Effects of International Food Price Shocks on Political Institutions in Low-Income Countries: Evidence from an International Food Net-Export Price Index

World Development Vol. 61, pp. 142–153, 2014 0305-750X/Ó 2014 Elsevier Ltd. All rights reserved. www.elsevier.com/locate/worlddev

http://dx.doi.org/10.1016/j.worlddev.2014.04.009

Effects of International Food Price Shocks on Political Institutions in Low-Income Countries: Evidence from an International Food Net-Export Price Index RABAH AREZKI a,c and MARKUS BRUECKNER b,* a International Monetary Fund, Washington, USA b National University of Singapore, Singapore c Brookings Institution, Washington DC, USA Summary. — We examine the effects of variations in the international food prices on political institutions in low-income countries. Our empirical analysis exploits that the economic impact of changes in the international food prices differs across countries depending on whether countries are net food importers or exporters. We construct an international food net-export price index that captures this heterogeneity. Our panel fixed effects analysis yields that in low-income countries within-country variations in the international food net-export price index are significantly negatively related to democratic institutions. We further explore the mechanisms driving this relationship along the economic and the conflict dimensions. Ó 2014 Elsevier Ltd. All rights reserved. Key words — food prices, polity, low-income countries

1. INTRODUCTION

Our main finding is a significant negative relationship between within-country variations in the international food net-export price index and democratic institutions in lowincome countries: a one standard deviation increase in the international food net-export price index significantly reduced low-income countries’ Polity2 score by over 0.05 units. This effect is equal to about 0.03 standard deviations of the within-country change in low-income countries’ Polity2 score during the sample period, which spans nearly four decades. We document that this finding is robust to different measures of democracy, time periods, and estimation techniques as well as the control for country-specific weather shocks. A telling country episode that fits the pattern documented by our regressions is Fiji during the sugar price boom of 2004–06. Sugar is Fiji’s largest commodity export, comprising more than 20% of total exports. The international price of sugar doubled over the 2004–06 period. In 2006 the military staged a coup, seizing executive authority from the democratically elected prime minister. 2 Fiji’s Polity2 score declined from 6 to 3 and remained there for the following years. Other country examples that fit this pattern are Thailand and Bangladesh; both of which are exporters of rice. Following the over 40% increase in the international rice price during 2004–06, the 2006 military coup in Thailand brought to an end over 15 years of free and fair elections. In Bangladesh, attempts at democratic rule after independence in 1971 were shattered when the military assumed firm control of the state in 1975, following the more than doubling of the international rice price during 1972–1974. 3 Our paper is most closely related to the literature on the effects that economic shocks have on political institutions.

It is often claimed by policy makers and the media that increases in international food prices put at stake the sociopolitical stability of the world’s poorest countries. Former World Bank’s President Zoellick for example claimed at the joint World-Bank International Monetary Fund (IMF) 2008 spring meeting that a drastic increase in food prices could mean “seven lost years” in the fight against worldwide poverty. At the same meeting former IMF’s managing director StraussKahn expressed that “. . . the consequences [of food price increases] on the population in a large set of countries . . . will be terrible . . . disruptions may occur in the economic environment . . . so that at the end of the day most governments, having done well during the last 5 or 10 years, will see what they have done totally destroyed, and their legitimacy facing the population destroyed also.” 1 The question of whether and to what extent variations in the international food prices affect the political institutions of the world’s poorest countries is therefore of high policy relevance. Yet, little formal empirical evidence exists on the link between political institutions in poor countries and international food price shocks. In order to advance our understanding of the link between international food prices and political institutions in poor countries, this paper explores empirically in a panel of over 60 low-income countries the within-country relationship between political institutions and a country-specific international food net-export price index that is driven by variations in the international food prices. The economic impact of variations in the international food prices varies across poor countries depending on whether these countries are net food importers or exporters. To capture this heterogeneity, we construct a country-specific food net-export price index that uses variations in the international food prices multiplied by countries’ food net-export shares. Because this index is country specific, we are able to apply in our empirical analysis rigorous panel data estimation techniques that account for both country and time fixed effects.

* We thank Caroline Silverman for excellent research assistance and Camelia Minoiu for providing us with her dataset on Gini coefficients. The views in this paper are those of the author(s) alone and do not necessarily represent those of the International Monetary Fund (IMF) or IMF policy. Final revision accepted: April 15, 2014. 142

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES

Acemoglu and Robinson (2001, 2006) develop a formal theory of democratization where temporary income shocks can give rise to a “democratic window of opportunity.” In their theory, temporary income shocks induce a change in political institutions (which are viewed as durable) because these shocks change de facto political power. Changes in political institutions thus occur because of changes in the temporary bargaining power of citizens for institutional change and the costs faced by the elite to extend the franchise. An important implication of the theory is that economic shocks which increase the threat of a revolution by the citizens must not necessarily lead to democratization: if the economic costs of democracy are large for the elite then the elite may well be willing to bear the costs of fighting the revolution. Motivated by the window-of-opportunity theory, we explore the mechanism through which changes in the international food net-export price index affect countries’ polity scores along the economic and the conflict dimensions. Along the economic dimension, our empirical analysis shows that within-country changes in the international food net-export price index are significantly positively related to growth in Gross Domestic Product (GDP) per capita and investment (the terms-of-trade effect). Despite these increases being associated with higher economic surplus, our results suggest that democratic institutions deteriorated during periods of increases in the international food price index. This, in turn, suggests that, beyond the increase in average income, increases in the international food net-export price index had important distributive consequences in low-income countries. Our empirical analysis shows that within-country changes in the international food net-export price index are significantly negatively correlated with private consumption and significantly positively correlated with income inequality. Thus, consumption in low-income countries decreased during times of increases in the international food net-export price index and there was also a widening in the gap between rich and poor. Beyond the economic dimension, we provide additional insights into the mechanism driving the significant negative relationship between variations in the international food netexport price index and low-income countries’ political institutions by exploring the conflict dimension. Using a variety of conflict indicators, we find that changes in the international food net-export price index are significantly positively correlated with intra-state conflict. Increases in the international food net-export price index lead to significant increases in the incidence of anti-government demonstrations, riots, and civil conflict. Quantitatively, the estimated effects are also sizable: for example, with respect to civil conflict incidence, we find that a one standard deviation increase in low-income countries’ international food net-export price index is, on average, associated with a 0.5 percentage point higher incidence of civil conflict. Hence, our empirical analysis yields that there is a significant positive link to indicators of intra-state conflict (i.e., the risk of revolution). A straightforward reading of our findings is that by widening the gap between rich and poor and by raising aggregate income, higher international food prices made it particularly costly for the elite to grant de-jure political rights in poor autocratic countries that are net exporters of food. The increase in economic surplus, that occurred during food price spikes, was then guarded by the elite from the people by reducing de-jure political rights. This, in turn, was associated with an increased risk of revolution. In the next section we expand the discussion of how our paper relates to the existing literature, and we also provide further intuition for our findings. The remainder of our paper is

