Journal Issue
Share
Article

Why Has the Grass Been Greener on One Side of Hispaniola? A Comparative Growth Analysis of the Dominican Republic and Haiti

Author(s):
International Monetary Fund. Research Dept.
Published Date:
June 2009
Share
  • ShareShare
Show Summary Details

The Dominican Republic and Haiti present a quasi-natural experiment; the two countries share the island of Hispaniola and are broadly similar in terms of geography and historical institutions, yet their growth performance has diverged remarkably since 1960, when the two countries had the same per capita real GDP, just below $800. However, by 2005, the Dominican Republic’s per capita real GDP had tripled to about $2,500, whereas that of Haiti had halved to $430 (Figure 1). Accordingly, the Dominican Republic and Haiti have been at opposite ends of the spectrum within Latin America and the Caribbean in terms of growth rates over the past 45 years, with the Dominican Republic achieving one of the highest average real GDP growth rates at above 5 percent and Haiti, the lowest at about 1 percent (Figure 1).

Figure 1.GDP Per Capita, and Real GDP Growth Rates in Latin America, 1960–2005

Source: World Bank, World Development Indicators.

What explains this divergence in per capita real GDP of the two countries? This paper seeks to answer this question by examining two main issues: (1) to what extent the divergence is the inevitable result of disparities in initial conditions, and (2) to what extent it is the result of differences in the policies pursued in each country since 1960. Drawing on the growth literature, the paper addresses these issues through a combined approach that includes a panel regression to study growth determinants across a broad group of countries, and a case study framework to better understand the specific policy decisions and external conditions that have shaped economic outcomes in the Dominican Republic and Haiti. To examine policy decisions, this paper uses growth determinants from the literature and introduces alternative variables of institutional quality and stabilization policies to help better explain the income divergence between the two countries. Furthermore, to facilitate comparisons, Latin America is used as a reference point throughout the paper.

When examining initial conditions, namely geography and historical institutions, we find great similarities between the Dominican Republic and Haiti, implying that initial conditions cannot explain their divergence in per capita real incomes. Moreover, based on the panel regression and case study, we find that policy decisions since 1960 have played a central role. In particular, the Dominican Republic has consistently outperformed Haiti and the rest of Latin America in terms of structural measures and stabilization policies, whereas Haiti has been subject to numerous political shocks that have severely affected its growth performance.

I. Literature Review

Only a few studies have compared the growth performance of the Dominican Republic and Haiti, and these studies have provided mostly qualitative discussions. Among the well-known ones is the chapter in Jared Diamond’s book Collapse: How Societies Choose to Fail or Succeed (2005). Although Diamond focuses on environmental policies, it can be inferred from his arguments that higher population density and lower rainfall have been the main factors behind the more rapid deforestation and loss of soil fertility on the Haitian side of Hispaniola, with adverse consequences for agricultural production and therefore growth performance. Similarly, Lundahl (2001) argues that Haiti is the poorest country in the western hemisphere because of the interplay between population growth and the destruction of \arable land. He explains that the increase in the rural labor force has led to an expansion of subsistence food crops to the detriment of export crops, in the context of decreasing international food commodity prices.

Other studies have found that economic performance in the Dominican Republic has been favored by political and macroeconomic stability. Bulmer-Thomas (2001) finds that, for the Caribbean in general, improvements in per capita GDP are linked to higher exports per capita, the quality of institutions, and stability of the macroeconomic framework. The World Bank (2006) also argues that the Dominican Republic experienced a more enabling environment for private investment than Haiti due to political stability and stable macroeconomic conditions over prolonged periods that allowed it to follow a more diversified and outward-oriented growth strategy. In addition, IMF (2001) argues that growth in the Dominican Republic during the 1990s was anchored by capital formation and strong productivity growth, whereas trade liberalization encouraged private investment and output growth.

This paper contributes to the study of growth performance of the Dominican Republic and Haiti by providing a comparative analysis of the two countries. In contrast to previous work, this paper relies on a combination of approaches, specifically growth accounting, panel regressions, and case studies. These approaches are complementary because growth accounting provides a broad overview of economic performance and resource endowments in the Dominican Republic and Haiti, the panel regressions provide a benchmark to compare the performance of the two countries with an “average economy” facing similar shocks, and the case study allows us to capture the heterogeneity in the conditions and processes that have governed the growth experience in each country.

II. Initial Conditions

We begin by examining two initial conditions that growth literature highlights as the most likely to influence long-term growth performance: geography and historical institutions. Although the absence of national accounts prior to 1960 does not allow us to determine exactly at what point in time the Dominican Republic started growing faster than Haiti, this analysis helps us determine if the divergence started long before 1960.

Geography

Geography plays a direct role in shaping a country’s growth performance. It determines the quality of natural resources, the productivity of land, the public health environment, and the extent to which a country can become integrated with world markets. When we compare the geographic characteristics of Haiti and the Dominican Republic based on measures widely used in the literature, we find no substantial differences between the two countries on these grounds, from which we can infer that geography cannot explain the growth divergence between the two countries.

Gallup, Sachs, and Mellinger (1998) assess the impact of geography on economic development and find that growth is favored in temperate regions, in coastal regions, and also in regions with high population density and good access to trade. However, these factors do not explain the divergence between Haiti and the Dominican Republic as they have the same location, equal ocean access, and similar climate. In addition, Haiti has historically had twice the population density of the Dominican Republic.1

As explained in the literature review, Diamond (2005) also addresses the issue of geography and argues that rapid deforestation, caused by lower rainfall and higher population density, has led to lower growth in Haiti compared with the Dominican Republic. However, a 1941 study on rainfall in Hispaniola did not find evidence that Haiti had lower rainfall than the Dominican Republic. On the basis of data for an average of 11 years, the study reveals that rainfall was comparable in the two countries (see Alpert, 1941). Likewise, had lower rainfall been an issue for Haiti, it would not have been one of the richest colonies in the French empire during the 18th century, when it produced about 40 percent of all the sugar and 60 percent of all the coffee consumed in Europe. In fact, deforestation on the Haitian side of Hispaniola can be considered a more recent phenomenon: even as late as 1960, the amount of arable land in both countries was comparable at about 20 hectares per person. Similarly, Diamond’s argument about the impact of high population density on land use cannot convincingly explain lower growth rates in Haiti compared with the Dominican Republic. Several studies have found evidence about the potential benefits of higher population density, including Gallup, Sachs, and Mellinger (1998) and Klasen and Nestmann (2004), who argue that population density generates the linkages, infrastructure, demand, and effective market size for technological innovations that fuel growth. Figure 2 illustrates the relationship between population density and growth. Between 1960 and 2005, many of the countries with the highest per capita real GDP growth had high population densities—in some cases even higher than that in Haiti—whereas countries with the lowest per capita real GDP growth had low population densities.2

Figure 2.Population Density, 2000

(People per square kilometer)

Source: World Bank, World Development Indicators.

