89687 MYANMAR: RICE PRICE VOLATILITY AND POVERTY REDUCTION Economic and Sector Work, Report No. 89687-MM MYANMAR: RICE PRICE VOLATILITY AND POVERTY REDUCTION October 4, 2014 Agriculture Global Practice East Asia and Pacific Region © 2014 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW, Washington, DC 20433 All rights reserved: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permission The material in this publication is copyrighted. Copying or transmitting any of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. • For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. • All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Contents LISTS OF BOXES, FIGURES AND TABLES ii ACRONYMS AND ABBREVIATIONS iv ACKNOWLEDGMENTS v EXECUTIVE SUMMARY xiv 1. INTRODUCTION 1 2. RICE PRICES AND POVERTY IN MYANMAR 3 3. RICE PRICE LEVELS AND VOLATILITY IN MYANMAR 6 4. DRIVERS OF RICE PRICE VOLATILITY IN MYANMAR 14 4.1 Volatility of world market prices 14 4.2 High seasonality of paddy production in Myanmar 17 4.3 Weak domestic market integration 19 4.4 Poor market transparency 21 4.5 Uncertainty over private rice stocks 23 4.6 Low diversification of exports 24 4.7 Uncertainty over export policy 25 5. INTERNATIONAL EXPERIENCE FROM MANAGEMENT OF RICE PRICE VOLATILITY 26 5.1 Government owned stocks 26 5.2 Export trade restrictions 30 5.3 Minimum farm prices 31 5.4 Safety nets 33 6. CONCLUSIONS AND RECOMMENDATIONS 35 7. REFERENCES 38 8. ANNEXES 41 8.1 Annex 1: Rice Prices in Myanmar 41 8.2 Annex 2: Price Volatility 53 8.3 Annex 3: Rice Market Integration in Myanmar 55 i Lists of Boxes, Figures and Tables Boxes Box 1: Alternative Methods for Measuring Price Volatility 7 Box 2: Examples of Price Transmission Elasticities in Asia 16 Box 3: Rice Market Segmentation in Myanmar 21 Box 4: Myanmar: Private Public Partnership in Rice Stock Management 30 Box 5: Lao PDR: Minimum Prices for Swine Products 33 Figures Figure 1: Myanmar: Nominal prices of Pawsan, Kyats/ton 7 Figure 2: Myanmar: Average nominal and real prices of Pawsan, Kyats/ton 8 Figure 3: Myanmar: Nominal prices of Manawthukha, Kyats/ton 9 Figure 4: Myanmar: Average nominal and real prices of Manawthukha, Kyats/ton 9 Figure 5: Myanmar: Nominal prices of Emata and Manawthukha in Yangon, Kyats/ton 11 Figure 6: Selected export rice prices, $/ton 13 Figure 7: Rice price volatility in selected Asian countries, % 15 Figure 8: Seasonality of real rice prices in selected countries 17 Figure 9: Myanmar: Seasonality of real rice prices 17 Figure 10: Paddy production by month in selected countries, % 18 Figure 11: Transport costs in selected countries, $ per ton per 100 km 19 Figure 12: US: Wheat stocks-to-use ratio and real prices, Jan. 1990-Aug. 2009 23 Figure 13: Nominal rice prices in selected countries, 2004-2013, $/ton 27 Figure 14: Philippines: Volatility of farm-gate, wholesale and retail prices, 1990-2011 31 Figure 15: Conceptual framework for assessing PT and market integration 56 Tables Table 1: Volatility of world cereal prices, % 5 Table 2: Myanmar: Average volatility of Pawsan prices, % 8 Table 3: Myanmar: Average volatility of Manawthukha prices, % 10 Table 4: Average volatility of low-quality rice in selected countries, % 11 Table 5: Myanmar: Average volatility of rice prices by variety, 2004-2013, % 12 Table 6: Myanmar: Average volatility of selected paddy and rice prices, 2009-2012, % 12 Table 7: Rice prices in selected countries, $/ton 13 Table 8: Myanmar: Formal rice exports by grade, ‘000 tons 15 Table 9: Estimates of the adjustment parameters between Yangon and Vietnam prices 16 Table 10: Myanmar: Ratio of maximum and minimum real prices, Manawthukha rice 17 Table 11: Myanmar: Estimates of the adjustment parameters of Pawsan prices, 2004-2013 20 ii Table 12: Myanmar: Estimates of the adjustment parameters of Manawthukha prices, 2004-2013 20 Table 13: Myanmar: Rice balance data from different sources, 2011/12, ‘000 tons 22 Table 14: Myanmar: Paddy production, million tons 22 Table 15: Myanmar: Formal rice exports by destination, 2005-2012, ‘000 tons 24 Table 16: Myanmar: Exports via border trade, 2009-2013, ‘000 tons 24 Table 17: Food as a share of CPI in selected countries, % 27 Table 18: Rice price volatility and paddy production in selected countries, 2004-2012 28 Table 19: Government spending on grain stocks in selected countries 29 Table 20: Volatility of paddy and rice prices in the Philippines and Thailand, % 32 Table 21: Myanmar: Summary of policy recommendations 37 Table 22: ADF test results: Vietnam and Yangon prices 57 Table 23: Johansen test trace results: Vietnam and Yangon prices 57 Table 24: Estimates of PT elasticities between Myanmar and world markets 57 Table 25: Estimates of the adjustment parameters between Yangon and Vietnam prices 58 Table 26: ADF test results: Domestic prices in Myanmar 58 Table 27: Johansen test trace results: Pawsan prices 59 Table 28: Johansen test trace results: Manawthukha prices 59 Table 29: Estimates of the adjustment parameters of Pawsan prices 60 Table 30: Estimates of the adjustment parameters of Manawthukha prices 60 iii Acronyms and Abbreviations ADB Asian Development Bank ADF Augmented Dickey-Fuller test ASEAN Association of Southeast Asian Nations CPI Consumer price index CSO Central Statistical Office of Myanmar EU European Union FAO Food and Agriculture Organization of the United Nations FAS Foreign Agricultural Service (of the USDA) FOB Free on board GDP Gross domestic product GIEWS Global Information and Early Warning System on Food Security Ha Hectare IMF International Monetary Fund LCU Local currency unit MIS Market Information System MOAI Ministry of Agriculture and Irrigation MOC Ministry of Commerce MWK Manawthukha OLS Ordinary least squares PT Price transmission Ton Metric ton Government Government of the Republic of the Union of Myanmar USDA United States Department of Agriculture VAR Vector autoregressive model VECM Vector error correction model $ United States dollar iv Acknowledgments This report was prepared by the World Bank’s Agricultural Global Practice and East Asia and Pacific Region by a team consisting of Sergiy Zorya (Senior Economist, TTL, GAGDR), David Dawe (Senior Economist, FAO), Theingi Myint (Associate Professor of Yezin Agricultural University), Friederike Greb (Post-Doctoral Researcher, Department of Agricultural Economics and Rural Development, Göttingen University), Stephan von Cramon-Taubadel (Professor of Agricultural Policy, Göttingen University), Daphne Aye (Rice Sector Specialist), and Paavo Eliste (Lead Rural Development Specialist, GAGDR). The task team is grateful to the Ministry of Commerce and the Myanmar Rice Federation for the information, discussions, and support provided during the preparation of the report. The excellent discussions of the initial findings in Nay Pyi Taw and Yangon in March and April 2014, and the final report in Yangon at the Myanmar Rice Federation’s Talk Series in August 2014 are particularly appreciated. The task team is also grateful to: the participants of the seminar in the World Bank Country Office in Yangon on May 5, 2014 and in the World Bank in Washington DC on June 3, 2014 for the excellent discussion and comments; Tom Slayton, an observer who has followed the world rice market for over 35 years, for his inputs on the rice sector; and Puja Vasudeva Dutta (Senior Economist, GSPDR) for her inputs on safety nets. Jose Cuesta Leiva (Senior Economist, GPVDR) and Robert Townsend (Senior Economist, GAGDR) served as peer reviewers. Ulrich Zachau (Country Director, EACTF), Abdoulaye Seck (Country Manager, EACMM), Kanthan Shankar (former Country Manager, EACMM), Nathan Belete (Practice Manager, GAGDR) Julia Fraser (Practice Manager, GEEDR), Louise Scura (Program Leader, EACTF), Steven Jaffee (Lead Rural Development Specialist, GAGDR), and Patrick Labaste (Sector Leader, GAGDR) supported this work and provided guidance during the preparation of the report and its dissemination. Amy Gautam edited the report, Tiina Joosu-Palu prepared various communication materials, and Miss Thiri (Program Assistant, EACMM) and Poonyanuch Chulsukon (Program Assistant, EACTF) provided logistical support. v vi inputs vii viii stocks privatestocks ix privat estocks x phytosanitary social net gov:stocks xi publicstocks social nets govt stocks xii safety net xiii EXECUTIVE SUMMARY Rice price volatility matters for poverty reduction in Myanmar Myanmar is a low-income agrarian country with a high poverty rate. The livelihood of many poor people depends on the performance of agriculture, especially the rice sector. Rice accounts for 70 percent of Myanmar’s total cultivated area and 30 percent of the value of its agricultural production. Increasing returns to rice production will be the key to increasing farm wages and incomes in the short to medium run. Higher rice production will also help maintain low food prices, improve food security, and reduce poverty, as an average household spends 61 percent of total household income on food, and rice is a major component of the food basket. Price fluctuations are a common feature of well-functioning agricultural markets. Price fluctuation should be expected in such markets, since output varies from period to period due to factors such as weather, pests and disease, and because demand and supply are inelastic in the short run. Moreover, some amount of seasonal and spatial price movements should be tolerated, since these usefully signal scarcity in the market and facilitate a supply response, foster arbitrage between surplus and deficit regions, as well as guide post-harvest handling, storage and trade decisions. However, in the case of Myanmar’s rice market, several factors serve to exacerbate price fluctuation and make them more pronounced and unpredictable (volatile) and lead to serious negative impacts for consumers and farmers. Rice price volatility is of concern to the Myanmar government given the high importance of rice for farm incomes and consumer expenditures, and thereby for food security and poverty reduction. On the production side, prices volatility inhibits farmers’ supply response and is a disincentive for greater use of purchased inputs and increased investments. Volatility can also discourage rice-producing farmers to diversify their cropping patterns to high-value crops if they cannot buy cereals for consumption at more predictable prices. On the consumption side, rice price spikes can cause increased food insecurity for those not wealthy enough to maintain consumption levels at the higher prices. For people spending 50 percent of their income on rice, a 20 percent temporary increase in rice prices would lead to an approximate 10 percent decline in effective income. This is a large shock, often equivalent to households’ spending on health and education. A good diagnostic is a precondition to identifying suitable remedies International experience shows that addressing price volatility requires a good diagnostic assessment of the actual situation. Detecting and addressing harmful price volatility as well as differentiating it from anticipated price fluctuations is not easy. In addition, there is a strong interplay between price volatility and price levels, and efforts to achieve price stability often distort price levels, which in turn can undermine the benefits of price stability. The international evidence suggests that stable prices per se do not generate long- term agricultural growth, and indeed can undermine growth if achieved through policies with distortionary side effects. That is why it matters how stable prices are achieved. This report examines the extent and nature of rice price volatility, its main sources, and outlines options for achieving sustainable outcomes in rice markets in Myanmar. The report draws on international experience from Asia and around the world for benchmarking purposes and for possible remedies. xiv Rice prices in Myanmar have been volatile, but the recent increase of price level is an emerging concern Rice prices of two major locally consumed rice varieties in Myanmar – Manawthukha and Pawsan – showed marked volatility during 2004-2013. During this period, the average price volatility was above the level of peer net rice-exporting countries, e.g., Cambodia, Thailand, and Vietnam. The volatility was especially high for Pawsan. Rice price volatility in Myanmar has declined over time with positive developments in the rice market, but remains relatively high. The average price volatility of both Pawsan and Manawthukha was about 35 percent lower during 2009-2013 than during 2004-2013. The recent economic liberalization and promotion of rice exports are likely to have contributed to this lower volatility. Also importantly, the rice market became more organized after the initial liberalization in 2004 when local barriers to trade were removed, helping bring down risks and volatility. Domestic and export rice prices are weakly integrated, largely a reflection of rice market segmentation and high logistical costs in Myanmar. For instance, the price volatility of exported Emata rice in Yangon was lower than that of Vietnamese and Thai export prices. Yet the relatively low volatility of Emata rice was not passed through to Manawthukha and Pawsan. Beyond price volatility, rising price levels are an emerging concern. Emata prices increased by 41 percent in nominal US$ terms between 2009 and 2013, much more than in the peer rice net-exporting countries. The increase of rice prices in Myanmar was a result of