Inequality and Human Development Luis F. LopezCalva - PowerPoint PPT Presentation
1 / 21
Title:
Inequality and Human Development Luis F. LopezCalva
Description:
Inequality and Human Development. Luis F. Lopez-Calva ... The exante emphasis, process-oriented approach, versus the expost nature of utilitarianism ... – PowerPoint PPT presentation
Number of Views:103
Avg rating:3.0/5.0
Title: Inequality and Human Development Luis F. LopezCalva
1
Inequality and Human DevelopmentLuis F.
Lopez-Calva
HDRO/RBA Regional Technical Workshop on Measuring
Human Development Nairobi September - 2007
2
Inequality and Human Development
- One one reads Sens work, based on which the
Human Development perspective has been
established, three things are obvious - The multidimensional nature of welfare and
development indicators - The fundamental importance of inequality
- The exante emphasis, process-oriented approach,
versus the expost nature of utilitarianism
3
Inequality and Human Development
- The HDI has been especially succesful at
- Revealing the multidimesionality story in the
policy sphere - Taking the focus out of simple growth and
trickle-down type of logic - It hasnt been so useful, however, to push the
agenda on inequality, even though it is at the
core of Sens framework on capabilities and
functionings
4
Paradoxes
- Looking at the HDI and the ranking based on it,
something looked interesting from the Mexican
perspective - Mexico has basically the same HDI as Cuba and
Trinidad y Tobago (among others, like Panama)
5
Paradoxes (2)
- Looking at the whole picture, it turns out that
Chile, Costa Rica, and Kuwait are close in
ranking and levels of HDI - Is this intuitive?
6
Across regions
7
Among People or Groups
8
Among Dimensions
- Gap Between Mexico City and Chiapas
- 10.1 in life expectancy
- 23.5 in educaction
- 52 in income
9
The Two Main Questions
- Decisions have been made regarding
- The dimensions that are included (health,
education, income) - The way they are aggregated
- Thus inequality can be accounted for by
- Looking into each dimension and then define
groups or individuals - Aggregating in a different way
10
Easy ways to do it
- Depending on data, estimate the quasi-HDI by
region, or groups (has been done in the Global
HDR and many national HDRs, as shown in the
Primer, Box 2.1 define income groups, look at
levels of mortality, for example) - Change dimensions completely (bad idea!, not an
HDI anymore, Box 2.4)
11
Disaggregation is not simple
- Demanding in terms of data
- There are statistical devices available
- Example, for income, use poverty mapping
- THUS, FIRST APPROACH ESTIMATE INDICATORS AT THE
DISAGGREGATED LEVEL, THEN COMPARE - (That does not incorporate inequality in the HDI,
only looks at disaggregations)
12
Second approach
- Adjust each dimension by inequality among people
- Example, in income, adjust by Gini Coefficient
(Hicks, 1997 not cited in primer!) - Once adjusted, add up in the traditional way
- Does not satisfy important properties, plus you
need information on the distribution by each
characteristic - Yet, it does incorporate inequality in the HDI
itself no inequality across dimensions
13
Third Approach
- Incorporate both inequality within dimensions
among people-- and inequality between dimensions
uneven development, satisfying desirable
properties - Aggregation through generalized means does this
(page 41 in primer, Mexicos case, Foster, et al.
2005 not cited either!)
14
In Mexico
15
Advantages
- It allows you to estimate the loss in HD due to
inequality - Even more, it allows you to decompose such loss
between one part due to inequality among people
or groups and inequality due to uneven development
16
Problems with third approach
- Demanding in terms of data
- Census data ad household surveys needed
- Under this general approach, traditional HDI is a
special case where society does not care about
inequality - Current gender adjusted HDI is also a special
case where there is inequality between groups
men, women but not within groups
17
Conclusions
- There are, summarizing, three approaches to look
at inequality in HD - One does not try to make the HDI sensitive to
inequality, just disaggregates regionally,
between groups, etc. - The other two try to actually adjust the HDI to
make it sensitive to inequality - Generalized means is by far the most solid way to
do it, but demanding in terms of data
18
Conclusions (2)
- All of them, useful and complementary
- It depends on data availability
- But also, what type of issues you want to address
19
- Appendix
20
Aggregation through generalized means
- The class of indices suggested is
- ?q(x) (x1q xnq)/n 1/q
- For all q ? 0 and ?q(x) (x1xn)1/n if q 0
- Generalized means makes the index sensitive to
inequality across dimensions and among
individuals
21
Interpretation
- The value of q is social aversion to inequality
- The generalization makes the traditional HDI just
a special case where q0, i.e., a case where
society does not care about inequality - The proposed index allows us to
- Punish the indicator of development if there is
inequality - Identify the development loss due to uneven
development (inequality across dimensions) and
the loss due to inequality among individuals,
which have very different policy importance
Recommended
CrystalGraphics Presentations
