Pro growth, Pro poor: Is there a trade off? J. Humberto Lopez The World Bank

Size: px

Start display at page:

Download "Pro growth, Pro poor: Is there a trade off? J. Humberto Lopez The World Bank"

Chrystal Barton
5 years ago
Views:

1 Pro growth, Pro poor: Is there a trade off? J. Humberto Lopez The World Bank

2 Motivation! PRSP/MDG focus on poverty reduction as main development objective:! Challenges for policy makers and operational staff on the impact of policy reforms on poverty.! Need to better understand the distributional implications of policy reforms:! Efforts to mainstream the Poverty and Social Impact Analysis (PSIA) of policy reforms.! Debate on the impact of structural adjustment on poverty:! Apparent huge variance in the perceptions of how a series of policies affect poverty.

3 Objectives/features Objectives! Analyze the joint impact that a set of policies have on inequaly and growth as a vehicle to infer their poverty impact:! Some policies can be good for growth but negatively affect inequaly.! How do we assess those policies when poverty reduction is our main objective?! Explore the possibily of different policy impacts in the short and long run:! A policy can be pro poor in the long run, but negatively affect poverty in the short run.! Policies can generate poverty dynamics. Features! It uses averages of non-overlapping overlapping five year periods:! Most studies on inequaly use unequal spaced spells.! It allows for fixed effects in dynamic panel framework:! Likely to be important due to religious and cultural consideration. on.

4 Two considerations! Is growth all that matters?! Should a pro poor strategy imply an improvement in income distribution?

5 Cross country evidence on pro-poor poor growth Negative Growth Inequaly Rises Posive Growth/Inequaly Rises Anti-Poor Recession Yrs g g20 Broadly Shared Growth Yrs g g20 Not Pro-poor By Any Definion Yrs g g20 Poland Korea, Rep Costa Rica Iran, Islamic Rep Taiwan, China Tanzania Slovak Republic Hong Kong China Bulgaria Niger Singapore Panama Sierra Leone China Nigeria Zambia Malaysia Dominican Republic Estonia Thailand El Salvador Latvia Maurius Senegal Russian Federation Brazil Etiopía Colombia Mexico Ecuador Philippines Chile Peru Negative Growth/Inequaly Falls Posive Growth/Inequaly Falls Pro-Poor Recession Yrs g g20 Pro-Poor Biased Growth Yrs g g20 Yrs g g20 Guyana Gabon Trinidad & Tobago Jordan Indonesia India Belarus Tunisia Bangladesh Madagascar Egypt, Arab Rep Nepal Ghana Jamaica Sri Lanka Honduras Hungary Bolivia Turkey Venezuela, RB Pakistan

6 Pro poor growth and inequaly: Make your choice!! Yp: : Growth in incomes of the poor! Y: Average growth Yp Y Case 1 2% 1% Case 2 4% 6%

7 What do we know? (i) Growth is good for poverty reduction. Growth Poverty Outcome

8 What do we know? (ii) Progressive distributional change is good for poverty reduction. Growth Poverty Outcome Income Distribution

9 What do we know? (iii) the relationship between growth and inequaly is weak, and if exists likely to go from inequaly to growth. Growth Poverty Outcome Income Distribution

10 This paper Growth Policy Poverty Outcome Income Distribution

11 This paper Growth Policy Poverty Outcome Income Distribution

12 Our main interest is in cases as: Growth Policy Poverty Outcome - Income Distribution

13 Poverty, growth and inequaly P P(Y,L(p)) L*(p) L(p) ψ(l(p) p) dp dx X P Y X X Y P Y Y P G X X G P G G P

14 Poverty, growth and inequaly! Changes in poverty would then be given by:! How policy X affects growth! How growth is translated into poverty reduction! How policy X affects income distribution! How changes in inequaly are translated into poverty reduction! If income can be approximated by a log normal distribution: P Y P G Y P G P f ( Y, G,) g( Y, G)

15 Theoretical elasticies under log-normaly Growth Elasticy PL b / Gini Inequaly Elasticy PL b / Gini

16 Impact of policies on growth and inequaly (1) y g y x g x y g 1 1 ω' β' y lim k x g lim k X k k δy αg ω' δ β' α 1 1 ω' x β' x ν µ i i τ η t t υ ε

17 Dynamics z(t)(1a)z(t-1)bx(t) Poverty Income (a-.1, b.3) Inequaly (a-.3, b.6)

18 Impact of policies on growth and inequaly (2) y g y x g x lim k lim k y g 1 1 β' y x g X (ω k k δy αg ξ ( β' α 1 1 β' ω' δ ) ξg β' ξβ' δα x ) ω' µ x i η t ν i τ ε t υ

19 Some inequaly Results Dependent Variable log(gini) Estimation Method OLS Fixed Effects 2SLS b A&B c Lagged Inequaly (-6.44) (-3.04) (-3.43) (-11.46) Summary Statistics N F (p val) Sargan (p val) 0.12 Second order correlation (p val) 0.53 a All regressions include time dummies. t-statistics in parentheses.

