wad ch1 state market and industrial policy

THE PREDOMINANT approach to economic policy in the 1950s and 1960s assigned the statel a substantial role in repairing market failures. In the industrialized countries the experience of the Great Depression and wartime
dirigisme2 provided the impetus. The approach was also taken up by economists dealing with "underdeveloped countries" and made the basis of a newly
emerging discipline of development economics. The early contributions to development economics concentrated on showing how the special circumstances
of underdeveloped countries-low private saving, dependence on primary
product exports, declining prices of exports in relation to imports, small internal markets, limited skills, few entrepreneurs adept at large-scale organization, and pervasive underemployment-required an even bigger role for the
state than in the more developed countries.
These circumstances meant that underdeveloped countries could not expect
to achieve full employment or earn enough foreign exchange to meet unrestrained demand in the context of an open economy, according to development
economists. Almost regardless of the exchange rate and even with wages no
higher than physical subsistence level, a shortage would remain of activities
producing goods or services saleable on the international market. Reliance on
the free market would perpetuate what was variously called dualism, marginalization, or the coexistence of precapitalist with capitalist forms of production. Hence the state should not only maintain macroeconomic balance and
supply "public" goods and services, but it should also undertake direct reI I use "government" to mean the executive branch and "state" to mean the wider structure
of governance institutions, but in practice the two words are often used interchangeably-less
serious a confusion for Taiwan and Korea than for many other countries.
2 Dirigisme in the French sense contains tbe idea of directional thrust, or orienting power, in
the hands of government (Petit Larousse 1975). I use dirigisme, guided market and governed
market more or less interchangeably, to mean strongly influenced rather than tightly controlled.
INDUSTRIAL POLICY 9
sponsibility both for augmenting the economy's investable resources and for
establishing a mechanism to transfer those resources into productive investment. Capital formation, in this view, was the engine of development. By
leaving significant amounts of an economy's resources underutilized, free
markets in less developed country (LDe) conditions, would generate less investment than was socially desirable and allocate it in less than socially desirable ways. Development economics thus restored capital formation to prominence, which having been at the heart of economic theory from the eighteenth
century to the First World War had then been displaced by issues of efficient
resource allocation. But it combined capital formation with an activist view of
the state in a way that classical economics had not. 3
The late 1960s and early 1970s saw a downgrading of the role of the state
in both developed and less developed countries. In the LDC context economists
presented three main kinds of evidence:
1. the use of the state to promote import-substituting industrialization during the
1950s and the 1960s had resulted in inefficient industries requiring permanent
subsidization, with little prospect of achieving international competitiveness
2. extensive government intervention tended to generate "rent-seeking" on a significant scale, that is, to divert the energies of economic agents away from production and into lobbying for increased allocations of government subsidies and
protection
3. some of the most successful LDcs-inc!uding Taiwan, Soulh Korea, Hong Kong,
and Singapore-had achieved extraordinary industrial growth by using an outward-oriented model driven by market incentives and a strong private sector"
The generally favorable effects that followed the gradual liberalization of
trade and capital flows among the developed countries during the 1960s
pointed to the same conclusion. And immediate fiscal conditions necessitated
smaller government; the 1973 energy price rise put acute pressure on state
budgets in most oil-importing countries, so that public expenditures had to be
curtailed.
Even where some form of government intervention could be justified by
market failure the actual effects of such interventions are often perverse, it
was said. Indeed, "government failure" is as pervasive and serious as market
failure, if not more so. A small but lively band of economists developed gen3 This approach was sometimes called "structuralism" because it emphasized the structural
impediments to the working of markets seen, for example, in large differences between sectors in
real product per worker. For the evolution of postwar thinking about the role of the state, I draw
on Brett 1985; Bienefe1d 1982; Lewis 1989. For early views which emphasize the employment
problem and/or the foreign exchange gap in underdeveloped countries, see Nurkse 1953; Rosenstein-Rodan 1943; Lewis 1955; Myrda11957; Chenery 1959; Singer 1949; Streeten 1964.
4 See in particular the seminal book by Little, Scitovsky, and Scotl1970; also Krueger 1974 on
"rent·seeking. "
10 CHAPTER I
eral arguments to show why such failure is an inherent tendency of all governments. S Other economists working on developing countries said that the governments of these countries were even more likely to fail than those in
developed countries, because, in Deepak Lal's words, "many developing
countries are closer in their official workings to the rapacious and inefficient
nation-states of 17th or 18th-century Europe, governed as much for the personal aggrandizement of their rulers as for the welfare of the ruled"
(1983:108). In these circumstances, he continued, it is the height of folly to
urge upon developing countries that "enlightened discrimination" toward foreign trade, transnational companies, technology, and the meeting of basic
needs prescribed by traditional development economics; discrimination is seldom enlightened.
The upshot of these new circumstances and arguments was that by the second half of the 1970s, if not earlier, the mainstream of thinking about development policy (especially in the English-speaking academic community and
international development agencies) had decisively shifted from the prescriptions of the 1950s and 1960s toward a "neoclassical" view of the appropriate
roles of markets and governments. The need for a special economics of development was denied (LaI1983).
In the neoclassical view, the engine of development is not so much capital
formation as efficient allocation of resources. Once institutional arrangements
are in place to generate an efficient allocation of resources investment can be
left to take care of itself. Whatever investment is generated by these arrangements constitutes-with some small exceptions-the social optimum. The
necessary institutional arrangements for generating efficient resource use are
competitive markets, particularly domestic markets integrated with international markets. Hence government should leave private producers operating
through market mechanisms to supply all but certain "public" goods. It
should limit its own activities to improving the functioning of markets and to
providing only those goods and services where the government has a clear
comparative advantage relative to private agents. The resources so released
can either be transferred to the private sector or used to improve the performance of the state's essential functions.
If prices reflect social opportunity costs, the underlying argument runs,
profit incentives will drive the economy to its maximum production potential.
So in a labor-abundant economy profit maximization by private producers will
lead to the choice of labor-intensive production techniques. In the absence of
trade distortions, exports of labor-intensive products will grow, generating
beneficial second-round effects on the rate of aggregate growth. As labor demand increases faster than the supply, labor surpluses will be absorbed, leading to a rise in real wages and a, change in comparative advantage guided by
the changes in factor scarcities.
~ For example, Wolf 1979; Buchanan and Tullock 1962.
INDUSTRIAL POLICY 11
The key development policy is therefore an outward-oriented trade regime,
characterized by low or negligible impediments to imports, relatively uniform
incentives for different production activities, and incentives for export sale
equal to the incentives for domestic market sale. These conditions will maximize the economy's income and growth (in world prices) by concentrating
resources on those activities in which the economy has a comparative advantage, leaving other forms of production to other nations. In addition, by expanding the proportion of the economy which is directly subject to international competitive pressures, the government's own ability to impose
"political" prices is weakened; hence producers' uncertainty about government policy is reduced.
Other poorer nations also benefit. For as a country experiences increases in
costs and improvements in technological capacity, so it exits light-manufactured exports-textiles, clothing, shoes, and simple consumer electronicsleaving them for the next tier, itself moving into more sophisticated products.
As Bela Balassa describes the sequence, the newly industrializing countries
(NICS) would "upgrade and diversify their exports in line with their
changing comparative advantage," leaving "countries at lower stages of industrial development to replace exports of unskilled-labor-intensive commodities from the newly-industrializing countries to industrial country markets"
( 1980:25-26).
THE PROPER ROLE OF GOVERNMENT
In the neoclassical view, the essential economic functions of government are
to:
1. maintain macroeconomic stability
2. provide physical infrastructure, especially that which has high fixed costs in relation to variable costs, such as harbors, railways, irrigation canals, and sewers
3. supply "public goods," including defense and national security, education, basic
research, market information, the legal system, and environmental protection
4. contribute to the development of institutions for improving the markets for labor,
finance, technology, etc.
5. offset or eliminate price distortions which arise in cases of demonstrable market
failure
6. redistribute income to the poorest in sufficient measure for them to meet basic
needs
This list of functions is uncontroversial as far as it goes. The controversy
comes at the step of recognizing market failures6 in practice and deciding what
to do about them.
