. 6
( 21)


characteristics of many of the less-developed societies which often render inapplicable the assump-
tion of an automatic, direct and smooth link between increased savings, increased investment, and
income growth.
20 However, s, the savings rate, is a constant value, and this rigidity is responsible for instability in the
Harrod-Domar formulation of the economic growth process.
21 This means that all production isoquants form right angles, and there is a constant K/L ratio which
is most efficient, that is, least cost, in production. This Harrod-Domar assumption also is known as
a Leontief, or fixed-proportions, production function.
22 Harrod (1948: 72“100) explained this instability as illustrating the Keynesian problem of booms
and busts, and the real-world reality that full-employment equilibrium without inflation was not an
automatic outcome of the workings of a capitalist market system. Also see Chapters 1 and 2 of Sen
(1970), which includes the original published journal contributions of Harrod (in 1939) and Domar
(in 1946).
23 Consider the following simple numerical example from Sen (1970: 13). If s = 0.2 and v = K/Q = 2,
then gw = s/v = 0.2/2 = 0.1, or 10 percent. A rate of growth of 10 percent would mean that if Qt’1 = 90,
then Qt = 100 (measuring growth as a proportion of Qt). If investors expect Qt = 100, then ”Kt = It =
Classical and neoclassical theories 139
v”Qt = 2 — 10 = 20. Given a savings rate of 0.2, the Keynesian income multiplier will be equal to 5,
so Qt = multiplier — ”spending = 5 — 20 = 100, so investor expectations are realized (St = It = 20) and
steady-state equilibrium is attained and maintained, as long as expectations remain the same.
However, imagine that investors expect Qt = 101, so that the expected ”Qt = 11. Then, ”Kt =
It = v” Qt = 2 — 11 = 22. By the multiplier formula, then actual Qt = multiplier — ”spending = 5 —
22 = 110. Investors will feel they have under-invested by anticipating too low a level of Qt and in
the next period will invest more, thus pushing the economy further away from equilibrium. The
same dynamic works if investors under-anticipate the level of GDP, with continued decreases in
GDP following.
If the labor market is also added to the model, such that the labor force grows yearly by the rate
n (= ”L/L), then the trick of reaching and maintaining equilibrium in the Harrod-Domar model
is exacerbated. Then, to avoid rising unemployment, even when ”Q/Q = s/v, this warranted rate,
gw, must equal the natural rate of growth, gL, of the labor force, n. Thus a steady-state equilibrium
requires gw = s/v = n = gL. The knife-edge equilibrium problem is even more exacting and the
possibility of steady-state equilibrium even more remote when a growing labor force is introduced
into the model.
24 In fact, a whole body of growth and income distribution literature “ so called Cambridge growth
models “ predates the Solow model, but this perspective had a smaller impact on development
economics than Solow™s formulation would have. The Cambridge models tended to focus on
savings behavior and the adjustment to equilibrium growth. See Sen 1970, esp. Chapters 2 and 3.

Adelman, Irma. 1961. Theories of Economic Growth and Development. Stanford: Stanford University
Basu, Kaushik. 1997. Analytical Development Economics: The Less Developed Economy Revisited.
Cambridge, MA: MIT Press.
Fisher, H.E.S. 1971. The Portugal Trade. London: Methuen.
Harrod, R.F. 1948. Towards a Dynamic Economics. London: Macmillan.
Hunt, E.K. 1979. History of Economic Thought: A Critical Perspective. Belmont, CA: Wadsworth.
Jones, Charles I. 1998. Introduction to Economic Growth. New York: W.W. Norton & Co.
Maddison, Angus. 1982. Phases of Capitalist Development. Oxford: Oxford University Press.
Mankiw, N. Gregory, David Romer and David N. Weil. 1992. “A Contribution to the Empirics of
Economic Growth,” Quarterly Journal of Economics 107 (May): 407“37.
Robinson, Joan. 1978. Aspects of Development and Underdevelopment. Cambridge: Cambridge
University Press.
Rogin, Leo. 1956. The Meaning and Validity of Economic Theory. New York: Harper & Bros.
Sachs, Jeffrey D. 2005. The End of Poverty. London: Penguin Books.
Sen, Amartya (ed.). 1970. Growth Economics. Harmondsworth: Penguin Books.
Smith, Adam. 1973. An Inquiry into the Nature and Causes of the Wealth of Nations. New York: The
Modern Library.
Solow, Robert. 1956. “A Contribution to the Theory of Economic Growth,” Quarterly Journal of
Economics 70 (February): 65“94.
5 Developmentalist theories of economic

after reading and studying this chapter, you should better understand:
• the concept of hidden development potential in less-developed nations;
• the possibility of market failure and the role of positive externalities in creating
virtuous circle effects;
• the importance of social overhead capital and a nation™s augmentable initial
endowments to growth;
• balanced versus unbalanced growth strategies and their shared paradigmatic
• the theory of export pessimism;
• backward and forward linkage effects and their key role in development;
• the idea of hidden comparative advantage;
• the potential role of surplus labor as a stimulant to growth in Lewis™s dualist
framework for transition;
• Rostow™s stages of growth theory, particularly the “take-off” stage.

After the Second World War, and particularly after the quick success of the United States-
financed Marshall Plan in helping to rebuild the European economies, several economists
who had been directly involved either in the Marshall Plan or with institutions such as the
United Nations and the World Banks turned their attention to the question of the economic
development of less-developed regions. Among these early pioneers of development thinking
were the Finnish economist Ragnar Nurkse, the Austrian economist Paul Rosenstein-Rodan,
the German-born economist Albert Hirschman, the West Indian and later Nobel Laureate
economist Sir Arthur Lewis, and the American economic historian Walt Whitman Rostow.
Only Lewis remained outside of the policy-making institutions in the late 1940s and early
1950s, but by 1957 he too was employed by the UN.
In a broad sense, the ideas of these early development economists were mutually
supportive. They formed a loose school of thought on the issue of economic development,
emphasizing a less theoretical and more historical and practical approach to the question of
how to develop “ particularly in relation to those who stressed the applicability of neoclas-
sical models, such as the Solow model discussed in the last chapter. Like any such school of
analysis, there were differences of emphasis and interpretation between these theorists. These
Developmentalist theories of economic development 141
differences are particularly striking in the work of W.W. Rostow, who stressed a descriptive
approach while emphasizing the near inevitability and predictability of economic develop-
ment, based on the premise that the industrial past of europe presents a rough picture of the
approaching future of the developing nations. The others emphasized analytical constructs
and were not striving to construct a megatheory of economic history. Yet they shared many
fundamental propositions. Above all, they coincided in believing, in Rostow™s words, that
“the tricks of growth are not that difficult” (Rostow 1960: 166). They also felt that the time
period necessary for achieving economic development in the less-developed world would be
relatively short, a matter of a decade or perhaps a generation, rarely more.
Rostow predicted that, with US aid, by 1970 nine nations (including India) would pass
the crucial threshold into a condition of self-sustaining economic growth. His bold ideas
heavily influenced US President Kennedy, who then made a proposal to the UN, which was
subsequently accepted, to name the 1960s the United Nations Decade of Development
(Toye 2004: 176“9). While this initiative raised expectations in the underdeveloped nations,
with its anticipation that aid-driven GDP growth could achieve a targeted rate of five percent
a year, the concept was largely a feat of public relations. By 1965 the US and other major
donor nations had lost their enthusiasm for the concept and aid began to decline.
Furthermore, all these economists shared, to different degrees, an affinity for the work
of British economist John Maynard Keynes, whose views on macroeconomics had swept
the economics profession in the late 1930s and 1940s. Thus, they emphasized aggregate
phenomena, such as the rate of saving, measured by the share of income not consumed in
gross national product (S/Y), and the rate of investment (I/Y), as fundamental variables, a
perspective which fits well, too, with the Solow-type model of the previous chapter. They
agreed with the Keynesian assumption that poor economic performance reflected a lack of
aggregate demand, rather than a shortage of, or limits to, resources. Keynes had come to this
conclusion based on his knowledge of the advanced capitalist nations, not from studying the
dualistic, less-developed economies to which this insight would be applied.
These early development economists also manifested a notable preference for industrializa-
tion as the driving force of economic growth, believing industrialization would release a tide
of prosperity lifting all other sectors of the economy. Finally, while these developmentalists
had a profound respect for market forces, they were not hesitant to advocate large-scale,
short-term governmental intervention into the economy, very much after the Keynesian
manner, if that might be expected to force economic growth. Markets were perceived as a
means to realizing the end of economic development; they were not an end in themselves.
Markets could achieve some objectives rather well, but there were other spheres in which the
market worked less well. Under certain conditions, an assertive, and even a leading, role for
government was to be encouraged and was perhaps necessary. In the long run, however, the
developmentalists expected that an economy would achieve its best results with a competi-
tive market interacting with a responsive and efficient governmental apparatus, and thus
the interventionist role of government in development would be reduced to its stabilizing
function, as in the already-developed nations. In this sense, the developmentalists had very
conventional economic ideas, but only in the very long run.
In this chapter, some of the leading theories of the developmentalists are examined. Their
theories and recommendations are more pragmatic and operational than the neoclassical or
classical formulations. The theories were devised with an eye to directly affecting public
policy in the less-developed countries. We shall see that their influence on the thinking of
many economists remains strong, though there have been, and need to be, further refinements
of their analyses.
142 The Process of Economic Development
The theory of the big push
One of the early theories about how a country might create the conditions for economic
progress, where growth and development had not already arisen spontaneously, was
formulated by Paul Rosenstein-Rodan on the basis of research he had conducted during
the Second World War. After analyzing the economic structures of a number of poor
eastern and South-east european nations, Rosenstein-Rodan drew a number of conclu-
sions which became basic building blocks for the field of development economics
emerging after the war.1
Rosenstein-Rodan was noted for his effort to call attention to the hidden potential for
economic development in less-developed regions. Much of his work centered on taking
advantage of the increasing returns that could be realized from large-scale planned industri-
alization projects that encompassed several major sectors of the economy simultaneously.
A “big push” of concurrent industrial investments could launch a chain reaction of virtuous
circles and complementary investments that would then ripple in many directions through
the economic system. Large-scale investments in several branches of industry would lead to
a favorable synergistic interaction between these branches and across sectors. If economic
development was to get a start in the now less-developed nations, Rosenstein-Rodan
argued, it would have to come from a concerted and substantial “push” from government
to create, effectively, an entire industrial structure in one huge and interlocked undertaking
(see Focus 5.1).