Historical Institutions

A growing body of literature argues that institutions are important for initiating and sustaining economic growth. Although institutions are clearly endogenous and evolve with economic performance, in this subsection we focus on the influence of historical institutions to understand if economic performance since the 1960s has been driven mainly by colonial legacies or rather by more recent policy developments. We find that the historical institutions of the Dominican Republic and Haiti were very similar leading into the 20th century, implying that this cannot fully explain the growth divergence.3

Acemoglu, Johnson, and Robinson (2001) argue in a seminal paper that colonial origin matters for growth. They find that Europeans were more likely to set up extractive institutions in places where they faced high mortality rates and could not settle, resulting in poor institutions that have persisted to the present. However, the Dominican Republic and Haiti had the same settler mortality rates as estimated by these authors. Therefore, it would be expected that both countries had equally extractive institutions (Figure 3). In terms of the impact of the colonial power, the literature does not provide evidence of significant differences between Spanish colonial rule in the Dominican Republic and French colonial rule in Haiti. Several studies have used dummy variables for French, British, and Spanish colonies to try to explain growth, corruption, and policy volatility, but have found only the British variable to be significant.4

Figure 3.Settler Mortality and Ethnolinguistic Fragmentation

Furthermore, Haiti and the Dominican Republic share common institutional features, such as those examined by La Porta and others (1998), who find that countries that are ethnolinguistically heterogeneous, use French or socialist laws, or have high proportions of Catholics or Muslims exhibit inferior government performance. The Dominican Republic and Haiti both have low ethnolinguistic fragmentation, both use French law, and both have mainly Catholic populations (Figure 3).

In fact, the quality of institutions was poor in both countries until early in the 20th century—at the time of the U.S. military occupation—with arguably greater political instability in the Dominican Republic. Between independence in 1804 and the U.S. military occupation in 1915, Haiti had 33 heads of state, with an average time in power of 3.4 years. Meanwhile, between independence in 1844 and the U.S. military occupation in 1916, the Dominican Republic had 61 heads of state, with an average time in power of only 1.2 years. Moreover, although it served U.S. interests, the U.S. military occupation of the island was linked to internal struggles and violence in both countries.5 Prior to the U.S. occupation in 1916, the assassination in 1911 of President Cáceres in the Dominican Republic led to various revolutions, economic chaos, and a near-collapse of government institutions. Similarly, before the U.S. intervention in 1915, six different Haitian presidents had been killed or forced into exile since 1911, with revolutions leading to economic disorder and growing external indebtedness.

In summary, strong similarities in the initial conditions in the Dominican Republic and Haiti indicate that these cannot explain the divergence in real incomes of the two countries since 1960. However, this finding does not imply that all conditions were identical going into the 1960s. In fact, by 1960, Haiti was already trailing the Dominican Republic in some social indicators, such as life expectancy and illiteracy rates.6 Nonetheless, we find that these differences are likely to have emerged closer to 1960 and were not inherited from the 19th or early 20th centuries. As mentioned earlier, lack of national accounts prior to 1960 does not allow us to determine the exact point in time when the Dominican Republic started to outpace Haiti, but differences are likely to have started to emerge in part as a result of the policies implemented following the U.S. military occupation. In general terms, the outcomes of the U.S. intervention in both countries were akin: order was broadly restored; the countries’ budgets were balanced and debts reduced; and infrastructure was expanded, including new roads, telephone connections, port facilities, and public health and education services. However, ensuing governments in Haiti practiced only rent-seeking behavior without efforts to maintain public infrastructure and social services, whereas the Trujillo regime in the Dominican Republic promoted agriculture, industry, and public works.

III. Analysis of Policies Pursued

In light of the similarity of initial conditions in Haiti and the Dominican Republic, we continue with an analysis of the policies pursued in each country since 1960, when national accounts became available, to shed more light on the factors contributing to the divergence in per capita real GDP. The analysis is based on a combination of growth accounting, panel regressions, and case studies.

Growth Accounting

A standard growth accounting exercise shows that total factor productivity (TFP) has been an important factor for both the Dominican Republic and Haiti (Figure 4).7 In broad terms, the Dominican Republic had favorable growth trends between 1960 and 2000, largely fueled by productivity gains and capital accumulation. In contrast, economic performance in Haiti has been dismal, with negative TFP in all four decades. Real growth in the 1970s, the only period in which Haiti had positive per capita GDP growth performance, was achieved through strong investment efforts. The empirical endogenous growth model explained below will provide greater insight into the factors that underlie both TFP and capital accumulation.

Figure 4.Contribution to GDP per Capita Growth

(Annual average over 10-year periods, in percent)

Note: Growth accounting uses data from IMF, International Financial Statistics, and World Bank, World Development Indicators.

Empirical Endogenous Growth Model

The empirical endogenous growth model identifies partly endogenous variables (structural policies, stabilization policies, and institutions) and exogenous variables (external conditions) (Figure 5). Our choice of growth determinants was guided by two criteria: that these variables are widely used in the growth literature and that they capture the diverse aspects of policies and shocks in Haiti and the Dominican Republic.

On the basis of this framework, a dynamic panel model of per capita real GDP growth is estimated to assess the relative importance of these growth determinants. The reduced-form equation builds on the work by Loayza, Fajnzylber, and Calderón (2005), modifying their specification to better explain developments in the Dominican Republic and Haiti, as will be described in detail below.