143

then organized as follows. In Section 3 we describe the data. In Section 4 we discuss our empirical results in more detail. In Section 5 we present additional robustness checks. In Section 6 we conclude. 2. RELATED LITERATURE On the empirical front, a number of recent papers have examined the link between exogenous income shocks and democratization. Burke and Leigh (2010) and Brueckner and Ciccone (2011) find that variations in year-to-year rainfall are significantly negatively correlated with democracy (the later authors show that this is particularly true in sub-Saharan African countries where the agricultural sector is large). Because year-to-year rainfall is a transitory shock to output the evidence provided in these papers supports the Acemoglu and Robinson window-of-opportunity theory. Our present paper seeks to complement the empirical studies by Burke and Leigh (2010) and Brueckner and Ciccone (2011). Variations in the international food prices are also of transitory nature and for the majority of low-income countries they constitute an exogenous economic shock. 4 In contrast to the literature on the democratic window of opportunity effect, another strand of the democratization literature has argued that income is positively related to democracy. Arguments for why income may be positively related to democracy go back at least as far as Lipset’s (1959) modernization hypothesis. Acemoglu, Johnson, Robinson, and Yared (2008, 2009) recently demonstrated that once country fixed effects are accounted for the relationship between income and democracy vanishes. In light of these authors’ findings, it is important to note that advocates of the modernization hypothesis are mainly concerned with how income is related to democracy in the long run. Thus, examining the modernization hypothesis based on within-country variation requires using time-series variation in income per capita that is of permanent nature. In this vein, Brueckner, Ciccone, and Tesei (2012) use variations in the international oil price. They find that these types of permanent income shocks are positively related with democracy. An important message from the above studies is that when examining the within-country relationship between income and democracy, it is important to distinguish between transitory and permanent shocks. The present paper’s results are based on using variations in the international food prices, which are predominantly of transitory nature. Therefore, the results in this paper should be compared with the results of papers on the relationship between transitory income shocks and democratic change (i.e., Burke & Leigh, 2010; Brueckner & Ciccone, 2011). Our findings on the conflict dimension make our paper also related to the literature on the determinants of state fragility (see, for example, World Bank, 2003, and references therein). Our paper contributes to that literature by focusing on the effects that variations in the international food net-export price index have on civil conflict—a focus that is unique, as no paper has examined yet for food commodities the effects that variations in these international prices have on civil conflict. 5 In addition to shedding novel light on the question of how international food price variations affect the likelihood of civil conflict in the world’s poorest countries, the paper also examines more minor forms of intra-state instability, such as, anti-government demonstrations and riots, which are of considerable interest in and of themselves from a political economy point of view.

144

WORLD DEVELOPMENT

The paper’s results also speak to the literature on the resource curse. This literature has been concerned with potentially adverse effects that mineral and fuel wealth have on economic growth, institutions, and civil conflict. 6 Our results yield that transitory revenue windfalls which accrue to food exporting countries during times of international food price booms have adverse effects on political institutions and increase the risk of intra-state conflict. Hence, we find that even for food commodities there exists an adverse relationship between increases in international prices and political institutions and conflict, although with regard to income the relationship is significantly positive. More narrowly, our paper contributes to the literature on food security. One of the issues that this literature is well aware of but struggling with, is that actual food production is endogenous to civil conflict and democratic change: country examples are indicative that the presence of civil war may be associated with an increase of domestic food prices. For example, in Darfur prices of the main food staples increased rapidly after widespread violence started in late 2003/early 2004 (see e.g., Brinkman & Hendrix, 2010). If this is indeed systematically the case across country-periods, then using domestic food prices to estimate the effect that higher food prices have on civil war will be complicated by a positive simultaneity bias. Our paper seeks to make a contribution to this literature on food insecurity by using variations in the international food prices which, for most low-income countries, are a plausibly exogenous source of variation in food expenditures. The focus in this paper is on low-income countries. And there are two main reasons for this. First, a large and active policy debate exists on the effects that food price variations have on state stability in the world’s poorest countries. The focus on low-income countries allows us to speak to this debate and address a question that is of high policy relevance. Second, only for countries that are price takers on the world food market does the correlation between international food prices and political institutions and conflict reflect a causal effect of food prices on these variables. The GDP of each low-income country is only a small fraction of world GDP. Thus, due to demand effects that arise from political change and intra-state conflict in low-income countries international food prices will be unaffected. Moreover, it is straightforward to exclude the handful of low-income countries that are potentially large food exporting countries. Therefore, focus on lowincome countries makes it credible that the correlation between variations in the international food net-export price index and political institutions reflects a causal effect of international food price shocks. 3. DATA (a) International food net-export price index The country-specific international food net-export price index that we use in our baseline analysis is constructed as: Y FoodPIi ; t ¼ FoodPricec ; thi;c c2C

where FoodPricec,t is the international price of food commodity c in year t, and hi,c is the average (time-invariant) share of net-exports of food commodity c in the GDP of country i during the period 1970–2007 period. 7 We use the average rather than the year-to-year variation in the net export share to reduce concerns that food exports and imports as a share of GDP are endogenous to year-to-year changes in countries’

