Agribusiness Management · Agri-sector specialisation · also AEC-515
What development means, how it is measured, the classic growth models, and the roles of state, market, civil society and international institutions.
In the exams: agricultural economics PhD entrance and fellowship examinations, development (heavy); central bank officer recruitment economic and social issues papers (reported)
Development economics studies how poor economies become richer and how the lives of their people improve: what development means, how it is measured, why some countries have grown fast while others have stalled, and what states, markets, civil society and international institutions can do about it.
For an agribusiness or agricultural economics professional the subject matters because most of the world's poor still live in rural areas and depend on farming, directly or indirectly, and because growth originating in agriculture has usually done more for the poor than growth elsewhere.
This manual covers the core examined worldwide: meanings and measures of development, classical and post-war theories, the Harrod-Domar, Solow and endogenous growth models, structural change, population and human capital, comparative experience and trade strategy, the state, markets, civil society and institutions, and international institutions and aid. The detailed agricultural theories (Lewis and Fei-Ranis dualism, Schultz, Mellor, induced innovation) and the design of farm price, land, credit and food policy belong to the separate subject Agricultural Development and Policy Analysis and are only referred to here.
In India the subject is examined heavily in the ICAR AICE-JRF/SRF Agricultural Economics paper (economic and agricultural development), in ASRB NET, and in the economic and social issues papers of NABARD and RBI.
Until the late 1960s development was usually equated with sustained growth of national income per head, together with the structural change that accompanies it: a falling share of agriculture in output and employment and a rising share of industry and services. Several countries grew in the 1960s while unemployment, poverty and inequality did not fall.
Dudley Seers (1969, "The Meaning of Development") asked three questions of any country: what has been happening to poverty, to unemployment and to inequality? If all three had fallen, there had been development; if one or two had worsened, it would be strange to call the result development even if income per head had doubled. The ILO's basic needs approach of the mid-1970s followed.
Amartya Sen (Commodities and Capabilities, 1985; Development as Freedom, 1999) gave the most influential reformulation. Commodities matter only for what they allow people to do and be. Functionings are the beings and doings a person values (being nourished, literate, part of community life); capabilities are the set of functionings a person can actually achieve, their real freedom to choose a life they value. Income is a means; capability is the end. Kerala in India and Sri Lanka, with high life expectancy and literacy at modest incomes achieved through public action, are the standard illustrations.
Denis Goulet (The Cruel Choice, 1971), in a framework popularised by Michael Todaro's textbook, named three core values of development: sustenance, the ability to meet basic needs; self-esteem, a sense of worth and of not being used as a tool by others; and freedom from servitude, the ability to choose.
Sustainable development entered the mainstream with the report of the World Commission on Environment and Development (the Brundtland Commission, Our Common Future, 1987), which defined it as development that meets the needs of the present without compromising the ability of future generations to meet their own needs. For agriculture, soil, groundwater and fisheries are capital stocks whose depletion is not development.
The eight Millennium Development Goals (2000 to 2015) began with halving extreme poverty and hunger. The seventeen Sustainable Development Goals, adopted by the United Nations in 2015 for 2030, are broader and apply to all countries; Goal 1 is no poverty and Goal 2 is zero hunger, which explicitly includes doubling the agricultural productivity and incomes of small-scale food producers and ensuring sustainable food production systems.
Economic growth is a sustained increase in real output, usually measured as the annual percentage change in real gross domestic product (GDP) or, for living standards, in real GDP or gross national income (GNI) per head. Development is broader: it includes growth but also changes in the structure of production and employment, in poverty and inequality, in health and education, and in institutions and freedoms. Growth is usually necessary for development in poor countries but not sufficient.
Per-capita growth is approximately total growth minus population growth. Exactly, (1 + g per head) = (1 + g of GDP) / (1 + n). A country whose GDP grows 7 per cent while population grows 2 per cent has per-capita growth of 1.07/1.02 - 1 = 4.9 per cent, close to the approximation of 5 per cent. Small differences in growth compound into large differences in levels. The rule of 70 gives the doubling time: about 70 divided by the percentage growth rate, so income doubles in about 35 years at 2 per cent a year and in about 10 years at 7 per cent.
Growth can occur without development when a few capture its gains, when it rests on a mineral enclave with few links to the rest of the economy, when it depletes natural capital, or when it creates few jobs.
Simon Kuznets (1955) suggested that inequality first rises and then falls as income per head rises, an inverted U, because labour moves from a low-income, low-inequality traditional sector into a higher-income, more unequal modern sector. Later evidence gives only weak support for a systematic Kuznets curve; East Asian economies such as Taiwan and South Korea grew fast with relatively low inequality, which is often attributed partly to early land reform and broad-based rural development.
Pro-poor growth is growth that reduces poverty substantially. The growth incidence curve (Ravallion and Chen, 2003) plots income growth at each percentile. The growth elasticity of poverty is the percentage change in the poverty headcount for a one per cent rise in mean income; it is larger in absolute value where initial inequality is lower.
Where growth originates matters. The World Bank's World Development Report 2008, Agriculture for Development, concluded that GDP growth originating in agriculture is at least twice as effective in reducing poverty as growth originating in other sectors, because the poor are concentrated in farming and rural labour markets.
Agriculture illustrates the growth-development distinction well. Its share of GDP falls as a country develops, so a declining share is a sign of development, not of failure; but a falling share of output with an unchanged share of employment means the gap between farm and non-farm labour productivity is widening, and rural incomes are falling behind. Suppose agriculture provides 18 per cent of GDP but employs 45 per cent of workers. Relative labour productivity in agriculture is 18/45 = 0.40 of the economy average, while in non-agriculture it is 82/55 = 1.49; a non-farm worker produces about 3.7 times as much as a farm worker. Such gaps, common in South Asia and sub-Saharan Africa, show why structural change of employment, not only of output, is central to development.
