One
Industrial Development and Liberalization
Anne O. Krueger
In the past four decades analysts have learned a great deal about the
development process, not just from theoretical models but also from the experience of
developing countries as they have sought more rapid growth and higher living standards for
their people. Among these experiences, none has contained more lessons than that of Korea.
Until the dramatic turnaround in Korean policies starting in the early
1960s, few believed that sustained real growth rates of 10 percent or more were
attainable. Fewer still believed that a poor developing country, with about 88 percent of
its exports in primary commodities, could within a short period of time become a
competitive force to be reckoned with by producers in developed countries. And it was
entirely unthinkable that Korea might ultimately even challenge Japan.
A number of factors joined to bring about the dramatic Korean
transformation. Analysts may disagree on the relative importance and contribution of some
of these factors, but all observers would agree that the wisdom, judgment, and pragmatism
(not to mention the willingness to contemplate and implement bold strategies) of Korean
policymakers was an essential ingredient without which the miracle could not have been
achieved. Kim Jae-Ik was one of the foremost of these policymakers. It is to honor his
memory that this chapter is therefore devoted to an effort to place the Korean experience
with industrialization and liberalization in analytical context, and to attempt to
synthesize some of the lessons of that experience that are relevant for other countries.
I am indebted to Larry Krause and Nam Chong
Hyun for helpful comments on an earlier draft of this paper.
The first section of the chapter sets up a basic model of an efficient
development path against which policy alternatives can be evaluated. The model, partly
inspired by the Korean experience, was developed earlier (Krueger 1977) and is summarized
here. The second section then examines how policies may affect growth, resource
allocation, and the efficiency with which resources are used in light of the model. It
identifies the types of policies that are likely to prove inimical to sustained growth as
well as the sorts of policies that are conducive to it according to the theory. A key
question is how the shift from growth-inhibiting to growth-enhancing policies can best be
accomplished. Much of the process of shifting is, in fact, the process of liberalization,
and theory cannot provide much of a guide to it. The reasons for this, and the Korean
experience, are the subject of section 3. A fourth and final section then sets forth some
hypotheses about the interaction between liberalization policies and industrial
development in the course of economic growth.
The Development Process
The basic model of the development process used here is a two-sector,
multicommodity, three-factor model of an efficient development path. It combines the Lewis
notion of development with a model of industrial development founded in traditional
Heckscher-Ohlin trade theory. With Arthur Lewis (Lewis 1954), the process of efficient
development is regarded as that of transforming the country's economic activity from being
largely rural and agricultural to being largely industrial and more urban. [1] For ease of
exposition, agricultural and rural are used synonymously throughout, as are industrial
and urban, although in principle (and sometimes in practice) there is no reason why
workers cannot change from working in agriculture to working in industry without moving to
an urban area.
In this section, focus is on the way development would proceed with an
efficient allocation of resources, which is assumed to take place. In section 2, the model
is used to investigate the effects of distortions, either policy-induced or exogenous. In
an important sense, the removal of distortions is the process of liberalization.
For simplicity, output of the agricultural sector is assumed to be
produced with labor and land. [2] Output of the industrial sector is produced with capital
and labor. Thus, although there are three factors of production, capital and land are each
specific to a particular sector and labor is the mobile factor of production. [3] All
savings is allocated to capital accumulation, and labor is allocated between agricultural
and industrial activities so as to equate the wage between the two sectors. [4]
There are any number, n, of possible industrial products, all of
which are assumed to be tradable on world markets. Their relative border, or
international, prices are therefore given through international trade. Because of this,
all producers of tradables face perfectly elastic demand curves and there can be no
monopoly profits.
In contrast to the assumption that there are many commodities produced
by the industrial sector, agricultural output is assumed to consist of a single
homogeneous commodity. When a high proportion of the total labor force is employed in
agriculture, it follows that the economy-wide wage is effectively determined by the
agricultural wage. At early stages of development, this is typically the case, and hence
the economy-wide wage can be thought of as being initially determined in the agricultural
sector.
Given international prices for industrial goods and the rural wage, the
determination of what industrial commodities will be produced under an efficient
allocation of resources is straightforward. Under the usual Heckscher-Ohlin assumptions
(including constant returns to scale, diminishing marginal product of each factor, and so
on) the many industries within the industrial sector can be uniquely ranked from
most to least labor-intensive. That is, the profit-maximizing or
cost-minimizing labor-capital ratio in each industry can be ascertained and the ranking of
industries by factor intensity will be the same for all wagerental ratios.
Under these circumstances, where capital will earn the highest return
will depend on the relative price of labor in the country: for the lowest-wage countries
(presumably those with the lowest marginal productivity of labor in agriculture, which is
the same thing as the highest man/ land ratios if all land is of equal quality), at the
early stages of development an efficient allocation of resources will imply that the
country will produce the most labor-intensive industrial goods, using very laborintensive
methods.
As additional savings permits an expansion of the industrial sector,
capital will continue to be allocated to the most labor-intensive industries, and labor
will migrate from rural areas to work with the capital. Should the agricultural wage
remain constant in the face of this migration (the pure Lewis case), factor-proportions in
the industry will remain unchanged as the industry expands in proportion to the rate of
increase of capital. If, however, outmigration from agriculture is rapid enough (and/or
productivity increases in agriculture outstrip population growth) so that the rural wage
rises, capital-intensity will increase in existing industries.
If the real wage keeps rising (at constant relative world prices), there
will come a point where it will be economically efficient for the next more
capital-intensive industries to begin operation and for the most labor-intensive
industries to close down. In the simplest version of the model, there are thus two
discrete stages of industrial development: a stage during which the country's manufactured
output is specialized in the existing industry mix and, with capital accumulation, the
industry expands, but possibly becomes more capital-intensive; and a stage at which the
industrial composition of output is shifting toward more capital-intensive activities.
During this stage, the real wage would remain constant and continuing capital accumulation
would imply a Rybczynski-like shifting of resources toward the more capital-intensive
industry and a contraction of the older, more labor-intensive industries. In reality, of
course, the existence of sunk costs in existing factories and transport costs (so that
c.i.f. and f.o.b. prices differ, thus conferring a margin of natural protection that
permits some change in domestic price) implies that the process will not be as
discontinuous as a strict interpretation of the model would suggest.
Among the properties of this model in its simplest form, two are of
particular interest for interpreting Korean development. First, there may be large
differences in initial conditions among developing countries. Some may be land-rich, and
thus have relatively high wages and a comparative advantage in somewhat more
capital-intensive goods even at early stages of development. Others, which are land-poor,
may have very low wages for a considerable period of their development. [5] There may thus
be rich developing countries (Argentina), poor developing countries (Korea in the 1950s),
rich developed countries (the United States) and poor developed countries (Japan). To be
sure, rich-poor and developed-developing are both continua, and any cut-off point between
them would be somewhat arbitrary. Nonetheless, in this framework there are at least two
dimensions to consider in assessing a country's development strategy.