socio-political environment. As a robustness check we will also present estimates based on two alternative indices: one that is constructed using the 1970 values of countries’ food net-export GDP shares and another that is constructed using the period t  1 time-varying values of food net-export GDP shares. We obtain data on annual international food prices for the 1970–2007 period from United Nations Conference on Trade and Development (UNCTAD) Commodity Statistics. Our data on the value of net food exports are from the National Bureau of Economic Research (NBER)-United Nations Trade Database (Feenstra, Lipsey, Deng, Ma, & Mo, 2004). We include as many food commodities in the index as possible, given the availability of NBER-UN data on imports and exports and UNCTAD data on international food prices. The food commodities included in our food price index are therefore beef, maize, rice, sugar, and wheat. In case there were multiple prices listed for the same commodity, we used a simple average of all the relevant prices. It is important to note that our index weights (exponentially) the international food prices by the country-specific food net-export GDP shares. This weighting scheme ensures that we take into account that for an exporter an increase in the price of the exported food commodity carries a positive terms of trade effect while for an importer it carries a negative terms of trade effect. The multiplicative exponential functional form of the international food net-export price index ensures that when we take the first difference of the logarithmic form of the price index, it is the percentage change in the international food prices that drives the variation in the index. Differences across food commodities that are due to differences in the measurement of the units of the commodities thus do not affect our results. In the robustness section, we will also present results that are based on a consumption-weighted food price index. The consumption-weighted food price index is constructed in exactly the same way as our food export price index with the only difference that hi,c represents the country-specific average expenditure share on food (these data are from FAO, 2010). (b) Political institutions Our main measure of political institutions is the Polity2 score of the Polity IV database (Marshall & Jaggers, 2010). The Polity2 score ranges from 10 to +10, with higher values indicating more democratic institutions. To examine also specifically the political competition and executive constraints channel, we use the polity sub-scores on constraints on the chief executive and political competition. According to the Polity IV project, constraints on the executive is a measure of the extent of institutionalized constraints on the decision making powers of chief executives and ranges from 1 to 7, with greater values indicating tighter constraints. Political competition measures the extent to which alternative preferences for policy and leadership can be pursued in the political arena. This indicator ranges from 1 to 10, with greater values denoting more competition. To explore the robustness of our results with respect to the democracy measure used we will also present estimates that use the political rights score from Freedom House (2010). This score ranges from 1 to 7. We rescale this score so that higher values denote more political rights. We will also present robustness checks that use as dependent variables the binary democracy indicators from Przeworski, Alvarez, Cheibub, and Limongi (2000) and Papaioannou and Siourounis (2008).

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES

(c) Intra-state conflict We use several indicators of intra-state conflict. The civil conflict incidence indicator variable is from the PRIO UPPSALA (2010a) database. According to PRIO UPPSALA (2010b, p. 1) civil conflict is a “contested incompatibility which concerns government and/or territory where the use of armed force between two parties, of which at least one is the government of a state, results in at least 25 battle deaths.” We also consider more minor forms of intra-state conflict such as riots and anti-government demonstrations, using data from Banks (2010) on the number of riots and anti-government demonstrations. (d) Macroeconomic variables To explore the economic dimension we use several macroeconomic indicators. Data on real GDP per capita, investment, private consumption, and government expenditures are from the Penn World Table, version 6.3 (Heston, Summers, & Aten, 2009). Data on income inequality as measured by the Gini coefficient are from the United Nations WIDER (2008) database. Descriptive statistics on the above variables are provided in Table 1. 4. BASELINE ESTIMATES Table 2 presents our baseline estimates of the average marginal effect that changes in the international food net-export price index have on changes in the Polity2 score. The panel comprises 60 low-income countries and spans the period 1970–2007. 8 We focus on the post-1970 period in order to ensure that the countries in our sample are independent and not under the colonial reign of the European powers. The baseline sample ends in 2007 to ensure that our estimates are not driven by the Great Recession, an atypical event; however, we will present as a robustness check estimates that cover the 1970–2010 period. The baseline reduced-form model that we estimate is: DPolityit ¼ ai þ d t þ bDlnðFoodPIit Þ þ zit

ð1Þ

We use the change in the log of the international food netexport price index in order to ensure that within-country

Table 1. Descriptive Statistics Variable

Mean

Polity2 1.92 DPolity2 0.17 GDP per capita 2,481 Dln(GDP per capita) 0.01 Cons/GDP 0.74 Dln(Cons) 0.01 Gov./GDP 0.19 Dln(Gov) 0.01 Inv/GDP 0.15 Dln(Inv) 0.01 Income Gini 0.43 No. of Riots 0.54 DNo. of Riots 0.002 No. of anti-government demonstrations 0.42 DNo. of anti-government demonstrations 0.007 Civil conflict incidence 0.27 DCivil conflict incidence 0.003

Std. Dev. Obs. 6.02 2.07 1,739 0.07 0.16 0.09 0.12 0.17 0.11 0.26 0.10 1.43 1.65 1.75 1.44 0.44 0.29

2,215 2,153 2,257 2,197 2,257 2,197 2,257 2,197 2,257 2,197 423 2,218 2,156 2,218 2,156 2,280 2,220

145

changes in the index are due to percentage changes in the international food prices. This has the advantage that differences across food commodities that are due to differences in the measurement of the units of the commodities do not affect our results. The error term zit is clustered at the country level so that it may be arbitrarily serially correlated. Column (1) of Table 2 presents estimates where the change in the Polity2 score is related to the change in the logarithm of the international food net-export price index, without controlling for country or year fixed effects. Column (2) adds to this regression country fixed effects and column (3) adds year fixed effects. The main result is a negative and highly significant effect of variations in the international food net-export price index on the Polity2 score. Quantitatively, the estimated coefficient implies that, on average, a one standard deviation (0.003) increase in the international food net-export price index significantly reduced the Polity2 score by over 0.05 units. This effect is equal to about 0.03 standard deviations of the change in the Polity2 score during the sample period. In column (4) we add lags of the international food netexport price index to check whether there are significant lagged effects on countries’ Polity2 scores. The estimated coefficients on the lagged index are negative, thus indicating that even after several years an increase in the international food net-export price index had a negative effect on the Polity2 score. However, statistically these lagged effects are not significant at conventional significance levels. Hence, we note that the main negative effect on the Polity2 score from international food price shocks is observed without any lags. To take into account dynamics in countries’ Polity2 scores, we show in columns (5) and (6) dynamic panel estimates that include on the right-hand side of the estimating equation the lagged level of the Polity2 score. Both the least squares and system-Generalized Method of Moments (GMM) estimates show that there is quite a bit of persistence in the dynamics of the Polity2 score. 9 The estimated convergence coefficient is 0.13 and implies a half-life of shocks to the level of the Polity2 score of around 4 years. More importantly, the dynamic panel data estimates confirm that there is a significant negative impact effect of variations in the international food net-export price index on the Polity2 score. In Table 3, we document that our finding of a negative relationship between the international food net-export price index and political institutions is robust to the use of alternative Polity IV democracy measures. In column (1), we present results when using the Polity2 score that excludes periods of interregnum and transition. 10 Columns (2) and (3) present results for the polity sub-scores on executive constraints and political competition, and columns (4) and (5) present the results for the democracy and autocracy score. 11 We find that regardless of which Polity IV measure is used our main result survives: increases in the international food net-export price index are associated with a significant deterioration of political institutions in low-income countries. Panel A of Table 3 shows that this result holds for the baseline 1970–2007 sample, and Panel B of Table 3 shows that it also holds in the 1970–2010 sample. In Table 4 we present results that use measures of political institutions from alternative data sources. Column (1) reports estimates where the dependent variable is the political rights score from Freedom House (2010). The estimated coefficient on the international food net-export price index is negative in sign and statistically significant at the 1% significance level. Quantitatively, the estimate implies that a one standard deviation increase in low-income countries’ international food net-export price index is associated with a significant decrease in these countries’ political rights scores of around 0.1

