Growth-Enhancing Policies Are Good for Poor People
Critics of market economies in general, and globalization in particular, contend that they have unleashed forces leading to large and pervasive increases in inequality within countries. In a recent paper we examine household survey evidence from 80 countries over the past four decades and find that this is simply not so. When inequality is on the rise, poor households benefit less from economic growth than wealthier ones.
In fact, average incomes of the poorest fifth of society rise proportionally with per capita income, indicating that inequality does not systematically increase with growth. Since few countries show significant trends in income inequality, on average economic growth has been the main driving force of poverty reduction in developing countries. A good example in the 1990s is Vietnam, which experienced rapid per capita GDP growth of 6 percent per year with no significant change in the distribution of income. This distributionally-neutral growth led to tremendous improvements in the material well-being of poor Vietnamese.
Of course there are deviations from this general relationship between growth in average incomes and growth in incomes of the poorest fifth of the population. Our research suggests that popular explanations for these deviations are not supported by evidence. In particular, a) the relationship between growth and incomes of the poor does not differ between periods of crisis and periods of normal growth; b) growth’s impact on the poorest quintile has not weakened in recent decades as globalization has become more pronounced; c) growth spurred by an open trade regime or other growth-enhancing policies such as good rule of law and macro stability does not in general have adverse effects on poor households. In fact such policies, on average, benefit poor households as much as the typical household, and some policies, notably stabilizing from high inflation, disproportionately benefit the poor.
Clearly, growth-enhancing policies such as sound rule of law, macro stability, and openness to trade are not all that is needed to improve the lives of the poor. But to the extent that such policies underpin growth they are a critical component of poverty reduction. Anyone who cares about the well-being of the poor should therefore support developing countries as they participate more in international trade and put in place a healthy environment for economic growth.
Growth is good for the poor
For this study we gathered household survey data for 80 countries for the past four decades. We define the poor as the bottom one-fifth of the income distribution, mainly because data on income distribution by quintile shares is most abundantly available. Of course there is a great deal of variation in incomes within the poorest quintile in a given country. However, we find very similar results using the more limited data on the poorest tenth of the population, which gives us confidence in the robustness of our findings.
Figure 1: Average Incomes and Incomes of Poor People
The basic relationship between the average income of the bottom quintile and per capita income is very striking (Figure 1). If the distribution of income were everywhere and always the same, the points would fit a perfect straight line. The data do not fit the line perfectly, but they fit it very well (the R-squared is 0.87). What this means quite simply is that poor people have higher income in rich countries not because they typically have a higher share of total income but because they happen to live in rich countries – that is, countries that have grown relatively rapidly over the past century. Even if the distribution of income in poor countries such as Ethiopia or Bangladesh were perfectly equal, annual incomes would only be $566 and $1407 respectively. Such incomes are still vastly lower than average incomes in the bottom quintile in a middle-income country such as Korea ($4650) or rich countries like the United States ($5263). 2 To us, this point should be non-controversial: it is not possible to make much progress against poverty in poor countries such as Bangladesh or Ethiopia by sharing the available pie more equally— a large pie is required.
Figure 1 reflects a very long-term relationship between growth and poverty. It is also important to see what is likely to happen over shorter periods of time. For this we have used the data in Figure 1 to calculate episodes of growth over periods of at least five years: we have 236 such episodes in which we can relate growth of income of the poor to growth of per capita GDP (Figure 2). Two important things to note here: first, the basic relationship is still one-to-one – that is, countries can expect that growth will raise the income of the poor proportionally; but, second, there is a lot more variation around this general relationship. Growth over a five-year period almost always has material benefits for the poor, but the extent of the benefit does vary around that one-to-one relationship.
Figure 2: Growth and Growth in Incomes of Poor People
Clearly it would be useful to know what causes growth to be more or less pro-poor. The main point of our study is to try to explain why some observations in Figure 2 are above the line (poor people benefit more than proportionally from growth) and some are below the line (poor people benefit less than proportionately from growth).
Openness is good for the poor
A number of popular ideas about globalization essentially are hypotheses trying to explain the deviations around this general relationship. For example: our data go back several decades. One idea about globalization is that it has changed the extent to which growth reaches poor people, which is the hypothesis that if we separate the earlier observations and the more recent ones we will find different relationships. Not so. Growth is just as pro-poor in the 1980s and 1990s as in earlier decades. Another idea is that globalization has made countries more vulnerable to crises that are particularly hard on the poor. Without taking a stand on whether globalization contributes to crises, we can ask whether incomes of the poor fall disproportionately when average incomes fall during a crisis. We find that they do not.
Figure 3: Trade and Inequality
Many of the demonstrators who disrupted the World Trade Organization’s meetings in Seattle last year argued that international trade and capital flows disproportionately benefit multinational corporations and the rich. In our framework, this is the hypothesis that growth spurred by greater openness will generate points “below the line” – that is, biased against the poor. We explored this idea with three measures of openness: actual trade volumes, a measure of open trade policy developed by Jeffrey Sachs and Andrew Warner, and a measure of capital controls from the International Monetary Fund. Contrary to popular belief, none of these measures of openness is related to inequality. The relationship between changes in actual trade volumes and changes in inequality (measured, for example, by the Gini coefficient) is particularly striking: it is simply not true that countries that trade more become more unequal (Figure 3).
Good policies are good for the poor
We next examined whether policies and institutions that are good for growth explained deviations from the one-to-one relationship. We focused primarily on a set of variables capturing sound macroeconomic policies and a strong institutional environment, in order to determine whether these were systematically associated with changes in income distribution. The only robust finding is that high inflation is particularly bad for the poor (it leads to slow growth for the whole economy and to especially slow or negative growth for income of the poor – that is, points “below the line”).
Figure 4: Growth and Distributional Effects of Policies on Incomes of Poor People
On the one hand it is disappointing that we cannot explain very well what makes growth more pro-poor. But, on the other hand, our findings do convey the good news that developing countries can pursue growth-enhancing policies, such as sound macroeconomic management and good rule of law without fear that these typically increase inequality. Since these policies raise average incomes without much effect on income distribution, they benefit poor people as much as the economy as a whole. We can summarize this finding in terms of the estimated impact of different policies on income of the poor–decomposing the effect into a component coming through growth and a component coming through distribution. Incomes of poor people will typically be increased by better rule of law, more openness to trade, and better macroeconomic management through less inflation and smaller government consumption (Figure 4). These policies raise incomes of poor people primarily by spurring overall growth. Their effects on incomes of poor people that operate through changes in the distribution of income are typically small, and with the exception of stabilizing from high inflation, statistically insignificant.
All this brings us back to the story of Vietnam in the 1990s: reforms returned land to family farms, liberalized prices and trade, and stabilized the macro economy. Poor rice farmers benefited immediately from these changes. Of the poorest 5 percent of families in a 1992 survey, 98 percent had higher income and real consumption six years later. The share of income going to the bottom 5 percent did not change over the period, but real income basically doubled for all income groups. Thus, the evidence is clear that growth-enhancing policies are good for the poor.
David Dollar is research manager, Development Research Group, The World Bank. Aart Kraay is senior economist, Development Research Group, The World Bank.
1. Dollar, David and Aart Kraay (2000). “Growth Is Good for the Poor”. Available at: www.worldbank.org/research/growth
2. These figures are based on per capita GDP for 1998 in current dollars adjusted for differences in purchasing power parity as reported by the World Bank. Incomes in the bottom quintile are based on a quintile share of 7 percent for Korea and 3.6 percent for the United States in the mid-1990s.