The Rise and Fall of the Washington Consensus as a Paradigm for Developing Countries
1. INTRODUCTION
Developing countries is an international practice. The essence of this practice is the mobilization and allocation of resources, and the design of institutions, to transform national
economies and societies, in an orderly way, from a state and status of being less developed
to one of being more developed. The agencies engaged in this practice include national
governments of less-developed countries, which have adopted ``development'' as a purpose to
which State power is put, and governments of richer countries, which disburse ocial development aid to support and in¯uence this process; a variety of non-governmental organizations concerned to animate and channel popular concerns; and international intergovernmental organizations, such as the organs of the United Nations and the World Bank, many of which have been expressly set up to resolve various development problems. Often it is the last group who have acted as the avant-garde of development practice. It is because of their activities, as well as the widespread tendency of governments to copy successful practice elsewhere, that it is appropriate to describe developing countries as an international practice. But it is by no means global in scope. Indeed the practice of developing countries is only done in a particular set of countries those which in the 1950s and 1960s were generally called ``underdeveloped'' or ``less developed'' countries, but which now generally identify themselves, and are identi®ed by others, as ``developing countries.''
This paper discusses trends in the body of knowledge which guides and justi®es the practice
of development. It examines, in particular, the ideas propagated by international development
agencies, and focuses on the shift in thinking which occurred in the 1980s with the
introduction and widespread adoption of an approach to the practice of developing countries
known as the ``Washington Consensus.'' In broad terms, this approach recommends that
governments should reform their policies and, in particular: (a) pursue macroeconomic
stability by controlling in¯ation and reducing ®scal de®cits; (b) open their economies to the
rest of the world through trade and capital account liberalization; and (c) liberalize
domestic product and factor markets through privatization and deregulation. Propagated
through the stabilization and structural adjustment policies of the International
Monetary Fund (IMF) and World Bank, this has been the dominant approach to development
from the early 1980s to the present. The paper examines the introduction of the Washington
Consensus as a paradigm shift, and assesses the con®guration of development
thinking in the 1990s and pressures for a further paradigm shift, particularly in the light
of the East Asian ®nancial crisis and recent attempts to construct a ``post-Washington
Consensus.'' The paradigmatic nature of the Washington Consensus is most clearly evident in the work of John Williamson (1990,1993,1997), who coined the name and also set out a speci®c formulation of the approach at the end of the 1980s. This formulation was founded on an attempt to summarize, with particular reference to policy reform in Latin America, ``the
conventional wisdom of the day among the economically in¯uential bits of Washington,
meaning US government and the international ®nancial institutions'' (Williamson, 1993, p.
1329). Williamson never explicitly identi®es the Washington Consensus as a paradigm. But the way he describes the approach conforms in many respects with Thomas KuhnÕs notion of
one. 1 Thus, he argued that the Washington Consensus is a ``universal convergence,'' and
that it constitutes ``the common core of wisdom embraced by all serious economists'' (Williamson,1993, p. 1334). He codi®ed the approach as a set of 10 axiomatic generalizations which, given certain values, are generally shared by scholars and practitioners concerned with
economic growth in developing countries; and he listed remaining analytical problems on
which normal economic science needs to focus Finally, he dismissed those who challenged the consensus view as ``cranks'' (p. 1330). As he put it, [T]he superior economic performance of countries that establish and maintain outward-oriented market economies subject to macro-economic discipline is essentially a positive question. The proof may not be quite as conclusive as the proof that the Earth is not ¯at, but it is suciently well established as to give
sensible people better things to do with their time than to challenge its veracity (p. 1330).
The structure of the revolution in thinking which occurred with the introduction of
Washington Consensus policies is usually seen as a shift from state-led dirigisme to marketoriented policies. Such a switch undoubtedly occurred. But it is not a sucient description of the nature of the change as a paradigm shift. As Kuhn shows, when paradigms change, there are usually signi®cant changes in the ``methods, problem-®eld, and standards of solution'' which are accepted by a community of practitioners (Kuhn, 1970, p. 103). As a consequence, ``the proponents of competing paradigms practice their trades in di€erent worlds...[they] see di€erent things when they look from the same point in the same direction'' (p. 150). In examining the introduction of the Washington Consensus as a paradigm shift, what matters is not simply the substantive di€erences with earlier approaches, but also the nature of the change in the disciplinary matrix and worldview. Here it will be argued that together with the swing to market-oriented policies, there was a deeper shift in the way development problems were framed and in the types of explanation through which development policies were justi®ed. This involved changes in the spatial and temporal frame of reference of development policy analysis. In brief, these changes were: the partial globalization of development policy analysis; and a shift from historicism to ahistorical performance assessment.
2. THE PARTIAL GLOBALIZATION OF
DEVELOPMENT POLICY ANALYSIS
Specifying development policy problems involves both explanations of development
trends and normative judgements about how the world should be. For each of these activities,
an important decision which must be made is deciding the policy frame, i.e. what elements
should be included when viewing a problem and what elements excluded. 2 The framing of
policy issues has various aspects but one which critically a€ects the practice of developing
countries is whether policy problems are seen within a global or national frame of reference.
