (Islam and Yokota, 2008). Some argue that its arrival can be seen in more assertive workers and wage rises. Others argue, however, that recent wage rises represent an upward shift in the whole labour supply curve, not a turning point, because the reserves of labour in agriculture are still vast and where labour’s productivity is less than 20% of the rest of the economy. Lewis gave several reasons why wages may rise before the turning point is reached (see below). Implicit in the Lewis model is the assumption that employment growth in the capitalist sec- tor will be proportional to the rate of capital formation. If profits are reinvested in labour-saving technology, however, this will not be so, and the rate of growth of employment in the industrial sector, as well as the rate of absorption from the agricultural sector, may be very low. It is also possible that the process of absorption may end prematurely before surplus labour in the subsistence sector is fully exhausted, owing to checks to the expansion of the capitalist surplus. Capital accumulation and labour absorption may be checked due to the expansion of the capitalist sector itself. For example, as the capitalist sector expands, the terms of trade may turn against it. If the demand for food expands faster than agricultural output, the capitalist sector will be forced to pay higher prices for food in exchange for industrial goods, reducing the size of the capitalist surplus. This will have two effects. First, if the capitalists are forced to pay higher prices for the goods they buy relative to those they sell, this means less saving for investment. The problem does not arise if productivity in agriculture is expanding rapidly, but Lewis recognized that the failure of peasant agriculture to increase its productivity has probably been the chief factor holding back the expansion of the industrial sector in many developing countries. If this is so, argue Lewis’s critics, the growth of non-farm employment can be said to depend on the growth of the agricultural surplus. This is, in fact, the starting point of neoclassical models of development (see Jorgenson, 1966), in contrast to classical models with their emphasis on surplus labour. The second effect arising from the expansion of the capitalist sector if there is a shortage of food is that the real wage may have to rise in industry, further squeezing the capitalist surplus. If labour is needed in agriculture to meet the demand for food, unlimited supplies of labour at a constant real wage may be very limited indeed. The assumption of an unlimited supply of labour is the central proposition underlying the classical approach to the theory of development, and Jorgenson has argued that the classical approach stands or falls by this hypothesis. Historically, of course, real wages have risen in agriculture and industry, and the capitalist sector has also expanded rapidly, which lends support to a middle view between the classical and neoclassical approaches. Lewis (1954) recognized the importance of both capital accumulation and food sup- ply, and it is this consideration that forms the basis of his argument for the balanced growth of the agricultural and industrial sectors. Capital accumulation in the industrial sector may also be checked for reasons unrelated to the expansion of the capitalist sector and its demand for food. For example, real wages may be forced up directly by trade unions, or indirectly through rising real wages in the subsistence sector due to increased agricultural productivity. Lewis (1954) states that:  
t
5 ) &
3 0 - &
0 '
" ( 3 * $ 6 -5 6 3 &
" / %
4 6 3 1 - 6 4
- " # 0 6 3 161 anything which raises the productivity of the subsistence sector (average product per person) will raise real wages in the capitalist sector, and will therefore reduce the capitalist surplus and the rate of capital accumulation, unless it at the same time more than correspondingly moves the terms of trade against the subsistence sector. Lewis reached this conclusion because one of the simplifying assumptions of his classical two- sector model is that the expansion of the capitalist sector is limited only by a shortage of capital, so that any increase in prices and purchasing power for farmers is not a stimulus to industrializa- tion but an obstacle to the expansion of the capitalist sector. How does this square with the idea of the agricultural sector providing a market for industrial goods, and the view of the World Bank (1979) that ‘a stagnant rural economy with low purchasing power holds back industrial growth in many developing countries’? The answer is that there does seem to be a contradiction, because the classical approach emphasizes supply to the exclusion of demand, or rather takes for granted that there will always be a market clearing price for industrial goods. In practice, there will always be a minimum below which the price of industrial goods cannot fall, set by the subsistence level wage in industry. Johnston and Mellor (1961) recognized this worrying feature of the Lewis model many years ago, when they perceptively remarked: ‘there is clearly a conflict between emphasis on agricul- ture’s essential contribution to the capital requirements for overall development, and emphasis on increased farm purchasing power as a stimulus to industrialization. Nor is there any easy rec- onciliation of the conflict.’ The challenge of reconciliation has never been taken up in a satisfactory way, but there is a resolution of the conflict if the complementarity between the two sectors is recognized from the outset, and it is remembered that there must be an equilibrium terms of trade that balances supply and demand in both sectors. The basis of a model of reconciliation is provided by Kaldor (1979). A model of the complementarity between agriculture and industry6 We have seen that agriculture provides the potential for capital accumulation in industry by pro- viding a marketable surplus. The greater the surplus, the cheaper industry can obtain food and the more saving and capital accumulation can be undertaken. This is the supply side. But industry also needs a market for its industrial goods, which, in the early stages of development, must largely come from agriculture. This is the demand side, and the higher the price of agricultural goods, the HSFBUFS
BHSJDVMUVSBM
QVSDIBTJOH
QPXFS
XJMM
CF
(JWFO
UIJT
DPOnJDU
CFUXFFO
MPX
GPPE
QSJDFT
CFJOH
good for industrial supply and high food prices being good for industrial demand, what is required is a simple model that brings together agriculture and industry in an equilibrium framework, where the terms of trade between agriculture and industry provide the equilibrating mechanism, ensuring that supply and demand grow at the same rate in each sector. Let us first model growth in the agricultural sector in relation to the terms of trade, then growth in the industrial sector, and then bring the two sectors together. Agriculture’s growth rate will be a function of how much it invests relative to output and of the productivity of investment. How much investment goods it obtains from industry in exchange for food that it ‘saves’ depends on the price of industrial goods relative to food; that is, on the terms of trade between industry and agriculture. The higher the price of investment goods, the lower the possible investment for a given amount of food and the lower the growth of supply capacity. This inverse relation between the industrial terms of trade (the price of industrial goods relative to the price of food) and the agricultural growth rate (gA) is shown in Figure 5.3. 162 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S Figure 5.3 Industrial terms of trade and agricultural growth rate Industrial terms of trade gA 0 Growth Industry’s growth rate will also be a function of its investment ratio and the productivity of investment. But there is a certain minimum to the terms of trade, below which industry would not be able to invest anything because all output would be required to pay for workers’ wage goods (food). If all wages are consumed, the cost of food input per unit of output in industry will depend on the real wage rate in industry divided by the productivity of labour, that is w/(O/L) 5 (W/O), where w is the real wage and W is the wage bill. Industrial prices must cover W/O, and this sets the lower limit to industrial prices relative to food prices. At the other extreme, industrial growth cannot exceed a certain maximum where the price of food is so low relative to indus- trial goods that all industrial goods are retained for investment in industry. The investment ratio approaches, in effect, 100%, and the upper limit to growth is given by the productivity of invest- ment. The positive relation between the industrial terms of trade and the industrial growth rate (gI) is shown in Figure 5.4. If we now assume for simplicity (although without loss of generality) that the income elasticity of demand for agricultural and industrial goods is unity, then at a given terms of trade, the rate of growth of agricultural output represents the rate of growth of demand for industrial goods, and the rate of growth of industrial output represents the rate of growth of demand for agricultural output, and where gA and gI cross, there will be balanced growth of agriculture and industry (g*) at equilibrium terms of trade (p*), as shown in Figure 5.5.7 In this model of the complementar­ ity between agriculture and industry, we can see the implications of what happens if the terms of trade are not in equilibrium, as well as the checks to the expansion of industry that Lewis mentioned. Figure 5.4 Industrial terms of trade and industrial growth rate gI Industrial terms of trade Minimum terms of trade 0 Maximum Growth growth rate  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 163 Figure 5.5 Growth equilibrium and disequilibrium Industrial terms of trade gI P1 P* P2 gA 0 g1 g2 g* Growth If the terms of trade are not in equilibrium – if the price of food is ‘too low’ or ‘too high’ in relation to industrial goods – then industrial growth is either demand constrained or supply con- strained. For example, if in Figure 5.5, the terms of trade were at P1, because the price of food was ‘too low’, industrial growth would be demand constrained to g1 by a lack of agricultural purchas- ing power over industrial goods. Industry could accumulate capital, but it could not sell its goods. Alternatively, if the terms of trade were below equilibrium at P2, industrial growth would be sup- ply constrained to g2 because the price of food would be ‘too high’, impairing capital accumula- tion in industry. Agriculture could buy, but industry could not supply. Growth is maximized