Food Security for Whom?

Transcription

1 Food Security for Whom? The Effectiveness of Food Reserves in Poor Developing Countries Randall Romero-Aguilar Seminar at Banco Central de Costa Rica June 02, 2014

2 Outline 1 Introduction 2 The model 3 Results 4 Conclusions

3 Introduction The Problem Despite rising production, food prices are higher and increasingly volatile Price Index Source: FAO's cereal price index Romero-Aguilar Food Security for Whom? 1 / 28

4 Introduction and its consequences More price uncertainty increase in risk for farmers. Increase in hunger among poor net food buyers who account for 88% of rural and 97% of urban poor households. More than 60 food riots in 30 different countries. Romero-Aguilar Food Security for Whom? 2 / 28

5 Introduction and its consequences More price uncertainty increase in risk for farmers. Increase in hunger among poor net food buyers who account for 88% of rural and 97% of urban poor households. More than 60 food riots in 30 different countries. Romero-Aguilar Food Security for Whom? 2 / 28

6 Introduction and its consequences More price uncertainty increase in risk for farmers. Increase in hunger among poor net food buyers who account for 88% of rural and 97% of urban poor households. More than 60 food riots in 30 different countries. Romero-Aguilar Food Security for Whom? 2 / 28

7 Introduction Undernourished people in the world Source: The Economist, Oct 10th 2012 Romero-Aguilar Food Security for Whom? 3 / 28

8 Introduction Food riots Algeria (4), Saudi Arabia (1) Haiti (5), Egypt (3), Mauritania (1), Sudan (1), Yemen (300+) Cote d'ivoire (1) Somalia (5) Oman (2), Morocco (5) Sudan (3) Tunisia (1) Egypt (800+) Iraq (29), Bahrain (31) Cameroon (40) Libya (10000+) Syria (900+) Yemen (12) Tunisia (300+) Uganda (5) Mozambique (6) India (1), Sudan (1) Food Price Index Mauritania (2) India (4) Somalia (5) Mozambique (13) Burundi (1) Source: Lagi, Bertrand, and Bar-Yam Romero-Aguilar Food Security for Whom? 4 / 28

9 Introduction Possible causes of high food prices Affecting supply: 1 rising oil prices; 2 declining stocks and reserves; 3 regional catastrophic weather; 4 export restrictions; 5 decline in productivity and R&D in agriculture. Romero-Aguilar Food Security for Whom? 5 / 28

10 Introduction Possible causes of high food prices Affecting supply: 1 rising oil prices; 2 declining stocks and reserves; 3 regional catastrophic weather; 4 export restrictions; 5 decline in productivity and R&D in agriculture. Affecting demand: 1 strong income growth in China and India; 2 biofuel production in the USA and Europe; 3 preventive imports surges; 4 speculation in financial markets. Romero-Aguilar Food Security for Whom? 5 / 28

11 Introduction The effects of export restrictions on rice prices Apr: Nigeria scraps 100% tariffs & imports 0.5 million mt Thai rice Mar: India, Vietnam, China & Cambodia impose tighter export restrictions & new Thai govt discusses possibility of ban Q1: Saudi Arabia subsidizes rice imports; Saudi imports from Thailand rise by nearly 90% after India s ban Jun: Record harvests, Japan allowed to re-export rice stocks, dollar strengthens, oil & other crop prices fall Jun: Cambodia removes ban Jun: Egypt announces reexport of rice from Sep. Sep: India lifts export ban on some higher quality varieties Jan: Egypt restrict exports, and China adds 10% tax on exports & rescinds VAT rebate Sep-Oct 2007, Vietnam and India place partial restrictions on exports Jan-Apr: Philippines buys normal annual quota injust 4 months, including govtto-govt deal with Vietnam Jan-Feb: Drought in Iran order 0.8 million mt of Thai rice Strong demand from energy exporters keeps rice prices 25-30% above 2007 levels Source: Headey (2011) Romero-Aguilar Food Security for Whom? 6 / 28

12 Introduction Food reserves as a solution? Several authors have called for food reserves. Objective: welfare? hunger? Intermediate target: price stabilization vs. humanitarian assistance. Romero-Aguilar Food Security for Whom? 7 / 28

13 Introduction Food reserves as a solution? Several authors have called for food reserves. Objective: welfare? hunger? Intermediate target: price stabilization vs. humanitarian assistance. Romero-Aguilar Food Security for Whom? 7 / 28

14 Introduction Food reserves as a solution? Several authors have called for food reserves. Objective: welfare? hunger? Intermediate target: price stabilization vs. humanitarian assistance. Romero-Aguilar Food Security for Whom? 7 / 28