Although Rosenstein-Rodan does not detail this point, one can sketch such virtuous circle
effects: large-scale investments in steel-making could lead to research in metallurgy which
would have “positive external” effects on companies which use metal products. Perhaps
stronger alloys could be found that could then be used in the metal-fabricating industries,
reducing wear and fatigue and downtime for the machines in this sector. All this could
reduce costs to another branch of industry, perhaps in railway equipment manufacturing.
Lower costs in the rail equipment could then be passed on to farmers, in the form of lower
transport costs.
Farmers, in turn, would now be able to invest in better mechanical equipment from the
metal-manufacturing industry, creating a further surge of positive ripple effects. Each branch
of industry, or at least many branches of industry, would be caught in a web of interacting
and mutually complementary activities. The more efficient are supply conditions, the lower
costs of production will be, and the greater the demand for the product. Cross-sector posi-
tive externalities will also be transmitted, for example, from industry to agriculture.
In recent years interest in Rosenstein-Rodan™s big-push theory has grown. His ideas were
formalized by Kevin Murphy, Andrei Shleifer and Robert Vishney (1989) and his views are
increasingly invoked by proponents of endogenous growth theory (see Chapter 8). This
more recent work tends to highlight the role of demand spillover effects which, like the
examples above, stress the virtuous circle effects which occur when an expanding manu-
facturing sector that raises productivity then stimulates income growth that, in turn, leads
to Increasing demand for the products of the expanding manufacturing sector. Increasing
growth in this manufacturing sector could lead to increasing demand for inputs that “
because they are produced on a larger scale “ lead to economies of scale in the production
of these inputs. This virtuous circle will then lower the costs of production for the manufac-
turing sector, which could lead to increasing demand and growth “ another virtuous circle!
Source: Hoff and Stiglitz 2001: 401“13
Developmentalist theories of economic development 143
While concentrating on the hidden potential of large-scale future investments, with each
successive increment to investment having an increasingly strong impact as output expanded
at a rising rate,2 Rosenstein-Rodan simultaneously maintained that these potential gains
could not be realized within a purely market frame of reference. Individual entrepreneurs
would be unlikely to invest enough to “push” the less-developed economy forward at its
maximum potential rate, because under the profit-and-loss calculations of private entrepre-
neurs, their frame of reference would be too limited. Profit-maximizing steel producers are
not concerned about whether their own private investments, if sufficiently large, will induce
other investments and technical change in metallurgy which will then make that industry
more profitable. Backward linkage effects which may be provoked by the investment actions
of the steel industry are not taken into consideration by private decision-makers in the steel
industry, because those firms cannot profit from these spin-off industries or even calculate
the likelihood of the emergence and success of such linked firms.
Using Rosenstein-Rodan™s terms, the steel industry cannot “appropriate” the future poten-
tial benefits to be gained in other sectors that are external to their business, and hence they do
not take these effects into account in making their private investment decisions. Because of
this information and appropriation failure, market decisions will lead to a sub-optimal level
of investment from the standpoint of society as a whole.3 Rosenstein-Rodan was convinced
that there were many such hidden potentialities for expanded production in less-developed
economies that went unexploited because of the inability of the market economy to coordi-
nate the multitude of simultaneous investment decisions that needed to be made.
Insufficient economic development would occur, because the private sector mechanisms
in place in less-developed societies lead to economic decision-making which is sub-optimal.
More investment was needed, and in many places at one time, in order to shift the economy
away from its low-level equilibrium trap and toward rapid and sustainable growth. Of partic-
ular importance to this process is the provision of social overhead capital or infrastructure:
roads, bridges, docks, communications systems, hospitals, schools, utilities, irrigation and
flood control projects, and so on, which also generate substantial positive external benefits
to society as a whole.

The market mechanism alone will not lead to the creation of social overhead capital,
which normally accounts for 30 to 35 percent of total investment. That must be spon-
sored, planned, or programmed (usually by public investment). To take advantage of
external economies (due to indivisibilities) requires an “optimum size” of enterprise to
be brought about by a simultaneous planning of several complementary industries.
(Rosenstein-Rodan 1984: 209)4

For example, if schools are built and operated under the profit motive, then they will be
available only for the child whose parents can pay. Bright and ambitious children of poor
parents will be less likely to gain needed skills, and society™s labor force will be under-skilled
and operating below its potential as a consequence. The hidden potential of the future labor
force may never be realized if the market is left to provide social overhead capital, such as
schools. This is the framework that Rosenstein-Rodan and others utilized when they argued
that the market mechanism will not adequately create social overhead capital.5
In terms of the sequencing of investment decisions, Rosenstein-Rodan prioritized social
overhead capital as an essential initial endowment, albeit one that nations have to actually
create. Social overhead capital is not an initial endowment in the same sense that, say,
land is.
144 The Process of Economic Development
Because of indivisibilities and because services of social overhead capital cannot be
imported, a high initial investment in social overhead capital must either precede or be
known to be certainly available in order to pave the way for additional more quickly
yielding directly productive investments.
(Rosenstein-Rodan 1976: 635)

Rosenstein-Rodan™s idea of the need for creating a “big push” of investment simultane-
ously in a number of branches of industry, and his emphasis on social overhead capital as
fundamental to the success of the development project in less-developed nations are his best-
known contributions to the literature, but they are not the whole of what he had to say about
the development process. Summing up his own contributions in the area of development
economics, Rosenstein-Rodan claimed that he had made four innovations.
First, he had stressed disguised unemployment, that is, those workers, particularly in
agriculture, who receive very low or no pay and whose work effort results in relatively little
increase in total output. Their labor could be tapped to create the vast public works of social
overhead capital which would be necessary for development, without reducing output in the
Second, by emphasizing the complementarity, and the external economies, of distinct
investments, Rosenstein-Rodan demonstrated that large-scale investments could have an
impact on overall economic growth greater than might be expected based on the calcula-
tions of individual entrepreneurs alone. It is necessary to take into account the positive
externalities of one investment on others and on the possibility of increasing returns from
successive units of investment. In order to achieve these serendipitous effects, however,
economic planning of a limited nature would be necessary. Key industries or branches of
industry would have to be targeted for expansion, and their initial investments would need
to be subsidized if they were to occur at all.
Rosenstein-Rodan™s third innovation was his emphasis on social overhead capital. Such
investments, he argued, should precede the expansion of consumer-goods manufacturing
investment if the latter is to be successful. As we shall discover in Chapter 8, this is a view
supported by recent research on endogenous growth models.
And fourth, a “big push” of investment through the economy could result in technological
external economies. These effects he defined in terms of workforce training. Large-scale
industrialization could contribute to a socially beneficial level of labor training that would
have spread effects to other sectors throughout the economy, whereas incremental, market-
driven development would not have the same impact, or at least dependence on the market
would result in sub-optimal social quantities of such training. Private businesses would not
invest in the socially optimal level of labor training, again because any individual employer
will be unable to appropriate the increases in income created by the new skill, especially if
a worker moves on to another employer, who would not need to make any investment to
benefit from the worker™s increased skill level. However, under the big-push approach, labor
training could be funded as part of a more general development plan. A broader time and
planning horizon could be entertained by government, which could determine the training
needs of an entire industrial complex and calculate the social profitability of any invest-
ment of additional educational expenditures and labor training. As we shall see in Chapter
8 in the discussion of endogenous growth theories, Rosenstein-Rodan was ahead of his
time in maintaining that appropriate labor training was of equal, or perhaps even greater,
importance than capital accumulation in the process of industrialization and economic
Developmentalist theories of economic development 145
a theory of balanced growth
Ragnar Nurkse, like Rosenstein-Rodan, emphasized above all the need for a coordinated
increase in the amount of capital utilized in a wide range of industries if the critical threshold
level of industrialization was to have a chance of being achieved. Nurkse agreed that a
massive injection of new technology, new machines, and new production processes spread
across a broad range of industrial sectors held the key to igniting the development process in
less-developed nations.7

Export pessimism and the need for domestic industrialization
This perspective of how to initiate rapid economic growth needs to be contrasted with what was,
in the 1940s, a received doctrine in trade theory: to foster economic progress, less-developed
regions were counseled to concentrate on increasing their exports of tropical products and raw
materials, products in which, it was suggested, such countries had a comparative advantage.
In Nurkse™s view, this rather standard prescription for accelerating economic growth in less-
developed countries was likely to yield meager results for two basic reasons. First, Nurkse
maintained that in future the world demand for tropical products and raw materials would
be relatively limited and slow to expand. An increase in supply under such conditions would
result in a decrease in the market price. The reduction in price could be of such a magnitude
that the total revenue received (= unit price — quantity of the product sold on the world market)
after an increase in supply could be less than the export income that was received prior to the
drive to increase such exports.8
Nurkse did not devote himself to proving this point; rather he seems to have utilized
this insight more as a working assumption based upon the weak pattern of prices for tradi-
tional primary exports from the less-developed nations he observed in the first half of the
twentieth century. Because of this break with the orthodox view that the colonial and post-
colonial regions had a comparative advantage in tropical products and raw material exports
which could be further exploited through even more ambitious and pragmatic economic
policy to expand such exports, Nurkse was branded an “export pessimist” (see Focus 3.5
for details).
The second reason for his rejection of the export-led road to development was based on
Nurkse™s interpretation of the propensity to import.9 In orthodox trade theory, it was assumed
that a less-developed nation with the ability to export either tropical products and/or raw
materials would use the income earned to import machinery, equipment, and manufactured
consumer goods for domestic consumption. Trade would balance, that is, the value of exports
would equal the value of imports, at least over an intermediate period of time. To challenge
orthodox assumptions, Nurkse utilized a socio-psychological theory which explains why
consumption continues to rise as income rises. This theory assumes that some “wants” are not
innate, but rather are socially created. In this framework, some new goods are “demonstrated”
to be desirable, because they are consumed by higher-income recipients in society. These
goods confer social status and are therefore sought by others with less income.10
Nurkse believed that the less-developed regions would be very vulnerable to the perni-
cious affects of this international demonstration effect. High-income consumers would spend
inordinately on imported luxury products to “keep up with the Joneses” of the richer nations.
Not only would there be an upward bias toward imports, especially of consumer goods, but
the already limited potential supply of savings in the less-developed nation that might have
been directed toward much needed domestic capital formation would be drawn down, as
146 The Process of Economic Development
consumption as a share of total national income rose. Furthermore, the drive to show status
through the importation of luxury commodities would conceivably cut into the ability of
the economy to purchase imported machinery for industry, as the two forms of demand for
foreign exchange competed for a limited stock of foreign exchange earned from primary
product exports.
Less-developed regions were poor, according to Nurkse, because productivity per worker
was low, and productivity, in turn, was low because savings were low, just as in the Solow
model. With a low capacity for savings, the level of investment would, by necessity, be low,
and consequently, with only a modest amount of capital equipment available to each worker,
the end result had to be a low level of per capita income because output per worker would
be low. Small, incremental increases in capital formation would not solve the problem, in
Nurkse™s view. The market-based approach would more than likely fail, because as an indi-
vidual business or a single industry alone attempted to raise its output level by increasing its
individual capital investment, it ran the risk of not finding a market for its product because of
the low level of overall average income. Alternatively, Nurkse emphasized that by attempting
to solve the problem of underdevelopment via an expansion on the supply side alone, that
is, through the expansion of production capacity, one ran the risk that the lack of demand for
new output would short-circuit the attempt to move the economy forward.
The only solution that Nurkse foresaw, as had Rosenstein-Rodan, was via balanced growth.
Large-scale increases in supply sweeping across a large number of industrial sectors would,
at the same time, be met by a large-scale increase in demand created by the same expansion.11
The essential demand-side stimulus would come from industries that were expanding as a
result of the overall, balanced investment program; they would need more inputs of raw
materials, intermediate or semi-processed products, and labor, and their act of buying inputs
would create income for their suppliers. This income would then be transposed into a further
expansion of demand by other firms and by workers in those firms buying the increased array
of domestic goods available. But this widespread expansion could happen only if the initial
effort at development was “balanced,” that is, only if supply increases were coordinated with
simultaneous demand increases across the economy.
Although Nurkse™s theory of balanced growth is very similar in many respects to the big-
push formulation of Paul Rosenstein-Rodan, Nurkse™s work was not merely a repetition. He
did not advocate planning, as did Rosenstein-Rodan, nor was his approach open to the charge
of being statist or of being dependent on the dominance of the public sector, a criticism that
might be leveled at Rosenstein-Rodan. Rather, Nurkse felt that dynamic fiscal policies could
have a very positive effect on the prospects for development without large-scale government
involvement in production decisions or large-scale planning projects. Specifically, Nurkse
advocated forced savings through an increase in taxes on upper-income recipients. The
government, then, could repress the level of consumption out of national income, thereby
increasing the level of overall savings. Then, the increased investment funds generated could
be allocated to the most promising industrial sectors, possibly via government-operated
development banks (see Focus 9.3 for details) designed to identify and promote industriali-
zation in the private sector or via private sector banks.
Industries would be encouraged to increase their capital formation and to raise their
productivity, both because of the availability of loans from the development banks and
because of the effects of infant industry protection, in which government would raise tariffs
against cheaply manufactured imports from the advanced nations that might compete with
the production of the new enterprises. Thus, both supply and demand factors would be
addressed. The supply of savings would be expanded, leading to an increase in the supply
Developmentalist theories of economic development 147
of available domestic output via enhanced capital formation. At the same time, a market
for domestically produced goods would be created, because potentially competing imports
would be deflected via tariffs to the purchase of lower-priced domestically produced goods,
a strategy which, later, became known as import substitution industrialization (discussed
in detail in Chapters 9 and 10).
Like Rosenstein-Rodan, Nurkse felt strongly that less-developed regions possessed the
hidden potential for greater progress; the resources and talents of society simply needed to
be coordinated and released.

