75648 Latin America and the Caribbean Region LCSSD Occasional Paper Series on Food Prices Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy Diego Arias Maxime Carneus March 17, 2011 2 The Unintended Consequences work has been of Food Subsidies: partly financed by the The Case Trust of the Fund forHaiti Rice Subsidy Environmentally and Socially Sustainable Development (TFESSD) LATIN AMERICA AND THE CARIBBEAN REGION LCSSD Occasional Papers Series UNINTENDED CONSEQUENCES of FOOD SUBSIDIES: THE CASE of the HAITI RICE SUBSIDY Diego Arias, Senior Agriculture Economist, LCSAR Maxime Carneus, LCSAR Authors would like to thank the comments received from: Gilles Damais, Julie Dana, Marie Chantal Messier, Nabil Chaherli, John Nash, Ethel Sennhauser, Marion Le Pommellec, and the LCSAR Team. March 17, 2011 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 1 March 17, 2011 Dear Colleagues : After falling sharply during the global financial crisis—in the second half of 2008—food and oil prices have resumed their upward trend. International food prices have risen almost to the levels of 2008 and some commodities like maize have reached record highs. Virtually all the commodity that matter for LAC countries are partaking in this strong wave of price increases. Moreover, the rate of price increases has accelerated in the last three to four months, with food price indices reaching roughly the same level as at their previous peak, in 2008. This increase in food prices presents some great challenges for some LAC countries, in particular in the Caribbean, but it also presents a great opportunity, as many LAC countries are net food exporters and are a food source for other Regions in the world. It is within this context that we are launching the Sustainable Development Occasional Paper Se- ries on Food Prices . We hope that this will contribute to add to the knowledge and exchange of innovative experiences in food policy and programs in LAC. The Occasional Paper Series on Food Prices is expected to include country-specific as well as regional analytical work related to food, logistics, and agriculture policy, and will seek to learn from the 2007-2008 food price crisis and put forward innovative concepts for improving the efficiency of food markets and for reducing the vulnerability to exoge- nous shocks in the food production and trade in the LAC Region. The series starts with papers on: (i) the impact evalu- ation of the 2008 food price subsidy on the rice sector of Haiti, (ii) an analysis of the transmission of international food prices to domestic markets in Central America, (iii) an assessment of the conditions for developing agriculture com- modity exchanges in LAC, (iv) a policy guidance for improving logistics and transport efficiency in the context of food prices, and (v) an analysis on logistics and grains in Argentina. We hope to continue publishing more papers soon to provide additional input to the debate as we take on the challenges and opportunities of the new dynamics in international food markets in the Region and elsewhere. Sincerely,   Laura Tuck Ethel Sennhauser Director Sector Manager Sustainable Development Department Agriculture and Rural Development Aurelio Menendez Jordan Schwartz Sector Manager Lead Economist Transport Sector Sustainable Development Department 2 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy TABLE of CONTENTS I. Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 II. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 III. The Relationship between International and Domestic Rice Prices . . . . . . . . . . . . . . . . . . . . 8 IV. Estimating the Impact of the Rice Price Subsidy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 V. Main Conclusions and Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 3 I. Rice Prices in Haiti Executive Summary Haiti is an important rice consumer, and a big rice import- sions. However it also produced a medium term increase (and not a reduction) in rice prices in local markets. By er. Around 86 percent of the Haitian population consumes estimating the relationship between international and rice. The poor urban families spend relatively more in rice local prices of imported rice before the rice subsidy pro- as a percentage of household income than other Haitian gram was implemented, this study was able to establish households. Haiti imports 70% of the rice it consumes. what would have been the domestic price of imported Before the food crisis (declared worldwide in March rice without the subsidy program and compare it with 2008) the price of rice in the local market was directly de- 1 the actual observed price. Before the announcement of termined by the import price, with no distortions. Haitian the rice price subsidy, the international price of rice (FOB rice importers purchase rice on a FOB Gulf basis. These