The 1960s look like the golden age of economic development. GDP in Latin America, East Asia, and the OECD countries grew on average at about 5 percent a year. South Asia and Sub-Saharan Africa trailed closely behind. Predictions for the future were frankly optimistic and---interestingly---rosier for Africa and somewhat bleaker for East Asia. Growth would continue (if not accelerate) and poverty would be sharply reduced.
But the next 20 years were turbulent---with many setbacks. In the 1970s and 1980s, the inter- regional growth gap widened dramatically. Roughly half the wider gap came from the unexpectedly fast growth in East Asia (including China). The other half came from the plummeting drop to negative growth rates in Latin America and particularly in Sub-Saharan Africa.
Growth has now resumed for the developing countries as a group. Will it last, and will it spread? Do past failures of prediction imply that we have learned no general lessons about the sustainability of the development process, its fundamental components, its distributional consequences? Do we understand when and why some countries---in the face of the same world environment---got into the wrong growth and poverty track for a long time? And what rises from the wreckage and the reform process as the correct blend of government and market? Michael Bruno, the Bank's Chief Economist and Vice President for Development Economics, took up these questions in his keynote address to the sixth Annual Conference on Development Economics.
What is most striking about the comparative developments of the past 20 years is the diversity of regional experience in a common turbulent environment. The East Asian "miracle" countries emerged as an apparent long-run model of development. They seem to have done almost everything right---and against them one can contrast other countries and ask what went wrong. There is very broad agreement about the remarkable shared growth---with rapid growth and poverty reduction reinforcing each other.
The most interesting and compelling part of the story (which certainly featured less in the development literature of 20 years ago) is the role of basic education in enhancing the quality of labor---complemented on the demand side by a pattern of export-led growth that made productive use of labor, the poor's most important asset. In particular, the emphasis on the education of---and on health spending for---mothers led to a drop in fertility rates and an implied increase in home-time spent on the education of the next generation of children. Supplementing this was an increased slice of basic education spending in an expanding resource pie.
The size of the effect is substantial. Based on a cross-country regression estimate, if the Republic of Korea had only the low school enrollment rate of Pakistan in 1960, its GDP per capita by 1985 would have been 40 percent lower than that actually attained. The role of the education of mothers can also be decomposed. Compare children's evolving basic home and public education in Korea with that of Kenya: in 20 years (1965-85) Korean mothers' enhanced primary education and the resulting drop in their fertility rates doubled the estimated gap in the time (weighted by education) mothers devote to their children. Korean mothers spent twice as much time with their children as Kenyan mothers did in 1965---four times as much in 1985.
Similarly, while Korea's lower birthrate kept a more or less stable number of children eligible for basic education over the period, in Kenya the number eligible almost doubled. Combining the effect of a larger percentage of GDP spent on basic education and the much faster economic growth rate gives a startling result: by 1985 Korea's public expenditure per child was 27 times that of Kenya's, up from being only three times higher in 1970. The stronger performance on education plus a high savings rate (cause or effect?) provided the bulk of the supply-side capital and labor accumulation for Asia's "miracle." Exports, meanwhile, delivered the demand side---along with the discipline of market prices.
Technological spillovers, contrary to what is sometimes claimed, do not appear to explain much of the very rapid East Asian growth. Measured total factor productivity for many of the East Asian countries does not account for more than 20 to 30 percent of GDP growth. And the part of this residual attributable to technological progress could be very much smaller.
Consider instead that the East Asians went through the thick and thunder of the 1970s and 1980s with better macro policy responses. Their deviations from rapid growth were shorter and their margins of underused capacity were smaller. Better macro policy may thus account for a big part of the observed difference between East Asia and the rest of the world in growth of total factor productivity.
Among the factors that may account for varying country performance over the past two decades, different domestic responses to external shocks may be most important. The protracted crises and their different resolutions have highlighted growth as a path-dependent process.
The shocks and the different responses were not predicted. But the unfolding of the crises vindicated the importance of macro fundamentals and highlighted the importance of micro market-oriented reforms---thus salvaging some major common policy lessons amid the diversity of country experience.
It is useful and conventional to distinguish two elements of the adjustment package to rescue an economy from a deep inflation (or balance of payments and stagnation) crisis and move it to relative price stability and the resumption of growth: one, stabilization, necessary to correct the underlying macroeconomic imbalance, and two, structural reform centering on the supply side, necessary to change incentives and restructure the distorted microeconomy. This usually encompasses a series of market-oriented reforms, such as trade liberalization, public sector reform, and reform of the financial sector. These also spell out a fundamental redefinition of the role of government.