Unbalanced growth
Not all developmentalist economists believed, however, that the resources needed for imple-
menting a big push or a balanced growth strategy actually were available, though ideally this
might be the optimum path in some abstract sense. One who voiced such concern was Albert
O. Hirschman. Like most of the pioneers in the field of economic development, Hirschman
was involved in the postwar economic reconstruction of europe. However, this experience
was followed by a four-year stint in Colombia, where his role as adviser to the National
Economic Planning Board arose as a result of the recommendation of the World Bank
(Hirschman 1984: 90). Hirschman™s experiences in Colombia were formative; he would
draw on a fund of experiences within this less-developed country to provide specificity to his
emerging ideas on development. His work since that time has continued to convey a sense
of immediacy and applicability that was at times lacking in the abstract and aggregative
approaches employed by Rosenstein-Rodan, Nurkse, and other developmentalists.
Because Hirschman employed the term unbalanced growth in his major work (1958) on
economic development, and because his seminal work came considerably later than the
ideas expressed by Rosenstein-Rodan and Nurkse, it has been commonly assumed that
Hirschman™s work was to be interpreted as an attack on the theory of big-push or balanced
growth. It is important, therefore, to note that Hirschman agreed with the vast bulk of the
ideas expressed by both Rosenstein-Rodan and Nurkse. He supported an “industrialization
first” strategy, and he firmly believed that the key to rapid industrialization was to be found
in large-scale capital formation in several industries and sectors. Hirschman also shared the
optimistic opinion that less-developed nations harbored significant hidden reserves of talent,
that potentially complementary relationships were waiting to be released, and that there
were major potential externalities which would be instrumental in speeding the thrust toward
industrialization. Hirschman™s own interpretation of the relationship of his work to that of
Rosenstein-Rodan and Nurkse was that he was a dissenter within the framework of the big-
push/balanced growth paradigm.
The less-developed economies did indeed need a big push; without it, there would be
either a snail™s-pace rate of economic and societal change, or perhaps no discernible progress
at all. But Hirschman advocated a big push for only a limited range of industries, with the
idea that by inducing development in key sectors first, overcapacity would be created in
these sectors, while supply bottlenecks would simultaneously increase production difficul-
ties elsewhere in the economic structure. These bottlenecks would create pressures for new
investments to resolve the supply inadequacies. In other words, Hirschman deliberately
advocated the unbalancing of the economy, creating disequilibrium situations, for two basic
First, he maintained that there were resource limits in the less-developed regions and that
this would necessitate prioritizing some areas of industry over others for the use of limited
148 The Process of Economic Development
investment funds. It was impossible to move forward on a “broad front” in all industries at
the same time as was envisioned in the big-push and balanced growth theories. Second, in
deliberately unbalancing the economy and in creating excess capacity in some areas and
intensifying shortages in other areas, he believed that the pressures created would result in
subsequent reactions that would speed the development process by opening up opportunities
for profit for new entrepreneurs.
In industries where overcapacity was generated, the output of these sectors would be made
cheaper than previously, as a result of economies of scale; as output grew, unit costs of
production would decrease as the firm moved down the average total cost curve. Hirschman
believed this decrease in costs, assuming these were passed on to the final consumer, would then
contribute to stimulating upstream investments. Hirschman™s theory might be illustrated with
the following example: by deliberately oversupplying electrical power, and thus lowering
its price to users, sectors of the economy which used large amounts of electrical power as
an input into their production process could be stimulated by this lowering of their marginal
and average costs. Hirschman argued that in conditions of limited resources, as applied in
the less-developed world, where it would be impossible to simultaneously increase electrical
power-generating facilities and still have sufficient investment funds to stimulate industries
that were intensive users of electrical power, it was the task of economic development econo-
mists to prioritize one of these two possible areas of growth. Then, rely upon the positive
effect of disequilibrium imbalances to push the economy forward as private entrepreneurs
responded to the possibilities created by bottlenecks via the market.
The priority sector could be the upstream or the downstream industry. excess capacity
in social overhead capital could lead to the rapid expansion of private sector investments,
which would then subsequently utilize the excess capacity generated in the public sector,
thus justifying its initial creation. On the other hand, were private sector investments to be
prioritized, the need for a rapid increase in social overhead capital would subsequently mani-
fest itself as the demand for electricity outstripped the supply; the profitability of more social
investment would be made manifest. Bottlenecks and shortages of some inputs would create
opportunities for profits for private entrepreneurs to fill in the gaps. These profits would
attract other investors in search of profit windfalls created by such bottlenecks. Investments
would flow into under-supplied sectors, where prices and profits were rising. Perhaps this
response would overshoot the needs of the market, thereby creating downstream opportuni-
ties for other businesses that could turn the new excess capacity and falling prices to their
Imbalances, or disequilibrium situations, would be conducive to further change; doing
things “the wrong way around” could provide greater benefit than any other strategy in
Hirschman™s view. Basically, what Hirschman was explaining was how a market system
responds to shortages and surpluses, but his contribution was to suggest how development
planners might utilize market disequilibriums to stimulate economic progress.

Backward and forward linkages
One of Hirschman™s best known and most creative ideas was that of industrial linkages.
When one industry expands, it requires inputs from other industries to be able to produce.
These are called backward linkages, that is, they are induced effects on the output of
supplying industries. For example, coal mining and iron ore mining constitute backward
linkages from a steel mill. On the other hand, when an industry sells and transports its
production to other firms and sectors in the economy, these are the forward linkages of
Developmentalist theories of economic development 149
the original producer, that is, the induced effects of the output of the first industry in the
direction of the final consumer. The metal fabrication industry and the chemical and paint
manufacturing industry which use the output of the steel industry as their inputs would be
forward linkages to the steel industry, and these industries might have further forward link-
ages to, say, the production of household stoves. Railroads or alternative forms of transport
would enter the example as both backward and forward linkages to steel production and at
each stage of production.
Thus, the production of one firm in one industry has a multiplicity of backward and forward
linkages with firms in other industries in the domestic economy and, perhaps, abroad as well.
In communicating the induced effects from one sector of the economy to another via short-
ages and excess capacity in Hirschman™s unbalanced growth process, the size of potential
backward and forward linkages were of paramount importance in evaluating where to locate
the initial investment. Development strategies could be built around the maximization of the
estimated stimulus of promoted industries in generating domestic backward and forward
Hirschman argued that the case could be made for large-scale capital-using projects, such
as steel mills, if these investments could stimulate significant backward and forward linkages.
Indeed, such investments could spark the creation of whole new industries, providing not
only increased output, but also increased employment and, with rising levels of production,
lower costs and lower prices to consumers as the benefits of economies of scale were reaped.
Nor would such large-scale capital-intensive investments necessarily displace workers, as is
sometimes alleged. In an empirical study which analyzed the relationship between industrial
structures and employment in Latin America, Hirschman found that:

once the indirect employment effects (via backward and forward linkages) are taken
into account, investment in large-scale (capital-intensive) industry turns out to be just
as employment-creating as investment in small-scale (labour-intensive) industry for the
industrially advanced countries of Latin America.
(Hirschman 1984: 97)

How might such linkages be measured? Even at the time Hirschman was writing, input“
output analysis of national economies was being elaborated, based on the pioneering work
of Wassily Leontief at Harvard. Using input“output tables, it is possible to calculate the
impact of a change in the output of one industry on supplying backward-linked producers
and simultaneously on the production of forward-linked industries that use the originating
output as inputs. An input“output table is a matrix showing the multiplier effects of the
impact on other industries per unit change of output in another industry, as well as on labor
use, imports, and final demand. For any country seriously thinking about stimulating devel-
opment, at least a simple input“output table and the required calculations are almost essential
for effective decision-making and monitoring of effects.

