Growth in the Dominican Republic and Haiti: Why has the Grass been Greener on One Side of

Hispaniola?

Laura Jaramillo and Cemile SancakWHD Workshop

November 17, 2006

The views expressed in this presentation are those of the authors and should not be attributed to the International Monetary Fund, its Executive Board, or its management.

MotivationHaiti and the DR: broadly similar in terms of geography and historical institutions, yet striking divergence in terms of growth performance.

GDP per capita (In constant 2000 U.S. dollars)

Dominican Republic

Haiti

Latin America and

Caribbean

0

500

1000

1500

2000

2500

3000

3500

4000

4500

1960 1966 1972 1978 1984 1990 1996 2002

Average Annual GDP Growth,1960-2004 (In percent)

0 1 2 3 4 5 6

HaitiJamaicaUruguay

ArgentinaNicaraguaDominica

BoliviaVenezuela

PeruTrinidad

El SalvadorLAC

HondurasGuatemalaColombia

ChileParaguayEcuadorMexico

Costa RicaPanama

BrazilDominican Rep.

OverviewCase study approach, using Latin America as a reference point Initial conditions

GeographyHistorical institutions

Analysis of policies since 1960, drawing on:Panel regressions in the literatureExamination of policies pursued in each country

Alternative variables that help explain growth divergence

Political instabilityMacroeconomic instability

Panel regression testing the proposed variables

Initial Conditions: GeographyGallup, Sachs and Mellinger (1998) and Diamond (2005) cannot explain the diverging growth rates in Haiti and DR.

Population density, 2000(People per square kilometer)

0 100 200 300 400 500 600

Central African RepublicNiger

SomaliaZambia

Congo, Dem. Rep.Madagascar

LiberiaUnited Arab Emirates

NicaraguaGuinea-Bissau

BotswanaIreland

MalaysiaThailand

ChinaJapan

Korea, Rep.Mauritius

MaltaSingapore

Latin AmericaHaiti

Dominican Republic

Countries with lowest GDP per capita growth rate between 1960 and 2005

Countries with highest GDP per capita growth rate between 1960 and 2005

Source: WDI and authors' calculations

Haiti and DR are under the same conditions in terms of location and climatic conditions.

Haiti, being half the size of DR, has historically had higher population density.

Both countries have much higher population density than the region.

Initial Conditions: Historical institutionsHaiti and DR had comparable historical institutions.

Colonial origin story (Acemoglu, Johnson, Robinson, 2005, 2001)Ethnolinguistic fragmentation, legal origin, religion (La Porta et al., 1998) No statistical differences between Spanish and French colonial rule. (Acemoglu, Johnson, Robinson, 2001; Sirimaneetham, 2006; Treisman, 2000).Common institutions at different points in time: Haiti ruled the DR for 22 years during the 19th century; U.S. military occupation of both countries in the 20th century.

Prior to 1960, political instability was high for both countries, in particular for Dominican Republic.

Average Settler Mortality Rate(Deaths per thousand mean strength)

0 50 100 150 200 250 300 350

Asia

South America

Caribbean

Central America

Northern Africa

Haiti

Dominican Republic

Subsaharan Africa

Index of Ethnolinguistic Fragmentation

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Dominican Rep.

Haiti

Industrial

Eastern Europe

Latin America

Middle East

Asia

East Asia

Africa

Framework: Empirical Endogenous Growth Model

Income

Factor endowments ProductivityEndogenous

Stabilizationpolicies

Geography

Partlyendogenous

External conditions

Exogenous

Structural policies Institutions

Income

Factor endowments ProductivityEndogenous

Stabilizationpolicies

Geography

Partlyendogenous

External conditions

Exogenous

Structural policies Institutions

Panel regression (Loayza, Fajnzylber, and Calderón, 2005)

Variation of the standard growth regression

tiittiT

titiCtititi Xyyyyy ,,'

1,1,1,1,, )( εηµβαα ++++−+=− −−−− where y : log of output per capita, X : a set of variables postulated as growth determinants,

Ty : trend component of output per capita, T

titi yy 1,1, −− − : the output gap at the start of the period,

tµ : a period-specific effect, iη : unobserved country-specific factors, and ε : the regression residual.

Explanatory variables in LFCTransitional convergence (initial GDP per capita) Cyclical reversion (initial output gap)Structural policies and institutions

Education (secondary school enrollment)Financial depth (private domestic credit/GDP)Government burden (government consumption/GDP)Public infrastructure (main phone lines per capita)Governance (ICRG index)Trade openness (trade/GDP)

Stabilization policies Lack of price stability (inflation rate)Cyclical volatility (standard deviation of output gap)Real exchange rate overvaluation (index of real exchange rate overvaluation)

External conditionsTerms-of-trade shocks (growth rate of terms of trade)Period shifts (dummy variables)

DR and Haiti compared to other LA countries Fit of the LFC model

The model does well overall in explaining the direction of changes in growth rates in Latin America. However, the model does not perform as well in explaining the magnitude of change, especially for Haiti.