Second, the model has some interesting implications for the evolution of
the pattern of trade, especially for a country that embarks on rapid development from a
highly labor-abundant initial situation, as did Korea. This may be regarded as the
economics of catch-up, or of growth sufficiently more rapid than that of the rest of the
world so that the country is "overtaking" countries initially higher up the
capital-intensity scale. On one hand, any catching-up country will experience rising
industrial and falling (because of assumed positive marginal product of labor with no
technical change) agricultural output. Thus, such a country would gradually shift from
being a net exporter of agricultural output to being a net importer. On the other hand, it
should also be noted that the determination of which goods are produced within the
industrial sector is independent of whether they are imported or exported. Indeed, in this
model of development, one would expect that new industries as they developed would at
first be import-substitutes, but as their output expanded they would become exportables,
and then later contract yet again and become import-competing before their costs became so
high as to make them uneconomic. Nonetheless the relative Special-
ization implied by the model does suggest that countries will tend to
be net exporters of a significant portion of their industrial outputs.
Further implications, especially with regard to the effects of
policy-imposed distortions and the benefits of liberalization, can be drawn. Before
turning to them, however, consideration should be given to some important factors omitted
from the model. These are human capital, entrepreneurship, and technology. [6] Although
their formal incorporation into the model is well beyond the scope of this chapter, each
has been sufficiently important in development, and especially in the Korean experience,
to merit attention as to how they relate to the model.
The importance of the accumulation of human capital is by now widely
understood. It is clearly an essential and important component of the growth process. More
highly educated and trained labor can be incorporated into development models in a variety
of ways: one can regard more educated workers as representing more efficiency units of
labor than do less educated workers; one can regard human capital as a separate factor of
production; one can model human and physical capital formation as two alternative uses of
savings with rates of return to each augmenting income streams of the capital owners (see
Kenen 1965 for an interesting model of this type). The first view is readily incorporated
into the model presented above: incomes would rise not only as workers moved to industrial
activities and the real wage rose, but also because the effective supply of labor
increased. In this view, the model's implications are entirely unchanged, although the
rate at which a country might "catch up" might be slower the more rapid the rate
of human capital accumulation.
The second interpretation would require adding a third factor of
production to either or both of the two sectors. Incorporating human capital as a specific
factor like physical capital used in industrial activities would not significantly alter
the model unless some commodities were more intensive in the use of human capital per unit
of output than were others. If, as seems plausible, this latter is the appropriate
interpretation, the straightforward ranking of industries by labor intensity is no longer
possible. While a more complex ranking can be technically derived, its intuitive
interpretation is no longer as clear. Clearly, for laborabundant, low-wage countries, one
would anticipate that industrial expansion would occur first in industries that were
unskilled labor-intensive relative to their use of either physical or human capital.
Based on the experience of rapidly developing countries, one might
conjecture that "catch-up" would first proceed in the more
physical-capital-intensive industries, leaving industries with high human-capital inputs
to a later stage of development. But careful scrutiny of the results that might arise from
formal derivation of such models of empirical analysis is beyond the scope of this essay.
To ignore the role of human-capital formation in the formal model, however, should not be
interpreted to imply that it is not important in the development process.
Consideration of entrepreneurship presents many of the same challenges
as does human capital. If some activities are more entrepreneurially intensive than
others, and entrepreneurship is another factor of production, any ranking of industries
would entail the same difficulties as does human capital. Worse yet, consideration of how
a "supply of entrepreneurship" changes over time would represent a formidable
task. A more promising alternative appears to be to regard entrepreneurship as something
needed in fixed amounts in each economic activity, with some participants in the labor
force choosing to become entrepreneurs rather than workers. If entrepreneurial
"experience" or "human capital" is gained on the job (perhaps as a
function of prior education and training) then the supply of entrepreneurship would grow
with development (both as more persons became entrepreneurs and as those involved in
entrepreneurial activities gained experience).
If each activity has one "entrepreneur," whose function is to
organize factors of production, deciding what to produce, how to produce it, and accepting
risk, then analysis is somewhat more straightforward. As long as entrepreneurship is
associated with each activity, the basic growth model sketched above is unaffected. If
"more entrepreneurship," in either quantity or quality terms, is associated with
capital-intensive activities than with labor-intensive ones, the "catch-up"
process would have yet one more factor contributing to growth. In fact, entrepreneurship
was very important in Korea. However, its role seems to have been largely in the
liberalization process, as will be discussed below.
There is, finally, technology. In a formal sense, technical change can
be represented in any model as an outward shift in the output possibilities attainable
with given inputs. In fact, however, for most developing countries, technical know-how
(perhaps not unlike complex machinery and equipment) cannot be acquired and profitably
used without the education and training of enough members of the labor force to master it.
In an important sense, therefore, the acquisition of technological
know-how can be regarded as a form of physical capital formation,
highly complementary to human capital formation.
Relevance of the Model For Economic Policy
As set forth, the model indicates the properties of an efficient
resource allocation, but has no policy implication as to how that allocation might be
achieved. In principle, it could be attained by central planning or by well-functioning
markets. In practice, most developing countries have sizable private industrial sectors,[7] although many have also established a number of state-owned
manufacturing enterprises. However as will be discussed in section 2.1, the activities
undertaken by those enterprises bear no resemblance to the pattern of industrial activity
suggested by the model.
Optimal and Actual Policy
To attain economically efficient resource allocation through private
markets, optimal trade policy in the context of the model of section 1 would consist of a
free-trade regime unless there were monopoly power in trade. This would permit relative
prices in world markets to be reflected to domestic producers. If there were domestic
market failures of any sort (see section 2.2 below), optimal policy would be to correct
them at the source, by, for instance intervening in domestic markets. Since market
failures typically occur when there are significant uncapturable externalities, public
goods, or major indivisibilities or economies to scale, governmental economic policy would
largely be addressed to correcting these, either by appropriate tax-subsidy policy or by
direct provision of the externality-generating, large-scale, indivisible projects. These
are usually collectively termed infrastructure, and efficient provision of those
servicesâcommunications, transportation, etc.âis essential to economic
growth.
Within the industrial sector, however, profit-maximizing firms would be
expected to choose the appropriate mix of outputs when confronted with prices of
tradables, labor, and capital services that appropriately reflected their opportunity
costs (in terms of agricultural production for labor and of alternative industrial
activities in the case of capital).
Beyond this, the model offers little guidance as to what optimal
economic policy should be, but provides considerable insight as to inappropriate policies.
In practice, in the early years of conscious development effort after World War II,
policymakers deliberately adopted policies designed to achieve results almost the opposite
of those that would emanate from the two-sector, three-factor trade model developed in
section 1. They typically used direct controls to influence resource allocation,
especially among industrial activities, toward "import-substitution."
Simultaneously, the relative pricesâreal exchange rates, exportables relative to
importables, wage rates, and capital servicesâthat confronted private producers
systematically discriminated against exports.
Some government economic activities were the deliberate choice of
policymakers. The establishment of state-owned enterprises (SOEs) to perform particular
manufacturing functions, especially those regarded as capital-intensive and requiring
large initial investments that the private sector was thought unwilling to undertake, was
often one such deliberate choice. Often, these investments were defended on "infant
industry" grounds, as it was initially believed that these industries would
eventually become competitive.
As a logical corollary of these choices, curbs were placed on many
aspects of private-sector activity: licensing of investments, required permissions for
expansion, price controls where it was believed that there was monopoly power, and so on.
