Poverty and Income Distribution
POVERTY AND INCOME DISTRIBUTION
Poverty and income distribution have risen to the top of the list of social issues in many countries. Since the 1970s the United States and United Kingdom have experienced increases in both poverty and economic inequality. But these countries are not unique; many developed countries have experienced at least modest increases in the inequality of income. As economies and labor markets become more international and countries wrestle with the social and economic consequences of aging populations, increased market work by women, and marital dissolution, public interest has come to focus on how successfully different social policies cope with inequality, poverty, and joblessness.
Definitions and Measurement
Poverty is measured by a lack of resources relative to needs. Resources can be measured by consumption, assets, or income, though most analysts prefer income because of the availability and comparability of relevant statistics. Needs measures can be either relative or absolute. Relative deprivation is almost always the preferred measure, both nationally and cross-nationally, because it examines deprivation subject to a household's social and economic context. There is no single best measure of absolute poverty for precisely this reason. Depending on the country, period, and context, the World Bank uses poverty lines of US $1.00, $2.00, or $3.00 per person per day. In contrast, the "official" poverty line for the United States is set at a level of $10.00 to $15.00 per person per day (depending on household size). Relative poverty can be defined using any of a variety of measures. The United States' "absolute" poverty line is approximately 40 percent of the median household income. Most international analysts, and many governments, choose a poverty line of 50 percent of the median. The European Union has chosen a line of 60 percent of the mean for measuring deprivation.
Since there are economies of scale in the consumption of most household goods, income and other measures of resources are usually adjusted for these differences by means of an equivalence scale. The equivalence scale measures the cost of providing an equal level of living for households that differ by characteristics such as household size or age of members. For instance, household size raised to the power 0.5 is a common equivalence scale adjustor. This adjustment says that if a single person needs 100 monetary units to be non-poor, a consumption group of four persons needs 100 × 4(.5) or 200 monetary units to be non-poor. Measures of poverty include the head count (the fraction poor), the poverty gap (the average additional income of those below the poverty line needed to bring them up to that line, and more sophisticated measures.
Income inequality refers to the distribution of income among households or persons. But the distribution of what, measured when, and among whom? Most analysts of inequality use a measure of disposable money income. For most households, the primary income source is market income, which includes earned income from wages, salaries, and self employment, and other cash income from private sources such as property, pensions, alimony, or child support. In calculating disposable income, governments add public transfer payments (e.g., retirement, family allowances, unemployment compensation, welfare benefits) and deduct income tax and social security contributions from market income. Most analysts measure income on an annual basis. This may be too long an accounting period for families that are severely credit-constrained and too short for those that can smooth consumption over several years, but almost all available surveys report income for the calendar year.
The usual answer to the question "distribution among whom?" is "among individuals." Most surveys focus on the individual as the unit of analysis and the household as the unit of income sharing. A household is defined as all persons sharing the same housing unit, regardless of any familial relationship. One therefore estimates individual disposable income by aggregating the income of all household members and using an equivalence scale to arrive at individual equivalent income per person equivalent income. Equal sharing of incomes within the household is assumed.
There exist many different summary measures of inequality, most of them based on the Lorenz curve, or other variants. Their usage is demonstrated below.
Databases for Measuring Poverty and Inequality
A heightened interest in poverty and inequality has led to greater efforts to assemble comparable crossnational measures of economic inequality–not an easy task, because the data that exist are not uniform in nature or purpose. Some national surveys are designed to collect income data and some to collect expenditure data. Some are longitudinal household panel surveys, while others are cross-sectional income or labor force surveys. For some countries, data are derived from income tax or administrative records. Despite the difficulties, projects such as the Luxembourg Income Study (LIS) and, to a lesser extent, the International Social Survey Program (ISSP) are helping create a richer body of comparative economic studies. It has become possible to provide a more complete picture of cross-country differences at many points in the income distribution, instead of merely providing snapshot comparisons of the "average" or "typical" family in different countries. Researchers have not only been able to address the factual question of whether inequality has grown in a particular country, but also to start to probe more deeply into the factors explaining changes in economic inequality.
The LIS provides standardized measures of poverty and inequality for a set of 25 high-income countries over the period from 1979 to 1999. These figures can be found on the Internet. As of 2002 it is the only such database. Data on poverty and inequality compiled from a number of different country sources are typically not fully comparable. Trend data, which rely on changes in the same measure within a single country, are likely to be more reliable, assuming that there is no substantial change in survey design or measures. The set of estimates briefly summarized in Figure 1 is based on the LIS.
Relative Differences in Poverty and Inequality Across Nations
A large body of research has documented comparative levels of poverty and inequality among countries and also the substantial increases in inequality in many countries over time. How do countries measure up? Figure 1 compares the distribution of disposable income in 22 countries for various years around 1995. It highlights the relative differences between those at the bottom and those at the top of the income distribution. It shows the ratio of the income of a household at the 10th percentile (P10) and a household at the 90th percentile (P90) to median income for each country. This indicates how far below or above the middle of the distribution the poor and the rich are located on the continuum of income. Figure 1 also shows the ratio between the incomes of those at the 90th and 10th percentiles (the "decile ratio"). This illustrates the size of the gap between the richest and the poorest in each country. LIS also uses the most common (Lorenz curve-based) measure of inequality–the Gini coefficient.