Gulf) determined the domestic price at a statistical con- FOB Gulf prices went from 17.75 to 43 US$/cwt between fidence level of over 99%. Using the pre-subsidy relation- mid 2007 and the beginning of the crisis. This increase was ship between international and domestic rice prices, the immediately reflected in the prices in Haiti’s local markets, analysis estimates that through June 2009 -- 1 year after which nearly doubled during the same period, corrobo- the subsidy program was concluded --rice prices paid rating the high correlation between the FOB and the by consumers were higher than they would have been domestic prices mentioned above. without the subsidy program. This excess was estimated to be 10% in June. Between April 2008 and June 2009, the The food crisis and the rice price Subsidy overall estimated surcharge to Haitian rice consumers Program was of approximately US$23 million. Possible explana- As a consequence of the crisis, food price inflation more tions for such an unintended consequence of the subsidy than tripled in Haiti from 6.4% in July 2007 to 20% in March program could be: (i) pricing decisions of rice importers 2008. This led the government to announce a subsidy have become more consultative, resulting in an oligopo- to reduce the price of rice to consumers to a fixed level listic element in the prices; and/or (ii) expectations of of 43 US$ per 50kg bag to control social unrest. The rice future losses from Government intervention policies in the price subsidy program was set as a temporary subsidy for rice market have been factored into profit margins of rice rice importers planned for 6 months (April – September importers and/or distributors in Haiti. 2008) and was designed to maintain the selling price from In the context of providing food subsidies to poor rice importers to distributors at 43 US$ per 50kg bag. The displaced households after the 2010 earthquake, public expectation was that rice distributors would then also policies could be designed to avoid such unintended maintain low prices to consumers. The expected impact consequences from happening again in the future. A sought by the Government of Haiti, was a drop in the direct subsidy to the targeted consumers (i.e. food vouch- price of imported rice in the local market. No subsidies ers) could have minimized distortionary effects in the were given to the price of locally produced rice. While the domestic rice market and directly boosted purchasing program was on-going, the subsidy period was revised power of low-income families. Using the US$17million to down to 5 months (ending in August 2008), with US$ 17 directly subsidize rice consumption of low income families million transferred from the Government to rice importers. through a voucher scheme would have: (i) yielded sav- According to the Central Bank of Haiti, the subsidy was ings of 40% on rice expenditures over a 5 month period; funded by a Budget Support Grant from the World Bank (ii) had an increased impact on nutrition of low-income (US$10 million) and a Grant from CARICOM/PetroCaribe households as it would have allowed consumers to (US$7 million). choose product and quality; and (iii) would have also al- lowed local rice farmers to benefit from the price subsidy Main Findings and Conclusions as market prices would have remained unchanged. The subsidy attenuated social unrest and political ten- 1  Locally produced rice was sold at a premium price given differences in quality with imported rice. 4 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy II. INTRODUCTION International food and oil prices spiked between the end As a consequence of the 2007-2008 Global Food Crisis of 2007 and beginning of 2008 (see Figure 1). Haiti was and the increase in the Food CPI, several protests against importing at that time (by November 2008) 53% of its food the Government took place in Haiti. Violent protests according to the World Food Program (WFP) and the began in early April in the city of Les Cayes and then in National Commission on Food Security (CNSA). In 2008 Port au Prince, the capital. Six (6) people died and several 80% of food imports into Haiti were rice imports, and rice shops were looted. By April 12, 2008, the Senate had given represented 7% of the Consumption Price Index (CPI). a vote of no confidence to Prime Minster Jacques Ed- This meant that the increase in international food prices, ouard Alexis who promptly resigned. Since: (i) 70% of the the 2007-2008 Global Food Crisis, led to an immediate rice consumed in Haiti is imported; (ii) 86% of Haitians con- increase in food inflation, going from 6.4% in July 2007 to sume rice, and (iii) households living on less than 2 dollars 20% in March 2008. a day (first 4 quintiles) consume 75% of the rice (see Table Figure 1 COMMODITY PRICES (5-DAY MOVING AVERAGE) OIL WTI IN CURRENT US$, WHEAT, COPPER