Explanations and normative judgements can each be elaborated within a national or global
frame of reference, and so the thinking which underpins the practice of developing countries
can be wholly national, wholly global, or some combination of both (Figure 1). The full
globalization of development policy analysis will be understood here to mean a shift from a
790 WORLD DEVELOPMENT national to a global frame of reference both for
explanations and normative evaluations. Before the propagation of the Washington
Consensus in the 1980s, mainstream explanations of the development process and evaluative
judgements of the goals of development were both conducted within a national frame of
reference. First, economic and social trends within countries were explained, in the mainstream, on the basis of conditions within the countries themselves, i.e. as a result of national factors. Particular external relations might be necessary to start the process, or to close ``gaps'' which threatened its breakdown. But the key ingredients of a successful development process were usually identi®ed through analyses of sequences of change within already industrialized countries, which were then applied in less developed countries without any reference to their di€erent external situation. Second, development policies were geared toward the achievement of national objectives. This orientation was often simply taken for granted in development policy analysis. But it was also in¯uenced, more or less strongly, by political and economic nationalism. According to Johnson(1967), key features of economic policy in new StatesÐnamely, the desire for greater selfsu ciency and early industrialization, the preference for economic planning and public control, and hostility to foreign investmentÐ
can all be traced to the mutual supporting relations between nationalism, aid policy, and ideas
about the development problem formed in the 1930s. Those ideas became part of a common understanding and language of national and international policymakers after WWII. There were, of course, major controversies both over the meaning of development and the means of achieving it. In the 1950s and 1960s there were debates about development strategy (for example, balanced or unbalanced growth), the nature of dualistic development processes, and the role of human capital. Moreover, in the 1970s the earlier focus on economic growth with structural change was strongly challenged by those who pointed to the need to focus on social objectives, notably income distribution, poverty, employment and basic needs satisfaction. 3 But these disputes actually served to reinforce the normative and explanatory frames of development policy analysis as being national. Whatever objectives were taken to be central, national objectives were the focal concern. Moreover the development strategy debates essentially examined the articulation and sequencing of internal (national) ingredients
which could facilitate or accelerate the national development process. An important countercurrent to mainstream development policy analysis before the 1980s came from structuralist and dependency theories elaborated in Latin America (see Kay, 1989). Like the dominant approach the normative concern of these theories was national, and indeed strongly informed by nationalist concerns. But their analytical perspective was global in scope and this
underpinned their critiques of mainstream thinking. Both structuralist and dependency theorists emphasized the importance of centerperiphery relations as determining or conditioning the national development process. But some strands within dependency theory,
Figure 1. Four main combinations of explanatory and normative framework in development policy analysis.
RISE AND FALL OF THE WASHINGTON CONSENSUS 791
instead of indicating how national development was a€ected by the articulation between internal and external factors, simply put forward an antithesis to the mainstream approach, arguing that external factors were the only ones that mattered, and then deduced that by delinking from the world economy, an ``authentic'' development process, solely founded on internal factors, could be made to occur. In the late 1970s and early 1980s, the growth
rate of most developing countries, with the notable exception of some countries in East
Asia, collapsed. The economic crises which beset most developing countries lent weight to
arguments that mainstream development practice had failed. But at the same time the East
Asian success neutralized those versions of dependency theory which argued that development would always be blocked on the periphery, and also Latin American structuralism, which allegedly was wedded to inwardoriented import-substitution policies in
contrast to East AsiaÕs alleged outward-orientation. In this situation, arguments which
emphasized the positive role of free markets in development attracted greater attention. These
ideas had always been an element within development policy analysis, represented, for
example, by early critiques of protectionism, such as G. Haberler and H. Myint, Milton
FriedmanÕs support of free enterprise, and P.T. BauerÕs dissection of mainstream thinking
(Bauer, 1971). The uptake of these ideas was not strong however until the late 1970s and
early 1980s, when a new approach to developing countries, which was later labeled the
Washington Consensus, emerged as the main alternative to national developmentalism. 4
The frame of reference for this new approach was, like the Latin American countercurrents
of the pre-1980s, partially global and partially national. But rather than combining normative
economic nationalism with a methodological internationalism, the Washington Consensus
was its mirror image. It combined normative economic internationalism with a methodologically nationalist form of explanation which attributed what was happening within countries mainly to national factors and policies (Figure 2). In this new approach, the key norms which played the decisive role in de®ning development practice were the norms of a liberal international economic order (LIEO). In most general terms, these norms involve a commitment to free markets, private property and individual incentives, and a circumscribed role for government. But they can be speci®ed in di€erent ways, according to di€erent interpretations of the precise content of the LIEO. For example, in the early 1980s, laissez-faire liberalism was strongly advocated. This entailed liberalization of both external and domestic economic relations. But at the start of the 1990s, this extreme market fundamentalism was softened with the emergence of the socalled market-friendly approach to development (see, notably, World Bank, 1991). This
Figure 2. The con®guration of development policy analysis: 1950±1990.
792 WORLD DEVELOPMENT continued strongly to advocate liberalization of external trade and capital movements. But, the scope of domestic economic liberalization was
limited, in particular, by recognizing more fully the legitimacy of state intervention in cases of
market failure. These norms were propagated through two types of persuasive argument: ®rst, arguments about the intrinsic ethical superiority of economic liberalism; and second, theoretical and empirical analyses which demonstrate that conformity to the norms of a LIEO (variously de®ned) would lead to better outcomes, not simply for the world community as a whole, but also for individual nation-states within it. The latter, which have served as the principal form of argument supporting the new approach, have mainly been articulated on a terrain in which promoting the national interest has been narrowly equated with promoting economic growth and increasing personal economic welfare. Important developmentalist concerns such as constructing national unity and realizing national sovereignty are thus excluded. On this narrowed ground, attention and publicity has been given to analyses which show that national policies which are in con¯ict with the norms of LIEO, including many elements at the heart of earlier development practice, such as protection of infant industries, managed interest rates and selective credit, have been harmful to national interests, and thus constituted domestic mismanagement and ``irrationalities.'' At the same time, the policies of the East Asian newly industrializing economies which had actually achieved rapid and sustained growth have been described in ways which suggest that they conformed to the requisite liberal norms. 5 For both con¯icting and conforming policies, their impact on the eciency of resource allocation has been identi®ed as the main mechanism by which domestic policies a€ect economic growth. While the normative frame of reference of the
new approach was global in scope, the explanatory arguments which sought to prove
the instrumental superiority of the LIEO were characterized by methodological nationalism.