the large public wage increases in recent years, the increased cross-border trade in rice with China, and economic transition in general. By the end of 2013, Myanmar rice export prices had essentially equalized with the export prices in Vietnam and Thailand. These higher prices not only put pressure on Myanmar exporters but also call for careful management of price volatility as direct price stabilization measures can further increase prices. Price volatility in Myanmar has been mainly determined by local forces Rice price volatility in local markets in Myanmar is higher than the price volatility in other countries. Myanmar’s status as a net rice exporter does not explain much of its high domestic price volatility. Rice exports only accounted for up to 9 percent of total rice production in recent years. The country is still weakly integrated with the world rice market; i.e., the world price’s volatility is not yet fully passed onto Myanmar. Most drivers of domestic rice price volatility in Myanmar were internal. Seasonal variations explain much of the volatility in rice prices in Myanmar. The high concentration of paddy production in November and December resulted in sharp drops of paddy and rice prices in December and January and sharp rises between May and October, depending on the market. Other factors amplify the seasonal volatility in rice prices. Weak domestic market integration has prevented price signals from being transmitted quickly between different markets. Weak domestic integration is a result of the fragmented rice market, poor roads and low phone connectivity, the improvement of which will take time. But it is also a result of weak market transparency. Data on production, consumption, exports, and stocks, even if available, are not accurate. As a result, market participants and the government often overreact to small changes in prices, which ignite volatility given the small size of private rice stocks. Weak geographic diversification of exports also contributes to price volatility in Myanmar. Most formal exports go to Africa, while China absorbs most of the cross-border exports. The latter are informal and reported to be highly unpredictable. xv Stabilizing rice prices requires long-term investments to address its structural causes Efforts to reduce price volatility focused on the main exacerbating factors would have the best chance of sustainable results, including measures in five broad areas: (i) raising rice productivity; (ii) spreading production more evenly over the year; (iii) improving post production handling, processing and storage; (iv) strengthening domestic market transparency and integration; and (v) adopting policies supportive of exports. Many of these factors require time to have impact on price volatility. But there are a number of policy actions that can be implemented quickly, at no or low cost to the public finance, and with a broad impact on the rice value chain’s participants (these are underlined in the text). These priority actions can help start addressing underlying driving forces of price volatility and lead the way to the needed longer-term investments and policies to significantly reduce volatility. A reduction in price volatility would require both increasing production and spreading production more evenly over the marketing year. This would, in turn, depend on farmers’ improved access to irrigation, seeds with different harvesting periods and growth duration, and farm advice on production technologies. Lower costs of doing business and improved access to finance for rice mills and traders would reduce storage costs and increase private stocks which would smooth price fluctuations. This could be facilitated through making the rice milling industry more attractive for foreign direct investment by removing the industry from the list of strategic sectors requiring special government approval and permitting all types of investments, not just “green” ones. Myanmar needs cost-efficient and competitive mills to turn paddy into valuable rice for both domestic and foreign markets and improve rice quality, and thereby reducing price volatility. A modernization of the rice industry can trigger lasting changes at the farm level (higher productivity, better paddy quality, and more evenly spread production) and allow the industry absorbing a potentially larger paddy production, as has happened in Cambodia, Thailand, and Vietnam. Modernized rice mills will be also better positioned to increase the level of private stocks which would further reduce price volatility. Investments in rural roads and telecom infrastructure would increase market integration: signals would pass more quickly from one market to another, including from world markets with lower price volatility than in Myanmar. Developing efficient transportation services for rural areas would increase rates of return in road infrastructure. More accurate and timely distributed market information would help the private sector and the government to make more rational and informed decisions. Diversification of exports to additional markets – both geographically and by quality – would further contribute to lower price volatility. This would require continuation of an unambiguous open trade policy. It could be further supported by investments in sanitary and phytosanitary infrastructure, ports, and less costly export procedures. While stronger integration of Myanmar with world markets will increase the risk of price volatility from external factors, it will also encourage investments in rice mills, private storage infrastructure, and export logistics, which will lower the risk of price volatility from domestic factors. xvi Short-term rice price stabilization measures should be used with caution since they can have unintended adverse impacts Effectively stabilizing rice prices will take time, resources, and commitment in Myanmar, as elsewhere. There are no shortcuts or substitutes for addressing long-term structural issues through investments in agriculture and infrastructure, an improved investment climate, and social safety nets. All of these policies and investments will indirectly reduce rice price volatility and mitigate the impact of residual price volatility. Lessons of experience in other countries indicate that direct short-term price stabilization measures are not cost-effective and often have unintended adverse impacts. Direct management of short-term rice price volatility rarely produces successful outcomes for food security and poverty reduction. Policy measures such as export restrictions, farm minimum prices, and government-owned stocks, currently being considered in Myanmar, may help reduce some price volatility but their costs will exceed their benefits. In Indonesia and the Philippines, for example, large public stock programs, along with monopoly over rice imports, have succeeded in reducing price volatility for consumers but not for farmers. In addition, these programs have pushed rice prices up, directly increasing urban consumers’ poverty, penalized the private sector, and crowded out more productive agricultural public expenditure. In Thailand, high minimum farm prices introduced in 2011 helped reduce price volatility for farmers but failed to reduce it for consumers and exporters. More importantly, they resulted in high rice prices, large fiscal costs, lower exports, deteriorated rice quality, increased production costs, and many other economic distortions. The takeaway for Myanmar from the experience of other countries is that there is a significant trade- off between lowering price volatility with short-term price stabilization measures and maintaining price (level) competitiveness. Some measures aimed at lowering price volatility such as large public stocks or farm minimum prices often put domestic prices above world market prices, and would thereby impede progress towards Myanmar’s aspiration to become a large exporter. Other measures, such as export restrictions, would also impede export competitiveness by depressing domestic prices below the world market level, hampering investments in rice mills and trade sectors and undermining Myanmar’s reputation as a reliable exporter. Temporary social safety nets have proven to be effective in mitigating the impact of temporary price increases on net consumers Social safety nets can be developed to mitigate the impacts of short-term price volatility on the most vulnerable. When rice prices sharply increase during off-seasons, public work (or cash for work) programs can help address seasonal unemployment and food insecurity in Myanmar. The international experience suggests that social safety nets can deliver good results even in low-income countries but they need to be carefully designed. For example, in many countries cash-based assistance, either through direct transfers to the most vulnerable or through public work programs, has proven to be more effective than food distribution from government-owned stocks at achieving most social and economic objectives. Myanmar would need to begin developing such safety nets as early as possible to be able to effectively mitigate negative impacts of rice price spikes on poor consumer in near future. xvii 1. Introduction 1. Rice is the main food staple in Myanmar. It accounts for 25 percent of the consumption of richer households and 50 percent of the consumption of poorer households. Rice prices affect poverty, which was estimated at 37.5 percent in 2010 (World Bank 2014c). Rice is also Myanmar’s main agricultural product and the second most important exportable agricultural commodity after beans and pulses, and is thus a very important income source for farmers. In 2010-2011, paddy planted area accounted for 70 percent of the total arable area, 30 percent of Myanmar’s gross agricultural output, and 95 percent of total cereal output, according to FAOSTAT. 2. Rice price volatility is therefore a matter of concern for everyone.1 Farmers are concerned about sharp price drops after the harvest and about price uncertainty. Consumers are hurt by price spikes and in general by higher off-season food prices. Exporters can be hurt by domestic price spikes when an export contract is already signed with a fixed price. As is the case in most Asian countries, the Government of the Republic of the Union of Myanmar (government) is concerned about price volatility, considering more stable prices to be a precondition for generating agricultural growth, reducing poverty, and maintaining political stability. 3. Rice price volatility can be driven by external and internal factors. World market price volatility could be partially responsible for the high price volatility in Myanmar. Yet given the (still) relatively weak integration of Myanmar’s markets with world markets, domestic factors are likely to play a much bigger role. 4. There is also a strong interplay between price levels and volatility. Often price stability measures elevate or depress average price levels. Therefore, making rice prices more stable requires a thorough analysis of the factors underpinning volatility and a careful choice of tools to affect it. Given Myanmar’s aspirations to become a large net exporter, stabilization efforts need to keep its rice prices at competitive levels vis-à-vis other exporters. 5. It is worth noting that not all volatility is harmful. Price fluctuations are a common feature of well- functioning agricultural markets. Price fluctuation should be expected in such markets, since output varies from period to period due to factors such as weather, pests and disease, and because demand and supply are inelastic in the short run. Moreover, some amount of seasonal and spatial price movements should be tolerated, since these usefully signal scarcity in the market and facilitate a supply response, foster arbitrage between surplus and deficit regions, as well as guide post-harvest handling, storage and trade decisions. Should volatility be a concern, attention needs to be paid to uncertain price movements, which are difficult to predict and disruptive for economic decisions. 1 In Myanmar, price volatility is also a concern for other agricultural commodities such as beans and pulses. However, for beans and pulses volatility is largely determined by external factors (e.g., price developments in India). Myanmar’s domestic prices closely follow prices in Mumbai. Few policy options are available in the short run to mitigate that volatility, and food security concerns for those crops are usually much lower than for rice. The focus of this report is therefore on rice. 1 6. Not all instruments are able to achieve lower price volatility without undermining food security.2 One lesson from the international experience on price stabilization is that choosing the right instruments is the key to translating price stability into higher food security. For example, more volatile prices combined with market-oriented trade policy are superior to less volatile prices achieved through state trade interventions. In Cambodia, for example, rice output grew by 6 percent annually between 2002 and 2012 compared to only 1 percent annually in Indonesia,3 although prices in Cambodia were more volatile. The difference between these two countries lies in their approaches to rice sector development. Cambodia has been pursuing an open trade policy while Indonesia has been pursuing an inward-oriented policy, with monopoly over trade and with large government-owned stocks. 