Agricultural growth also affects food prices and the real incomes of poor net buyers of food, including urban workers and landless labourers. Rising demand for livestock products, fruit, vegetables and processed food means agribusiness grows even as farming's share shrinks.
Gross domestic product measures the value of final output produced within a country; gross national income adds net primary income from abroad and is the World Bank's preferred basis for classifying countries. Per-capita GNI divides by population. The World Bank classifies economies each year as low-income, lower-middle-income, upper-middle-income or high-income using GNI per capita converted by the Atlas method, which smooths exchange rates with a three-year average adjusted for inflation differences. The thresholds are revised every year.
For comparisons of living standards and poverty, PPP conversion is used instead. A PPP conversion factor is the number of units of a country's currency needed to buy the same basket that one US dollar buys in the United States. If a household earns 150,000 units of local currency a year, the market exchange rate is 80 per dollar and the PPP factor is 20 per dollar, its income is 1,875 dollars at market rates but 7,500 PPP dollars, four times more, because local prices of non-traded goods are low.
Income measures omit non-market production, including subsistence farm output that is not well recorded and household work, mostly done by women; they ignore distribution; count defensive spending as output; ignore depletion of natural capital; and say nothing directly about health, education or freedom.
The United Nations Development Programme's Human Development Index (HDI) was introduced in the first Human Development Report in 1990, led by Mahbub ul Haq with Amartya Sen's conceptual input. It combines three dimensions: a long and healthy life (life expectancy at birth), knowledge (mean years of schooling of adults aged 25 and over, and expected years of schooling for children of school-entry age), and a decent standard of living (GNI per capita in PPP dollars).
Each dimension index is (actual - minimum) / (maximum - minimum) using fixed goalposts. In the method used since the 2010 report, with goalposts set in 2014, life expectancy runs from 20 to 85 years, mean years of schooling from 0 to 15, expected years from 0 to 18, and GNI per capita from 100 to 75,000 PPP dollars. The income index uses natural logarithms, reflecting diminishing returns of income to capabilities: (ln GNI - ln 100) / (ln 75,000 - ln 100). The education index is the arithmetic mean of the two schooling indices. Since 2010 the HDI is the geometric mean of the three dimension indices, so poor performance in one dimension cannot be fully offset by high performance in another; before 2010 an arithmetic mean was used. Worked example: life expectancy 70 gives (70 - 20)/65 = 0.769; mean schooling 8 and expected schooling 12 give (8/15 + 12/18)/2 = 0.600; GNI of 10,000 gives (ln 10,000 - ln 100)/(ln 75,000 - ln 100) = 0.696; HDI = (0.769 x 0.600 x 0.696) to the power 1/3 = 0.685.
Related indices: the inequality-adjusted HDI (IHDI, from 2010) discounts each dimension by its inequality; the Gender Development Index compares female and male HDI; the Gender Inequality Index (from 2010) covers reproductive health, empowerment and labour-market participation. An earlier composite, Morris D. Morris's Physical Quality of Life Index (1979), averaged infant mortality, life expectancy at age one and basic literacy with equal weights and left out income altogether.
The global Multidimensional Poverty Index (MPI), developed by the Oxford Poverty and Human Development Initiative with UNDP and first published in 2010, replaced the earlier Human Poverty Index. It uses the Alkire-Foster counting method on household survey data. There are three equally weighted dimensions and ten indicators. Health (weight 1/3) has nutrition and child mortality, each 1/6. Education (1/3) has years of schooling and school attendance, each 1/6. Living standards (1/3) has six indicators, each 1/18: cooking fuel, sanitation, drinking water, electricity, housing and assets.
Each person receives a deprivation score, the weighted sum of the indicators in which the household is deprived. A person is multidimensionally poor if the score is at least one third; a score of one half or more marks severe multidimensional poverty, and a score between one fifth and one third marks vulnerability. A household deprived in nutrition (1/6) and in cooking fuel, sanitation and housing (3 x 1/18 = 1/6) has a score of exactly 1/3 and is counted as poor.
The MPI is the product of two numbers: the incidence H, the share of people who are multidimensionally poor, and the intensity A, the average deprivation score among the poor. If 30 per cent of people are poor and on average they are deprived in 45 per cent of weighted indicators, MPI = 0.30 x 0.45 = 0.135. National MPIs, such as NITI Aayog's for India, use different indicators from the global version.
Income or consumption poverty is measured against a poverty line z. An absolute line is the cost of a fixed basket of basic needs; a relative line is a fraction of mean or median income (the European Union uses 60 per cent of the median). The World Bank's international extreme poverty line has been revised with each new round of PPP data, from 1.90 dollars a day at 2011 PPPs to 2.15 dollars at 2017 PPPs and, from June 2025, 3.00 dollars at 2021 PPPs; national lines differ and are usually preferred for domestic policy.
The Foster-Greer-Thorbecke (1984) family is P-alpha = (1/N) sum over the poor of ((z - y_i)/z) to the power alpha. With alpha = 0 it is the headcount ratio H, the share of the population below the line. With alpha = 1 it is the poverty gap index, the average shortfall of the poor from the line expressed as a share of the line and averaged over the whole population (non-poor count as zero). With alpha = 2 it is the squared poverty gap, which weights larger shortfalls more and so measures severity.