Changing the social organization of the labor process
Hirschman advanced an additional reason for promoting a capital-intensive, unbalanced
industrialization program in less-developed nations. Many social scientists had argued that
an attitude of “achievement” needed as a precondition for industrialization was missing in
both the labor force and in management in less-developed nations. It had been suggested
that standards in the workplace were exceedingly lax in less-developed countries, and that
150 The Process of Economic Development
neither workers nor managers were willing to take responsibility for errors in production.
Slack management techniques often made it impossible to assign culpability when tasks
were left uncompleted or were not completed within the time-norm set for a particular
task. Hirschman did not take issue with this characterization of the work place in the
less-developed nations. Rather, he noted that with the introduction of more advanced,
machine-paced techniques, it would become easier both to calculate reasonable work-norms
and to evaluate both success and failure in completing tasks (see Focus 5.2).
Hirschman thus advocated new forms of production on the shop floor that might “hot-house”
the completion of the less productive handicraft and manufacture stages of industrialization
and allow the less-developed countries to move quickly to the machino-facture stage and
its higher level of productivity. Under simple, relatively labor-intensive manufacture, the
human operative has a great deal of control over the pacing of and output of a machine, since
the worker chooses how quickly to work and how much effort to put into the production
Under machino-facture, or more capital-intensive production techniques, however,
norms and rates of production are pre-determined to a great extent by the pace at which the
machines are engineered to operate. Workers and managers are faced with a situation that is
much more “all or nothing”: maintain the pace of work and the quality of output determined
by the machines, or risk losing employment. This situation, argued Hirschman, forces a
change in the labor process which could lead to a rapid rise in productivity and could force
institutional and behavioral changes that would be conducive to further economic devel-
opment. This is another example of a Hirschmanian “pressure point,” or disequilibrium
process, designed to disrupt the production process and society in a way that promotes a
positive outcome.
New attitudes and expectations regarding the labor process, both at the level of the shop
floor and in management, could be inculcated as a by-product, or positive externality, of
this more capital-intensive industrialization as the pace of work is increasingly dictated by
machines. Both traditional labor practices and often ritual management responses would be
made untenable with the new rules of the game, and a new cadre of workers and managers
would be created as a complementary effect of industrialization, with positive and cumulative

A study of Mexican corporations conducted by the international consulting company,
Vertex, may illustrate the significance of achievement attitudes. In comparing output per
worker in Mexico with similar firms in advanced nations, Vertex found that productivity
was 50 percent below international work-norms. In the most complex operations relating
to production and maintenance, productivity was only 40 percent of what might be antici-
pated elsewhere.
Only 55 percent of the work day was devoted to work; 17 percent of time was spent
in office gossip and coffee drinking, and 28 percent of the day was lost to (1) inefficiency
of personnel, (2) communication problems, (3) repeating work because of errors, and (4)
repeating instructions to employees. Among the difficulties cited by Vertex were the lack of
motivation of workers as a result of the unwillingness of management to delegate authority,
the lack of communication skills and proper training of workers, and the high turnover of
workers with minimum loyalty to the firm. These conditions, they stated, tended to create
apathy and negligence and an “it can™t be done” mentality in the workforce.
Source: Crevoshay (1994: 11)
Developmentalist theories of economic development 151
spin-off effects for other industries. Hirschman felt that innovative attitudes toward efficiency
and responsibility on the job would also be transmitted to society at large. A system built
upon merit and performance would eventually threaten the outmoded social structure built
upon privilege and ceremony, a system which far too often remains a source of inefficiency
in the less-developed world.

Antagonistic growth
Like all the developmentalist pioneers, Hirschman was extremely optimistic about the
possibility for progress in the less-developed world in the 1950s. In his Strategy of Economic
Development (1958: 5), he stated that:

development depends not so much on finding optimal combinations for given resources
and factors of production as on calling forth and enlisting for development purposes
resources and abilities that are hidden, scattered, or badly utilized.

In a self-review of his own work in the 1980s, Hirschman struck a more sobering note. He
defended his argument that development via excess capacity, or unbalanced growth, could
be a viable strategy, while acknowledging that problems arising from resource scarcity also
need to be given a more central role in conceptualizing the development process. Under
conditions of resource scarcity “ and all less-developed nations face scarcity, be it of invest-
ment funds or of skilled labor “ an over-emphasis on a certain sector, such as industry, can
mean that another sector, especially agriculture, fails to receive the inputs and support it
needs to progress at a reasonable or desirable pace. Thus unbalancing development in one
sector can leave another sector worse off, leading to what Hirschman termed an antagonistic
growth process. In such a situation, Hirschman warned, further economic growth along the
same lines will serve only to exacerbate existing levels of economic inequality. And this, of
course, can lead to difficult if not explosive political struggles. So, both efficient allocation
and effective reallocation of resources must be considered at the same time. It is not one or
the other that is most important for development.

Growth with unlimited supplies of labor
Another of the most important pioneers of early development economics is Sir Arthur W.
Lewis, who, along with Gunnar Myrdal, is one of only four development economists to have
been awarded the Nobel Memorial Prize in Economic Science. Lewis™s most cited work, and
one of the best-known models in development economics, is his classic article on unlimited
supplies of labor (1954). From 1970 to 1974, Lewis, who was born on St. Lucia in the Carib-
bean, was President of the Caribbean Development Bank, having previously held high-level
positions at the UN in the area of development policy.
Like the other developmentalists discussed in this chapter, Lewis was quite optimistic
that hidden reserves of strength could be tapped in the less-developed nations, and that by
doing so, economic development could rapidly be promoted. He also shared their convic-
tion that industrialization was the route the less-developed nations needed to pursue to
escape poverty and reach a higher level of economic and social progress. Lewis™s reasons for
supporting industrialization, however, were quite distinct. He was not an “export pessimist.”
In fact, Lewis produced a major research work (Lewis 1969), the purpose of which was to
demonstrate that tropical products and raw materials exports, the traditional primary exports
152 The Process of Economic Development
of less-developed nations, were not subject to falling international prices resulting from
supposed limits of the advanced nations to absorb these products. On the contrary, he drew
the conclusion that rising incomes and rising levels of production in the already-developed
nations would call forth a stronger demand for tropical products and raw materials. Thus, the
promotion of such exports promised higher levels of export income in future.
Despite this latent potential, Lewis nonetheless insisted that the wage level of the less-
developed nations was moving upward at much too slow a pace; workers™ incomes in the
less-developed nations were falling further behind that of their counterparts in the developed
nations. Lewis believed this growing disparity was the result of differences in the productive
structures existing between the two areas. The already-developed nations had large indus-
trial and manufacturing sectors, where many workers were employed, and relatively small
agricultural sectors, using a relatively small proportion of the labor force. Just the reverse
structure prevailed in the less-developed nations, where most of the labor force was occupied
in rural areas, with agricultural production their primary activity.
Higher wages were paid to workers in the manufacturing sector compared to agricul-
ture in both the developed and less-developed nations, though the gap was smaller in the
already-developed nations, because productivity per worker was higher in both the industrial
and agricultural sectors. especially low incomes prevailed in the agricultural sector of the
less-developed countries, where most of the population lived and worked, since output per
worker also was quite low, primarily because of the lack of capital and the relatively primitive
technologies in use. Thus the higher average income in the already-developed nations was a
structural function of having more workers in the higher-productivity, higher-wage industrial
sector relative to the less-developed regions.
While the less-developed nations often were portrayed by economists and policy-makers
as having a comparative advantage in the production of tropical agricultural goods and raw
materials for export that should continue to be exploited, Lewis suggested that they also
had a potential, hidden, dynamic comparative advantage in some types of manufacturing.
At the time, this was still a somewhat unconventional view as applied to the promotion
of manufacturing. It arose from his observation that wages in the manufacturing sector in
less-developed nations were relatively low compared to those of the advanced nations. Since
wages were an important component of costs in labor-intensive manufacturing processes,
such as textile production, if the less-developed regions could restructure their economies
toward this type of manufacturing, they could perhaps create comparative advantage based
on their relatively lower wage costs.12
Lewis actually was an “export optimist,” believing that the small net addition to global
manufacturing exports coming from less-developed regions would be easily absorbed by a
growing world market. Thus a higher level of manufactured exports from the less-developed
nations need not spark a defensive reaction in the advanced nations in terms of new tariffs
and other barriers, because Lewis believed that the increases in labor-intensive manu-
facturing exports added by the less-developed nations to total exports would be dispersed
throughout the global economy. He thus did not believe that any developed country would
face a serious threat from manufactured export competition coming from the less-developed
regions. Consequently, the developed nations would not resort to protectionism to stop the
flow of new manufacturing exports.
Lewis wanted to advocate shifting labor away from agriculture and into industry. But, as
a well-trained orthodox economist, he had been taught that switching labor from agricul-
ture to industry would mean that agricultural production must surely decline with such a
reallocation, assuming the marginal product of labor in agriculture to be greater than zero.
Developmentalist theories of economic development 153
Consequently, food prices might be expected to rise, as fewer farmers would be producing
less output for a growing number of non-agricultural workers.13 With rising food prices,
industrial wages would need to rise to ensure at least a subsistence wage, and the potential
comparative advantage of the less-developed country in producing labor-intensive manufac-
turing goods for export would disappear with the rising wages. Was there no way out?

Surplus labor
It was at this point that Lewis brought into development economics an important construct
which had been widely utilized by Keynesian economists in analyzing the Great Depression
(1929“39) within the industrial countries in writing about disguised unemployment. What if
labor in the agricultural sector was being utilized in an extremely inefficient manner, to the
degree that, by taking agricultural workers out of this sector and employing them in industry,
agricultural production would not decline at all, while industrial output was increased with the
influx of greater employment? What if there were actually a surplus of agricultural workers,
such that by transferring some labor from agriculture to industry the remaining workers could
work longer hours, or more efficiently, and total agricultural production could remain constant
or even rise?14 Or, alternatively, agricultural producers who had been selling to the export
market could replant their fields with an eye to the potential profits created by the growing
domestic market resulting from the process of industrialization. In any event, Lewis reasoned,
if there was surplus labor in agriculture, then that “hidden reserve” could be tapped for indus-
trialization, and development perhaps would not prove to be so difficult to attain after all.
If industrialists were to pay a wage somewhat, say 30 percent, above the average wage
prevailing in agriculture to cover the costs and discomforts of migrating to industrial areas
and to compensate for the higher cost of urban living, then industrialists could hire all the
labor they might want at a constant wage, as long as surplus labor conditions prevailed in
agriculture. Industrialists could look forward to a double advantage. First, the absolute level
of wages would be above but close to subsistence, yet domestic wages would be far below
the wage prevailing in the advanced nations. Second, as industry shifted to higher and higher
levels of production over time, more and more surplus agricultural laborers would be brought
into the industrial sector. But wages in that sector would not have to rise at all, because the
cost of food, the basic determinant of the wage level, would remain constant until the labor
surplus was exhausted.