These controls extended to agriculture via pricing policy intended to keep food cheap for
urban workers and via government marketing agencies that had monopolies on the purchase of
agricultural outputs and supplies of agricultural inputs. They also extended to labor
markets, as minimum wage legislation, social insurance provisions, training and housing
requirements for workers, and guarantees against layoffs all raised the cost of hiring
labor. Likewise, the banking and financial system was heavily regulated, with the intent
being to channel "low-cost credit" into those lines of economic activity deemed
compatible with rapid growth. As a consequence, interest rates paid by those producers
with access to credit were well below those reflecting the opportunity cost of capital,
and were often negative in real terms. In most of these cases, motives for intervention
centered on a belief in market failure. Establishment of SOEs was undertaken on infant
industry or indivisibility grounds. Agricultural marketing boards were intended to prevent
"monopolistic exploitation" of small peasants; labor markets were regulated in
the belief that individual employers had considerable monopsony power with respect to
their labor forces, and so on. Credit
rationing was used to "direct" investment to those
industries deemed to be "high priority" for development; they were usually
capital-intensive, and often loss-making, despite high walls of protection through tariffs
or quantitative restrictions.
There were many unanticipated consequences of these policies.
Inflationary pressures were often considerably greater than expected; with domestic price
levels rising more rapidly than international prices, governments adhered to fixed nominal
exchange rates, which therefore represented increasing overvaluation in real terms.
Reluctance to devalue was based partly on the notion that export earnings of developing
countries could not grow ("elasticity pessimism") and partly on the idea that
increasing the price of foreign exchange would make capital goods more expensive and
thereby discourage investment.
Foreign-exchange earnings (which originated almost exclusively in
nongovernmental activities) almost always fell short of expectationsâ usually
growing less rapidly than national income, and often declining as real exchange rates
appreciated. Demand for foreign exchange, associated with the heavy investment
requirements of capital-intensive industries and with overvalued exchange rates, grew much
more rapidly. Balance-of-payments crises were the almost universal consequence of these
phenomena. Typical responses included "too little too late" devaluations,
combined with stringent exchange controls and reliance upon import licensing to restrain
demand for imports.
As is only too well known, the consequent impetus to "import
substitution" went far beyond anything that might have been defended on infant
industry grounds, or that might have been intended under the initial policy design.
Policies were often internally conflicting, and results were often very different from, if
not opposite to, those intended. Strict rationing of imports of intermediate goods and raw
materials often gave domestic producers determinate output levels and shares of the
domestic market. With import prohibitions prevailingârationalized on the ground
that domestic production already existedâforeign competition could not provide a
spur, and capital/output ratios rose sharply. Economies that had initially been dependent
on the international market for consumption and investment goods became instead dependent
for employment and output, as factories could not produce the consumer and capital goods
without imported raw materials and intermediate goods.
The disincentives for hiring labor associated with labor legislation and
the relative cheapness of capital for those with access to it combined
to yield powerful incentives for using capital-intensive techniques of
production. Consequently, industrial employment grew much more slowly than industrial
output, as much new investment was allocated to labor-saving, rather than
output-expanding, investments. Moreover, this cost structure obscured the social
profitability of producing more laborintensive goods and thus contributed to a resource
allocation very different from that suggested by the model of section 1.
Each of these policies had direct costs in terms of inefficiency and
forgone growth, as will be discussed in greater detail in section 2.3. However, they also
interacted in ways that magnified their impactâ overvalued exchange rates further
reduced returns to agricultural producers; their supply response further shrank export
earnings and induced policymakers to further restrict imports on balance-of-payments
grounds; that, in turn, provided yet further discrimination against agriculture, and so
on.
The combined effect of these policy inefficiencies cumulated over time,
and in country after country growth rates fell despite increases in savings rates. In
early thinking about development, emphasis had been placed almost exclusively on resource
accumulation as a means of attaining economic growth. A major lesson arising from
experience with the types of policies described above has been that attaining economic
efficiency is at least as important as resource accumulation. Moreover, economic
inefficiency can result not only from "market failure," as was recognized
earlier, but also from "government failure." The analysis of economic policy and
liberalization in developing countries therefore has several parts. First, a framework is
required for analysis of "market failure" and "government failure";
that is the topic of section 2.2. Next, there is analysis of how "government
failure" may impede growth, the topic of section 2.3. Finally, there is analysis of
the ways in which liberalization, or the removal of controls that inhibit economic
activity, may affect growth, which is covered in section 2.4.
Interactions of Policies and Growth
For purposes of analyzing policy, it is useful to distinguish between
policy impacts on resource accumulation and policy impacts on resource
allocationâalthough clearly there are a large number of policies that affect both
simultaneously. An efficient resource allocation is characterized by equalities between:
(1) the domestic marginal rate of transformation (DMRT) and international marginal rate of
transformation (IMRT) in production, (2) The DMRT and the domestic marginal rate
of substitution (DMRS) in consumption, and (3) the marginal rate of
substitution between factors of production in all uses. For resource accumulation,
necessary conditions would include the real return to investment at least being reflected
to savers.
Bhagwati (1971) classified departures from efficient markets
(distortions) as being of two types: exogenous and policy-induced. For present purposes,
exogenous distortions can be regarded as "market failures," while
"policy-induced distortions" can be regarded as instances of "government
failure." Distortions are defined as phenomena that result in failure of the
optimality equalities to hold. Examples of government failure distortions would include
the imposition of tariffs when a country has no monopoly power in trade (thereby breaking
the equality between IMRT and DMRT), effective minimum wage legislation that results in a
higher wage in the formal sector of the economy than in the informal (so that the marginal
rate of substitution between labor and capital differs between the two sectors), and tax
structures that drive a wedge between factor or goods payments as perceived by the two
sides of the market, to name just a few.
Market failure distortions are those that are independent of policy.
These might include the existence of externalities in production or consumption
uncorrected by appropriate taxes or subsidies, monopoly power in trade (which is a case of
externality as viewed from a national perspective) under a regime of free trade, economies
to scale under laissez-faire, or the absence of well-functioning markets (as is often
thought to be the case with financial markets or labor markets in developing countries).[8]
For purposes of analyzing policy, it is evident that there are four ways
in which policy can affect growth, either positively or negatively. Positively, inherent
market failures can be corrected so that either (1) resource allocation or (2) resource
accumulation decisions more correctly reflect the relevant trade-offs. The converse would,
of course, be if government intervention in fact moved the trade-offs as reflected
to decision makers even further away from their true relations.
Negatively, "government failure" can cause trade-offs as perceived by private
agents to diverge from social opportunity costs between (3) present and future consumption
(resource accumulation) and/or (4) production and consumption activities. The positive
side of these policies would be liberalization of one or more markets.
Policies of the first type might include government introduction of
agricultural research institutions, provision of mass vaccination programs, and provision
of elementary education opportunities. To be sure, these activities could be so
inefficiently undertaken as to yield a negative real rate of return, in which case the
intervention would lead away from, rather than toward, an efficient allocation of
resources. Generally, however, the evidence suggests that rates of return on these
activities are high.
Many observers believe that financial markets are undeveloped in most
developing countries and that the absence of appropriate financial instruments is a strong
disincentive to additional savings. If this is so, an example of the second kind of
positive government action would be policies designed to create a framework for more
efficient financial intermediation. Creating a legal framework for financial markets, or
otherwise encouraging their more efficient functioning, would then permit increased
resource accumulation. There might, of course, he some degree of market failure, and yet
the policies intended to remedy it (such as credit rationing) might lead to a worse
outcome than the original market imperfection; this would be an instance of a negative
government action of the second type.
The third kind of policy action would be of the opposite type:
governmental regulations that provide a disincentive for savings, such as interest rate
ceilings, are a good case in point. Savers confronted with low or negative real returns
would save less, thereby reducing the rate of resource accumulation.