at P* We can now examine what happens if there are shifts in the curves. Clearly, shifts in the curves will cause both the growth rates and the equilibrium terms of trade to vary. An improvement in agricultural productivity that shifts gA outwards will mean both higher industrial growth and an improvement in the industrial terms of trade. The importance of agricultural productivity improvement could not be better illustrated. An improvement in industrial productivity will shift gI outwards, which will also mean higher industrial growth but at the expense of worse terms of trade for industry.8 If there is a tendency for real wages in industry to rise commensurately with productivity increases, however, the gI curve will remain stable and the terms of trade will never move against industry in favour of agriculture unless agricultural productivity falls and the gA curve shifts inwards. The checks to industrial expansion in Lewis’s model are easily illustrated. A rise in the real wage in industry will shift the gI curve inwards, which will choke industrial expansion unless an equivalent increase in agricultural productivity shifts the gA curve outwards (see the earlier quote from Lewis). A final implication of the model is that if, through time, agriculture is subject to diminishing returns, productivity in agriculture will fall, shifting inwards the gA curve and reducing the rate of industrial growth. If the gI curve is relatively stable, industrial growth depends fundamentally on the rate of land-saving innovations (technical progress) in agriculture to offset the effect of diminishing returns. Rural–urban migration and urban unemployment Lewis (1954) spoke of an urban–rural wage differential of approximately 30% to attract labour to the industrial sector. What has happened in recent years, however, is that the urban–rural wage differential has widened considerably beyond this level – there has been rural–urban migration on 164 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S an unprecedented scale, but the expansion of the industrial sector has not generated sufficient employment for all those available to work. The urban–rural wage differential in China is over 300%, in Guatemala 300%, and in Vietnam 210% (World Bank, 2007). Migration has thus served to transfer unemployment from rural to urban areas, as described in Chapter 3. In 2008, for the first time in human history, the number of people living in urban areas exceeded those living in the rural sector. The informal economy of the urban sector harbours the bulk of unemployed labour in transition from the rural sector into industrial employment. The conclusion to be drawn is that the expected value of the urban wage, notwithstanding the probability of long spells of unemployment, still exceeds the wage in the rural sector, and as long as it does so, the process of migration will continue. In these changed circumstances, development theory has focused its attention in recent years on urban unemployment and policies to combat it. Most of the models of the rural–urban migration process are pessimistic about reducing the level of urban unemployment by conventional means such as subsidies to labour or public works programmes in the urban areas. The reason is that migra- tion from the land is made to be a function not only of the actual urban–rural wage differential but also of the level of employment opportunities. More employment opportunities reduce unemploy- ment immediately but encourage more migration. It thus becomes an empirical question whether increasing the rate of growth of employment in urban areas will actually reduce unemployment. New migrants may exceed the number of new jobs created. The very real possibility exists that urban areas may be caught in a ‘high level unemployment equilibrium trap’ as long as surplus labour on the land remains and development policy concentrates new activity in established urban (industrial) centres. One of the earliest and simplest models of the rural–urban migration process, which is also operational in the sense of being testable, is that of Todaro (1969, 1971). Let us consider its main features and implications. The supply of labour to the urban sector is assumed to be a function of the expected urban– rural wage differential (d), where the expected urban–rural wage differential is equal to the actual urban wage times the probability of obtaining a job in the urban sector minus the average rural wage. Thus: S 5 fs(d)(5.1) where S is the supply of labour to the urban sector and: d 5 wπ 2 r(5.2) where w is the urban real wage, r is the average rural wage, and π is the probability of obtaining a job in the urban sector. The probability of obtaining a job in the urban sector is assumed to be directly related to the rate of new job creation and inversely related to the ratio of unemployed jobseekers to the num- ber of existing job opportunities,9 that is: gN gN p5 5  (5.3) W2N U where g is the net rate of new urban job creation, N is the level of urban employment, W is the total urban labour force,10 and U is the level of urban unemployment. Substituting equation (5.3) into equation (5.2) gives: wgN d5 2 r(5.4) U  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 165 If it is assumed that migration will come to a stop when the expected urban wage equals the rural wage (that is, when d 5 