## _wp1162 - 2007. Our main finding is that in Low Income Countries increases in the international

## Source details

**Canonical URL:** [_wp1162 - 2007. Our main finding is that in Low Income Countries increases in the international](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1162.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1162.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1162.pdf.json)

---

### I. Introduction
- Research question: whether variations in international food prices affect intra-state stability and political institutions, especially in the world's poorest countries.
- Sample and period: panel of over 120 countries during the 1970-2007 period.
- Identification strategy: construct a country-specific food price index driven by international food price variation and exploit within-country variation using panel data techniques that account for unobservable cross-country heterogeneity and common year shocks.
- Key previewed findings:
  - Increases in international food prices lead to a significant deterioration of democratic institutions in Low Income Countries.
  - A one standard deviation increase in the international food price index significantly reduced Low Income Countries' polity score by about 0.03 standard deviations on average.
  - Food price increases significantly increase the incidence of intra-state conflict (anti-government demonstrations, riots, civil conflict) in Low Income Countries; no significant effect in High Income Countries.

### II. Data
- Food price index construction:
  - Country-specific international food price index FoodPrice_c,t weighted by θ_i,c (average time-invariant net-exports of food commodity c in country i GDP); exponential form motivated by log-linearizing GDP and taking total differentials.
  - Weights: average (time-invariant) net export GDP shares to reduce endogeneity concerns from year-to-year variation in net export shares.
  - Commodities included: beef, maize, rice, sugar, and wheat.
  - Data sources: annual international food prices 1970-2007 from UNCTAD Commodity Statistics; net food export values from NBER-United Nations Trade Database.
  - Multiple prices for same commodity averaged when present.
- Democracy and intra-state conflict measures:
  - Democracy: revised combined Polity score (Polity2) from Polity IV; range -10 to +10; democracies if Polity2 > 0, autocracies if Polity2 < 0; deep democracies if Polity2 > 6; deep autocracies if Polity2 < -6.
  - Polity sub-scores used: constraints on the chief executive (range 1 to 7) and political competition (range 1 to 10).
  - Civil conflict indicator = 1 if country experienced intra-state conflict with more than 25 battle deaths (PRIO Uppsala 2010).
  - Riots and anti-government demonstrations data from Banks (2010).
- Group summary statistics (selected):
  - Low Income Group: mean Polity2 = -2.42; Std. Dev. 6.02; Obs. 2676.
  - Low Income Group: mean incidence of civil conflict = 0.25; Std. Dev. 0.43; Obs. 2676.
  - Middle Income Group: mean Polity2 = 2.34; Std. Dev. 6.89; Obs. 1106; mean civil conflict = 0.14; Std. Dev. 0.35; Obs. 1269.
  - High Income Group: mean Polity2 = 4.74; Std. Dev. 7.89; Obs. 1435; mean civil conflict = 0.06; Std. Dev. 0.24; Obs. 1435.

### III. Estimation Strategy
- Baseline econometric model:
  - ΔPolity_i,t = γ(ΔFoodPI_i,t) + α_i + β_t + u_i,t (country fixed effects α_i; year fixed effects β_t; errors clustered at country level).
- Primary focus: Low Income Countries sample (World Bank 2010 classification, includes Low Income and Lower Middle Income as "Low Income Countries").
- Identification: within-country variation in food price index; comparison by restricting to Middle and High Income Countries.