Three nested necessary conditions link growth with poverty reduction and with the elements of adjustment:
All three, in addition to basic education, are necessary---but by no means sufficient---conditions for more rapid growth. Narrowing the gap between necessity and sufficiency remains a major research task that may only in part turn out generally applicable results. But even the three conditions have been---and sometimes still are---subjects of heated dispute, especially the link between poverty and growth.
The controversy on the "tradeoff" between growth and equity endures, though its intensity and form have varied. In the golden age of growth at the end of the 1960s and in the early 1970s, the debate was over "redistribution with growth." Policy actions, often misguided, led to an undifferentiated increase in public social spending that largely bypassed the poor. Increased spending led to larger public sector deficits and to deepening macroeconomic imbalances when worldwide shocks impinged on country performance.
In the adjustment to new external shocks (such as the debt crisis), the attack has centered on the social consequences of adjustment. One form of the attack implies that poverty can be reduced only by direct antipoverty targeted policies. The rhetoric comes close to denying any link between growth and sustainable poverty alleviation. By implication, it belittles the view that adjustment is a fundamental step toward renewed long-term growth after a crisis.
It is hard to see how a given (or diminishing) pie can be redistributed in favor of the poor---and yet make them absolutely better off in income or wealth. Direct action on health, education, and nutrition could always improve the quality of life for the poor, even if their improved human capabilities may not translate into higher income or wealth. But even that requires public expenditure---unlikely from a diminished aggregate pie.
The evidence on this is clear, and here again recent research has helped clarify and solidify our understanding. Across several countries over a long period, the poverty gap drops with positive mean growth at an elasticity of about -2: that is, 5 percent annual GDP growth reduces the poverty gap by 10 percent a year.
East Asia's shared growth shows that growth and poverty alleviation can go well together. Estimates may differ about the relative strength of the two-way causal link between growth and poverty. The argument is that initial conditions---both existing human capital and greater equality at the outset---help explain post-1960 growth at least as much as growth accounts for subsequent better distributional outcomes. Better distribution enhances political stability, which in turn makes for less political disruption of sound macro policies. But experience elsewhere in Asia and more recently in Latin America shows that poverty measures are strongly correlated with the ups and downs of growth.
This evidence is silent on two crucial aspects. It does not argue that growth is sufficient for poverty alleviation. Nor does it support the view that pro-poor government policies---other than those that enhance growth---are not important for poverty reduction. On the contrary, patterns of growth can clearly be shown to be more or less poverty-enhancing, depending on initial conditions.
World Development Report 1990 has already stressed the importance of primary education and basic health care both for the benefits they confer and for their role in promoting labor-intensive growth. Moreover, the Latin American experience shows that growth can falter and become increasingly unequal when policy is biased against the rural sector through mistaken pricing and protection policies, overvalued exchange rates, or the neglect of rural infrastructure.
A new consensus has emerged on the roles of the private and public sectors---and it influences the design of policies aimed at poverty alleviation. With a more neutral structure of microeconomic incentives, price stability and growth work in favor of the poor on two important counts. First, the poor benefit from increased demand for labor and thus from greater private consumption. Second, with incomes growing, governments are more capable (and arguably more willing) to finance public action that is pro-poor.
One key question is how to make public action on poverty more cost-effective. This applies to the provision of both infrastructure and social services, and is an area requiring much more study. Unresolved is the issue of group-targeted programs (for example, women, families with many children, or specific poverty-stricken regions) versus universal provision of social services. It raises difficult questions of political economy and any proposal is fraught with problems in implementation.
There is also the difficulty in evaluating the effectiveness of proposed policies. Much is known about poverty, how to measure it, and its relationship to growth and macro policies---much less about developing effective policies and institutions whose objective is the optimal reduction of poverty.
Macro fundamentals and the reform of micro incentives work together in adjustment and in the resumption of growth after a prolonged crisis. And the experience of the past 25 years vindicates the strength of some old and by now well-honed ideas.
High inflation in Latin America (as well as in Israel and Eastern Europe) has dealt a mortal blow to the bubble theory of inflation---that is, the idea that inflation could be self-driven by expectations without underlying roots in fiscal imbalances. It was relatively easy to fall into the trap that nonmonetary and nonfiscal means could be used for stabilization because of the apparent empirical existence of multiple inflationary equilibria---the fact that different inflation rates may be consistent with the same budget deficit. It is interesting that countries hardly ever learn from other countries' past mistakes and have to repeat the same painful experience before they turn around. Russia may be no exception.