0), we can derive from equation (5.4) the equilibrium level of unemploy- ment as: wgN Ue 5 (5.5) r It can be seen from equation (5.5) that a reduction in the actual urban wage will reduce the equi- librium level of unemployment, and a rise in the rural wage will also reduce it, but (paradoxically) an increase in the rate of new job creation will raise the equilibrium level of unemployment by increasing the probability of obtaining a job and encouraging migration. Whether policies such as wage subsidies can reduce unemployment therefore depends on whether the increase in the demand for labour as a result is greater or less than the induced supply. From equation (5.5), we can solve for the equilibrium ratio of unemployment to employment and give some quantitative content to the model. Dividing both sides by N gives Ue/N 5 wg/r. Thus, for example, if the industrial wage is twice as high as the rural wage (w/r 5 2), and g 5 0.05, the equilibrium ratio of unemployment to employment will be 10%. To consider the policy implications more fully, and to answer the question: Under what condi- tions will the actual level of urban unemployment rise?, let us suppose that the rate of urban job creation is a function of the urban wage, w, and a policy parameter, a (e.g., a government policy variable to increase employment). Thus: 'g g 5 fd(w,a). 0(5.6) 'a If the growth of urban labour demand is increased, the response of labour supply can be written as: 'S 'S 'd 'g 5 (5.7) 'a 'd 'g 'a Now, from equation (5.4) by partial differentiation, we have: 'd N 5 w (5.8) 'g U Substituting equation (5.8) into equation (5.7) gives: 'S 'S wN 'g 5 (5.9) 'a 'd U 'a There will be an increase in the absolute level of urban unemployment if the increase in supply in response to a policy change exceeds the increase in the absolute number of new jobs created, that is, if: 'S wN 'g 'g. N (5.10) 'd U 'a 'a 166 F A C T O R S I N T H E ­D E V E L O P M E N T P R O C E S S Now, cancelling N and ∂g/∂a from both sides and multiplying both sides by d/w and U/W, the condition for unemployment to increase becomes: 'S/W d U.  (5.11) 'd/d wW or substituting equation (5.2) into equation (5.11): 'S/W wp 2 r U. (5.12) 'd/d w W In words, equation (5.12) says that unemployment will increase in the urban sector as a result of a policy change to increase employment if the elasticity of the urban labour supply (by migra- tion) with respect to the urban–rural wage differential exceeds the expected urban–rural wage differential as a proportion of the urban wage times the unemployment rate. Equation (5.12) is clearly testable. It transpires, in fact, that equation (5.12) is satisfied with a very low elasticity. For example, suppose that the actual urban wage is twice the rural wage,11 that the probability of obtaining a job in the urban sector is 0.8 and that the unemployment rate is 10%, then the level of unemployment will increase if the elasticity of the urban labour supply with respect to the expected urban–rural wage differential is 0.03. Note that the growth of total labour supply as a result of migration (∂S/W) is not the same thing as the rate of growth of migration (∂S/S), so that the elasticity of supply with respect to a change in job opportunities is not the same as the elasticity of migration with respect to a change in job opportunities. We could, however, convert equation (5.12) into the elasticity of migration with respect to ∂d/∂ by multiplying both sides of equation (5.10) by U/S instead of U/W. This would give: 'S/S wp 2 r U. (5.13) 'd/d w S Since the ratio of unemployment to migration (U/S) is much higher than U/W, the elasticity of migration itself would have to be higher than the elasticity of labour supply for unemployment to increase following a job expansion programme. If, as before, we assume that w/r 5 2, π 5 0.8 and, say, U/S 5 2, the migration elasticity would have to exceed 0.6 for unemployment to rise. In principle, this elasticity is easy to estimate by specifying a migration function in which migration is a function of the expected urban–rural wage differential, holding constant other factors affecting migration. For an interesting case study on Tanzania, see Barnum and Sabot (1977), who estimate an elasticity of migration with respect to the urban wage itself, holding other things constant, of between 0.7 and 2.0.12 Disguised unemployment: types and measurement We must now examine more critically the classical assumption of unlimited supplies of labour, defined as labour’s marginal product below the subsistence wage. If the marginal product of labour in the rural sector is positive (which is not precluded in Lewis’s model as long as it is below the subsistence wage), the withdrawal of labour from the subsistence  5 T H E R O L E O F A G R I C U LT U R E A N D S U R P L U S L A B O U R 167 sector will reduce total output. To argue that development via unlimited supplies of labour is feas­ ible and relatively painless, one must implicitly assume that the marginal product of labour is virtu- ally zero. The term disguised unemployment is usually defined loosely in this way. But the question arises of how workers can survive on the land if their marginal product is zero, or even positive but below subsistence. Who would employ such labour? Would output in the subsistence sector really remain unaffected if substantial quantitie