Stages, big push and poverty traps

W. W. Rostow (1960) argued, in The Stages of Economic Growth: A Non-Communist Manifesto, that every society is in one of five stages, in this order: (1) the traditional society, with limited technology and a mainly agricultural output; (2) the preconditions for take-off, in which science and trade begin to change attitudes and institutions; (3) take-off; (4) the drive to maturity, when the economy extends modern technology across its industries; and (5) the age of high mass consumption, when durable goods and services dominate. The exam favourite is the order (UOH 2018) and the number: five (JNU 2020). He said a student of underdeveloped areas is mainly concerned with the preconditions and the take-off (UOH 2021).

Take-off is the short period, a decade or two, in which growth becomes normal. Rostow describes it, as usually summarised, by three conditions: the rate of productive investment rises from about 5 per cent or less to over 10 per cent of national income; one or more substantial manufacturing leading sectors develop with a high growth rate; and a political, social and institutional framework emerges that exploits the expansion. For a leading sector he required that the market for its product expands rapidly, that new techniques are introduced to raise productivity, that it has a continual supply of capital from ploughed-back profits and that it generates secondary expansion in other sectors (UGC NET 2025). In terms of Harrod–Domar’s g = s/v, with a capital–output ratio of 3 a rise in investment from 5 to 12 per cent of income raises growth from 1.7 to 4.0 per cent.

Critics say the stages are descriptive and not a theory (nothing says what moves a country from one stage to the next), that real countries did not follow a common path, and that the dating of take-off is arbitrary. Alexander Gerschenkron (1962) offered the best reply from economic history: the later a country industrialises, the more it differs. A backward country can borrow technology and may need substitutes for missing prerequisites: investment banks in Germany, the state in Russia, to do what accumulated private capital did in Britain.

Nurkse’s vicious circles. Ragnar Nurkse (1953) described a poor economy as caught in two circles. On the supply side: low real income means low saving, which means little capital per worker, which means low productivity and so low real income. On the demand side: low real income means a small market, which means little inducement to invest, so capital and productivity stay low. Nurkse put more weight on the second: “the inducement to invest is limited by the size of the market” (UOH 2016).

Rosenstein-Rodan’s big push. Paul Rosenstein-Rodan (1943), writing about Eastern and South-Eastern Europe, argued that a poor region needs a large programme of investment, a “big push”, not piecemeal investment. His reasons are three indivisibilities: in the production function (above all social overhead capital such as railways, power and ports, which comes in large lumps and has to be there before directly productive activities); in demand (new industries create each other’s markets: workers in a new shoe factory buy the products of other new factories, though none would pay alone); and in the supply of saving (a rise in income needs a high marginal saving rate to be turned into investment). Since piecemeal investment cannot do this, the theory says it will not promote development (the “NOT consistent” statement in the opening question). It calls for comprehensive planning (UGC NET 2024).

Leibenstein’s critical minimum effort. Harvey Leibenstein (1957) argued that an underdeveloped economy is in a low-level equilibrium trap: small gains in income are offset because population grows in response (population growth is his endogenous variable, UOH 2018), pushing income back. To escape, the economy needs a critical minimum effort: a big enough one-time rise in investment to raise income per head faster than population can respond.

Murphy, Shleifer and Vishny (1989) turned the big push into a model with multiple equilibria. Here is its simplest version. A firm that modernises pays a fixed cost c (the indivisibility) and gains m for every share-point of other firms that also modernise, through a bigger market and cheaper inputs. If a share x of firms has modernised, one firm’s net gain is mx − c. It is positive only when x exceeds the critical share c/m. Below it, no firm wants to go first and the economy slides to the trap. Above it, modernising spreads to everyone. Both ends are stable; the threshold between them is not.

The lab shows the gain curve and the path of the share over time. In the preset “Trapped” a quarter of the firms have modernised and the threshold is 40 per cent: the share falls to zero. In “A big push” the same economy starts at 50 per cent and takes off. A subsidy that lowers the fixed cost does the same job as a bigger push, by lowering the threshold. And with little spillover (m below c) there is no high equilibrium at all, which is the strongest challenge to the argument: a big push helps only when the complementarities are real.

If the problem is coordination, the answer is to invest everywhere at once. This is balanced growth, associated with Nurkse (UOH 2021 asks who proposed it) and Rosenstein-Rodan: expand many industries in step, so that each provides the market for the others. Albert Hirschman (1958) disagreed. A poor country does not have the money, the skills or the decisions to do everything at once; its scarce resource is the ability to make development decisions. The better strategy is unbalanced growth: invest in a few sectors with strong linkages, create shortages and surpluses on purpose, and let these pressures draw in further investment. Hans Singer also argued for unbalanced growth (JNU 2019 asks for Hirschman and Singer).

Linkages are the heart of Hirschman’s strategy. A backward linkage is the demand an industry creates for its inputs: an automobile industry stimulates steel production (UOH 2011). A forward linkage is the supply of its product as an input to other industries: expanding the tractor industry raises potential farm output (UOH 2011). Investment should go where total linkages are strongest, because it then induces more investment. The belief behind it is that once certain industries develop, linkage effects induce others to follow.

Gunnar Myrdal (1957) rejected the idea that market forces tend towards equilibrium between regions. His principle of circular and cumulative causation says that a change in one direction sets off other changes that reinforce it. A region that gets ahead attracts capital, skilled labour and demand from the others: backwash effects. Some benefits flow outward, such as demand for the others’ goods and diffusion of knowledge: spread effects. In poor countries, Myrdal argued, backwash effects dominate, so regional inequality widens (the model that “shows the emergence of regional inequalities as a consequence of the development process”, UOH 2012), and public policy has to correct it. He wrote Economic Theory and Under-developed Regions (1957), Asian Drama: An Inquiry into the Poverty of Nations (1968) and An American Dilemma (1944): UGC NET asks which three of five titles are his.

The modern literature restates these ideas with explicit models. Azariadis and Stachurski (2005) review the ways a poverty trap can arise: increasing returns, credit constraints, externalities and the like, each of which produces two stable equilibria. The hard question is whether real economies are in traps. Kraay and McKenzie (2014) examine the modern evidence on S-shaped savings at the country level, big-push theories based on coordination failures, hunger-based traps and occupational traps, and conclude that these traps are rare and largely limited to remote or otherwise disadvantaged areas, with geographic traps the most likely form. Their policy conclusion differs from the call for a big push in aid: the more likely traps call for less traditional measures, such as promoting migration. The practical message is that a big push is a hypothesis to be tested for each constraint, not a general law.

Nobel Memorial Prizes: Myrdal 1974 (shared with Hayek), Milton Friedman 1976, Arthur Lewis 1979 (shared with Schultz), Lawrence Klein 1980, Abhijit Banerjee 2019 (with Duflo and Kremer). Joan Robinson’s The Accumulation of Capital (1956) belongs to the Cambridge tradition of the growth module.

Questions on Rostow, the big push, Nurkse, Leibenstein, Hirschman and Myrdal. Questions on the dual economy are in Lesson 3; the dependency school and strategy in Lesson 6.