146

WORLD DEVELOPMENT Table 2. Food Prices and Political Institutions DPolity2

DlnFoodPI, t

(1) LS

(2) LS

(3) LS

(4) LS

(5) LS

(6) SYS-GMM

17.62*** (3.51)

17.78*** (3.63)

16.34*** (3.43)

16.781*** (4.74) 4.46 (1.09) 4.39 (0.82)

12.43*** (3.75)

16.24*** (4.25)

0.14*** (7.18) Yes Yes 2,153

0.13*** (4.29) Yes Yes 2,153

DlnFoodPI, t  1 DlnFoodPI, t  2 Polity2, t  1 Country Fe Year Fe Observations

No No 2,153

Yes No 2,153

Yes Yes 2,153

Yes Yes 2,101

Note: The method of estimation in columns (1)–(5) is least squares; column (6) system-GMM. The t-values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable is the change in the Polity2 score. The explanatory variable is the change in the log of the international food net-export price index. *** Significantly different from zero at the 1% significance level.

Table 3. Food Prices and Political Institutions (Alternative Polity IV Democracy Measures) DPolity (1)

DExconst (2)

DPolcomp (3)

DDemoc (4)

DAutoc (5)

Panel A: 1970–2007 period DlnFoodPI 19.44*** (7.25) Country Fe Yes Year Fe Yes Observations 1,998

10.31*** (10.60) Yes Yes 1,998

7.81*** (6.59) Yes Yes 1,998

8.47*** (6.46) Yes Yes 1,998

10.97*** (7.62) Yes Yes 1,998

Panel B: 1970–2010 period DlnFoodPI 19.44*** (7.25) Country Fe Yes Year Fe Yes Observations 2,162

10.31*** (10.60) Yes Yes 2,162

7.81*** (6.59) Yes Yes 2,162

8.47*** (6.46) Yes Yes 2,162

10.97*** (7.62) Yes Yes 2,162

Note: The method of estimation is least squares. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity score; column (2) the change in the executive constraints score; column (3) the change in the political competition score; column (4) the change in the democracy score; column (5) the change in the autocracy score. All scores exclude values that are recorded as 66, 77, and 88. The explanatory variable is the change in the log of the international food net-export price index. *** Significantly different from zero at the 1% significance level.

standard deviations. Columns (2) and (3) report results that use the Przeworski et al. (2000) and Papaioannou and Siourounis (2008) democracy measures. Again the coefficient on the international food net-export price index is negative and statistically significant. Quantitatively, the estimated coefficient implies that a one standard deviation increase in lowincome countries’ international food net-export price index is associated with a decrease in the incidence of democracy of about half a percentage point. As a first step to explain the negative relationship between variations in the international food net-export price index and political institutions, we document in Table 5 the effects that variations in this index have on macroeconomic variables. In column (1), we show that increases in the international food net-export price index lead to significant increases in GDP per capita. Primarily, this positive relationship is the consequence of the terms of trade effect. The estimated coefficient implies that, on average, a one standard deviation increase in the international food net-export price index induced an increase in GDP per capita of around 0.3%.

Given the significant positive link with GDP per capita, our finding of variations in the international food net-export price index being significantly negatively correlated with democratic institutions can be interpreted as resonating the findings in Burke and Leigh (2010) and Brueckner and Ciccone (2011). Indeed, if we use the international food net-export price index as an instrument for GDP per capita, then our IV estimates are of similar magnitude as those reported in Burke and Leigh (2010) and Brueckner and Ciccone (2011). Appendix Table 12 shows these estimates. An instrumental variables estimate yields that a 1% transitory drop in GDP per capita due to a decrease in the international food net-export price index increases the Polity2 score by around 0.2 units (0.1 standard deviations). 12 In column (2) of Table 5 we show that there is a significant positive effect of variations in the international food netexport price index on investment and in column (3) we show that this effect is positive as well for real per capita government consumption expenditures. However, column (4) of Table 5 shows that private consumption decreased significantly. The

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES

147

Table 4. Food Prices and Political Institutions (Democracy Measures: Freedom House, Przeworski et al., and Papaioannou and Siourounis)

DlnFoodPI Country Fe Year Fe Observations

Freedom house (1)

Przeworski et al. (2)

Papaioannou and Siourounis (3)

19.72*** (14.38) Yes Yes 1,922

1.56*** (3.08) Yes Yes 1,817

1.57*** (2.06) Yes Yes 1,879

Note: The method of estimation is least squares. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Freedom House political rights score. The score is rescaled so that higher values denote stronger political rights. In column (2) the dependent variable is the democracy indicator of Przeworski et al. (2000). In column (3) the dependent variable is the democracy indicator of Papaioannou and Siourounis. The explanatory variable is the change in the log of the international food net-export price index. *** Significantly different from zero at the 1% significance level.

Table 5. Food Prices and Macroeconomic Outcomes

DlnFoodPI Country Fe Year Fe Observations

DlnGDP (1)

DlnInv (2)

DlnGov (3)

DlnCons (4)

DGini (5)