AND SOYBEAN ARE INDEX NUMBERS: 1/01/05=100 450 150 Wheat 140 400 130 350 120 110 300 Copper 100 250 90 80 200 70 Soybean 150 60 50 100 Petroleum (rhs) 40 50 30 Jan 05 Mar 05 May 05 Jul 05 Sep 05 Nov 05 Jan 06 Mar 06 May 06 Jul 06 Sep 06 Nov 06 Jan 07 Mar 07 May 07 Jul 07 Sep 07 Nov 07 Jan 08 Mar 08 May 08 Jul 08 Sep 08 Nov 08 Jan 09 Mar 09 Source : Bloomberg Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 5 Table 1: Rice Expenditures by Quintile Quintile % of national rice expenditures 1 (poorest) 15% 2 18% 3 21% 4 21% 5 (richest) 25% Source: Demombynes, G. Calculations of 2001 Household Survey (IHSI.) Notes: Quintiles shown were calculated using per capita household income and were calculated for individuals, ie each quintile includes one-fifth of the population of the individuals. 1 below); the new President Preval immediately reacted, sought external financing to keep domestic rice prices announcing the (indirect) fixing of the price of rice to Hai- from further increasing for 6 months (April to October tian consumers by providing a subsidy to rice importers. 2008), and the Ministry of Economy and Finance (MEF) This was done through private-public agreements on a was able to secure budget support from the World Bank fixed price to be sold to local rice distributors. The decision and CARICOM of US$10 and 7 million respectively. to implement an indirect subsidy was made based on the The Presidential Commission met every 3 to 4 weeks to fact (as is shown in the next section) that import prices report on the current rice stocks and incoming vessels with had in the past been transmitted fully and immediately rice, and to agree on a subsidy level per vessel in order to rice consumers. Thus, a subsidy to the price of rice at to ensure a fixed price to be sold to local rice distributors. the level of the importers was expected to be passed on With the signed minutes of the meetings agreement of the immediately into benefits (savings) to rice consumers. Presidential Commission, each importer had to present The Government announced the subsidy after April documentation to the Central Bank of Haiti (BRH), show- 12, 2008, fixing (indirectly) a price between rice importers ing that the rice was in fact imported into the country and and local distributors at US$43/50kg bag. This was the av- sold to local distributors at the agreed price. Box I shows erage price of imported rice sold to consumers observed the details on the information collected during the first in Port au Prince since 2006. The main objective was to meeting of the Presidential Commission on April 14, 2008. reduce social tensions among the low income urban As Figure 4 shows, the price to consumers never reached population. The Government also prohibited rice exports the intended target price of US$43/50kg (equivalent of to the Dominican Republic in order to avoid re-export of HTG94/6lbs) bag set by the Government, and there was subsidized rice. The subsidy scheme was implemented by no way for the Government to enforce the final price be- forming a public-private sector Presidential Commission ing sold to consumers. between the Central Bank of Haiti (BRH), the President’s Office and rice importers. The subsidy scheme initially 6 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy Box I – Details on the structure of rice price subsidies Subsidies to food importers were agreed on individually, importer by importer, according to the price of the rice stored in Haiti as well as the price paid for the rice in transit to Haiti. The documents necessary for the rice importers to obtain the subsidy were: (i) the customs and shipping documents; and (ii) the signed copy of the minutes of the meeting of the Presidential Commission mentioning that specific shipment to be subsidized. Here below is an example of the minutes of the first meeting with the details of the subsidy of the rice in storage by importers (Table 1) and the rice in transit to Haiti (Table 2) Table 1. Subsidies for Stored Rice by Importers (April 14, 2008) Importateur Volume souscrit Volume Pieces soumises Subvention par sac Montant de la (PV 14 et 15 avril déclaré(Base Subvention par sac Subvention dû à 2008) pieces soumises) l’importateur connaissement bordereau de maritime douane MANUTHOR S.A. 2100 TM (42,000 2086.0 TM (41.738 ok ok $5.00 $208,690.00 sacs) sacs) NABATCO 3869,9 TM (77398 5125.4 TM ok ok $5.00 $386,990.00 sacs) (102508 sacs) RICECO PAP 6300 TM (126 000 7770.4 TM ok ok $5.00 $630,000.00 sacs) (155408 sacs) RICECO PAP 2100 TM (42 000 Docs incomplets ok $3.00 $126,000.00 sacs) TCHAKO S.A. 14275 TM (285 11405 TM (228 ok ok $3.00 $684,300.00 500 sacs) 100 sacs) LAROCHE CAP 850 TM (17000 873.95 TM (17479 ok ok $4.00 $68,000.00 sacs) sacs) LUCKY CAP 6206 TM (124 120 Docs manquants Aucun Aucun $4.00 $496,480.00 sacs) $2,600,460.00 Table 2: Subsidies for Import