That is to say, in explaining economic trends within countries, they partitioned in¯uences
into external and internal factors and attributed most of what was happening to internal
(national) factors and, in particular, to domestic policy. 6 In making the case for trade
liberalization and export promotion, for example, conditions of global demand are
generally ignored and, through the ``small country'' assumption, it is typically assumed
that foreign markets are always available, and at prices largely independent of a countryÕs
exports. Empirically, the most common approach to prove the dynamic bene®ts of
outward-orientation has been crosscountry regression analyses which establish the statistical
relationships between indicators of national economic change and a series of national variables, which include, in particular, indicators of national policy. The essence of this methodology is areal correlation between dependent and independent variables, to identify the extent to which variation in the former between a given set of national territories matches variation in the latter between the same territories. This can be done at a certain point in time or for periods of time (e.g. by using growth rates over 20 years). In either case, speci®c histories are ®ltered out and it is assumed that relationships which pertained in the past will continue into the future. Economic trends are necessarily attributed to the behavior of the national factors. In the 1990s, changes in the nature of the external environment are increasingly being used to explain why liberalization, coupled with the right macroeconomic fundamentals, ``works.'' Thus it is argued that in an increasingly globalized world economy, in which there is the globalization of production systems, increasing reliance on trade and increased availability of external ®nancial ¯ows, countries which do not follow Washington
Consensus policies will be especially penalized, as they will be cut o€ and thus excluded from
the intensifying (and implicitly bene®cial) global ®eld of ¯ows. Concomitantly, those
countries which do follow the right policies will be rewarded, as they can capture foreign direct investment which brings technology and market access, and they can also supplement
national savings with international capital ¯ows, thus reaping the bene®ts of the new
external environment. In this way, the case for liberalization is rooted in the rhetoric of the
globalization. But the analysis remains methodologically nationalist as it retains the
distinction between external and internal (national) factors, and still attributes country
trends largely to domestic policy (see, for example, IMF, 1997; World Bank, 1997).
Globalization is something which is happening to the external economic environment of
countries; it is outside them. RISE AND FALL OF THE WASHINGTON CONSENSUS 793
3. THE SHIFT FROM HISTORICISM TOAHISTORICAL PERFORMANCE
ASSESSMENT
The curious combination of global liberalism 7 and methodological nationalism which
underpins the way in which development is seen in the new paradigm has been buttressed by a second key shift which occurred in development policy analysis at the end of the 1970s.
This can be characterized as a shift from historicism to ahistorical performance assessment.
Theorizing on development strategy from the 1950s to the 1970s was historicist in the general
sense that it was founded on an attempt to understand rhythms, patterns and laws of
development. 8 This understanding was based on historical analysis of long-term sequences of
economic and social change, which had occurred in the past in already-industrialized countries and which were expected to re-occur, particularly if the right policy interventions
were made, in ``less developed'' countries. Such theorizing most typically understood development as a societal and economy-wide transition from a ``traditional'' (rural, backward, agricultural) society to a ``modern'' (urban, advanced, industrial) society. This process was
seen as a sequence of stages of growth, a process of modernization, or recurrent patterns
of structural transformation. 9 All countries were expected to go through such patterns of
development, and development agencies sought to ensure or accelerate the arrival of a better
future for whole societies through interventions in these long-term processes of historical
transformation. With the shift to ahistorical performance assessment, the focal object of enquiry has been to describe and explain national ``performances'' of various types. Not surprisingly but now taken-for-granted, the key word in the discourse propagated by international development agencies since the start of the 1980s has been ``performance.'' Attention has been particularly paid to economic performance, but also agricultural performance, industrial performance, trade performance, ®nancial performance, ®scal performance, poverty performance, human development performance and so on. Using these various standards, countries have been partitioned into good and bad performers, and ranked according to their performance in various new leagues of nations. Moreover comparative performances have been explained by reference to national factors and national policy.
It is according to these performance standards that past development policies have been
criticized because they do not ``work'' and narratives have been constructed about the
e€ectiveness of the Washington Consensus. A succession of countries which have undertaken
policy reform in the requisite way and achieved good short-term growth results have also been identi®ed as, and dubbed, ``success stories.'' These stories have acted as exemplars for the new paradigm, providing not only practical rules-of-thumb guidance on how policy reform should be undertaken, but also proof of the validity of the Washington Consensus.
The transition from historicism to a historical performance assessment started in the 1970s,
and was initially animated by those who sought to re®ne the de®nition of development by
adding social aspects. E€orts to measure poverty based on the quality of life and satisfaction
of basic needs were particularly important in this regard. Michael LiptonÕs book Why Poor People Stay Poor was a key text in propagating a performance-oriented approach. The uptake of the notion of urban bias, a concept which was forged within debates about how to achieve redistribution with growth but which became central to the neoliberal paradigm, can be attributed to its performance-based de®nition, and the vitriolic debates of the late 1970s, particularly with Byres, can be interpreted as an attempt to sustain a historicist view (see, for example, Byres, 1979). In the 1980s, these initial moves toward performance assessment were overtaken by, and later incorporated in, the discourse and practice of structural adjustment.