Worked example with z = 100 and ten incomes 40, 60, 80, 90, 110, 120, 150, 200, 250 and 300. Four people are poor, so H = 0.40. Their normalised gaps are 0.6, 0.4, 0.2 and 0.1, summing to 1.3, so the poverty gap index is 1.3/10 = 0.13. The squared gaps are 0.36, 0.16, 0.04 and 0.01, so P2 = 0.57/10 = 0.057. The minimum cost of eliminating poverty with perfect targeting is PG x z x N = 0.13 x 100 x 10 = 130. The income gap ratio, the average gap among the poor only, is PG/H = 0.325.
Amartya Sen (1976) noted that the headcount ignores how far below the line the poor are, and can fall when money moves from a very poor person to someone just below the line, lifting that person out, although the poorest is worse off. Sen proposed two axioms. The monotonicity axiom says that, other things equal, a fall in a poor person's income must raise the poverty measure. The transfer axiom says that a transfer from a poor person to a richer person must raise the measure. The headcount violates both; the poverty gap satisfies monotonicity but not the transfer axiom, since a transfer between two poor people leaves the total gap unchanged; the squared poverty gap satisfies both. The Sen index and the Watts index also satisfy both.
A further useful property is subgroup decomposability: overall poverty is the population-weighted sum of poverty in subgroups. The FGT measures are additively decomposable. Such decompositions usually show that landless agricultural labourers and marginal farmers account for most of the poor.
The Lorenz curve plots the cumulative share of income received against the cumulative share of population, ranked from poorest to richest. Perfect equality gives the 45-degree line; the further the curve sags below it, the more unequal the distribution. The Gini coefficient is the area between the line of equality and the Lorenz curve (A) divided by the whole triangle below the line of equality (A + B). Since A + B = 1/2 on a unit square, Gini = 1 - 2B, where B is the area under the Lorenz curve; it ranges from 0 (equality) to 1 (one person has everything).
Worked example: quintile shares of 5, 10, 15, 25 and 45 per cent give cumulative shares of 0.05, 0.15, 0.30, 0.55 and 1.00. With the trapezoidal rule, B = 0.2 x [(0 + 0.05) + (0.05 + 0.15) + (0.15 + 0.30) + (0.30 + 0.55) + (0.55 + 1.00)]/2 = 0.31, so Gini = 1 - 0.62 = 0.38. If the poorest half of the population earns 20 per cent of income, the two-segment Lorenz curve gives B = 0.05 + 0.30 = 0.35 and Gini = 0.30.
Other measures are also used. The Kuznets ratio is the top 20 per cent's share over the bottom 40 per cent's; the Palma ratio is the top 10 per cent's share over the bottom 40 per cent's. The Theil index, from information theory, is additively decomposable into within-group and between-group inequality, which the Gini is not (unless subgroups do not overlap in income).
Good inequality measures satisfy anonymity, scale independence (doubling all incomes does not change it), population independence and the Pigou-Dalton transfer principle (a transfer from richer to poorer reduces measured inequality). The Gini satisfies all four. When Lorenz curves cross, rankings are ambiguous. Land Ginis, from the distribution of farm area, are common in agricultural economics.
Adam Smith (The Wealth of Nations, 1776) located growth in the division of labour, which raises productivity and is limited by the extent of the market; in capital accumulation, which finances more specialised production; and in a system of natural liberty with secure property and low barriers to trade.
David Ricardo (Principles of Political Economy and Taxation, 1817) built growth around land. As population grows, cultivation extends to poorer land or is intensified on existing land, so the marginal product of labour on land falls. Rent, the differential return to better land, rises; food prices and hence the subsistence wage in money terms rise; profits, the source of accumulation, are squeezed until accumulation stops at a stationary state. Hence his case for repealing the Corn Laws. Diminishing returns in agriculture are thus the brake on growth in the classical model; technical change in farming, which Ricardo underrated, later released that brake.
Thomas Robert Malthus (An Essay on the Principle of Population, 1798) argued that population tends to grow geometrically while food supply grows at best arithmetically, so incomes are pushed back to subsistence by positive checks (famine, disease) unless preventive checks (later marriage) intervene. Technical change and the demographic transition broke the trap.
Karl Marx saw capitalism as accumulating capital and raising productivity while concentrating ownership and generating crises. Joseph Schumpeter (The Theory of Economic Development, 1911) placed the innovating entrepreneur, and new combinations financed by bank credit, at the heart of development.
Walt Rostow (The Stages of Economic Growth, 1960) proposed five stages: the traditional society; the preconditions for take-off; the take-off, in which the rate of productive investment rises from about 5 per cent or less to over 10 per cent of national income and one or more leading manufacturing sectors grow rapidly; the drive to maturity; and the age of high mass consumption. Rising farm productivity was a precondition.
Paul Rosenstein-Rodan (1943) argued for a big push: because of indivisibilities in social overhead capital, complementarity of demand and indivisibility in the supply of savings, many industries must be started together. Ragnar Nurkse (1953) described the vicious circle of poverty: low income means low saving and low demand, hence low investment and low productivity, hence low income again; he advocated balanced growth across many sectors, including agriculture, to create mutual markets.
Albert Hirschman (The Strategy of Economic Development, 1958) replied that poor countries lack the resources for balanced growth; development should proceed through a chain of disequilibria, investing first in key sectors with strong backward linkages (demand for inputs) and forward linkages (supply of outputs to other sectors). He judged agriculture's linkages weak, a view later challenged by evidence on rural consumption linkages. Harvey Leibenstein (1957) argued for a critical minimum effort large enough to overcome income-depressing forces; Richard Nelson (1956) modelled a low-level equilibrium trap in which population growth swallows small income gains. Gunnar Myrdal (1957) described circular and cumulative causation, in which growing regions draw labour and capital from lagging ones (backwash effects) more than they stimulate them (spread effects).
