The main purpose of this paper is to investigate how the political instability in 2009 affected the Madagascar vanilla exports. Madagascar vanilla sector has experienced robust growth over the past decade which contributes to both the economy and the job creation, especially in northeast region where Vanilla was planted. However, this growth faced a serious problem after the political instability in 2009. This paper shows that the suspension of Africa Growth and Opportunity Act (AGOA), due to the political instability, had a large negative impact on vanilla exports. To estimate its impacts on vanilla exports, the international trade data permit us to isolate those impacts from various factors between Madagascar and other vanilla exporting countries by using the different-in-different (DID) technique. We take the sample of 2007 to 2011 by using Tobit model with random effects in addition to OLS with fixed effect in order to observe how far political instability hampers the vanilla exports in Madagascar. Through analysis, we find that the political instability has a negative effect on economic growth in terms of Madagascar vanilla exports.
In today’s world, political instability remains the most questionable issues and hard to tackle. Too many countries have faced this turmoil especially in Sub-Saharan Africa (SSA). Madagascar, which is located in the Southern Indian Ocean, has experienced cycles of political instability over the last five decades. However, it is well-known by its nature and among the one enormous potential on cash crop, namely vanilla, coffee, and cloves. And despite its recent progress, Madagascar is shaken by a profound political instability in 2009. Yet, in the context of domestic political turmoil together with the global economic crisis, the Malagasy economy has not been growing since the second quarter of 2009. Weak governance and inefficiency of political parties remain the key constraints to Madagascar’s development. The governance and financial management challenges have been both the cause and consequence of this crisis. The absence of concrete and effective follow-up within government, the inappropriate mingling of public and private interests and institutional weaknesses has reduced the level of efficiency, transparency and liability of government action. Therefore, Malagasy people have felt unsatisfied and powerless and even have lost their trust and hope on their leaders. And this led people to oust President Marc Ravalomanana in military-backed coup in March 2009 and replaced by Andry Rajoelina who took power to lead a transitional government. The international organization has suspended most aid, except on humanitarian and poverty grounds, on which the country is highly dependent. In January 2010, Madagascar couldn’t have access on AGOA as the American government revoked the country’s eligibility. And like most other donors, the World Bank has suspended new lending and grants.
Vanilla is a member of the orchid family, which is a native of Mexico. It was brought to Madagascar in the 1890s by French colonists, and is generally produced in the North-East coast of Madagascar. Most of the country’s vanilla beans are grown in the humid and wet climate of the rainforest covering the SAVA (Sambava, Antalaha,Vohémar, and Andapa) region, where about 70% of the population depend on this lucrative and fragrant spice. Thus, the principal economic motor for the SAVA region has been the vanilla industry, with production of green vanilla increasing from 4000 to 10,000 tons between 1994 and 2004 [
At the end of the year 2007, Andry Rajoelina was elected as a mayor of the capital of Madagascar (Antananarivo). However, there were problems happened between the mayor and the President Marc Ravalomanana in 2008. And different angles of unpopular policies ratified by the president were criticized by Andry Rajoelina. In the same year, the government apparently delayed to pay the funds for various local government infrastructure projects in the capital, and even diverted investment away from the city. And in too many other ways, they refused to facilitate cooperation between the municipal and the central governments. Many people said that the government’s strategies were intended to undermine Andry Rajoelina. He also criticized the President Ravalomanana who agreed that large parts of Malagasy land are leased to the company Daewoo from Korea. Furthermore, the mayor of the capital found a linked business interests between President Marc Ravalomanana and his party TIM (Tiako i Madagasikara) which needs to be controlled. The Malagasy political instability became very serious when a young mayor was self-proclaimed as the president of Madagascar on March 21st 2008.
Political instability is the qualitative phenomena which are difficult to measure quantitatively and not easily to define.
Political stability provided a description of a well-functioning government while political instability is the propensity of a government collapse either because of conflicts or rampant competition between various political parties [
1) Economic Inequality
It is important to state that the power given to the government influences the leaders of some countries to make a change in economic and social policies for their personal interests, not for their nations. Such an anti-social behavior would harm the economic performance hence the birth of economic inequality. With a polarized and inequality distribution of national resources, people will find reasons to pursue their interest outside the normal channels of both political representation and market activities. Therefore, people in unequal societies, as Perotti (1996) [
2) Income
In reference to Blanco and Grier (2009) [
3) Inflation
A possible macroeconomic factor which influences the government to be stable or unstable is based on inflation rates in a country. Cukierman and Tabellini (1992) [
4) Social Inequality
According to Blanco and Grier (2009) [
In this section, we will present previous literature made on relationship between political instability and economic growth which has been discussed by various researchers and is not new at all. The most prominent literatures are reviewed as follows: Political instability can affect economic growth in different number of channels. According to Aisen and Veiga (2013) [
- Regime-related instability: Coups d’état, governmental crisis, purges, cabinet changes.
- Instability induced by civil-society: assassinations, general strikes, guerilla warfare, riots, revolutions, anti-governmental demonstrations.