Adjustment involved improving the performance of national economies by increasing the
eciency of resource allocation. The central criterion used to measure performance was
current or recent GDP growth rate, and macroeconomic stability, indicated by ®scal and external payments balance and low in¯ation. The dynamics of long-term transformations
of economies and societies slipped from view and attention was placed on short-term
growth and re-establishing ®nancial balances. The shift to ahistorical performance assessment
can be interpreted as a form of the postmodernization of development policy analysis.
It re¯ects, in particular, the questioning of grand narratives of historical transformation
which was central to the appeal of the post- modern ethos in the 1980s. 10 Before the shift,
development agencies acted as handmaidens of ``progress,'' ``modernization,'' ``industrialisation,'' or the emancipation of people from oppression, exploitation, disease and drudgery. After it most agencies re-oriented their work to monitor and seek to improve ``performance,'' often through local problem-solving and local social engineering designed to make economic and social institutions ``work'' better. Adjustment also entailed the abandonment of grand long-term government-directed designs for whole societies and a shift to decentralized decision-making, laissez-faire and local social engineering. But ironically, this shift away from holism could not be achieved without a holistic approach. Everything has been made subject to the rules and discipline of the market. The vision of the liberation of people and peoples, which animated development practice in the 1950s and 1960s, has thus been replaced by the vision of the liberalization of economies. The goal of structural transformation has been replaced with the goal of spatial integration. 4. THE CONFIGURATION OF
DEVELOPMENT POLICY ANALYSIS IN
THE 1990S
The collapse of communism in Eastern Europe and the Soviet Union has served as
con®rmation of arguments which predicted the impossibility of central planning and reinforced the apparent superiority of a market-oriented development approach. Since the late 1980s however there have developed two important challenges to the Washington Consensus. The ®rst is the UNDP's sustainable human development (SHD) approach. This approach takes up some of the themes of the UNICEF critiqueof the dominant approach, Adjustment with a Human Face, originally published in 1987, and has been elaborated through the annual Human Development Report, which ®rst appeared in 1990 (UNDP, Various years). The second is a latent ``Southern Consensus,'' which is founded on analyses made from the perspective of countries undertaking late industrialization and seeking to catch up with richer countries in the global economy. This Southern Consensus does not exist as a political reality. Nor has it, as yet, been articulated analytically. Its existence is apparent however in the convergence between the policy conclusions of Latin American neostructuralism, initially set out by ECLAC in 1990, and the deeper understanding of East Asian development models, which is described in ESCAP (1990), but has been most thoroughly reconstructed by UNCTAD in its annual Trade and Development Report (particularly 1994, part 2, chapter 1; 1996, part two; 1997, part 2, chapters V and VI; and 1998, part 1, chapter 3). 11 These two challenges to the Washington Consensus have shaped development thinking and practice in di€erent ways. Indeed development policy analysis is now characterized by a double dialectic. The clash between the Washington Consensus and the sustainable human development approach acts to reinforce and conserve the key elements of the current paradigm, and in particular its ahistorical approach and its combination of normative internationalism with methodological nationalism, whilst the clash between the Washington Consensus and ideas within the two strands of the Southern Consensus serves to undermine these elements and creates tensions and pressures for a further paradigm shift. The key feature of the sustainable human development approach which distinguishes it from the Washington Consensus, is that it espouses a di€erent set of values. Whereas the Washington Consensus focuses on the promotion of GDP growth, and has been implemented through a top-down, donorconditionality- driven and outside-expert-led, approach, the sustainable human development approach argues that the ultimate test of development practice is that it should improve the nature of peopleÕs lives, and advocates that it should be founded on participation and a more equal partnership between developing countries and aid donors.
This ``people-centered'' approach, which explicitly identi®es itself as an alternative
paradigm (see, for example, ul Haq, 1995, Part I), has been quite in¯uential. An important
strand of development research in the early 1990s has sought to refute its challenge by
showing that Washington Consensus policies in fact serve to reduce poverty, increase employment and can, in themselves, deliver growth with equity, and that therefore social concerns are already adequately addressed by the mainstream approach. But the SHD alternative has promoted the introduction of poverty reduction as a key goal of development practice and increasing attention to possible LIEO-compatible relaxation of Washington Consensus policies in order better to achieve poverty objectives (see World Bank, 1990).
These changes have certainly made the Washington Consensus more humane. But at
the same time, the SHD approach has had the e€ect of conserving key features of the worldview of the dominant paradigm. Although its di€erent values have emphasized di€erent
indicators and weighting systems, particularly to capture levels of human development and
poverty, these measures have reinforced a focus on short-term performance assessment. The
substitution of multidimensional indicators of poverty for simple income poverty, for example, has added greater reality to the description of deprivation and more leverage for moral outrage, but at the cost of crippling e€ective analysis of the dynamics of change. Signi®- cantly also, the analytical basis of the SHD approach, which is itself somewhat loose, is methodologically nationalist. A central focus is the mismatch between economic growth
performance and social performance and the ways in which domestic policy can rectify this
mismatch to deliver more social achievements for any given level of GDP per capita. Even the apparent di€erence in values between the SHD approach and the Washington Consensus is less clear-cut than it appears. This applies whether human development is speci®ed rigorously, as in Amartya SenÕs capability approach which underpins the human development index, or through a vaguer focus on decentralization and participation. SenÕs capability concept emphasizes freedom of choice which is quite consonant with the liberal perspective. 12 Moreover the project of making economic and social institutions work better through decentralization and the use of local knowledge, indigenous management practices and the participation, not of the masses, but of ``local people'' and ``small communities,'' can be, and has easily been, fused into a kind of neoliberal populism. 13
Whereas the SHD approach has made a moral critique of the Washington Consensus,the two strands of the Southern Consensus, Latin American neostructuralism and East Asian developmentalism, remain focused on economic growth as the central objective. 14
They o€er however a di€erent economic analysis of how growth occurs in late industrializing
countries and on this basis propose a di€erent policy orientation to the dominant paradigm.