### IV. Main Results
- Effect on democracy:
  - Negative, highly significant effect of food price increases on Polity2 in Low Income Countries.
  - Quantitative magnitude: a one standard deviation increase in the food price index reduced the Polity2 score by about 0.03 standard deviations in Low Income Countries.
  - Standard deviations reported: standard deviation of the log-change in the food price index is 0.004; standard deviation of the change in the Polity2 score is 1.99.
  - Lag structure: FoodPI t-1 = -5.536 (t-value -1.41) and FoodPI t-2 = -2.761 (t-value -0.68) reported (not statistically significant at conventional levels). Polity2, t-1 coefficient = -0.131*** (t-value -8.31) in SYS-GMM.
  - Table 2 FoodPI coefficients (selected LS/SYS-GMM): Column (1) -16.379*** (t-value -3.05); Column (2) -16.901*** (t-value -3.32); Column (3) -15.093*** (t-value -3.19); Column (4) -14.801*** (t-value -3.84); Column (5) -12.165*** (t-value -3.34); Column (6) -12.144** (t-value -2.60). Observations typically 2676; Countries 61.
  - Dynamic panel estimates: estimated convergence coefficient = -0.13 implying a half-life of shocks to the level of the polity score of around 4 years. Dynamic estimates confirm significant negative impact effect of food price increases on polity score.
- Robustness across democracy measures:
  - Results robust when using Polity, Exconst, Polcomp, Democ, Autoc.
  - Table 3 FoodPI coefficients (LS): Polity -18.717*** (t-value -6.44); Exconst -7.836*** (t-value -6.02); Polcomp -9.703*** (t-value -11.07); Democ -8.029*** (t-value -5.72); Autoc 10.687*** (t-value 6.85). Countries 61; Observations 2496.
- Macroeconomic channels:
  - Table 4 FoodPI coefficients (LS):
    - GDP (log-change real per capita GDP): 1.122*** (t-value 3.39).
    - Inv (investment): 1.978*** (t-value 3.19).
    - Gov (government expenditures): 1.175*** (t-value 4.31).
    - Cons (private consumption): -0.684** (t-value -2.54).
    - Gini: 0.228* (t-value 1.82).
  - Interpretation: food price increases raise average income in net food exporters (terms-of-trade effect) and increase investment and government expenditures, while reducing private consumption and increasing income inequality.
- Intra-state conflict:
  - Table 5 FoodPI coefficients and significance:
    - Demonstrations: 4.899*** (t-value 2.75).
    - Riots: 4.693*** (t-value 4.93).
    - Expropriation: 32.997*** (t-value 9.58).
    - Civil Conflict (LS): 1.259** (t-value 2.06).
    - Civil Conflict (Logit): 180.014* (t-value 1.68) — note conditional logit estimates do not represent marginal effects.
  - Quantitative magnitude for demonstrations/riots: a one standard deviation increase in the food price index increased the number of anti-government demonstrations and riots by about 0.01 standard deviations.
  - Overall: food price shocks are significant determinants of intra-state conflicts in Low Income Countries.
- Comparison by income groups:
  - Food price increases had no significant effects on democratic institutions or intra-state conflict measures in High Income Countries (Table 8).
  - Middle Income Countries: no consistent significant increase in intra-state conflicts associated with food price increases.

### V. Robustness Checks
- Excluding large food producers (Table 6):
  - Excluded countries: China, Guatemala, India, Indonesia, Pakistan, Thailand, Uganda, Ukraine and Vietnam (these produce more than 3 percent of world food production).
  - Results persist: Demonstrations 5.615*** (t-value 3.75); Riots 5.187*** (t-value 5.40); Expropriation 32.378*** (t-value 9.80); Civil Conflict 1.264** (t-value 2.06); Civil Conflict Logit 253.145* (t-value 1.87). Countries 51 (most columns).
  - Supports exogeneity assumption of international food prices for most Low Income Countries.
- Pre- vs. post-1990 comparison (Table 7):
  - FoodPI main effects (pre/post pooled): Demonstrations 3.348* (t-value 1.80); Riots 5.009*** (t-value 4.42); Expropriation 11.157* (t-value 1.68); Civil Conflict 1.044* (t-value 1.87).
  - FoodPI*Post 1990s interaction coefficients: Demonstrations 11.021** (t-value 2.31); Riots -2.254 (t-value -0.53); Expropriation 30.856*** (t-value 3.15); Civil Conflict 1.539 (t-value 0.71).
  - Findings: (i) significant positive average effect of food prices on intra-state conflict for pre-1990 period; (ii) evidence that for the post-1990 period the effect has become significantly larger for some outcomes (e.g., demonstrations, expropriation).
- High and Middle Income Countries (Table 8):
  - High Income Country estimates not significant and often imprecise (e.g., Demonstrations -11.390 (t-value -0.17); Riots 18.129 (t-value 0.33); Civil Conflict -0.292 (t-value -0.08)).
  - Middle Income Country estimates largely not significant (e.g., Demonstrations 10.625 (t-value 1.43); Riots -3.402 (t-value -0.71); Civil Conflict 0.997 (t-value 1.15)).
  - Interpretation: Low Income Countries are more vulnerable because food expenditures are a larger share of the consumption basket for the extremely poor.