Rice in Transit (April 14, 2008) IMPORTER VOLUME FROM US (Metric Tons) ACRA 7200 (purchased price USD 56-57, arrival july 08), subsidy USD14-15/bag Tchako S.A. 11500 (purchased price USD 51) , subsidy USD7/bag ACRA 11500 (purchased price USD 51, subsidy USD9/bag Rice co, S.A. 4000 (purchased price USD 51), subsidy USD9/bag Rice co, S.A. 1500 (purchased price USD 51), subsidy USD9/bag Manuthor S.A. 6300 (purchased price USD 51), subsidy USD9/bag Manuthor S.A. 2700 (purchased price USD 56), subsidy USD13/bag Manuthor S.A. 2000 (purchased price USD 51), subsidy USD9/bag Tchako S.A. 11500 (purchased price USD 55-56), subsidy USD13-14/bag Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 7 III. the RELATIONSHIP BETWEEN INTERNATIONAL and DOMESTIC RICE PRICES A. International Rice Prices Government subsidy was in place during a period when Haiti mainly imports its rice from the US, and therefore the international prices were decreasing. international reference price for the rice imported into Figure 2 also shows the CIF prices at the Port au Prince Haiti is FOB Gulf. Figure 2 below shows that the internation- port. These are the prices recorded at customs when the al price increase started at the end of 2007 and reached rice was imported into the country. One can observe a its maximum point in May 2008. It is important to note strong correlation between FOB Gulf and CIF PAP rice that when the Haitian Government implemented the rice prices with a lag of approximately 4 to 5 months. This is price subsidy, this was the time period when international the time taken for the purchase to be made, the ship- the rice price was at its maximum, meaning that the ping and handling of the rice from the US to Haiti and Figure 2: Evolution of International Rice Prices - FOB Gulf (USD/cwt*) and CIF Haiti (USD/MT) INTERNATIONAL RICE PRICES (FOB GULF VS. CIF HAITI) 1600 50.00 45.00 1400 40.00 1200 35.00 USD/MT 1000 30.00 25.00 800 20.00 600 15.00 400 10.00 5.00 200 0.00 0 6/1/2006 2/1/2006 4/1/2006 8/1/2006 10/1/2006 12/1/2006 2/1/2007 4/1/2007 6/1/2007 8/1/2007 10/1/2007 12/1/2007 2/1/2008 4/1/2008 6/1/2008 8/1/2008 10/1/2008 12/1/2008 4/1/2009 6/1/2009 2/1/2009 CIF Port au Prince (USD/MT) FOB GULF (USD/cwt) Source : Creed Rice Market Report (2009) and SGS (2009). (*) cwt = 112 lbs. 8 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy the final declaration of the merchandise at the port. This and it is interesting to note the evolution of prices before, strong relationship between FOB and CIF prices is an initial during and after the price subsidy (two vertical lines). Be- indication that in Haiti, movements in domestic prices fore the subsidy, local rice had a substantial premium over of imported rice can be attributed to movements in the imported rice (almost 50%). However during the subsidy FOB Gulf price. In order to better assess this relationship, a program was in place, this premium decreased, and after closer look at domestic rice prices in Haiti is needed (see the subsidy program ended, we observe that domestic Section c. below). rice prices decreased and that imported rice became equally or more expensive than domestic rice. B. Domestic Rice Prices The objective of this study is not to analyze the rela- There are two types of rice sold in Haiti, imported and tionship between locally produced rice and imported domestic (locally produced) rice. Imported rice repre- rice, but between international and domestic prices of sents 70% of the rice consumed in Haiti, however there imported rice. However, it is important to note that there are significant quality differences between the imported 2 are important implications in terms of the impact of the and domestic rice. This means Haitian consumers do not subsidy program on domestic rice production. In fact, it substitute easily one type of rice for another, in particular appears that the subsidy has caused a reduction in the households in urban areas with relatively higher incomes. premium charged for the locally produced rice. Although Imported rice prices more than doubled between it is not the scope of this analysis, it would be important December 2007 and August 2008 while domestic rice to see how this subsidy program impacted the local rice prices remained fairly stable. Figure 3 shows the evolution production. of domestic rice prices. The rice price subsidy was imple- Prices of imported rice vary within Haiti, but are highly mented from April 12 to mid August 2008 (see vertical lines correlated across the country (see Table 2). Correlations in Figure 3), and the target price announced by the Gov- are all above 0.68, with many above 0.803. For the pur- ernment was US$43/bag, which translates to HTG94/6lb pose of this study, we have used the prices from the cities or US$2.35/6lb (horizontal dotted line in Figure 3). Figure 3 of Cap Haitien, Les