Dualism is the coexistence in one economy of two sectors with different technology, organisation and productivity. J.H. Boeke described social dualism in colonial Indonesia, arguing that the traditional sector responded to different motives from the capitalist sector. Benjamin Higgins described technological dualism: capital-intensive modern industry with fixed factor proportions beside labour-intensive traditional agriculture. W. Arthur Lewis (1954) formalised the labour-surplus dual economy, in which the modern sector draws labour from agriculture at a constant wage until the surplus is exhausted; that model and its extensions by Fei and Ranis, Jorgenson and Harris-Todaro are treated fully in Agricultural Development and Policy Analysis.
Structural change is the reallocation of output and employment across sectors. Colin Clark and Allan Fisher observed a shift from primary to secondary to tertiary activities as income rises; Hollis Chenery and Moshe Syrquin (Patterns of Development, 1975) estimated typical cross-country patterns: as income per head rises, the shares of agriculture in output and employment fall, industry's share rises and then levels off, investment and saving rates rise, and urbanisation and the demographic transition proceed. Engel's law is one driver.
Latin American structuralists, notably Raul Prebisch and the United Nations Economic Commission for Latin America, argued that developing economies have rigidities (inelastic food supply, concentrated land, foreign-exchange shortages) that markets cannot quickly remove, which justified planning and import substitution. Dependency theorists (Andre Gunder Frank, Theotonio dos Santos) argued that the periphery is kept underdeveloped by its relationship with the core.
Roy Harrod (1939) and Evsey Domar (1946) independently built a growth model from Keynesian foundations. Saving is a fixed share s of income: S = sY. Investment adds to the capital stock: the change in K equals I (ignoring depreciation). Output is proportional to capital through a fixed capital-output ratio v = K/Y, so that the change in K equals v times the change in Y; v at the margin is the incremental capital-output ratio (ICOR). Setting saving equal to investment gives sY = v times the change in Y, so the growth rate of output is g = s/v. With depreciation at rate delta, g = s/v - delta.
With s = 24 per cent and v = 4, output grows at 6 per cent a year. To reach a target growth rate, the required saving rate is g times v: an 8 per cent target with an ICOR of 4 needs investment of 32 per cent of GDP. With s = 25 per cent, v = 5 and depreciation of 2 per cent, g = 5 - 2 = 3 per cent. Per-capita growth is approximately s/v - n, where n is population growth. The ICOR can be estimated from data as investment share divided by growth rate; an economy investing 30 per cent of GDP and growing at 6 per cent has an implied ICOR of 5, and one that invests 200 in a year and raises output by 50 has an ICOR of 4.
Harrod distinguished three growth rates. The actual rate g = s/v_actual, where v is the realised capital-output ratio. The warranted rate g_w = s/v_r, where v_r is the capital-output ratio that firms desire; at g_w, firms are satisfied with the investment they have made and wish to continue at the same rate. The natural rate g_n is the maximum sustainable rate allowed by growth of the labour force and labour-augmenting technical progress.
Two problems follow. The first is instability, the knife edge. If actual growth exceeds warranted growth, capital looks too small, firms invest more and demand rises further; if it falls short, investment and growth fall further. Deviations are self-reinforcing. The second is the coincidence problem: because s, v_r and n are determined independently, there is no reason why g_w should equal g_n. If g_w exceeds g_n (for example s = 15 per cent, v_r = 3 gives g_w = 5 per cent while the labour force grows at 3 per cent), the economy cannot sustain its warranted rate; capital outruns labour and there is a tendency towards stagnation and chronic excess capacity. If g_w is below g_n, capital grows too slowly to employ the growing labour force, and unemployment rises, which Harrod thought typical of developing countries with high population growth.
The Harrod-Domar framework is still used for quick planning arithmetic and in the two-gap model of aid (Chenery and Strout, 1966), in which the investment needed for a target growth rate is compared with domestic saving to find the saving gap, and required imports are compared with export earnings to find the foreign-exchange gap. With a 6 per cent target and an ICOR of 4, investment must be 24 per cent of GDP; if domestic saving is 18 per cent, the saving gap is 6 per cent of GDP, to be filled by foreign capital or aid.
Its limitations are serious. Capital and labour cannot substitute for each other because v is fixed, so the model cannot explain how economies absorb labour growth. It has no role for technical progress except through an exogenous change in v. The ICOR is not a technological constant but an outcome of policy, the sectoral mix and capacity utilisation; it varies widely across countries and periods. William Easterly (1999) showed that the predicted link between aid-financed investment and growth fails in cross-country data.
In agriculture, capital is largely land improvement, irrigation, herds and roads with long gestation and weather-dependent returns, which a single economy-wide ICOR hides.
Robert Solow (1956) and Trevor Swan (1956) replaced the fixed capital-output ratio with a neoclassical production function Y = F(K, AL) that has constant returns to scale, diminishing returns to each input, and smooth substitution between capital and labour. Per effective worker, y = f(k) with k = K/AL. With a Cobb-Douglas function, y = k to the power alpha, where alpha (between 0 and 1) is capital's share of income.
Saving is a constant fraction s of output and all saving is invested. Labour grows at rate n, labour-augmenting technology A at rate g, and capital depreciates at rate delta. The fundamental equation of the model is: the change in k = s f(k) - (n + g + delta) k. The first term is actual investment per effective worker; the second is break-even investment, the amount needed to keep k constant as capital wears out and is spread over more effective workers. Where the two are equal the economy is in steady state, k*.