1.04*** (5.08) Yes Yes 2,197

2.30** (2.53) Yes Yes 2,197

1.20*** (4.37) Yes Yes 2,197

0.80** (3.31) Yes Yes 2,197

0.228* (1.81) Yes Yes 423

Note: The method of estimation is least squares. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the log of real per capita GDP; column (2) the change in the log of real per capita investment; column (3) the change in the log of real per capita government expenditures; column (4) the change in the log of real per capita private consumption; column (5) the change in the Gini coefficient. The explanatory variable is the change in the log of the international food net-export price index. * Significantly different from zero at the 10% significance level. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

estimate implies that a one standard deviation increase in the international food net-export price index induced a decrease in real consumption per capita of about 0.2%. This significant decrease in private consumption indicates that beyond the positive effect on GDP per capita, variations in the international food net-export price index carry significant compositional effects. 13 To explore this channel further, we report in column (5) of Table 5 estimates of the effect that variations in the international food net-export price index have on low-income countries’ Gini coefficient. We find that increases in the international food net-export price index are significantly positively related to the Gini coefficient: a one standard deviation increase in the international food net-export price index induced a significant increase in the Gini coefficient of about 0.07 percentage points. This suggests that variations in the international food net-export price index induced also significant within-country variations in the income gap between rich and poor. In Table 6, we examine the effects that variations in the international food net-export price index have on measures of intra-state conflict. Column (1) reports effects on the number of anti-government demonstrations and column (2) reports effects on the number of riots. Our main finding is that increases in the international food net-export price index significantly increase both the number of anti-government demonstrations and the number of riots. Columns (3) and (4) show that there is also a significant positive effect of variations in the international food net-export price index on the incidence of civil conflict. 14

5. ROBUSTNESS CHECKS We report in this section a number of additional robustness checks. In Table 7 we show that there continues to be a significant effect when we exclude the handful of low-income countries that are large food suppliers to the world food market. 15 The identifying assumption in the previous tables was that low-income countries are price takers on the international food market. This seems a plausible assumption for the majority of the low-income countries, as these countries produce and consume individually only a very small fraction of world food production. Thus, the fact that there continues to be a highly significant effect of changes in the international food net-export price index when excluding potentially large food producing countries from the low-income countries sample is reassuring that our baseline assumption of international food prices being exogenous is reasonable. 16 Another interesting issue that we are able to examine with our panel data is whether the relationship has changed over time. One particular reason for this could be the end of the Cold War (see also Fearon & Laitin, 2003). In Table 8 we report estimates from an interaction model where the marginal effect of the international food net-export price index is allowed to differ for the preand the post-1990 period. Our main findings from these interaction regressions is that: (i) for the pre-1990 period, there is a significant negative effect of the international food net-export price index on political institutions and there is also evidence of a significant positive effect on intra-state conflict; (ii) there is some evidence that for the post-1990 period the effects on intra-state conflict have become significantly larger (see e.g., columns (2) or (5)).

148

WORLD DEVELOPMENT Table 6. Food Prices and Intra-State Conflict Civil conflict

DDemonstrations

DRiots

(1) LS

(2) LS

(3) LS

(4) Logit FE

3.92*** (2.70) Yes Yes 2,156

4.56*** (3.98) Yes Yes 2,156

1.67** (2.01) Yes Yes 2,280

229.68* (2.44) Yes Yes 2,280

DlnFoodPI Country Fe Year Fe Observations

Note: The method of estimation in columns (1)–(3) is least squares; column (4) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the number of anti-government demonstrations; column (2) the change in the number of riots; columns (3) and (4) an indicator variable that is equal to unity if the country experienced a civil conflict. The explanatory variable is the change in the log of the international food net-export price index. * Significantly different from zero at the 10% significance level. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

Table 7. Robustness Check I: Excluding Large Food Producers

DlnFoodPI Country Fe Year Fe Observations

Civil conflict

DPolity

DDemonstrations

DRiots

(1) LS

(2) LS

(3) LS

(4) LS

(5) Logit FE

18.88*** (7.31) Yes Yes 1,812

4.35*** (3.69) Yes Yes 1,814

5.39*** (6.02) Yes Yes 1,814

1.52** (2.43) Yes Yes 1,900

229.18** (1.98) Yes Yes 1,900

Note: The method of estimation in columns (1)–(4) is least squares; column (5) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity2 score; column (2) the change in the number of anti-government demonstrations; column (3) the change in the number of riots; columns (4) and (5) an indicator variable that is equal to unity if the country experienced a civil conflict. The explanatory variable is the change in the log of the international food net-export price index. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

Table 8. Robustness Check II: Are the Post-1990s Different?

DlnFoodPI DlnFoodPI * Post 1990s Country Fe Year Fe Countries

Civil conflict

DPolity

DDemonstrations

DRiots

(1) LS

(2) LS

(3) LS

(4) LS

(5) Logit FE

19.77*** (3.59) 23.38 (1.11) Yes Yes 2,153

2.20 (1.48) 11.70*** (2.83) Yes Yes 2,156

4.95*** (3.78) 2.69 (0.61) Yes Yes 2,156

1.15** (2.13) 3.67 (0.98) Yes Yes 2,280

148.54 (1.39) 457.02** (1.96) Yes Yes 2,280

Note: The method of estimation in columns (1)–(4) is least squares; column (5) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity2 score; column (2) the change in the number of anti-government demonstrations; column (3) the change in the number of riots; columns (4) and (5) an indicator variable that is equal to unity if the country experienced a civil conflict. The explanatory variable is the change in the log of the international food net-export price index. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

In Table 9 we report estimates that use a consumptionweighted food price index. This index is constructed in exactly the same way as our international food net-export price index (see Section 3). The main result is that variations in the consumption-weighted food price index do not significantly correlate with the polity score; nor do they significantly correlate with measures of intra-state conflict. On the other hand,

the effects of variations in the international food net-export price index continue to be quantitatively large and statistically significant. The most straightforward interpretation of this finding is that what matters in the group of low-income countries (where the average and median share of consumption expenditures on food already exceeds 50%) is countries’ food net-export status.

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES

149

Table 9. Robustness Check III: Effects of a Consumption-Weighted Food Price Index

DlnFoodPI DlnFoodPI ExpenditureWeighted Country Fe Year Fe Countries

Civil conflict

DPolity

DDemonstrations

DRiots

(1) LS

(2) LS

(3) LS

(4) LS

(5) Logit FE

16.45*** (3.54) 0.23 (0.26) Yes Yes 2,153

3.78*** (2.71) 0.30 (1.26) Yes Yes 2,156

4.57*** (4.21) 0.01 (0.04) Yes Yes 2,156

1.58** (2.01) 0.12 (1.30) Yes Yes 2,220

220.26** (2.28) 0.52 (0.41) Yes Yes 2,220

Note: The method of estimation in columns (1)–(4) is least squares; column (5) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity2 score; column (2) the change in the number of anti-government demonstrations; column (3) the change in the number of riots; columns (4) and (5) an indicator variable that is equal to unity if the country experienced a civil conflict. DlnFoodPI refers to the change in the log of the international food export price index. DlnFoodPI ExpenditureWeighted refers to the change in the log of a food price index that is constructed by exponentially weighting the international food prices with the consumption expenditure share on food. See Section 3 in the manuscript for further details. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