### VI. Conclusion
- Core empirical finding: during times of international food price increases, political institutions in Low Income Countries significantly deteriorated.
- Mechanism: international food price increases in Low Income Countries significantly increased the likelihood of civil conflict and other civil strife (anti-government demonstrations, riots).
- Macroeconomic summary:
  - International food price increases induced a significant increase in real per capita GDP and real per capita investment in net food exporters (the “terms of trade effect”).
  - Simultaneously, international food price increases induced a significant decrease in real per capita consumption and a significant increase in income inequality.
  - Net effect: real macroeconomic impacts beyond average per capita income, associated with decreased consumption and increased gap between rich and poor.
- Policy implication raised (not addressed within the paper): what can and should be done by developed countries and international organizations in response to drastic increases in international food prices?

*Source: _wp1162 - 2007. Our main finding is that in Low Income Countries increases in the international (PDF chapter/section).*

### 2007. Our main finding is that in Low Income Countries increases in the international

### _wp1162 - 2007. Our main finding is that in Low Income Countries increases in the international

### I. Introduction
- Main stated finding: in Low Income Countries increases in the international food prices lead to a significant deterioration of democratic institutions and a significant increase in the incidence of anti-government demonstrations, riots, and civil conflict.
- In High Income Countries variations in the international food prices have no significant effects on democratic institutions and measures of intra-state conflict.
- Empirical results point to a significant externality of variations in international food prices on Low Income Countries' social and political stability.

### II. Data
- (Section listed in Contents, page 5.)

### III. Estimation Strategy
- (Section listed in Contents, page 7.)

### IV. Main Results
- In Low Income Countries:
  - increases in the international food prices lead to a significant deterioration of democratic institutions.
  - increases in the international food prices lead to a significant increase in the incidence of anti-government demonstrations, riots, and civil conflict.
- In High Income Countries:
  - variations in the international food prices have no significant effects on democratic institutions and measures of intra-state conflict.
- Overall interpretation:
  - the empirical results point to a significant externality of variations in international food prices on Low Income Countries' social and political stability.

### V. Robustness Checks
- (Section listed in Contents, page 10.)

### VI. Conclusion
- (Section listed in Contents, page 11.)

### References .............................................................................................................

### _wp1162 - References .............................................................................................................

### I. INTRODUCTION
- Research question: whether variations in international food prices affect intra-state stability and political institutions, especially in the world's poorest countries.
- Sample and period: panel of over 120 countries during the 1970-2007 period.
- Identification strategy: construct a country-specific food price index driven by international food price variation and exploit within-country variation using panel data techniques that account for unobservable cross-country heterogeneity and common year shocks.
- Key takeaway preview:
  - Increases in international food prices lead to a significant deterioration of democratic institutions in Low Income Countries.
  - A one standard deviation increase in the international food price index significantly reduced Low Income Countries' polity score by about 0.03 standard deviations on average.
  - Food price increases significantly increase the incidence of intra-state conflict (anti-government demonstrations, riots, civil conflict) in Low Income Countries; no significant effect in High Income Countries.
- Relation to literature:
  - Complements Acemoglu and Robinson (2001, 2006) theory and empirical work on economic shocks and democratic change.
  - Extends conflict literature by focusing exclusively on food commodity international prices and their link to civil conflict risk.
  - Addresses endogeneity concerns in food insecurity literature by using international food price variation as a plausibly exogenous source of variation.