20 Some growth results Dependent Variable log(y) Estimation Method OLS Fixed Effects 2SLS b A&B c Inial Income (1.09) (-5.38) (-1.64) (-3.20) Inequaly (-1.41) (1.03) (-1.48) (.18) Summary Statistics N F (p val) Sargan (p val) 0.03 Second order correlation (p val) 0.25

21 More growth results Dependent Variable log(y) Sample Low incomemiddle Income High Income Inial income (-6.08) (-0.81) (-2.76) Inequaly (.80) (.68) (-.27) Summary statstics Number of countries Number of observations Sargan (p val) Second order correlation (p val)

22 Variables and expected impact Variable Growth Inequaly! Education -! Infrastructure -! Financial development! Government size - -! Trade?! Governance -! Inflation -! External imbalances -! Cyclical Volatily -! Banking crisis - -

23 Growth and Inequaly Results Variable Growth Inequaly! Education -! Infrastructure -! Financial development! Government size - -! Trade! Governance??! Inflation -! External imbalances -! Cyclical Volatily -! Banking crisis - -

24 Growth Growth log(gini) log(gini) log(gini) lagged Inequaly (-13.32) (-13.99) (-11.98) Inequaly (1.66) Inial GDP per capa (-3.80) (-13.44) (-2.24) (-2.42) Inial output gap (-8.52) (5.84) (0.77) Education (6.70) (3.99) (-2.77) (-1.60) (-1.72) Financial depth (4.28) (4.04) (2.83) (2.99) (2.91) Trade openness (3.14) (0.69) (3.04) (2.70) (2.69) Government burden (-3.18) (-4.11) (-2.71) (-2.17) (-2.17) Public Infrastructure (2.71) (0.27) (-3.32) (-1.72) (-1.59) Governance (-0.68) (4.63) (1.74) (1.89) (2.02) Price Stabily (-1.89) (-3.08) (2.16) (2.97) (2.87) Cyclcical Volatily (-3.76) (-4.30) (1.41) (2.14) (2.22) External imbalances (-3.90) (-3.36) (-0.32) (-0.47) (-0.52) Banking crisis (-7.42) (-4.11) (-4.02) (-4.17) (-4.23) External condions (4.98) (6.32) (1.87) (1.30) (1.05)

25 Main empirical results! At the general level:! Growth does not seem to affect inequaly! Inequaly does not seems to affect growth! Improvements in education, infrastructure and lower inflation rates would lead to both growth and progressive change > to poverty reduction.! Improvements in government size (reductions), financial development and trade would lead to growth but negatively affect income distribution > possible trade offs between growth and poverty:! in the short run the distributional impact dominates and poverty might increase.! In the long run the growth effects dominates and poverty would decrease.

26 Short-run run Elasticies Education Finan. Development PL a /Gini PL a /Gini Trade Openness Government Burden PL a /Gini PL a /Gini Infrastructure Inflation PL a /Gini PL a /Gini

27 Long-run Elasticies Education Finan. Development PL a /Gini PL a /Gini Trade Openness Government Burden PL a /Gini PL a /Gini Infrastructure Inflation PL a /Gini PL a /Gini

28 Conclusions! We find evidence of inequaly convergence, and that the speed of this convergence is faster than the one for per capa income (potential dynamics in poverty).! We do not find evidence that inequaly as such does affect growth.! Pro growth strategies are likely to be pro poor.! However, many policies can possibly have significantly different short-run run and long-run poverty impacts.! Need to assess the possible polical economy implications of negative short-run run effects.! Need to expand the use of poverty and social impact analysis techniques.! Need to consider the design of compensatory mechanisms that address these issues.

29 Thank you!

30 Trade etc Inequaly All Li Mi Hi Growth All Li Mi Hi

Appendix. Table S1: Construct Validity Tests for StateHist

Appendix. Table S1: Construct Validity Tests for StateHist Appendix Table S1: Construct Validity Tests for StateHist (5) (6) Roads Water Hospitals Doctors Mort5 LifeExp GDP/cap 60 4.24 6.72** 0.53* 0.67** 24.37** 6.97** (2.73) (1.59) (0.22) (0.09) (4.72) (0.85)