7. The bottom line is that stable prices per se do not generate growth. This is especially so when the instruments used to achieve stability undermine private investment, worsen food availability and access to food, drain fiscal resources, and crowd out investment in public goods. When this happens, stable prices may actually increase poverty, distort economic decisions, and eventually lower the quality of growth. 8. Against this background, the report seeks to assist the Government of Myanmar in studying the causes of rice price volatility and identifying options for reducing rice price volatility to achieve food security objectives. Chapter 2 explains the role of rice prices (i.e., their level and stability), in reducing poverty and enhancing food security in Myanmar. Chapter 3 reviews rice price developments in Myanmar, including the interplay between rice price level and price volatility. Chapter 4 analyzes the drivers of price volatility, considering both external and internal factors. Chapter 5 presents lessons from international experience, through the lens of their suitability for Myanmar to achieve its food security and poverty reduction objectives. Chapter 6 concludes and provides policy recommendations. 2 Price volatility is only one pillar of food security. Food security encompasses four pillars: food availability, access to food, food use, and stability, according to the 2009 World Summit on Food Security. Availability (supply) and access to food (demand) are of similar and often greater importance than price stability. 3 Estimated using data from the Foreign Agricultural Service (FAS) of the US Department of Agriculture (USDA) at http://apps.fas.usda.gov/psdonline/psdquery.aspx. 2 2. Rice Prices and Poverty in Myanmar 9. Myanmar is a low-income country with a high poverty rate. Poverty is estimated to have been 37.5 percent in 2010 (World Bank 2014c), higher than previously reported (25.6 percent). The majority of the poor (76 percent) live in rural areas. The highest number of poor live in the Delta and the Dry Zone, two major agricultural regions. Many small farmers are poor, but the rural landless are among the poorest. Urban poverty is also high, estimated at 34.6 percent. Even Yangon, the commercial hub of the country, has a poverty rate of 34 percent. 10. Many farmers are poor or vulnerable, living close to the poverty line and at high risk of falling into poverty. Their pathway out of poverty depends on income generation from agriculture. Myanmar is still an agrarian country. According to official national accounts estimates for 2010, the agriculture sector employed 52 percent of the workforce and generated 37 percent of GDP. Rice is the main agricultural product, accounting for about 70 percent of total cultivated area in the country, 30 percent of agricultural production, and 97 percent of total grain production by value in 2010-2011, according to FAOSTAT. Thus, the key to increasing agricultural incomes at this point is increasing returns to rice production. Without higher economic returns to rice production it is hard to imagine a substantial increase in agricultural incomes in the medium run.4 With the lowest average paddy yield in the Association of Southeast Asian Nations (ASEAN) (World Bank 2014b), Myanmar is well positioned to significantly raise yields, thereby increasing the incomes of farmers and the many landless who earn their income from seasonal farm work. 11. Consumers also depend on agriculture through the food price link. The majority of the urban population is net buyers of rice. In addition, rural landless and small farmers with less than 1 hectare (ha) of agricultural land (accounting for about 30 percent of farm households) are also net rice buyers. There is no estimate of the number of net buyers versus net sellers of rice in Myanmar, but based on international experience it is likely that the country has more buyers than sellers. Myanmar consumers spend an average of 61 percent of their total expenditures on food (World Bank 2014c).5 The expenditure share of rice ranges from 25 percent for better off people to 50 percent for the poorest.6 Higher rice prices reduce effective purchasing power, leaving less money available to spend on foods with the essential vitamins, minerals, and amino acids that are lacking in rice. Higher rice prices also result in less money available to spend on other essential non-food items. 12. Agricultural development has been crucial for poverty reduction and improved food security in Asia. Between 1990-1992 and 2010-2012, higher agricultural output was the largest contributor to the reduction of undernourishment in Asia and the large contributor to poverty reduction in the region (Warr 2013). On the other hand, higher food prices significantly increased the rate of undernourishment. Therefore, to improve food security and reduce poverty, higher agricultural production and lower food prices are needed. What was true for many middle-income Asian countries in the past is true for Myanmar today. 4 In contrast to Myanmar, the importance of rice in most Asian countries has reduced over time. Between 1961 and 2007, the share of rice in agricultural production in East Asia declined from 19 percent to 8 percent, while in Southeast Asia this decline was from 40 percent to 32 percent. Jaffe et al. (2014) use the metaphor of rice being produced under the widening shadow of skyscrapers. Many countries therefore have moved from rice- based strategies to reduce food security and poverty. As Myanmar’s situation is different, it still needs to pay high attention to the rice sector to reduce poverty. 5 In Myanmar, the poorest 20 percent of households spend on average 74 percent of total expenditure on food. The richest 10 percent spend 39 percent on food. 6 The 2010 poverty survey does not estimate the share of rice in food expenditures. Anecdotal estimates range between 25 percent and 50 percent. 3 13. Food prices are important for achieving food security and thus reducing poverty. The definition of food security has evolved over time. During the 1950s, food security was generally equated with a secure and adequate food supply. Food aid programs during the 1960s emphasized the need to ensure people’s physical access to food as the expanded supply of food during the Green Revolution did not eliminate persistent hunger and periodic cases of famine (Dawe and Jaffee 2014). Since the 1990s, the concept of food security has evolved beyond the aspects of supply and demand to draw attention to the effective utilization of food – that is its contribution to nutritional status. In the late 2000s, the large volatility of global food prices raised the importance of price volatility in ensuring food security. 14. Rice price volatility affects both supply and demand aspects of food security in Myanmar. On the supply side, farmers benefit from stable prices because they are protected from periods of unusually low prices. With more stable prices, farmers in general can accelerate their supply response through better access to finance, greater use of purchased inputs, and ultimately increased investments. Less volatility can encourage rice-producing farmers to diversify their cropping patterns to high-value crops if they can buy cereals for consumption at more predictable prices. On the demand side, sharp rice price increases can cause food insecurity for those not wealthy enough to maintain consumption levels at the higher prices. Rice is likely to account for 50 percent of expenditures of the poorest quarter of Myanmar’s population. For these people, a 20 percent increase in rice prices would lead to an approximate 10 percent decline in effective income. This is a large shock, often equivalent to their spending on health and education (World Bank 2014c). More stable prices also allow consumers to diversify their diets and increase their intake of proteins, vitamins, and minerals – crucial for reducing malnutrition. 15. In addition, where food is a large share of household expenditures as in Myanmar, less volatile rice prices can generate economy-wide gains. If more stable prices help induce a supply response, this brings the average level of prices down. In return, low rice prices effectively increase real wages for employees without increasing the nominal wages paid in the industrial and service sectors. In conjunction with other factors, the combination of low nominal and high real wages stimulates job creation and economic growth necessary for sustainable poverty alleviation. On the other hand, uncertain prices slow farmers’ supply response, leading to longer periods of lower production and higher foodgrain prices. When rice prices rise, workers in Asian countries, where rice is a major staple, need higher wages to keep real incomes from falling as they pay for more expensive food (Timmer 2004; Dawe and Timmer 2013). As higher nominal wages discourage investments, the end result is often a slowdown in the productivity growth essential for food security and poverty alleviation. 16. The strategic policy question is how to reduce price volatility. This is not a trivial question because many past efforts to stabilize rice prices in Asia have failed to help generate growth and reduce poverty, especially when higher agricultural production was stimulated by artificially high rice prices or when consumers were protected by export restrictions. More stable prices per se do not guarantee the realization of economic gains. Success depends on how stability is achieved and the complementary measures taken. If domestic prices are stabilized for long periods of time without taking into account changes in world market prices, domestic prices often end up either at high levels in net importing countries (above import parity) or at very low levels in net exporting countries (below export parity). When this happens, stable prices may actually increase poverty, distort economic decisions, and eventually lower the quality of growth (World Bank 2012b; Warr 2013). Low rice prices discourage agricultural growth and high rice prices discourage consumption (and directly increase poverty). This strong interlinkage between price level and price volatility explains why more stable prices (food stability) at the expense of higher prices (food affordability) or lower prices (food availability) may actually reduce food security. 17. Some price volatility is common to agricultural markets, for example, seasonal volatility. Smoothing out this volatility with short-term measures is not desirable as it requires long-term structural improvements; for example, production spread out more evenly over the year and low marketing costs. Thus, domestic price volatility close to international price volatility can be acceptable, especially given that global rice price volatility is generally lower than that of global wheat and maize prices (Table 1). 4 Table 1: Volatility of world cereal prices, %7 Wheat Maize Rice, Thailand Rice, Vietnam 2004-2008 7.2 7.5 8.0 5.2 2009-2013 7.6 6.9 4.5 6.1 2004-2013 7.3 7.2 6.5 5.7 Note: Wheat (US), no. 1, hard red winter, ordinary protein, export price delivered at US Gulf ports for prompt or 30-day shipment; Maize (US), no. 2, yellow, free-on-board (FOB) US Gulf ports; Rice (Thailand), 5% broken, white rice, milled, indicative price based on weekly surveys of export transactions, government standard, FOB Bangkok; Rice (Vietnam), 5% broken, FOB Hanoi. Source: World Bank Pink Sheets. 18. Price volatility above the level observed in reference international markets requires the most attention. Such volatility can be reduced but this requires solid analysis of driving forces of volatility and identification of the best remedies. Direct price stabilization usually leads to price distortions; i.e., low prices in the case of export restrictions or high prices in the case of government-owned stock programs. This is harmful to food security in the long run, suggesting that rice price stability needs to be addressed carefully, mostly through indirect measures. 19. It is important that price stability is not equated with price support to either producers or consumers. Farm support can lead to short-term production benefits but in the long run, more rice production can increase farmers’ income and reduce consumers’ food expenditures only when rice production is the most beneficial use of land (i.e., farmers are not forced to grow rice instead of another crop) and is accompanied by lower costs of production (through an increase in total factor productivity). The next chapters go systematically through these issues for Myanmar. 7 Price volatility is defined as the standard deviation of the logarithm of monthly price returns. 5 3. Rice Price Levels and Volatility in Myanmar 20. Rice is not a uniform commodity. There are many varieties of rice across Asia.8 Myanmar also has many rice varieties, which often not only complicates market monitoring and price stabilization but also makes it difficult to aggregate paddy into uniform lots, control rice quality, ensure efficient milling, and trade at a profit. Two rice varieties for Myanmar are selected as representative domestic prices for this report. The first is Pawsan, the most popular fine-quality aromatic rice variety, which is marketed from Pathein, Shwebo, and Yangon to all other markets in Myanmar. However, this variety is not widely produced and is relatively thinly traded. The second is Manawthukha (MWK), a lower-quality Letyezin rice of fair eating quality and reasonable price. It is the most preferred variety of middle-class consumers in the country, produced in all paddy growing areas and marketed across all regions in Myanmar. In addition to MWK and Pawsan, there is Emata,9 which is exported as 25% broken rice. Emata comprises more than 90 percent of Myanmar’s rice exports (see Table 5 in World Bank 2014b) and is used in the report as a representative export price. 