### II. DATA
- Food price index construction:
  - Country-specific international food price index: FoodPrice_c,t weighted by θ_i,c (average time-invariant net-exports of food commodity c in country i GDP); exponential form motivated by log-linearizing GDP and taking total differentials.
  - Weights: average (time-invariant) net export GDP shares to reduce endogeneity concerns from year-to-year variation in net export shares.
  - Commodities included: beef, maize, rice, sugar, and wheat.
  - Data sources: annual international food prices 1970-2007 from UNCTAD Commodity Statistics; net food export values from NBER-United Nations Trade Database.
  - Multiple prices for same commodity averaged when present.
- Measures of democracy and intra-state conflict:
  - Democracy: revised combined Polity score (Polity2) from Polity IV; range -10 to +10; democracies if Polity2 > 0, autocracies if Polity2 < 0; deep democracies if Polity2 > 6; deep autocracies if Polity2 < -6.
  - Polity sub-scores used: constraints on the chief executive (range 1 to 7) and political competition (range 1 to 10).
  - Intra-state instability:
    - Civil conflict indicator = 1 if country experienced intra-state conflict with more than 25 battle deaths (PRIO Uppsala 2010).
    - Riots and anti-government demonstrations data from Banks (2010).
- Group summary statistics (Panel A of Table 1 referenced):
  - Low Income Group: mean Polity2 = -2.4; mean incidence of civil conflict = 0.25.
  - Middle Income Group: mean Polity2 = 2.3; mean incidence of civil conflict = 0.14.
  - High Income Group: mean Polity2 = 4.7; mean incidence of civil conflict = 0.07.

### III. ESTIMATION STRATEGY
- Baseline econometric model: ΔPolity_i,t = γ(ΔFoodPI_i,t) + α_i + β_t + u_i,t (country fixed effects α_i; year fixed effects β_t; errors clustered at country level).
- Primary focus: Low Income Countries sample (World Bank 2010 classification, includes Low Income and Lower Middle Income as "Low Income Countries").
- Identification: within-country variation in food price index; checks by restricting to Middle and High Income Countries for comparison.

### IV. MAIN RESULTS
- Effect on democracy (Table 2 summary):
  - Negative, highly significant effect of food price increases on Polity2 in Low Income Countries.
  - Quantitative magnitude: a one standard deviation increase in the food price index reduced the Polity2 score by about 0.03 standard deviations in Low Income Countries.
  - Standard deviations: the standard deviation of the log-change in the food price index is 0.004; the standard deviation of the change in the Polity2 score is 1.99.
  - Lag structure: lagged food price index coefficients are negative but not statistically significant at conventional levels; main effect is on impact.
  - Dynamics: dynamic panel estimates (least squares and system-GMM) include lagged polity level; estimated convergence coefficient = -0.13 implying a half-life of shocks to the level of the polity score of around 4 years. Dynamic estimates confirm significant negative impact effect of food price increases on polity score.
- Robustness to alternative democracy measures (Table 3 summary):
  - Results robust when using Polity2 excluding interregnum/transition, executive constraints, political competition, democracy and autocracy scores (both range 0 to 10).
  - Conclusion: consistently, higher international food prices lead to deterioration of political institutions in Low Income Countries.
- Macroeconomic channels (Table 4 summary):
  - Column (1): increases in food prices lead to significant increases in GDP per capita (terms-of-trade effect for net food exporters).
  - Column (2): significant increase in private investment.
  - Column (3): significant increase in real per capita government expenditures.
  - Column (4): significant decrease in private consumption.
  - Column (5): significant increases in Gini coefficients (data from United Nations WIDER (2008)).
  - Interpretation: food price increases raise average income in net food exporters but reduce private consumption and increase income inequality, suggesting distributional and compositional effects (increased relative deprivation).
- Intra-state conflict (Table 5 summary):
  - Food price increases significantly increase the number of anti-government demonstrations and number of riots.
  - Quantitative magnitude for demonstrations/riots: a one standard deviation increase in the food price index increased the number of anti-government demonstrations and riots by about 0.01 standard deviations.
  - Significant increase in expropriation risk measured by PRS expropriation score.
  - Higher food prices significantly increase the incidence of civil conflict (including conditional logit fixed effects estimates for binary dependent variable; note conditional logit estimates do not represent marginal effects).
  - Overall: food price shocks are significant determinants of intra-state conflicts in Low Income Countries.

### V. ROBUSTNESS CHECKS
- Excluding large food producers (Table 6):
  - Excluded countries: China, Guatemala, India, Indonesia, Pakistan, Thailand, Uganda, Ukraine and Vietnam (these produce more than 3 percent of world food production).
  - Finding: significant effects of food prices on measures of intra-state conflict persist after excluding these large food suppliers, supporting the exogeneity assumption of international food prices for most Low Income Countries.
- Pre- vs. post-1990 comparison (Table 7):
  - Interaction model allowing marginal effect of food prices on intra-state conflict to differ for pre-1990 and post-1990.
  - Findings:
    - (i) Significant positive average effect of food prices on intra-state conflict for pre-1990 period.
    - (ii) Evidence that for the post-1990 period the effect of food price changes on intra-state conflict has become significantly larger (see e.g. columns (1) and (3)).
- High and Middle Income Countries (Table 8):
  - Food price increases did not lead to a significant increase in incidence of intra-state conflicts in High Income and Middle Income Countries.
  - Interpretation: Low Income Countries are more vulnerable because food expenditures are a larger share of the consumption basket for the extremely poor.

