THINGS ARE ALWAYS GETTING WORSE

Much of the current discourse on international development appears steeped in a deep pessimism about the low-income world. Many casual observers appear to assume that international development trends are inevitably negative, despite the healthy rates of economic growth that prevailed in much of the low-income world during the last decade, at least before the current recession took hold. Echeverri-Gent's article seems informed by this pessimism when it implies that all forms of inequality are getting worse. In fact, it is important to distinguish between several different types of inequality. The claim that inequality is worsening is usually made with data about international inequality, in other words the inequality between the mean per capita national incomes of the countries of the world. There, the growing gap between rich and poor countries appears incontrovertible (Pritchett Reference Pritchett1997). As Pritchett suggests, in striking contradiction of neo-classical economic theories of growth that predicted international income convergence, the nineteenth and twentieth centuries witnessed striking divergences. With a small number of exceptions, mostly in East Asia, the poor countries either grew no faster, or actually grew more slowly and got poorer, relative to the rich countries. The poorest countries grew the most slowly and remain mired in devastating poverty. The failure of development in this “bottom billion” should be considered among the most pressing public policy issue for the international community in the coming years (Collier Reference Collier2007).

Most analysts, including Echeverri-Gent, say little about quality-of-life issues and non-income international inequality, though arguably they have at least as much political salience as income inequality. In fact, one of the empirical puzzles of the contemporary era is the appreciable cross-national convergence in quality-of-life indicators, from infant mortality and life expectation to educational attainment, even as national incomes diverge (Kenny Reference Kenny2005). Thus, even as the poorest countries have fallen further behind the West in terms of per capita income, they have managed to continue to improve social indicators faster than in the rich countries of the Organisation for Economic Co-operation and Development (OECD). Convergence on this score is to be sure highly imperfect, and the quality of life of broad segments of the population in many low-income countries remains shockingly low. Nonetheless, this is an important component of the inequality picture and should not be ignored, both because it mitigates the ambient pessimism about the possibility of progress in these countries, and because, as I argue below, it may well have significant political implications.

So much for international inequality. Many observers also suggests that another kind of inequality, interpersonal income inequality within developing countries, is increasing sharply, but as a general proposition of a global trend, this is more problematic. The World Bank paper cited in this context by Echeverri-Gent (Ferreira and Ravallion Reference Ferreira and Ravallion2008), based on probably the best current cross-national database on in-country poverty and inequality, detects a slight overall trend in growing income inequality, but only between the mid-1990s and the mid-2000s, and cautions against over-interpreting the trend, given the fact that the data come only from some 49 countries in the developing world (Ferreira and Ravallion Reference Ferreira and Ravallion2008, 8). Longer-term trends are harder to discern in the low-income countries, in part simply because comparable longitudinal data do not exist, and in part because the partial and unsatisfactory evidence available suggests no single trend across the last few decades across region or income level. In the case of the 48 countries of sub-Saharan Africa, for instance, income-distribution data from before 1965 exist for only 16 countries, mostly from surveys of suspect quality. Indeed, as I have argued elsewhere (van de Walle Reference van de Walle2009), the paucity of data points before the 1990s means there is simply not enough comparable longitudinal income-distribution data from sub-Saharan Africa to assert a clear regional trend in the post-colonial period with any confidence. Data are more plentiful for middle-income developing countries, such as those in Latin America, but there is a danger in generalizing about all developing countries from data that overemphasize only the richest of such countries.

In any event, rather than rapid change in the levels of domestic inequality, what is striking about them is the combination of the relative stability of national indicators of inequality and the variation in their levels across countries. Thus, the change in developing-country Gini coefficients between the mid-1990s and the mid-2000s reported in Ferreira and Ravallion (Reference Ferreira and Ravallion2008) very rarely exceeds .05; on the other hand, Ginis vary from under .35 to over .65 for the countries with data during this period. One can argue about the evolution of inequality in Brazil (the same World Bank Povcalnet data suggest the Gini went from .615 in 1995 to .570 in 2004), but the more salient fact must be that compared to some other middle-income countries, such as Indonesia (.344 in 1993 and .343 in 2002), Brazil's high inequality appears to be a persistent feature of its political economy. Bigger changes in inequality over time are rare, and typically associated with radical domestic policy changes, such as revolutionary land reform, which fundamentally alter the distribution of individual assets.

In addition, the stability in national inequality levels over time is often quite striking, not least because it suggests that the underlying causes of inequality are quite old, and that powerful path dependencies mitigate the impact of short-term policy shifts in the present era. In the case of sub-Saharan Africa, the tentative evidence (van de Walle Reference van de Walle2009) suggests that variation in colonial experiences largely shapes current levels of inequality, despite half a century of policy changes since independence. In a very suggestive analysis, Huillery (Reference Huillery2006) argues that variations in inequality in francophone Africa today are largely explained by patterns of public investment under French colonialism in the first decades of the twentieth century.