For Cobb-Douglas, k* = (s / (n + g + delta)) to the power 1/(1 - alpha) and y* = (s / (n + g + delta)) to the power alpha/(1 - alpha). Example without technical progress: y = k^0.5, s = 0.2, n = 0.02, delta = 0.03. Then k* = (0.2/0.05)^2 = 16, y* = 4, consumption per worker = 0.8 x 4 = 3.2, and investment sy* = 0.8 equals break-even (n + delta)k* = 0.8. With alpha = 1/3, s = 0.25 and n + g + delta = 0.05, k* = 5^1.5 = 11.18 and y* = 5^0.5 = 2.24.
A higher saving rate raises the steady-state levels of capital and output per effective worker but not the long-run growth rate. In steady state, output per worker grows at g, the rate of technical progress, whatever the saving rate; a rise in saving produces a temporary burst of faster growth while the economy moves to the higher level. With alpha = 1/3, the elasticity of y* with respect to s is alpha/(1 - alpha) = 0.5, so doubling the saving rate raises steady-state output per worker by a factor of 2^0.5, about 1.41, not 2.
The golden-rule saving rate (Edmund Phelps, 1961) maximises steady-state consumption per effective worker. It is where the marginal product of capital equals n + g + delta; with Cobb-Douglas this gives s = alpha. Saving above it is dynamically inefficient.
Absolute convergence is the prediction that poor countries grow faster than rich ones unconditionally; it holds among similar economies (US states, OECD members since 1950) but not across the world as a whole. Conditional convergence is the Solow prediction properly stated: each economy converges to its own steady state, so poorer countries grow faster only after controlling for saving, population growth and technology (Barro, 1991; Mankiw, Romer and Weil, 1992). The speed of convergence in the Cobb-Douglas case is about (1 - alpha)(n + g + delta); with alpha = 1/3 and n + g + delta = 0.06, it is 4 per cent a year, giving a half-life of ln 2 / 0.04, about 17 years.
Solow (1957) used the production function to decompose growth: growth of output = alpha times growth of capital + (1 - alpha) times growth of labour + growth of total factor productivity (TFP). TFP growth is calculated as a residual, the Solow residual. If output grows 6 per cent, capital 8 per cent and labour 2 per cent with alpha = 0.4, TFP growth is 6 - 3.2 - 1.2 = 1.6 per cent. Solow found that most of the growth of output per worker in the United States in the first half of the twentieth century came from the residual, not capital deepening. Alwyn Young's 1990s studies argued that East Asian growth came mostly from factor accumulation rather than TFP, a contested finding.
The basic Solow model predicts differences in income per head smaller than those observed, given realistic capital shares. N. Gregory Mankiw, David Romer and David Weil (1992) added human capital as a separate accumulated input: Y = K^alpha H^beta (AL)^(1 - alpha - beta). With alpha and beta around one third each, the elasticity of steady-state income with respect to the physical investment rate rises to alpha/(1 - alpha - beta) = 1, explaining much more of cross-country income variation.
Endogenous growth theory, from the mid-1980s, explains long-run growth from within the model. The simplest version is the AK model (associated with Sergio Rebelo, 1991), Y = AK, where K is a broad measure of capital including human capital and knowledge. With no diminishing returns to broad capital, growth equals sA - delta, so a higher saving rate raises the growth rate permanently, and there is no convergence.
Paul Romer (1986) argued that knowledge created by one firm's investment spills over to others, so that while each firm faces diminishing returns, the economy as a whole does not. Romer (1990) modelled the deliberate creation of new ideas by profit-seeking research firms. Ideas are non-rival and partly excludable, so their production needs some market power to be financed. Robert Lucas (1988, "On the Mechanics of Economic Development") put human capital accumulation at the centre, with external effects of average human capital on everyone's productivity.
Unlike in Solow, research, education, openness and market size can change long-run growth. In agriculture they underlie the case for public research and the value of international spillovers of crop varieties. Romer shared the 2018 Nobel memorial prize with William Nordhaus.
The demographic transition describes a shift from high birth and death rates to low ones. In the first stage both are high and population grows slowly; in the second, death rates fall with better nutrition, sanitation and medicine while birth rates stay high, so population grows fast; in the third, birth rates fall with urbanisation, female education, lower child mortality and the rising cost of children; in the fourth, both are low. At 2.5 per cent, population doubles in about 70/2.5 = 28 years.
The dependency ratio is the population aged 0 to 14 plus the population aged 65 and over, divided by the population aged 15 to 64. A country with 30 per cent children, 8 per cent elderly and 62 per cent of working age has a dependency ratio of 38/62 = 61 per cent. As fertility falls, the working-age share rises for a few decades, creating a demographic dividend that can raise saving and output per head if the extra workers find productive jobs; East Asia is the usual example, and South Asia and parts of Africa are in or approaching this window.
Views on population and development have changed. Pessimists (Malthusians, and the Solow steady state) see rapid population growth lowering income per head; optimists (Julian Simon, Esther Boserup) stress its stimulus to innovation and agricultural intensification. The modern view treats high fertility more as a symptom of low development, reduced by female education, child survival and family planning.
Theodore Schultz (1961, "Investment in Human Capital") and Gary Becker (Human Capital, 1964) treated education, training and health as investments that raise future productivity and earnings. Jacob Mincer (1974) estimated the earnings function: the log of earnings regressed on years of schooling and experience; the schooling coefficient is the approximate private rate of return to a year of schooling. A coefficient of 0.08 means each extra year raises earnings by about 8 per cent (exactly e^0.08 - 1 = 8.3 per cent). Estimates vary by country and period, and social returns differ from private ones.
Health and nutrition are both outcomes and inputs of development. The efficiency-wage hypothesis in its nutrition form (Harvey Leibenstein, later formalised by others) holds that very poor workers' productivity depends on their calorie intake, so employers may pay above the market-clearing wage and labour markets may exclude the undernourished. Child stunting has lasting effects on schooling and adult earnings. Biofortified crops and school feeding draw on this evidence.
