Does aid reduce inequality? Evidence for Latin America David - - PowerPoint PPT Presentation

Does aid reduce inequality? Evidence for Latin America

David Castells-Quintana AQR-IREA. Universidad de Barcelona, José María Larrú Universidad CEU San Pablo larram@ceu.es

MOTIVATION & FRAMEWORK

POVERTY AID GROWTH INEQUALITY income

• utcomes /
• pportunities

within between +

• ?

+/-? + ??

OUTLINE

• INEQUALITY IN LATAM: LIT. REVIEW
• AID AS POSSIBLE DETERMINANT OF INEQUALITY

– Lit. review – Theoretical channels

• EMPIRICAL EVIDENCE

– Data and trends (Gini and AID/GDP) – Correlation analysis – Cross-country evidence

• Cross-section and pooled
• Panel data:

– pooled OLS (triennial data) – -System-GMM (triennial & annual data)

• CONCLUSIONS

INEQUALITY in LATAM

Average Gini 1990-2008

58,27 56,36 56,06 55,80 55,66 55,36 55,25 55,24 54,29 53,24 52,60 52,25 50,60 50,55 49,90 48,36 46,71 44,88 43,95 10 20 30 40 50 60 70 B r a z i l B

• l

i v i a G u a t e m a l a P a r a g u a y C

• l
• m

b i a E c u a d

• r

H

• n

d u r a s P a n a m á C h i l e N i c a r a g u a M é x i c

• L

A T A M

• 1

8 P e r u E l S a l v a d

• r

D

• m

. R e p . A r g e n t i n a C

• s

t a R i c a V e n e z u e l a U r u g u a y G ini

Lower Middle Income Countries in yellow

Changes in Inequality: 1990-2008

• 6,63

0,63 0,39

• 5,77
• 1,30
• 6,24
• 6,16
• 2,92

2,69

• 1,66
• 3,24
• 3,33
• 3,19

1,81 2,33 1,11 4,71 6,98

• 8,66
• 8,51
• 7,32
• 6,90
• 6,43
• 6,16
• 4,56
• 4,51
• 4,20
• 4,06
• 3,86
• 3,51
• 2,95
• 2,36
• 2,33
• 1,22

0,00

• 6,24
• 10,00
• 8,00
• 6,00
• 4,00
• 2,00

0,00 2,00 4,00 6,00 8,00 ECU PER ARG ELS VEN PAR BRA PAN BOL MEX NIC GTM CHI HON URU RDOM CRIC COL G in i p ercen tag e p o in ts 2008-1990 2008-max

CAUSES

• Macro-policies

– Trade openess

• Fiscal policy

– Progressive taxation – Condional cash tranfers (focused)

• Lower skilled labour premium

– Increase average years of schooling – Employment for low-skilled workers (maquiladoras) – Higher minimum wages

• External flows

– FDI & remittances

OUR SAMPLE

AID in LATAM

DATA for OUR SAMPLE

• LIT. REVIEW
• POSITIVE:

– Bornschier et al. 1978; Layton & Nielson 2008; BjØrnskov 2010 – Herzer & Nunnenkamp 2012 (cointegration)

• No effect:

– Dolan & Tomlin 1980; Chong et al. 2009 (N=112

T=1972-2001)

• NEGATIVE:

– Cuesta et al. 2006; Tezanos et al. 2013

AID & INEQUALITY: CHANNELS

AID INCOME DISTRBUTION BILATERAL MULTILATERAL MULTI‐BI‐LATERAL

(pooled & common funds)

Central state Non‐central state NGOs SOCIAL ECONOMIC INFRASTRUCTURE PRODUCTION MULTISECTOR

education health

elite middle class poorest

Political reforms (voice, participation, civil rights, security, better democracy and institutions) income poverty

Projects & programmes Budget support Debt relief climate investment micro businesses food and humanitarian aid Fiscal policy (budget constraint)

In‐kind transfers Credits Cash transfers Knowledge & know‐how

water and sanitation

Technichal assistance Macro‐stability

agents instruments sectors Programs & projects Policies and political context

SELECTED CHANNELS

• Focus on the poorest zones, groups, sectors
• Better governance: trade unions -> minimum

wages, collective negotiations…

• Macro-stability: multilateral programmes

– Lower inflation – Better real exchange rate & terms of trade – Attract FDI

• Debt relief frees-up resources
• Direct Budget Support (CCTs programmes)
• Technical cooperation (fiscal reforms)

CORRELATION

By country: 8 cases <0 URU=-5.1 (R2=0.45) Brazil ≈0 9 cases >0 ECU=+6.4 (R2=0.46)

OUR MODEL

• CONTROLS:

– Chong et al. (2009):

• inflation; liquid liabilities; literacy; GDPpc; agriculture and

industry (%V.A.)

– Domestic redistributive policies:

• Public expenditure (and social exp.)

– Labour institutions and education

• Minimum wages (formal); unemployment; Gini-education

– External redistributive flows:

• Terms of trade; FDI; remittances

– Political context:

• polity 2

CORRELATION MATRIX

RESULTS: cross-section

Results: pooled cross-section

Results: SYS-GMM

ROBUSTNESS

• Gini:

– 2008 vs average – SEDLAC: Gini, Theil, Atkinson, E(0) but 230

• bservations vs our 323 (correlation=0.992)
• Aid

– Per capita vs ODA/GDP(%) – Gross disbursements (with & without debt relief)

CONCLUSION

• Consider aid as an egalitarian external flow
• Multiple channels:

types, donors, instruments, sectors.

• Complex transmissions
• Caveats:

– Measurement erros (Gini annual) – Overall effect vs. Country specific effect