"Concepts and Operationalisation of Pro-Poor Growth: A Usable PPG Index" Inequality Mario Negre UNU-WIDER, Helsinki 20 Sep 2013
Introduction 1/2 Inequality increasingly recognised as key factor for Poverty Reduction Ravallion: i) g 7x more Pov reducing when Ineq ii) Ineq IMF: Ineq WB, UNICEF, UNDP, Pov reduction rate longer g spells Mounting academic evidence (economic, governance, stability, etc) 2
Introduction 2/2 Growth: a necessary but not sufficient condition for poverty reduction Growth: Shared-g Pro-poor g Inclusive-g Inclusive development (Multidimensional) 3
Defining PPG Approaches: Absolute Relative Poverty reducing growth Disproportionally benefiting the poor Absolute approach reductio ad absurdum: Pro-Poor = Pro-Nonpoor! 4
PPG Operationalisation Shapley Decomposition (1953; Game Theory) Shorrocks (1999) generalisation for Poverty: It calculates the the marginal impact on poverty of eliminating each contributing factor in sequence, and then assigns to each factor the average of its marginal constributions in all possible sequences. Exact Decomposition 5
PPG Indices Kakwani and Pernia s (2000) + per capita g (γ > 0 ) - per capita g (γ < 0) PEGR Limitations: Lack of comparability (+ vs. - g) Φ when: G 0 (γ > 0) and G - R (γ < 0) Other problems of lack of meaning and monotonicity in PEGR 6
A New PPG Index 1/4 Φ = - R general for any poverty measure or income level Based on PPG relative approach PPG should satify these porperties: i) It is + (-) when γ poor > γ > γ nonpoor (γ poor < γ < γ nonpoor ) ii) It is zero when γ poor = γ = γ nonpoor 7
A New PPG Index 2/4 What about using the Headcount Index (H)? If all the poor increase their income without going above the poverty line and the nonpoor experience no change it should be PPG but R H =0 because H didn t change Thus, it doesn t satisfy desirable properties 8
A New PPG Index 3/4 What about using the Poverty Gap (PG)? Not only does it satisfy the properties but it can be mathematically linked to: g poor > g > g nonpoor Substituting in it can be shown that (For the case of constant population and number of poor) 9
A New PPG Index 4/4 R PG < 0 R PG > 0 If G > 0 Pro-Poor Growth Pro-NonPoor Growth Anti-Poor/NonPoor Recession (also cases with positive per capita growth but higher population growth) PPG Rate: γ * = γ νr PG 10
Pro-Poor Performance Φ t = ϕ / t to compare pro-poorness over time Φ that would have beeen necessary to meet a given target PG reduction (halving, eradicating) Φ that would be necessary to achieve a target given a forecast economic growth Poverty Gap Target-based Pro-Poor Performance Trend 11
Example - Honduras ϕ PG > 0 ϕ PG < 0 Pro-Poor Growth Pro-Nonpoor Growth 12
Halving 1992 s PG by 2015 with 2004-7 growth 13
Why using this PPG Index? Because it s based on the relative approach it works it measures: if g is pro-poor how pro-poor it is it s understandable for the policy level: how distributional shifts affect the PG It focuses on what happens to the poor (or those below a chosen income) unlike Gini It s a monitoring and assessment tool across countries and over time It s calculable with current datasets (income) 14
Thank You mario.negre@die-gdi.de http://www.die-gdi.de 15