Policies of the fourth kind would include the imposition of tariffs or
quotas on imports (in the absence of monopoly power in trade), minimum-wage legislation
that resulted in substantial unemployment and induced firms to use more capital-intensive
techniques and less labor than they otherwise would, and controls on producer prices of
particular agricultural commodities.
Although each of the examples cited above would probably have its
primary impact either on resource accumulation or on resource allocation, many
policy-induced distortions directly affect both accumulation
and utilization; policy changes that improve resource allocation,
either by improving functioning of markets or by mitigating the impact of existing
policies, probably also affect rates of resource accumulation. Thus, interest-rate
ceilings probably affect not only the willingness of individuals to save, but also the
behavior of financial intermediaries and borrowers: either financial intermediaries charge
market rates of interest for their loans and reap rents that themselves become a source of
distortions, as large numbers of those intermediaries compete for deposits, or ceilings
are also imposed on interest rates that may be charged to borrowers with consequent
misallocation of scarce resources. Moreover, insofar as economic efficiency increases, it
is reflected in higher real rates of return to investment (and higher payments to other
factors of production) and thus in higher growth.
Impact of Government Failures on Efficiency and Growth
As already mentioned, active government intervention was based largely
on the belief that private market failure was all-pervasive. Once begun, however, there
were a number of unintended policy outcomes, many of which led to further interventions.
Often policies were far different than those that might originally have been anticipated.
For example, "foreign-exchange shortage" led to import restrictions far more
severe and protective than those envisaged under the aegis of infant industry protection.
Moreover, activist governments, and especially direct controls, led to
the creation of many valuable property rights that were allocated by government officials.
Import licenses, investment licenses, permissions to expand capacity, sales of commodities
at controlled prices, subsidies on inputs such as fertilizer, and agricultural credit all
were highly valuable to the recipient. Administration of these allocations was costly,
both in terms of the drain on scarce administrative and bureaucratic talent, and in terms
of the resources then devoted to attempting to capture licenses and permissions.
Certainly, the implicit assumption underlying the rationale for government intervention in
the case of market failure was that these interventions were costless. But, on the
contrary, they were often very expensive.
Although governments have intervened and "failed" in many
fields of economic activityâfor example, energy pricing, transport charges, and
pricing and delivery of communicationsâthere are several broad areas in which
intervention has had overriding macroeconomic effects on efficiency and growth. They are:
agriculture, industry, the trade and pay-
ments regime, the labor market, and capital markets. [9] These are
areas of intervention that have been most frequently encountered in developing countries,
and that have had demonstrably harmful effects on growth rates in many instances. For
present purposes, the discussion will be confined to the negative effects on growth and
resource allocation of intervention in these five areas, in line with the model of growth
presented in section 1.
2.3.1. Intervention in Agriculture.
A strict interpretation of the model presented in section 1 would
suggest that the role of agriculture in growth is severalfold: (1) to generate savings
enough to permit capital accumulation in the industrial sector; (2) to provide a highly
elastic source of labor to the urban sector; and (3) although the model does not
explicitly derive a trade balance, it is apparent that at early stages of growth, imports
of manufactures would greatly exceed exports, and that these would be financed by exports
of agricultural commodities. In the growth process, incomes accruing to those employed in
agriculture would rise as the land/man ratio rose with outmigration, although in a more
complete model of the agricultural sector, investment in agriculture, technical change,
and other productivity-raising activities would also raise incomes.
In reality, of course, agricultural growth has yet other functions. It
became well understood in the marketed-surplus discussion of the 1950s and the 1960s[10] that growth of agricultural output is a prerequisite for
development at its early stages, and that policies I hat provide sufficient disincentives
to agricultural production can effectively thwart growth prospects. In addition to the
savings, labor supply, and foreign-exchange earning functions identified by the model,
there are other reasons for this. First, such a very high fraction of population and
output originate
in agriculture at the early stages of development that failure to
generate any increase in agricultural output effectively implies no growth. Second, the
model abstracts from population growth. The fact that in most developing countries
populations are growing implies that diminishing marginal product will lower rural real
wages in the absence of measures that raise agricultural output (either through increasing
land use or by increasing yields). [11] As if these arguments were not enough, imagine
that in one way or another, the industrial labor force could grow with stagnant or
declining total agricultural output. Either exports would have to decline sharply with
industrial growth, as urban demand for agricultural output rose, or prices of food would
have to rise sharply. Neither of these outcomes would be compatible with long-term growth.
Finally, it is inconceivable that there could be sustained growth of the
industrial labor force with declining food production at early stages of development; it
has already been pointed out that agricultural products would be the exportables. With no
increase in food production, and rising urban demand for food, the price of food and hence
urban labor costs would surely rise, thereby choking off prospects of further industrial
growth.
Some governments, usually weighing this last argument very heavily,
attempted to contain industrial wage costs by legislating price controls on food
commodities for urban areas. [12] Because the budgetary costs of financing these
below-market prices would have been excessive, many countries instead followed policies of
keeping producer prices low. Others, usually countries with specialized export crops,
attempted to achieve the savings from agriculture by direct or indirect taxation (through
agricultural marketing board policies and also through overvaluation of the exchange rate,
which serves as an implicit tax on exports) of agricultural exports. In both instances,
the result has been a shift by producers to noncontrolled (sometimes nonmarketed) crops.
In extreme cases, exports and output of exportables have fallen sharplyâ Ghanaian
cocoa may be the best known example.
It is a reasonable hypothesis that sufficient repression of the
agricultural sector at early stages of development in effect foredooms any reasonably
satisfactory development effort until policies toward that sector are reformed.
2.3.2. Distortions in Industry.
As indicated earlier, the major stated rationales for direct
intervention in the industrial sector (as contrasted with indirect intervention through
the trade and payments regime, which is analyzed in section 2.2.3) were three, not
necessarily consistent, beliefs: (1) the infant industry argument that some industries
might prove economic in the longer run, but that owing to externalities, those incurring
the initial high costs of production would be unable later to reap the returns, (2) a
deep-seated suspicion of the monopoly position of private producers, and (3) the view that
large-scale industrial ventures would not be undertaken by private entrepreneurs because
the initial costs and risks would be too great to be acceptable. These second and third
beliefs both focused on some aspects of entrepreneurship and implicitly implied that
government entrepreneurship would function more efficiently than did private.
While market failure was also thought to occur because of
indivisibilities and economies to scale, those phenomena presumably underlay one of the
other three reasons for direct intervention. [13]
Here, attention is focused on the ways in which, in practice, government
failure led to inefficient industrial growth. In some countries, governments established
public sector enterprises to produce a variety of industrial commodities. Often these
commodities were precisely the capital-intensive ones that are inconsistent with efficient
resource allocation. Often, too, production techniques were highly capital intensive.
Again in contrast to the model of section 1, most of these public enterprises were
established and managed on the assumption that they would sell only in the domestic
market. In many public sector enterprises, this immediately implied uneconomically small
production runs, and considerable excess capacity owing to lack of domestic demand and/or
sufficient foreign exchange to purchase raw materials and intermediate imports.
In part because public sector enterprises can obtain their financing
from government treasuries, there appears to have been little attention to cost control,
with deficits of sizable proportions financed by the
government. Because public sector enterprises were seen as a potential
instrument of government, politicians often loaded extra objectives on these agencies,
including the imperative to employ redundant workers, to select those politically favored
for posts for which other, better qualified, candidates were available, and so on. In
practice, SOEs were not innovative, competitive, and entrepreneurial, but were rather
genuine high-cost monopolies, unable to change their ways.