15 Introduction Some questions If a country decides to operate a grain reserve what is the optimal size of the reserve? how is the country's hunger rate affected by its operations? is it better to store cash? how does the optimal operation of the reserve change if the objective is to maximize 'social welfare'? Romero-Aguilar Food Security for Whom? 8 / 28

16 Introduction Some questions If a country decides to operate a grain reserve what is the optimal size of the reserve? how is the country's hunger rate affected by its operations? is it better to store cash? how does the optimal operation of the reserve change if the objective is to maximize 'social welfare'? Romero-Aguilar Food Security for Whom? 8 / 28

17 Introduction Some questions If a country decides to operate a grain reserve what is the optimal size of the reserve? how is the country's hunger rate affected by its operations? is it better to store cash? how does the optimal operation of the reserve change if the objective is to maximize 'social welfare'? Romero-Aguilar Food Security for Whom? 8 / 28

18 Introduction Some questions If a country decides to operate a grain reserve what is the optimal size of the reserve? how is the country's hunger rate affected by its operations? is it better to store cash? how does the optimal operation of the reserve change if the objective is to maximize 'social welfare'? Romero-Aguilar Food Security for Whom? 8 / 28

19 Introduction Motivation: Is grain storage a good idea? The logic behind grain storage is simple: Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

20 Introduction Motivation: Is grain storage a good idea? The logic behind grain storage is simple: Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

21 Introduction Motivation: Is grain storage a good idea? The logic behind grain storage is simple: Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

22 Introduction Motivation: Is grain storage a good idea? Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! To answer this, a model should quantify the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

23 Introduction Motivation: Is grain storage a good idea? Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! To answer this, a model should quantify the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

24 Introduction Motivation: Is grain storage a good idea? Seven years of abundance followed by seven years of famine What if country never has years of abundance? Opportunity cost of storing grain is very high! To answer this, a model should quantify the increase on national hunger induced by an international crisis; to what extent a reserve alleviates this increase, and at what cost. Romero-Aguilar Food Security for Whom? 9 / 28

25 Outline 1 Introduction 2 The model 3 Results 4 Conclusions

26 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption grain x g Utility U(f i, n i) food f i else n i vegetable x v Romero-Aguilar Food Security for Whom? 10 / 28

27 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption food = dprice α income η Romero-Aguilar Food Security for Whom? 10 / 28

28 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption [θgrain σ 1 σ food f i ] + (1 θ)veget. σ 1 σ σ 1 σ Romero-Aguilar Food Security for Whom? 10 / 28

29 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption Romero-Aguilar Food Security for Whom? 10 / 28

30 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption Romero-Aguilar Food Security for Whom? 10 / 28

31 The model Key features of the model Nested utility: two goods, two food ingredients Constant demand elasticities Substitution between ingredients Intertemporal, two grain prices Heterogeneous households: log-logistic income distribution Log-logistic food consumption Romero-Aguilar Food Security for Whom? 10 / 28

32 The model Hunger changes in response to food prices [ ( cp α (Gπ) η ) ] 1/Gη 1 Γ(P) = 1 + ζy η sin η (Gπ) B B = hungry because of high price A = hungry even when food is cheap Density A Hunger threshold tranquil period food crisis 0 c 2c 3c 4c Food consumption Romero-Aguilar Food Security for Whom? 11 / 28

33 The model Empirical relevance of the model Food adequacy x it and undernourishment Γ it in ECOWAS and ASEAN ( ) log Γit 1 Γ it = di b f log x it + ɛ it Model approximates FAO's hunger estimates reasonably well. FAO data Fixed-effects Prevalence of undernourishment ECOWAS countries Benin Cape Verde Gambia Ghana Guinea Guinea Bissau Mali Niger Nigeria Senegal Sierra Leone Togo Average Dietary Energy Supply Adequacy Romero-Aguilar Food Security for Whom? 12 / 28

34 The model Empirical relevance of the model Food adequacy x it and undernourishment Γ it in ECOWAS and ASEAN ( ) log Γit 1 Γ it = di b f log x it + ɛ it Model approximates FAO's hunger estimates reasonably well. FAO data Fixed-effects Prevalence of undernourishment ASEAN countries Cambodia Indonesia Lao People s Democratic Republic Philippines Thailand Viet Nam Average Dietary Energy Supply Adequacy Romero-Aguilar Food Security for Whom? 12 / 28

35 The model Government problem: objectives and tools Government runs a grain stockpile to deal with price fluctuations. Two alternative objectives: welfare vs. hunger One policy tool: tariff on grain imports Two state variables: initial stock and grain price ( V s, pg ) { = max r(τ, pg ) + δ E V (s, p )} g τ [ ( )] subject to s = (1 φ) s + 1 p Υ τ, p g g 0 ) π ij = Pr (p g = p j p g = p i Romero-Aguilar Food Security for Whom? 13 / 28