6 Market failure is defined as a situation in which the market system produces an allocation of
resources which is not Pareto-efficient-it is possible to find ways of changing resource allocation
12 CHAPTER J
For example, many neoclassical economists would accept that markets for
technology development, manpower training, and credit to small firms and
exporters may fail seriously enough to warrant some offsetting government
intervention. 7 With respect to technology development market failure may occur because knowledge leaks. If a firm invests in acquiring technological mastery over new processes it will have trouble keeping the benefits to itself.
Other producers can' 'reverse-engineer" products. Employees can leave their
firm and apply on the new job what they learned on the old. So while the firm
bears the cost of the innovation it does not capture the full social benefit, and
hence underinvests in technological effort. Also, some forms of technological
development may require a minimum scale of effort, which a single firm acting on its own may not be able to mount. In cases where technology investments are suboptimal, tax incentives may be used to increase the implicit private rates of return on investment, while in cases where private scale is too
small, government coordination of public and private technology development
activities may be justified. The government also has a role in formulating and
enforcing intellectual property rights, in the form of patents and copyrights.
With respect to manpower development market failure may occur if the
benefits to society of having an educated popUlation exceed the benefits which
educated individuals can capture in the form of higher incomes. These benefits
may include a healthier population, lower fertility, and a better informed citizenry. Education and training are generally difficult to finance privately because of the absence of nonfamily institutions that lend to individuals against
uncertain and distant future income returns. Hence market forces generally
lead to underinvestment in human capital, which a government may correct
by a variety of subsidies.
With respect to credit allocation to small firms, market failure may occur if
large firms have control over credit markets and apply bargaining power to
obtain loans on privileged terms. Also, large firms may get privileged terms if
they have an implicit government guarantee, in the sense that they are more
likely to be rescued by government than small firms. These cases can provide
a rationale for compensating government intervention to increase small firms'
access to credit. Likewise, if action to remove exchange rate overvaluation
and protection is ruled out, cheap export credit may be justified to offset higher
private profitability of domestic market sales.
Such arguments can be used to provide a justification for a functional or
horizontal industrial policy which is consistent with the principles of neoclasso as to make some consumer(s) better off and none worse off. Hence individual self-seeking
behavior by consumers and firms will not achieve the highest level of welfare for society as a
whole insofar as market failure is prevalent. It is predicted to occur in the presence of monopoly
and oligopoly, externalities, public goods, and common property resources, in each of which
individualistic behavior leads to suboptimal results. See Bannock, Baxter, and Rees 1978:287.
7 See World Bank 1987b.
INDUSTRIAL POLICY 13
sical economics. It is difficult to use those principles to justify a sectoral or
industry-specific industrial policy, however. A sectoral industrial policy aims
to direct resources into selected industries so· as to give producers in those
industries a competitive advantage. It therefore aims to produce a different
profile of industries compared to what would result from the decisions of unguided, unstimulated market agents on their own.
In the case of "sunrise" or "infant" industries such justification must rest,
in economic terms, on positive side-effects between firms or on economies of
scale. Positive interfirm side-effects, or "externalities," may take the form of
benefits created by a firm, in the form of goods, services, or technological
capacity, for which it is not fully compensated in market transactions. Or they
may take the form of benefits from an investment activity carried out by one
firm which are dependent upon complementary investments by other firms, in
which case present market prices are unlikely adequately to signal the interdependence that exists among these investment decisions. Externalities may
provide a prima facie case for intervention to offset suboptimal private investment through government coordination of investment decisions or incentives
for additional investment. Economies of scale, as distinct from externalities,
occur when costs per unit fall as output expands, generating an imperfectly
competitive industrial structure as those with lower costs are able to drive
others out of business. Here the prima facie case might call for government
ownership, or price regulation, or measures to help firms travel down their
declining cost curves. In "sunset" industries the justification has to rest on a
demonstration that the pace or patterns of decline generated by the market are
in some way socially inefficient or unacceptedly disruptive, for example, with
respect to employment.
However, neoclassical economics teaches that market failure of a kind that
could be improved upon by a sectoral industrial policy is rare. Indeed, some
economists think that there are few inherent market failures and that existing
market imperfections are often due to government actions which distort markets. They are pessimistic about politicians' willingness to resist the temptation to misuse economic powers, and still more pessimistic about the ability
of governments to detect opportunities which private entrepreneurs have
missed. Other neoclassical economists are more impressed by the infant industry case, but see the proper role of government as being confined to the
provision of nondiscriminatory and nondiscretionary industrial promotion policies for the set of infant industries. All would agree with Assar Lindbeck that
"in market-oriented economies, the role of government planning and public
finance is largely to 'plan' the physical, social and psychological environment
of private agents rather than to plan what these agents are supposed to do"
(1986:8). In most LDCS application of this principle would mean a shrinkage
in the size of government and a concentration of public attention on a much
14 CHAPTER 1
more limited set of tasks than at present. It would mean substantial economic
liberalization and privatization.
WHAT DOES the NEOCLASSICAL CONFIDENCE REST ON?
Theory
The source of these policy prescriptions can be traced to the theory of comparative advantage. This theory shows that anyone nation will be better off;i
in the sense of enjoying more output, if it concentrates on those activities inJ
which its costs are relatively cheapest. Resources will be so allocated provided
that international market forces are allowed to detennine the relative prices of
internationally tradable goods in the domestic economy. AndJb_at re9.!!i!emen!, in t.l:![D.,.-Calis . .forJree trade,..2r a close approximation to it, with low or
no impediments to imports and with relative prices that give no more incentive
to sell on the domestic market than to sell abroad.
However, th~_,,!heory._oLc.omparativ.e .. adyantag.e_c.oxeJ".IL.onLy-the .. effe.c.ts_of
once-an41Qr;illl..-,:,~<Inges i~JJ<!ge. restric.tions. ltdoes not specify a.c\l~ual
mech;;tism linkingrea"frzation of comparative adyagtage to higher growth. A
leading prop6ii"enfadmitnliiiP'ln"iis'pr~sent stat~, trade theory provides little
guidance as to the role of trade policy and trade strategy in promoting growth.
There is nothing in theory to indicate why a deviation from the optimum
should affect the rate of economic growth" (Krueger 1980:288). Attempts to
make the theory more dynamic, as in the stages of growth approach (Balassa
1981), show how changes in a country's educational and capital stocks produce changes in comparative advantage; but the changes in the stocks are
taken as independent of the changes in the pattern of industrial specialization.
Neoclassical economists instead rely on ad hoc factors to make the link between freer markets and higher growth. Some have suggested that opening the
economy to international competitive pressures assists technical change, economies of scale, and indigenous entrepreneurship. Others say that a free trade
regime is an important component of the task of confining government to its
proper place, by making it more difficult for the government to go beyond the
bounds of providing those public goods essential for civil life (Lal and Rajapatirana 1987; Krueger 1980). But these explanations shift the burden of argument well beyond the theory of comparative advantage to issues of government failure and technological change which have received little theoretical
attention within the neoclassical framework (Frans man 1986).8
Some economists, indeed, have recently provided theoretical arguments
which suggest that in certain conditions with real world analogues freer trade
may not promote faster growth. In conditions of increasing returns to scale
8 For a seminal attempt to theorize the connection between "dynamic comparative advantage"
and growth from a non-neoclassical perspective, see Chenery 1961.
INDUSTRIAL POLlCY 15
and imperfect world markets, a country's growth can be faster if it restricts
trade to some degree. Paul Krugman concludes that "the new thinking about
trade makes one thing clear: the idealized theoretical model on which the
classical case for free trade is based will not serve us anymore. The world is
more complex than that, and there is no question that the complexities do
open, In principle, the possibility of successful activist trade or industrial policy" (1986: 15). One could still argue that the qualifications only apply in conditions not found in most developing countries, so that for them the classical
case for free trade still holds. But to repeat, the classical case for free trade as
a means to higher growth is shaky.
