konjunkturforschungsstelle swiss institute for business cycle research do imf and world bank...

KonjunkturforschungsstelleSwiss Institute for Business Cycle

Research

Do IMF and World Bank Influence Votingin the UN General Assembly?

Axel Dreher and Jan-Egbert Sturm

International Political Economy Society Inaugural Conference November 17-18, 2006

Prof. Dr. Jan-Egbert Sturm / [email protected]


Research

Literature overview

G7 UN Voting

IMF / WB

a

b dc



Research

Hypotheses

1. Cultural and political proximity increases voting coincidence

2. Countries depending on foreign support are more likely to vote in line with G7 countries

3. () Bilateral foreign aid increases the probability that a recipient country votes in line with the donor

4. Trade flows increase/decrease the probability that a country votes in line with its partner country

• () IMF and World Bank loans increase the probability that a recipient country votes in line with the institutions’ major shareholders



Research

Main Hypothesis

() IMF and World Bank loans increase the probability that a recipient country votes in line with the institutions’ major shareholders

IBRD: International Bank for Reconstruction and Development IDA: International Development Association

IMF World Bank

Concessional - net flows

- amount agreed

net flows(IDA)

Non-concessional - net flows

- amount agreed

net flows(IBRD)



Research

Main Hypothesis

() IMF and World Bank loans increase the probability that a recipient country votes in line with the institutions’ major shareholders

World Bank Number of technical loans programs starting Number of adjustment programs starting Number of other programs starting

IMF Start of EFF and SBA program (non-conc.) Start of SAF and PRGF program (conc.)



Research

Data

Years: 1970-2002

Recipient-countries: up to 188

Donor-countries: each individual G7 country & weighted-average G7



Research

Comparing inline voting behavior across donor countries, 1970-2002

Variables corrected for country- and year-specific effects

Each cell based on ca. 5,000 country/year observations

CAN FRA GBR DEU ITA JPN USA

Canada (CAN) 0.95 0.96 0.92 0.99 0.97France (FRA) 0.98 0.91 0.97 0.91UK (GBR) 0.92 0.97 0.92Germany (DEU) 0.93 0.89Italy (ITA) 0.96Japan (JPN)United States (USA)

0.750.710.770.710.740.73

0.79

G7

0.950.960.990.920.970.92



Research

Voting coincidence across different regions

Regional classification based on CIA World Fact Book

CAN DEU FRA GBR ITA JPN USA G7

Rest of the world 0.45 0.42 0.36 0.35 0.44 0.48 0.20 0.36

Europe, Western 0.73 0.71 0.64 0.65 0.73 0.71 0.46 0.65Asia, Central & Western 0.60 0.60 0.55 0.54 0.60 0.62 0.37 0.54Oceania 0.61 0.57 0.52 0.52 0.60 0.62 0.39 0.52Europe, Central & Eastern 0.59 0.59 0.53 0.51 0.59 0.60 0.34 0.52South America 0.57 0.53 0.46 0.46 0.55 0.60 0.30 0.46Central & Middle America 0.55 0.52 0.45 0.45 0.54 0.58 0.31 0.46Caribbean 0.53 0.50 0.45 0.43 0.52 0.56 0.30 0.44North America 0.54 0.50 0.44 0.43 0.52 0.58 0.26 0.43Africa 0.51 0.48 0.43 0.41 0.50 0.55 0.27 0.42Middle East 0.52 0.48 0.42 0.41 0.50 0.55 0.27 0.42Asia, Eastern & Southern 0.52 0.48 0.43 0.41 0.50 0.56 0.25 0.41



Research

Extreme Bounds Analysis

Estimate equation of the following formYit = Mit + Fit + Zit + uit

Re-specify Z-vector holding M and F constant Go through all combinations of the Z vector Examine all estimated coefficients

Levine and Renelt (1992): look at extremes Sala-i-Martin (1997): look at distribution

How to specify baseline model (M-vector)? General-to-specific methodology (Temple, 2000)



Research

Results for baseline model

Voting coincidence increases when cultural and political proximity increases (hypothesis 1) dependence upon foreign support increases (hypothesis 2)

CAN FRA GBR DEU ITA JPN USA G7

p-value Hausman 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

Adj. R20.86 0.89 0.89 0.86 0.87 0.83 0.82 0.89

#Obs 4286 4286 4286 4286 4285 4286 4286 4286#Cnt 177 177 177 177 177 177 177 177Period 73-01 73-01 73-01 73-01 73-01 73-01 73-01 73-01

National capability -7.147 -7.506 -7.610 -6.476 -7.304 -7.064 -7.932 -7.686(-9.98) (-11.29) (-10.62) (-7.34) (-10.28) (-10.26) (-8.90) (-10.70)

Democracy [t-1] 0.013 0.012 0.012 0.013 0.013 0.010 0.010 0.012(16.43) (15.64) (14.95) (12.90) (16.09) (12.68) (9.73) (14.50)



Research

Testing the main hypothesis for the G7

IMF World Bank

conc. flows 0.094 IDA flows (conc.) -0.110(0.92) (-1.39)

conc. flows agreed -0.011(-0.28)

non-conc. flows 0.197 IBRD flows (non-conc.)