Comparative development asks why countries that looked similar in 1950 diverged so far. The East Asian group (Japan earlier, then South Korea, Taiwan, Hong Kong, Singapore, later China and Vietnam) combined high saving and investment, rapid expansion of schooling, early land reform in Japan, Korea and Taiwan, stable macroeconomic policy, export orientation, and active but selective state support for industry.
Latin America grew under import substitution until the 1970s, then lost a decade to the 1980s debt crisis, and remains highly unequal, partly from concentrated land ownership. Sub-Saharan Africa stagnated from the mid-1970s to the mid-1990s amid commodity price falls, conflict, overvalued exchange rates and heavy taxation of export crops.
Agriculture's role differed sharply. China's rural reforms from 1978, above all the household responsibility system, which returned farming decisions and residual income to households, together with higher state procurement prices, raised farm output quickly; Justin Lin's widely cited estimate attributes roughly half of output growth in 1978 to 1984 to the change in incentives. Vietnam's Doi Moi reforms from 1986 turned a rice importer into one of the world's largest rice exporters. The Asian Green Revolution raised yields of rice and wheat from the late 1960s. In much of sub-Saharan Africa cereal yields rose far less, fertiliser use stayed low and irrigation limited, which is one reason structural transformation there has been slower.
Import-substituting industrialisation (ISI) protects domestic manufacturing behind tariffs, quotas and overvalued exchange rates; Latin America, India and much of Africa followed it from the 1950s to the 1980s. It built capacity but often produced high-cost industry, foreign-exchange shortages and an anti-export bias that indirectly taxed agriculture. Export-oriented industrialisation (EOI) in East Asia used world markets to discipline firms and earn foreign exchange.
Raul Prebisch and Hans Singer (1950) argued that the net barter terms of trade of primary commodity exporters tend to decline over the long run, because demand for primary products grows more slowly than income (Engel's law), because of synthetic substitutes, and because industrial productivity gains are kept as higher wages and profits while primary-sector gains are passed on as lower prices. The net barter terms of trade are the export price index divided by the import price index, times 100; export prices at 90 and import prices at 120 give 75. The income terms of trade multiply this by the export volume index, measuring the import-purchasing power of exports; with export volume at 110 the index is 82.5.
Dutch disease describes how a resource boom appreciates the real exchange rate and squeezes other traded sectors, including agriculture. Jeffrey Sachs and Andrew Warner (1995) found that resource-rich economies grew more slowly on average, although countries such as Botswana, with strong institutions, avoided the curse.
Across countries a few regularities hold. First, successful structural transformation has almost always been preceded or accompanied by rising agricultural productivity, which frees labour, keeps food cheap, provides raw materials and savings, and creates rural demand for manufactures. Second, where agriculture was heavily taxed, directly or through overvalued exchange rates and industrial protection, farm growth suffered, as World Bank studies led by Krueger, Schiff and Valdes documented.
Recent work adds nuance. Dani Rodrik and Margaret McMillan have shown that labour moving from low- to high-productivity sectors raised growth in East Asia, but in parts of Latin America and Africa after 1990 labour moved into low-productivity informal services, so structural change reduced growth.
For agribusiness, the fastest-growing parts of food systems are often beyond the farm gate, and smallholder participation depends on contracts, standards, finance and producer organisation.
The case for state action in development starts from market failure. Public goods such as agricultural research, rural roads, flood control and basic statistics are non-rival and non-excludable, so markets under-supply them. Externalities, such as pollution, soil erosion that affects downstream farms, or knowledge spillovers, drive a wedge between private and social returns. Information problems, especially asymmetric information between lenders and borrowers or buyers and sellers, lead to credit rationing, adverse selection and moral hazard; Joseph Stiglitz and Andrew Weiss (1981) showed that credit may be rationed even in equilibrium. Coordination failures arise when an investment is profitable only if others also invest, which is the modern reading of the big push (Murphy, Shleifer and Vishny, 1989). Missing rural insurance markets force costly self-insurance.
Government failure is the counterpart. Governments have information limits and self-interested officials; rent-seeking (Anne Krueger, 1974) wastes resources in competition for licences, quotas and subsidies; and state-owned enterprises may be run for patronage rather than efficiency. Robert Bates (Markets and States in Tropical Africa, 1981) explained why African governments taxed dispersed smallholders through marketing boards to favour organised urban groups. Michael Lipton (Why Poor People Stay Poor, 1977) described the same pattern as urban bias.
The practical question is which institution can solve which problem at acceptable cost.
In the 1950s and 1960s many developing countries, influenced by the big push, Harrod-Domar planning, structuralism and the Soviet example, gave the state the leading role. India's Second Five-Year Plan (1956), built on P.C. Mahalanobis's model, emphasised heavy industry and public-sector investment.
After the debt crisis of the 1980s, the pendulum swung toward markets. IMF and World Bank structural adjustment programmes required stabilisation, devaluation, liberalisation and privatisation. John Williamson (1989) summarised the policies that Washington institutions then agreed on in ten points that became known as the Washington Consensus: fiscal discipline, reordering public expenditure priorities toward health, education and infrastructure, tax reform broadening the base, market-determined interest rates, competitive exchange rates, trade liberalisation, openness to foreign direct investment, privatisation, deregulation and secure property rights. In African agriculture, abolishing marketing boards often raised export-crop prices but left gaps in input supply and credit in remote areas.
The post-Washington Consensus, associated with Stiglitz and with the World Bank's World Development Report 1997 (The State in a Changing World), recognised that markets need effective states: the question is the capability of the state and matching its role to that capability. Growth diagnostics (Hausmann, Rodrik and Velasco, 2005) recommend identifying the binding constraint in each economy rather than applying a uniform list.