Evidence
Yet even if the causal mechanisms linking nearly free trade with higher growth
are not well understood, neoclassical proponents claim that the record overwhelmingly supports such a connection. For example:
"The evidence is quite conclusive: countries applying outward-oriented development strategies had a superior performance in terms of exports, economic
growth, and employment whereas countries with continued inward orientation
encountered increasing economic difficulties." (Balassa 1981: 16--17)
"Detailed and historical studies have provided an impressive empirical validation of the theoretical case for the view that free trade remains the best
policy for developing (and developed) countries." (LaI1983:27-28)
"It seems to be as firm a stylized fact as any in the economics of developing
countries: a sustained movement to an outward-oriented trade regime leads to
faster growth of both exports and income." (Lal and Rajapatirana 1987:208)
• 'The case for liberalizing financial and trade control systems and moving back to
a nearly free trade regime is now incontrovertible." (Lal 1983:32)
"Experience has been that growth performance has been more satisfactory under
export promotion strategies than under import substitution strategies. There is little doubt about the link between export performance and
growth rates." (Krueger 1980:288-89)
, Jagdi~h:ahagwaii"i~probablY ,~orn,qtto_claimthat beliefin-thesuperiority of
----What fie Te'rms'''the "export promotion" strategy ove.r.JPJ',"iJ;npQrLsubstitution" strategy is all but unlversafamong"economists', "insofar as any kind of
consensus can ever be fou~di;;-·our'trib~;;"(l986:93Y.'What Bhagwati calls
export promotion, and others call outward orientation, is close to, though not
identical with, free trade, a point to which we return.
The consensus among economists has become the standard recipe of the
multilateral lending agencies, notably the International Monetary Fund (IMF)
and the World Bank. In a recent survey of the world economy, the IMF said:
16 CHAPTER 1
Another noteworthy element of policies in developing countries has been an increasing emphasis on a wide range of structural adjustment measures intended to
improve the allocation of resources and enhance growth prospects. For the most
part, these measures have involved restructuring prices and price-setting procedures towards a more market oriented approach. These measures are already
leading to more robust development in some countries, and should bring further
benefits as their efforts become more widely felt. (1985:66)
The World Bank, speaking of the same reforms, goes on to assure its clients
that, "however difficult the policy reforms may have been to adopt, in due
course they will create much easier conditions for developing countries than
would prevail if the reforms were diluted or abandoned" (1985:146). Both
statements express a ringing confidence in the force of the evidence.
Two kinds of evidence are involved. One is the cross-sectional study of the
relationship between "inward" and "outward" orientation, on the one hand,
and growth on the other, or some more dis aggregated specification thereof.
The other kind is the in-depth study of one or more key cases, the key cases
being the East Asian NICS, including South Korea, Taiwan, Hong Kong, and
Singapore. These are judged to have been the most successful developing
countries of the postwar era, and therefore to show what policies and other
conditions make for superior development performance. Most of the rest of
this book relates to the second kind of evidence. Here I discuss the first. The
question is whether the cross-sectional evidence provides good grounds for
confidence in the neoclassical prescriptions. I do not attempt to appraise a vast
literature, but merely to indicate grounds for doubt.
Bela Balassa has carried out much research on trade regimes which all
seems to support the neoclassical position. But Colin Bradford has applied to
one of Balassa's studies a simple test (1987). Taking Balassa's eight inwardoriented and six outward-oriented countries over the 1970s, Bradford found
their price distortion scores as calculated by the World Bank (1983). For the
overall price distortion index, as well as for the two most important components of the overall index (exchange rate distortion and interest rate distortion), he found virtually no difference between the averages for the outwardand the inward-oriented countries. There are also virtually no differences in
savings rate and growth of gross domestic product (GDP). To compare averages for such a small sample is admittedly a crude test. But the failure of
Balassa's argument to pass it is reason to reject his own conclusion that "the
evidence is quite conclusive: countries applying outward-oriented development strategies had a superior performance in terms of exports, economic
growth, and employment whereas countries with continued inward orientation
encountered increasing economic difficulties" (1981: 16-17).
Other studies have also found only a weak connection between export (or
outward) orientation and growth. Studies by Rostam Kavoussi (1985) and by
INDUSTRIAL POLICY 17
Hans Singer and Patricia Gray (1988) find export orientation to be positively
correlated with growth performance only when world demand is growing fast
(as in 1967-73). When world demand is growing slowly (as in 1973-77 and
1977-83) the correlation is slight, suggesting that in such conditions the gains
from export orientation are offset by negative effects. In all conditions of
world demand, the correlation is weaker for poorer countries than for richer
countries. 9
Howard Pack examines the connection between trade regimes and technical
efficiency to see whether export-oriented trade regimes go with faster growth
in technical efficiency, as some neoclassical economists have said. He concludes that, "there is no clear confirmation of the hypothesis that countries
with an external orientation benefit from greater growth in technical efficiency
in the component sectors of manufacturing" (1986:33). Even Jagdish Bhagwati, a long-time proponent of outward-orientation, has recently concluded
that there is little systematic evidence to support an outward orientation (which
he calls "export promotion") over import substitution on grounds of scale
economies, innovation, savings, or technical efficiency (1988:38-40).
The World Bank's 1987 World Development Report makes a more ambitious attempt to demonstrate the virtues of outward orientation. Going beyond
the dichotomy, it classifies forty-one developing countries into strongly outward-, moderately outward-, moderately inward-, and strongly inward-oriented categories for 1963-73 and 1973-85. It measures each category against
six criteria of macroeconomic performance for both periods, or twelve indicators in all. The theory predicts an even progression in performance, the
strongly inward doing worst, the strongly outward doing best.
The results show that the moderately inward-oriented countries do better
than the moderately outward-oriented countries on six out of twelve indicators. IO The moderately inward-oriented countries do better on inflation and
savings in both periods, and better on the rate of growth of GDP and of gross
9 See also Jung and Marshall 1985, who use n performance-based rather than incentive-based
measure of "export led." They find that at the level of indivjdual industries in thirty-seven countries output growth tends to cause export growth more frequently than the other way around,
including in Taiwan, Korea, and Brazil. On the export growth link, see also Havrylshyn and Alikhalli (1982:661): "the link between rapid growth of manufactured exports and rapid growth of
income is by no means automatic"; and Michaely (1977:52): "growth is affected by export performance only once countries achieve some minimum level of development"-perhaps under
"bad" import substitution policies. Presumably the effects of exports on overall growth in Taiwan, Korea, and Japan depended on their being manufactured exports, a point which is obscured
when countries are asked to adopt an export orientation per se.
10 The moderately outward-oriented countries include Brazil, Israel, Malaysia, Thailand, and
others. The moderately inward-oriented countries include El Salvador, Honduras, Kenya, Mexico, Nicaragua, the Philippines, Senegal, Yugoslavia, and others. The strongly inward-oriented
countries include Argentina, Bangladesh, the Dominican Republic, Ethiopia, Ghana, India, Peru,
Sudan, Tanzania, Zambia, and others. The named countries are all those which appear in their
category in both periods.
18 CHAPTER 1
national product (GNP) per capita in the more recent period. They do worse on
incremental capital output ratio and growth of manufactured exports in both
periods. For the more recent period, 1979-85, the moderately inward-oriented
do better on four indicators out of six, which is striking because the world
economy over the next ten to fifteen years will probably be more like the
1973-85 period than the 1963-73 period in terms of instability and low growth
of demand.
It is true that the averages for the strongly and moderately outward-oriented
cases combined are substantially better than those for the combined inwardoriented categories. But the strongly outward-oriented cases include only Korea, Hong Kong, and Singapore. (Taiwan was not in the sample because the
World Bank does not recognize Taiwan as a separate entity.) Since weighted
averages are used and since the size of the Korean economy swamps the other
two, the results from the strongly outward-oriented category are largely the
results from one country (except for export growth). Only anthropologists are
allowed to draw sweeping conclusions from a sample of less than two.
Besides which, it is striking that all the strongly outward-oriented cases
happen to be East Asian. This raises the possibility that the causes of exceptional macroeconomic performance have more to do with East Asianness, or
(absent Hong Kong) with industrial policies and managed trade, than with the
attributes of strongly outward orientation.