21. The monthly wholesale price time series data used for this report cover the period from January 2004 to December 2013. The price data come from the Market Information System (MIS) of the Ministry of Agriculture and Irrigation (MOAI).10 Four domestic markets are included in the analysis.11 Yangon and Pathein/Delta represent rice surplus areas, while Mandalay and Myingyan represent rice deficit areas. Pathein is located in Ayeyarwady Region; Yangon is the central market located in Yangon Region, which is also the major rice market for domestic and international trade. Mandalay is the focal market in central Myanmar and Myingyan is located in upper Mandalay Region. The wholesale prices of Emata are available only in Yangon and they are used to represent the export price.12 22. Paddy price data series are limited and thus used for analysis only in Chapter 3. Paddy price data are from the Ministry of Commerce (MOC) and are available for Emata and MWK, but not for Pawsan. These price data cover Pathein and Yangon, but not Mandalay or Myingyan. Time series data for paddy prices are only available from January 2007 to December 2012. 23. The analysis of domestic prices in Myanmar is conducted in both nominal and real terms. The Myanmar consumer price index (CPI) from the International Monetary Fund (IMF) and the Central Statistical Office (CSO) of Myanmar is used to deflate nominal prices into real ones. Unless otherwise indicated, the volatility of prices in Myanmar is presented in real terms (CPI-adjusted) for comparability of volatility over time. Price volatility in this report is defined as the standard deviation of the logarithm of monthly real price returns (Box 1). International price comparisons are made in US dollars ($), and international price volatility is measured in real local currency units (LCUs). The price data for other countries are from the FAO’s Global Information and Early Warning System on Food Security (FAO-GIEWS). 8 See World Bank (2014b)’s Annex 2 for a description of the complexity of the world rice market. 9 Emata is a grouping of rice varieties based on length-breadth ratio. Other groupings include Ngasein, Letywezin, Medon, and Byat. 10 Annex 1 presents the price data used for this analysis. 11 There are no price time series data available for other markets and types of rice for the period from January 2004 to December 2013. Unfortunately, there are also no price time series data for the Muse market on the border with China. Information on price volatility on that border is important for understanding price volatility across Myanmar. 12 The wholesale Emata Yangon price is used in this report as a proxy for export price. The FOB Yangon export prices were available to the team from May 2011 only. In 2012-2013, the average margin between these two prices was $55 per ton (see Figure 6). Note that actual prices received by exporters may even differ from the reported FOB prices due to the nature of export licensing in Myanmar. The export licenses include information on indicative prices for export transactions, but sometimes traders informally sell rice at different prices depending on the market situation. 6 Box 1: Alternative Methods for Measuring Price Volatility Price volatility can be measured in different ways. One alternative to the indicator used in this report is a relative price spread, defined as the difference between maximum and minimum prices divided by the average price. This definition is criticized for using only two price points out of many available observations. The second alternative is a coefficient of variation, defined as the ratio of the standard deviation to average prices. Sometimes the coefficient of variation is adjusted by a time trend. The first and second indicators tend to underestimate volatility for higher- value commodities compared to lower-value commodities, making it less suitable for the analysis of low- and higher-value rice carried out in this report. 24. The analysis begins with a review of rice price levels in Myanmar. In nominal terms, the prices of Pawsan rose on all markets in Myanmar over the past 10 years. From January 2004 to December 2013, they doubled in Yangon and more than tripled in Myingyan (Figure 1). Typically, prices were lowest in Pathein, the main producing area, and highest in Myingyan, the main deficit area. Prices in Pathein were also the lowest because farmers there are reported to harvest paddy early with high moisture content (to prepare fields for beans and pulses as a second winter crop) and they often include foreign matter to inflate the weight. 25. Pawsan prices were highly seasonal. They followed a predictable, intra-annual pattern. Every year they dropped between December and February and peaked between August and November. During the observed period, maximum prices in real terms were 40 percent above minimum prices on average. Figure 1: Myanmar: Nominal prices of Pawsan, Kyats/ton Figure 1: Myanmar: Nominal prices of Pawsan, Kyats/ton 900,000 800,000 700,000 Myingyan 600,000 Mandalay Prices, Kyats/ton 500,000 Average Yangon 400,000 Pathein 300,000 200,000 100,000 - Ju -04 N -04 Ap -04 Se r-05 Fe -05 Ju 06 D l-06 M c-06 O -07 M t-07 Au -08 Ju -09 N -09 Ap -09 Se r-10 Fe -10 Ju 11 D l-11 M c-11 O -12 M t-12 Au -13 Ja -08 13 b- b- g- n n ov p ay ar n n ov p ay ar g c c e e Ja Source: MIS/MOAI (2014). 7 26. In real terms, the increase in Pawsan prices was much smaller. The simple average price in the four markets increased by 28 percent from January 2004 to December 2013, compared to a 252 percent increase in nominal terms (Figure 2). In the period 2004-2008, real prices increased by 43 percent. However, in the period 2009-2013, they fell by 5 percent. While these are wholesale prices, this suggests that, at least for Pawsan during the past five years, farmers faced constant output prices while production costs, especially labor costs and those of tradable inputs, reportedly grew steadily. Figure 2: Myanmar: Average nominal and real prices of Pawsan, Kyats/ton 900,000 Average Real Prices (in 2013 contant prices) 800,000 700,000 600,000 Prices, Kyats/ton 500,000 400,000 300,000 Average Nominal Prices 200,000 100,000 - 04 04 5 06 06 7 08 08 9 10 10 1 12 12 3 -1 -0 -0 -0 -1 n- p- n- p- n- p- n- p- n- p- ay ay ay ay ay Ja Se Ja Se Ja Se Ja Se Ja Se M M M M M Source: MIS/MOAI (2014), IMF (2014), and CSO (2014). 27. The average volatility of Pawsan prices was 10.5 percent during the observed period (Table 2).13 The price volatility was 30 percent higher during 2004-2008 than during 2009-2013. It peaked in 2008 and then subsequently went down. The lowest volatility was observed in Yangon and the highest in Myingyan. Table 2: Myanmar: Average volatility of Pawsan prices, % Table 2: Myanmar: Average volatility of Pawsan prices, % Time periods Yangon Pathein Mandalay Myingyan Average for all markets Jan. 2004-Dec. 2008 10.4 10.8 12.1 15.5 12.2 Jan. 2009-Dec. 2013 7.6 10.3 6.5 9.6 8.5 Jan. 2004-Dec. 2013 9.1 10.5 9.6 12.9 10.5 Source: Authors’ estimates based on data from MIS/MOAI (2014). 28. The prices of MWK, the medium-quality rice, were half the level of Pawsan prices. But in nominal terms they increased more than the prices of Pawsan. The MWK prices more than tripled in Yangon, Pathein, and Mandalay and increased six-fold in Myingyan (Figure 3). Similar to Pawsan, the average MWK prices were lowest in Pathein and highest in Myingyan during the observed period. 13 Annex 2 presents the tables with price volatility in nominal and real prices in all locations using alternative methods of coefficient of variations and price returns as described in Box 1. 8 Figure 3: Myanmar: Nominal prices of Manawthukha, Kyats/ton 700,000 600,000 500,000 Myingyan Prices, Kyats/ton Mandalay 400,000 Average 300,000 Yangon Pathein 200,000 100,000 - Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Source: MIS/MOAI (2014). 29. In real terms, the price increase in MWK was 51 percent from 2004 to 2013. The increase in real MWK prices in both subperiods, 2004-2008 and 2009-2013, was about the same, at 25 percent (Figure 4). Figure 4: Myanmar: Average nominal and real prices of Manawthukha, Kyats/ton Figure 4: Myanmar: Average nominal and real prices of Manawthukha, Kyats/ton 600,000 Average Real Prices (in 2013 contant prices) 500,000 400,000 Prices, Kyats/ton 300,000 200,000 Average Nominal Prices 100,000 - 04 04 5 06 06 7 8 08 9 10 10 1 12 12 3 -1 -0 -0 -0 -0 -1 n- p- n- p- p- n- p- n- p- ay ay ay n ay ay Ja Se Ja Se Ja Se Ja Se Ja Se M M M M M Source: MIS/MOAI (2014), IMF (2014), and CSO (2014). 9 30. Similar to Pawsan prices, MWK prices exhibited a seasonal pattern. They tended to drop in December and January and peaked between August and November. During 2004-2013, the ratio of maximum to minimum prices in real terms averaged 35 percent. 31. MWK prices were less volatile than Pawsan prices. This could be due to the wider cultivation of MWK in Myanmar. Between January 2004 and December 2008, the volatility of MWK averaged 10.1 percent, declining to 6.3 percent between January 2009 and December 2013 (Table 3). During the whole observed period (January 2004 to December 2013), the average volatility was 8.8 percent, or 20 percent below the average volatility of Pawsan prices. Table 3: Myanmar: Average volatility of Manawthukha prices, % Time periods Yangon Pathein Mandalay Myingyan Average for all markets Jan. 2004-Dec. 2008 11.0 10.2 10.6 10.8 10.6 Jan. 2009-Dec. 2013 5.9 5.7 6.5 8.4 6.7 Jan. 2004-Dec. 2013 8.8 8.2 8.7 9.6 8.8 Source: Authors’ estimates based on data from MIS/MOAI (2014). 32. It is worth noting that the price volatility of both Pawsan and MWK rice has decreased over time. This could be a result of more liberal and organized trade during 2009-2013 compared to 2004-2008. While the rice sector liberalization started in 2004, it took many years to replace the system of trade and price controls and few state trading enterprises with the current system of open borders, no price controls, and many private traders. This transition seems to be complete and is reinforced by the absence of local barriers to trade, market-based price mechanisms, support for cross-border trade, especially with China, and promotion of formal exports. 33. The prices of Emata (25% brokens), the main rice quality exported, closely followed the prices of MWK. MWK was slightly more expensive than Emata. Between January 2004 and December 2006, the price gap between them was 13,000 Kyats per ton. It surged to 84,000 Kyats per ton in 2007-2008, but starting from January 2009 the nominal gap between these prices stabilized at 30,000 Kyats per ton or about 13 percent during 2011-2013 (Figure 5). The gap was closed in December 2013. 10 Figure 5: Myanmar: Nominal prices of Emata and Manawthukha in Yangon, Kyats/ton 600,000 500,000 400,000 Prices, Kyats/ton 300,000 MWK-Yangon Emata-Yangon 200,000 100,000 - Ju -04 N -04 Ap -04 Se r-05 Fe -05 Ju 06 D l-06 M c-06 O -07 M t-07 Au r-08 Ju -09 N -09 Ap -09 Se r-10 Fe -10 Ju 11 D l-11 Ja -08 M c-11 O -12 M t-12 Au r-13 13 b- b- g- n n ov p ay n n ov p ay g c c a a e e Ja Source: MIS/MOAI (2014). 34. The price volatility of Emata appears to be lower than the price volatility of comparable rice types in Cambodia, Thailand, and Vietnam, the top regional exporters. Measured in real LCUs, the average volatility of wholesale prices in all countries was lower in 2009-2013 compared to 2004-2008 (Table 4). During 2009-2013, the volatility of Myanmar’s Emata was the lowest, at 5.1 percent, compared to the price volatility of similar rice at 6.8 percent in Cambodia, and 6.9 percent in Vietnam, and 8.2 percent in Thailand. Table 4: Average volatility of low-quality rice in selected countries, % Time periods Myanmar, Yangon- Cambodia, mixed Thailand, 25% Vietnam, 25% Emata rice brokens brokens Jan. 2004-Dec. 2008 6.0 9.4* 8.7 8.7 Jan. 2009-Dec. 2013 5.1 6.8 8.2 6.9 Note: * From January 2006 to December 2008. For Cambodia, wholesale prices are used due to the lack of the export price time series data for 2004-2013. Thailand and Vietnam prices are export FOB prices. Source: Authors’ estimates based on data from MIS/MOAI (2014), FAO-GIEWS (2014), IMF (2014), and CSO (2014). 35. Among all rice types in Myanmar analyzed in this report, the prices of Emata were the least volatile. They were 57 percent less volatile than MWK and 88 percent less volatile than Pawsan (Table 5). This could be because Emata has been linked to less volatile world market prices, which cannot be said about Pawsan, for example. Pawsan is not yet exported in large quantities and thus Myanmar cannot “import” lower volatility from the world market. It is not even traded domestically in large quantities; of all the reviewed rice varieties, Pawsan is the most exposed to local shocks, which are likely to cause its higher volatility. 11 Table 5: Myanmar: Average volatility of rice prices by variety, 2004-2013, % Rice variety Volatility Emata 5.6 Manawthukha 8.8 Pawsan 10.5 Source: Authors’ estimates based on data from MIS/MOAI (2014), IMF (2014), and CSO (2014). 