The developmental state is a state with the autonomy and capacity to direct industrialisation, as in Japan (Chalmers Johnson, MITI and the Japanese Miracle, 1982), South Korea and Taiwan. Peter Evans (Embedded Autonomy, 1995) argued that effective states combined a meritocratic, coherent bureaucracy insulated from capture with close ties to business that provided information.
In agriculture, most successful cases combined markets with strong public action. The Asian Green Revolution relied on public research (national systems linked to the international centres), public extension, public or cooperative irrigation, input supply and credit, and output price support that reduced risk, while farm production itself remained private.
Current thinking gives the state public goods, regulation and safety nets, leaves most allocation to markets, and relies on producer organisations and firms to coordinate value chains. Randomised evaluation of such choices was recognised by the 2019 Nobel memorial prize to Abhijit Banerjee, Esther Duflo and Michael Kremer.
Douglass North (Institutions, Institutional Change and Economic Performance, 1990) defined institutions as the rules of the game in a society, the humanly devised constraints that shape interaction. They include formal rules, informal norms and their enforcement. Organisations, such as firms, cooperatives, political parties and government agencies, are the players. Institutions matter because they determine transaction costs, the costs of finding partners, negotiating, monitoring and enforcing agreements. Ronald Coase (1937, 1960) introduced transaction costs to explain the existence of firms and the importance of property rights.
Daron Acemoglu, Simon Johnson and James Robinson (2001, "The Colonial Origins of Comparative Development") used settler mortality as an instrument for colonial institutions: where Europeans faced high mortality they set up extractive institutions designed to transfer resources, and where they settled they set up institutions protecting property; these institutions persisted and explain much of today's income differences. In Why Nations Fail (2012) Acemoglu and Robinson contrast inclusive economic and political institutions, which secure property, enforce contracts and allow entry, with extractive ones controlled by a narrow elite. The three shared the 2024 Nobel memorial prize. Rival explanations stress geography (Jeffrey Sachs) and culture; the settler mortality data remain debated.
Agrarian institutions are a central case: land tenure and titling, tenancy and sharecropping contracts, interlinked credit-labour-output contracts and water rights all shape farm investment. Secure, transferable rights generally raise investment, though evidence on formal titling programmes is mixed.
Social capital refers to the networks, norms and trust that enable people to act together. James Coleman (1988) and Robert Putnam (Making Democracy Work, 1993, a study of Italian regions) made the concept popular; Putnam found that regions with dense civic associations had better-performing regional governments.
Mancur Olson (The Logic of Collective Action, 1965) showed that large groups find it hard to organise because members free-ride, one reason dispersed farmers are politically weaker than concentrated industries. Garrett Hardin's tragedy of the commons (1968) predicted overuse of open-access resources. Elinor Ostrom (Governing the Commons, 1990; Nobel memorial prize 2009) documented many cases of irrigation systems, forests, fisheries and pastures that communities managed sustainably for long periods, and identified design principles for durable commons institutions, including clearly defined boundaries, rules suited to local conditions, participation of users in making rules, monitoring accountable to users, graduated sanctions, low-cost conflict resolution, recognition of the right to organise and, for larger systems, nested enterprises. Her work distinguishes open access from common property, which is managed by an identifiable group with rules.
Civil society is the space of voluntary association between the household, the market and the state: non-governmental organisations (NGOs), community-based organisations, cooperatives, self-help groups, trade unions, farmers' associations, religious bodies and the media. It delivers services where the state is weak, organises the poor, represents interests, monitors governments and firms, and pilots approaches that governments later scale up.
Several well-known cases come from South Asia. BRAC, founded in Bangladesh in 1972, grew into one of the largest NGOs in the world, delivering health, education, microfinance and livelihood programmes. The Grameen Bank, formalised in 1983 by Muhammad Yunus, pioneered group lending without collateral to poor rural women; Yunus and the bank shared the 2006 Nobel Peace Prize. Randomised evaluations have found modest average effects of microcredit on household income. India's dairy cooperatives on the Anand pattern and the self-help group-bank linkage programme begun by NABARD in 1992 show member organisations linking small producers to markets and finance; these are covered in the cooperatives and rural development subjects.
NGOs may, however, answer more to donors than members and can be captured by local elites.
The International Monetary Fund and the International Bank for Reconstruction and Development (IBRD) were agreed at the Bretton Woods conference in 1944. The IMF conducts surveillance and lends to countries with balance-of-payments problems, usually with conditions. The IBRD lends to middle-income and creditworthy lower-income governments. The World Bank Group now has five institutions: the IBRD; the International Development Association (IDA, 1960), which gives concessional loans and grants to the poorest countries; the International Finance Corporation (IFC, 1956), which invests in the private sector; the Multilateral Investment Guarantee Agency (MIGA, 1988), which insures investors against political risk; and the International Centre for Settlement of Investment Disputes (ICSID, 1966). The World Bank in the narrow sense means IBRD plus IDA.
GATT (1947) was succeeded by the WTO in 1995. Regional banks: the Inter-American Development Bank (1959), the African Development Bank (1964) and the Asian Development Bank (1966). Newer ones: the New Development Bank set up by the BRICS countries (agreement 2014, operations from 2015) and the Asian Infrastructure Investment Bank (operations from 2016).
The United Nations system contributes UNDP (1965; Human Development Reports), UNICEF, the Food and Agriculture Organization (FAO, 1945; Rome), the World Food Programme (WFP, operating from the early 1960s; food assistance), and the International Fund for Agricultural Development (IFAD, created in 1977 after the 1974 World Food Conference, which finances rural poverty projects). CGIAR, formed in 1971, links international agricultural research centres such as IRRI (rice) and CIMMYT (maize and wheat).