Cayes and Port au Prince as repre- shows the local and imported price of rice in Cap Haitian, senting national domestic prices of imported rice. These Figure 3 DOMESTIC VS IMPORTED RICE PRICES IN CAP-HAITIEN (HTG/6LB) 250 200 150 100 50 0 23 avril 04 21 juin 04 16 aout 04 24 avril 06 12 octobre 04 7 decembre 04 7 novembre 05 14 aout 06 9 octobre 06 28 mars 05 23 mai 05 18 juillet 05 12 septembre 05 3 janvier 06 27 fevrier 06 19 juin 06 4 decembre 06 21 mai 07 10 septembre 07 5 novembre 07 31 decembre 07 11 aout 08 6 octobre 08 31 janvier 05 29 janvier 07 26 mars 07 16 juillet 07 25 fevrier 08 21 avril 08 16 juin 08 1 decembre 08 26 janvier 09 23 mars 09 18 mai 09 Imported rice Local rice Subsidized price Source : CNSA (2009) 2  he main difference between domestic and imported rice is that Haitians prefer the taste of local rice, as well as the % of broken rice is lower than T the imported rice. 3 Data is from the National Food Security Commission (CNSA) of Haiti, which publishes biweekly data on its website: www.cnsahaiti.org.  Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 9 Table 2: Correlations of Domestic Prices of Imported Rice in Major Urban Cen- ters in Haiti (2004-2009) Port-au- Cap- Haitien Cayes Jeremie Gonaives Jacmel Hinche Port de Paix Ouana- Prince minthe Port-au- 1.00 0.95 0.97 0.93 0.85 0.95 0.94 0.81 0.94 Prince Cap- Haitien 1.00 0.95 0.93 0.74 0.92 0.95 0.82 0.93 Cayes 1.00 0.93 0.79 0.94 0.93 0.78 0.94 Jeremie 1.00 0.68 0.94 0.93 0.78 0.93 Gonaives 1.00 0.77 0.70 0.74 0.81 Jacmel 1.00 0.94 0.83 0.94 Hinche 1.00 0.76 0.93 Port de Paix 1.00 0.86 Ouanaminthe 1.00 cities were chosen because: (i) these are 3 major cities to the volumes imported. During the period of the subsidy for which data was mostly complete; (ii) these three cities program, a total of 107,000 MT of rice were imported, have a high correlation of domestic prices of imported which represents an average of 27,000MT/month. Some rice with the rest of the cities within Haiti; and (iii) there importers have reported that approximately 5% of the are a few importers in Cap Haitian and Port au Prince, the imported (subsidized) rice did cross the border into the two largest cities of the country, who have anecdotally Dominican Republic, even with the re-export prohibi- been identified as the domestic market “price setters’’. tions. But since there is no verifiable way of documenting Unfortunately, rice consumption volumes are not the flows to the Dominican Republic during the period available by city, so a weighted average at the national (or even after), and given the estimated relative low level is not possible. Nevertheless, we have estimated vol- volumes, these adjustments were not considered in the umes of imported rice consumed within Haiti according analysis. Figure 4 INTERNATIONAL VS. DOMESTIC RICE PRICES IN HAITI 6.00 5.00 4.00 USD/6lb 3.00 2.00 1.00 0.00 1/5/2005 1/5/2006 1/5/2007 1/5/2008 1/5/2009 Domestic Rice (Imported @PAP) FOB Gulf Source : CNSA (2009) and Creed Rice Market Report (2009) 10 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy C. Estimating the relationship between and b is the coefficient that explains the degree of inte- international and domestic rice prices gration between markets. If b is close to 1, this represents Figure 4 shows the international (FOB Gulf) and domestic a very strong relationship between markets (perfect price (at Port au Prince) import rice prices. The Figure points transmission), while if b is close to 0, there’s little relation- out by vertical lines the beginning and end of the period ship. Since rice importers sell the rice locally at the cost of of the subsidy program. It appears that the relationship replacement of their stock, there is no time lag between between international and domestic rice prices changed international and domestic prices. Even though Figure 2 during the subsidy period. shows a lag between FOB Gulf and CIF Haiti prices, do- In order to determine the relationship between inter- mestic prices of imported rice follow the level of FOB Gulf national prices and domestic prices of imported rice in prices as importers charge the price needed to secure Haiti before the subsidy program, we used the following new imports (replacement cost). function following the law of one price4 : Based on equation (1) we ran a regression for the pe- riod of January 2005 to April 2008 (before the subsidy pro- Pd = a + bPi (1) gram began) for three cities within Haiti: Port au Prince, Les Cayes and Cape Haitian. The results of the regression Where Pd is the domestic price of imported rice (as are in Table 3 below. We observe that the coefficients are reported by CNSA), Pi is the international price of rice pur- statistically significant and the R2 are quite large, signal- chased by Haitian importers (FOB Gulf), a is the constant ing that the estimates