Another striking feature of development pessimism is the view that economic growth has and can have only a limited effect on global poverty and inequality. The implication is that poor countries can't win—their poverty and inequality levels result equally from economic growth and its absence. A long economic tradition going back at least to Simon Kuznets (Reference Kuznets1955) argues that economic growth increases inequality. Yet, in fact, the evidence from the last 50 years suggests no straightforward relationship between growth and inequality (e.g., Bruno, Squire, and Ravallion Reference Bruno, Squire and Ravallion1997). In some cases, growth spurts have been associated with increases in inequality, but in just as many, the impact on distribution has been negligible or positive. The relationship varies according to both the level of development, and at different points in the income distribution (Voitchovsky Reference Voitchovsky2005).

Echeverri-Gent's article argues rightly that high inequality reduces the decline in poverty brought about by economic growth, but it fails to add that during the course of the last 30 years, economic growth has been associated with substantial declines in the rate of poverty, involving many hundreds of millions of people. The World Bank thus estimates that the percentage of people living under the so-called one-dollar-a-day poverty line was cut by more than half between 1981 and 2004 (Ferreira and Ravallion Reference Ferreira and Ravallion2008). Though much of this decline can be attributed to India and China's tremendous economic success over this period, the decline of poverty is fairly broadly distributed. Even sub-Saharan Africa, where poverty levels were growing through the mid-1990s, has since benefited from reductions in the poverty rate, due to its recent economic success (Ndulu Reference Ndulu2007). To be sure, these gains are vulnerable and should not obscure the reality that absolute poverty remains much too high, but they represent an important progress.

The significant rates of economic growth in the low-income world over the course of the last decade, adding to the earlier economic miracles in China and India, have led to a decline in a third kind of inequality, which economists call “global” income inequality, relating to the distribution of individual income across the world. It appears to be declining since the early 1990s, after over a century of steady increases (Firebaugh Reference Firebaugh2003; Milanovic Reference Milanovic2005). Echeverri-Gent rightly expresses shock at the mind-boggling differences in income between the world's richest and poorest people, and growing global inequality may have been one of the truly significant trends of the nineteenth and twentieth centuries. On the other hand, as Firebaugh (Reference Firebaugh2003) has argued, the last 20 years have witnessed an increasingly sharp shift in this trend, and the decline in poverty in the developing world, and the concomitant rise of a massive middle class in the developing world, most notably in China and India, has huge implications for contemporary international political economy. China and India's newfound diplomatic and economic leverage are perhaps the most dramatic consequence of this rise, but there are others.

ALL BAD THINGS GO TOGETHER

Robert Packenham (Reference Packenham1973) chides the modernization literature for mistakenly believing that all good things went together in the development process. The more common mistake today may well be the opposite view that there are simple causal linkages between undesirable things in the development process. Above, I tried to dispel the tendency to view poverty and inequality as interchangeable phenomena that always co-vary. Another common empirical claim about bad things being causally linked is the argument that domestic and global inequality causes political instability and violence.

Recent work suggests a causal linkage between extreme poverty and poor economic performance, on the one hand, and instability in the developing world, on the other. For instance, Collier (summarized in Collier Reference Collier2007) links low GNP per capita with state failure. Boix (Reference Boix2003) argues that high levels of inequality lower the likelihood of successful democratization. Cheibub et al. (Reference Cheibub, Przeworski, Neto, Papaterra and Alvarez1996) show that the stability of low-income democracies relies on sustained rates of economic growth.

The cross-national literature on intra-state violence has not found nearly as compelling a link between inequality and intra-state violence, however (e.g., Fearon and Laitin Reference Fearon and Laitin2003), and has increasingly turned its attention to other causes. Similarly, and more recently, Kaufman (Reference Kaufman2009) questioned the effects of economic inequality for a wide set of political outcomes in Latin America. Again, the absence of clear correlation in the statistical work may be due to the low quality of the inequality data, and to the paucity of comparable historical income-distribution data. Nonetheless, other reasons should lead us to question such a clear causal relationship. One lies in the different historical origins of income inequality; the political salience of domestic inequality may vary, depending on its causes, and its interactions with other variables, from the distribution of assets, notably land, to region, race, and ethnicity. For example, in some countries, inequality is largely spatial, with income varying by region because of historical reasons or basic resource endowments; in other countries, stratification is sustained at the local level and correlated with caste or race. These very different inequality dynamics can have very different political implications, despite not being picked up by aggregate inequality measures, such as the Gini coefficient.

Another complication may be provided by the complex relationship between social mobility and inequality. The existence of social mobility, or even the perception of its existence, may legitimate situations of intense inequality, an argument that has been made about the United States. Echeverri-Gent's argument, following Tilly (Reference Tilly1998), points to an inequality trap, in which yesterday's inequality promotes tomorrow's.

This view accords with my argument above about the long-term persistence of inequality, but it should be nuanced. Bénabou (Reference Bénabou, Bernanke and Rotemberg1996) has provided evidence for the existence of “inequality convergence”—in other words, that the level of inter-country inequality has tended to converge over the course of the last few decades, as high-inequality countries have tended to become more equal, even as low-inequality countries have tended to become more unequal. The reasons for this are contentious, but at the very least this evidence should lead us to caution about too simple an endogenous model of self-sustaining inequality.