Orthodox fiscal and monetary components have become a necessary and generally accepted part of any stabilization package. But there are some important new modifications. Since the mid-1980s an added twist for countries with prolonged inflationary inertia has been the multiple anchor approach. It calls for a wage and exchange rate freeze (with or without price controls) to accompany the orthodox components of a heterodox stabilization package. Another relatively new emphasis calls for bolstering the independence of central banks, a stabilization tool with a structural reform element.
The other important legacy of recent reform experience has to do with the microeconomy of market incentives and the redefinition of the role of government and the private sector. Here again are old elements and new. The advantage of trade-determined market discipline for efficiency and distributive (political economy) reasons was well established more than 25 years ago. It proved effective in the more successful growth experiences of the Asian newly industrializing economies, in isolated Latin American countries (Colombia), and in Israel.
The crisis of the past two decades produced additional hard evidence on the resilience of an outward orientation in the face of external shocks. And the very recent brand of reform experiments in Latin America (Argentina, Chile, and Mexico) and Eastern Europe (Czechoslovakia and Poland) incorporates an even more extreme form of quick-fix import liberalization. Indeed, in these later approaches, the pace of liberalization is clearly more heavily influenced by considerations of political economy than by economic arguments alone.
Government divestiture is providing new lessons about its two main aspects: the divestiture of production (mainly through privatizing state-owned enterprises) and the demonopolization of the financial sector. A new paradigm is emerging from the pervasive mismanagement of public enterprises and political patronage in both industrial and industrializing countries, culminating in the most extreme case in the collapse of the centrally controlled economies. Private ownership of the means of production---except in a diminishing share of infrastructure investment, natural monopolies, and so on---is the new ideal.
But is ownership all that matters? The answer that is beginning to be distilled from experience is more qualified. Changes in ownership do matter in the long run. And in the transition economies, privatization is the only way to move from central state control to a market economy with private sector enfranchisement. But hard budget constraints and the discipline imposed by the market are critical even in the short run. These forces can work---or fail---under either government ownership or private ownership.
Ownership comes up in two forms. The first is the intermediate stage of commercializing state-owned enterprises. As in China and Poland, the credible imposition of market discipline, even before the transfer of ownership, creates the incentives for productivity and profit enhancement through better management. In Poland, managers (and workers) may change their behavior either because they have a stake in future profits under an impending privatization program or because they are signaling their value as managers in a future market for their services.
The flip side is that a mere formal change of ownership does nothing to change incentives when the budget faucet stays open. Consider an authoritarian ruler distributing ownership of enterprises among cronies or close family associates (the Philippines under Marcos). Or consider quasi privatization in Russia, where ownership passes into the hands of managers and workers---yet the enterprise may for a time continue to milk the state budget.
There also are many examples of private enterprises mismanaged under soft budget constraints. Redrawing the boundary between the public and the private sectors as part of structural reform may thus start with the clear demarcation of future ownership of enterprise debt even before proceeding on the asset side.
With time, more will be learned from comparison of the varying restructuring experiences of countries in Eastern and Central Europe. For example, to resolve bad debts, the most advanced restructuring economies in transition have opted for different channels. The Czechs expect private owners to take responsibility. Poland relied on the banking system. And Hungary is relying on liquidation and bankruptcy procedures.
Two issues require more research. What is the correct sequence of reforms in the financial and production sectors? What are the appropriate regulatory mechanisms for newly liberalized economies?
On the first, experience from Latin America and Africa suggests that financial sector reform fails when public sector enterprises are not restructured pari passu (as they continue to be a drag on the bad loan portfolio of the banks). But commercial banks may help induce changes in the corporate governance of nonfinancial enterprises in the immediate post-privatization stage of a transition economy. For that matter, what is the relevant model for banks? Should they be allowed to hold equity as in the continental or Japanese models? When and where is one or the other argument dominant?
On the second, the fact that greater micro market liberalization has to go hand in hand with tougher supervision at the government level, especially for the financial sector, is a paradoxical lesson that not just the newly reforming countries are learning (see the recent major financial supervision failures in Japan, the United Kingdom, and the United States).
What are the necessary and sufficient conditions for adjustment that lead to sustainable growth? Put differently, what is the minimum package of reforms that have to be undertaken? We seem to understand the issues, but we know little about how to handle them for policy design. Even when macro fundamentals are in place, and structural reforms are under way, sustainable growth takes a long time to resume.