produced by the equation and measuring average logistic and import costs between the coefficients results are very accurate in determining the US market (FOB Gulf) and the different cities within Haiti, actual domestic price of imported rice in Haiti. Table 3: Regression coefficient estimates of equation (1) a b t stat R2 Port au Prince 0.44 1.75 35.07 0.88 Les Cayes 0.48 1.80 24.39 0.79 Cap Haitien 0.60 1.83 31.20 0.86 Author’s calculations 4 W  e used a simple OLS because 70 to 80% of the rice consumed in Haiti is imported rice coming from the US and bought at FOB Gulf quoted prices. After discussing with importers, they also mentioned that the price they charged distributors (and thus final consumers) of rice was the replace- ment cost. This led us to test for a simple regression on absolute changes in prices, where the constant (alpha) in the OLS regression was seen as the transport/logistics/import costs and the dependent variable coefficient (beta) the import tariff and % margin charged by importers and price distributors. After testing for different lags, it was clear that the strongest relationship existed with no lag (proving the replacement cost descrip- tion). The OLS presented high R2 and statistically significant (over 99%) coefficients, so we did not go further into looking for further methods for testing time series. In other words, the simple OLS immediately showed the strong and direct relationship of international and domestic prices of rice before the subsidy. Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 11 IV. ESTIMATING the IMPACT of the RICE PRICE SUBSIDY Looking at simple correlations between international and In order to measure the impact of the subsidy pro- domestic prices before, after, and during the subsidy gram, we estimate the savings to Haitian consumers by program, it is interesting to note (see Figure 5) that the cor- comparing the actual price of imported rice observed relation becomes negative during the subsidy program. in the three selected cities during the subsidy program In theory, if the subsidy program fixed the domestic price with the estimated domestic price calculated based on of rice, the correlation should be close to 0. The fact that the international price at the time following equation (1) the correlation is negative and highly negative, points out and the coefficients in Table 3. This enables us to estimate to the fact that not only domestic prices were decreasing what would have been the domestic price of imported when international prices were increasing, but the con- rice without the subsidy program and compare it with the trary as well: domestic prices were increasing when inter- actual observed price. Figure 6 shows both estimated and national ones were decreasing. This is reflected in Figure 4, actual domestic prices of imported rice before, during where we observe towards the end of the subsidy period and after the subsidy program. that international prices were decreasing while local ones were increasing. Figure 5: Correlation between international and domestic rice prices in Haiti be- fore, during and after the subsidy program (2005-2009) 1.50 1.00 FOB Gulf vs P-au-P 0.50 0.00 FOB Gulf vs Cap-H -0.50 FOB Gulf vs Cayes -1.00 All the period Before subs During subs. After subs 12 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy Figure 6 PRICE OF IMPORTED RICE IN P-AU-P (USD/6LB) 6.00 5.00 4.00 B USD/6lb A 3.00 2.00 1.00 0.00 1/5/2005 1/5/2006 1/5/2007 1/5/2008 1/5/2009 Price of imported rice in P-au-P (USD/6lb) Price estimated of rice imported in P-au-P (USD/6lb) By measuring the difference between the actual and multiply the difference between actual observed domes- estimated price of imported rice, we are able to obtain tic prices and estimated prices (prices estimated using the benefits (or costs) of the subsidy program to Haitian the pre-subsidy relationship between international and rice consumers. Figure 6 shows that during the first part of domestic prices) by the volume of rice imported during the subsidy program, actual domestic prices were below that period. Although there is no way to verify actual rice estimated prices, meaning that the subsidy program volumes sold to Haitian consumers each week, we as- enabled Haitian consumers to buy rice at prices that sume that consumers purchased the same total amount were lower than would have otherwise been without any of rice every week (during 2008 the average national subsidy. However, towards the end of the subsidy period, weekly consumption of imported rice was 6100MT). In actual prices rise above the estimated price. This situa- other words, we assume a low price elasticity of imported tion continued up to June 2009, although prices gradually rice demand. However, this is likely not the case. Low converged toward the end of the period. This means that income