The Lewis surplus labor model
We can formalize the Lewis model along the following lines. Lewis presumed that the typical
less-developed nation was dualistic, not only in having two key sectors, but in the sense
that these sectors had little interconnection. There was a traditional, low-productivity rural
and predominantly agriculture sector, where the great bulk of the population worked and
produced what it consumed. But there also existed (or there could be created) an incipient
modern capitalist sector, where production was more technologically driven and, accordingly,
worker productivity was higher than in the traditional sector. The modern sector bought food,
and perhaps other inputs, from the traditional sector for use in the production process, and
the traditional sector provided labor to industry in the cities, but otherwise the links between
the two sectors were weak.
It was in the labor supply link between the two sectors that Lewis found a transformation
dynamic. His model can be explained by examining Figure 5.1(a) and (b). Figure 5.1(a)


Units of labor
in agriculture


Figure 5.1(a) Lewis™s surplus labor model: agriculture.



1 2
w w

w w

Units of labor
0 in industry
I1 I2
L1 L2

Figure 5.1(b) Lewis™s surplus labor model: industry.
Developmentalist theories of economic development 155
shows the marginal product (MPL) and average product (APL) of labor curves in agriculture.
Since Lewis assumes a surplus of labor in agriculture, it can be presumed that the MPL = 0,
so that LA workers are employed in that sector. However, unlike the usual neoclassical
assumption that workers are paid their marginal product, which, in this case, would mean
agricultural workers would be paid nothing “ clearly an impossibility “ workers actually
receive a wage, wA, equal to their average product when LA workers are employed. Why? In
the traditional sector, it is presumed income is shared by the members of extended families.
One can think of the production process being organized around the household, rather than
by and for individual decision-makers. Work often is done collectively on family farms,
where the marginal product calculation of the optimal use of labor inputs would be a wholly
alien concept.15 All family members may contribute to production in their own fashion; and
all share in the fruits of the labor process more or less equally, regardless of the individual
contribution to production.
If the industrial sector pays a wage wI that is above wA, then labor will be attracted from
agriculture to industry (Figure 5.1(b)). Industrial capitalists, who are presumed to be profit-
maximizers, will hire LI1 workers: additional labor will be used until the industrial wage is
equal to the MPL1 in the industrial sector. The industrial sector™s total output is equal to area
0M1P1LI1; the workers™ share of that income is equal to area 0wIP1LI1, while the capitalists™
share is area wIM1P1.
When this profit or surplus is reinvested, in whole or in part, depending on other costs
and considerations, the addition of new physical capital, and the technology embodied in
that capital, will shift labor™s marginal product curve in the industrial sector upward and
outward, since the effect of more capital and more technology is to increase the produc-
tivity of labor. With increased investment, and hence the new MPL curve, M2PL2, and given
the labor surplus which keeps the industrial wage at wI, LI2 workers now will be employed
in the industrial sector. Thus, with continuing reinvestment from the profits of the modern
sector, the transfer of labor from agriculture to industry is accomplished in the Lewis model,
especially if the capital“labor ratio in industry does not rise very much “ that is, as long
as production remains labor-intensive. employment in industry will rise, along with total
national output.
The process of transferring labor from agriculture to industry will slow and eventually
come to an end, of course. As labor leaves agriculture, the marginal product of labor, and
its average product, must eventually rise as the labor surplus is exhausted. In Lewis™s view,
the upward pressure this puts on wages in agriculture will force producers in that sector to
become more productive via the adoption of better technologies, thus forcing the “moderni-
zation” of the primary sector as well. Of course, this process happens gradually, rather than
discretely, as some agricultural producers embrace modern methods of production earlier
than others, but the effect of a growing scarcity of labor in agriculture will require the use of
greater amounts of capital and technology to save on labor. In the process, productivity and
incomes in agriculture also will rise.
To keep this “virtuous circle” of labor transfer going once started, there would have to
be more and more capital formation in manufacturing capacity, which would necessitate a
higher level of savings which then could be transformed into investment. In Lewis™s view,
only the fledgling capitalist sector would save. Large landowners, monopoly bankers, mine
owners, and other wealthy strata of traditional society, including the political elite, would be
more likely to squander their economic surplus in ostentatious consumption and/or capital
flight out of the country. Only by increasing the share of national income which accrued to
industrial capitalists, Lewis reasoned, could the less-developed regions move forward, and
156 The Process of Economic Development
this would be accomplished by the transfer of labor from agriculture to industry, shown in
Figure 5.1(b).

The distribution of income
As Lewis defined the problem, his surplus labor model suggested a very rapid dynamic
process would unfold. A significant and rapidly rising share of national income would be
shifted to the national, industrial capitalist strata. As this class increased its productive
investments by plowing back its profits into new investments in pursuit of ever-greater profit,
total national output would rise. But since wages would not rise as long as the labor surplus
remained, a growing share of a growing total national income would accrue to the capitalist
class. They, in turn, motivated only to increase production and profits further, would reinvest
at an accelerated rate, thereby ensuring that national income would rise further. A perpetual-
motion machine would be put into play, moving faster as time went on.
What would happen, though, when the unlimited supply of laborers was finally depleted?
Lewis was unconcerned. At that point, the objective of the transformation of the economy
would have been achieved. Wages for workers would rise, the standard of living would
improve, and the gap between the poor and rich nations would have closed considerably.
This Lewis saw as the inevitable and desirable end of the process he envisioned.
Lewis often has been accused of advocating a worsening of the distribution of income
as a means to promote development. In his model, the share of income going to relatively
well-off capitalists rises over time, as can be seen in Figure 5.1(b). Meanwhile, a small gap
opens between the average wage in the agricultural sector and that of the industrial sector, to
the disadvantage of agricultural workers, whose income remained relatively stagnant as long
as there was surplus labor. Lewis was well aware of this criticism, but he felt that it missed
the mark. Painful as it might be to contemplate, Lewis found no other way to foster growth.
He pointed out that he was not advocating a worsening of the distribution of income. What
he was advocating was economic development and a general rise in the standard of living,
and he could see no other way to exploit the “hidden surplus” of disguised unemployment in
agriculture without such an adverse, but temporary, increase in inequality between agricul-
ture and industry (see Focus 5.3).16
Joan Robinson put the matter well in another context: “The misery of being exploited by
capitalists is nothing compared to the misery of not being exploited at all” (Robinson 1966:
46). In other words, the wage of the industrial worker is likely to be higher, and the standard
of living better, than for the rural peasant or rural worker. The transition from agricultural
poverty to a higher standard of living in industrial production was desirable, even if it engen-
dered temporary inequality. One suspects that Lewis would have agreed with Robinson™s
pithy comment.
Lewis subsequently broadened his definition of what contributed to a labor surplus, or
disguised unemployment, to include:

1 individuals unemployed because of technical change in agriculture and industry;
2 the underemployed in rural areas;
3 the movement of women from the household to the labor force;
4 the surplus labor generated by rapid increases in population.

Of these factors, he considered the last the most powerful force for creating a labor surplus
(Lewis 1984: 133).17
Developmentalist theories of economic development 157

The Lewis model of unlimited supplies of labor is but one, though perhaps the best known,
of a number of dualist models examining the transfer of labor from agriculture to industry.

The Fei-Ranis model: This was an extension and elaboration of the basic Lewis model,
a major distinction being an upward-sloping labor supply curve after some amount of
labor has been extracted from agriculture to industry. The Lewis model also envisages
such a tendency, though it is not formalized as it is in the Fei-Ranis model. This was a
theoretical contribution, and it does not detract from the basic Lewis conclusion that
labor will be attracted from rural areas to urban industrial centers by a wage higher
than that paid in agriculture, and that it is that transfer which contributes to the desired
structural transformation of production (which will be examined in Chapter 9).

The idea that total national output and income per person could be increased by such
a strategy of labor transfer, where the marginal product of labor in industry exceeded the
marginal product of labor in agriculture, is the key insight of both the Lewis and Fei-Ranis

The Harris-Todaro model: Though traditionally the Harris-Todaro model is not catego-
rized as a dual economy model, but rather one explaining rural“urban migration and
urban unemployment and underemployment, it is in the tradition and spirit of the
Lewis model. The Todaro model envisages workers in agriculture rationally choosing
to migrate to urban, industrial centers in the pursuit of wages that, with some prob-
ability, are expected to be higher than in rural areas. Urban wages are higher on
account of many forces: higher productivity of workers, higher living costs of cities,
and, perhaps, the “wedge” of unionization pushing wages above the market-clearing

As a consequence of a “too”-high differential between urban and rural wages, too
many migrants continue to be attracted to urban areas relative to the number of urban,
industrial jobs available in the formal sector. As a consequence, many migrants are forced
into informal sector employment in low-productivity, low-income jobs: domestics, street
vendors, beggars, jugglers, newspaper vendors, day laborers, and so on.
Further, urban slums emerge and grow, as the urban formal sector wage wedge relative
to agricultural wages continues to draw migrants hoping to find formal sector employment.
Thus the disguised unemployment or underemployment of some rural migrants becomes
open unemployment, or the disguised underemployment of the urban informal sector.
Sources: Fei and Ranis 1964; Harris and Todaro 1970

Utilizing the economic surplus
Although Lewis is best known for his article on unlimited supplies of labor, he later took
a somewhat different approach to the development problem. He felt that more attention
should be paid to the inordinately high level of consumption as a share of total income in
the less-developed regions. Much of the income generated in less-developed countries was
squandered in conspicuous consumption, and not just by capitalists, who were too small in
number to have much of an impact and who, in any case, were presumed to save and invest
much of their income in pursuit of future profit. Rather, it was residual classes, like land-
owners and plantation barons, as well as new groups such as financial manipulators and a
158 The Process of Economic Development
political elite who were skimming off part of aggregate income that could have been used
for investment, who engaged in superfluous consumption (Lewis 1976: 257).
In order to reduce this waste, Lewis advocated raising the tax burden on the top 10, or
perhaps the top 20 percent of income recipients to the point that government would receive
20 percent of national income. The state, in turn, would devote roughly 60 percent of those
revenues, or 12 percent of national income, to basic public services, such as schools, hospi-
tals, social security, and so on, and 40 percent of tax revenues (8 percent of national income)
to public capital formation or social overhead capital. Thus, the mature version of the Lewis
model should include two forms of investment that would be promoted: private-sector
investment in manufacturing and industry deriving from the private capitalist, and public
investment in social overhead capital, such as roads, communications systems, energy, and
so on, deriving from government decisions as to priorities. Of the two, Lewis felt that the
role to be played by the state in taxing the unproductive elite and allocating national income
to socially productive purposes to be far more important in future.
By 1984, Lewis had determined that political and economic matters could not be sepa-
rated. Development was as much a matter of “getting public policy right,” as of providing a
constructive environment for the private sector, a view with a strong Keynesian resonance.
Achieving development, Lewis seemed to say, was as much a question of political will as
it was of finding the technical means. Sir Arthur continued to advocate a large increase
in savings and investment, but he also emphasized that the only way to achieve this was
to reduce the unproductive and wasteful consumption levels at the top of the income
pyramid, especially of unproductive classes who continued to control significant economic
and political resources.