Most measures of inequality, including those presented in Figure 1, are conducted on a relative basis within nations. (With careful use of Purchasing Power Parties, one may also be able to compare income distributions and percentiles of the distribution among similarly developed countries in real income terms.) The measures describe relative social distance. They are easy to understand but focus on only a few points in the distribution of income.
Figure 1 shows that the United States has an exceptionally large gap between the rich and the poor when compared to other advanced market economies. A low-income American at the 10th percentile in 1997 had an income that was 38 percent of median income, whereas a high-income American in the 90th percentile had an income that was 214 percent of the median. The high-income American had an income nearly six times as much as the low-income
American, even after adjustment for taxes, transfers, and family size (the decile ratio is 5.63). In contrast, in the other countries in Figure 1, the income of the poor averaged 51 percent of the median income; that of the rich, 184 percent. The average rich person in these countries had only 3.6 times the income of the average poor person.
The countries in Figure 1 fall into clusters. Inequality is lowest in Northern Europe (the Scandinavian countries and Finland, and also Luxembourg and the Netherlands), where the income of those at the 10th percentile averages 57 percent of the median. A number of Central and Western European countries come next (Germany, Belgium, Austria, Switzerland, and France–plus Taiwan, included for comparison). Israel and the United Kingdom have the highest levels of inequality outside the United States. In some countries–Italy, Ireland, Israel, and the United Kingdom–the incomes of the richest people, those at the 90th percentile, are more than 200 percent of median income. This is not very different from the United States; the United States differs, above all, in the relative disadvantage of its poorest residents.
Table 1 shows poverty rates (fraction of persons below 50 percent of median income) in these same
countries for all persons, children, and the elderly. Once again, these figures show that the United States stands apart from the other countries in the study, with the highest levels of poverty for the total population (16.9%) and for children (22.3%). In fact, more than one child in five fell below the poverty line in the United States in 1997. Only Australia had a higher percentage of elderly persons below the poverty line (nearly 30% of elderly Australians were living in poverty in 1994). At the other extreme, only3.9 percent of Luxembourgers (1994), 2.6 percent of Swedish children (1995), and 2.7 percent of elderly Swedes were below the poverty line in their countries. Average 1990s poverty rates for the Table 1 countries apart from the United States are 9.5 percent for the total population, 10.8 percent for children, and 11.6 percent for the elderly.
Poverty and inequality measures differ across countries (e.g., compare the extremes, Luxembourg and the United States, in Figure 1). However, the Northern European countries tend to have the lowest levels of poverty, followed by Central Europe and then Southern Europe and the English-speaking countries (United States, United Kingdom, and Australia).
Extensions and Summary
Poverty and income distribution are concrete and valid measures of economic status. Broader measures of well-being may also include such items as health status and literacy, and are especially appropriate for developing countries. An example of such a measure is the Human Poverty Index, published in the United Nations Human Development Report. A broader analysis of global poverty is contained in the 2000/2001 World Development Report, issued by the World Bank. The effect of inequality and poverty on economic growth, crime, and related social outcomes is also a growing field of inquiry. Using measures such as those described above, and developing more datasets like the LIS, should provide a clear picture of how well the world does in combating poverty and in understanding the effects of both poverty and inequality on social well-being.
Atkinson, Anthony B., Lee Rainwater, and Timothy M. Smeeding. 1995. "Income Distribution in OECD Countries: Evidence from the Luxembourg Income Study (LIS)." Social Policy Studies 18.
Gottschalk, Peter, and Timothy M. Smeeding. 2000. "Empirical Evidence on Income Inequality in Industrialized Countries." In Handbooks in Economics, Vol. 1: Handbook of Income Distribution, ed. Anthony B. Atkinson and François Bourguignon. Amsterdam: North-Holland.
Ishikawa, Tsuneo. 1996. Data runs conducted by the Ministry of Welfare, Japan.
Jäntti, Markus, and Sheldon Danziger. 2000. "Income Poverty in Advanced Countries." In Handbooks in Economics, Vol. 1: Handbook of Income Distribution, ed. Anthony B. Atkinson and François Bourguignon. Amsterdam: North-Holland.
Smeeding, Timothy M., Lee Rainwater, and Gary Burtless. 2002. "United States Poverty in a Cross-National Context." In UnderstandingPoverty, ed. Sheldon H. Danziger and Robert H. Haveman. New York and Cambridge, MA: Russell Sage Foundation and Harvard University Press.
Smeeding, Timothy M., and Lee Rainwater. 2001. "Comparing Living Standards Across Nations: Real Incomes at the Top, the Bottom and the Middle." Syracuse, NY: Luxembourg Income Study. <http://www.lisproject.org/publications/liswps/266.pdf.>.
United Nations Development Programme. 2001. "Human Development Report 2001." <http://www.undp.org/hdr2001/>.
World Bank. 2000. "World Development Report 2000/2001: Attacking Poverty." <http://www.worldbank.org/poverty/wdrpoverty/>.
David K. Jesuit
Timothy M. Smeeding