Figure 5.Endogenous Growth Model

The Loayza, Fajnzylber, and Calderón (2005) Model

Loayza, Fajnzylber, and Calderón (2005) (hereafter LFC) use the following variation of the standard growth regression:

where y is log of output per capita, yT represents the trend component of output per capita, yi,t1yi,t1T is the output gap at the start of the period, X is a set of additional variables postulated as growth determinants, μt is a period-specific effect, ηi represents unobserved country-specific factors, and ∊ is the regression residual. The inclusion of the initial output gap as an explanatory variable controls for cyclical output movements to differentiate between transitional convergence (initial GDP per capita, yt–1)—based on the conditional convergence hypothesis—and cyclical reversion to the long-run trend (yt1T).Table 1 describes the growth determinants included in the LFC model, which can be grouped as in the proposed empirical endogenous growth model.8x

Table 1.Variables Used in Loayza, Fajnzylber, and Calderón (2005) Model
Transitional convergence (initial GDP per capita)
Cyclical reversion (initial output gap)
Structural policies
Education (secondary school enrollment)
Financial depth (private domestic credit/GDP)
Government burden (government consumption/GDP)
Public infrastructure (main phone lines per capita)
Trade openness (structure-adjusted 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)
Systemic banking crises (frequency of years under crisis)
Institutions
Governance indicator (International Country Risk Guide)
External conditions
Terms of trade shocks (growth rate of terms of trade)

The LFC model is estimated using a system generalized method of moments (GMM) estimator for a sample of 79 countries over the period 1961–99.9 The regression analysis is conducted using 5- and 10-year averages, but for conciseness, we will refer only to the 10-year results.

Figure 6 shows that, in broad terms, the LFC model does well in explaining the direction of the changes in growth rates in Latin America and the Caribbean. The results show that the Dominican Republic, Haiti, and Latin America and the Caribbean all experienced a decline in their growth rates in the 1980s, whereas economic recovery in the Dominican Republic was stronger than the average economic recovery of Latin America and the Caribbean in the 1990s. However, the LFC model does not perform as well in explaining the magnitude of change, especially for Haiti. Therefore, to improve the fit of the model, we propose some modifications to the LFC specification.

Figure 6.Actual Changes in Growth Rates vs. Those Projected by Loayza, Fajnzylber, and Calderόn (LFC) (2005)

(In percentage points)

Improving the Fit of the LFC Model

Looking more deeply into the behavior of the different growth determinants for the Dominican Republic, Haiti, and Latin America and the Caribbean, we identify variables in the LFC model that can be enhanced to improve the fit of the model. In particular, we propose alternative measures for institutional quality and stabilization policies.

Institutional Quality

LFC uses an International Country Risk Guide (ICRG) index as a measure of institutional quality that captures the prevalence of law and order, quality of bureaucracy, absence of corruption, and accountability of public officials. However, this variable is not found to have a statistically significant impact on economic growth. This may be the case because the time series is too short, as ICRG data are only available from 1984 onward, and because the ICRG is a subjective indicator that measures perceptions of change in institutional quality, not the actual change. Furthermore, for the particular case of Haiti, the ICRG indicator does not provide an adequate picture of important political and institutional developments because it shows improvement in institutional quality during the most politically unstable periods, that is, the mid-1980s and early 1990s.

Another factor—directly linked to institutional quality—that has been found to have a negative impact on growth is political instability (Barro, 1991; Corbo and Rojas, 1993). We propose using changes in political regime as a proxy for political instability, by constructing a variable that measures the absolute magnitude of regime changes in each period—regardless of the type of regime (democratic or autocratic)—on the basis of information from the Polity IV database.10 In this way, the measure tests the relationship between political instability and growth without making a judgment on whether a democratic or authoritarian regime is better for fostering economic growth, given the mixed empirical evidence on the link between types of political regimes and economic performance.11

Figure 7 shows the contrast between Haiti and the Dominican Republic in terms of political instability. Haiti has faced much more political instability than the rest of Latin America, with several democratically elected regimes quickly turned over by coups or elected leaders becoming more authoritarian. On the other end of the spectrum, the Dominican Republic has been among the most stable countries in the region, in particular since 1970, when the Dominican Republic’s growth rate began to outpace that of Latin America.12

Figure 7.Magnitude of Regime Changes, 1970–2003

(Higher value indicates more political instability)

Source: Polity IV database.

Macroeconomic Stability

We find that the stabilization measures used by LFC have several drawbacks related to measurement difficulties. In particular, a credible estimation of potential output is complex, especially in countries that face structural breaks (through conflict or structural reform), making the output gap difficult to measure. Determining the real exchange rate equilibrium to calculate overvaluation of the exchange rate is also tricky. These issues are evident in the case of Haiti during the 1980s, where continuous poor economic performance resulted in low output gap volatility, implying better stabilization policies according to LFC variables, but clearly contrary to overall developments in the economy (Figure 8).

Figure 8.Loayza, Fajnzylber, and Calderόn (LFC) (2005) Stabilization Policies

Note: For illustrative purposes only, this figure combines the effects of the stabilization variables used by the LFC model, that is, inflation, cyclical volatility, and real exchange rate overvaluation. These variables were used independently in the LFC regression analysis.

Alternatively, to get an overall picture of the stance of stabilization policies, we propose a measure of macroeconomic instability that combines inflation, fiscal deficit, exchange rate volatility, and international reserves losses. The need to look at various factors simultaneously to determine the stance of macroeconomic policies has been underscored by Fischer (1993) and Sahay and Goyal (2006). A combined indicator is considered to be more appropriate because any variable taken in isolation provides only partial information. For example, inflation gives a good indicator of the stance of monetary and fiscal policies but may be biased by price controls. However, price controls would show up as higher fiscal deficits when funds are transferred to price control agencies. Moreover, although price controls might keep inflation down, uncertainty and lack of confidence in financial policies would put pressure on the exchange rate. Exchange rate pressures may not be evident in the context of a fixed exchange rate regime, but the authorities’ efforts to maintain foreign exchange stability would show up as changes in international reserve holdings.

More specifically, a macroeconomic instability index (mi) is constructed as the weighted sum of inflation rates and exchange rate volatility minus reserve accumulation as a percent of base money at the beginning of the period and minus the fiscal balance as a percent of GDP. Therefore, a higher value for the index indicates more instability. Each variable is weighted by the inverse of its standard deviation. Standardizing the variables ensures that all the components of the index have equal sample volatilities so that movements in the index are not solely driven by the most volatile component. Although the methodology for constructing the mi index draws from indicators of speculative pressure in the crisis literature, to our knowledge, this particular index has not been used elsewhere.13 The index is constructed as follows:

where mi is the macroeconomic instability index for country i at time t, cpi is the consumer price index, er is the exchange rate of national currency to U.S. dollar, res is the stock of international reserves, bm is the base money, fbal is the fiscal balance, gdp is the nominal GDP, and a is the standard deviation of each variable in the numerator.