Nowadays in most underdeveloped countries people know what economic growth
requires; the difficulty is to make available the quarter of the national income which it
costs. Personal consumption which should only be 75 per cent of the national income is
nearer 85 per cent, leaving for the public services and for capital formation together only
about 15 per cent instead of the 25 per cent they need.
(Lewis 1984: 256)

The legacy of the Lewis model
The Lewis model has continued to have an important influence in development economics
(we shall use it again in Chapter 9). Subject as it was to a great deal of critical scrutiny, it
is not surprising that many objections were raised. Most telling were two. First, the model
ignored institutional factors which influence the level of wage determination in the industrial
sector. Governmental labor standards, including minimum wages, and unions are absent
from the model. In fact, many less-developed nations have introduced relatively advanced
labor legislation, and unions often have been able to negotiate a wage far above that deter-
mined by the free play of market forces.
Many of these institutional factors were introduced by, or were a reaction to, foreign
multinational firms in mining and agriculture. These firms could easily afford the increase
in their costs which would improve working conditions. However, via “target bargaining,”
such improved conditions can quickly become the bargaining norm for unions and workers
in other industrial sectors not linked to the multinationals. The end result often has been to
ratchet up wages for those workers in the industrial sector who have permanent jobs (i.e. who
are not hired on a day-by-day or per-job basis). Thus rather than a constant industrial wage
Developmentalist theories of economic development 159
with some premium above the agricultural wage, industrial workers in some countries have
been able to achieve substantial increases in wages, thereby eroding the potential comparative
advantage in wage costs and undercutting Lewis projections for economic development by
reducing the absorptive capacity of the industrial sector.
A second major objection concerned the socially virtuous behavior which Lewis assumed
the capitalist strata would engage in, that is, the continued reinvestment of earnings in new
production. Some have argued that the native capitalist strata may short-circuit the growth
process through capital flight, rather than plowing profits back into production; of course
that possibility certainly exists. Lewis assumed that capitalists would have a high propensity
to reinvest and that their earnings would not leak out of the country via capital flight or via
the conspicuous consumption of luxury imports. Whether capital flight or reinvestment take
place, however, is not something that can be assumed. Governments interested in promoting
economic and social development can help to create an internal economic environment attrac-
tive to domestic investors, particularly by keeping the inflation rate relatively low and stable.
There are other aspects of the internal balance that are important, but there is no guarantee that
capitalist profits will be reinvested, especially in an increasingly global capitalist economy.
There can be no question that capital flight played a major role in many less-developed
nations in the 1980s, often contributing to an external debt crisis. However, to criticize the
Lewis model in this context would appear to be inappropriate, for two reasons. First, Lewis™s
model was developed in the early 1950s, when international currency and financial markets
were in a shambles, and most nations maintained strict currency controls. Second, given
Lewis™s strong advocacy of governmental intervention in order to tap the economic surplus
of high income recipients, it is doubtful that Lewis would oppose the reinstitution of currency
controls to block capital flight. Nations, such as Brazil, that imposed currency controls in the
1980s suffered relatively little capital flight.

Stages of growth theory
The last developmentalist theory to be examined in this chapter is Walt Whitman Rostow™s
stages of growth analysis (1960). Rostow™s writing on the economics of what he called the
“take-off” into sustained growth quickly became influential, in large part because of his
remarkable ability to use metaphors, such as that of the take-off, and his deft compaction of
european economic history. Like Marx before him, Rostow sought a universal interpretation
of history, and this he provided in his stage model. He argued that all nations pass through
five phases: the traditional society, the preconditions for take-off, the take-off, the drive to
maturity, and the age of mass consumption. Rostow built his theoretical analysis upon the
history of Britain, as had Marx. In doing so, he utilized a framework that most economists
and other social scientists knew quite well. The plausibility of his argument seemed to many
to be well-anchored in historical dynamics, both because it seemed to fit quite well the British
experience, which had long been the basis for countless generalizations in economics, and
because many economists did not have a ready grasp of the economic history of the less-
developed regions. Thus, the model projected by Rostow seemed to generally conform to
what many economists knew, or at least believed, to be true.

Stage 1: traditional society
In defining his first stage of historical and economic development, Rostow was rather vague.
He likened traditional society to that of medieval europe, and more broadly to any society
160 The Process of Economic Development
that was pre-Newtonian. That is, traditional society was pre-scientific. Scientific progress
might occur from time to time, but there was no systematic mechanism which led to the intro-
duction of scientific knowledge into the production process on a continual basis. Traditional
society was dominated by a perspective which Rostow defined as “long-term fatalism.”
According to his formulation, traditional society was predominantly agricultural, with land-
holders playing a dominant role in the determination of political and economic power.
Although Rostow attempted to demonstrate a general theory of historical stages, it would
seem his sketch of traditional society best fits Europe prior to the sixteenth century, during
its feudal period. He made virtually no effort to extend his analysis to the Third World.
Were these vast regions, from 1500 to 1800, similar in any meaningful way to Europe, circa
1400? Certainly the information surveyed in Chapter 3 on colonialism demonstrates that
this vast region was not traditional or unchanging “ on the contrary, the changes imposed
by colonialism were revolutionary. Rostow neglected to forge the link between traditional
(European) society and the societies in the Third World “ because none exists.

Stage 2: the preconditions for take-off
After his brief sketch of traditional society, Rostow moved forward to the second stage, the
preconditions for take-off. Here, under one stage category, we find two processes at work:
the beginnings of a sweeping destruction of traditional society, and the gathering of societal
forces which will propel it forward into the subsequent take-off stage. But in his emphasis on
the destruction of traditional society, usually through an outside source, probably colonialism
according to Rostow, he blurred the line between a nation which becomes colonized and
the colonizing nation itself. The presumption is that both colonizer and colonized are swept
forward through this stage, both benefiting from events which stimulate development. But,
in the case of the colonized nations, Rostow fails to entertain the likelihood that the process
of destruction will be so thorough that the colonized society will be set on a path that does
not lead to take-off, but to stagnation.
Thus the processes of both entering and leaving a major transformational epoch are
commingled in the same stage, without a detailed analysis as to how these two processes
unfold. Rather than doing so, Rostow presents a shopping list of changes which he expects
to arise during this stage, without much apparent regard to either causality or sequence. He
states that new types of entrepreneurs and managers will appear in the private and public
sectors, banks will appear and investment will increase, particularly in infrastructure.
Modern businesses will be created which will make use of new and sophisticated methods
of production. As this process unfolds in the colonial or post-colonial regions, “reactive
nationalism” sets the less-developed region on a new course, the drive for modernization
(see Focus 5.4).

Stage 3: the take-off into sustained growth
Our brief review of Latin America in the nineteenth century in Focus 5.4 suggests that neither
the role of “reactive nationalism” nor the existence of the profit motive appeared to be suffi-
cient conditions to launch Latin America into its take-off stage. Rostow does not explain the
movement from one stage to the next, and since he does not provide his reader with an inter-
pretation of the nineteenth-century economic history of Latin America that would support his
views, the Rostovian framework may be less universal than Rostow had hoped. Nonetheless,
given the importance of Rostow™s work in the field of development economics, it is useful
Developmentalist theories of economic development 161

In attempting to test Rostow™s hypotheses regarding the preconditions stage, the situation
in Latin America in the early nineteenth century would appear to be an important case.
Here we find, however, that the breaking of the colonial bonds did not lead to a full rupture
with the past. Factions within the new nationalist elite fought among themselves for polit-
ical control for another half-century, and then split into independent nations that mirrored
the separated colonial vice-royalties that had kept the colonies divided from each other
prior to independence. Moreover, the new nationalist elite classes were not interested in,
or were not capable of, transforming their newly independent countries along the path
that had been followed in Europe and the United States, that is, following a dynamic capi-
talist and industrial revolution. Rather, the goals of these new elite classes were relatively
limited. They wished to gain the class privileges Spanish colonial policy had for so long
reserved exclusively for pure-blooded European immigrants. Such a backward-looking
elite was content to continue the pattern of exporting primary commodities begun under
Spanish rule. This new dominant class of large landowners, merchants, and politicians
was certain to enrich itself through the expansion of such exports.
As we have noted in Chapter 3, throughout nearly the entire nineteenth century, raw
material prices soared and the terms of trade moved, perhaps fortuitously, in favor of such
products. Trade with the advanced industrial nations permitted the nationalist leaders to
import the manufactured luxury goods to which they aspired as emblems of their social
status. With easy access to vast reaches of land, much of it expropriated from the Catholic
Church and native Indians, the new nationalist elite was able to prosper by producing in
the same technologically backward manner while utilizing more land, that is, using exten-
sive forms of production.
Thus the Latin American elite bypassed one of the prime defining characteristics of the
Rostovian second, preconditions, stage; they were not forced to utilize the latest techno-
logical advances in an effort to make each unit of land more productive, that is, they did
not pursue intensive production methods. Contrary to Rostow, there was an obvious lag in
the development of Latin America™s essential infrastructure, such as banks, communica-
tions systems, and roads. And this, in turn, tended to reinforce the lag in the moderniza-
tion of the productive apparatus, that is, a delay in the introduction and use of machinery,
equipment, knowledge, and managerial strategies in tropical agriculture, mining, farming
and ranching, let alone in industry. It was in the period after 1870 that Latin America™s
pernicious pattern of limited export diversity was consolidated. In some countries this was
manifested by mono-export production.
Source: Dietz 1995: Chapter 1

to briefly analyze what he considers to be the key stage in the development process: the
take-off. This stage is defined as emerging under the following simultaneous conditions:

(1) a rise in the rate of productive investment from, say, 5 percent or less to over 10
percent of national income; (2) the development of one or more substantial manufac-
turing sectors with a high rate of growth; [and] (3) the existence or quick emergence of a
political, social and institutional framework which exploits the impulses to expansion.
(Rostow 1960: 39)

Furthermore, Rostow states that there must be a sweeping reallocation of resources devoted
now to:
162 The Process of Economic Development
building up and modernizing the three non-industrial sectors required as the matrix for
industrial growth: social overhead capital; agriculture and foreign-exchange earning
sectors, rooted in the improved exploitation of natural resources. In addition, they must
begin to find areas where the application of modern techniques is likely to permit rapid
growth rates, with a high rate of plow-back of profits.
(Ibid.: 193)

The take-off is to occur in the space of roughly twenty to twenty-five years. According to
Rostow™s dating, India began its “take-off” in 1952. Thus India, with a per capita income in
2000 of only $460, ranked thirty-sixth in terms of the poorest countries in the world, should
in fact have had by then a relatively strong economy. Yet in the period 1980“91, decades
past the presumed take-off stage, India™s per capita growth rate was a disappointing 0.7
percent per year (per capita growth rose to an impressive 4.2 percent per year in the 1990s).
Following the take-off, growth at rates well above the population growth rate was expected
to be the normal condition. The take-off into sustained growth had faltered in India, appar-
ently “ an event that the stage model cannot even consider.

