Source: Carnegie
Inequality and Global Coordination
American University Dean's Colloquium on Globalization in Crisis, December 18, 1998
We start today from the important observation that the world economy has been chronically unstable, and that improved domestic policies along with better international coordination can reduce that instability and improve the human condition. In my remarks I want simply to note that the world has also been chronically unequal in the distribution of economic well-being, across and within countries, and to propose that we add this additional reality to the background against which we consider remedies for strengthening the world economy. One way to frame the issue is to posit that a more equitable growth process would also be a more stable one¾ so that contrary to conventional wisdom, the right policy mix should combine the growth and stability objectives with more concerted efforts to improve equity. If reduced inequality contributes to reduced instability, then not only could we have the best of both worlds (growth with equity), we are shortsighted if we fail to aim more explicitly for the best of both worlds. Indeed the current global financial crisis could then be viewed as a wake-up call, not only for rethinking the international financial architecture, but for rethinking at a more fundamental level how to mainstream the equity objective into what we might call more broadly the international economic architecture.
I have five parts to these remarks: 1) the distinction between constructive and destructive inequality, 2) a few facts about inequality in the world, 3) thoughts about the impact of globalization on inequality across and within countries, 4) why inequality matters¾ why not so what?, and 5) initial ideas about the implications for global economic coordination.
Constructive vs. destructive inequality. Inequality is not really bad of course. Some inequality of income across individuals may be the healthy outcome of differences in ambition, motivation or willingness to work. This constructive inequality provides incentives for mobility and rewards high productivity. But inequality can also reflect deep and persistent differences across individuals in economic opportunities. This destructive inequality usually takes the form of persistent differences in access to the assets that generate income¾ including not only land and physical and financial assets, but education and good health. It is likely to be associated with persistent poverty and immobility for members of certain groups within a society. That destructive inequality undermines economic growth and efficiency, by reducing incentives for individuals to work, to save, to invest, to innovate, is obvious ¾ even a tautology. The problem for policy is that traditional efforts to remedy destructive inequality via redistribution of assets or compensatory programs can create the wrong incentives, undermining constructive inequality (the alleged failing of the pre-reform, U.S. welfare program), or can take the form of populist transfers that will be fiscally irresponsible and thus ultimately destabilizing (populist bouts in Latin America). In both circumstances traditional redistributive policies undermine the growth process that is ultimately the more effective remedy for improving the human condition. Also inhibiting policy efforts to attack inequality may be the prevailing assumption, at least among economists, that though inequality is an unfortunate outcome of all sorts of economic phenomena, it is not, fundamentally, a cause of economic problems.
The facts: inequality across the world. The conventional economic growth model assumes eventual convergence in income levels between poor and rich countries, as countries at lower levels of income per capita exploit higher returns to inputs and grow faster. The facts belie the assumption however, at least for the last 100 years. The ratio of income of the richest to the poorest country in the world has risen from about 9 to 1 at the end of the nineteenth century to 60 to 1 today. China and India illustrate the difficulties of arguing for any foreseeable convergence in income of poor and rich countries. For the last 15 years these two nations have experienced faster income growth than the rich countries, yet it would take them almost a century of constant growth at rates higher than those in today’s industrialized countries just to reach current U.S. income levels. Aside from the examples of China and India, both of which have been relatively closed economies, the only other examples of convergence are for those countries that have aggressively sought commercial links to the outside world ¾ Japan, beginning in the Meiji era, the poorer countries of Western Europe during the nineteenth century, and then again during the postwar period of European integration (but not, notably, in the interwar period when economic links disintegrated), and the so-called miracle economies of East Asia for three decades prior to the recent crisis.
Despite the economic logic of expected higher returns to capital and other inputs the smaller their base, which should ensure higher growth rates in poorer countries, it appears that institutional and policy failures and lack of a critical mass of infrastructure and human capital can condemn poor economies to low growth, if not indefinitely then for very long periods indeed, implying continuing and even growing inequality between today's rich and today's poor countries. Meanwhile, because population growth has been and will be much more rapid in today's poor countries, there is a stunning mismatch across countries between income growth and population growth. As a result, more and more of the world’s population lives in countries that generate less and less of the world’s wealth.
Within countries, inequality also at best persists and in some cases has been worsening. Between 1973 and 1998, inequality steadily worsened in the United States, with the average income of the poorest 20 percent of households declining steadily and the wage of white male high school graduates falling 15 percent (from 1973 to 1993), while the income of the richest 20 percent of households rose 15 percent. The inequality of wealth, a forerunner of income inequality, almost certainly has increased in China, Russia, and the until recently prospering countries of East Asia. Inequality of income in Latin America is higher than anywhere in the world. It got worse in the 1980s, when growth was low or negative in most countries, and has not gotten better (though poverty has been very modestly reduced) in the growth years of the 1990s. In a study done at the World Bank of changes in income distribution across 45 countries, Deininger and Squire report that only eight, including Japan and three European nations, showed any improvement in income distribution over any time period, and this improvement was minimal.