From the Southern perspective, national economic growth involves a process of catching-
up, in which national enterprises build up production capabilities and international
competitiveness in a range of activities undertaken in more advanced countries. The structure
of the economy changes as the relative importance of agriculture and natural resource exploitation declines while that of manufacturing activities increases, and as production
progresses from less to more skill-, technology-, and capital-intensive activities. At the macrolevel, growth, structural change and productive upgrading is driven by a rapid pace of capital accumulation, which depends on increased domestic savings, investment, and exports,
linked together in a virtuous circle of cumulative causation (ECLAC, 1990, pp. 48±49;
ESCAP, 1990, pp. 13±14, 115, 151; UNCTAD, 1996, pp. 108±112). At the microlevel, this
process is founded on imitation, adaptation and learning of internationally available Technologies in order to reduce costs, improve quality, and introduce goods and services not
existing in the country, and the di€usion of best practices from more advanced to less advanced enterprises within the country, including from foreign-owned to locally-owned ®rms (ESCAP, 1990, pp. 15±17 and pp. 92±95; ECLAC, 1990, pp. 64±71).
An important feature of the Southern Consensus is that it rejects the idea that growth
with late industrialization can be animated using a general blueprint. Policy measures have
to be adapted to initial conditions and the external environment, and change over time as
an economy matures (ECLAC, 1990, pp. 97± 102; UNCTAD, 1996, pp. 133±134; ESCAP,
1990, pp. 21±23, 140±141). It is possible however, to identify some general policy
orientations which apply in all circumstances. 15
First, the process of growth and structural change is best achieved through the ``strategic
integration'' of the national economy into the international economy rather than either delinking from the rest of the world or rapid across-the-board opening up of the economy to
imports and external capital. This means that the timing, speed and sequencing of opening, in
relation to di€erent types of international ¯ows, should be decided on the basis of how they
support the national interest in terms of promoting economic growth and structural
change (Singh, 1994). Multilateral norms are not disregarded (ECLAC, 1996, p. 86;
UNCTAD, 1996, pp. 156±157). As far as possible, however, import liberalization should
be gradualÐto enable national enterprises to build up production capabilities and thus face
external competitionÐand selective. Tari€s should also be complemented by special
measures to promote exports (ECLAC, 1990, pp. 103±107; ECLAC, 1995, chapter VI; and
for East Asian policies, UNCTAD, 1994, pp. 58±59). Capital account liberalization should
also be gradual and should be managed, in coordination with domestic ®nancial development,
to ensure that capital ¯ows are, as much as possible, additional to, rather than asubstitute for, domestic resources, that they support increased investment rather than consumption, and that they do not undermine macroeconomic stability (ECLAC, 1995, pp. 285±291; UNCTAD, 1998, pp. 75±76, 101± 106). Inward FDI should support the build-up of domestic production capabilities and exports, and this is not automatic but requires speci®c domestic policies (ESCAP, 1990, p. 132; ECLAC, 1990, p. 45; UNCTAD, 1996, pp. 131±133).
Secondly, growth and structural change is best promoted through a combination of a
macroeconomic policy and what Latin American neostructuralists describe as a ``productive
development policy.'' The macroeconomic policy is growth-oriented. It seeks to reduce
in¯ation and ®scal de®cits, but also aims to ensure full utilization of production capacity
and encourage the pace of capital formation (ECLAC, 1996, chapter V; ESCAP, 1990, pp.