### VI. CONCLUSION
- Core empirical finding: during times of international food price increases, political institutions in Low Income Countries significantly deteriorated.
- Mechanism: food price increases in Low Income Countries significantly increased the likelihood of civil conflict and other civil strife (anti-government demonstrations, riots).
- Macroeconomic summary:
  - International food price increases induced a significant increase in real per capita GDP and real per capita investment in net food exporters (the “terms of trade effect”).
  - Simultaneously, international food price increases induced a significant decrease in real per capita consumption and a significant increase in income inequality.
  - Net effect: real macroeconomic impacts beyond average per capita income, associated with decreased consumption and increased gap between rich and poor.
- Policy implication (raised as a natural question): what can and should be done by developed countries and international organizations in response to drastic increases in international food prices? (Not addressed within scope of this paper.)

*Source: _wp1162 - References .............................................................................................................*

### REFERENCES

### _wp1162 - REFERENCES

### References (Selected)
- Arezki, R. K. Deininger, and H. Selod. 2011. “Interest in Large-Scale Land Acquisition for Agribusiness Investment: Extent and Determinants and the ‘Global Land Grab.” Forthcoming Policy Research Working Paper, World Bank, Washington, DC.
- Acemoglu, D. and J. Robinson (2001). "A Theory of Political Transitions." American Economic Review 91: 938-963.
- Acemoglu, D. and J. Robinson (2006). Economic Origins of Dictatorship and Democracy. New York, Cambridge University Press.
- Berger, H. and M. Spoerer (2001). "Economic Crises and the European Revolutions of 1848." Journal of Economic History 61: 293-326.
- Brinkman, H. and C. Hendrix (2010). "Food Insecurity and Conflict: Applying the WDR Framework." in World Development Report 2011.
- Bruckner, M. and A. Ciccone (2010a). "International Commodity Price Shocks, Growth, and the Outbreak of Civil War in Sub-Saharan Africa." Economic Journal 120: 519-534.
- Bruckner, M. and A. Ciccone (2010b). "Rainfall and the Democratic Window of Opportunity." Econometrica, forthcoming.
- Burke, P. and A. Leigh (2010). "Do Output Contractions Trigger Democratic Change?" American Economic Journal: Macroeconomics 2: 124-157.
- Fearon, J. and D. Laitin (2003). “Ethnicity, Insurgency and Civil War.” American Political Science Review 97 (1): 75-90.
- Heston, A., R. Summers and B. Aten (2009). "Penn World Table Version 6.3", Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, August 2009.
- Marshall, M. and K. Jaggers (2010): Polity IV Project: Dataset Users’ Manual. Center for Global Policy, George Mason University (www.cidcm.umd.edu/polity). [Polity IV Data Computer File, Version 2010. College Park, MD: Center for International Development and Conflict Management, University of Maryland.]
- Miguel, E., S. Satyanath, and E. Sergenti (2004). "Economic Shocks and Civil Conflict: An Instrumental Variables Approach." Journal of Political Economy 112: 725-753.
- UNU-WIDER (2008). World Income Inequality Database. Version 2.0c. Online database.
- 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).