0.861(2.38) (4.75)

non-conc. flows agreed 0.208(3.41)

SAF & PRGF (conc.) 0.002(0.59)

EFF & SBA (non-conc.) 0.012(4.75)

techn. projects 0.004(2.04)

adjust. projects 0.005(3.22)

other projects 0.001(2.28)



Research

CDF(0) Results of EBA

CAN FRA GBR DEU ITA JPN USA G7Baseline variables (4,525 regressions per cell) National capability 0.93 0.99 0.99 0.90 0.96 1.00 0.99 0.97Democracy [-1] 1.00 0.98 0.98 1.00 1.00 1.00 0.94 0.96

Main hypothesis: IMF & World Bank support (4,089 regressions per cell)IMF conc. flows 0.91 0.75 0.73 0.82 0.91 0.90 0.73 0.74IMF non-conc. flows 0.81 0.82 0.81 0.80 0.83 0.84 0.80 0.79IMF conc. flows agreed 0.62 0.62 0.67 0.67 0.66 0.69 0.66 0.60IMF non-conc. flows agreed 0.89 0.92 0.91 0.87 0.93 0.81 0.72 0.92IMF SAF & PRGF (conc.) 0.82 0.85 0.86 0.66 0.85 0.83 0.81 0.84IMF EFF & SBA (non-conc.) 0.96 0.93 0.93 0.89 0.97 0.93 0.76 0.93IDA flows (conc.) 0.80 0.78 0.71 0.84 0.77 0.81 0.79 0.74IBRD flows (non-conc.) 0.82 0.90 0.92 0.87 0.86 0.89 0.99 0.93WB techn. projects 0.86 0.97 0.89 0.91 0.94 0.74 0.80 0.85WB adjust. projects 0.99 1.00 1.00 0.98 1.00 0.96 0.89 0.98WB other projects 0.81 0.77 0.77 0.83 0.80 0.90 0.91 0.75



Research

Testing the extended models

CAN FRA GBR DEU ITA JPN USA G7

p-value Hausman testAdj. R2

#Obs#CntPeriod

IMF EFF & SBA

IBRD flows

WB adjust. projects

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.000.82 0.86 0.86 0.86 0.84 0.82 0.79 0.863578 3578 3578 3578 3578 3578 3578 3578154 154 154 154 154 154 154 154

73-00 73-00 73-00 73-00 73-00 73-00 73-00 73-00

0.011 0.008 0.009 0.009 0.010 0.007 0.005 0.008(4.28) (3.67) (3.54) (3.30) (4.17) (2.89) (1.51) (3.29)0.701 0.636 0.717 0.770 0.720 0.620 0.783 0.772(3.94) (3.91) (4.02) (3.81) (4.09) (3.69) (3.49) (4.33)0.004 0.005 0.005 0.004 0.004 0.002 0.003 0.005(2.46) (3.34) (3.10) (1.97) (2.72) (1.17) (1.53) (3.03)

USA

0.000.824249177

73-01

0.851(3.90)



Research

Robust to changes in dependent variable?

How to define voting coincidence? Barro & Lee = BothYes + BothNo + Abs. Thacker: = BothYes + BothNo + ½ Abs. Kegley & Hoock: = BothYes + BothNo

Which votes to count? Also almost unanimous votes? All votes or only Keyvotes? Include “dominant” topics? (Israel: 20% of all votes)

How to weigh the past? No weights, or slowly decaying weights



Research

Conclusions Hypotheses 1 to 4

The probability that a recipient country votes in line with the donor …

1. significantly increases with cultural proximity

2. significantly increases with dependency on foreign support

3. is not significantly related to bilateral foreign aid

4. is somewhat related to bilateral trade flows



Research

Conclusions Main Hypothesis

Concessional loans are not correlated with voting coincidence in the UN General Assembly

Non-concessional IMF and World Bank programs and loans matter for voting in the UN General Assembly number of IMF EFF and SBA programs flow of World Bank IBRD loans

Also the start of World Bank adjustment programs matter

For the US, only IBRD loans matter for the voting behavior of recipient countries

konjunkturforschungsstelle swiss institute for business cycle research do imf and world bank...

Documents

recipient country votes

world bank loans

international bank

partner country imf

net flows ibrd slide

world bank influence

individual g7 country

average g7 slide