Figure 9 illustrates the behavior of the mi index for the Dominican Republic and Haiti. Since the 1970s, the Dominican Republic has outperformed Haiti and Latin America on average in terms of stabilization policies. Meanwhile, Haiti’s performance in terms of stabilization policies was better than that of the region until the 1990s. This was largely the result of a conservative monetary policy, which consisted of a de facto currency board arrangement until 1979 and a tight policy stance during the 1980s with high reserve requirements and high real interest rates.14 Further details on stabilization policies in both countries are provided in the following subsection.

Figure 9.Indicator of Macroeconomic Instability

Note: Macroeconomic instability indicator is calculated by using data from IMF, International Financial Statistics.

Panel Regression

We use dynamic panel analysis to test the significance of the two proposed variables discussed above. We follow the basic approach of LFC, modifying the explanatory variables by (1) replacing the ICRG index with the regime change variable based on the Polity IV database; and (2) replacing the four LFC stabilization policy variables (inflation rate, standard deviation of output gap, real exchange rate overvaluation, and systemic banking crises) with the composite macroeconomic instability index (mi). We use the same 79-country sample, with 5- and 10-year intervals between 1961 and 1999, but discuss only 10-year results for simplicity and conciseness.15 The model is estimated with a GMM-IV system estimator using lagged levels and first differences of the variables as instruments and imposing robust two-step standard errors (Table 2).16

Table 2.Results of Panel Regressions on Real Per Capita GDP
Growth DeterminantsGMM-IV

System

Five-Year

Periods
GMM-IV

System

10-Year

Periods
Transitional convergence (initial GDP per capita, in logs)–0.0245–0.0495
[0.023][0.089]
Cyclical reversion (initial output gap, in logs)–0.2273–0.0167
[0.027][0.818]
Structural policies
Gross secondary school enrollment (in logs)0.01860.0145
[0.093][0.483]
Private domestic credit/GDP (in logs)0.00380.0092
[0.514][0.118]
Trade/GDP (in logs)0.01390.0101
[0.139][0.269]
Government consumption/GDP (in logs)–0.0208–0.0031
[0.064][0.819]
Main phone lines per capita (in logs)0.01120.0246
[0.174][0.002]
Stabilization policies
Macroeconomic instability index–0.0024–0.0021
[0.063][0.211]
Institutions
Polity IV index–0.0016–0.0003
[0.004][0.620]
External conditions
Growth rate of terms of trade0.01970.0592
[0.520][0.323]
Period shifts
1966–700.0058
1971–750.0038
1976–80–0.0010
1981–85–0.0227**
1986–90–0.0161
1991–95–0.0230*
1996–99–0.0287**
1970s–0.0081
1980s–0.0288***
1990s–0.0403**
Intercept0.08510.3010*
Specification tests (p-values)
Hansen test of overidentifying restrictions1.0000.412
First-order serial correlation0.0000.065
Second-order serial correlation0.2780.159
Source: Authors’ estimation using Arellano-Bond dynamic panel data estimation with robust two-step standard errors.Note: GMM = generalized method of moments. p-values are in brackets. *, **, and *** denote significance at 10 percent, 5 percent, and 1 percent, respectively, and refer only to the period shifts. Significance of coefficients is lower for the estimation with 10-year periods largely due to lower degrees of freedom. In addition, imposing robust two-step standard errors considerably lowers significance, especially for the estimation for 10-year periods. Data and programs are available at the IMF Staff Papers website.

Growth Determinants

The analysis of changes in growth rates between decades reveals that the model improves the fit for both the Dominican Republic and Haiti, as well as for Latin America (Figure 10). In the case of the Dominican Republic, the model is able to better predict the changes in growth trends in the 1970s and 1980s, though some of the growth boom of 1990s remains unexplained. For Haiti, the model provides a better explanation of developments over the entire period.

Figure 10.Actual Changes in Growth Rates vs. Those Projected by Loayza, Fajnzylber, and Calderόn (LFC) (2005) and Jaramillo and Sancak (JS) (this paper)

(In percentage points)

On the basis of these results, we use a case study approach that identifies the main determinants of growth for each period and for each country to provide a flavor of the specific policies being captured by the model. In general, structural policies have been the key determinant of growth in both the Dominican Republic and Haiti, followed by political stability and stabilization policies.

Dominican Republic

During the 1970s, average GDP per capita growth rates in the Dominican Republic increased to just above 5 percent from close to 3 percent in the 1960s. The improvement in growth rates is explained in most part by progress made on structural measures, in particular education and credit to the private sector, and by enhanced political stability (Figure 11). On the structural front, improvements in secondary school enrollment during the 1970s outpaced the average for Latin America and the Caribbean, which allowed the Dominican Republic to “catch up” with the average enrollment levels of the region. Specifically, the government introduced major curriculum reforms at the primary and secondary levels to make schooling more relevant to work life, resulting in improvements in access to education. Also during this period, credit to the private sector as a percent of GDP grew more rapidly in the Dominican Republic than that in the region, albeit still at lower levels than the Latin America and the Caribbean average. Commercial bank credit received a boost in the late 1960s when effective reserve requirements were lowered. The new system allowed banks to discharge part of their required reserve obligation by lending a certain minimum of their reserve liabilities to “productive” sectors. By end of 1973, all banks were under the new system. In terms of political stability, Joaquin Balaguer remained the president of the country for most of the 1970s, contrasting with the political unrest of the mid-1960s.

Figure 11.Dominican Republic: Changes in Growth Rates by Component

(In percentage points)

As with most of Latin America and the Caribbean, the 1980s were years of economic turmoil. Average growth of GDP per capita in the Dominican Republic was basically flat. The sharp contraction in growth rates during this decade can be explained by unfavorable external conditions and by the deterioration of stabilization policies. This took place despite considerable improvements in structural measures, particularly trade openness.17 During this period, large fiscal deficits contributed to monetary expansion and inflation pressures, which in turn exacerbated the distortions created by extensive price controls. The inconsistency between exchange rate policy and other financial policies generated overvaluations of the peso that resulted in sizable depreciations, and the central bank steadily lost official reserves.