In many countries SOEs existed side by side with private sector firms.
Because of suspicions of private economic activity, however, much of it was heavily
regulatedâespecially for larger enterprises. Often regulations governed the types
of economic activities that could be undertaken privately. Those private enterprises
undertaking activities deemed "priority for development" were then accorded
special privileges, including access to rationed credit at low or negative interest rates,
import licenses for capital goods to be purchased at overvalued exchange rates, tax
holidays, and other valuable, but incentive-distorting, privileges. One consequence was
that these private enterprises had much the same incentive to employ capital-intensive
techniques as did their SOE counterparts.
Requirements that private firms apply for various licenses and permits,
meet various restrictions, provide services for their labor forces, and otherwise meet
regulations can impede efficiency to a greater or lesser extent, and can slow down growth
of industrial output. This result obtains because investment per unit of output rises (as
it does with anything that slows down or reduces the efficiency of the process of plant
construction and equipment installation) and because the composition of investment shifts
toward activities with lower rates of return.
2.3.3. The Trade and Payments Regime.
The model of section 1 implies that industrial growth can most rapidly
be achieved through reliance on the international market. There are two reasons for this.
First, capital will have a higher marginal product in the more labor-intensive industries.
Second, a given savings rate will result in a greater upward shift in the demand for labor
under free trade than it would in a closed, or protected, economy because investment can
be more heavily concentrated in labor-intensive industries when output of those industries
can be sold abroad in exchange for capital-intensive goods that would otherwise have to be
produced domestically.
While direct regulation of agricultural and industrial production has
had high costs in a large number of developing countries, the evidence
suggests that in most situations the negative impact on growth of
direct intervention was generally less than was the impact of the trade and payments
regime. Moreover, once restrictive trade and payments regimes were in place (often imposed
in response to immediate balance-of-payments pressures), they became subject to
manipulation for a variety of purposes.
Whereas many of the regulations directly governing agricultural and
industrial production were the intended consequence of belief in market failure, many of
the inefficiencies associated with highly restrictive trade and payments regimes were the
unintended consequence of policies undertaken for other purposes. Moreover, if there was
ever a clear-cut case of government failure, it was in the trade and payments regimes:
whatever the original motive for intervention, the evolutions of trade and payments
regimes over time demonstrated that particular types of controls, once in place, bring
into being political pressures to manipulate their uses.
The foreign-exchange market in and of itself can be more or less
distorted. In practice, however, restrictive trade regimes have affected not only the
foreign-exchange market, but also the market for industrial and agricultural commodities.
For countries that have been able to achieve a rate of growth of agricultural output that
permits a satisfactory rate of industrial growth, at least in principle, the foreign-trade
regime has proved to be the instrument most governments have used to influence (both
deliberately and with unintended side effects) the market for industrial output.
While there are instances of some degree of monopoly power in trade, few
if any regimes intervene in ways that would be consistent with correcting an exogenous
distortion. Almost all distortions have been policy-induced, although they have not all
been deliberate. Many highly restrictive trade regimes have been put in place as a
(distorting) policy response to an excess demand for foreign exchange and as a
not-necessarily-recognized alternative to alteration of the exchange rate. The existence
of an overvalued exchange rate and a highly restrictive import-licensing regime has then
had pervasive consequences for both the agricultural and the industrial output in many
developing countries.
Indeed, it can be argued that many of the apparent inefficiencies of the
industrial sectors in developing countries have resulted from trade and payments regimes
that have provided individual producers with quasi-monopoly positions, sheltering them
from the international market and preventing domestic competition (partly because of the
small
size of the domestic market) through import and investment licensing
procedures designed to prevent "excess capacity" and to accord individual firms
"fair shares" of available imports of intermediate goods and raw materials.
Moreover, the highly protected domestic market has served as a sizeable
disincentive for exporters, and countries with highly restrictionist trade and payments
regimes have typically found the commodity composition of their exports heavily centered
on raw materials in whose international price there is a significant component of rent, as
new resources have been channeled almost exclusively into production for sale on the more
profitable, highly protected, domestic market. (See Krueger 1984 for a fuller analysis of
the costs of these trade regimes.)
2.3.4. The Labor Market.
A crucial implication of the multicommodity, two-sector, three-factor
model is that a major gain from trade for developing countries derives from their ability
to export "embodied labor" and to import "embodied capital." Since the
competitive positions of individual firms are determined by the profitability of
alternative activities, it is readily apparent that economic efficiency would obtain when
firms in countries with very high labor-land ratios found it profitable to produce and
export very labor-intensive commodities. This would occur when the wage rate (or wage
structure, in a more general model) reflected the relative abundance of labor.
If firms in labor-abundant countries are confronted with relatively high
labor costs, they will find it profitable to produce goods that are less labor-intensive,
and comparative advantage would not be realized. If real wages are legislated, the outcome
might even be a "reversal" of comparative advantage, with exports of more
capital-intensive goods and imports of labor-intensive commodities. In such a
circumstance, of course, there would either be open unemployment in the economy or there
would be a wage differential between regulated activities and the rest of the economy.
(See Magee 1973 for analysis of "reversal" of comparative advantage under a wage
differential model and Brecher 1974 for analysis of the rigid real wage-unemployment
case).
Hence, in a market economy, efficient use of the international economy
to accelerate industrial growth could be thwarted by regulation of wages as much as by
protection of domestic production and exchange rate overvaluation. If wages are
effectively set at levels significantly above those at which a sufficient labor supply
would be forthcoming from the agricultural sector, it could preclude reliance on the
interna-
tional market and the type of outward-oriented development implied by
the model of section 1.
It has already been shown that many developing countries have attempted
to regulate wages and conditions of employment. In some cases, these regulations have been
largely ineffective. [14] In other cases, however, they have been enforced, at least over
the "formal sector," which encompasses all sizable enterprises. Since there is a
reasonable presumption that most industrial exports will be produced by firms large enough
to be visible, and hence subject to these regulations, there is a presumption that highly
restrictive labor legislation can preclude the development of production for export of
manufactures in countries that might otherwise have had a significant comparative
advantage in export of labor-intensive manufactures.
In countries with highly restrictive trade and payments regimes,
regulations effectively raising real wages to levels that lead to open urban unemployment
can significantly retard growth. But even with an open trading regime, the potential
benefits of the international market can be choked off if wages are set significantly
above levels consistent with comparative advantage.
2.3.5. Capital-Market Interventions.
The frequently encountered phenomena of credit rationing, interest
ceilings, and fragmented and regulated financial markets have already been mentioned. Some
(see, for example, McKinnon 1973) believe that the costs of these regulations have been so
high as to dwarf the effects of the four types of distortions discussed thus far.
Clearly, if capital services are artificially cheap to some producers,
there will be an excess demand for them. If some investable resources are directed to
recipients who would not find it profitable to pursue their activities at a realistic
interest rate, there will necessarily be others who would have been able to find
profitable activities at that rate but are excluded by whatever rationing mechanism
supports the credit rationing. In those instances, the consequences of credit rationing
will be identical to those of regulating real wages at levels inconsistent with
comparative advantage, and the costs of interest ceilings and credit rationing,
from the viewpoint of the model set forth in section 1, will be the
same as those of labor-market interventions, so the analysis need not be repeated here.
