Transcript Regional Trends for South, East, Southeast Asia & Oceania

Global Patterns of Income and Health:
facts, interpretations and policies
By Augus DEATON, October 2006
Research Program in Development Studies, Center for
Health and Wellbeing, Princeton University
Presentation by Stéphanie Carret
10.11.09
1
The planning for today
1.
2.
3.
4.
Review of the paper: main ideas
Analysis of illustrative graphs
Other views on the subject
What questions can we raise?
2
Dataset used
• Different dataset are used, especially since it’s difficult to compare
health levels at an international scale
 No natural metric of health (which is not the case for income)
 How to measure inequality with arbitrary chosen measures of health?
• Most of the data used extends between 1960 and 2000
 Penn World Table, WDI, WHO and World Bank for China & India
• What are the main health measurements used?




Life expectancy
Infant & child mortality rates (and absolute levels)
Types of widespread deadly diseases worldwide
GDP and growth rate
• The Preston curve focuses on the year 2000
 Relates GDP/capita to life expectancy
• The author shows that it’s essentiel to make a difference between
proportionnate (ratio) and absolute (level) rate of decline of infant
3
and child mortality
Main ideas (1)
• Main indicators for inequality
 Poverty; income & wealth inequality
 Health inequality
 In an overall, population in poorer countries have shorter life
expectancy and suffer/die from more diseases than in wealthier
nations
 BUT exceptions: Cuba, Sri Lanka, Costa Rica…
• Historically, the first noticeable global divergence in income &
health levels: 18th century
 This income inequality trend did not reverse
 But it did for health (beginning 20th & post WWII)
4
Main ideas (2)
• Between 1950 and 1990, convergence in international life
expectancy which does not mean a link between growth
rate/higher income and health levels
 In poorer countries, infant & child mortality rates have decreased
 In wealthier countries, mortality rates reduced for elderly people
 Between 1990 and today, there has been a big break on
international increase of life expectancy because of Aids/HIV
• India & China, health improvements were not correlated with
economical growth, or only in small increases
• Income poverty and health poverty are positively correlated
 Composite measure of wellbeing (income/capita X life expectancy) >
separate inequalities of income or health
 Other componants impact on income & life expectancy
• Education, political & civil rights, democracy…
5
Main ideas (3)
• Deaton argues that life expectancy is misleading when used
in international comparison
 Different patterns of mortality & mortality reduction (children vs
elderly mortality)
 How to compare different progress in saving lives?
 Inequality reduction in life expectancy does not mean reduced
health inequality
• Decrease of mortality in wealthier countries increased inequality
 Data poorly measured (sub-saharian Africa)
 Lack of vital registration in many countries
• Income growth improves population health?
 Direct effect: poverty reduction
 Indirect effect: health poverty reduction
6
Main ideas (4)
• Different papers, different variables explaining mortality
 Poverty & lack of income (Preston)
 Wagstaff shows the link between high child mortality with lower
expenditure/capita
 Moser, Leon & Gwatkin show the same but linked to durable
good ownership (wealth measurement)
 Nutritional Engel Curve
 « Wealthier is healthier » by Pritchett & Summers show
undoubtedly a strong link between income/capita & child mortality
• Also no evidence of good effects of government spending on health
 Vulnerability of poor countries facing many diseases
• As a conclusion, it seems that economic growth heals
deprivation in health but it’s far from being the only factor




Education, especially for women
Institutionnal factor: health services and healthcare
Reduction in child malnutrition
Lowering fertility rates
7
Graphic Analysis: The Preston
Curve, 2000
8
Life expectancy and infant mortality
9
Standard deviations of measures of health
& income, 1960 to 2004
10
Death & Poverty worldwide
11
Proportionnal changes: Infant mortality &
per capita economic growth
12
Proportionnal & Absolute
13
Economic Growth & Life expectancy
14
Income & Infant mortality, India & China
(UN Data)
15
Another view: paper analysis
•
•
•
« The effect of health on economic growth: a production function
approach », by D.Bloom, D.Canning, J.Sevilla, in Harvard School of Public
Health, 2004
Data used: 1960-1990, output in GDP, panel of countries
Use of a P° function model of aggregate economic growth, with 2
microeconomics variables of human capital: work experience and
health
 Labor quality is essentiel to economic growth
•
In the function: output is a function of its inputs and their linked
efficiency (TFP)
 Physical capital, Labor and human capital with 3 dimensions: education,
experience and health
 Growth can be decomposed in 2 sources: level of inputs and growth in TFP
•
Life expectancy effect on growth appears to be a real labor
productivity effect
 1 year improvment in life expectancy contributes to an increase of 4% in output
 Probably effect on life cycle savings & capital accumulation
 Growth rates & effeciency of inputs also influence output level
16
Questions to be asked
• According to you, what would be the best health measurement?
What factors would you chose to put in this measurement?
• Do you think the poor countries would perform any better
regarding the Aid/HIV disease if pharmaceuticals lobbies
allowed treatment generics?
• How income influences health and how health impacts income &
growth?
 Are poor countries poor because of their poor health?
 Historically no significant increase in growth comparing to the huge
improvement in health in wealthier counties
17
Questions?
Thank you.
18