In the 1990s, the Dominican Republic recovered the GDP per capita growth rates of the 1970s, just under 5 ½ percent of GDP. The primary contributors to growth were structural measures, in particular important improvements in trade and infrastructure, supported by the implementation of corresponding stabilization policies. During this period, the Dominican Republic embarked on a comprehensive economic program that included liberalization of the exchange rate, prices, and interest rates, as well as fiscal consolidation. This led to a sharp reduction in inflation, foreign exchange rate stability, and declining public external debt. Meanwhile, trade openness jumped as a result of the elimination in 1993 of export restrictions (such as export licensing and minimum export prices for agricultural products) and all export taxes, as well as the reduction in tariffs. In addition, a new foreign direct investment (FDI) law was approved in 1996 that facilitated operations of foreign firms, reduced sectoral restrictions, and liberalized repatriation of capital and utilities. This led to a substantial increase in FDI, in particular in the tourism sector. Furthermore, infrastructure in the Dominican Republic grew at a faster rate than the average for the region, with the expansion of telephone lines and electricity generation, as well as construction in tourism resort areas (including an international airport and roads).

Haiti

The 1970s was the only decade when the change in GDP per capita growth rates in Haiti was positive (Figure 12). From negative growth rates in the 1960s, growth in the 1970s averaged just above 2.5 percent. The main contributors to growth were structural measures, in particular trade openness, education, and credit to the private sector. In 1971, Haiti freed exchange transactions for all current payments and capital investment while eliminating restrictive practices and controls (including the exchange surrender requirement for exports and delays in remittances for service payments). This favored the export environment, and the assembly sector in Haiti began to grow. In addition, the Haitian government began to reform its educational system by unifying educational administration, introducing the use of Haitian Creole as the language of instruction in the first four grades, as well as implementing school nutrition programs. Meanwhile, credit to the private sector was encouraged by a 1972 reform that gave banks permission to extend medium-term credit to industrial and export sectors.

Figure 12.Haiti: Changes in Growth Rates by Component

(In percentage points)

The economic contraction in Haiti during the 1980s was stronger than that in Latin America and the Caribbean, at an average of—2.4 percent in Haiti compared with—0.9 percent for the region. Although external conditions were adverse, growth performance was affected by increasing political instability and a weakening of stabilization policies. Political turmoil increased in the early 1980s due to growing discontent with the Duvalier regime. The departure of Duvalier in 1986 was followed by successive failed elections and coups d’état. Between 1986 and 1990, there were six different heads of state. Political instability affected macroeconomic performance, as fiscal revenues began to weaken while the government pursued heavy outlays on construction works, defense, and loss-making public enterprises. Extrabudgetary spending continued to expand in the face of civil disturbances, as external concessional assistance declined. The public sector increasingly relied on central bank financing that led to official reserve losses, and external payments arrears emerged, putting rising pressures on the exchange rate and domestic prices.

GDP per capita growth in the 1990s continued the same negative trend as in the 1980s, with an average growth rate of—2.4 percent. Growth performance over the period was dominated by a major shock to the economy: following the coup d’état against Jean Bertrand Aristide, the United States and the Organization of American States imposed a trade embargo in 1991, aggravated by a UN-sponsored oil embargo in 1993. Trade openness declined during the embargo years from about 40 percent of GDP to an average of 25 percent. Assembly sector exports were severely affected. At the end of the 1980s, Haiti assembled a wide variety of light manufactures, such as baseballs and electrical switches, together with apparel products. With the embargo, the assembly industry collapsed, and following the lifting of sanctions in 1994, only the garment sector reestablished itself. Employment in the assembly sector fell from 46,000 in the late 1980s to just 5,000 in 1995.

IV. Conclusions

This paper tries to explain why the Dominican Republic and Haiti have experienced a striking divergence in growth performance despite their broad similarities in terms of geography and historical institutions. We examine both the initial conditions and the different policies pursued in each country since 1960 in order to identify the factors that may have contributed to this divergence.

When examining initial conditions, namely geography and historical institutions, we find strong similarities between the Dominican Republic and Haiti, implying that initial conditions cannot fully explain the divergence in real incomes of the two countries. Although, by 1960, Haiti was already trailing the Dominican Republic in some social indicators, differences were not inherited from the historical legacies of the 19th or early 20th centuries. Although lack of national accounts prior to 1960 does not allow us to determine the exact point in time when the Dominican Republic started to outpace Haiti, disparities are likely to have started to emerge in part as a result of the policies implemented in each country following the U.S. military occupation.

The panel regression and case study approach allow us to conclude that policy decisions since 1960 have played a central role in the growth divergence between Haiti and the Dominican Republic. In general, structural policies have been the key determinants of growth in both the Dominican Republic and Haiti, followed by political stability and stabilization policies. In particular, we find that the Dominican Republic has consistently outperformed Haiti and Latin America and the Caribbean in terms of implementation of structural measures, stabilization policies, and political stability. Meanwhile, Haiti has lagged the region in implementing structural policies while being subject to numerous political shocks that have severely affected its growth performance.

Although this paper identifies the kind of policies linked to growth performance in both countries of the Hispaniola, future work would be needed to understand why decision makers in the Dominican Republic chose and were able to implement superior policies to those in Haiti. Further research could also provide insight into the sources of the persistent political turmoil in Haiti, as well as the circumstances that allowed the Dominican Republic to surmount the political instability of the 19th and early 20th centuries.