36 The model Government problem: objectives and tools Government runs a grain stockpile to deal with price fluctuations. Two alternative objectives: welfare vs. hunger One policy tool: tariff on grain imports Two state variables: initial stock and grain price ( V s, pg ) { = max r(τ, pg ) + δ E V (s, p )} g τ [ ( )] subject to s = (1 φ) s + 1 p Υ τ, p g g 0 ) π ij = Pr (p g = p j p g = p i Romero-Aguilar Food Security for Whom? 13 / 28

37 The model Government problem: objectives and tools Government runs a grain stockpile to deal with price fluctuations. Two alternative objectives: welfare vs. hunger One policy tool: tariff on grain imports Two state variables: initial stock and grain price ( V s, pg ) { = max r(τ, pg ) + δ E V (s, p )} g τ [ ( )] subject to s = (1 φ) s + 1 p Υ τ, p g g 0 ) π ij = Pr (p g = p j p g = p i Romero-Aguilar Food Security for Whom? 13 / 28

38 The model Reward function r(τ, P), by objective Objective, V Reward function, r(τ, pg ) [ ] 1 ρ Hunger, Γ 1 1 ρ 1 Γ(τ, pg ) ] Utility, S(v i ) 1 ρ 1 1 ρ [v(τ, S pg ) Romero-Aguilar Food Security for Whom? 14 / 28

39 The model Solving the model: The food crisis in Haiti Calibration of parameters: Haiti Γ 2011 = 44.5% Imports 70% of cereals consumed p g increased 85% during crisis Food Crisis in Haiti: Dec2007-Mar2008: rice price doubles Early April 2008: violent protests in Port-au-Prince April 12: Prime Minister Jacques Adouard Alexis ousted Romero-Aguilar Food Security for Whom? 15 / 28

40 The model Solving the model: The food crisis in Haiti Calibration of parameters: Haiti Γ 2011 = 44.5% Imports 70% of cereals consumed p g increased 85% during crisis Food Crisis in Haiti: Dec2007-Mar2008: rice price doubles Early April 2008: violent protests in Port-au-Prince April 12: Prime Minister Jacques Adouard Alexis ousted Romero-Aguilar Food Security for Whom? 15 / 28

41 The model Food reserve in Haiti Jul2013: gov't begins construction of reserve, tonnes The construction of this strategic reserve reflects the desire of my Government to promote national agricultural production, stabilize the market price of commodities and combat food insecurity. Indeed, the fight against hunger and extreme poverty constitutes the main pillars of government action. Romero-Aguilar Food Security for Whom? 16 / 28

42 The model Food reserve in Haiti Jul2013: gov't begins construction of reserve, tonnes The construction of this strategic reserve reflects the desire of my Government to promote national agricultural production, stabilize the market price of commodities and combat food insecurity. Indeed, the fight against hunger and extreme poverty constitutes the main pillars of government action. Romero-Aguilar Food Security for Whom? 16 / 28

43 The model Solving the model: Numerical methods Numerical solution: Collocation method (dpsolve solver in CompEcon) Chebyshev polynomials with 12 nodes for continuous state s t One discrete variable, price, with values 1.0 and 1.85 Once solved, run Monte Carlo simulations to assess performance of the policy Romero-Aguilar Food Security for Whom? 17 / 28

44 The model Solving the model: Numerical methods Numerical solution: Collocation method (dpsolve solver in CompEcon) Chebyshev polynomials with 12 nodes for continuous state s t One discrete variable, price, with values 1.0 and 1.85 Once solved, run Monte Carlo simulations to assess performance of the policy Romero-Aguilar Food Security for Whom? 17 / 28

45 Outline 1 Introduction 2 The model 3 Results 4 Conclusions

46 Results The effects of crisis, without policy Variable p L p H % Price of grain Price of food Food consumption Grain consumption Vegetable consumption Hunger rate (%) Romero-Aguilar Food Security for Whom? 18 / 28

47 Results Storage policy Next period stock (weeks of grain imports) tranquil period food crisis Initial Reserve (weeks of grain imports) Romero-Aguilar Food Security for Whom? 19 / 28

48 Results Effects of storage policy on hunger 54 without policy = 53.8 % Hunger rate (percent of population) tranquil period food crisis Full stock = 5.1 weeks 44 without policy = 44.5 % Initial Reserve (weeks of grain imports) Romero-Aguilar Food Security for Whom? 20 / 28

49 Results Long-term distribution of grain reserve In half of the crisis, the reserve would be empty! 1.00 Cummulative probability tranquil period food crisis Initial Reserve (weeks of grain imports) Romero-Aguilar Food Security for Whom? 21 / 28

50 Results Long-term distribution of hunger 2.0 The reserve would fail at preventing extreme hunger. 1.5 Density Hunger rate (percent of population) Romero-Aguilar Food Security for Whom? 22 / 28