[
This paper used mainly secondary data which were sourced from National Institute of Statistics of Madagascar (INSTAT) [
As we try to identify the impacts of political instability on vanilla exports, we require information from abroad to distinguish the effects of political instability from financial crisis. To estimate the impacts of political instability on vanilla exports, the international trade data allow us to isolate those impacts from various factors to Madagascar and other vanilla exporting countries by using the different-in-different (DID) technique. To avoid extraordinary fluctuation in export, we selected the exporting countries that take share in USA and EU markets with more than 0.1%. Our methodology requests the assumption that domestic political instability systematically and differently affected vanilla exports from Madagascar compared with other countries. The suspension of AGOA due to the political instability is the most likely factor to cause differential changes in exports between Madagascar and other countries. So our analysis is restricted to the period before 2009 or vanilla export to the EU market. Access to the EU market applied to least developed countries (LDC) changed in 2011 with the relaxing of the rule of origin from double to single transformation, which is equivalent to the one applied to Madagascar. While it affects performance of Madagascar’s exporters to the EU market, it does not affect our analysis until 2010. As we know that some Madagascar companies export Vanilla to USA and other markets. Our fundamental assumption is that the suspension of AGOA systematically and differently affects changes of vanilla exports to the USA market compared with those to the EU market. We apply the difference-in-differences-in-differences (DIDID) method in order to avoid the possible bias. This method takes the difference in changes of vanilla export volume in the USA market as well as in the EU market resulting from Madagascar and the same difference from other exporting countries. We consider only two periods 2009 and 2010, and two countries for simplicity, it is expressed as follows:
D I D I D = [ ( ln V 2010 MD,USA − ln V 2009 MD,USA ) − ( ln V 2010 MD,EU − ln V 2009 MD,EU ) ] − [ ( ln V 2010 Other,USA − ln V 2009 Other,USA ) − ( ln V 2010 Other,EU − ln V 2009 Other,EU ) ] (1)
where:
ln V 2010 MD,USA = the log of the value of Vanilla imports from Madagascar to the USA in 2010;
ln V 2010 Other,USA = the log of the value of Vanilla imports from another country to the USA in 2010;
ln V 2010 MD,EU = the log of the value of Vanilla imports from Madagascar to the European Union in 2010 and;
ln V 2010 Other,EU = the log of the value of Vanilla imports from another country to the European Union in 2010.
This operation allows for changes in demand in markets irrelevant to suspension of AGOA to be excluded from our estimation. For factory-level changes, DIDID cannot be applied and its estimation, based on DID, is susceptible to possible violation of the assumption mentioned earlier. Complete differences in characteristics between control and treatment groups, namely exporters to the USA and those to other markets are controlled.
Following are the hypotheses of this study:
➢ H1: Political instability affects negatively Madagascar vanilla exports to USA and EU markets in 2009.
➢ H2: Suspension of AGOA deteriorates Madagascar vanilla export to USA.
The main objective of the current study is to test the hypothesis that the political instability affects negatively Madagascar vanilla exports to USA and EU markets. The vanilla exports from Madagascar to the USA and EU markets decreased 4.9% in 2008, and this reduction reached significantly 18.0% in 2009 and 38.5% in 2010 [
ln V v , x , m , t = α 0 + α 1 Crisis x , t + β year t + Vanilla ∗ Country ∗ Market v , x , m . (2)
where:
lnV = represents the log of the value of Vanilla imports in the USA or EU markets at the six-digit different quality of vanilla in the Harmonized System (HS);
Crisis = a dummy variable and takes one for imports from Madagascar after 2009;
Year = a dummy for a set of years;
Vanilla*country*market = represents the fixed effect, and v, x, m, and t indicate respectively a product at the six-digit level, country, market, and time1.
The DID estimate is α1, and a triple interaction term is incorporated to control the possible impact of AGOA suspension. We also estimated the above model using import value only in the EU market, dropping the triple interaction term. We include the observations with zero import value to estimate the extensive and intensive margins. But, we also need to insert 1 (dollar) for these observations in order to get the logarithm. Tobit model with Random Effect (RE) is applied in addition to OLS with Fixed Effect (FE) to seeing the sensitivity of including these censored observations [
In
To analyze the measurement of the impact of AGOA suspension, DIDID using imports in the USA and EU vanilla markets from six low-income countries is applied in the following form:
ln V v , x , m , t = β 0 + β 1 ( U S A m × M G x ) + β 2 ( U S A m × y r 2010 t ) + β 3 ( M D x × y r 2010 t ) + year t + Vanilla ∗ Country ∗ Market v , x , m .(3)
1Compared with the standard DID model, we added year dummies and applied the vanilla-country-market fixed effect rather than the product fixed effect and market and country dummies to allow for flexibility in the base level.