Actual vs. Projected Changes in Growth Rates1980s vs. 1970s

-6.00

-5.00

-4.00

-3.00

-2.00

-1.00

0.00

Latin America DominicanRepublic

Haiti

Actual

1990s vs. 1980s

-1.00

-0.50

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

Latin America DominicanRepublic

Haiti

Projected

Improving the fit of the LFC model

Can we explain the higher growth rates in the DR and the consistently lower growth rates in Haiti compared to Latin America?The LFC model provides a good fit in terms of structural policies. We propose measures that could enhance the fit of the model:

Political stabilityStabilization policies

Structural measuresStructural policy variables provide a good indication of performance of Haiti and the DR compared to Latin America.

DR has outperformed Haiti and Latin America since the 1970s.Despite improvements in some periods, structural policies have been weaker in Haiti.

Structural measures: Combined indicator of trade openness, credit to the banking sector, government consumption, and secondary education

0.00

0.02

0.04

0.06

0.08

0.10

1960s 1970s 1980s 1990s

Dominican Republic

HaitiLatin America

Structural measures: TradeOn trade, both Haiti and the DR have liberalized imports but DR has done a better job of boosting exports.

Trade openness(As a percent of GDP)

0

5

10

15

20

25

30

35

40

45

50

1960 1965 1970 1975 1980 1985 1990 1995 2000

DR

HaitiLatin America

DR implemented policies to promote exports of goods through free-trade zones and of services through FDIpolicy.

Despite some improvement in the late 1970s, Haiti was severely affected by the trade embargo in the 1990s.

Political shocksProblem with statistical significance of governance variable could be explained by:

A short time seriesSubjective indicator

An alternative measure of institutional quality is political instability.

Barro (1991) finds a negative relationship between political instability and investment and growth, due to the adverse effects of political instability on property rights. Corbo and Rojas (1993) also find that political instability has a highly negative effect on investment rates and growth by creating an environment of high uncertainty.

Political ShocksUsing the episodes of regime change as a proxy for political instability shows that Haiti has been the most unstable and DR among the most stable in the region.

Magnitude of regime changes 1970-2003

0 10 20 30 40 50 60

Costa RicaVenezuelaColombiaHonduras

Dominican RepMexico

BrazilParaguay

El SalvadorPanama

GuatemalaBolivia

NicaraguaChile

EcuadorUruguay

PeruArgentina

HaitiGovernance Indicator based on ICRG(+ indicates better governance)

Dominican Republic

Haiti

0

2

4

6

8

10

12

14

16

1984 1986 1988 1990 1992 1994 1996 1998 2000

Macroeconomic stabilityDrawbacks of stabilization measures used by LFC:

Difficulties in measuring potential output. Difficulties in determining the equilibrium exchange rate.Inflation provides only partial information about economic policies.

A composite indicator would be more appropriate as it provides a better overall picture of stabilization policies. Index constructed as the weighted sum of:

InflationFiscal deficitExchange rate volatilityInternational reserve losses

This index has not been used elsewhere, but the literature indicates the need to look at various factors to determine a country’s policy stance.

Fischer (1993) identifies the inflation rate, budget balance, and the black market exchange premium as basic indicators of macroeconomic policy. Sahay et al. (2006) look at the relationship of growth with volatility of inflation, the exchange rate, and fiscal balances.

Macroeconomic stabilityGiven what we know about the relative performance of DR and Haiti, the composite index provides a good description of macroeconomicpolicy developments.

Macroeconomic instability index(+ indicates more instability)

0.0

0.5

1.0

1.5

2.0

2.5

1960s 1970s 1980s 1990s

Latin America

Haiti

Dominican Republic

LFC stabilization policies(+ indicates poor policies)

10.5

11.0

11.5

12.0

12.5

13.0

1960s 1970s 1980s 1990s

Latin America

Dominican Republic

Haiti

Panel regression with alternative variables Fit of the JS model

The model improves the fit for both the Dominican Republic and Haiti, as well as for Latin America.

Dominican Republic: Actual and projected changes in growth rates between decades

-5

-4

-3

-2

-1

0

1

2

3

4

5

1970s 1980s 1990s

Actual

Proj (LFC)

Proj (JS)

Haiti: Actual and projected changes in growth rates between decades

-7

-5

-3

-1

1

3

5

1970s 1980s 1990s

Actual

Proj (LFC)

Proj (JS)

Dominican Republic. Changes in growth between decades

The main contributors to improvements in growth rates were:Structural policies in the 1970sExternal conditions, transitional convergence, and structural policies in the 1980sStructural policies in the 1990s

Dominican Republic: Changes in growth rates between decades

-8

-6

-4

-2

0

2

4

6

1970s 1980s 1990s

Actual change Projected change Transitional convergence Cyclical reversion

Structural measures Stabilization policies External conditions

Haiti. Changes in growth rates between decadesThe structural measures and stabilization policies have not had a substantial contribution to growth, except in the 1970s.

Haiti: Changes in growth rates between decades

-8

-6

-4

-2

0

2

4

6

1970s 1980s 1990s

Actual change Projected change Transitional convergence Cyclical reversion

Structural measures Stabilization policies External conditions

Conclusions

The growth divergence between the Dominican Republic and Haiti is not explained by differences in the initial conditions but rather by different policy decisions.

The DR has consistently outperformed Haiti and the region in terms of structural measures and stabilization policies.

Haiti has been subject to numerous political shocks that have severely affected its growth performance.