17±19). The productive development policy involves a range of measures, coordinated with
the trade policy, which are designed to improve the supply capabilities of the economy as a
whole and also speci®c sectors within it, and to help private enterprise identify and acquire
competitive advantages. These measures are founded on a dynamic interpretation of the
principle of comparative advantage. In this forward-looking approach, the opportunities of
current relative cost advantages are exploited to the full, but e€orts are made at the same time to promote investment and learning in economic activities where comparative advantage can
realistically be expected to lie in the immediate future as the economy develops and as other
late industrializing countries catch up (ESCAP, 1990, pp. 148±149; OECF, 1991; UNCTAD,
1996, pp. 112±123; ECLAC, 1995, pp. 132±135, 159). Elements of a productive development policy include: technology policy, ®nancial policy, human resource development, physical infrastructure development, and industrial organization and competition policy (UNCTAD,
1994, pp. 57±69, ECLAC, 1990, pp. 107±148, ECLAC, 1995, pp. 161±190; ESCAP, 1990,
chapter V, pp. 149±150). These elements can form part of, but they should not be simply
equated with, a selective industrial policy. They are directed at improving productivity and
competitiveness in agriculture and natural-resource based activities as well as manufacturing
(ESCAP, 1990, pp. 22, 70±75; ECLAC, 1990, pp. 126±137). They entail a mix of sectorallyneutral as well as selective policies. Moreover their main goal is to accelerate the rate of capital accumulation and learning throughout the economy. Third, the successful implementation of these development policies requires governmentbusiness cooperation within the framework of a pragmatic developmental State. The policies
are implemented, as far as possible, through private initiative rather than public ownership,
and through the market mechanism rather than administrative controls. But government plays
a key role both in animating the ``animal spirits'' of the private sector and harnessing the
aggressive pursuit of pro®ts, which are the motor of the system, to the realization of the
national interest. This requires the enhancement of state capacities rather than state
minimalism. Policy should be formulated by a capable and pragmatic economic bureaucracy
which, through various formal and informal ties with business, develops a common vision of
development objectives and targets, and a common understanding of how these can best
be achieved (ECLAC, 1990, pp. 94±96; Evans, 1998). But government must ensure that any
support or protection for the private sector is conditional on investment, export or productivity
targets, and also temporary. Policies should also focus on overcoming speci®c
problems which impede the achievement of national development objectives, notably,
missing markets and the lack of an entrepreneurial base, imperfections in technology and
capital markets, risks of exporting, and dynamic complementarities between sectors
which render competitiveness systemic rather than just dependent on ®rm-level capabilities
(UNCTAD, 1994, pp. 50, 69; ECLAC, 1995, pp. 152±157; ECLAC, 1996, Box VI.1; JDB/
JERI, 1993, pp. 53±56). Fourth, distributional dimensions of the growth process are managed in order to ensure the legitimacy of the overall growth process. This is primarily achieved through a production- oriented approach rather than redistributive transfers. That is to say, the main bases for a more equitable and inclusive growth process are wide asset ownership and the expansion of productive employment. Important policies in this regard are: agrarian reform and rural development policies; high rates of re-investment of pro®ts and the establishment of pro®trelated payment systems; support for small and medium enterprises, particularly through ®nancial policies; and broad-based human resource development (Campos & Root, 1996; ECLAC, 1992, pp. 15±27; UNCTAD, 1997, pp. 183±189).Finally, regional integration and cooperation policies are identi®ed as an important element
of strategic integration (ECLAC, 1990, chapter VI; ECLAC, 1994, pp. 9±19; ESCAP, 1990, pp. 24±25; UNCTAD, 1996, Part II, chapter 1, especially pp. 75±79, 92±94). Such policies
should support the goal of increased international competitiveness, for example, by
promoting regional production chains, and also nurture the development of regional markets in order to reduce demand-side constraints on growth. These substantive features of the Southern Consensus arise because Latin American neostructuralism and East Asian developmentalism are rooted in a totally di€erent worldview to the Washington Consensus (Figure 3). This does not reject performance standards as a guide to policy, but actions are founded on historical analysis, particularly of long-term processes of late industrialization in the periphery of the world economy. A global analytical perspective is adopted and this has a
realist rather than idealist view of the way in which market economies work. This recognizes
vulnerabilities associated with integration into the international economy and also external
constraints due to restrictions in access to advanced country markets, falling terms of
trade for primary commodities and simple manufactures, cartelization in global markets,
diculties in gaining access to technology, and instabilities of the international ®nancial
system. Finally, the approach is normatively rooted in a distinctive form of economic
nationalism. This is not ideologically committed to self-suciency or public ownership, nor
hostile to foreign ownership in and of itself. It does not seek the appearance of catching up,
through either imitating consumption standards, or setting up showcase industries. It
respects multilateral rules and arrangements, engaging in their design, negotiation and
interpretation. But its aim is to build international competitiveness as part of a long-term
national economic project founded on the development of national capabilities.
Of the two strands of the Southern Consensus, the challenge from the East Asian development
models has proved to be most powerful because these models have, in terms of their
performance and according to the criterion of economic growth, ``worked'' spectacularly well.
Since the early 1990s, the major fault line in development policy analysis has thus been the
discrepancy between the policies which have been pursued in rapidly growing and industrializing East Asian economies and the policies advocated by the Washington Consensus. 16 Figure 3. The con®guration of development policy analysis: 1990s and beyond.
Kuhn argues that the questioning of a paradigm begins when anomalies arise between
paradigmatic expectations and actual events, and shows that numerous ad hoc modi®cations
typically are made to maintain an old paradigm before the accumulation of anomalies requires, and the availability of a superior alternative paradigm enables, a paradigm shift. With increasing awareness of the discrepancy between Washington Consensus recommendations and East Asian development practices, such a process has occurred with the Washington Consensus. The discrepancy has been a key factor which has impelled the shift in the Washington Consensus from laissez-faire liberalization to the market-friendly approach. But more fundamental change has, at the same time, been slowed by semantic ambiguities,
particularly centred on the key words ``outward- oriented'' and ``openness'' (see Gore,
1996a), and also further work to re-describe the East Asian experience as being compatible with the norms of the market-friendly LIEO. The World BankÕs East Asian Miracle studyÐ
which was prompted by disagreements between the Japanese government and the World Bank
on speci®c development policy mechanisms and which Wade (1996) has explicitly dubbed an
exercise in the ``art of paradigm maintenance'' Ðis a particularly signi®cant example of
the latter (World Bank, 1993). These re-descriptions have, like earlier characterizations,
now been shown to have inconsistencies and ambiguities (Amsden, 1994;
Rodrik, 1994). But the debate has taken yet another turn with the ®nancial crisis in East
Asia, and the apparent fall of the newly industrializing economies which hitherto had
been claimed on all sides as ``legitimating angels.''
