270 likes | 435 Views
Globalization and Inequality: Comments on Ravallion. Lant Pritchett, KSG, CGD, WB 13 May 2004. Two points. Current “globalization” talk is, and is bound to be, nonsense.
E N D
Globalization and Inequality:Comments on Ravallion Lant Pritchett, KSG, CGD, WB 13 May 2004
Two points • Current “globalization” talk is, and is bound to be, nonsense. • Most economists do not, and should not, believe in PWPL (Poverty with low poverty lines) as a social objective—and this is a defensible value judgment—an array of poverty lines is appropriate
Globalization and Inequality • Real globalization—which included mobility of people--would reduce inequality in well being across spatial regions: full stop. • We don’t have anything like globalization: • Nation-states have proliferated (more sovereigns, more borders, more monies) • Movement of people has been off the agenda.
Number of new countries added per year(Source: Braun, Hausmann and Pritchett, “The Proliferation of Sovereigns”)
Fraction of inequality due to across versus within countries (Source: Bourguignon and Morrison)
Wages Zombies Ghosts Population With negative region specific shocks of “desired” populations one either gets large changes in growth of populations (ghosts) or growth of income per head (zombies)
Large movements of population within (large) countries, large differences in income per head across countries
X and inequality “Does the set of processes called X in which (a) the number of nation-states rises, but (b) transport costs (including of information) and (c) some nation-states pursue modestly more liberal policies of some types towards the movement of goods and (perhaps) capital but not labor lead to Increase or decrease inequality across nations/individuals?
“Globalization” and Y • It is not clear “globalization” is the key phenomena of the 20th century—nationalism which is deeply anti-global is. • It cannot be the case that we have an answer as to the impact of “globalization” on Y (where Y=growth, inequality, the environment)—an ill-posed question.
PWLPL vs. SWF • Does X affect “poverty”? • FGT(α,PL) requires a poverty line • Gains to income/expenditure above the PL count zero towards reduction in poverty.
Contribution of income gains at various points in the income distribution to welfare measures(Source: Kraay 2004)
Value judgments and poverty • I personally do not have an objective function in which income/consumption gains count for zero above 1$/day or 2$/day. • The question is: where can the poverty line be drawn such that zero is even a reasonable approximation?
Arguments against low poverty line • Non-paternalism—there is no evidence of a sharp reduction in increases in subjective well-being w.r.t. expenditures around 1 or 2 $/day. • Non-money metric indicators—relationship between income/expenditures and mortality/health/education does not get lower anywhere near 1 or 2 $/day • Non-money metric indicators indicate the “rich” (upper 20 percent) in poor countries are enormously worse off than the “poor” in rich countries.
Arguments against low poverty line • Non-paternalism—there is no evidence of a sharp reduction in increases in subjective well-being w.r.t. expenditures around 1 or 2 $/day. • Non-money metric indicators—relationship between income/expenditures and mortality/health/education does not get lower anywhere near 1 or 2 $/day • Non-money metric indicators indicate the “rich” (upper 20 percent) in poor countries are enormously worse off than the “poor” in rich countries.
Arguments against low poverty line • Non-paternalism—there is no evidence of a sharp reduction in increases in subjective well-being w.r.t. expenditures around 1 or 2 $/day. • Non-money metric indicators—relationship between income/expenditures and mortality/health/education does not get lower anywhere near 1 or 2 $/day • Non-money metric indicators indicate the “rich” (upper 20 percent) in poor countries are enormously worse off than the “poor” in rich countries.
Infant mortality has (roughly) the same elasticity w.r.t. GDP per capita over a large range of (average) incomes
Arguments against low poverty line • Non-paternalism—there is no evidence of a sharp reduction in increases in subjective well-being w.r.t. expenditures around 1 or 2 $/day. • Non-money metric indicators—relationship between income/expenditures and mortality/health/education does not get lower anywhere near 1 or 2 $/day • Non-money metric indicators indicate the “rich” (upper 20 percent) in poor countries are enormously worse off than the “poor” in rich countries.
Infant mortality of the richest 20 percent in selected countries
Malnutrition of the richest 20 percent in selected countries
Arguments against low poverty lines (con’t) • Revealed preference within and across countries • Poverty lines rise with income both within countries and over time • There is absolutely no compelling factual or value argument for low poverty lines. • LPL are not “biologically” based, e(p,Upov) • Use of LPL is fine, but let us not confuse type I and type II error—almost certainly no one below a LPL is “not poor” but many (billions) are above LPL and are poor by reasonable definitions.
A proposal • For 1$/day poverty we use the term “destitution” • For 2$/day we use the term “extreme poverty” • Each country uses its own ‘national” poverty line • We use the capital weighted average of the World Bank shareholder poverty lines—something like 15$/day—to define “global poverty”
World Distribution of Income and “Poverty” National poverty lines Global Poverty Extreme Poverty Destitution
Poverty is a social construct • This is not about having a “poverty focus” or not having a “poverty focus” this is about choosing a poverty line. • For global discussions there is no reason to insist on a low poverty line as the only or even the major metric for success. • I come not to bury poverty but to raise it.