The resumption of profit-motivated, real, long-horizon private investment (as against portfolio investment or speculation) depends on how investors evaluate the future of an economy that is emerging from a crisis and that has a bad track record. Governments can help by providing complementary infrastructure, covering start-up costs, and alleviating the worst social hardships. The irreversibility of reforms (both economic and political) plays the dominant role in forming investors' expectations.
The most important---and hardest---service for a government to deliver is the irreversibility of a new policy environment and the credibility of the reform effort. A supposed positive-sum game (that is, a game in which society in the aggregate stands to gain) is not enough. Government resolve and the widespread shared ownership of a reform effort, while still in progress, are the great imponderables. Experience is accumulating from successes and failures in more and more countries, yet evaluation for policy purposes escapes precision.
Institutions matter, and they may vary substantially, as recent cross-country experience demonstrates. So, similar policy advice (as for the introduction of regulatory agencies and rules) may result in different policy outcomes in different institutional settings.
Progress in economic theory does not always conform to progress in policy practice. Practitioners and policymakers reared in the cultural and political traditions of any particular country know the importance of institutions in shaping what is feasible (that is, both the enhancing and the constraining factors) for recommended policies. The problem arises when attempts are made to replicate or transfer policy lessons from one setting to another, which emissaries of a multilateral institution are naturally tempted to do. Not all policy applications are institution-sensitive, but quite a few are.
But is there a general institutional theory that could be empirically applied or used as a tool kit for predictive modeling? Not yet. But this does not mean that the study of institutions and organizations should not be taken seriously. At a minimum, the positive economics agenda should involve a careful and systematic comparative study of markets and the institutional constraining factors in different countries. The normative implication of this agenda would then (in the absence of generally distilled theories) bring cumulative international experience to bear when advice is given for the application of a policy in a particular country.
Consider three very different recent lines of inquiry. Much progress has been made in recent years in understanding the relationship of central bank independence and price stability. One interesting finding is the difference between legal independence (based on an index derived from formal statutes, i.e., the written "rules") and actual independence (measured by, say, the rate of turnover of central bank governors). The first is a good predictor of price stability for industrial countries but a very poor one for developing countries. The second turns out to be a much better predictor for developing countries. This is consistent with the intuition that it is the enforceability (which includes the general social and political acceptance) of legal rules that matters.
Another example is a recent World Bank comparative study of regulatory agencies in the telecommunications industry, from which some general lessons can be learned about the applicability of different regulatory institutions to different countries depending on the commitment mechanisms and the complexity of regulations in different settings.
Third and last is a study comparing the efficiency of the civil service, in the different rates of mobility and rules of promotion, as observed in the irrigation systems of India and Korea. A good understanding of the rules governing the behavior of those in charge of policy implementation in any country should be a prerequisite for giving advice. For example, irrespective of the results of the debate about the role of micro interventions in East Asia, it would be of small practical service for Sub-Saharan Africa unless it can explicitly tackle the different institutional foundations of the civil service in these countries.
By all indications, world trade will grow considerably faster than output. The demise of Comecon, the successful completion of the Uruguay Round, and the trade-increasing regional trade arrangements all point to an improved harmonization of institutions and regulations. Private capital flows and the flows of ideas have increased rapidly, and the environment is increasingly competitive.
Will the current upbeat assessment of a stable and competitive world economy prove more accurate than the earlier ones? Is there indeed a better capacity to prescribe and implement policies that lead to sustained growth and reduced poverty? A sober assessment of past predictions should instill humility. But the upside is that there is a hard core of knowledge---small and increasing---that has been sustained and buttressed through the turbulence.
Because development and adjustment are inherently uncertain and complex, it is futile to attempt to get at necessity and sufficiency for the general case. A less ambitious and more realistic research agenda would reduce the area of ignorance that lies between necessity and sufficiency and sift out from theory and empirical evidence all that is general and invariant.
There should be an attempt to minimize the residual country-specific or situation-specific factors (recognizing that they will always be there to qualify the efficacy of our policy prescriptions). For a better understanding of the difference between necessity and sufficiency in the prescription of development policy, there is a need for deeper knowledge of institutions and political factors. And this knowledge has to be made operational across an increasing array of countries and experiences.
Drawn from Michael Bruno, "Development Issues in a Changing World: New Lessons, Old Debates, Open Questions," keynote address at the World Bank's sixth Annual Conference on Development Economics. (See box for publication information.)