Critical responses to the concept of the take-off
As intuitively appealing as Rostow™s list of conditions for take-off may be, it is discon-
certing to note that a number of development economists who have reviewed the histor-
ical record have found that the concept does not accord with the history of most of the
nations which have purportedly moved beyond take-off into “self-sustained growth.” For
example, Albert Fishlow argues that, in the now advanced nations, there was no major
abrupt jump in either the rate of investment or the rate of growth of output for most nations
(stage 3 above). Rather, there was a gradual speed-up in the rate of investment and growth
in most countries, and a sharp rise in investment and growth in only some (Fishlow 1976:
84“5). Simon Kuznets also argued (1971a) that a review of the economic history of the
now-developed nations showed no sudden significant rise in the rate of savings during
what might be considered their take-off stage. Kuznets further pointed out that when the
now-developed nations moved into the take-off stage, they did so at per capita income
levels much higher than those prevailing in the less-developed world currently (Kuznets
1971b: 224).
Gerald Meier elaborated on this point by concentrating his analysis on the agricultural
sector. He drew a contrast between the robust agricultural sector of the nations which went
through a take-off in the eighteenth, nineteenth, and early twentieth centuries, such as
Britain, France, Germany, the United States, Canada, and Australia, with the weak agricul-
tural sectors generally prevailing in the less-developed regions.

It is fairly conclusive that productivity is lower in the agricultural sector of underdevel-
oped countries than it was in the pre-industrialization phase of the presently developed
countries. Although direct evidence of this is unavailable, it is indirectly confirmed by
data suggesting that the supply of agricultural land per capita is much lower in most
underdeveloped countries today than it was in presently developed countries during
their take-off, and that there is a wider difference between per worker income in agricul-
ture and nonagricultural sectors in the underdeveloped countries today than there was in
the preindustrial phase of presently developed countries.
(Meier 1976: 95)
Developmentalist theories of economic development 163
Meier pointed to another important difference between the conditions prevailing in the
present-day less-developed regions compared to those that prevailed when the now-devel-
oped nations entered into their initial period of rapid development: population pressures
were relatively moderate in the past, whereas today an annual population rate of growth of
1.5“3 percent necessitates a much higher level of investment in order to move the economy
forward fast enough just to keep per capita income constant. That is, what would have been
considered a remarkably fast rate of aggregate economic growth during Britain™s industrial
revolution, 3 percent per year, is often the minimum rate of aggregate growth that must be
attained in many less-developed nations today in order simply to maintain the existing, low
standard of living per capita.
Furthermore, migration played a tremendous role in the economic performance of the
now advanced nations during their early industrial period. For some nations, like the United
States, Canada, and Australia, the influx of trained, ambitious young immigrants was a clear
economic boon. At the same time, the out-migration of young workers from Europe tended
to eliminate both potential unemployment and social problems that might have arisen from
structural unemployment. Lacking surpluses of labor, many of the now-developed nations
had a strong incentive to adapt new machinery and equipment which would dynamize the
productive process.
Criticisms of the take-off have continued to be published. Nicholas Crafts™ research
confirmed the analysis of Fishlow and Kuznets. On the basis of more recent work, he suggests
that we “discard Rostow™s linear model”:

Rostow™s notion of the takeoff seems to be completely discredited. GDP growth [in
Britain 1780“1830] exhibited at steady acceleration over perhaps half a century ¦ and
there is no sign of the rapid doubling of the investment rate postulated by Rostow. The
notion of a leading sector has also fared badly.
(Crafts 2001: 312)

Stages 4 and 5: maturity and high mass consumption
Rostow™s last two stages, maturity and high mass consumption, are defined sequentially as:

• a period wherein growth is sufficiently high so that there is significant increase in per
capita income. The economy becomes diversified and technologically sophisticated,
such that the society can now produce anything, but not everything, it chooses;
• a subsequent period where production is largely for the purpose of consumption, with
relatively little concern for the need to further build production capabilities. Society is
now devoted to the pleasures of consumer choice, the pursuit of security, and the enjoy-
ments of the arts and leisure.

Rostow™s legacy
In spite of the fact that it has been his fate to serve as a lightning-rod for criticisms
from virtually all schools of thought in development economics, Rostow clearly made
a powerful contribution. He forced other economists to review the experiences of the
now-developed nations and to demonstrate the tremendous gulf that exists between the
historical conditions which gave rise to the developmental success stories of the eight-
eenth, nineteenth, and twentieth centuries and the experiences with patterns of distorted
164 The Process of Economic Development
development, stagnation, and economic decay that prevail in the less-developed world
today. Rostow also opened the debate to another question in development economics. Did
colonialism lead to the entrenchment of backward socio-economic forces and processes
in less-developed nations which could not be displaced easily once political independence
was achieved? A full exploration of this matter will be left for the following chapter, where
we will review the work of a number of analysts who clearly argue that Rostow™s main
analytical error was to be found in his failure to incorporate the retarding and inhibiting
forces of colonialism into his model.
While today little remains of Rostow™s analysis which is of general use in the field of
development economics, Rostow was clearly a pioneer in opening up new areas for study,
debate, and analysis. Without his “big-picture” approach, many major issues might not have
received a critical airing. Furthermore, Rostow™s willingness to express his ideas within the
difficult terrain of political economy forced those who would refute him to consider a broad
range of factors at the analytical points of intersection of historical dynamics, political proc-
esses, and economic forces.

Questions and exercises
1 Contrast Nurkse™s “export pessimism” with Lewis™s views on development. In what
respects do their apparently contrasting views on exports actually coincide, and where
do differences remain?
2 In what sense would you argue that the economists discussed in this chapter formed a
school of thought? What ideas did they tend to share?
3 How can a fair test of Rostow™s stages model be formulated? Analyze the history of a
specific economy to see if such a test can be made.
4 Why and how did Hirschman argue that by putting things the wrong way around, by
actually creating disequilibrium, economic development could be promoted?
5 Why might unbalanced growth be easier, and less costly, for a poor economy to follow
than a balanced growth strategy?
6 Briefly explain the ideas of virtuous circles. Can you give two different examples of
virtuous circles that might affect a less-developed economy? Summarize the various
forms of positive external effects and virtuous hidden effects which Rosenstein-Rodan
utilized to argue that development could be achieved quicker than one might expect.
Can you speculate on what a “vicious circle” might be?
7 What did Lewis mean when he wrote that there was a surplus of labor in agriculture?
How does one measure that surplus? To what standard is labor in surplus, that is, in
surplus relative to what?

1 Rosenstein-Rodan became an influential policy-maker after the war. He held a top-level post
within the World Bank from 1947 until 1953, and from 1962 to 1966 he served on a key directive
committee of the United States-sponsored development programme for Latin America, known as
the Alliance for Progress.
2 This is the situation of increasing returns to successive inputs of investment, so that if investment
increases by x per cent, output rises by more than x percent. For an aggregate production function of
the form Q = f(K, L), where Q is total output, K is capital and L is labor, this means that both fK, fL >
0, but also that fKK, fLL > 0, that is, diminishing returns to the inputs to production have not yet been
reached. If one draws the aggregate production function, it will have both a positive and increasing
Developmentalist theories of economic development 165
slope. Interestingly, the possibility of increasing returns, which seems to go against the grain of so
much of both classical and neoclassical economic thinking and the law of eventually diminishing
marginal return, is one of the pillars of the new, endogenous theories of growth considered in
Chapter 8 that have become increasingly influential since the late 1980s.
3 Rosenstein-Rodan™s argument illustrates an important example of a larger phenomenon in
economics, called market failure. Whenever there is a divergence between private and social
benefits, as in this instance, and/or private and social costs, an unfettered market economy may
fail to produce the socially optimal level of output. What is desirable is to have the marginal social
benefits of any action equal to the marginal social costs, but private calculations of benefits and
costs may, and often do, differ from the social values. Basically, Rosenstein-Rodan was arguing that
the inability of any single private entrepreneur to appropriate all the social benefits “ in this case
profits “ of an action will result in an under-estimation of the total value of any private action. One
entrepreneur™s private investment decision, such as that of the steel firm, creates positive externali-
ties that accrue to other potential entrepreneurs, such as the metallurgy industry, and/or society in
the form of increased opportunities, higher demand, and lower costs that resulted from the decision
of another. Government intervention may be required in such circumstances, particularly when
many persons or firms are involved, if the social and private benefits and costs are to be equated,
and if the optimal and socially desirable level of output is to be reached.
4 The term “indivisibilities” was another of Rosenstein-Rodan™s favorites. Unlike neoclassical
economic analysis, which assumes that capital can be combined with labor in precisely optimal
amounts on the assumption that there is an infinitely divisible set of combinations of capital and
labor available, the concept of indivisibility is intended to illustrate production situations where
fixed, minimum amounts of capital (or labor) are necessary. A little less, and the product cannot
be produced. For example, in building a steel bridge, one cannot simply and infinitely substitute
labor for capital inputs and still produce the bridge; obviously labor cannot completely substitute
for capital and other inputs, like steel or bolts. The bridge will be engineered in such a way that
a specific amount of structural steel will be needed; an amount somewhat less and there will be
no bridge at all. Likewise in oil drilling, the drilling company either buys all of a drilling rig, or
none. It is not a divisible item. In general, social overhead capital tends to be of this nature. Often
Rosenstein-Rodan referred to the “lumpiness” of capital in this context.
5 Of course, government-sponsored projects can lead to over-investment or under-investment in
social overhead capital. The developmentalists did not naïvely believe that every action of govern-
ment was per se justified. If government does not employ transparent methods whereby officials
can be held accountable for their actions and their spending of public funds, then the government
itself can become one of the primary sources of social inefficiency. Without an efficient government
bureaucracy, the state itself often becomes an arena where individual fortunes are amassed through
the manipulation of public funds. Unfortunately, in many less-developed nations the most promising
avenue for upward social mobility lies within the governmental apparatus where accountability is
nearly non-existent and corruption is rife. This barrier to progress is one we shall have occasion to
comment on again later in discussing “economic rents” and the relative economic success of the
East Asian nations in recent decades.
6 This theme of “surplus” labor in agriculture is one that recurs again and again in the development liter-
ature. One of the leading theories of development, that of Sir Arthur Lewis, considered below, makes
this basic assumption central to the structural transformation required for economic development.
7 Nurkse is best known for his book Problems of Capital Formation in Underdeveloped Countries
(1953). His remarkable essay “Patterns of Trade and Development” (Nurkse 1962), which consti-
tuted an attack on the idea of trade as the “engine of growth,” was finished only a month prior to his
untimely death in 1959.
8 This adverse effect of a lower price and greater quantity will occur, assuming demand to be constant,
as supply increases if the demand for the good is price inelastic. In such cases, the larger quantity of
export sales will be insufficient to compensate for the lower price, and hence total export revenues
will decline.
9 The propensity to import is technically defined from the statement, M = mY, where M is the value
of imports purchased, Y is national income (GNP or GDP), and, m, which has a value 0 < m < 1, is
the “marginal propensity to import,” that is, it is the proportion of income that society chooses to
spend on imported goods and services. This proportion depends upon the level of average income,
the income distribution of society, and social and cultural factors.
166 The Process of Economic Development
10 In modern economic analysis, such consumption items are referred to as positional goods.
11 It is not sufficient to simply produce more to have economic growth; if the increase in output is to
be sustainable, it must find a market and be sold, or capitalist enterprises will stop producing.
12 This is simply an extension of the insight that was formalized early in the 1920s in the Heck-
scher-Ohlin theory of trade, which suggested that countries with an abundance of one factor of
production over another, would, with free trade, tend to export those goods using the abundant,
that is, relatively cheaper, input because that is where their comparative advantage would exist vis-
à-vis other nations. Thus less-developed economies, with their abundant labor and scarce capital,
could be expected to export those goods, be they agricultural or industrial, that were labor-intensive
in their production and, by the Stolper-Samuelson theorem of international trade, this would be
expected, over time, to lead to the equalization of income for the different factors of production
within and among nations, assuming free mobility of capital and labor and perfect competition.
13 In effect, as labor left agriculture, the supply of agricultural output might be expected to decrease as
the quantity of labor, L, in agriculture falls, while the demand for agricultural goods would, at best,
stay the same, and might even be expected to rise if workers in industry have rising incomes. Thus,
from simple supply and demand analysis, if the supply of agricultural output decreases (the supply
curve shifts inward), while the demand rises (an outward shift), the equilibrium price of agricultural
products must increase, given the assumptions.
14 In effect, what if the marginal product of labor, MPL, in agriculture, at the current level of labor
usage, is such that MPL ¤ 0? In such a case, extracting L from agriculture will not reduce agricul-
tural output, if MPL = 0, and will result in an increase in agricultural output if MPL < 0. As long as
the MPL in manufacturing > MPL in agriculture, a shift of labor from agriculture to manufacturing
will increase aggregate output.
15 In this instance, given wA, which can be interpreted as the subsistence wage, the optimal quantity
of labor to be employed in agriculture would be L O, which is clearly less than LA, the actual level of
employment with the household calculation of labor usage in which income is shared and average
income is distributed among family members.
16 Such an outcome in the transition from a surplus labor economy might be one explanation for the
Kuznets inverted-U hypothesis considered in Chapter 2, which also prognosticated a worsening of
income distribution with economic growth, up to a threshold level of per capita income, after which
the income distribution might be expected to improve.
17 Two very important theoretical models related to the Lewis model are the Fei-Ranis and the Todaro
models. These are briefly explained in Focus 5.3.