Table 10. Robustness Check IV: Controlling for Weather Conditions

DlnFoodPI Country Fe Year Fe Rainfall and temperature controls Countries

Civil conflict

DPolity

DDemonstrations

DRiots

(1) LS

(2) LS

(3) LS

(4) LS

(5) Logit FE

17.16*** (3.78) Yes Yes Yes 2,153

3.97*** (2.80) Yes Yes Yes 2,156

4.68*** (3.88) Yes Yes Yes 2,156

1.58** (2.18) Yes Yes Yes 2,220

210.26** (2.23) Yes Yes Yes 2,220

Note: The method of estimation in columns (1)–(4) is least squares; column (5) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity2 score; column (2) the change in the number of anti-government demonstrations; column (3) the change in the number of riots; columns (4) and (5) an indicator variable that is equal to unity if the country experienced a civil conflict. DlnFoodPI refers to the change in the log of the international food export price index. See Section 3 in the manuscript for further details. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

Table 11. Robustness Check V: Effects in Middle- and High-Income Countries Civil conflict

DPolity

DDemonstrations

DRiots

(1)

(2)

(3)

(4)

(5)

11.34 (1.42) Yes Yes 910

2.99 (0.67) Yes Yes 910

2.21 (1.27) Yes Yes 988

228.10 (1.20) Yes Yes 988

Panel A: Middle-income countries DlnFoodPI 28.33*** (4.42) Country Fe Yes Year Fe Yes Observations 845 Panel B: High-income countries DlnFoodPI Country Fe Year Fe Observations

(1)

(2)

(3)

(4)

(5)

37.63 (1.15) Yes Yes 951

1.16 (0.02) Yes Yes 1,000

13.09 (0.47) Yes Yes 1,000

2.06 (1.17) Yes Yes 1,178

688.86 (0.91) Yes Yes 1,178

Note: The method of estimation in columns (1)–(4) is least squares; column (5) maximum likelihood. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity2 score; column (2) the change in the number of anti-government demonstrations; column (3) the change in the number of riots; columns (4) and (5) an indicator variable that is equal to unity if the country experienced a civil conflict. The explanatory variable is the change in the log of the international food net-export price index. *** Significantly different from zero at the 1% significance level.

150

WORLD DEVELOPMENT

Recent work by Hsiang, Meng, and Cane (2011) has shown a strong link between El Nino and civil conflict. In order to outrule that our estimates are picking up the effects that countryspecific climate change may have on international food prices, we report in Table 10 estimates that control for within-country changes in rainfall and temperature. The main result is that there continues to be a strong and highly significant effect of the international food net-export price index on low-income countries’ Polity2 score and indicators of intra-state conflict when controlling for country-specific weather conditions. To complete the picture, we report in Table 11 estimates for the sample of middle- and high-income countries. The main finding is that variations in the international food net-export price index are not significantly correlated with the incidence of intra-state conflict. This is true for the middle-income group as well as for the high-income group. With respect to political institutions, variations in the international food export price index are negatively correlated with the Polity2 score in the middle-income group, but not in the high-income group. Given that the assumption of price takership for the middle and, in particular, for the high-income countries is more questionable we believe that the causal direction of these correlations must be interpreted with some caution. Moreover, low-income countries feature a significantly higher incidence of intra-state conflicts and lower Polity2 scores than middleand high-income countries (Acemoglu, Johnson, Robinson, & Yared, 2008; World Bank, 2003). 6. CONCLUSION This paper examined effects of international food price shocks on indicators of political institutions in low-income countries. It exploited that the economic impact differs across low-income countries depending on whether these countries are net food importers or exporters. Using an international food net-export price index the paper’s main finding was that in low-income countries, which are net food exporters, political institutions significantly deteriorated during times of international food price increases. The paper documented that this finding is robust to alternative model specification and democracy data; it also explored mechanisms behind this significant negative relationship along the economic and the conflict dimensions. The finding of a significant negative relationship between the international food net-export price index and political institutions supports models which predict transitory economic shocks to create a window of opportunity for democratic change. From a macroeconomic perspective, it is worthwhile to restate that international food price increases induced in the net food exporting countries a significant increase in real per capita GDP and real per capita investment (the “terms of trade effect”). At the same time, international food price increases induced a significant decrease in real per capita consumption and a significant increase in income inequality. Thus, increases in the international food prices

had real macroeconomic effects that went beyond average per capita income: they were associated with a significant decrease in consumption and a significant increase in the gap between rich and poor. All in all, our empirical results are broadly consistent with the often made claim by policy makers and the press that food price increases put at stake the socio-economic and political stability of the world’s poorest countries. Arguably a large share of the variation in the international food prices is due to changes in the demand and supply of the High- and Middle-Income Countries. A natural question beyond the scope of this paper is what can and should be done by the developed world and international organizations in response to drastic increases in international food prices. In response to the episode of soaring food prices which started at the end of 2007, the World Bank and the International Monetary Fund have put in place emergency loans facilities in order to help Low-Income Countries. In the longer run, Low-Income Countries will need to increase their domestic production of food to insure food security through modernizing their agriculture sector. While High and Middle-Income countries have not been subject to social instability compared to Low-Income Countries, they did not remain insensitive to the recent soaring food prices. Indeed, High and Middle Income have shown rising interest in acquiring pasture land in a number of Low-Income Countries including in the African continent in order to insure their own food security (Arezki, Deininger, & Selod, in press). Despite the fact that foreign investment is often seen as a source of productivity gains for the recipient countries, in the particular circumstances some commentators have raised the fear of a “global land grab” given the lack of transparency surrounding a number of large scale land investment deals. Those suspicions have in turn led to social unrest including Madagascar where it prompted a coup. It thus remains to be seen how those potentially large foreign investment in pasture land in Low-Income Countries will play out in the long run. As an avenue for future research, it may be of interest to explore how the effects of international food prices on political institutions depend on factors that go beyond countries’ netexport structures such as insulation policies and degree of market integration. Indeed the pass-through of international food prices on domestic food prices is likely to vary across countries depending on cross-country differences in insulation policies and market integration (see, for example, Martin & Anderson, 2012, or Anderson & Nelgen, 2012). While an indepth study of the roles that policies and market integration play in the context of international food price shocks and political institutions is beyond the scope of the current paper, future research may build on our paper by exploring how, and to what extent, the relationship between the international food net-export price index and political institutions differs depending on cross-country differences in insulation policies and market integration.