It should be pointed out, however, that the costs of "financial
repression," to use McKinnon's phrase, would presumably rise over time with capital
accumulation in the two-sector, multicommodity, three-factor model. This would follow
simply by virtue of the fact that the costs of given divergences between rates of return
to capital-starved and capitalabundant activities would become greater as the size of the
capital stock increased.
Liberalization and Economic Growth
To this point, a model of efficient outward-oriented growth has been
developed, and the resultant theory of economic policy briefly sketched. That theory has
two sets of implications for policy. First, it points to the importance of an open trade
regime (which in turn requires a realistic exchange rate) and well-functioning factor
markets as essential elements of a development policy and, by implication, assigns to the
state important responsibilities for delivery of reliable infrastructural services to
support relatively rapid export growth. Second, and the obverse of the first part, it
suggests the kinds of costs that may result from government failures of the type described
in section 2.3.
While the model provides a guide to appropriate economic policy, a key
question remains. What are the implications for a policymaker desirous of achieving more
rapid growth if he starts in a situation characterized by the types of interventions
outlined in section 2.3. and is, in addition, constrained by the political process
(perhaps because of the built-in vested interests and political pressures that surround
interventions of the type described) as to the number and magnitude of reforms that can be
undertaken?
3.1â
Theory and Liberalization
Theory provides very little guidance in addressing this question. In
the first place, the theory of the second best suggests that in general one cannot be
certain that liberalization of any particular market will increase welfare in the presence
of other, unremovable, distortions (given that total liberalization of all markets will
lead to the very best outcome). Second, insofar as there are costs to adjustment (which
would be the main analytical rationale for reducing the speed of adjustment), it is en-
tirely an empirical question as to how sizable these might be.
Furthermore, the magnitude of those costs might well differ from country to country as a
function of the prior history of attempted liberalizations, the credibility of the
policymakers, the initial set of policies that are in place, and the timing of the
liberalization effort in relation to the cyclical phase of the international economy.
The empirical evidence, however, suggests that most market
liberalizations that were sustainable have had at least some welfare-improving effects.[15] And, for purposes of the present discussion, it will be assumed
that any liberalization will produce some positive benefits to the efficiency of resource
allocation and to growth. [16]
The very concept of "liberalization" is associated with the
removal of intervention and the shifting toward reliance on market forces to achieve
objectives. Liberalization is therefore a process of removing the types of interventions
detrimental to growth, described in section 2.3. Except in the sense that the theory
described in sections 1 and 2.1 suggests that removal of controls may contribute to
growth, theory can provide little guidance to policy for such questions as:
(1) Starting from an initial situation of multiple interventions and
controls, if there is a constraint on the number of fronts on which action can
simultaneously be taken, which markets should be liberalized first?
(2) Is it preferable to liberalize all markets at about the same rate,
or entirely to liberalize first one market and then another?
(3) What are the costs of adjustment associated with liberalization and
can these costs be reduced if, as may be unrealistic, policy-makers can announce a
credible and unalterable schedule of liberalization to permit more gradual adjustment?
(4) Is there a critical minimum amount of liberalization that must be
undertaken in order to be reasonably confident that the benefits will exceed any
adjustment costs?
None of these questions can be answered in theory. Adjustment costs
might be negligible or highly significant; they might differ from firm to firm, industry
to industry, or country to country. The speed of reaction might also vary widely according
to circumstances.
Hence, efforts to understand the process must rely on empirical
evidence. This can be of two kinds. Efforts can be made to ascertain the structure of a
particular economy, focusing on estimation of the likely magnitude of the key response
parameters. There are several difficulties with this. One problem is that the types of
growth-inhibiting policies described above significantly reduce the response of the
economy to such changes as do occur. And, since in most instances liberalization is an
event taking place outside the range of observations of past economic behavior, empirical
evidence, or even inference ex ante about the likely response in a particular country, is
exceptionally difficult to obtain, and professional economists can reasonably disagree as
to its implications.
The second kind of evidence as to the probable magnitude of the
parameters is from countries that have previously liberalized, or attempted to liberalize,
one or more markets. It is in this context that the Korean experience provided a valuable
laboratory in which to examine the liberalization process.
Lessons from the Korean Experience
In the 1950s the Korean economy was subject to many of the policies
described in section 2.1. [17] Macroeconomic imbalances were severe, with inflation
averaging 35 percent annually (and peaking at 85 percent), government budget deficits
ranging around 5 percent of GNP, investment averaging around 11 percent of GNP, and
domestic savings around 3 percent of GNP. Aid inflows covered much of this imbalance, and
were around 10 percent of GNP in the mid to late 1950s.
The government's efforts were largely directed at containing
inflationary pressures, along with postwar reconstruction. As part of the
anti-inflationary policy, nominal exchange rates were held fixed for long intervals before
adjustments were undertaken; these adjustments nonetheless left the exchange rate
chronically overvalued. In response to chronic excess demand for foreign exchange, the
government resorted
to multiple exchange rates, high tariffs, and quantitative
restrictions on imports. Exports were consequently discriminated against, and were only 3
percent of GNP as late as 1960, when imports were 10 percent of GNP.
Again motivated largely by a desire to contain inflation, the Korean
government depressed the prices of agricultural commodities, especially rice, a policy
made possible by large quantities of PL 480 grain. The nominal purchase price of rice was
already depressed in the mid 1950s, and it remained constant between 1956 and 1960 despite
an increase of 23 percent in the wholesale price index over that period.
Import-substitution industries were developed behind high walls of
protection, with average annual growth rates in excess of 15 percent in real manufacturing
value-added for the last half of the 1950s in production of paper products, chemical and
petroleum products, rubber products, basic metals, and metal products, including machinery
and transport equipment. These compared with average annual growth rates of real
manufacturing value-added of less than 10 percent for textiles and leather products, two
sectors that were major sources of export growth in later decades. (Data are from Kim and
Roemer 1979.)
Interest rates were kept well below inflation rates, and credit
rationing prevailed; financial markets in Korea in the 1950s were therefore highly
repressed and fragmented. Only in the labor market does Korean policy in the 1950s appear
to have been less distortionary than that of many other developing countries. The labor
market appears to have been relatively unregulated in the 1950s, as well as in the period
of Korea's rapid growth (see Kim and Roemer 1979, 162).
These distortions were reflected in generally slow growth in Korea.
Despite the huge inflow of aid and the end of the Korean War in 1953, which should have
permitted fairly rapid growth with reconstruction, the average annual rate of growth of
real GNP was only 4. 1 percent, which, with flows of migrants from the north and high
birth rates after the war, resulted in per capita income growth of only 1. 7 percent
annually for the years 1953â55.
The sequence of reforms cannot be neatly demarcated. Policy changes
began in 1958, but it was not until 1964 that the features of the new policy regime were
reasonably well defined and understood. Starting in 1958 there was a considerably greater
adjustment in the nominal exchange rate and depreciation of the won in real terms, but it
was not until 1964 that the exchange rate was unified and permitted to float to maintain a
virtually constant real value. And, starting in 1957â58,
serious efforts were undertaken to achieve greater macroeconomic
stability. The inflation rate as measured by the GNP deflator slowed drastically,
averaging 3.8 percent annually in 1957â60 compared to a 36 percent annual average
from 1953 to 1957. However, it rose again to an average of 22 percent over the
1960â64 period, when further fiscal and interest rate reforms were introduced.