One cannot deny that elites at the very least attempt to structure institutions they dominate in a self-interested manner, with the objective to reproduce their own economic and political power. Nonetheless, it is far less clear that they are able to do so. Even when they are united, which is sometimes far from the case, elites often fail to understand their true economic interests and may be tied to anachronistic assets of diminishing value. The institutions they create may actually serve their economic and political interests poorly.

Perhaps for these reasons, there is surprisingly little quantitative cross-national evidence of a positive correlation between high levels of inequality and perceptions about the degree of social mobility. Not all relatively equal societies have a high level of social mobility (think India, for example), while not all highly unequal societies have low levels of social mobility (think Brazil). If people's political attitudes are shaped as much by their beliefs about their future prospects as by their understanding of their own relative income in the present, then beliefs about social mobility may be more important to political outcomes than inequality per se (Kaufman Reference Kaufman2009).

The combination of high degrees of international and domestic inequality should not lead one to the thesis of a causal linkage between these two forms of inequality: the notion that high levels of international inequality are conditioned by the political and institutional imbalances they generate, and that this dynamic transfers inequality from the national to the international level and back. The organizations of the international community, it is sometimes suggested, reinforce inequality within the developing world, even as they maintain international inequality (Held and Kaya Reference Held and Kaya2006). There is a kernel of truth in such a general claim, insofar as the domestic and international political economies are today inextricably linked. In addition, it helps to explain the persistence of domestic inequality over time that I noted above. Nonetheless, it is important to stress the limits of such an argument, which again has bad things going together, in this case global, international and national inequality.

It is important not to assume that global inequality and domestic inequality co-vary. Echeverri-Gent is right to argue that the agendas of some international organizations such as the WTO have been driven by rich-country concerns, and have lacked what Wade (Reference Wade2003) terms a “developmental” dimension. Of course, the impact of WTO policies on global and international inequality is already hard to assess and the source of much contention (e.g., Harrison Reference Harrison2007); their effects on domestic inequality within the developing world is even less clear. Again, it is important not to assume domestic, global, and international inequality co-vary.

One example of what critics of international institutions appear to have in mind, though they rarely mention it, is provided by the fact that the ability of contemporary low-income governments to accumulate and sustain enormous levels of public debt has been enhanced by the current set of international institutions. By providing resources to governments, the international financial institutions on balance probably helped maintain the domestic status quo, in a way that would not have been true in an earlier era when governments would have had to adjust more quickly to the vicissitudes of the business cycle. International support thus clearly helped maintain political stability in Africa through the era of structural adjustment: between 1979–1999, the region combined the lowest numbers of executive turnovers of any region in the world, with the worst economic record and very rapid accumulations of international debt (van de Walle Reference van de Walle2001).

On the other hand, other current dynamics lead to skepticism about such a transfer of international inequality to the domestic level. The system of international organizations does result in the mobilization of tens of billions of dollars for low-income countries each year, much of it with an explicit anti-poverty focus. The aid system is often woefully inefficient and it has been often too accommodating to local power structures (Easterly Reference Easterly2001; van de Walle Reference van de Walle2005), but on balance, it has been a force for poverty alleviation and the formulation for public policies that are less regressive than those pursued by many low-income governments.

I agree with Echeverri-Gent that wealthy-country experts have dominated many of the epistemic communities that shape the dominant policy paradigms within the international community, but I would resist the implication that increasing the number of experts from poor countries would necessarily lead to a leftward movement of the social welfare function of these communities. Certainly, in the area of international foreign assistance, I would argue they have on balance served to promote the legitimacy of genuine poverty reduction and greater equality in the public policies of developing countries than had been generated by domestic policy processes. Much of the pro-poor development agenda that has appeared in the last 30 years has originated in the donor organizations, including the moving of resources from tertiary to primary education, attempts to lessen the urban bias in the delivery of services and infrastructure, or the increased emphasis for preventive rather than curative medicine. Targeting the poor is notoriously difficult, and these policy shifts are not without problematic dimensions, but they represent explicit attempts to increase the incomes and choices of poor people in developing countries.

CONCLUDING REMARKS

In sum, this article wholeheartedly agrees with Echeverri-Gent that current levels of inequality represent a real concern, not only in normative terms, but also because of the negative political effects they can have. The unequal distribution of wealth within and across borders is the premier public policy problem of our era, and it needs to be a much greater focus of political science research than has been the case. In that sense the APSA task force on inequality was both welcomed and long overdue.

At the same time, it is important to avoid the view that the problem is unambiguously getting worse, or that linkages between different dimensions of the problem make a worsening almost inevitable. Not all types of inequality are undergoing the same evolution, and the links between them are complex and hardly straightforward. Some forms of inequality may actually be improving. The lack of progress on some problems does not prevent real progress on others. Resisting the “everything always gets worse” and “all bad things go together” views does not make one a Panglossian optimist, or blind one to the enormous costs of global poverty and inequality, which are real and pressing. The complexity of social reality should not lead us to complacency, but should lead us to improved efforts to effectuate change.