Globalization and inequality. Of course we cannot blame globalization for all the inequality in the world. Some inequality across and within nations is due to longstanding structural factors which reflect history, geography and predictable human behavior. (For example, that educated women in the U.S. now earn high incomes and tend to marry educated and high-income men has increased household income inequality in the U.S.) But there are indications that globalization, even as it expands the potential for healthy and sustainable growth through increased trade and capital flows, generating jobs in the industrialized economies and encouraging investment in developing economies, is a mixed blessing. Along with its benefits comes the potential not only for more instability but for increased inequality.
Within the OECD countries, globalization can exacerbate inequality in two ways. First, it puts pressure on the wages of the unskilled and creates job insecurity, because of the viable threat of firms to move jobs and capital to lower-wage settings. Second, because capital and skilled labor are mobile across borders and can flee nation-based taxes, globalization makes it more difficult for governments to finance social insurance and other programs to ease its strains on people, without adding to the tax burden on unskilled labor itself.
What about emerging and transitional markets? In those countries the market reform processes associated with globalization (the so-called Washington Consensus), have also had their downside. Though trade liberalization, for example, has almost certainly reduced disequalizing rents to insiders, it also seems to be associated with increasing wage dispersion, at least in Latin America¾ apparently because the combination of technology change with the globalization of markets has increased the demand for skilled labor faster than the supply, so that the premium to skilled labor has gone up. Other reforms, such as privatization, turn out to pose grave risks of concentrating wealth unless done well and with the full complement of regulation; Russia is only the most extreme example. More fundamentally, it should not be surprising that market reforms, by improving market signals, reward those who already have assets. If the initial distribution of assets¾ be they financial assets, property, or education¾ is unequal, then the reforms are likely to be disequalizing, at least for an initial period. Compounding the problem, high initial inequality of assets and income can easily infect the implementation of market reforms that are well-designed on paper, so that the reforms themselves, rather than enhancing competition, fall victim to the disequalizing rent-seeking they are meant to correct.
The emerging markets face the additional cost of globalization associated with their vulnerability to volatility in the international market. This vulnerability, unfortunately, can itself contribute to inequality. High inflows of capital generate inflationary pressure and hurt labor-intensive agricultural and manufactured exports. In Asia, the poor probably gained less during the asset boom, and then lost more with the bust. During periods of capital outflows, high interest rates to protect currencies hurt small enterprises more than big ones. In fact, austerity policies that the global capital market demands of developing countries in times of stress imply fiscal discipline and high interest rates, when under the same conditions the OECD economies would benefit from automatic countercyclical stabilizers. Furthermore, we know now that the effects of unemployment and bankruptcy on the poorer half of the population can have permanent effects; in Mexico for example, increases in child labor force participation and reduced enrollment in school during the 1995 crisis have not been reversed. Similarly a collapse in employment opportunities for labor force entrants can have life-time effects on job possibilities and income-earning potential for the affected cohorts.
Finally, for the poorest of the developing countries (Africa) the cost of globalization is the likelihood that positive effects of globalization will completely pass them by¾ even as their citizens continue to be exposed to the higher consumption standards visible thanks to globalized communications.
Why inequality matters. If part of the fallout of globalization is greater inequality, that is worth understanding in itself, as there are surely ways to minimize any disequalizing pressures. In addition, inequality, whether exacerbated by globalizing forces or not, is worth attention if it contributes to undermining the positive benefits of globalization. I can think of three examples.
First, inequality may affect the stability of the global economy. In his paper for this conference, Colin Bradford points out that of 50 developing countries, 12 that are among the largest and the least poor received the bulk of capital inflows in the early 1990s. He suggests that capital simply couldn’t find enough developing countries with potentially adequate returns to ensure global diversification of risks. The poverty of most of the other 38 countries, in the form of inadequate infrastructure, poor human capital, weak institutions, lack of property rights, and so on, translates into low expected returns to new investments.
Second, high income inequality may directly glow growth. Analysis of the determinants of growth for developing countries suggests that those countries with greater inequality have grown more slowly over long periods. A simple example: Latin America, with the highest levels of inequality of land and education in the world, has had among the lowest annual average growth rates in the postwar period¾ 2 percent annual per capita income growth, including the high-growth years of the 1960s and 1970s. In East Asia, economies like Taiwan, Korea, Malaysia, and Indonesia, with much more equal distribution of land and massive efforts to provide education to all, have had almost 30 years of average annual per capita growth close to 6 percent, until this crisis.