Thereafter, it remained moderate, averaging 12 percent annually over the years
1964â73.
Similarly, by the early 1960s, quantitative restrictions on imports were
almost entirely abolished insofar as they affected producers of exportables. Exporters
could import any intermediate input used in production and were, in addition, the only
ones eligible to receive import licenses for other commodities, mostly luxury consumer
goods. It was not until 1967, however, that the government shifted from a positive list of
eligible imports to a negative list of ineligible imports and began significantly to
reduce tariffs. The import liberalizationâtariff reduction process continues to
this day. [18]
For agriculture, suppression of producer prices remained the order of
the day until the late 1960s, although the rate of increase of agricultural production in
the 1960s exceeded that of the 1950s, presumably because a more realistic exchange rate
and import liberalization had reduced the extent of discrimination against agriculture. By
the late 1960s, however, it became apparent that rural farmers' incomes were not rising as
rapidly as urban ones, and policies were shifted toward inducements to agriculture. In the
Korean case it seems evident that liberalization of the trade regime, and especially
removal of the bias toward import substitution, spurred industrial growth, and that rapid
industrial growth in turn "spilled over" via increased demand for agricultural
products (and increased demand for urban labor, which led to rural outmigration and rapid
increases in real urban wages). [19] Until the 1970s Korean agriculture benefited from
liberalization chiefly through the effects of a more realistic real exchange rate and from
rising urban demand. Thereafter, efforts to support rural incomes led to a reduction in
the disparity between rural and urban incomes. [20]
As to financial liberalization, it has not been fully accomplished to
this date. Major reforms in 1964 greatly reduced the ability of the government to borrow
or print money and closed the budget deficit. Simultaneously, ceilings on nominal interest
rates were raised sharply, so that real interest rates were positive in the late 1960s. In
the early 1970s, however, they were again controlled, despite a sharp increase in the
inflation rate.
If one were to attempt to characterize the Korean experience in general
terms, it would be that the sharpest policy reversal, and earliest liberalization, was in
the trade and payments regime, and especially in removing the bias of the regime against
the international market. The response of the economy to that liberalization was very
rapid and far greater than had been anticipated. Although many opposed the removal of
protection for domestic industry in the late 1950s, opposition appears to have been
sharply reduced by the unanticipated magnitude of the success. That export earnings in
dollar terms could grow by more than 35 percent annually over the fifteen-year interval
following 1960 attests to the magnitude of that response. That it exceeded all
expectations is evidenced not only by statements of Koreans, but by the fact that a
fiveyear plan introduced in 1962 had to be abandoned because its goals were all achieved
in the first half of its intended life.
Consistent with the model set forth in section 1, Korea's exceptionally
rapid growth in the early 1960s was spurred by the growth of the industrial sector, which
in turn was led by the growth of exports of laborintensive goods. As Mason et al. (1980)
point out, the leading growth sector was industry; agriculture benefited and grew largely
because of the spillover of demand in the rapid-growth environment. That labor-intensive
industry was where comparative advantage lay initially cannot be doubted: textile exports
rose from $4.1 million in 1961 to $54.5 million in 1965; lumber and plywood from $1.2
million to $18.2 million; and metal products from $1.6 million to $17.8 million: later on
electronics products also entered the list (see Krueger 1984, 101). By almost any method
of measurement, the role of import substitution in growth was larger than that of export
expansion in the 1950s, whereas export expansion accounted for more than twice as much of
the rapid growth as did import substitution in the 1960s and 1970s. For the period
1963â73, one estimate put the contribution of export expansion to growth at 40. 1
percent of total growth compared to 9. 9 percent for imports. [21]
Interestingly, many of the new industries that arose in response to
export incentivesâwigs, plywood, electronics, and so onâwere initially
developed as export industries. To the extent that they were sold on the domestic market,
it was secondary right from the start.
Another episode in Korea's remarkable experience serves as partial
verification of the model. That is, in the second half of the 1970s, the government
decided to undertake substantial investments in the heavy and chemical industries, which
was, in effect, a reversal of the policies (which turned out to be temporary) that had
served the economy so well over the preceding years. This program was decided upon based
on the belief that the Korean economy had by that date progressed far enough to sustain
that sort of development. However, it became evident that the Korean economy was not yet
ready: as Kim Kihwan points out: "These policy changes were achieved at the cost of a
high rate of inflation. . . . In addition, these changes resulted in
serious imbalances in the economy. The most serious imbalances were an over-investment in
heavy industries and under-investment in light industries: extensive price distortions and
lack of competition due to government controls; and a rise in real wages which exceeded
productivity improvement. These imbalances led to a weakening of export competitiveness,
and brought about a slowdown in the overall growth in the economy" (Kim 1985, 15).
Only in 1981â82 were the policymakers able to resume the outward-oriented stance of
the earlier era and to return to growth. That Korea's comparative advantage did not lie in
the capital-intensive heavy industries even after a decade and a half of rapid growth and
rising real wages became manifestly evident.
One final point needs to be mentioned: examination of the history of
policy change in Korea strongly suggests that "liberalization" was not a
once-and-for-all policy action. Rather, it was a continuing process, with difficulties
encountered at each step. Yet, the Korean experience since the mid 1950s strongly suggests
that rapid growth was not, in any sense, assured after the initial liberalization: rather,
as restrictive policies increasingly appeared to represent impediments to further growth,
policymakers grappled with the problem of how to liberalize further still
A second factor of importance in Korea was that the labor market was
relatively free and therefore did not have to be liberalized. And while financial
liberalization was far from complete, the reforms that were undertaken implied far smaller
subsidization of successful borrowers than had earlier been the case.
Timing and Sequencing of Liberalization
Given the remarkable transformation of the Korean economy, it is
relatively straightforward to infer what happened there. Quite clearly the gains from
liberalization were enormous, and the shift toward reliance on the international market,
with accompanying liberalization of other parts of the economy, permitted a sustained rate
of growth that had earlier been regarded as unattainable under any circumstances.
One question remains: what lessons does the Korean experience provide
for countries whose economies are still subject to the same sorts of controls as was the
Korean economy of the 1950s? On one hand, the evidence appears overwhelming that the
potential gains from a major shift in policy stance could be very large. The real and more
difficult question is what sorts of liberalization measures are most likely to be
successful in achieving significant increases in efficiency and growth at least cost.
Some lessons seem to follow straightforwardly from the model and from
the Korean experience. Attention turns first to them. Thereafter, some more conjectural
hypotheses as to the liberalization process, which are not inconsistent with the Korean
experience but can in no way be verified by it, are put forth in the hope that further
research, perhaps involving comparative analysis, might shed light.
Turning first to those conclusions that seem fairly solidly based, the
first, and probably most significant, conclusion is that liberalization of the trade and
payments regime could not have succeeded had the authorities not simultaneously ensured
the continuing maintenance of a realistic exchange rate. [22] Second, and closely related
to the first, the Korean export drive could not have succeeded had imports, and the trade
and payments regime more generally, not been liberalized. This was essential both in order
to ensure that potential producers would be willing to take the risks associated with
exporting (which would have appeared far less attractive had there remained a sheltered
domestic market) and to ensure that exporters could efficiently obtain needed imports in
order to compete with suppliers in other countries. Third, Korea's labor market was not
subject to heavy regulation, which was undoubtedly a necessary condition for the early
success in exporting labor-intensive goods. Had there been minimum-wage legislation and
other policies that sharply increased the cost of employing labor, it is difficult 10
imagine
that Korea could have been successful, especially in the early stages
of the export-oriented drive.
