Transcript Maurice Kugler HDRO - Human Development Report

Maurice Kugler
HDRO
2nd Conference on Measuring Human Progress
4-5 March 2013, New York
2010 ADJUSTMENTS OF THE HDI
• The same dimensions but
– New indicators:
• Expected years of schooling, Mean years of schooling
• Gross national income
– New functional form
• Geometric mean
• Normalization based on data
• Calculated for 187 countries in 2011 (compared
to 177 in 2009)
2
GNI
Definition (2011 HDR): “GNI per capita based on constant 2005
international $”
• From World Bank:
-GNI in current international $
-GDP in current international $
-GDP in constant 2005 international $
NB: GNI in constant international $ only available for a subset of
countries
• HDRO:
-Estimates GNI in constant 2005 international $ by using GDP
real/nominal factors
-For 2005, GNI in constant 2005 international $ is the same as
GNI in current international $
3
• From IMF WEO:
-Projection of GDP in current international $ for 2011,
2012,…
-Growth rates of real GDP
• HDRO:
Estimates GNI in constant 2005 international $ for 2011 and
2012 by applying growth rates derived from GDP in current
international $
4
EYS
UNESCO’s Institute for Statistics – Expected Years of Schooling
Definition: “The total number of years of schooling a child of
certain age can expect to receive in the future, assuming that
the probability of his/her being enrolled in school at any
particular age is equal to the current enrolment ratio for that
age”
2011 HDR:
# 16 countries had EYS>16
# 3 countries had EYS >18
# 1 country had EYS >20
Reasons
Capping at 18
5
FUNCTIONAL FORM PROS AND CONS
o Geometric mean advantages:
• Imperfect substitutability
• Rewards balance
o Critiques:
• Lack of subgroup decomposability
• “Troubling” trade-offs:
- HDI is a capability index, not a utility function.
- Asymmetric treatment income (logged):
means, not end.
- HDI as an ordinal index, not cardinal.
6
NORMALIZATIONS
o Natural zeroes:
• Are $100 per year for per capita GNI, 20 years for life expectancy
and 0 mean years of schooling sensible lower bounds?
• Engage in normative discussion to diminish ad-hocness!
• Upper bounds:
- Constantly shifting normative structure based purely on data changes
is misleading
- Aspirational levels ought to be discerned normatively from first
principles if possible
• Revision frequency:
- As data fluctuate and upper bounds are updated annually, there are
unsubstantiated ranking variations from year to year.
- Infrequent revisions, 5-10 years apart, aid consistency.
7