Third, inequality puts the market and globalization model at risk of political backlash. Along with its partners¾ social exclusion, poverty in OECD countries, wage discrimination and job uncertainty and insecurity ¾ inequality threatens politically the market reforms and the global integration of markets which are ultimately the best hope for improving people’s lives. You know the signs of backlash: capital controls in Malaysia, rejection by Congress of fast-track legislation for free trade negotiations, ruble-printing in Russia, the battle in Mexico over the government's technically sound but politically disastrous proposal to more explicitly transfer the costs of the 1995 bank failures to the taxpayers. It doesn't help that among the international solutions to the financial crisis have been a style of IMF rescue that for all its merits seems to bail out the big guys and imposes bankruptcy and unemployment on the little guys. For the IMF, this is not just a problem of moral hazard. That the current arrangements force a tradeoff between risking global financial collapse and being fair has become a political problem.
Implications for global policy. There are a number of implications for global policy. These include but go beyond current discussion about the international financial architecture.
First, continuing inequality across countries suggests there is still a compelling logic for sizeable and stable public transfers from rich to poor countries. This is neither a new nor an exciting point, but that makes it no less central. Public resource transfers to developing countries are tiny today. The World Bank’s net transfers in the last several years have been negative $1 to 2 billion annually (though with the much larger disbursements to Korea and other Asian countries in 1998, net transfers may have become positive.) Transfers of the regional development banks and through bilateral development assistance are positive but small. They pale compared to the kinds of transfers that postwar Europe received under the Marshall Plan, and that Japan, Korea and Taiwan received from the U.S. in the 1950s. Yet we face a world today in which there is good evidence that without external capital poor countries simply do not catch up to rich countries in income terms in any reasonable time period. Even in the postwar period of rapid growth of some developing countries, convergence of per capita incomes has been more the exception than the rule. True: economists have shown that there is conditional convergence, i.e. that under the right conditions – with good policy especially open markets, prior investment in human capital, strong institutions, and small but effective government¾ poor countries can and have caught up. But it seems to take an unusual confluence of good politics and prior institutional strength in poor countries to generate these particular conditions. Few poor countries in fact have these conditions; it is the nature of their poverty not to have them. In their absence, private returns to investment are uncertain and low, and the risks and transactions costs of private lending discourage all but short-term lending with high country risk premia.
This is not to say that private transfers will not occur at all. However, it does suggest that public transfers could catalyze or crowd in more private transfers to more countries. Those public resources are critically needed to support the institutional and human capital investments that will not attract private resources (since private agents cannot easily capture the benefits of such investments), but that will set the stage for private investment and lending.
So stable net public transfers to countries exercising sensible policy choices should be bigger, and more consciously structured to crowd in private investments whenever possible. They are key to the diversification of private risks across developing country markets, and to the reduced volatility of total transfers that can make open capital markets more politically and socially viable. They are in the enlightened self-interest of the richer countries. They are currently small (or when large fulfilling more political functions as with the large U.S. transfers to Egypt, Israel and the Palestinians) in part because of free-rider problems among the richer countries (the Nordics who are currently the most generous will only go so far on their own) and because of a history of reliance on what has become a more and more reluctant U.S. role. Only at the global level can the coordination problem or market failure that reduces from its optimum the size of these transfers be addressed.
Second, the global community would benefit from automatic Keynesian-style stabilizers for emerging market economies that are hit by global liquidity crises. These stabilizers should play a role analogous to the fiscal and monetary countercyclical policies that OECD countries implement during recessions. The IMF could provide such stabilizers in the form of transfers triggered under certain conditions. These transders would have an objective different from that of traditional IMF support. Their purpose would not be to build up reserves and thus market confidence, but to support current spending – on unemployment, public works and possibly credit programs, and on pre-existing health, education and other social programs. They would have to take the form not of short-term loans at high cost (the current approach for emergency IMF lending into a crisis) but long-term loans at relatively low cost. The implied fiscal stimulus would have to be temporary – as appropriate in the face of a liquidity crisis.
On the one hand, countries would not be eligible for such automatic stabilizer resources unless they were already demonstrating fiscally responsible policies. They would need the ability to manage a minimal deficit net of these short-run transfers; the political ability to define clear sunset provisions and to phase out such programs; and the long-run fiscal capacity to finance the resultant debt service. On the other hand, the implication of automaticity is that such transfers would be triggered quickly given prespecified conditions, without additional "conditionality" related to structural reforms. Korea, Thailand and Malaysia would have benefitted from such automaticity in 1998, as would have Mexico in 1995.