51 Results Cash vs. grain reserve? In this scenario, a grain reserve outperforms a cash reserve Value Function (certainty equivalent, percent not hungry) tranquil period food crisis Initial reserve (weeks of grain imports) Reserve Grain Deposit Romero-Aguilar Food Security for Whom? 23 / 28

52 Results Food storage vs. fighting poverty Resources used for grain reserve might be better spent at promoting growth. 100 Hunger rate (percent of population) tranquil period Haiti's income food crisis Y=131.5 'Best crisis' 48.1' Income per capita (US dollars per quarter) Romero-Aguilar Food Security for Whom? 24 / 28

53 Results Price stabilization vs. safety net? Income redistribution, targeting the poor, may have a better outcome. 100 Hunger rate (percent of population) G=0.54 food crisis tranquil period 'Best crisis' 48.1' 0 Haiti's Gini Income inequality (Gini coeffcient) Romero-Aguilar Food Security for Whom? 25 / 28

54 Results Summary statistics for other scenarios Scenario 1: Scenario 2: Scenario 3: Scenario 4: Scenario 5: (baseline) ρ = 3.0 ψ = 4 φ = 0.05 P H = 1.60 Variable Stat. p L p H p L p H p L p H p L p H p L p H Tax rate, % Initial stock End stock Food price Hunger rate, % min mean max min mean max min mean max min mean max min mean max Romero-Aguilar Food Security for Whom? 26 / 28

55 Density Scenario 1: Baseline Scenario 2: higher aversion Scenario 3: longer crisis Scenario 4: higher cost Scenario 5: lower PgH Hunger rate (percent of population) Policy A: Hunger reduction with Grain reserve Policy B: Hunger reduction with Cash reserve Policy C: Welfare maximiz. with Grain reserve PriceState tranquil period food crisis

56 Outline 1 Introduction 2 The model 3 Results 4 Conclusions

57 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

58 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

59 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

60 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

61 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

62 Conclusions Conclusions The optimal grain storage policy would not fully stabilize food prices. would not prevent extreme hunger, yet it would reduce its frequency. is very sensitive to key parameters (price process, storage costs) might be outperformed by policies that attack poverty directly. in many cases, no better than accumulating financial assets. is more active when objective is avoiding extreme hunger. Romero-Aguilar Food Security for Whom? 28 / 28

63 Appendix Additional material buyers maize wheat rice soybeans Net buyers of staple foods Historical real prices of maize Historical real prices of wheat Historical real prices of rice Historical real prices of soybeans Romero-Aguilar Food Security for Whom? 29

64 Appendix Net buyers of staple foods Net buyers of staple foods All households Poor households Urban Rural All Urban Rural All (Percentage) Albania, * * * Bangladesh, Ghana, * 69.1 * Guatemala, Malawi, Nicaragua, Pakistan, Tajikistan, Viet Nam, Unweighted average * Insufficient data. Source: FAO (2008) Source: FAO. Romero-Aguilar Food Security for Whom? 30

65 Appendix Historical maize price Maize (US), no. 2, yellow, f.o.b. US Gulf ports Price Mar 2013 US$ / mt Source: Nominal prices from World Bank's GEM database, deflated by U.S. CPI Romero-Aguilar Food Security for Whom? 31

66 Appendix Historical wheat price Wheat (US), no. 1, hard red winter, ordinary protein, export price delivered at the US Gulf port for prompt or 30 days shipment Price Mar 2013 US$ / mt Source: Nominal prices from World Bank's GEM database, deflated by U.S. CPI Romero-Aguilar Food Security for Whom? 32

67 Appendix Historical rice price Rice (Thailand), 5% broken, white rice (WR), milled, indicative price based on weekly surveys of export transactions, government standard, f.o.b. Bangkok Price Mar 2013 US$ / mt Source: Nominal prices from World Bank's GEM database, deflated by U.S. CPI Romero-Aguilar Food Security for Whom? 33

68 Appendix Historical soybeans price Soybeans (US), c.i.f. Rotterdam Price Mar 2013 US$ / mt Source: Nominal prices from World Bank's GEM database, deflated by U.S. CPI Romero-Aguilar Food Security for Whom? 34

69 Appendix References I Food and Agriculture Organization of the United Nations (2008). The State of Food Insecurity in the World High food prices and food security -- threats and opportunities. FAO. ISBN: Headey, Derek (2011). ``Rethinking the global food crisis: The role of trade shocks''. In: Food Policy 36.2, pp Lagi, Marco, Karla Z. Bertrand, and Yaneer Bar-Yam (2011). ``The Food Crises and Political Instability in North Africa and the Middle East''. In: ArXiv e-prints, p. 15. arxiv: Romero-Aguilar Food Security for Whom? 35