36. Paddy prices have been more volatile than rice prices. During 2009-2012, paddy price volatility was consistently higher than rice price volatility for all regions (Table 6). For MWK, for example, the difference ranged from 40 percent in Pathein to 110 percent in Yangon. Addressing price volatility in Myanmar would therefore require a high attention to the issues at the farm-gate level. Table 6: Myanmar: Average volatility of selected paddy and rice prices, 2009-2012, % Pathein, MWK Yangon, MWK Yangon, Emata Paddy prices 7.0 7.8 8.3 Rice prices 5.0 3.7 5.4 Source: Authors’ estimates based on data from MIS/MOAI (2014), MOC (2014), IMF (2014), and CSO (2014). 37. While volatility is a concern, it is also important to consider the increase in rice price levels in Myanmar. During 2011-2013, $-denominated rice prices from Thailand and Vietnam decreased while those from Myanmar increased (Figure 6), making it potentially difficult for Myanmar to compete if this trend continues. In the second part of 2013, Emata FOB prices even exceeded Vietnamese export prices.14 If Myanmar improved the quality and reliability of its exports compared to Vietnam, the reduced price differential would not necessarily be a problem and could be viewed as a desirable outcome. However, such improvements do not appear to have taken place in Myanmar yet (World Bank 2014b). 38. The recent increase in domestic prices in Myanmar was larger than that in other regional rice- exporting countries and close to that in rice importing ones, which are not concerned about the competitiveness of their domestic prices. Between January 2009 and December 2013, wholesale Emata prices in Yangon expressed in nominal $ terms rose by 41 percent, much more than the wholesale prices in the capitals of Cambodia, India, Vietnam, and Thailand (Table 7). The price increase in Myanmar even exceeded the increase in prices in some net importing countries such as Bangladesh and the Philippines. There is little space for further domestic price increases in Myanmar without erosion of its rice price competitiveness unless the quality of rice and reliability of exports are improved. In the summer of 2014, for example, rice exports from Yangon could not compete with cheaper-priced exports from Vietnam and India. The management of price volatility, therefore, requires a careful analysis. The next chapter presents a review of the driving forces of price volatility in Myanmar. 14 The FOB prices for Emata/Yangon are available only from May 2011 from MOC. 12 Figure 6: Selected export rice prices, $/ton 900 800 700 Nominal Prices, USS/ton 600 Thailand 500 Vietnam 400 Myanmar wholesale prices 300 Myanmar FOB prices 200 100 0 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Note: Thailand and Vietnam prices are export prices for 25% brokens. Myanmar prices are wholesale and FOB prices for Emata in Yangon, 25% brokens. Source: MIS/MOAI (2014), MOC (2014), and FAO-GIEWS (2014). Table 7: Rice prices in selected countries, $/ton 2009 2010 2011 2012 2013 Increase, 2009- 2013 Exporters Cambodia 373 439 460 445 437 17% India 348 420 418 386 396 14% Myanmar 259 301 345 344 364 41% Vietnam 384 387 467 397 363 -5% Thailand 460 444 511 560 504 10% Importers Bangladesh 293 403 409 302 372 27% China 410 437 561 609 601 46% Indonesia 476 617 699 749 753 58% Philippines 593 630 673 713 743 25% Note: Cambodia price is Phnom Penh, mix rice, wholesale. India price is New Delhi, coarse rice, wholesale. Myanmar price is Yangon, 25% Emata, wholesale. Vietnam price is 25% brokens FOB. Thailand price is 25% brokens FOB. Bangladesh price is Dhaka, coarse rice, wholesale. China price is Hunan, Indica, first quality. Indonesia price is national average, medium quality, retail. Philippine price is national average, regular milled, wholesale. Source: FAO-GIEWS (2014). 13 4. Drivers of Rice Price Volatility in Myanmar 39. There are a number of potential drivers of rice price volatility in Myanmar. Some are external, such as the volatility of world market prices. Yet given the low share of rice exports in total production (ranging from 7 to 9 percent depending on the year) and the large share of local production in domestic consumption, most drivers of price volatility are likely to be internal. The most significant candidates are: (i) the high concentration of paddy production within a few months; (ii) weak domestic market integration; (iii) poor market transparency; (iv) low private rice stocks; (v) limited export diversification with regard to quality and markets; and (vi) uncertain export policy. These drivers are discussed below in turn. 4.1 Volatility of world market prices 40. Due to the nature of the world market for rice, the price volatility of net exporters is higher than that of net importers. The world rice market is thin, segmented, and highly distorted. This is because most large importers such as China, Indonesia, the Philippines, and increasingly African countries insulate domestic markets from the world market through high import tariffs and import monopolies. They pursue rice self-sufficiency policies and import only when domestic prices reach very high levels within a short period of time (i.e., the price elasticity of import demand from the world market is very low). This leads to unpredictable volatility passed through to the world markets, to be absorbed by exporters (top exporters are India, Vietnam, Thailand, Pakistan, and increasingly Cambodia). As a result, the price volatility in net importing countries tends to be much lower than that in net exporting countries (Figure 7).15 41. Therefore, as a net exporter, Myanmar is bound to face higher price volatility than that observed in net importing countries. In the current situation, its domestic price volatility is actually expected to decrease through the stronger integration of Myanmar with the world trading system. But its domestic price volatility may also increase if world market prices are more volatile in the future. Domestic price volatility will eventually be determined by the interplay between many internal and external factors that could compensate for greater exposure to future world market volatility. Thus, this does not need to stop Myanmar from integrating with the world trading system, which would bring many other economic benefits such as incentives for investing in rice mills and export trading and smoother transmission of price and product quality information between Myanmar rice buyers and producers. 15 The example of India, a net exporter with low volatility, is less relevant to Myanmar because the former is a populous country with an inward-oriented food security and trade policy that generates exports mainly to dispose of excess government stocks, rather than using exports as a vehicle for agricultural growth. This is reflected in India’s very small share of exports in domestic production, relative to that in other exporting countries. The current Indian export drive beginning in the fourth quarter of 2011 is an exception, as overseas shipments are principally coming out of the marketplace. 14 Figure 7: Rice price volatility in selected Asian countries, % Philippines (regular milled) Importers China (Japonica) Indonesia (IR 64 III) Vietnam (25% broken) Cambodia (Rice Mix) Exporters Thailand (25% broken) Myanmar (Pawsan) Myanmar (Manawthukha) Myanmar (Emata) 0% 2% 4% 6% 8% 10% 12% Note: Volatility is measured in real LCU prices. Prices in Myanmar, Philippines, and Thailand are from 2004 to 2013; in Indonesia, from 2004 to 2012; in China, from 2005 to 2013; in Cambodia, from 2006 to 2013; in Vietnam, from 2008 to 2013. Source: Authors’ estimates based on data from FAO-GIEWS (2014) and MIS/MOAI (2014). 42. Future price volatility can particularly decline for Pawsan, high-quality rice. The volatility of Pawsan prices has been very high, almost twice as high as exported Emata (Table 5). One reason is that Pawsan is still hardly exported (see “Fragrant” column in Table 8) and is thus exposed mainly to local supply and demand shocks. During 2004-2013, the average volatility of fragrant rice on world markets (proxied by the Bangkok 100% fragrant FOB price) was only 5.1 percent or almost half of the price volatility of Pawsan. For the Pawsan variety, therefore, the strategy for reducing price volatility could be stronger integration with the world trading system. Table 8: Myanmar: Formal rice exports by grade, ‘000 tons Fragrant 100% S 5% 10% 15% Parboiled 25% Brokens Other Total 2010 * 2 12 9 5 0 456 0 * 485 2011 * 5 20 1 12 1 737 39 1 816 2012 * 1 53 0 6 8 504 30 2 605 Note: * Less than 500 tons. Non-fragrant rice is mainly Emata varieties. Rice 15% brokens means 85% head rice and 15% brokens rice, and so on. The category “Brokens” does not include any head rice. Source: Slayton & Associates. 43. To a lesser extent, volatility can also decline for MWK prices. MWK prices were more volatile than both the price of Emata (which is exported) and world market prices. With more integration with world markets, MWK price volatility could decline in the future. 15 44. The price transmission analysis carried out for this report shows that integration of the Yangon market of Emata and MWK with the world market, proxied by the Vietnam 25% brokens prices, was apparently non-existent during 2004-2013.16 Table 9 shows the parameters of price adjustments: α1 for an adjustment in Vietnamese prices to changes in Yangon prices and α2 for an adjustment of Yangon prices to changes of Vietnamese prices. There was no significant adjustment in Yangon prices of either Emata or MWK17 to the changes in Vietnamese prices during the observed period. On the other hand, Vietnamese prices reacted to changes in Yangon prices. On average, it took 6-7 months for Vietnam (with α1=0.14) to restore half of the price deviation from equilibrium of the price relation with Myanmar. Table 9: Estimates of the adjustment parameters between Yangon and Vietnam prices Market pairs 2004-2013 α1 α2 Manawthukha Vietnam-Yangon -0.16*** -0.03 Emata Vietnam-Yangon -0.14*** 0.00 Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates based on data from MIS/MOAI (2014) and FAO-GIEWS (2014). 45. The elasticity of price transmission (PT) between Yangon and Vietnam markets averaged 0.51. This indicates how much Vietnam prices changed between 2004 and 2013 when Yangon prices either increased or decreased. This elasticity is small compared to those of Myanmar’s exporting peers (Box 2). Rice-exporting countries usually have a higher elasticity than rice-importing countries. This low PT elasticity indicates that price volatility in Myanmar has so far been driven mainly by domestic factors. Box 2: Examples of Price Transmission Elasticities in Asia The PT elasticity for net rice exporters is often close to one. Gilbert and Moran (2010) found the elasticity of transmitting Thai to Vietnamese prices to be 0.821. This means that a 10 percent change in Thai prices leads to an 8.2 percent change in Vietnamese prices. For net rice importers, PT elasticity is usually lower as these countries protect domestic markets by tariff and non-tariff barriers. During 1990- 2010, the average PT elasticity in Philippines was 0.23, while Indonesian rice prices were not integrated with world market prices at all (Ghoshray 2011). A World Bank study on PT elasticity in global markets estimated that the long-term PT elasticity for rice in Asia ranged from 53 percent using FAO GIEWS prices to 60 percent using meta-analysis based on the literature (Greb et al. 2012). 46. Therefore, stronger integration of Myanmar’s markets with world markets could eventually reduce domestic price volatility. But domestic price volatility is unlikely to decline below the prevailing volatility in the world market. That volatility would need to be accepted and tolerated in Myanmar to maintain export competitiveness and sustain the country’s long-term food security. 16 Annex 3 presents the methodology and more detailed results of the market integration in Myanmar. It should be noted that the quality of data in Myanmar is very poor and these price data are not an exception. The results of the analysis, therefore, need to be interpreted with care. 17 The MWK prices are used in the analysis of international price integration due to their close link with Emata. 16 4.2 High seasonality of paddy production in Myanmar 46. The volatility of both rice prices in Myanmar has a strong seasonal nature. Between 2004 and 2008, the highest MWK prices in real terms were 40 percent larger on average than the lowest prices. Between 2009 and 2013, the ratio of maximum to minimum real prices declined to 1.29. The ratio was lowest in Yangon and highest in Myingyan (Table 10). Prices were usually lowest in December and January in all markets and highest between August and November, depending on the market. Table 10: Myanmar: Ratio of maximum and minimum real prices, Manawthukha rice Yangon Pathein Mandalay Myingyan Jan. 2004 – Dec. 2008 1.29 1.41 1.44 1.53 Jan. 2009 – Dec. 2013 1.20 1.31 1.30 1.34 Jan. 2004 – Dec. 2013 1.24 1.36 1.37 1.44 Source: Authors’ estimates based on data from MIS/MOAI (MIS/MOAI 2014), IMF (2014) and CSO (2014). 48. The average difference between trough and peak wholesale prices (as measured by seasonal factors) is higher in Yangon than in several other major Asian cities (Figure 8). However, domestically the seasonality of prices is lowest in Yangon. The seasonality of rice prices is much stronger in Pathein, Mandalay, and Myingyan, especially reflected in the depth of the trough, when farmers without storage capacity must sell their harvest (Figure 9). This implies very high seasonal price volatility in Myanmar. 