Appendix I
Table A1.Chronology of Political Events
DominicanRepublic
1821–43In 1821, the Dominican Republic gains independence from Spain but is soon after invaded by Haitian leader Jean-Pierre Boyer, uniting the island for the next 22 years.
1844On February 27, 1844 (Dominican Independence Day), Juan Pablo Duarte executes a bloodless coup in Santo Domingo. During the next 16 days, all of the eastern towns announce their decision to separate from Haiti.
1844–65General Pedro Santana Familias and Buenaventura Baez Mendez dominate the political scene. In 1861 Santana agrees to the Dominican Republic’s annexation by Spain. However, on March 3, 1865 the annexation is annulled, and Spain withdraws its soldiers.
1865–78Political turmoil and corrupt governments dominate this period.
1879–82Two dominant parties—the Azules and the Rojos—emerge. For the next three years, the Azules control the government.
1882–99General Ulises Heureaux comes to power. He retains power through fraudulent elections and army control. Heureaux borrows vast sums from U.S. investors at high interest rates. He also replaces the National Bank with the U.S.-owned and operated San Domingo Improvement Company. Heureaux is assassinated in July 1899.
1900–05Soon after Heureaux’s death, the Dominican Republic could not repay its debts. The U.S. government intervenes by taking control of the customs houses in 1905, guaranteeing repayment of all loans.
1905–11Ramon Cáceres is elected president. He constructs the railway and highways, rebuilds docks, improves the postal service, installs telegraph lines, and funds new schools. He increases export taxes on Dominican sugar to pay for public improvements, which angers plantation owners. Cáceres is shot dead in 1911.
1911–16Various revolutions follow Cáceres’ death. U.S. President Woodrow Wilson, concerned about U.S. national security, threatens to send marines if elections are not held. Juan Isidro Jinenez is elected president but is soon impeached before Congress. The United States offers support, and though Jimenez only requests weapons, the marines are sent in.
1916–24The Dominican Republic comes under U.S. control for eight years. In particular, the Americans control the budget. The American troops leave by 1924, as Woodrow Wilson is no longer in power, World War I is over, and the United States is considerably less concerned with the Dominican Republic’s strategic importance.
1924–30President Horacio Vasquez has a progressive government, building roads creating access to the countryside, schools, and irrigation and sanitation services.
1930–61Rafael Leonidas Trujillo, chief of the National Police (which later became the National Army), forces Vasquez to resign. Trujillo holds an election for which he is the sole candidate. Trujillo rules the Dominican Republic with an iron fist from 1930 to 1947 and indirectly thereafter until his assassination in 1961. Trujillo amasses a personal fortune by establishing monopolies that his family controls. He carries out programs of public works and construction. He also presses for industrial progress, and scores of factories are opened. Agricultural production improves and the economy flourishes.
1962–66President Joaquin Balaguer is in office at the time of Trujillo’s assassination. Elections are organized. In 1962 Juan Bosch Gaviño is elected, but is toppled in a military coup in 1963. Bosch and a group of supporters who call themselves the Constitutionalists take to the streets and seize the National Palace. To reinstate order, 24,000 U.S. soldiers are ordered to the Dominican Republic until new elections in 1966.
1966–78Balaguer defeats Bosch in national elections. Balaguer purges the military and uses the National Police to curtail nonmilitary opposition. His reelections in 1970 and 1974 are mostly accomplished through intimidation. The economy expands rapidly, benefiting from favorable world prices for sugar. However, by the late 1970s, plunging sugar prices and rising oil costs bring the Dominican economy to a standstill.
1978–86Silvestre Antonio Guzman defeats Balaguer in the elections. He leads a corrupt government. Public works programs are brought to a halt, and the administration borrows heavily from abroad. Guzman’s popularity diminishes rapidly. Salvador Jorge Blanco is elected in 1982.
1986–96Balaguer is reelected for a fifth term. He runs his government like a dictatorship, intimidating political rivals. He reverses the adjustment program under Blanco, leading to a sharp depreciation of the peso and annual inflation of 60 percent. By 1990, 900,000 Dominicans move to New York, fleeing the economic situation.

Balaguer rigs the 1990 and 1994 elections. However, in 1994 the military threatens to intervene. Balaguer agrees to cut his last term short and hold elections 18 months later.
1996–2000Leonel Fernandez wins the presidency. He presides over strong economic growth, privatization, and structural reform.
2000–04Hipolito Mejia becomes president. The events of September 11, 2001 and the ensuing slowdown in the world economy take a toll on economic growth. In 2003, failures of three large banks due to accounting malpractices and mismanagement lead to a banking crisis.
2004–06Fernandez is reelected in 2004. The economy recovers at high growth rates.
Haiti
1804The ex-slaves of the French colony of Saint-Domingue declare independence after defeating the French army at Vertiere. They name their new nation Haiti, which means “mountainous” in the language of its original Arawak inhabitants.
1804–20Jean Jacques Dessalines proclaims himself emperor, but is assassinated in 1806. His death leads to civil war between the south under General Pétion and the north under Henry Christophe.
1820–43Faced with a rebellion by his own army, Christophe commits suicide, paving the way for Jean-Pierre Boyer to reunify the country and become president of the entire republic in 1820. President Boyer invades Santo Domingo following its declaration of independence from Spain, controlling the entire island until 1844. In 1938, France recognizes Haitian independence in exchange for an indemnity of 150 million francs to compensate for the losses of French planters during the 1803 revolution.

In 1843, Boyer flees the country following a revolution to overthrow him from the presidency.
1844–1915Haiti sees 22 heads of state, most of whom leave office by violent means. The United States and other slaveholding states do not recognize Haiti until 1862.
1915–34Concerned about German economic influence in the country and continued political instability, the U.S. military occupies Haiti after the last in a series of short-lived presidents is torn to pieces by a mob. The U.S. marines establish control over customshouses and port authorities, and create the Haitian National Guard.
1934–56Once the Americans leave, political instability resumes, with regime changes taking place through military coups. In 1937, thousands of Haitians living near the border of the Dominican Republic are massacred by Dominican soldiers under the orders of President Trujillo.
1957–71Francois Duvalier (Papa Doc) is elected president in the country’s first universal suffrage election. Reelected in fraudulent elections in 1961, he declares himself president for life in 1964 and, until his death, holds on to power as a ruthless dictator with the help of a paramilitary force known as Tontons Macoutes.
1971–86Jean-Claude Duvalier (Baby Doc) becomes the new president for life after his father’s death in 1971.

After relatively favorable economic performance in the 1970s, the macroeconomic situation deteriorates sharply in the early 1980s, affected by the recession in the United States and economic mismanagement. Social discontent culminates in public protests and violence, and Duvalier flees to France in 1986.
1986–94Starting in 1986 there are several brief attempts at civilian democracy, each terminated by a military coup.