consumers tend to have relatively higher price Haitian consumers, since the end of the subsidy period, elasticity of demand of those products that compose a and up to the middle of 2009, were purchasing imported large portion of their consumption basket as they are very rice at prices above what they would have paid without sensitive to such prices changes. Therefore, the estimates the subsidy program. Also, Figure 6 shows that actual ob- on the benefits of the rice price subsidy are likely to be served prices of imported rice after the subsidy program over-estimated if in fact the volumes of imported rice sold was implemented displayed a larger volatility, implying were reduced given the drop in demand. an increase in price uncertainty by rice consumers. The Table 4 below shows the estimated price difference standard deviation of prices of imported rice before the (benefits/costs) in the sample cities and the estimated subsidy program was 0.42 US$/6lbs, but after the sub- national value of consumer savings (costs) of purchased sidy program, the standard deviation increased to 0.68 rice. We observe that during the period of the subsidy, US$/6lbs. The estimated price of imported rice (without consumers saved a total of approximately US$17million the subsidy program) displayed a lower standard devia- (area A in Figure 6), while after the subsidy program end- tion of 0.62 US$/6lbs. ed, the cost to consumers has been US$40million (area B In order to estimate the total value of savings or ad- in Figure 6). ditional costs of the subsidy to Haitian consumers, we Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 13 Table 4: Estimated Consumer Savings (costs) from the Rice Subsidy Program Average Consumer Savings Total consumer savings (USD/6lbs) (USD) During the Subsidy Period 0.48 17,135,276 (April-August 2008) After the Subsidy Period -0.46 (40,359,616) (September,2008-June 2009) Total -0.19 (23,224,339) The analysis shows that as of June 2009, domestic prices among them (collusion) on the prices to be sold to dis- of imported rice were still 10% higher than they would tributors, which could have very well continued after the have otherwise been without the subsidy program. Since subsidy program ended. Another potential cause is that the implementation of the subsidy program, Haitian rice importers and/or distributors have been charging a consumers paid US$23million more for imported rice than premium to consumers in order to make-up for past and/ they would have paid if the program had not been imple- or potential future Government interventions in the do- mented (area A-B in Figure 6). The subsidy to the price of mestic rice market. Many rice importers complained that rice produced consumers’ savings of up to 30% and con- although they did receive the subsidy payment, these sumer over-costs of up to 33% after the end of the subsidy were not easy to obtain, and payments were received period (see Figure 7). In addition to the cost to consum- late and after much bureaucratic processes that cost ers, one must incorporate the cost to Haitian taxpayers of importers’ resources and time. Nevertheless, the main US$17million which were transferred to rice importers by welfare benefits from the program were the subsidence the Government during that same period. of social unrest (although this is questionable given that A potential cause for the higher than estimated do- international rice prices did decline after the April 2008 mestic prices of imported rice is that, during and after the violent protests); and the apparent benefits to importers subsidy program, the rice importers were able to com- from the subsidies. pare prices of incoming shipments and reach agreements Figure 7: Consumer Savings (over costs) on Rice Purchases in Haiti (April, 2008 to June 2009) 40% 30% 20% 10% 0% 08 08 08 08 08 08 08 08 08 09 09 09 09 09 -10% 20 20 20 20 20 20 20 20 20 20 20 20 20 20 / / / / / / / / / / / 6/ 6/ 6/ 16 16 16 16 16 16 16 16 16 16 16 -20% /1 /1 /1 1/ 4/ 5/ 2/ 6/ 9/ 3/ 7/ 4/ 8/ 5/ 11 12 10 -30% -40% 14 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy V. MAIN CONCLUSIONS and POLICY IMPLICATIONS The findings of this study show that although the rice actors of the rice import supply chain, potentially beyond price subsidy program did produce the intended sav- the end of the subsidy program. ings to Haitian consumers during the 4-month period of These findings point to the need to review public poli- the subsidy program (April-August 2008), this intervention cies and programs that intend to lower food costs to low- caused medium term distortions in the domestic market income consumers. Although at the time of the 2007-2008 of imported rice such that domestic prices of imported Global Food Crisis, providing a subsidy to Haitian Rice rice have risen beyond the price that consumers would Importers