*Source: _wp1162 - REFERENCES (PDF).*

### Table 1: Descriptive Statistics — Key values by income group
- Panel A: Low Income Countries
  - Polity2: Mean -2.42; Std. Dev. 6.02; Obs. 2676
  - Executive Constraints: Mean 3.76; Std. Dev. 3.01; Obs. 2544
  - Political Competition: Mean 3.07; Std. Dev. 1.91; Obs. 2544
  - GDP Per Capita (PWt 6.3): Mean 2444; Std. Dev. 1739; Obs. 2627
  - Gini: Mean 0.42; Std. Dev. 0.10; Obs. 423
  - Riots: Mean 0.48; Std. Dev. 1.43; Obs. 2642
  - Anti-Government Demonstrations: Mean 0.45; Std. Dev. 1.75; Obs. 2642
  - Civil Conflict: Mean 0.25; Std. Dev. 0.43; Obs. 2676
- Panel B: Upper Middle Income Countries
  - Polity2: Mean 2.34; Std. Dev. 6.89; Obs. 1106
  - Executive Constraints: Mean 5.92; Std. Dev. 3.55; Obs. 1077
  - Political Competition: Mean 4.57; Std. Dev. 2.23; Obs. 1077
  - GDP Per Capita (PWt 6.3): Mean 8002; Std. Dev. 5041; Obs. 1122
  - Gini: Mean 0.43; Std. Dev. 0.11; Obs. 373
  - Riots: Mean 0.77; Std. Dev. 1.96; Obs. 1171
  - Anti-Government Demonstrations: Mean 0.59; Std. Dev. 1.77; Obs. 1170
  - Civil Conflict: Mean 0.14; Std. Dev. 0.35; Obs. 1269
- Panel C: High Income Countries
  - Polity2: Mean 4.74; Std. Dev. 7.89; Obs. 1435
  - Executive Constraints: Mean 7.40; Std. Dev. 3.88; Obs. 1421
  - Political Competition: Mean 5.42; Std. Dev. 2.36; Obs. 1421
  - GDP Per Capita (PWt 6.3): Mean 20687; Std. Dev. 12077; Obs. 1372
  - Gini: Mean 0.28; Std. Dev. 0.06; Obs. 459
  - Riots: Mean 0.79; Std. Dev. 2.84; Obs. 1384
  - Anti-Government Demonstrations: Mean 0.56; Std. Dev. 2.60; Obs. 1384
  - Civil Conflict: Mean 0.06; Std. Dev. 0.24; Obs. 1435

### Table 2: Food Prices and Political Institutions — Polity2 regressions (selected coefficients)
- Estimation methods: LS in columns (1)-(5); SYS-GMM in column (6).
- FoodPI, t: Coefficients and significance
  - Column (1): -16.379*** (t-value -3.05)
  - Column (2): -16.901*** (t-value -3.32)
  - Column (3): -15.093*** (t-value -3.19)
  - Column (4): -14.801*** (t-value -3.84)
  - Column (5): -12.165*** (t-value -3.34)
  - Column (6): -12.144** (t-value -2.60)
- FoodPI, t-1 and t-2 reported in column (6): coefficients -5.536 (t-value -1.41) and -2.761 (t-value -0.68) respectively.
- Polity2, t-1: Column (6) coefficient -0.131*** (t-value -8.31); and -0.141*** (t-value -4.68) reported elsewhere.
- Country FE: Varies by column (No/Yes). Year FE: Varies by column.
- Observations: 2676 (most columns), 2589 in column (4).
- Countries: 61.
- Note: Dependent variable is the change in the Polity2 score. Explanatory variable is the log-change in the food price index. *Significance levels: * 90 percent, ** 95 percent, *** 99 percent.*

### Table 3: Food Prices and Political Institutions — Alternative measures
- Dependent variables across columns: Polity, Exconst, Polcomp, Democ, Autoc.
- FoodPI coefficients (LS):
  - Polity (1): -18.717*** (t-value -6.44)
  - Exconst (2): -7.836*** (t-value -6.02)
  - Polcomp (3): -9.703*** (t-value -11.07)
  - Democ (4): -8.029*** (t-value -5.72)
  - Autoc (5): 10.687*** (t-value 6.85)
- Country FE: Yes. Year FE: Yes.
- Countries: 61. Observations: 2496.
- Note: All scores exclude values recorded as -66, -77, and -88. Explanatory variable is the log-change in the food price index. Significance markers as above.

### Table 4: Food Prices and Macroeconomic Outcomes — Key coefficients
- Dependent variables: GDP (log-change real per capita GDP), Inv (investment), Gov (government expenditures), Cons (private consumption), Gini.
- FoodPI coefficients (LS):
  - GDP (1): 1.122*** (t-value 3.39)
  - Inv (2): 1.978*** (t-value 3.19)
  - Gov (3): 1.175*** (t-value 4.31)
  - Cons (4): -0.684** (t-value -2.54)
  - Gini (5): 0.228* (t-value 1.82)
- Country FE: Yes. Year FE: Yes.
- Countries: 61 for columns (1)-(4); 411 for Gini (column 5).
- Observations: 2444 for columns (1)-(4); 57 for Gini.
- Note: Explanatory variable is the log-change in the food price index. Significance markers as above.