The results do appear to support the neoclassicallWorld Bank position with
respect to the strongly inward-oriented countries, which performed worse than
the moderately inward-oriented countries on all twelve indicators. But as Hans
Singer has shown, it cannot be concluded that they performed worse because
of their strongly inward orientation (1988). They are much poorer than the
moderately inward-oriented countries, with weighted average incomes little
more than half. Indeed, over the whole sample per capita income is a better
discriminator with respect to growth performance than trade regime. The results are consistent with the proposition that the poorer countries find it more
difficult and less beneficial to maintain an outward orientation, especially because of the negative effect on growth of greater instability of exports and
imports.
There is a further problem concerning the very meaning of the terms. Outward and inward orientation are taken to be near synonyms for free trade and
controlled trade, respectively. The World Development Report 1987 defines
strongly outward-oriented countries as those where' 'trade controls are either
nonexistent or very low. There is little or no use of direct controls and
licensing arrangements. " Defining the moderately outward-oriented category the report adds another criterion, namely, that the range of effective
protection rates to different sectors of the economy is low. By contrast, the
moderately inward-oriented category is defined to include countries where
"the overall incentive structure distinctly favors production for the domestic
INDUSTRIAL POLICY 19
market. The average rate of effective protection for home markets is relatively
high and the range of effective protection rates relatively wide. The use of
direct import controls and licensing is extensive " (l987a:82). The definitions make a sharp distinction between moderately outward- and moderately
inward-oriented cases, which makes the lack of difference in performance all
the more striking.
Korea is the prime example of a strongly outward-oriented country. But
does Korea really meet the criteria for a strongly outward-oriented case? Anticipating chapter 10, we can say here that several studies provide detailed
evidence of selective trade controls in Korea. The rate of effective protection
for manufacturing was 49 percent in both 1978 and 1982, according to Young
(1984), which is by no means low. The locus classicus of the view that Korea
has had a relatively free trade regime, the study by Larry Westphal and Kwang
Suk Kim (1982), suffers from serious methodological problems and uses data
from only one year, as long ago as 1968. And even if we take their findings at
face value, it turns out that Korea had nearly as much variation in effective
protection to different manufacturing sectors as Colombia, and more than Argentina; Korea had the second highest amount of interindustry dispersion in a
six-country comparison (Balassa, et a1.1982).11 This would preclude Korea's
classification as even a moderately outward-oriented country in the World
Bank study, because even moderately outward-oriented countries are defined
to have little variation in effective protection rates to different industries.
Without Korea the strongly outward-oriented category contains only the Hong
Kong and Singapore minnows.
Another source of cross-sectional evidence is the direct correlation between
price distortions and economic growth. The neoclassical argument says that
lower price distortions cause higher growth, while higher price distortions
have an adverse effect on growth. This argument was given prominence in the
World Bank's World Development Report 1983, which presents the results of
a correlation between price distortion scores and growth rates for a sample of
developing countries. The key finding is that "ttle average growth rate of those
developing countries with low distortions in the 1970s was about 7 percent a
year-2 percentage points higher than the overall average. Countries with
high distortions averaged growth of about 3 percent a year, 2 percentage points
lower than the overall average" (1983:61). This conclusion does not survive
even casual scrutiny. By far the most influential component of the overall
distortion index is the exchange rate. If exchange rate distortion is an important cause of poor growth performance, it should make its effect felt through
lagging export volume. But there is no statistically significant relationship between the growth of export volume and the exchange rate distortion index.
" The Korean study shows much the same problems as the companion Taiwan study by Lee
and Liang (1982), discussed in chapter 5. See also tables 3.2 and 3.3; and chapter 10.
20 CHAPTER 1
The same holds for the relationship between agricultural growth and the measured distortion of agricultural prices. "It is difficult to argue that one is measuring one's distortions right if they fail to exert an influence where they are
most relevant," Albert Fishlow observes. He concludes that "the widely publicized World Bank results are inadequately founded on a distortion index that
has limited analytic content" (1985: 140).
Indeed, some evidence suggests that deliberate "distortion" of some prices
may help growth. Colin Bradford compares the price of investment goods relative to consumption and government goods in a number of countries, and
finds that the newly industrialized countries are distinguished from other developing countries and from some of the poorer industrialized countries (such
as Hungary, Italy, and Spain) by a relatively low price of investment goods.
He concludes:
These patterns suggest the possibility that NIC growth and export performance in
manufactures have been accelerated by public policies that have lowered the cost
of investment goods. These policies could have been in the form of domestic
monetary policy affecting interest rates and credit allocations to industrial investors and borrowers or in the form of direct subsidies affecting the price of domcstically produced investment goods. Such monetary and fiscal policies would have
the effect of stimulating greater demand and supply of investment goods, which
in tum spurs capital accumulation, industrialization and structural change.
(1987:309)
What about the effects of trade liberalization, of a change in trade regime
toward free trade? Recall the claim that this relationship-of a sustained
change generating faster export and income growth-"seems to be as firm a
stylized fact as any in the economics of developing countries" (Lal and Rajapatirana 1987:208). But the evidence in support of the claim conflates the
effects of trade policy changes with the effects of commonly associated
changes, such as macroeconomic stabilization. One study which attempts to
identify the impact of trade liberalization alone concludes that "if truth-inadvertising were to apply to policy advice, each prescription for trade liberalization should be accompanied with a disclaimer: 'Warning! Trade Liberalization cannot be shown on theoretical grounds to enhance technical efficiency;
nor has it been empirically demonstrated to do so' " (Rodrik 1988b:28).
Even where individual countries improve their performance after a trade
liberalization (and certainly some do), this does not provide firm grounds for
recommending trade liberalization as a general prescription. For one thing, the
direction of causation is unclear. The liberalization of trade may itself be facilitated by the circumstances conducive to rapid growth, and even though
under such conditions it helps to reinforce that growth, this beneficial consequence of freer trade cannot be considered universal. Indeed, when the conditions for rapid growth are no longer given, the social and political conse-
INDUSTRIAL POLICY 21
quences of entering further liberalization could become destabilizing
(Bienefeld 1982:34). In conditions of widespread unemployment, trade tends
to become more adversarial than in full employment conditions. It becomes a
means of allocating unemployment across countries, as well as the more familiar means to balance the conunodity bundle in each economy (Drucker
1986).
There is also the fallacy of composition. Even the present very low levels
of developing country penetration of industrial country markets for manufactured goods have triggered protectionist reactions. Presumably these reactions
are part of the reason why the share of imports from developing countries in
relation to total consumption of manufactured goods in industrial countries has
remained roughly constant over the past decade, at just over 1.5 percent
(North-South 1980:176; UNCTAD 1988:4; Cline 1982b). If many more countries increased their share of industrial country markets without displacing existing suppliers-which is the implication of a generalized export orientation
strategy for developing countries-the protectionist reaction could be expected to be much stronger.
So the cross-country evidence relevant to neoclassical trade and industrial
policy prescriptions is by no means unambiguous. 12 It supports instead a cautious assessment of what a free trade approach can be expected to achieve for
a broad cross-section of developing countries. The state of the evidence is no
better than that on the effects of IMF stabilization programs, about which a
recent in-house survey concludes: "Little empirical evidence exists on the
long-run effects of Fund programs, and none at all on the effects of various
combinations of stabilization policies on economic development. Even
the informal evidence that is available is ambivalent on the relationship between financial stability and economic development" (Khan and Knight
1985:7, emphasis added)-this after nearly forty years of Fund stabilization
programs. The paper was published by the IMF in the same year as the earlierquoted IMF report on the world economy, which avowed that "these [marketoriented] measures are already leading to more robust development in some
12 Rati Ram (1986), using a very large sample of liS countries, finds that government size is
positively correlated with growth in almost all cases; that the externality effect of government size
is generally positive; that the positive effect of government size on growth may be stronger in
lower-income contexts; and several more relationships which confute for the neoclassical argument. I do not give much weight to these results. however. because I cannot specify a mechanism
by which such an aggregate variable as government size might have a positive impact on growth.