Crafts, Nicholas. 2001. “Historical Perspectives on Development,” pp. 301“34 in Gerald Meier and
Joseph Stiglitz (eds.), Frontiers of Development Economics. Oxford: Oxford University Press
Crevoshay, Fay. 1994. “Complejos, Confusiones y Autoritarismo, Barreras para la Aplicación de la
Calidad Total en M©xico,” El Financiero (21 abril): 11.
Dietz, James L. (ed.). 1995. Latin America™s Economic Development, 2nd edn. London and Boulder,
CO: Lynne Rienner Publishers.
Fei, John C.H. and Gustav Ranis. 1964. Development of the Labour Surplus Economy. New Haven,
CT: Yale University Press.
Fishlow, Albert. 1976. “Empty Economic Stages?,” pp. 82“9 in Gerald Meier (ed.), Leading Issues in
Economic Development, 3rd edn. Oxford: Oxford University Press.
Harris, J.R. and M.P. Todaro. 1970. “Migration, Unemployment, and Development: A Two-Sector
Analysis,” American Economic Review 60 (March): 126“42.
Hirschman, Albert O. 1958. The Strategy of Economic Development. New Haven, CT: Yale University
””. 1984. “A Dissenter™s Confession,” pp. 87“111 in Gerald Meier and Dudley Seers (eds.), Pioneers
in Development. Oxford: Oxford University Press.
Hoff, Karla and Joseph Stiglitz. 2001. “Modern Economic Theory and Development,” pp. 389“459
in Gerald Meier and Joseph Stiglitz (eds.), Frontiers of Development Economics. Oxford: Oxford
University Press.
Developmentalist theories of economic development 167
Kuznets, Simon. 1971a. Economic Growth of Nations. Cambridge, MA: Harvard University Press.
””. 1971b. “Notes on Stage of Economic Growth as a System Determinant,” pp. 243“68 in Alex-
ander Eckstein (ed.), Comparison of Economic Systems. Berkeley, CA: University of California
Lewis, W.A. 1954. “Economic Development with Unlimited Supplies of Labour,” Manchester School
of Economic and Social Studies 22 (May): 139“91.
””. 1969. Aspects of Tropical Trade. Stockholm: Almqvist and Wiksell.
””. 1976. “The Cost of Capital Accumulation,” pp. 256“7 in Gerald Meier (ed.), Leading Issues in
Economic Development, 3rd edn. Oxford: Oxford University Press.
””. 1984. “Development Economics in the 1950s,” pp. 121“37 in Gerald Meier (ed.), Pioneers in
Development. Oxford: Oxford University Press.
Meier, Gerald. 1976. “Future Development in Historical Perspective,” pp. 93“9 in Gerald Meier (ed.),
Leading Issues in Economic Development, 3rd edn. Oxford: Oxford University Press.
Murphy, Kevin, Andrei Shleifer and Robert Vishny. 1989. “Industrialization and the Big Push,” Journal
of Political Economy 97 (October): 1003“26.
Nurkse, Ragnar. 1953. Problems of Capital Formation in Underdeveloped Countries. New York:
Oxford University Press.
””. 1962. “Patterns of Trade and Development,” pp. 282“336 in Gottfried Haberler and Robert Stern
(eds.), Equilibrium and Growth in the World Economy. Cambridge, MA: Harvard University Press.
Robinson, Joan. 1966. Economic Philosophy. Harmondsworth: Penguin.
Rosenstein-Rodan, Paul. 1976. “The Theory of the ˜Big Push™,” pp. 632“6 in Gerald Meier (ed.),
Leading Issues in Economic Development, 3rd edn. Oxford: Oxford University Press.
””. 1984. “Natura Facit Saltum,” pp. 207“21 in Gerald Meier and Dudley Seers (eds.), Pioneers in
Development. Oxford: Oxford University Press.
Rostow, Walt. 1960. The Stages of Economic Growth: A Non-Communist Manifesto. Cambridge:
Cambridge University Press.
””. 1984. “Development: The Political Economy of the Marshallian Long Period,” pp. 229“61 in
Gerald Meier and Dudley Seers (eds.), Pioneers in Development. Oxford: Oxford University Press.
Toye, John. 2004. The UN and Global Political Economy. Bloomington, IN: Indiana University Press.
6 Heterodox theories of economic

after reading and studying this chapter, you should better understand:
• the importance of the distinction between the center and the periphery in structuralist
• the Prebisch-Singer hypothesis on declining terms of trade for primary product
exporters and the debate surrounding it;
• the role of import substitution industrialization (ISI) according to the ECLA econ-
omists and their subsequent critique of this policy;
• Ayres™s concept of inhibiting institutions and the importance of education and
technology to the institutionalist perspective;
• Gunnar Myrdal™s seminal ideas about spread effects and backwash effects as
examples of cumulative causation;
• the distinction between associated dependent development and the dependency
perspective of underdevelopment;
• Furtado™s structuralist perspective on dependency;
• Baran™s view of the equilibrium trap of underdevelopment;
• the distinction between stagnationist dependency analysis and the classical
progressive Marxist view of development as presented by Bill Warren.

This chapter discusses and analyses the ideas of economists and social scientists who have
broken with economic orthodoxy and moved beyond the framework of the developmentalist
economists considered in the previous chapter. These heterodox economists do not believe
that relatively minor changes in economic conditions, such as an increase in foreign aid or
a sudden increase in investment, will be sufficient to create the “big push” or the “take-off”
into sustained growth, as did the developmentalists. In fact, many of the heterodox econo-
mists would argue that such limited changes, within the context of the existing structures
and institutions prevailing in less-developed societies, might result in a strengthening of
backward socio-economic frameworks, consolidating adverse path dependence. For the
heterodox economists, the changes required to propel the development process forward are
more fundamental, more sweeping, and more profound.
Included among the heterodox grouping are a number of thinkers who hold no more in
common than the fact that they vigorously dissented from the general premises and propositions
Heterodox theories of economic development 169
of the developmentalist economists, and of orthodox economists, as well. In this chapter, then,
we will trace the ideas of the Latin American structuralists, represented by Raúl Prebisch
and Hans Singer; the institutionalists, represented by Clarence Ayres and Gunnar Myrdal; the
dependency school, represented by Paul Baran and Fernando Henrique Cardoso; and the clas-
sical Marxist approach in its more modern form, represented by Bill Warren.
The heterodox thinkers reached their maximum point of influence within the field of
development economics in the late 1960s and early 1970s.1 Indeed, the impact of some of
their views was so profound that their ideas exerted some indirect impact on the leading
development institutions, most notably the World Bank. In the 1970s, these institutions
embraced the basic needs (BN) approach, which is discussed in Chapter 17. As we shall see,
the heterodox thinkers had much more ambitious hopes for change than those embodied in
the limited goals of the basic needs approach.2

The Latin american structuralists
In 1948, as the result of a Chilean initiative, the UN agreed to form the Economic Commission
for Latin America, known best by its acronym, ECLA. Unlike the UN™s more technically
oriented Economic Commission for Asia and the Far East, created in 1947, or the Economic
Commission for Africa (1958), the ECLA was destined to become a center of advocacy for a
distinct Third World perspective and a hotbed of controversy. It was in and around the work
of ECLA that a Latin American school of structuralist economics was forged. The structuralists
argued that the less-developed countries of the periphery were structurally and institution-
ally different from the developed nations of the center in ways that made some aspects of
both orthodox economic theory and developmentalist theory inapplicable. In particular, the
Latin American structuralists were quite suspicious of the Ricardian theory of comparative
advantage and the alleged benefits of free trade among nations that supposedly derive from
specialization and trade.
In introducing the possibility of conflictive or adversarial relations in trade, the Latin
American structuralists challenged the “harmony of interests” assumption in market transac-
tions that had been a cornerstone of economic thinking since Adam Smith (see Chapter 4).
As Gabriel Palma explains, structuralism is concerned with the totality of a social system and
the many forms of interaction of the component elements within that system.

The principal characteristic of structuralism is that it takes as its object of investigation
a “system”, that is, the reciprocal relations among parts of a whole, rather than the study
of the different parts in isolation. In a more specific sense this concept is used by those
theories that hold that there are a set of social and economic structures that are unobserv-


. 6
( 21)