In September 1991, Jean-Bertrand Aristide, a popular priest elected in 1990, is forced to flee the country only nine months after taking office. The United States and the Organization of American States respond with a trade embargo, and in 1993 a UN-sponsored oil embargo is imposed. In September 1994, Aristide is reinstalled in office with the help of a U.S.-led military intervention.
1995–96Aristide disbands the army. Aristide supporters win parliamentary elections. As, according to the constitution, Aristide cannot serve a consecutive second term, he is succeeded by a close ally, René Préval.
1997–99Following a period of political deadlock, Préval dissolves parliament and rules by decree.
2000–03Aristide is reelected president, but the international community criticizes irregularities during the parliamentary elections. Political stalemate with the opposition leads to the freezing of foreign aid.
2004–05Amid deteriorating economic conditions, growing dissatisfaction with the government culminates in an armed rebellion. President Aristide resigns in February 2004 and leaves the country. A transition government is formed to lead the country to elections and an interim multinational force arrives.
2006Elections are held in February 2006, and René Préval is reelected president.
Appendix II
Table A2.Variable Definitions and Sources
VariableDefinitionSource
Real GDP per capita1985 PPP adjusted U.S. dollars (in logs)Penn World Tables 6.1 (Heston, Summers, and Aten, 2002)
Transitional convergenceInitial GDP per capita (in logs)
Cyclical reversionInitial output gap The difference between the log of actual GDP and the log of trend GDP at the beginning of the period. Trend growth is determined using a Baxter-King filter.
EducationGross secondary school enrollment (in logs) The gross enrollment ratio is defined as the total enrollment, regardless of age, expressed as a percentage of the official school-age population for a given level. Gross enrollment ratio can be over 100 percent due to the inclusion of over-aged and under-aged students.UNESCO and LFC
Financial depthPrivate domestic credit/GDP (in logs) Ratio to GDP of the stock of claims on the private sector by deposit money banks and other financial institutions.IMF, International Financial Statistics (IFS)
Government burdenGovernment consumption/GDP (in logs).IFS
Public infrastructureMain phone lines per capita (in logs) The number of telephone mainlines per 1,000 people.World Bank, World Development Indicators (WDI)
Trade opennessThe residual of a regression of the log of the ratio of exports and imports to GDP, on the logs of area and population, a dummy for oil-exporting countries, and a dummy for landlocked countries.Authors’ estimates based on IFS and WDI data
Terms of tradeChange in terms of trade Log difference of the terms of trade. Terms of trade is estimated as customary.IMF, World Economic Outlook (WEO)
Alternative Measures
Political instabilityChange in political regime The absolute magnitude of regime changes in each period, as determined by the Polity IV polity scale.Authors’ estimates using Polity IV database.
MIMacroeconomic instability index Weighted sum of inflation rates and exchange rate volatility minus reserve accumulation as a percent of base money at the beginning of the period and minus the fiscal balance as a percent of GDP. Each variable is weighted by the inverse of its standard deviation.Authors’ estimates based on IFS data
Inflation rateAnnual percentage change in the consumer price index (in logs)Authors’ estimates based on IFS data
Exchange rate volatilityAnnual percentage change of the foreign exchange rate (in logs)Authors’ estimates based on IFS data
International reserve accumulationAnnual change in net international reserves as a percent of base money at the beginning of the period (in logs)Authors’ estimates based on IFS data
Fiscal balanceGovernment fiscal balance/GDP (in logs)Authors’ estimates based on IFS data
Appendix III
Table A3.List of Countries Included in the Panel Regression
AlgeriaKenya
ArgentinaKorea, Republic of
AustraliaMadagascar
AustriaMalawi
BangladeshMalaysia
BelgiumMexico
BoliviaMorocco
BotswanaNetherlands
BrazilNicaragua
Burkina FasoNiger
CanadaNigeria
ChileNorway
ColombiaPakistan
Congo, Democratic RepublicPanama
Congo, RepublicPapua New Guinea
Costa RicaParaguay
Côted’Ivoire
DenmarkPhilippines
Dominican RepublicPortugal
EcuadorSenegal
Egypt, Arab RepublicSierra Leone
El SalvadorSouth Africa
FinlandSpain
FranceSri Lanka
GambiaSweden
GhanaSwitzerland
GreeceSyria, Arab Republic
GuatemalaThailand
HaitiTogo
HondurasTrinidad and Tobago
IcelandTunisia
IndiaTurkey
IndonesiaUganda
Iran, Islamic RepublicUnited Kingdom
IrelandUnited States
IsraelUruguay
ItalyVenezuela
JamaicaZambia
JapanZimbabwe
Jordan
References

    Acemoglu, Daron, SimonJohnson, and James A.Robinson,2001, “The Colonial Origins of Comparative Development: An Empirical Investigation,” American Economic Review, Vol. 91, No. 5, pp. 1369101.

    Alpert, Leo, 1941, “The Areal Distribution of Mean Annual Rainfall over the Island of Hispaniola,” Monthly Weather Review, Vol. 69, No. 7, pp. 2014.

    Barro, Robert,1991, “Economic Growth in a Cross-Section of Countries,” Quarterly Journal of Economics, Vol. 106, pp. 40743.

    Bulmer-Thomas, Victor,2001, “The Wider Caribbean in the 20th Century: A Long-Run Development Perspective,” Integration and Trade, Vol. 5, No. 15, pp. 556.

    Corbo, Vittorio, and PatricioRojas,1993, “Investment, Macroeconomic Stability, and Growth: The Latin American Experience,” Revista de Análisis Económico, Vol. 8, No. 1, pp. 1935.

    Diamond, Jared,2005, Collapse: How Societies Choose to Fail or Succeed (New York, Viking Penguin Group).

    Eichengreen, Barry, and others1995, “Exchange Market Mayhem: The Antecedents and Aftermath of Speculative Attacks,” Economic Policy, Vol. 10, No. 21, pp. 249312.

    Fischer, Stanley,1993, “The Role of Macroeconomic Factors in Growth,” NBER Working Paper No. 4565 (Cambridge, Massachusetts, National Bureau of Economic Research).

    Gallup, John Luke, JeffreySachs, and AndrewMellinger,1998, “Geography and Economic Development,” NBER Working Paper No. 6849 (Cambridge, Massachusetts, National Bureau of Economic Research).