where:
USA = a market dummy (equal to 1 if observations represent import values in the USA);
USA and EU market (2007-2009) | EU market (2007-2010) | |||
---|---|---|---|---|
OLS FE 1 | Tobit RE 2 | OLS FE 3 | Tobit RE 4 | |
MD*post 2009 | −0.1381 (0.1871) | −0.4632** (0.2352) | −0.3730* (0.2191) | −0.6451** (0.2910) |
yr2007 | −0.1062** (0.0541) | −0.1321** (0.0662) | −0.1563** (0.0691) | −0.1880** (0.0830) |
yr2009 | −0.0121 (0.0562) | 0.0243 (0.0641) | 0.1424 (0.0623) | 0.1910*** (0.0701) |
yr2010 | 0.2761*** (0.0701) | 0.3490*** (0.0820) | ||
_cons | −3.6381*** (0.0310) | −4.4112*** (0.1621) | −3.2873*** (0.0432) | −3.8340*** (0.1821) |
Fixed/Random effect | vanilla*country *market | vanilla*country *market | vanilla*country | vanilla*country *market |
Marginal effect of political crisis | −12.90% | −33.82% | −31.11% | −45.33% |
R2 | 0.0009 | 0.0090 | ||
Log likelihood | −20164.571 | −13215.270 | ||
N | 8283 | 8283 | 5732 | 5732 |
Source: Author’s Estimation. aRemark: Numbers in parentheses represent clustered standard errors for OLS and bootstrap standard errors for the Tobit model. The marginal effect of political crisis is obtained by [Exp (coef.) − 1]*100, where the coefficient of Tobit models is transformed to indicate marginal effect on observed (censored) dependent variable meanwhile *, **, and *** indicate respectively p < 0.1, 0.05, and 0.01.
yr2010 = a dummy for year 2010 when AGOA is suspended and;
MD = represents a dummy for a country from which a product is exported (it is 1 if an observation represents vanilla exports from Madagascar and 0 if it represents vanilla exports from the other countries) and the triple DID estimate is β1.
DIDID | DID | |||
---|---|---|---|---|
OLS FE (2) | Tobit RE (1) | OLS FE (4) | Tobit RE (3) | |
USA*MD* yr2010 | −1.0121*** (0.3670) | −1.7730*** (0.5462) | ||
USA*yr2010 | −0.2492*** (0.0951) | −0.3351** (0.1122) | −1.2610*** (0.3551) | −2.5213** (0.6021) |
yr2007 | −0.1061 (0.0542) | −0.1332** (0.0641) | −0.1453 (0.2023) | −0.2430 (0.3331) |
yr2009 | −0.0321 (0.0552) | −0.0401 (0.0641) | −0.1690 (0.1981) | −0.3112 (0.3150) |
yr2010 | 0.2470*** (0.0632) | 0.3322*** (0.0702) | −0.1401 (0.2770) | −0.0800 (0.4152) |
MD*yr2010 | −0.3281 (0.2452) | −0.5310 (0.331) | ||
_cons | −3.6381*** (0.0350) | −4.4230*** (0.1461) | −6.5521*** (0.1272) | −8.5070*** (0.5311) |
Fixed/Random effect | vanilla*country *market | vanilla*country *market | vanilla*country | vanilla*country |
Marginal effect of AGOA suspension | −63.60% | −78.12% | −71.70% | −81.6% |
R2 | 0.0009 | 0.0210 | ||
Log likelihood | −25997.480 | −3592.81 | ||
N | 11044 | 11044 | 1592 | 1592 |
Source: Author’s Estimation. aRemark: Numbers in parentheses represent clustered standard errors for OLS and bootstrap standard errors for the Tobit model. The marginal effect of political crisis is obtained by [Exp (coef.) − 1]*100, where the coefficient of Tobit models is transformed to indicate marginal effect on observed (censored) dependent variable meanwhile *, **, and *** indicate respectively p < 0.1, 0.05, and 0.01.
than estimate by DIDID, and thus, adverse impact will be overestimated. Finally, we run DID model and found that estimated effect is larger by 4.5% to 12.7% than those in the DIDID model (column 3 and 4).
Although, the vanilla production displayed robust growth in 2003, overcoming the market liberalization in 1990, this sector experienced a critical effect after the domestic political instability in 2009. Our analysis indicates that the political instability has negatively affected the vanilla exports (reduction of Vanilla exports by 31% to 45%). And it is more drastic when the main importers (USA) of Madagascar vanilla have suspended the duty-free access (the AGOA) to the USA vanilla market due to the government instability or political crisis. Furthermore, the estimated results show that suspension of AGOA caused exports to the USA vanilla market to fall by 64% to 78% which reveals that political instability reduces significantly economic growth.
In terms of vanilla production, Madagascar has never been able to industrialize this sector even though we have been trading with USA and EU for several decades. At the absence of the processing industry of its products, therefore, the researcher recommends for stronger policies and regulations in efforts from the competent authority to revive the vanilla sector in order to boost the vanilla exportation not only to USA and EU but also worldwide. Furthermore, the Malagasy government needs to intervene and have bilateral talks with other countries in efforts to lure overseas investors who can input technological and innovation aspects in vanilla production by setting up the factory of transformation of vanilla in Madagascar.
Andriamahery, A. and Zhou, J. (2018) The Impact of Political Instability on Madagascar Vanilla Exports. Open Journal of Social Sciences, 6, 27-38. https://doi.org/10.4236/jss.2018.64003