Official development assistance (ODA), as defined by the OECD Development Assistance Committee (DAC), is flows from official agencies to countries and multilateral institutions eligible under DAC rules, whose main objective is the economic development and welfare of developing countries and which are concessional in character. A United Nations resolution of 1970, following the Pearson Commission, set a target for donors of 0.7 per cent of gross national income; only a few donors, mostly in northern Europe, have met it consistently.
The economic case for aid in the Harrod-Domar tradition is filling gaps. The two-gap model of Hollis Chenery and Alan Strout (1966) holds that growth can be limited by the saving gap (investment needed minus domestic saving) or by the foreign-exchange gap (imports needed minus export earnings), whichever binds; aid fills the binding gap.
Evidence on aid and growth is disputed. Craig Burnside and David Dollar (2000) found that aid raises growth only in countries with good fiscal, monetary and trade policies, which influenced selectivity in allocation; William Easterly, Ross Levine and David Roodman (2004) found that the result did not survive with more data. Jeffrey Sachs (The End of Poverty, 2005) argued for a large increase in aid to escape poverty traps; Easterly (The White Man's Burden, 2006) and Dambisa Moyo (Dead Aid, 2009) argued that aid can weaken accountability and institutions. Other concerns are fungibility (aid for one purpose frees domestic funds for another), Dutch disease from large inflows, and volatility.
Donors and recipients responded with principles for aid effectiveness. The Paris Declaration on Aid Effectiveness (2005) set five principles: ownership (recipient countries set their strategies), alignment (donors support those strategies and use country systems), harmonisation (donors coordinate and simplify), managing for results, and mutual accountability. Accra (2008) and Busan (2011) followed.
Conditionality, the attachment of policy conditions to loans, was central to structural adjustment. Conditions without domestic support were often not implemented, hence the emphasis on ownership and the poverty reduction strategy papers linked to debt relief under the HIPC initiative (1996).
Agriculture's share of ODA fell substantially from the 1980s to the early 2000s, as donors shifted toward social sectors and governance; the World Development Report 2008 and the food price crisis of 2007 to 2008 renewed attention, leading to commitments such as the L'Aquila Food Security Initiative (2009) and African governments' Comprehensive Africa Agriculture Development Programme (CAADP, 2003), under the Maputo Declaration target of allocating at least 10 per cent of public spending to agriculture. Food aid has shifted from donor-surplus shipments toward cash and local purchase, which avoid depressing local farm prices.
Learn each model as an equation and a diagram: g = s/v with its knife edge; the Solow fundamental equation with its steady state, golden rule and convergence speed; the growth-accounting identity. Practise constructing an HDI from raw data with the goalposts, an MPI deprivation score, the three FGT indices from a list of incomes and a Gini from quintile shares; these short calculations recur. Keep a dated list of authors and books (Seers 1969, Rostow 1960, Nurkse 1953, Hirschman 1958, Harrod 1939, Domar 1946, Solow 1956, Romer 1986/1990, Lucas 1988, MRW 1992, North 1990, AJR 2001, Ostrom 1990) and the founding dates of the World Bank Group, regional banks and UN agencies. Be able to explain, with a country example, where agriculture fits into each theory.
In a bank, development finance institution, NGO or agribusiness firm working in lower-income markets, this subject helps read country risk and opportunity. A credit or investment officer uses GNI per capita, poverty and inequality data, and the pace of structural change to size demand for processed food, inputs and services. A programme officer writes proposals and results frameworks against SDG, MPI and poverty-gap indicators and must compute them correctly from survey data. Someone working with a development bank or donor-funded project must understand ownership, alignment and results-based financing, and the appraisal logic that treats smallholder income, nutrition and women's employment as outcomes alongside financial return.
At senior level, in government, a development bank or a large agribusiness, the subject frames strategy. A chief economist judges whether a country's growth is broad-based enough to support mass-market food demand, whether its institutions protect contracts and land rights, and whether aid, remittances or a resource boom are appreciating the exchange rate and hurting farm exports. A policy adviser weighs public investment in research, roads and irrigation against subsidies, using evidence on returns and on growth-poverty elasticities by sector. Leaders negotiating with the World Bank, IFAD, regional banks or bilateral donors need to know how these institutions allocate funds, what conditions they attach and how results are verified.
Research frontiers include identifying the causal effect of institutions (the settler-mortality debate), measuring and explaining structural change that reduces rather than raises productivity, the agricultural productivity gap between farm and non-farm workers and how much of it reflects measurement and selection, TFP measurement in agriculture, convergence clubs, the aid-growth relationship with better identification, the design and axiomatics of multidimensional poverty measures, and the external validity of randomised evaluations. A senior researcher is expected to combine growth theory with micro evidence, to use panel and quasi-experimental methods, and to advise on how development goals and indicators are defined and monitored.
The manual gives no current values that change each year, such as World Bank income thresholds, the latest HDI or MPI ranks, ODA volumes or country growth rates; readers should take these from the latest World Bank, UNDP, OPHI and OECD releases. Detailed dual-economy models (Lewis, Fei-Ranis, Harris-Todaro), Schultz, Mellor and induced innovation, and Indian poverty-line committees are left to Agricultural Development and Policy Analysis to avoid duplication. Endogenous growth models are described but their optimisation algebra is not derived; open-economy growth, growth with natural resources and optimal growth (Ramsey-Cass-Koopmans) are not covered. Evidence summaries (Kuznets curve, aid and growth, microcredit, East Asian TFP) state the main positions without settling contested questions. Country examples are illustrative, not full development histories.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.