    Herrera, Santiago, and ConradoGarcia,1999, “User’s Guide to an Early Warning System for Macroeconomic Vulnerability in Latin American Countries,” Policy Research Working Paper No. 223 (Washington, World Bank).

    Heston, Alan, RobertSummers, and BettinaAten,2002, Penn World Table Version 6.1, (Center for International Comparisons at the University of Pennsylvania (CICUP), October).

    International Monetary Fund (IMF),2001, “The Dominican Republic: Stabilization, Reform and Growth,” Occasional Paper No. 206 (Washington, International Monetary Fund).

    Kaminsky, Graciela, and Carmen Reinhart,1999, “The Twin Crises: The Causes of Banking and Balance of Payments Problem,” American Economic Review, Vol. 89, No. 3, pp. 473500.

    Klasen, Stephan, and ThorstenNestmann,2004, “Population, Population Density, and Technological Change,” CESifo Working Paper No. 1209 (Munich, CESifo Group).

    La Porta, Rafael, FlorencioLopez-de-Silanes, AndreiShleifer, and RobertVishny,1998, “The Quality of Government,” NBER Working Paper No. 6727 (Cambridge, Massachusetts, National Bureau of Economic Research).

    Loayza, Norman, PabloFajnzylber, and CésarCalderón,2005, Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations, and Forecasts (Washington, World Bank).

    Lundahl, Mats,2001, “Poorest in the Caribbean: Haiti in the Twentieth Century,” Integration and Trade, Vol. 5, No. 15, pp. 177200.

    Sahay, Ratna, and RishiGoyal,2006, “Volatility and Growth in Latin America: An Episodic Approach,” IMF Working Paper 06/287 (Washington, International Monetary Fund).

    Sirimaneetham, Vatcharin,2006, “Explaining Policy Volatility in Developing Countries,” Bristol Economics Discussion Papers No. 06/583 (Bristol, University of Bristol).

    Treisman, Daniel,2000, “The Causes of Corruption: A Cross-National Study,” Journal of Public Economics, Vol. 76, No. 3, pp. 399457.

    WorldBank,2006, “Haiti: Country Economic Memorandum,” Latin America and Caribbean Region (Washington, World Bank).

Laura Jaramillo and Cemile Sancak are economists in the IMF’s Western Hemisphere and Fiscal Affairs Departments, respectively. The authors would like to thank Guy Meredith, Jeromin Zettelmeyer, and Andy Wolfe for their encouragement and support. We also thank Dan Holmes, Juan Climent, Chris Towe, Luis Cubeddu, and participants at the Western Hemisphere Department seminar and growth workshop for their useful comments. Volodymyr Tulin provided valuable research assistance.

Haiti is about half the size of the Dominican Republic, but has roughly the same population.

Looking at GDP growth by sector in Haiti and Dominican Republic over the period 1960–2000 reveals that performance of the primary sector did not drive overall growth performance in either country. In the Dominican Republic, the primary sector explains about 12.5 percent of the total average growth rate, whereas the secondary and tertiary sectors explain 32.5 and 55 percent, respectively. Similarly, in Haiti, the primary sector explains less than 10 percent of the total average growth rate, whereas the secondary and tertiary sectors explain about 45 percent each.

Appendix I provides a chronology of political events for both the Dominican Republic and Haiti.

U.S. expansion into the Caribbean Basin—at a time when the United States was pursuing the construction of the Panama canal—was supported by the Monroe Doctrine, originally intended to keep European nations out of Latin America, and Theodore Roosevelt’s corollary to this doctrine, which stated that the United States had a moral mandate to enforce “proper” behavior among Latin American countries.

In 1960, life expectancy at birth was 44 years in Haiti compared with 54 years in the Dominican Republic. The under-5 mortality rate per thousand children was 253 in Haiti compared with 149 in the Dominican Republic. The adult illiteracy rate above age 15 (data available for 1970) was 78 percent in Haiti and only 33 percent in the Dominican Republic.

The capital stock data were constructed using aggregate investment figures. We use a perpetual inventory method to compute capital stocks, with a share of capital income in national output of 0.4 (different values do not alter the conclusions). We do not control for capacity utilization or quality of human capital, due to data limitations.

See Appendix II for details on definitions and sources.

See Appendix III for the list of countries included in the panel regression.

The Polity IV database provides a unified polity scale that ranges from +10 (strongly democratic regime) to –10 (strongly autocratic regime) on an annual basis for all countries between 1800 and 2004.

For completeness, we also estimate the model directly using the polity scale from the Polity IV database. The interpretation of this measure goes beyond political instability and focuses on the effect of democracy (or autocracy) on growth by taking into account the direction of regime change as well as its magnitude.

Since 1960, Haiti has experienced numerous regime changes, including following the fraudulent elections of Papa Doc (1961); the death of Papa Doc (1971); the freeing of political prisoners and loosening of control over the press by Baby Doc (1977); the departure of Baby Doc (1986); failed elections and coups d’état (1988–91); return of Aristide (1994); the dissolving of Parliament (1999); and irregular elections (2000). Regime changes in the Dominican Republic have been few, especially since 1970, and include the death of Trujillo (1961); coup and civil war (1962–63); Balaguer’s defeat in the elections (1978); and Balaguer’s agreement to cut his term short following rigged elections (1994). See Appendix I.

See Eichengreen and others (1995), Kaminsky and Reinhart (1999), and Herrera and Garcia (1999). Sensitivity analysis indicated that the panel regression results are largely robust to the choice of weighting scheme.

Until 1979, the central bank law stipulated that the amount of currency issued had to be fully covered by foreign reserves in order to preserve the exchange rate parity.

Both period intervals yield similar results; the coefficients have identical signs and are broadly of the same order of magnitude; however, the significance of these coefficients is lower for the 10-year estimation largely due to lower degrees of freedom.

The model is also estimated by replacing our measure of political instability with the polity scale from the Polity IV database, as discussed in footnote 11. This measure has the expected positive sign (not reported in Table 2); however, it is significant neither for 5- nor 10-year intervals.

The increase in trade openness is explained by improvements in exports arising from greater trade benefits to the U.S. market, as the Caribbean Basin Initiative came into effect in 1983, and by the approval of the 1983 Free Trade Zone Law that provided incentives for export industries, including 20-year tax exemptions.

Other Resources Citing This Publication