49. Because many factors other than seasonality impact prices, there is also volatility in the magnitude of the price rise from year to year. For example, while prices typically rise substantially from January to March, that is not the case every year in every market. If farmers decide to hold their production for sale later in the year, the price rise may not compensate them for the storage costs incurred. Figure 8: Seasonality of real rice prices in Figure 9: Myanmar: Seasonality of real rice selected countries prices Source: Authors’ presentation based on data from FAO-GIEWS (2014) and MIS/MOAI (2014). 17 50. A main driver of seasonal price volatility appears to be the very high concentration of paddy harvest in the last quarter of the year. Most rice varieties in Myanmar have long durations. Due to this and the lack of efficient irrigation,18 more than 70 percent of paddy is produced in just two months – November and December (Figure 10). The dry season crop, which accounts for 8 percent of the total harvest, is so minor that its peak harvest (March-April in the Delta region, and the beginning of July in the Central Dry Zone) has a modest impact on prices. The lack of plantings with varieties of different durations prevents farmers from spreading the harvest more evenly over time, putting significant downward pressure on paddy prices during the main harvest. Figure 10: Paddy production by month in selected countries, % 90 Myanmar Philippines Thailand 80 70 Distribution of production, % 60 50 40 30 20 10 0 Q1 Q2 Q3 Q4 Note: Data for Myanmar are the average of 2003-2007. Data for Philippines and Thailand are the average of 2008-2012. Source: ASEAN Food Security Information System and FAO. 51. Harvests in other countries tend to be spread more evenly over the year as plantings of non- photoperiod-sensitive varieties predominate and more irrigation is available. In the Philippines, as with Thailand’s non-glutinous rice, paddy production is much more evenly distributed (Figure 10). In the Philippines, for example, 23 percent of paddy is produced in the first quarter of the year, 21 percent in the second quarter, 20 percent in the third quarter, and 37 percent in the fourth quarter. In Bangladesh, due to the availability of seeds with different durations and access to irrigation, the dominant boro crop accounts for only 60 percent of total rice production (the average for 2006/07-2008/09).19 Vietnam’s dominant crop, the winter-spring harvest, accounts for 50 percent of the total crop, spreading from March to May. The summer- winter harvest accounts for 28 percent and the 10th Month (mua) harvest accounts for 24 percent. 52. Thus, to reduce seasonal price movements in Myanmar, the best remedies are better availability of non-photoperiod-sensitive rice varieties with different growth durations to fit different cropping systems; improved irrigation systems; and enhanced farm advisory services assisting farmers to adopt new technologies and better market their produce. 18 About 23 percent of paddy land in Myanmar is classified as irrigated, according to MOAI. Yet in the irrigation sites visited by the World Bank in 2013 and 2014, only 30-40 percent of these areas had well-functioning irrigation coverage in the dry season. 19 Quarterly production data for Bangladesh and Vietnam were not available to the team, so these countries are not shown in Figure 10. 18 4.3 Weak domestic market integration 53. Strong market integration ensures that price signals quickly reach more producers to inform their production and marketing decisions, thereby increasing the responsiveness of the food system to price changes. By linking farmers more closely with consumers, marketing systems also better transmit signals to farmers on new market opportunities and guide their production choices to meet consumers’ preferences at competitive prices. Thus, stronger market integration helps reduce price volatility and increases the efficiency of markets. 54. As discussed below and detailed in Annex 2, domestic market integration in Myanmar appears to be weak. The transmission of price signals between surplus and deficit regions is slow due to many factors, including the existence of many local varieties that lead to rice market segmentation (Box 3), poor market transparency (see Chapter 4.4), poor quality of rural infrastructure (and thus high transport costs; see Figure 11), inadequate transportation services, and weak phone connectivity. Myanmar has the lowest road density in Southeast Asia, with 40 kilometers per 1,000 square kilometers: this is 50 percent lower than that of Lao PDR, which has the second lowest road density in the region. Furthermore, only 40 percent of main roads and 13 percent of rural roads are paved to all weather standards (ADB 2012). Its road transport costs have recently increased due to compliance with ASEAN transportation rules. Myanmar’s rural cell phone penetration is only 4 percent (see Table 90 in LIFT 2012). This stands in sharp contrast to the rest of Southeast Asia, where penetration is close to 100 percent. Figure 11: Transport costs in selected countries, $ per ton per 100 km 25 20 Costs in $ per ton per 100 km 15 10 5 0 Myanmar (2013) Cambodia (2009) Cambodia (2013) Vietnam (2013) Thailand (3013) Vietnam (2013), water transport) Source: World Bank (2014a) and World Bank (2014b). 55. The price integration of domestic market pairs is stronger than the integration with world markets, but is still relatively weak. For Pawsan, only Pathein and Myingyan markets seem to be strongly integrated, with price changes in one market reflected in other markets (Table 11). Partial price adjustments are also observed in the Mandalay market. 19 Table 11: Myanmar: Estimates of the adjustment parameters of Pawsan prices, 2004-2013 Market pairs α1 α2 Yangon-Pathein 0.00 0.09* Yangon-Mandalay -0.07 0.11 Yangon-Myingyan -0.09 0.12 Pathein-Mandalay -0.17 0.27** Pathein-Myingyan -0.23*** 0.24** Mandalay-Myingyan -0.28*** 0.13 Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates using the MIS/MOAI prices (2014). 56. For MWK, only Mandalay and Myingyan markets are clearly integrated (Table 12). Other markets displayed a variable degree of integration. Prices in Yangon react only to prices in Myingyan but not to prices in Pathein or Mandalay, a strange result given the closer distance between these markets and Yangon. Prices in Pathein adjust to changes in prices in Mandalay, and Mandalay prices react to changes in prices in Myingyan, but cross-market PT remains weak. Table 12: Myanmar: Estimates of the adjustment parameters of Manawthukha prices, 2004-2013 Market pairs α1 α2 Yangon-Pathein -0.09 0.11 Yangon-Mandalay -0.02 0.12 Yangon-Myingyan -0.15*** -0.07 Pathein-Mandalay -0.07 0.39*** Pathein-Myingyan -0.29*** -0.08 Mandalay-Myingyan -0.34*** -0.18* Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates using the MIS/MOAI prices (2014). 57. This weak market integration slows spatial arbitrage and may lead to higher price fluctuations. When prices rise in rice deficit markets such as Mandalay and Myingyan, farmers and traders in Yangon and Pathein are not immediately aware of this increase or are unable to ship rice to those markets quickly. The delays in the movement of rice from surplus to deficit areas leads to excessive price volatility even when there is sufficient rice availability in surplus areas. The weak domestic market integration also makes price stabilization difficult. If the government decides to stabilize prices in Mandalay, for example by releasing stocks, there is no guarantee that more stable prices in Mandalay will automatically stabilize prices in other markets. This needs to be borne in mind when addressing price volatility in Myanmar. 58. Going forward, stronger domestic market integration is vital for reducing domestic price volatility. The best ways to improve domestic market integration are by promoting more uniform seed varieties to reduce market segmentation, increasing the density of rural roads, especially all-weather roads, promoting the development of efficient transport services, and by implementing policy changes that facilitate the development of a competitive mobile phone network. 20 Box 3: Rice Market Segmentation in Myanmar In Myanmar, overabundance of varieties complicates aggregation along the value chain. It reduces the quality of consignments and ultimately increases market segmentation. The Beale classification system, established in 1927 and still in existence today, groups rice varieties based on paddy length and ratio of length and breadth. The categories are Emata, Ngasein, Letywezin, Meedon, and Byat. Most international traders, however, are only familiar with the first two groups – Emata and Ngasein. While there has been a narrowing of the number of varieties in the intervening period of the last three-quarters of the century, 250 varieties with plant characteristics are recorded by the seed division of MOAI, and 1,074 varieties listed just by name, grain type, and location are regarded as local varieties. This high rice market segmentation contributes to the weak market integration. Source: World Bank 2014a. 4.4 Poor market transparency 59. Timely provision of accurate market information is an important lubricant for the smooth functioning of agricultural markets. It is well recognized around the world that public access to timely and accurate agricultural outlooks and other market information such as prices reduces market uncertainty. On the other hand, poor market information increases uncertainty and ignites volatility. Poor market transparency was among the major factors of the global foodgrain price spikes in 2008 and 2010 (World Bank 2012a). Accurate information on production, consumption, exports, prices, and stocks is therefore very important to mitigate price volatility and inform policy decisions. 60. Yet the reliability and timeliness of such information is very poor in Myanmar compared to its peers in Southeast Asia. This makes its market integration weaker and price volatility excessive. Myanmar does not have a rice supply-demand balance trusted by market participants. The government neither produces nor publishes commodity forecasts, and the information on different parts of the rice supply- demand balance from various sources is conflicting (Table 13). The MOAI reports only paddy production, rice consumption, and export figures. The MOC reports official rice exports, which account for only half of total exports as described in Chapter 4.6 below. The commodity balance produced by FAS-USDA is publicly available and the most complete, but even that information requires further improvement. 61. The largest unknown is paddy production. According to MOAI, the country produced 31.5 million tons of paddy in 2011/12 (Table 14). The FAS-USDA’s alternative estimate is 17.9 million tons, or about half of the official estimate. USDA’s production figures are based on periodic travels to the key production areas where interviews are conducted with both farmers and millers. Price changes are also taken into account in arriving at USDA’s production estimates. Official wet season acreage figures are accepted as part of its calculations, but it comes up with its own estimates of plantings of the much smaller dry season crop. It is important to note that domestic and foreign analysts who have examined the two series find the USDA figures to be a closer fit with reality. 21 Table 13: Myanmar: Rice balance data from different sources, 2011/12, ‘000 tons MOAI MOC USDA Area harvested (‘000 ha) 8,000 7,030 Beginning stocks n/a 485 Paddy production 32,064 17,927 Milled production 19,328 11,473 Exports 707 821 1,357 Domestic consumption and residual 11,675 10,200 Ending stocks ~7,000 401 Yield (tons of paddy per ha) 3.83 2.55 Source: MOAI, MOC, and FAS-USDA (2014). 62. Estimates of rice consumption also vary widely. In 2011/12, the total population of Myanmar was estimated at 64 million. Using the data from MOAI in Table 13, per capita rice consumption was 182 kg. Using the FAS-USDA data, per capita consumption was lower, at 159 kg. This difference is equivalent to 1.5 million tons, more than the total exports in 2011/12. According to the World Bank, per capita rice consumption is even lower: average rice consumption in 2010 was found to be 145 kg per capita, with 117 kg in urban areas and 155 kg in rural areas (World Bank 2014c). Compared to the MOAI estimate, actual total consumption could have been 2.4 million tons smaller. Using the provisional results of the 2014 population census, which estimate the total population of Myanmar at 51.4 million, makes the total rice consumption even smaller: 9.4 million tons using MOAI’s per capita rice consumption data and 7.5 million tons using World Bank data. This very large margin of error adds uncertainty to the market. The lack of reliable official production and stock data handicaps governmental efforts to implement any rice export policy that seeks to maximize domestic price stability while increasing farm incomes and improving food security. Table 14: Myanmar: Paddy production, million tons Period Official Data USDA Data Difference % Difference Avg ‘75-76/’79-80 9.79 9.79 0.00 0 Avg ‘80-81/’84-85 14.08 11.05 3.03 27 Avg ‘85-86/’89-90 13.81 12.14 1.67 14 Avg ‘90-91/’94-95 15.40 14.20 1.20 8 Avg ‘95-96/’99-00 17.90 16.17 1.73 11 Avg ‘00-01-’04-05 22.59 18.03 4.55 25 Avg ‘05-06/’09-10 31.06 18.09 12.97 72 2008-09 32.57 17.50 15.07 86 2009-10 32.68 18.19 14.49 80 2010-11 32.58 17.28 15.30 88 2011-12 31.50 17.93 13.57 76 Source: FAOSTAT (2014) and FAS-USDA (2014). 