One needs to specify government activity in line with a theoretical reason why the specified activity might be important for growth. The difficulties of trying to base policy generalizations on
empirical evidence are seen in the fact that Ram (1986) and Landau (1986) published papers
within a few months of each other with diametrically opposed econometric findings about the
correlation between size of government in GNP and growth (Helleiner 1988). In formulating the
argument set out in the text. I have benefited from discussions with Manfred Bienefeld (see
Bienefeld 1988).
22 CHAPTER 1
countries and should bring further benefits as their efforts become more widely
felt. "
THE FREE MARKET (FM) THEORY OF EAST ASIAN SUCCESS
We tum now to the East Asian NICS. With some stretching of the category,
pre-I970 Japan can be included as well. An abundant literature attributes the
industrial success of the five NIcs-Japan, Taiwan, Korea, Hong Kong, and
Singapore-to their reliance on free markets. For example, Hugh Patrick declares himself to be "of the school which interprets Japanese economic performance as due primarily to the actions and efforts of private individuals and
enterprises responding to the opportunities provided in quite free markets for
commodities and labor. While the government has been supportive and indeed
has done much to create the environment of growth, its role has often been
exaggerated" (1977:239, emphasis added).I3 Referring to the five, Edward
Chen asserts that" state intervention is largely absent. What the state provided
is simply a suitable environment for the entrepreneurs to perform their functions." Such practices as "directing resources to the desired channels by state
intervention" are part of central planning and have no part in the development
of the East Asian five (1979:41, emphasis added). Hence, according to Chen,
the hypergrowth of the five demonstrates that "the free market environment
provides the necessary mechanism to gear the economies towards their optimal points on the production possibilities frontier" (1979: 185). Shifting into
even higher rhetorical gear, David Aikman claims that Taiwan and Hong Kong
"demonstrate just how faithful, consciously or not, the rulers of these two
countries have been to American conceptions of free enterprise" (1986: 116).
Milton and Rose Friedman, in Free to Choose, make the same point on a still
grander scale: "Malaysia, Singapore, Korea, Taiwan, Hong Kong and Japan-relying extensively on private markets-are thriving. By contrast,
India, Indonesia, and Communist China, all relying heavily on central planning, have experienced economic stagnation" (1980:57).
According to this "free market" (FM) theory, East Asia does better than
other newly industrializing countries because the East Asian state interferes
hardly at all in the working of the market. The other countries have been held
back from the development they would have achieved in the "normal" course
of events by excessive state intervention, especially in foreign trade. Or in a
slightly different version of the argument, John Fei claims that "the basic
causation of success of the [East Asian] NICS on the policy front, can be traced
to the lessening of government interferences in the market economy during the
I) David Henderson (1983:114) claims that "the real explanation for the Japanese economic
miracle is the country's laissez-faire policies on taxes, antitrust, banking and labor," which
prompts Chalmers Johnson to wonder whether Henderson understands French (1985:3). See also
Hosomi and Okumura 1982:150.
INDUSTRIAL POLICY 23
E-O [Export-Oriented] phase. In Taiwan and Korea, interference with the market was considerably less as compared to other worse offenders in the near
NICS and the Latin American countries (1983:34, emphasis added).
THE SIMULATED FREE MARKET (SM) THEORY OF EAST ASIAN SUCCESS
Some neoclassical economists conclude that the governments of East Asia did
more than just liberalize markets and lower distortions. In their view the governments also intervened more positively to offset other distortions, both those
caused by other policies (e.g., import controls) and those remaining from government failure to change distortion-inducing institutions directly (e.g., segmented financial markets). Frederick Berger states the argument as follows: "I
believe that the crux of the Korean example is that the active interventionist
attitude of the State has been aimed at applying moderate incentives which are
very close to the relative prices of products and factors that would prevail in
a situation offree trade. . It is as though the government were 'simulating'
a free market" (1979:64, emphasis added).
This is similar to Gary Saxonhouse's argument (1985) that Japan's peculiar
institutional features are merely the functional equivalent of different arrangements in other countries. Its industrial policy, for example, is but a substitute
for information which is provided by better-developed capital markets in the
West.
Jagdish Bhagwati endorses a further type of government intervention in
support of what he calls the "export promotion" (EP) strategy. An EP strategy
is a set of policies which results in the average effective exchange rate for
importables being approximately equal to that for exportables. The most important thing the government of an underdeveloped country can do to promote
growth, he implies, is to maintain an EP strategy, and this requires government
intervention. "The Far Eastern economies (with the exception of Hong Kong)
and others that have come close to the EP strategy have been characterized by
considerable government activity in the econ<?mic system. In my judgement,
such intervention can be of great value, and almost certainly has been so, in
making the EP strategy work successfully" (1988:33). However, Bhagwati's
desirable interventions are restricted to those which increase producers' confidence in the government's commitment to the EP strategy. "By publicly supporting the outward-oriented strategy, by even bending in some cases towards
ultra-export promotion, and by gearing the credit institutions to supporting
export activities in an overt fashion, governments in these [Far Eastern] countries appear to have established the necessary confidence that their commitment to the EP strategy is serious, thus inducing firms to undertake costly investments and programs to take advantage of the EP strategy" (p. 34). He
mentions in passing that the EP strategy does not preclude import substitution
in selected sectors but gives no attention to this combination.
22 CHAPTER I
countries and should bring further benefits as their efforts become more widely
felt. "
THE FREE MARKET (FM) THEORY OF EAST ASIAN SUCCESS
We tum now to the East Asian NICS. With some stretching of the category,
pre-1970 Japan can be included as well. An abundant literature attributes the
industrial success of the five NIcs-Japan, Taiwan, Korea, Hong Kong, and
Singapore-to their reliance on free markets. For example, Hugh Patrick declares himself to be "of the school which interprets Japanese economic performance as due primarily to the actions and efforts of private individuals and
enterprises responding to the opportunities provided in quite free markets for
comltJodities and labor. While the government has been supportive and indeed
has done much to create the environment of growth, its role has often been
exaggerated" (1977:239, emphasis added).!3 Referring to the five, Edward
Chen asserts that "state intervention is largely absent. What the state provided
is simply a suitable environment for the entrepreneurs to perform their functions." Such practices as "directing resources to the desired channels by state
intervention" are part of central planning and have no part in the development
of the East Asian five (1979:41 , emphasis added). Hence, according to Chen,
the hypergrowth of the five demonstrates that "the free market environment
provides the necessary mechanism to gear the economies towards their optimal points on the production possibilities frontier" (1979: 185). Shifting into
even higher rhetorical gear, David Aikman claims that Taiwan and Hong Kong
"demonstrate just how faithful, consciously or not, the rulers of these two
countries have been to American conceptions of free enterprise" (1986: 116).
Milton and Rose Friedman, in Free to Choose, make the same point on a still
grander scale: "Malaysia, Singapore, Korea, Taiwan, Hong Kong and Japan-relying extensively on private markets-are thriving. By contrast,
India, Indonesia, and Communist China, all relying heavily on central planning, have experienced economic stagnation" (1980:57).
According to this "free market" (FM) theory, East Asia does better than
other newly industrializing countries because the East Asian state interferes
hardly at all in the working of the market. The other countries have been held
back from the development they would have achieved in the "normal" course
of events by excessive state intervention, especially in foreign trade. Or in a
slightly different version of the argument, John Fei claims that "the basic
causation of success of the [East Asian] NICS on the policy front, can be traced
to the lessening of government interferences in the market economy during the
13 David Henderson (1983:114) claims that "the real explanation for the Japanese economic
miracle is the country's laissez·faire policies on taxes, antitrust, banking and labor," which
prompts Chalmers Johnson to wonder whether Henderson understands French (1985:3). See also
Hosomi and Okumura 1982:150.
INDUSTRIAL POLICY 23
E-O [Export-Oriented] phase. In Taiwan and Korea, interference with the market was considerably less as compared to other worse offenders in the near
NICS and the Latin American countries (1983:34, emphasis added).