The World Bank and the regional development banks have supported loans for "social safety net" programs, and should continue to do so. But in their case support has too often been triggered suddenly at times of crisis (as it was in Mexico and Argentina in 1995, and in Asia in 1998) when the situation is hardly conducive to design of new, serious programs that are appropriate for the long run. The banks’ support should come in the context of the design and implementation of a permanent safety net, with full emphasis on structural issues including design, evaluation, and visible arrangements for phasing out as well as phasing in what should be only temporary bouts of countercyclical spending. The IMF's support should finance that spending when a liquidity crisis would otherwise require a developing country to tighten unreasonably its fiscal stance, exacerbating rather than offsetting economic and social costs.
This approach would address the risk that the fiscal and monetary discipline which the markets demand of emerging markets, which even if apparently sensible for each individual country, will add up to global deflation at times of financial panic. It would also address the criticism raised by Mahbub ul-Haq some years ago and more recently by Joseph Stiglitz, that the World Bank and the IMF have for the last two decades been the instruments of deflationary policies in poor countries – with social costs to the countries as well as the economic risk that such discipline can involve a tradeoff, i.e. that the economic policies meant to strengthen market confidence in a country and its currency may also increase bankruptcies and unemployment and thus perversely feed an additional loss of market confidence.
I recognize there are difficult issue of definition, eligibility, and so forth in such a proposal for a kind of global Keynesianism. The point now is to put the issue on the table, not to resolve it.
Third, the international financial institutions could pay more attention to the political reality of the inequality of wealth and of assets in many developing countries. There is evidence that this inequality discourages growth for reasons noted above, and that the market reforms the international institutions support can actually exacerbate income inequality, at least in the short run, undermining their potential growth gains and generating political and social problems, unless those reforms are more explicitly designed and implemented to be equity enhancing. Why shouldn’t there be more international support for land reform, for education that reaches the poor? Why shouldn’t there be more loan conditionality emphasizing the slashing of subsidies to insiders? Land reform and elimination of subsidies are difficult political choices, and often there are sound technical arguments against their rapid and oversimplified implementation. But they do deserve more concerted and more official attention than they have gotten for the last several decades. If the international institutions are to use their financial resources to push for policy reforms in poor countries, they should be pushing among other things for policies to reduce destructive inequality. One example: recent initiatives of multilateral debt relief for poor countries could be grounded in explicit attention to issues of distribution within benefiting countries.
Fourth, the OECD countries in their own self-interest could revisit their trade stance as it affects developing countries. Protection of agriculture and of textiles discriminates against poor countries and against the poor within countries. The head of the WTO has proposed elimination of tariffs on all imports of the world’s poorest 50 countries. The benefits would be enormous within those countries as well as to consumers everywhere. On the part of the developing countries, fresh thinking is needed on how to assist the rich countries manage the increasing political pressure to link trade with developing countries to improved environmental and labor standards in these developing countries. It is no longer sufficient for developing countries to argue that such pressure is merely reflecting the interests of rich country protectionist groups. The reality is that in a world of heightened global integration, with all its benefits, the arguments for more global cooperation on protection of the environment and the rights of labor must be taken seriously. More active leadership from the largest and richest of the developing countries on the substantive issues in global fora would make it easier to segregate the standards question from the issue of trade sanctions.
Fifth, the question of how governments can continue to finance the social compacts that have made unfettered markets politically palatable (and that have in some cases by fortunate coincidence also financed real increases in productivity of people) in the face of footloose capital and the international mobility of highly skilled labor, deserves attention. The risks of a race to the bottom in the form of tax and regulatory competition are well understood, but the actual evidence and the potential policy remedies in the hands of governments, either individually or collectively, are far from clear. Should there not be some more systematic assessment of the nature and the size of the problem? Is this an issue for the World Bank research complex, or for the UNDP? Should it be on the agenda of the World Bank’s Development Committee?
This brings me to my final implication for global policy – a point that has been raised recently by Peter Sutherland as Chairman of the Board of the Overseas Development Council. It is high time that the world had a forum in which the developing countries were adequately represented – for coordination of macroeconomic policies, and for discussion of trade, labor, environment and the international financial architecture. For a global economy in which the interests of developing and developed countries are increasingly intertwined – as the recent financial crisis all too well demonstrates – global governance is needed. Global governance requires a forum in which the benefits of repeated interaction and reciprocity can be realized.