22 4.5 Uncertainty over private rice stocks 63. The large uncertainty over production and consumption discussed above makes it difficult to estimate ending stocks, yet rice stocks are crucial for understanding and acting on price volatility. When stocks are large relative to consumption (use), prices are usually less volatile in anticipation of possible stock release. But when stocks are low, particularly private stocks, food prices are often more volatile (World Bank 2012a). Gouel and Jean (2012) provide a theoretical framework for the large role played by stocks and trade in determining food price volatility. Figure 12 illustrates this relationship between stocks and price volatility using the US wheat market as an example. When the stock-to-use ratio is large, a 5 percent reduction in stocks leads to a small increase in prices compared to the price increase due to stock’s reduction when the stock-to-use ratio is low. Figure 12: US: Wheat stocks-to-use ratio and real prices, Jan. 1990-Aug. 2009 2.0 1.8 1.6 US$/bu, CPI deflated 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Stock-to-use ratio (US) Source: Cafiero and Schmidhuber (2011). 64. Based on the MOAI data, derived from Table 13, ending stocks in Myanmar are estimated at 7 million tons or 60 percent of the stock-to-use ratio, a very large amount that is not consistent with high domestic price volatility. The USDA’s more modest estimate of ending stocks (0.4 million tons or 4 percent of the stock-to-use ratio) is a better reflection of the actual situation (low stocks and high price volatility). It is observed in many countries and globally that when the stock-to-use ratio goes below 15 percent, prices become highly volatile, which is more consistent with the current situation in Myanmar. 65. Higher private stocks would help reduce rice price volatility. More profitability and predictability for rice mills and traders would be instrumental in increasing private stocks. The best ways to increase rice stocks in Myanmar would be through: increasing market transparency; committing to open trade policy; improving access to working and investment capital to rice mills and traders, including through the promotion of foreign direct investment; allowing foreign direct investments in the existing rice mills (“brown” projects), not just in new (“green”) projects; and undertaking reforms that could reduce capital interest rates (to lower storage costs). These measures would stimulate more storage and reduce price volatility. 23 4.6 Low diversification of exports 66. In addition to exporting mainly low-quality non-fragrant rice (recall Table 8), the geographic diversification of Myanmar’s exports remains limited. There are essentially two big buyers – African countries for official exports (Table 15) and China for unofficial cross-border exports (Table 16). The price volatility of formal exports appears to be low as discussed in Chapter 3. The largest source of unpredictability and the consequently harmful volatility seems to be emanating from the cross-border trade with China. Table 15: Myanmar: Formal rice exports by destination, 2005-2012, ‘000 tons 2005-09 2008 2009 2010 2011 2012 Destination average EU * * * 0 12 28 Former Soviet Union * 0 2 11 19 44 Africa 261 196 899 318 506 460 Asia 129 403 150 156 276 72 Total 391 598 1,052 485 816 605 Note: * Less than 500 tons. Source: Slayton & Associates. 67. The border trade is not officially recognized by China. Thus, Myanmar’s exports to China are unofficial. This trade offers many opportunities. Chinese buyers are described as not choosy, with an estimated 70 percent of the rice bought in Myanmar being 25% brokens. Only 10 percent of the rice is 5% brokens and the remaining balance comprises brokens and glutinous rice. By trading informally, Myanmar can also bypass the import tariff rate quota and compliance with sanitary and phytosanitary rules (World Bank 2014b). Table 16: Myanmar: Exports via border trade, 2009-2013, ‘000 tons China Thailand Total 2009/10 68 11 79 2010/11 0 0 0 2011/12 136 0 136 2012/13 752 8 760 Source: MOC (various documents). 68. But informal trade brings uncertainty and volatility. Reportedly, there are some 200 traders involved in Myanmar, but only about 10 Chinese buyers. Trucks with rice are reported to have been periodically seized by Chinese customs officials before they reach Ruili (4 km from Muse across the border), which could be part of the negotiating tactics of the Chinese buyers. More and more Myanmar traders complain about this uncertainty and the risks of trading over the border. 69. Neighboring Cambodia has managed to improve quality and diversify its rice exports in a very short time, and Myanmar can learn from Cambodia’s experience. Cambodia’s rice exports increased from 12,600 tons in 2009 to 378,850 tons in 2013, with more than half of that going to the European Union (EU) under the Everything but Arms Agreement (World Bank 2014a). From being a small exporter of low- quality rice, Cambodia has emerged as a reliable exporter of high-quality aromatic rice, able to meet the stringent food safety requirements of developed countries. Cambodia has also increased its export of paddy to Vietnam and Thailand, from 1.4 million tons in milled equivalent in 2011 to 1.6 million in 2013, pragmatically using the lower milling and logistical costs across the border to the benefit of its farmers. High attention to quality and diversification has paid off, and the next generation of reforms is now focusing on improving rice 24 mills’ access to finance (particularly via a system of warehouse receipts), further reducing export processing costs, and investing in energy and transport infrastructure, all of which will help sustain this success, and all of which are also necessary in Myanmar. 4.7 Uncertainty over export policy 70. Although many barriers to exports were recently abolished in Myanmar, the private sector remains cautious given the potentially conflicting objectives of rice export promotion and price stabilization. This significantly increases uncertainty for millers and traders. Having been previously coerced to sell stocks at a loss, millers and traders remain cautious when domestic prices show a sustained increase. At the same time, international buyers are wary about whether the government will allow its purchases to be executed. 71. Currently, there are no formal export restrictions. But when domestic prices rise to high levels, there is always a risk of exports being temporarily restricted. The most recent export bans were in (i) 2004, after only a little more than half of the licensed exports of 0.5 million tons had been executed; (ii) 2008, when the government signaled that it would allow about 0.4 million tons of rice to be exported by 22 private companies but suspended their licenses after cyclone Nagris hit; (iii) 2011, when exports were limited in the lead-up to the 2011 election; and (iv) 2013, when an informal agreement with millers and exporters was reached to release stocks and restrict exports. 72. On one hand, export bans or even the risk of imposing them may have helped reduce the price volatility of exported rice. As shown in Table 5, the price volatility of Emata (which is exported) was lower than the volatility of Pawsan and MWK, domestically consumed rice varieties. On the other hand, lower volatility of export prices does not seem to be passed through to the prices of locally-consumed rice. High domestic price volatility has prevailed in spite of the reduction in the volatility of export prices. 73. While helping to contain some short-term price volatility of export prices, export bans harm farmers, millers, and traders, especially when they are applied in an ad hoc manner. They increase uncertainty and tend to push down average farm prices, thus reducing long-term food availability, an important precondition for food security. Export bans also slow down modernization of the milling and trade sectors, resulting in inefficient logistics and thus a sector that is unable to mitigate large drops in paddy prices after the harvest. More rapid paddy milling and rice exporting could have partially mitigated these price drops, but the rice industry in Myanmar is not currently positioned to accelerate the movement of rice from fields to local and world markets (World Bank 2014b). Export bans would further hamper structural transformation. 74. International experience shows that targeted protection of the poor through safety nets can be a better alternative to mitigate the impacts of price spikes. This is because such a policy intervention does not create uncertainty for rice exporters and traders. If Myanmar is perceived as an unreliable exporter because of frequent bans, this will lower export prices and push down the prices received by farmers. In the long run, therefore, export bans and other restrictions20 yield more costs than benefits and thus need to be avoided. 20 A variable export tax is often considered a market-friendly alternative to export bans. In theory, a variable export tax follows world market prices up and down, helping to stabilize prices without completely stopping trade. In practice, effective operation of such a variable tax for rice requires strong market monitoring capacity and protection of the tax rate-setting mechanism from political influence to avoid prohibitively high rates that would effectively turn the tax into a ban. Due to these reasons, variable export taxes for rice are rarely used in Asia. 25 5. International Experience from Management of Rice Price Volatility 75. Most Asian countries highly value rice price stability. Price stability is considered a public “good”. But the international experience suggests that not all instability or volatility is “bad”. Stable prices do not mean fixed prices. Full price stabilization for agricultural products is neither achievable nor desirable. Seasonal and spatial price movements are common for agriculture, as they are the key to fostering arbitrage between surplus and deficit regions and they underline seasonal storage and trade decisions. But countries often put more emphasis on price stability than on price levels and food affordability (i.e., other elements of food security), which results in the loss of competitiveness, food insecurity, and other costs. 76. Should volatility be a concern, attention needs to be paid to large price movements, which are both difficult to predict and disruptive for economic decisions. Such volatility is bad for both producers and consumers. An example of such disruptive volatility is the global foodgrain price spikes at the end of the 2000s. During 2007-2013, world rice prices were much higher and more volatile21 than during 2000- 2006, reaching high levels similar to those experienced in the 1970s, which induced the imposition of costly policies that were subsequently difficult to remove (World Bank 2012a). Concerns over similar distortive policy responses to price volatility in the recent period remain high. 77. The strategic policy question is how to reduce price volatility. This is not a trivial question because many past efforts to stabilize rice prices in Asia have failed to help generate growth and reduce poverty. More stable prices per se do not guarantee the realization of economic gains. Success depends on how stability is achieved and the complementary measures taken. If domestic prices are stabilized for long periods of time without taking into account changes in world market prices, domestic prices often end up either at high levels in net importing countries (above import parity) or at very low levels in net exporting countries (below export parity). When this happens, stable prices may actually increase poverty, distort economic decisions, and eventually lower the quality of growth (World Bank 2012b). This interlinkage between price level and price volatility explains why more stable prices (food stability) at the expense of higher prices (food affordability) or lower prices (food availability) may actually reduce food security. 5.1 Government owned stocks 78. The Philippines and Indonesia, countries with less volatile rice prices than Myanmar, provide relevant examples. As net importers, the Philippines and Indonesia have used a mix of government-owned stocks and monopoly over import of rice for many decades to stabilize prices. They have succeeded in bringing volatility below the level observed on the world market (recall Figure 6). Yet these more stable prices are kept at a high level, well above the market prices in exporting countries (Figure 13). In most other countries where domestic food prices are stabilized through the use of large government-owned stocks and monopoly over trade, this is the most common outcome, creating many problems (World Bank 2012b). 21 During 2007-2013, the average volatility of the global grain price index was twice as high as during 2000-2006, see World Bank (2012a). 26 Figure 13: Nominal rice prices in selected countries, 2004-2013, $/ton 950 850 Philippines Norminal Prices, US$/ton 750 650 Indonesia 550 Thailand 450 350 Vietnam 250 150 Ju 04 N 04 Ap 04 Fe -05 Ju 6 D 6 O 07 M -07 Au -08 N -09 Fe -10 Se 5 Ap 09 Ju 1 D 1 O 12 M 12 Au -13 M -06 Ja 08 Ju 09 Se 0 M -11 13 1 l-1 0 l-0 r-0 r-1 b- n- n- - b- - - - - g- n- g- ec ov p ay ct ar n ov p ay ct ar ec Ja Source: FAO-GIEWS (2014). 