THE SIMULATED FREE MARKET (SM) THEORY OF EAST ASIAN SUCCESS
Some neoclassical economists conclude that the governments of East Asia did
more than just liberalize markets and lower distortions. In their view the governments also intervened more positively to offset other distortions, both those
caused by other policies (e.g., import controls) and those remaining from government failure to change distortion-inducing institutions directly (e.g., segmented financial markets). Frederick Berger states the argument as follows: "I
believe that the crux of the Korean example is that the active interventionist
attitude of the State has been aimed at applying moderate incentives which are
very close to the relative prices of products and factors that would prevail in
a situation offree trade. It is as though the government were 'simulating'
a free market" (1979:64, emphasis added).
This is similar to Gary Saxonhouse' s argument (1985) that Japan's peculiar
institutional features are merely the functional equivalent of different arrangements in other countries. Its industrial policy, for example, is but a substitute
for infonnation which is provided by better-developed capital markets in the
West.
Jagdish Bhagwati endorses a further type of government intervention in
support of what he calIs the "export promotion" (EP) strategy. An EP strategy
is a set of policies which results in the average effective exchange rate for
importables being approximately equal to that for exportables. The most important thing the government of an underdeveloped country can do to promote
growth, he implies, is to maintain an EP strategy. and this requires government
intervention. ''The Far Eastern economies (with the exception of Hong Kong)
and others that have come close to the EP strategy have been characterized by
considerable government activity in the economic system. In my judgement,
such intervention can be of great value, and almost certainly has been so, in
making the EP strategy work successfully" (1988:33). However, Bhagwati's
desirable interventions are restricted to those which increase producers' confidence in the government's commitment to the EP strategy. "By publicly supporting the outward-oriented strategy, by even bending in some cases towards
ultra-export promotion, and by gearing the credit institutions to supporting
export activities in an overt fashion, governments in these [Far Eastern] countries appear to have established the necessary confidence that their commitment to the EP strategy is serious, thus inducing finns to undertake costly investments and programs to take advantage of the EP strategy" (p. 34). He
mentions in passing that the EP strategy does not preclude import substitution
in selected sectors but gives no attention to this combination.
24 CHAPTER 1
This "simulated free market" (SM) theory differs from the FM theory in
tenns of the distinction between a free (or liberal) trade regime and a neutral
trade regime. The fanner is one with no or few impediments to imports; the
latter is one where any incentive for domestic producers to sen on the domestic
market rather than export, because of protection, is offset by export subsidies.
This means that, overall, a U.S. dollar of exports fetches, in local currency,
the same as a U.S. dollar of imports, when all export subsidies and tax credits
and all import premia resulting from quantitative restrictions and tariffs are
included. So a neutral trade regime may go with some government intervention, including protection of the domestic market. The important point, according to this theory, is that the incentive effect of such protection in biasing
sales toward the domestic market should be offset, in aggregate, by export
promotion measures. The Far Eastern countries have managed to do this, according to Bhagwati, which is a large part of the reason why they have been
so successful compared to others which have not.
However, the proponents of the SM view have shown little interest in analyzing the nature of government intervention in East Asia, though they recognize its existence. And they also place primary causal weight on the character of the trade regime for explaining economic perfonnance. For both
reasons, the SM theory can be considered a variant of the core neoclassical
theory, which links economic success to self-adjusting markets.
THE GOVERNED MARKET (GM) THEORY OF EAST ASIAN SUCCESS
Over the past decade or so, another stream of literature has emphasized the
directive role of the state in East Asia. Parvez Hasan,14 for example, writing
of South Korea, draws attention to an "apparent paradox":
[T]he Korean economy depends in large measure on private enterprise operating
under highly centralized government guidance. In Korea the government's role is
considerably more direct than that of merely setting the broad rules of the game
and in influencing the economy indirectly through market forces. In fact, the government seems to be a participant and often the determining influence in nearly
all business decisions. (1976:29)
Edward Mason and associates come to a similar conclusion in their study of
government-business relations:
The rapid economic growth that began in South Korea in the early 1960s and has
accelerated since then has been a government-directed development in which the
principal engine has been private enterprise. The relationship between a government committed to a central direction of economic development and a highly dynamic private sector that confronts the planning machinery with a continually
changing structure of economic activities presents a set of interconnections diffi-
'4 A World Bank staff member.
INDUSTRIAL POLlCY 25
cult to penetrate and describe. Planning in South Korea, if it is interpreted to
include not only policy formulation but also the techniques of policy implementation, is substantially more than "indicative." The hand of government reaches
down rather far into the activities of individual firms with its manipulation of
incentives and disincentives. At the same time, the situation can in no sense be
described in terms of a command economy. (1980:254, emphasis added)
Much the same has been said of Japan. According to this interpretation, the
Japanese were the first to recognize that international competitive advantage
could be deliberately created by government not just to nurture a few infant
industries to supply the domestic market but to push broad sets of industries
toward areas of growth and technological change in the world economy. In
the words of a vice-minister of MITI (Ministry of International Trade and Industry):
The MITl decided to establish in Japan industries which require intensive employment of capital and technology, industries that in consideration of comparative
cost of production should be the most inappropriate for Japan, industries such as
steel, oil-refining, industrial machinery of all sorts, and electronics. From a
short-run, static viewpoint, encouragement of such industries would seem to conflict with economic rationalism. But, from a long-range viewpoint, these are precisely the industries where income elasticity of demand is high, rechnological
progress is rapid, and labor productivity risesfas!. It was clear that without these
industries it would be difficult to employ a population of 100 million and raise
their standard of living to that of Europe and America with light industries;
whether right or wrong, Japan had to have these heavy and chemical industries.
Fortunately, owing to good luck and wisdom spawned by necessity, Japan
has been able to concentrate irs scant capital in strategic industries.
(OECD 1972:15, emphasis added)
Henry Rosovsky went so far as to say of Japan that it "must be the only capitalist country in the world in which the Government decides how many firms
should be in a given industry and sets out to arrange the desired number"
(1972:244).
Chalmers Johnson has sketched a model of the "capitalist developmental
state," based on the institutional arrangements common to the high-growth
East Asian capitalist countries (1981, 1982, 1983). These arrangements are
characterized, he says, by the following features: IS
1. The top priority of state action, consistently maintained, is economic development, defined for policy purposes in terms of growth, productivity, and competitiveness rather than in terms of welfare. The substance of growthlcompetitiveIS I have slightly revised the presentation of Johnson's model and have omitted one of his
points, that the state supervises a heavy and consistent investment in education for all the people
(which relates to investment allocation rather than to organizational arrangements).
26 CHAPTER I
ness goals is derived from comparisons with external reference economies which
provide the state managers with models for emulation.
2. The state is committed to private property and the market and limits its interventions to conform with this commitment.
3. The state guides the market with instruments formulated by an elite economic
bureaucracy, led by a pilot agency or "economic general staff."
4. The state is engaged in numerous institutions for consultation and coordination
with the private sector, and these consultations are an essential part of the process
of policy formulation and implementation.
5. While state bureaucrats "rule," politicians "reign." Their function is not to
make policy but to create space for the bureaucracy to maneuver in while also
acting as a "safety valve" by forcing the bureaucrats to respond to the needs of
groups upon which the stability of the system rests: that is, to maintain the relative autonomy of the state while preserving political stability. This separation of
"ruling" and "reigning" goes with a "soft authoritarianism" when it comes to
maintaining the needs of economic development vis-a.-vis other claims, and with
a virtual monopoly of political power in a single political party or institution over
a long period of time.
This picture of a centralized state interacting with the private sector from a
position of preeminence so as to secure development objectives has been
called the "developmental state" theory of East Asian industrial success
(Johnson 1982; White 1988). It is not, however, much of a theory. Its specification of institutional arrangements is descriptive rather than comparativeanalytic, so what the developmental state is contrasted with is not clear. It also
says little about the nature of policies and their impact on industrial performance. Indeed, Johnson's institutional arrangements are for the most part as
consistent with simulated free market policies as with more directive ones. I
now propose a "governed market" theory which builds on both the idea of
the developmental state and on the older development economics' understanding of the nature of the development problem.