### Table 5: Food Prices and Intra-State Conflict — Selected results
- Dependent variables: Demonstrations, Riots, Expropriation, Civil Conflict (columns (1)-(5)).
- Estimation: LS for columns (1)-(4); Logit for column (5).
- FoodPI coefficients:
  - Demonstrations (1): 4.899*** (t-value 2.75)
  - Riots (2): 4.693*** (t-value 4.93)
  - Expropriation (3): 32.997*** (t-value 9.58)
  - Civil Conflict (4): 1.259** (t-value 2.06)
  - Civil Conflict Logit (5): 180.014* (t-value 1.68)
- Country FE and Year FE: Yes across columns.
- Countries: 61 (except Expropriation uses 53 in column (3)).
- Observations: 2450 (columns (1)-(2), (4)-(5)); 1059 (column (3)).
- Note: Expropriation score re-scaled so higher values denote higher risk of expropriation. Explanatory variable is the log-change in the food price index.

### Table 6: Robustness Check I — Excluding Large Food Producers (Low Income Group)
- Excluded countries: China, Guatemala, India, Indonesia, Pakistan, Thailand, Uganda, Ukraine, and Vietnam.
- FoodPI coefficients (LS/Logit variants):
  - Demonstrations (1): 5.615*** (t-value 3.75)
  - Riots (2): 5.187*** (t-value 5.40)
  - Expropriation (3): 32.378*** (t-value 9.80)
  - Civil Conflict (4): 1.264** (t-value 2.06)
  - Civil Conflict Logit (5): 253.145* (t-value 1.87)
- Countries: 51 (columns (1)-(2), (4)-(5)); 44 (column (3)).
- Observations: 2060 (columns (1)-(3)); 2102 (columns (4)-(5)).
- Note: Estimation methods and significance markers as noted previously.

### Table 7: Robustness Check II — Post-1990s interaction
- FoodPI main effects:
  - Demonstrations (1): 3.348* (t-value 1.80)
  - Riots (2): 5.009*** (t-value 4.42)
  - Expropriation (3): 11.157* (t-value 1.68)
  - Civil Conflict (4): 1.044* (t-value 1.87)
  - Civil Conflict Logit (5): 111.21 (t-value 0.96)
- FoodPI*Post 1990s interaction coefficients:
  - Demonstrations: 11.021** (t-value 2.31)
  - Riots: -2.254 (t-value -0.53)
  - Expropriation: 30.856*** (t-value 3.15)
  - Civil Conflict: 1.539 (t-value 0.71)
  - Civil Conflict Logit: 475.859 (t-value 1.57)
- Countries: 61. Observations: 2450 (except Expropriation 1059).
- Note: Estimation methods and significance markers as noted previously.

### Table 8: Robustness Check III — Effects in Middle and High Income Countries
- Panel A: High Income Countries (selected FoodPI coefficients)
  - Demonstrations (1): -11.390 (t-value -0.17)
  - Riots (2): 18.129 (t-value 0.33)
  - Expropriation (3): -112.305 (t-value -1.22)
  - Civil Conflict (4): -0.292 (t-value -0.08)
  - Civil Conflict Logit (5): -98.047 (t-value -0.11)
  - Countries: 1357 (cols (1)-(3)); 1414 (cols (4)-(5)). Observations: 34 (cols (1)-(2),(4)-(5)); 33 (col (3)).
- Panel B: Middle Income Countries (selected FoodPI coefficients)
  - Demonstrations: 10.625 (t-value 1.43)
  - Riots: -3.402 (t-value -0.71)
  - Expropriation: 28.295 (t-value 0.80)
  - Civil Conflict: 0.997 (t-value 1.15)
  - Civil Conflict Logit: 37.729 (t-value 0.14)
  - Countries: 26 (most columns); Observations and panel counts vary (e.g., Year FE 1046 or 1060 etc.).
- Note: Estimation methods and significance markers as noted previously. Explanatory variable is the log-change in the food price index.

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1162.pdf_