5. THE COMING PARADIGM SHIFT
The ®nancial crisis in East Asia is signi®cant for the future directions in development
thinking and practice. Economic growth has fallen dramatically in developing countries and,
just as there was during the crisis of the early 1980s, there is now increasing reason to call
into question the e€ectiveness of dominant policies. Commentators of every persuasion
have been quick to argue that events con®rm their analysis. Some of those who support the
Washington Consensus have reversed their earlier description of East Asian policies as
market-friendly, and identi®ed domestic mismanagement, in the guise of crony capitalism
and excessive government intervention, as responsible for the crisis. On the other side, it is
argued that the crisis is mainly due to speculative ®nancial ¯ows and contagion. But
domestic policy, particularly fast ®nancial liberalization, is also said to have played a role.
The abandonment of government coordination of capacity expansion has led to overinvestment, and the lack of government supervision of the scale of the foreign debts of domestic companies has precipitated overexposure toexternal debt. Finally, the IMF bailout packages are said to have exacerbated the problem. At best they are seen as a misdiagnosis; at
worst, an attempt to use the crisis further to impose in a deeper way LIEO norms on domestic economic activity. Although these debates are still playing themselves out, it is becoming increasingly unconvincing to attribute the crisis solely to domestic mismanagement (see, for example, Chang, Palmer & Whittaker, 1998), or analytically to separate external and internal factors. Moreover the Washington Consensus has cracked in the practical sense that real di€erences of opinion have emerged in Washington, between the IMF and the World Bank, on the causes of the crisis and how best to handle it. One important opinion-leader, Paul Krugman (1995), has already written the obituary of the Washington Consensus. After the Mexican crisis of 1994, he argued that the major mechanism through which its policies have worked is a speculative bubble in emerging markets in which policy reforms attracted private capital ¯ows, and the attraction of the ¯ows stimulated policy reforms, and that this bubble had now burst. In e€ect, he exposed market-friendly policies as actually being markets-friendly ± ®nancial markets, that is. Similarly, Joseph Stiglitz (1998a,b) has argued that there is a need for a ``post-Washington Consensus,'' a new paradigm. This should seek to achieve broader objectivesÐembracing a focus on the living standards of people and the promotion of
equitable, sustainable and democratic development. It should use a wider range of instruments
to build markets as well as to correct market failure, and to foster competition as
well as liberalization and privatization. It should also adopt limited forms of regulation,
if necessary controlling short-term international capital ¯ows. Finally change should not
be imposed from outside but requires owner-
RISE AND FALL OF THE WASHINGTON CONSENSUS 799
ship, participation, partnership and consensusbuilding. It may be too early yet to announce the fall of the Washington Consensus. Stiglitz's proposed new paradigm contains some important shifts on values, continuing the incorporation of the goals and implementation style advocated by SHD, and, perhaps more signi®cantly, it argues for a return to the notion of a development
strategy, based on a long-term perspective, respecting historical speci®cities and with a
more holistic approach centred on the transformation of societies. Development should no
longer be a monopoly of economists. But the proposed post-Washington Consensus consensus can also be interpreted as simply a change to preserve the old order by making it more e€ective as well as more humane. In elaborating the new paradigm, Stiglitz (1998b, p. 34) explicitly states that a key task is to lessen the momentum of an expected swing of the
pendulum of opinion against openness. The proposal retains a strong commitment to the
fundamental principles of a LIEO founded on open trade regimes, competitive markets and
open societies. But, by de-linking trade and ®nancial liberalization and then analytically
separating short-term from long-term international capital ¯ows, it reduces the risk that in
the aftermath of the ®nancial crisis the liberalization of external economic transactions will
be called into question as a whole. Through this analytical splitting, what previously was propagated as a total package can now be taken to be a more ¯exible menu of options, and any
possible backlash against liberalization can be more easily contained. Whether or not StiglitzÕs courageous intervention is a rupture with the past or the preservation
of the old regime, more profound change is inevitable. This is because the forced
marriage of global liberalism and methodological nationalism, the latter providing the
empirical justi®cation for the internalization in domestic policy of the prescribed international
norms of the former, is inherently unsustainable. The only circumstances under which
methodological nationalism is a completely coherent approach to explanation is if national
economies are completely isolated and closed from outside in¯uences. The more that the
norms of a LIEO are adhered to, the more that national economies become open to outside
in¯uences, the less tenable methodological nationalism becomes as a form of explanation.
The dominant paradigm is thus unstable. Its ideology and methodology are in contradiction.
The coming paradigm shift will be driven by the main ``workable'' alternative, East Asian
models, politically strengthened through their convergence with Latin America neostructuralism, and extended to Africa and the least developed countries. But while this approach can o€er a more e€ective way of developing countries than the Washington Consensus, it does not, as it stands, provide an ideal alternative paradigm. This is not because the current ®nancial crisis has somehow nulli®ed the development transformation which has
occurred in East AsiaÐthough the crisis demands closer consideration of the issue of
``development strategy in the age of global money.'' 17 Rather it is because it remains a
moot point whether it is possible to achieve similar results to those achieved by East Asian
countries in their high-growth period, given the widespread, simultaneous adoption of past
East Asia-type policies. Moreover, though exaggerated, some new global rules, particularly
concerning technological borrowing and adaptation, may inhibit the replication of some
of these policies. In the future, the full globalization of development policy analysis seems inevitable (Figure 3). This will entail the explanation of national development trends in a global context, and also the elaboration of alternative normative principles for the international regimes which constrain and enable national policy choices. Signs that such a spatial frame shift is now occurring are evident in diverse and unconnected analytical arenas. These include: attempts to link international trade theory to labor market performance (Wood, 1994); the development of the new economic geography (Ottavino & Puga, 1998) and sociological analysis of global production chains (Gere, 1995); work on global environmental commons; and the emergence of social exclusion as a concept of deprivation (Gore, 1996b). The spatial frame shift is likely to be linked to the re-introduction of a historical perspective, which is already becoming evident, for example, in analyses of the history of globalization of economic activity (Bairoch, 1993; Bairoch & Kozul-Wright, 1998; Brenner, 1998). But with the rejection of grand narratives, bringing history back in should not presage a return to the old teleological historicism, but rather identify alternative situations and possible development paths, and thereby inform a 800 WORLD DEVELOPMENT
pragmatic commitment to progressive change in favor of present as well as future generations.