The governed market (GM) theory says that the superiority of East Asian
economic performance is due in large measure to a combination of: (1) very
high levels of productive investment, making for fast transfer of newer techniques into actual production; (2) more investment in certain key industries
than would have occurred in the absence of government intervention; and (3)
exposure of many industries to international competition, in foreign markets
if not at home. These are the proximate causes. At a second level of causation,
they are themselves the result, in important degree, of a set of government
economic policies. Using incentives, controls, and mechanisms to spread risk,
these policies enabled the government to guide~r govern-market processes
of resource allocation so as to produce different production and investment
outcomes than would have occurred with either free market or simulated free
INDUSTRIAL POLICY 27
market policies. At the third level of explanation, the policies have been permitted or supported by a certain kind of organization of the state and the private sector. Let us specify the policies and the organizational arrangements in
more detail.
Johnson's picture of the developmental state can be recast to fit with concepts developed elsewhere in political science for comparing political regimes. The relevant distinctions are "democratic versus authoritarian" and
"pluralist vs. corporatist." The first refers to the rules by which rulers are
chosen. In democratic regimes the rulers are chosen by a process much influenced by popular preferences, while in authoritarian regimes they are selected
by methods which give relatively little scope for the expression of popular
sentiment. The second distinction refers to relations between interest groups
and the state. In pluralist regimes, interest groups are voluntary associations,
free to organize and gain influence over state policy corresponding to their
economic or political resources. The process of government consists of the
competition between interest groups, with government bureaucracies playing
an important but not generally dominant role. In corporatist systems the state
charters or creates a small number of interest groups, giving them a monopoly
of representation of occupational interests in return for which it claims the
right to monitor them in order to discourage the expression of "narrow," conflictful demands. The state is therefore able to shape the demands that are
made upon it, and hence-in intention-maximize compliance and cooperation (Schmitter 1974; Stepan 1978; Zeigler 1988).
In these terms, the United States is the example par excellence of a pluralist
democracy; Korea and Taiwan are examples of authoritarian corporatism;
Austria and Switzerland illustrate democratic corporatism; and Japan illustrates corporatism combined with arrangements for selecting rulers which are
intermediate between democratic and authoritarian, or what Johnson calls
"soft authoritarian."
The corporatist and authoritarian political arrangements of East Asia have
provided the basis for market guidance. Market guidance was effected by augmenting the supply of investible resources, spreading or "socializing" the
risks attached to long-term investment, and steering the allocation of investment by methods which combine government and entrepreneurial preferences.
In particular, the governments guided the market by: (I) redistributing agricultural land in the early postwar period; (2) controlling the financial system
and making private financial capital subordinate to industrial capital; (3) maintaining stability in some of the main economic parameters that affect the viability of long-term investment, especially the exchange rate, the interest rate,
and the general price level; (4) modulating the impact of foreign competition
in the domestic economy and prioritizing the use of scarce foreign exchange;
(5) promoting exports; (6) promoting technology acquisition from multinational companies and building a national technology system; and (7) assisting
28 CHAPTER I
particular industries. (For Japan post-1970173 we would also have to include
industry-specific policies to ease decline. Throughout the reference is to Japan
before this time. 16)
I am especially interested in the policies to assist particular industries. This
is not because I think that industry-specific policies were causally more important than the others. But they were important enough, and yet have been
almost completely ignored in most of the economics literature about the Taiwanese and Korean "economic miracles." Neglect of these policies matters
particularly because it is in the histories of specific industries that one can most
clearly see the government in action.
However, the existence of sectoral policies does not in itself mean that they
produced significantly different outcomes from free market or simulated free
market policies. They might merely put the government's seal of approval on
some private sector projects by way of mild assistance for something that private firms would have done anyway in response to price signals alone. In that
case we could dismiss sectoral policies as mere "hand-waving" or "windowdressing. "
To clarify the issue, let us distinguish between leading the market and following the market. Sectoral policies lead the market when the government
takes initiatives about what products or technologies should be encouraged,
and puts public resources or public influence behind these initiatives. A clear
case is where the government proposes a project to private firms, the private
firms decline, and the government goes ahead through a public enterprise. On
the other hand, sectoral policies follow the market when the government
adopts the proposals of private firms about new products and new technologies. If private firms propose to make the quantum jump from fabrication of
16K to 64K DRAM chips and ask for government assistance, then government
assistance follows the market.
Leading and following should be qualified by the degree of additionality.
When government helps firms to do what they would have done anyway, this
is-with apologies to the English language-"small followership." When
government assists firms significantly to extend the margin of their investments, this is "big followership." We can use "big leadership" to refer to
government initiatives on a large enough scale to make a real difference to
investment and production patterns in an industry, and "small leadership" to
refer to government initiatives which on their own carry too few resources or
too little influence to make a difference.
The FM and SM theories of East Asian industrial success can accommodate
the fact of sectoral policies by saying or implying that they constitute merely
"small followership." The GM theory says that the governments' industry-
" By 1970-73 Japan's protection-promotion system began to be substantially dismantled.
These years mark the end of an era-the end of catch-up, the beginning of large balance-ofpayment surpluses, and a shift in industrial policy from growth, industrial investment, and export
objectives, toward freer trade and social-overhead investment.
INDUSTRIAL POLICY 29
specific policies went beyond "small followership," to either or both "big
leadership" and' 'big followership." Interventions of these types suggest that
the production and investment outcomes were different from what would have
occurred with free market or simulated free market policies, the difference
being greater for big leadership than for big followership. The fact of a difference does not in itself imply that the difference helped or hindered development. Whether it helped or hindered has to be established independently.
Let us now summarize the main differences between the GM theory and the
other two. The PM and SM theories emphasize efficient resource allocation as
the principal general force for growth, and therefore interpret superior East
Asian performance as the result of more efficient resource allocation than in
other LDCS or NICS. This more efficient resource allocation comes from more
freely functioning markets, including closer integration of domestic product
markets into international markets. Hence these countries show the virtues of
"getting the prices right," where "right" means domestic prices in line with
international prices. The GM theory, on the other hand, emphasizes capital
accumulation as the principal general force for growth, and interprets superior
East Asian performance as the result of a level and composition of investment
different from what FM or SM policies would have produced, and different,
too, from what the "interventionist" economic policies pursued by many
other LDCS would have produced. Government policies deliberately got some
prices "wrong," so as to change the signals to which decentralized market
agents responded, and also used nonprice means to alter the behavior of market agents. The resulting high level of investment generated fast turnover of
machinery, and hence fast transfer of newer technology into actual production.
The FM and SM theories are silent on the political arrangements needed to
support their policies. The GM theory emphasizes the developmental virtues of
a hard or soft authoritarian state in corporatist relations with the private sector,
able to confer enough autonomy on a centralized bureaucracy for it to influence resource allocation in line with a long-~erm national interest17-which
sometimes conflicts with short-run profit maximizing. The state's steering of
resource allocation is the economic counterpart to its political restrictions on
"free trade" in interest groups.
THE QUESTION OF EVIDENCE
There is an unavoidable ideological loading in debate about the role of the
state, for the issues lie uncomfortably close to the heart of the ideological
11 Many writers in the tradition of political pluralism deny any meaning to the notion of the
general good or the national interest, other than whatever happens to be the balance of demands
between competing interest groups. I use "national economic objectives" or "the national interest" to refer to interests or objectives that are broadly rather than narrowly shared and enduring
rather than short-term. Cf. Dahl and Lindblom 1963:501.
30 CHAPTER 1
dispute in which the superpowers have clothed their global rivalry. There are
also disciplinary interests at stake, as each discipline or subdiscipline tries to
emphasize the importance of that factor-free markets, the state, culture, or
something else-in which it can claim a comparative analytical advantage.