NOTES 1. http://edition.cnn.com/2008/WORLD/americas/04/14/ world.food.crisis/. 2. A mechanism that might explain this outcome is that the increase in the sugar price increased the costs of democracy for the elite. By seizing political control, the elite increased its chances of capturing rents associated with the higher sugar prices.

3. More recently, the so-called food and fuel crisis of 2008 has had devastating consequences on low-income countries. Renewed surges in food prices have plunged many people into poverty, including the Horn of Africa in 2011. Some commentators argued that food price surges have been an important contributing factor behind conflicts in Sub-Saharan Africa and the uprisings in the Middle East and North Africa (the Arab Spring). Several recent reports by the Food and Agriculture Organization

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES (FAO) document the link between recent food price surges and poverty and social unrest (e.g., FAO, 2008, 2011). Several recent academic papers also study the role of food price surges on conflicts and political stability in a specific region or country including in West Africa (Flores, 2004), Afghanistan (D’Souza & Jolliffe, 2013) and the Middle East and North Africa (Lagi, Bertrand, & Bar-Yam, 2011). 4. During the 1970–2009 period the AR(1) coefficient on the international maize price is 0.67; the international wheat price 0.61; the international rice price 0.65; the international sugar price 0.55; and the international beef price 0.76. The Dickey-Fuller test rejects the null hypothesis of a unit root in the international maize price (p-value 0.00); the international wheat price (p-value 0.09); the international rice price (pvalue 0.00); the international sugar price (p-value 0.00); and the international beef price (p-value 0.06). In contrast, for comparison purposes, the AR(1) coefficient on the international oil price during that period is 0.99 and the Dickey-Fuller test does not reject the null hypothesis of a unit root (p-value 0.99). 5. Besley and Persson (2008) and Brueckner and Ciccone (2010) have examined effects of an overall commodity price index on civil conflict; these authors did not focus or carry out a separate analysis for the effects of food prices.

151

Paraguay, Philippines, Rwanda, Senegal, Sierra Leone, Somalia, Sri Lanka, Sudan, Tanzania, Thailand, Togo, Tunisia, Uganda, Ukraine, Vietnam, Zambia, Zimbabwe. 9. In the system-GMM estimation we use the second and third lag in order to keep the number of instruments low. The p-value on the test for second order serial correlation is 0.58 and the p-value from the Hansen J test is 0.45. 10. These periods are coded in the executive constraints and political competition sub-scores as 77, 88, 66. The revised combined Polity2 score integrates these values by assigning interregnum periods the value of zero and linearly interpolating transition periods. 11. Both the democracy and autocracy score range between 0 and 10, with higher values indicating stronger democratic (autocratic) institutions. 12. We note that the instrumental variables estimates imply strong exclusion restrictions. As this paper is about the reduced-form relationship between variations in the international food net-export price index and political institutions we do not wish to emphasize these IV estimates. 13. An interpretation of the negative consumption response is that when the international food prices increase, the substitution effect outweighs the income effect in food exporting countries.

6. See Van der Ploeg (2011) for a review of this literature. 7. This functional form of the price index follows common practice in the literature. See, for example, Collier and Goderis (2007) and the references cited therein. 8. With low-income countries we refer to the group of countries that is identified by the World Bank (2010) as Low-Income Countries and Lower Middle-Income Countries. We use the largest possible sample given data on democracy and food exports. The countries in the low-income sample are: Angola, Bangladesh, Bolivia, Burkina Faso, Burundi, Cambodia, Cameroon, Central African Republic, Chad, China, Democratic Republic of Congo, Republic of Congo, Cote d’Ivoire, Djibouti, Ecuador, Egypt, El Salvador, Ethiopia, Gambia, Ghana, Guatemala, Guinea, Guinea-Bissau, Guyana, Haiti, Honduras, India, Indonesia, Iran, Jordan, Kenya, Madagascar, Malawi, Mali, Mauritania, Mongolia, Morocco, Mozambique, Nicaragua, Niger, Nigeria, Pakistan, Papua New Guinea,

14. Note that we present in column (4) conditional logit fixed effects estimates to take into account the binary nature of the dependent variable. Estimates from the conditional logit fixed effects regression do not represent marginal effects because this would require knowledge of the distribution of the country fixed effects (e.g., Wooldridge, 2002). 15. The excluded countries are China, Guatemala, India, Indonesia, Ivory Coast, Pakistan, Thailand, Uganda, Ukraine, and Vietnam. These low-income countries produce a significant (more than 3%) share of world food production and might therefore have an effect on world food prices. 16. The data on rainfall and temperature are from Matsuura and Willmott (2009).

REFERENCES Acemoglu, D., Johnson, S., Robinson, J., & Yared, P. (2008). Income and democracy. American Economic Review, 98, 808–842. Acemoglu, D., Johnson, S., Robinson, J., & Yared, P. (2009). Reevaluating the modernization hypothesis. Journal of Monetary Economics, 59, 1043–1058. Acemoglu, D., & Robinson, J. (2001). A theory of political transitions. American Economic Review, 91, 938–963. Acemoglu, D., & Robinson, J. (2006). Economic origins of dictatorship and democracy. New York: Cambridge University Press. Anderson, K., & Nelgen, S. (2012). Trade barrier volatility and agricultural price stabilization. World Development, 40, 36–48. Arezki, R., Deininger, K., & Selod, H. (in press). What drives the global land rush? World Bank Economic Review. Banks, A. (2010). Cross-national time-series data archive. Binghamton NY: Databanks International Data File. Besley, T., & Persson, T. (2008). The incidence of civil war: Theory and evidence. NBER working paper #14585. Brinkman, H., & Hendrix, C. (2010). Food insecurity and conflict: Applying the WDR framework. In World development report 2011.

Brueckner, M., & Ciccone, A. (2010). International commodity price shocks, growth, and the outbreak of civil war in sub-Saharan Africa. Economic Journal, 120, 519–534. Brueckner, M., & Ciccone, A. (2011). Rainfall and the democratic window of opportunity. Econometrica, 79, 923–947. Brueckner, M., Ciccone, A., & Tesei, A. (2012). Oil price shocks, income, and democracy. Review of Economics and Statistics, 94(2), 389–399. Burke, P., & Leigh, A. (2010). Do output contractions trigger democratic change?. American Economic Journal: Macroeconomics, 2, 124–157. Collier, P., & Goderis, B. (2007). Commodity prices, growth, and the natural resource curse: Reconciling a conundrum. CSAE working paper series no. 276. D’Souza, A., & Jolliffe, D. (2013). Conflict, food price shocks, and food insecurity: The experience of Afghan households. Food Policy, 42(C), 32–47. Food and Agriculture Organization, U.N. (2008). The state of food insecurity in the World 2008: High food prices and food security – Threats and opportunities. Food and Agriculture Organization, U.N. (2010). FAO statistics household survey database. [Online Data].