At the more conjectural level, other hypotheses emerge. Note in
particular that Korea did not initially liberalize all markets entirely. Indeed, the
outward-oriented export strategy had been in effect at least four years before there was
any degree of financial liberalization and, except to exporters, imports were not entirely
liberalized for an even longer period.
It is tempting to postulate that successful economic growth brings with
it an increasing complexity of the economy, which it surely did in the Korean case. With
that increasing complexity, a system of direct controls probably becomes more difficult to
administer, and the cost of errors probably increases. This is not only because the range
of particular decisions that has to be taken increases (as the number of products,
intermediate goods, types of labor, etc., increases), but also because the transparency of
the effects of particular controls diminishes to policymakers and the public alike. This
diminishing transparency means that the regulators cannot readily know the implications of
their decisions, and simultaneously that individual entrepreneurs have more opportunity
for misrepresenting their situations in order to avail themselves of the profitable
opportunities inherent in a system of direct controls. The consequences of these two
phenomena combined are probably increasing complexity of controls and increasingly costly
mistakes as growth progresses.
Because of the role of the international market as a regulator of
economic behavior and the increasingly profitable opportunities that a direct-controls
regime creates, it is likely that direct controls over foreign trade will become
increasingly costly the further industrial development progresses. Part of the observed
rapidly rising capital/output ratios in countries clinging to direct control systems may
be a reflection of this phenomenon, rather than only of the rising capital intensity of
new import-substitution lines.
By a similar line of reasoning, it may be that credit rationing and
interest-rate ceilings do not, at early stages of development, impose very high direct
costs, because of the transparency of the opportunities for very highly profitable
investments. However, as the complexity of the industrial structure increases, and the
contribution of capital as a factor of production rises (because there is more of it), the
costs of misallocation of scarce capital may rise, while the quality of the information on
which decisions as to credit allocations are based diminishes.
Using the Korean experience as a model, it seems evident that credit
rationing could not have imposed too high a cost on the economy in the early 1960s. The
fact is that exporting was highly profitable. Resources should have been drawn into export
industries. Credit was allocated preferentially to exporters, and this simple rule
probably ensured a reasonably efficient allocation of investable resources, although there
were undoubtedly inefficiencies and discrimination against small firms. As development
progressed, however, a sufficiently high fraction of all investment was in exportable
activities, so that the simple rule probably failed to perform as well. When
differentiation had to be made among exportable activities as to which were likely to
yield higher rates of return, allocation rules would have resulted in larger and more
costly mistakes.
Based on the Korean experience, therefore, one might tentatively
conclude that liberalization and reduction of the bias of the trade regime, and especially
the regime as it affects exporters, is an essential prerequisite for satisfactory
industrial growth. Especially for countries with a high labor/land ratio and therefore a
strong comparative advantage in very labor-intensive industries, a reasonably free labor
market is also essential. Financial liberalization, while clearly important for sustaining
it, may not be as necessary initially if a government starts in a situation where many
markets are regulated. Sustained growth will nonetheless ultimately require liberalization
of the financial markets; their liberalization remains a major challenge for Korean
policymakers.
References
Bell, Clive, and T. N. Srinivasan. 1985. "Agricultural
Credit Market in Punjab: Segmentation, Rationing, and Spillover." DRD Working Paper
no. 7. Mimeographed.
Bhagwati, J. N. 1971. "The Generalized Theory of Distortions
and Welfare." In J. N. Bhagwati, R. W. Jones, R. A. Mundell, and J. Vanek, eds., Trade,
Balance of Payments and Growth, 69â90. Amsterdam: North-Holland Publishing Co.
Brecher, Richard. 1974. "Minimum Wage Rates and the Pure
Theory of International Trade." Quarterly Journal of Economics 88 (February):
98â116.
Carvalho, Jose L., and Claudio Haddad. 1981. "Foreign Trade
Strategies and Employment in Brazil." In Anne O. Krueger, ed., Trade and
Employment in Developing Countries: Individual Studies, 29â81. Chicago:
University of Chicago Press.
Corbo, Vittorio, and Jose Miguel Sanchez. 1985. "Adjustments
by Industrial Firms in Chile during 1974â82." In Vittorio Corbo and Jaime de
Melo, eds., Scrambling for Survival: How Firms Adjusted to the Recent Reforms in
Argentina, Chile and Uruguay. World Bank Staff Working Paper no. 764. Washington, D.
C.: World Bank.
Frank, Charles R., Jr., Kim Kwang-Suk, and Larry E. Westphal. 1975. South
Korea. Vol. 7 of Foreign Trade Regimes and Economic Development. New York:
National Bureau of Economic Research.
Hansen, Bent, and Karim Nashashibi. 1975. Egypt. Vol. 4 of
Foreign Trade Regimes and Economic Development. New York: National Bureau of
Economic Research.
Hayami Yujiro and Vernon W. Ruttan. 1985. Agricultural
Development: An International Perspective. Baltimore: Johns Hopkins University Press.
Jones, Ronald W. 1971. "A Three-Factor Model in Theory,
Trade, and History." In J. N. Bhagwati, R. W. Jones, R. A. Mundell, and J. Vanek.
eds., Trade, Balance of Payments and Growth, 437â59. Amsterdam:
North-Holland Publishing Co.
Kenen, Peter B. 1965. "Nature, Capital and Trade. "Journal
of Political Economy 73 (October): 437â60.
Kim Kihwan. 1985. The Korean Economy: Past Performance,
Current Reforms, and Future Prospects. Seoul: Korea Development Institute.
Kim Kwang-Suk and Michael Roemer. 1979. Growth and Structural
Transformation. Studies in the Modernization of the Republic of Korea, 1945-1975.
Cambridge, Mass.: Council on East Asian Studies, Harvard University.
Krueger, Anne O. 1977. Growth, Distortions and Patterns of
Trade among Many Countries. Essays in International Finance no. 40. Princeton:
International Finance Section, Department of Economics, Princeton University.
. 1984. "Comparative Advantage and Development Policy Twenty
Years Later." In M. Syrquin, L. Taylor, and L. E. Westphal, eds., Economic
Structure and Performance. 135â56. Orlando, Fla.: Academic Press.
Krueger, Anne O. and Baran Tuncer. 1982. "Growth of Factor
Productivity in Turkish Manufacturing Industries." Journal of Development
Economics 11 (December): 307â25.
Lewis, W. Arthur. 1954. "Economic Development with Unlimited
Supplies of Labor." Manchester School 22 (May): 139â91.
McKinnon. Ronald. 1973. Money and Capital in Economic
Development. Washington, D. C.: Brookings Institution.
Magee, Stephen R 1973. "Factor Market Distortions,
Production and Trade: A Survey," Oxford Economic Papers 25 (March):
1â43.
Mason, Edward S., Kim Mahn Je, Dwight H. Perkins, Kim Kwang Suk,
and David C. Cole 1980. The Economic and Social Modernization of the Republic of Korea.
Cambridge, Mass.: Council on East Asian Studies, Harvard University.
Stiglitz, Joseph E. 1985. Economics of Information and the
Theory of Economic Development. NBER Working Paper no. 1566. Cambridge, Mass: National
Bureau of Economic Research.
World Bank. 1983. World Development Report, 1983.
Washington, D. C.: World Bank.