seemed to be an efficient and quick mecha- have faced without a subsidy program. Furthermore, the nism to lower domestic rice prices, it proved to distort lo- actual prices consumers faced after the subsidy program cal rice markets and produced an over-cost of imported was implemented were much more volatile than the esti- rice purchases, hurting low income households relatively mated price without a subsidy program, pointing to also more. A direct transfer from taxpayers to rice consum- an increased consumer uncertainty about local market ers would have yielded better results. The US$17million of prices for rice. The estimated domestic prices of imported public resources spent in the subsidy program would have rice in a “without subsidy� scenario are based on statisti- translated in a food voucher of over US$4/month5 per cally robust parameters calculated based on the relation- household for 5 months for the families living under the 2 ship of international and domestic rice prices before the dollars/person/day poverty line. This US$4/month repre- subsidy program was put in place. sents 40% of monthly rice expenditures of a low-income There are several potential explanations for this price household in Haiti. behavior post-subsidy, but two of the most likely accord- Using a targeted food voucher as a subsidy mecha- ing to interviews undertaken in August 2009 with the main nism is not only more efficient, but can have higher rice importers are: (i) that price importers and/or distribu- impact on nutrition as it can be used for a wide variety tors have (across the board) increased margins in order of food products. Of course, a food voucher would have to either recover past losses during the subsidy scheme needed an institutional delivery mechanism to be in and/or build additional margins in case the Government place before the 2007-2008 Global Food Crisis. This under- intervenes in the market again; and/or (ii) the creation scores the importance of having such delivery mecha- of the presidential commission where rice importers were nism in place now given the expected medium to long encouraged to share market price and volume informa- term food and nutritional needs of the Haitian population tion led to practices of collusion and price fixing among in the aftermath of the recent earthquake. 5 This estimate includes the overhead cost of running the program, estimated at below 10% as per international experience.  Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy 15 Bibliography Arias, Diego; Peña, Hector; Nguyen, Giau; and Stutton, Christopher (2008). “Haiti’s Food Support Structure: Recommen- dations for Overcoming the Food Crisis.� Latin American and the Caribbean Economist Association Review. Banque de la République d’Haïti – BRH (2008). « Subventions du Riz au Consommateurs Haïtiens via les Importateurs: Mise en Œuvre et Financement. » Report. FAO (2006). Agriculture Price Transmission in Latin America and the Caribbean in the Context of Trade Liberalization. Santiago, Chile (www.rlc.fao.org/prior/desrural/fao-bid/). Foellmi R. and J. Zweimueller (2004). “Income distribution and Macroeconomics: The role of market product power.� Karagiannis G., T. Palivos, and C. Papageorgiou (2004). “Variable Elasticity of Substitution and Economic Growth: Theory and Evidence.� Peña, Hector, y Arias, Diego (2010). An assessment of price wedges in Central America. Draft. 16 Unintended Consequences of Food Subsidies: The Case of the Haiti Rice Subsidy About the series: The LCSSD Occasional Paper Series is a publication of the Sustainable Development Depart- ment (LCSSD) in the World Bank’s Latin America and the Caribbean Region. The papers in this series are the result of economic and technical research conducted by members of the LCSSD community. The series addresses issues that are relevant to the region’s environmental and social sustainability; water, urban, energy and transport sector development; agriculture, forestry and rural development; as well as cross-cutting topics related to sustainable develop- ment such as climate change; logistics; crime and violence; and spatial economics. While all papers in this series are peer reviewed and cleared by the LCSSD Economics Unit on behalf of the Director of LCSSD, the findings, interpretations, and conclusions expressed in this paper, as in all publications of the LCSSD Occasional Paper Series, are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations or to members of its Board of Executive Directors or the countries they repre- sent. The World Bank does not garantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. For environmental sustainability reasons as well as ease of distribution, the primary vehicle for the dissemination of this series is electronic, with papers available on the following web- site: www.worldbank.org/lcssdpapersseries