The values which will glue together the new way of seeing the world are, like the methods of
global analysis, as yet unclear. The most likely prospect is that we shall be blown into the
future facing backward, embracing a form of embedded communitarian liberalism, which
seeks to reconcile the achievement of national, regional and global objectives, and to marry
universal values with a respect for diversity. But this is still waiting to be born.
NOTES
1. That is, a constellation of beliefs, values, techniques and group commitments shared by members of a given community, founded in particular on a set of shared axioms, models and exemplars (see Kuhn, 1970). The term ``paradigm'' is used in this sense throughout this paper.
2. For an extended discussion of the importance of frames in policy analysis, see Schon and Rein (1994). The notion of the frame is also pivotal in Amartya SenÕs work on development evaluation, though he uses the term ``informational basis'' of evaluative judgements
rather than ``frame.'' 3. For deeper discussion of these debates, and the role of international development agencies in them, see Arndt (1987), chapters 3 and 4.
4. This was a complex historical process. As Kuhn (1970) explains, the timing of paradigm shifts is in¯uenced not simply by scienti®c and policy debate, but also broader political and ideological con®gurations. These broader changes, which include the election of conservative
political leaders in the United Kingdom, United States and Germany in the late 1970s and early 1980s, will not be dealt with here. For a subtle account, which locates changes in development thinking and practice within a broader counter-revolution against Keynesian
economic policies, see Toye (1993). 5. For these two lines of argument, see various World
Development Reports, particularly World Bank (1983, 1986, 1987). The last, as well as criticizing deviant policies, is an exemplar of the mobilization of East Asian experience to support key principles of a LIEO. 6. For an extended discussion of methodological nationalism, see Gore (1996a). 7. The term ``global liberalism'' is used here as shorthand
for various types of LIEO, which may or may not allow a circumscribed role for national government intervention in market processes. 8. The term ``historicism'' is used here in the most general sense given by Popper (1960, p. 3). It does not imply that planning which aims at arresting, accelerating or controlling development processes is impossible, though some historicists would adopt this stronger position (Popper, 1960, pp. 44±45). 9. Exemplars are Rostow (1960) and Chenery and Syrquin (1975). 10. Lyotard (1984) sees the main criterion which is used to legitimate knowledge after the questioning of the grand narratives as ``performativity,'' which is understood as assessment of the performance of systems in terms of the best input/output relations (p. 46). 11. Various academic books and articles are associated with these policy reports. Key elements of Latin American neostructuralism, which developed as a response to the weaknesses of both neoliberalism and importsubstitution
industrialization, are set out in Bitar (1988), Ffrench-Davies (1988), Sunkel and Zuleta
(1990), Fajnzylber (1990) and Sunkel (1993), and are surveyed in Kay (1998). A Japanese view of the contrast between East Asian developmentalism and the Washington
Consensus is set out in OECF (1990), whilst Okudo (1993) and JDB/JERI (1993) discuss the Japanese approach, focusing on two important policy mechanisms which diverge from the tenets of the dominant approachÐtwo-step loans and policy-based lending. UNCTADÕs reconstruction of East Asian developmentalism, which was elaborated independently of Latin
American neostructuralism, draws on analyses of the Japanese development experience, particularly Akamatsu (1961, 1962) and Shinohara (1982), and key elements are set out in Akyuz and Gore (1996) and Akyuz (1998). 12. For an outline of this approach see, inter alia, Sen (1993), and an analysis of the limits of its moral individualism is made in Gore (1997). 13. For examples of a loose approach to poverty analysis based on the concept of sustainable human development, see UNDP (1995a,b); but Banuri et al. (1994) attempt to give a more rigorous speci®cation of the concept through the notion of social capital. An
interesting recent development has been to link sustainable human development to the promotion of human rights discourse, which some see as an alternative global ethics to neoliberalism. The increasing incorporation of the voice of nongovernment organizations (NGOs) into or alongside UN social deliberations is also a€ecting the SHD approach. A good discussion of some of the notions which animate these discussions is Nederveen
Pierterse (1998). 14. It is dicult to identify an African strand to the Southern Consensus, but Mkandawire and Soludo (1999) seek to develop an African alternative to the Washington Consensus, and UNCTAD (1998, part 2) has drawn implications of the East Asian evelopment experience for Africa. 15. There are some divergences between the East Asian
and Latin American approaches. The latter gives more prominence to environment and democracy, is less committed to aggressive sectoral targeting (ECLAC, 1996, pp. 70±71; Ocampo, 1999), and has a more re®ned policy analysis of the process of ®nancial integration
than East Asian developmentalism (ECLAC, 1995, Part 3). But their similarities, and common disagreements with the Washington Consensus, are more striking.
16. For an interesting alternative interpretation of this fault line, see Yanagihara (1997) who contrasts an ingredients approach and a framework approach and seeks ways of synthesizing them. 17. To paraphrase Yanagihara and Sambommatsu (1996). approach to structural adjustment ± proposal from a major partner. OECF Discussion Paper No. 1.
Hiç yorum yok:
Yorum Gönder