But serious differences of opinion among scholars remain, which reflect the
methodological difficulties with any theory which posits government steerage
of the market as an independent variable. In part, the difficulties stem from
the absence of an economic theory of sufficient credibility to provide a legitimate base from which technical economic analysis can act as a constraint on
admissible arguments. In part, also, the difficulties stem from the same problem as with virtually all interesting social science questions: the absence of a
counterfactual. The question of whether measures designed by a well-meaning
bureaucracy can achieve results superior to those which a more liberal market
system would produce is impossible to answer conclusively; what would have
happened in the absence of the intervention is always unknown. Nonetheless,
economists from Adam Smith onwards have not hesitated to make strong assertions, both positive and negative, about the effectiveness of government
intervention without offering serious evidence to support their claims.
Serious direct evidence would have to separate out the impact of industrial
policies from that of macroeconomic and other policies. Macroeconomic policies affect aggregate demand, but they also affect different industries differently although not intended to produce such differential effects. Industrial policies, on the other hand, are intended to affect production and investment
decisions of decentralized producers. And industrial policies, as we have seen,
come in two broad kinds-functional and sectoral-whose respective impacts
also need to be distinguished. In addition to macroeconomic and industrial
policies, there are also public goods policies, or what Adam Smith called the
three "duties of the sovereign" --defense, law and order, and physical infrastructure. We should further distinguish policies aimed at changing income
and asset distributions, whose results then affect the political feasibility of
different macropolicies, especially ones for adjusting to economic austerity.
The relationships among macroeconomic, industrial, public goods, and distributional policies is best thought of in terms of overlapping circles. The effects
of one set are highly contingent upon the effects of the others. In particular,
macroeconomic fiscal and monetary policies affect the general thrust of government economic actions toward growth and competitiveness vis-a.-vis consumption and redistribution, and so have an important bearing on both the
speed and composition of industrial growth. 18
To assess the impact of industry-specific policies, one obvious method
would be to take a number of industries and examine case by case the connection between promotion measures and subsequent growth. But the problems
" For example, Scott 1985.
INDUSTRIAL POLICY 31
include holding other things constant between high-assistance and low-assistance industries; commensurating different types of assistance (e.g., subsidies, protection, antitrust exemptions); and obtaining information on promotional measures which are not contained in financial disbursements or legal
directives (such as the "announcement effect" of loans from the Japan Development Bank on the direction of commercial bank lending, which enables
the Japan Development Bank to influence credit allocation with only a small
amount of lending). Then there is the problem of interpretation: if industries
which receive a lot of assistance grow more slowly than those which do not,
does this indicate the failure of assistance or does it indicate targeting of industries that need assistance as a condition of subsequent fast growth? And
always one needs to make an assumption about what would have happened in
the absence of government help. Even for as nicely specific a policy as performance requirements on foreign direct investment (such as export ratios,
local content conditions, requirements for ownership, and employment based
on national origin) it has proved impossible to answer the question of how the
requirements affect trade and investment. Assumptions have to be made about
what other countries do when one country raises or lowers its performance
requirements. The requirements may be redundant in that investors would
meet them anyway or they may not be enforced. 19
Again, take selective credit policies, used by many governments to steer
resources into certain uses. In order to be effective, selective credit policies
have to cause a net increase in credit to the priority use, and this net increase
has to bring about a reallocation of real resources. To cause a net increase, the
government or the central bank has either effectively to control and monitor
the behavior of financial institutions or provide them with incentives to raise
lending for the priority use. To bring about a change in resource allocation,
the priority borrowers should not simply substitute cheap credit through the
concessional scheme for more expensive credit which they would have obtained without the scheme, without altering their production plans. Finally,
the effectiveness of selective credit needs to be weighed against the costs, in
terms of the effects on other participants in financial markets of raising the
extra revenue needed to pay for the subsidy. Obtaining information on these
three sets of conditions for success is difficult, to put it mildly. 20
Even if studies of selective intervention showed effectiveness at the industry
level, they would still leave open the question of whether the population as a
whole would have been better off in welfare terms developing industries other
than those particular ones. There is a cost to selective promotion, which is the
cost of diverting resources from currently profitable production (e. g., textiles)
to production that might be profitable in the future (e.g., sophisticated elec19 See Guisinger 1986:170.
20 See Ghamen and Rajaram 1987, for Korea and Tunisia.
32 CHAPTER 1
tronics). Industry optimality does not establish national optimality, and national optimality may be defined not only in terms of present or future consumption but also in terms of the competitive strength of national industries in
relation to other countries. It is tempting to use aggregate production function
analysis to estimate the extent of "national" optimality, with the size of the
residuals indicating the maximum possible extent of the government's contribution. The problem is that the size of the residuals depends on how the production function is specified, which is a matter of very subjective judgment.
The • 'national" issue can be got at another way, by comparing countries
which in many important respects are similar but where the role of government
has been significantly different (Japan and Italy in the postwar period, for
example [Boltho 1981]). The hazards are obvious.
Faced with the manifold difficulties in determining the economic effects of
government attempts to steer the market, we can use more indirect evidence
to take the debate forward.
First, we can establish the extent to which the key neoclassical growth conditions have been present over time: to what extent trade has been free, the
exchange rate in equilibrium, the labor market competitive, and interest rates
high enough to reflect the real scarcity of capital. In other words, we can assess
the degree of price "distortions" in different product and factor markets. The
neoclassical (both PM and SM) presumption is that the lower these distortions,
the faster the growth; so high-growth East Asia should show low distortions.
Second, we can get evidence of sectoral industrial policies by examining
the histories of particular industries to see what kinds of activity the government was undertaking. We need to know how much control it exercised over
investment decisions, and to what extent it was responsible for taking initiatives about products or production processes that private firms would not have
undertaken at about the same time without assistance. The GM theory leads us
to expect big followership or big leadership in some important industries.
Third, if the government is to exercise leadership it must have instruments
for affecting investment decisions. The instruments might include trade controls, foreign exchange controls, export incentives, selective credit allocation,
tax incentives, public enterprises, as well as other means of punishing firms
that do not comply. Such instruments are needed for getting prices "wrong"
and in other ways altering market behavior. If we do not find such instruments
we can discount claims that the government has an important role. Of course
it is not enough to establish that the instruments exist; they may exist on paper
only, or they may dedicate only a trivial amount of resources, or they may
entail too little change in costs and prices to have a resource-pulling effect. By
looking at the "input" side we can get some sense of whether it is plausible
to say that the policies have significant effects on the' 'output" side. The GM
theory leads us to expect substantial variation between industries in the effects
INDUSTRIAL POLICY 33
of government policies on relative prices, corresponding to the objectives of
government promotional activity.
Fourth, we also have to be able to identify the institutional locus of the
instruments--one or more central agencies vested with the powers to plan and
coordinate within parts of the economy and with some responsibility for industrial success. Likewise we should find evidence of national goal-setting,
going beyond the practice of Anglo-American economies and the principles
of neoclassical economics, with goals relating substantively to industrial structure and international competitiveness as well as to macroeconomic balance,
adequate market infrastructure, and "fair" competition.
Finally, the configuration of agencies, national goals, and industrial policy
instruments is likely to be more effective where political power is relatively
unified, and unified around groups of people who are committed to industrialization. This is what the GM theory predicts. If instead we find that political
power is distributed in a pluralistic or fragmented way, with different constellations of domestic and foreign interest groups exerting pressure on different
public policy issues, or if we find that political power is unified around groups
of people whose interests are hostile to industrialization, then whatever the
government claims to be the case we can discount the argument that its planning and coordination have helped industrialization. Relatedly, the evidence
for the GM theory is stronger insofar as we can show why the "needs of political survival" are in line with the "needs of productive investment" rather
than opposed-or why the political leaders do not adopt a Marcos- or Mobutulike strategy of plundering the Treasury and pulverizing state-implementing
agencies in disregard of the consequences for economic development. This
involves the political calculations that shape industrial policies.
And that, roughly, is the sequence of chapters 3 to 9. Chapter 10 then brings
information and argument from these earlier chapters together with material
from Korea, Japan, and Hong Kong to address the question of how important
governing the market has been in capitalist East Asia. Chapter 11 goes on to
ask what lessons might be drawn from the East Asian experience for development strategy in other developing countries.
But before all this, we need to examine economic performance in East Asia,
particularly Taiwan's. How good has it been?