152

WORLD DEVELOPMENT

Food and Agriculture Organization, U.N. (2011). The state of food insecurity in the World 2011: How does international price volatility affect domestic economies and food security?. Fearon, J., & Laitin, D. (2003). Ethnicity, insurgency and civil war. American Political Science Review, 97(1), 75–90. Feenstra, R., Lipsey, R., Deng, H., Ma, A., & Mo, H. (2004). World trade flows: 1962–2000. NBER working paper no. 11040. Flores, M. (2004). Conflicts, rural development and food security in West Africa. Working papers 04-02. Agricultural and Development Economics Division of the Food and Agriculture Organization of the United Nations (FAO-ESA). Freedom House (2010). Freedom in the World country ratings. [Online Database]. Heston, A., Summers, R., & Aten, B. (2009). Penn World table version 6.3. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, August 2009. Hsiang, S., Meng, K., & Cane, M. (2011). Civil conflicts are associated with the global climate change. Nature, 476, 438–441. Lagi, M., Bertrand, K. Z., & Bar-Yam, Y. (2011). The food crises and political instability in North Africa and the Middle East. arXiv:1108.2455, August 10, 2011. Lipset, S. (1959). Some social prerequisites for democracy: Economic development and political legitimacy. American Political Science Review, 53, 69–105. Marshall, M., & Jaggers, K. (2010). Polity IV project: Dataset users’ manual. Center for Global Policy, George Mason University. Polity IV data computer file, version 2010. College Park, MD: Center for International Development and Conflict Management, University of Maryland. Available from www.cidcm.umd.edu/polity. Martin, W., & Anderson, K. (2012). Export restrictions and price insulation during commodity price booms. American Journal of Agricultural Economics, 94, 422–427.

Matsuura, K., & Willmott, C. (2009). Terrestrial air temperature and precipitation: 1900–2009 gridded monthly time series, version 2.01. University of Delaware. Available from http://climate.geog.udel.edu/ ~climate/. Papaioannou, E., & Siourounis, G. (2008). Economic and social factors driving the third wave of democratization. Journal of Comparative Economics, 36, 365–387. PRIO/UPPSALA (2010a). Armed conflict database. Available from http:// www.prio.no/Data/Armed-Conflict/UCDP-PRIO/. PRIO/UPPSALA (2010b). Armed conflict dataset codebook. Available from http://www.prio.no/Data/Armed-Conflict/UCDP-PRIO/. Przeworski, A., Alvarez, M., Cheibub, J., & Limongi, F. (2000). Democracy and development: Political institutions and the well-being of the World, 1950–1990. Cambridge, UK: Cambridge University Press. UNU-WIDER (2008). World income inequality database. Version 2.0c. [Online database]. Van der Ploeg, F. (2011). Natural resources: Curse or blessing?. Journal of Economic Literature, 49, 366–420. Wooldridge, J. (2002). Econometric analysis of cross section and panel data. Cambridge, Mass.: MIT Press. World Bank (2003). Breaking the conflict trap: Civil war and development policy. Oxford University Press. World Bank, 2010. World Development Indicators. Washington, D.C.: World Bank.

APPENDIX A See Tables 12 and 13.

Table 12. Food Prices and Political Institutions. (Food Price Index Based on 1970 Net-Export Shares; Time-Varying Net-Export Shares)

Panel A: 1970 net export shares DlnFoodPI [1970 net-export shares] Country Fe Year Fe Panel B: 1-year lagged time-varying net-export shares DlnFoodPI [1-year lagged time-varying net-export shares] Country Fe Year Fe

DPolity (1)

DExconst (2)

DPolcomp (3)

DDemoc (4)

DAutoc (5)

36.47*** (4.30) Yes Yes

21.42*** (3.57) Yes Yes

6.64** (2.05) Yes Yes

29.73*** (5.75) Yes Yes

4.63 (1.18) Yes Yes

7.27*** (3.20) Yes Yes

3.21** (2.11) Yes Yes

3.64*** (6.90) Yes Yes

2.91*** (2.81) Yes Yes

4.35*** (3.45) Yes Yes

Note: The method of estimation is least squares. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The dependent variable in column (1) is the change in the Polity score; column (2) the change in the executive constraints score; column (3) the change in the political competition score; column (4) the change in the democracy score; column (5) the change in the autocracy score. All scores exclude values that are recorded as 66, 77, and 88. The explanatory variable is the change in the log of the international food net-export price index. ** Significantly different from zero at the 5% significance level. *** Significantly different from zero at the 1% significance level.

EFFECTS OF INTERNATIONAL FOOD PRICE SHOCKS ON POLITICAL INSTITUTIONS IN LOW-INCOME COUNTRIES Table 13. GDP Shocks and Political Institutions. (IV Estimates)

DlnGDP First stage DlnGDP DlnFoodPI Kleibergen Paap F-stat Country Fe Year Fe Observations

DPolity (1) 2SLS

DExconst (2) 2SLS

DPolcomp (3) 2SLS

DDemoc (4) 2SLS

DAutoc (5) 2SLS

16.83*** (3.72)

8.93*** (2.82)

6.76*** (3.36)

7.34*** (3.74)

9.50** (2.60)

1.16*** (3.37) 11.37 Yes Yes 1,996

1.16*** (3.37) 11.37 Yes Yes 1,996

1.16*** (3.37) 11.37 Yes Yes 1,996

1.16*** (3.37) 11.37 Yes Yes 1,996

1.16*** (3.37) 11.37 Yes Yes 1,996

Note: The method of estimation is two-stage least squares. t-Values shown in parentheses are based on Huber robust standard errors that are clustered at the country level. The excluded instrument is the change in the log of the international food net-export price index. The dependent variable in column (1) is the change in the Polity score; column (2) the change in the executive constraints score; column (3) the change in the political competition score; column (4) the change in the democracy score; column (5) the change in the autocracy score. All scores exclude values that are recorded as 66, 77, and 88. * Significantly different from zero at 90% confidence. ** Significantly different from zero at 95% confidence. *** Significantly different from zero at 99% confidence.

Available online at www.sciencedirect.com

ScienceDirect

153