79. The first problem is that high prices hurt consumers, especially the poor. Unlike in developed countries, most consumers in developing countries spend a large share of their income on food (Table 17).22 Higher rice prices force them to spend even more. In addition, high rice prices tend to lower wage competitiveness. Eventually, the pace at which Indonesia and the Philippines are able to bring down domestic prices sustainably will have a bearing on their capacity to expand into any internationally competitive, labor- intensive activity. Table 17: Food as a share of CPI in selected countries, % Country Share (%) Myanmar 61.0 Bangladesh 58.8 Philippines 46.6 India 46.2 Vietnam 39.9 Indonesia 36.2 Thailand 33.0 United States 14.8 Euro zone 14.0 Japan 25.9 Source: ADB (2011), World Development Indicators, and World Bank (for Myanmar). 22 Note that the share of food in the CPI understates the importance of food to the poor, because the poor spend a much larger share of their income on food than do the middle class and the rich. 27 80. The second problem is that high rice prices do little to accelerate agricultural growth in these countries. High consumer rice prices trickle down to farmers in both Indonesia and the Philippines and higher paddy prices help induce some supply response (although some of the supply response comes at the expense of other crops). But this happens only in the short-term, because higher paddy prices lead to higher production costs as time goes on. In the Philippines, for example, labor costs account for 60 percent of paddy production costs and when rice prices go up, labor costs follow. Input suppliers adjust prices for seeds and fertilizers and land owners raise rental fees, aware of the short-term increases in rice farmers’ purchasing power. Therefore, in spite of the 40 percent increase of farm-gate paddy prices in the Philippines from 2005 to 2010, the profit/cost ratio of paddy producers remained almost unchanged, at 35 percent (BAS 2013). During 2004-2012, paddy production grew by only 2.5 percent annually, not enough to meet the country’s rising consumption requirements (Table 18). In Indonesia, paddy production grew by only 0.8 percent annually during the same time period. To induce a continuous supply response in such a situation, rice prices need to be adjusted upwards continuously, which is not possible from consumers’ point of view. 81. The relationship between price volatility and agricultural growth is clearly complex. In Cambodia and Vietnam, for example, in spite of more volatile prices, paddy production grew much faster (Table 18). This does not imply that there is a positive relationship between volatility and growth. To the contrary, rice production could have been higher in Cambodia and Vietnam had their prices been more stable. What it does imply is that how stability is achieved really matters. Measures that contribute to more stable prices but undermine private sector development or crowd out public spending away from delivery of public goods, as is the case in both Indonesia and the Philippines (World Bank 2007; 2009) are not advisable. On the other hand, predictable trade policy as in Cambodia (Slayton and Muniroth 2012) and investments in rural development as in Vietnam can counteract some of the negative effects of price volatility and help encourage agricultural growth. Table 18: Rice price volatility and paddy production in selected countries, 2004-2012 Country Price volatility (%) Annual growth of production (%) China 3.5 2.8 Philippines 3.4 2.5 Indonesia 5.2 0.8 Thailand 6.4 1.5 Vietnam 8.3 3.2 Cambodia 8.0 5.2 Source: Price volatility data are from Figure 7. Paddy production data are from FAS-USDA (2014). 82. The third problem is high budget costs. Government-owned stocks are expensive. They are a recurrent expenditure for national budgets. In many instances public agencies tend to buy high, sell low, and store long, incurring large costs.23 In 2008/09, India spent 1.5 percent of GDP on government-owned stocks of rice and wheat (Table 19), though it is likely to have recovered some of the costs by selling stocks. During 2008-2010, Indonesia spent 0.5 percent of GDP on its government-owned rice stocks. In 2004/05, the Philippines spent 0.4 percent of GDP, and scaled it up to 1 percent in 2009 (ADB 2011; World Bank 2012b). In comparison, in 2003 these countries spent only 0.06 percent of GDP on agricultural research 23 The international experience shows that in contrast to private firms who seek to maximize their profit by buying low, selling high and storing short, and by this also helping to stabilize prices, public sector programs often have multiple objectives, including farm support and protection of consumers. As a result, public programs often pay above-market prices to farmers and sell stocks at below-market prices to consumers. In addition, it takes longer than for private firms to authorize purchases and releases of public stocks, increasing storage costs for the program. 28 and development (Pardey et al. 2006). In the Philippines, during 2009-2012, expenditures on irrigation and rural roads (key to reducing production and marketing costs) were only 0.17 and 0.05 percent of GDP, respectively. Due to the large spending on government-owned stocks, the Philippine National Food Authority is the most indebted state-owned corporation in the country. In 2012, the gap between its liability and assets was $3.6 billion, or 1.4 percent of GDP (DBM 2013). The high fiscal costs of buffer stocks, therefore, have opportunity costs in terms of growth that offset the benefits of stable prices. Table 19: Government spending on grain stocks in selected countries Country Spending on public stocks Spending on agriculture Spending on (% of GDP) (% of GDP agricultural R&D excluding [A]) (% of GDP) [A] [B] [C] India 1.0% (2004/05) - 1.5% (2008/09) 1.2% (2008/09) 0.06% (2008/09) Indonesia 0.5% (2008-10) 0.8% (2008) 0.05% (2003) Philippines 0.4% (2005-06) – 1.0% (2009) 0.8% (2005) 0.06% (2002) Zambia 0.3% (2009) – 1.9% (2011) 0.6% (2010) 0.15% (2010) Source: World Bank (2012b). 83. The fourth problem is the crowding out of private investments in trade and storage. In addition to a state monopoly over rice imports in both Indonesia and the Philippines, the low transparency and high uncertainty about imports, procurement, release, and rotation of stocks reduce private sector investment in many segments of the rice value chain. As a result, the private sector underperforms in terms of creating jobs and driving economic growth, which is needed for sustained poverty alleviation and shared prosperity. The extent of private sector underinvestment became obvious when the Philippines opened imports to the private sector in 2011. Despite concerns about the inability of the private sector to carry out import operations, private firms imported 660,000 of the 770,000 tons of total rice import; the Philippine National Food Authority imported only 110,000 tons compared to 2, 240,000 tons in 2010. Private import has not only brought budget savings but also reduced price volatility and stimulated creation of new jobs. 84. The fifth problem is the difficulty in choosing the right type of rice variety for stocking and price stabilization. Rice is not a uniform commodity. In Myanmar, 250 rice varieties with plant characteristics are recorded by the seed division of MOAI, and 1,074 varieties listed just by name, grain type, and location are regarded as local varieties (see Box 3). Substitution in consumption among different varieties is not perfect. Procuring one variety for the government stocks may not allow its effective consumption in other areas where different varieties are usually consumed. 85. In net exporting countries such as Myanmar, an additional problem originating from high prices is the loss of international competitiveness. In contrast to net importers, net exporters cannot elevate domestic prices above the world market level because they will not be able to export without extremely large subsidies. Large public stock programs are therefore not suitable for net exporters such as Myanmar (see also Box 4). The good news is that net exporters do not need large public stock programs as they are not vulnerable to the export bans of other countries. The two major Southeast Asian rice-exporting countries, Thailand and Vietnam, have both historically held low stocks. The current situation in Thailand is an exception as the government tried (and failed) to raise world market prices, as opposed to trying to stabilize domestic prices.24 Thus, historical Thai experience from the 1950s to 1970s is more relevant to Myanmar today than is the current Thai rice policy. 24 See the World Bank (2014b) for the description and analysis of the rice pledging scheme in Thailand. 29 Box 4: Myanmar: Private Public Partnership in Rice Stock Management In Myanmar, the government does not hold public stocks but works with the Myanmar Rice Federation to coordinate private stock releases by the private sector during rice price spikes. In summer 2013, for example, rice mills were reported to have collectively released stocks in the Mandalay region helping to put downward pressure on wholesale prices. Although this program does not have a large distortionary impact on the market, due to its low size, and rice mills sell at market prices, it may have a negative effect on rice mills’ profits, who could have benefitted from higher prices. This partnership would need to be carefully managed to avoid pressures on rice mills to sell rice at the below-market prices and focus on information sharing and coordination. It is also important to realize that such a program will continue having a limited (e.g., short-lived and localized) impact on price volatility for most markets in the country until (i) storage costs are significantly reduced so that rice mills and traders can store more paddy and rice and release stocks when prices rise, and (ii) market transparency is improved so that the private sector rationally releases or buys stocks depending on the price outlook, thus bringing stability to the market through normal business transactions. 86. The international experience shows that in the short run, net exporting countries are left with very few options. They include: (i) trade restrictions and (ii) minimum farm prices, both to directly reduce price volatility, and (iii) social safety nets, to mitigate the impacts of volatility. Only the last one, safety nets, offers a long-term sustainable solution. The following analysis explains why. 5.2 Export trade restrictions 87. The benefits of export bans, taxes, or quotas rarely exceed their costs. Often they are not even effective at containing domestic price volatility. The export bans in Myanmar used frequently in the past did not help reduce domestic price volatility or reduce volatility of export prices much below the level prevailing in the world market (as discussed in Chapter 3). Myanmar aspires to increase rice exports to 2 million tons by 2014/15 and 4 million tons by 2019/20. The current export volume is around 1.3 million tons, and any use of the above trade policy instruments would make it impossible to achieve these export targets. Having been previously coerced to sell their stocks at a loss and facing frequent export bans, Myanmar millers and traders are suspicious of any restriction on trade. Even if trade measures are designed and communicated in a transparent manner (for example, a variable export tax), the private sector is likely to remain risk averse for many years to come, offering lower paddy prices to farmers in case of the use of export restrictions. 88. Another problem with the use of trade policy to stabilize prices is that it diverts attention from the key causes of volatility. As discussed in Chapter 4, rice price volatility is a result of many structural factors. Trade policy does not address any of them. Most difficult but important is tackling volatility at the farm level. In the Philippines, for example, in spite of the successful stabilization of consumer rice prices, paddy prices remain volatile (Figure 14) due to ineffective management of irrigation systems, and underinvestment in rural infrastructure, research and extension services, and other productivity-inducing activities (World Bank 2007). Paddy prices in Myanmar are volatile for similar reasons, compounded by the heavy concentration of production in just two months of the year. As a result, the growth of agricultural production remains low, despite the large amount of funds spent on stabilization of consumer prices. Unless public investments address the key causes of volatility at the farm-gate, stabilizing consumer prices alone will continue to undermine growth and poverty reduction as discussed above. 30 Figure 14: Philippines: Volatility of farm-gate, wholesale and retail prices, 1990-2011 40 35 Volatility 30 3.0% Prices, pesos/kg 25 Retail 20 Wholesale 15 10 Paddy Volatility: 11.6% 5 0 M 91 Ap 2 M 93 D 00 Ap 5 Au 09 Fe 90 Ju 4 Ju 5 Au 6 Se 97 N 9 Ja 1 M 04 Ju 7 O 8 Fe 03 M 06 Ju 8 10 -9 9 l-9 -9 -0 -0 -0 -9 9 0 - l- n- b- r- n- g- b- p- n- r- n- g- ov ar ct ec ar ay ay Ja Source: Authors’ estimates based on data from the Philippine Bureau of Agricultural Statistics. 5.3 Minimum farm prices 89. Some countries intervene at the farm-gate level through minimum prices for stabilization. While price support is popular with farmers, it creates many problems. This is especially true for exporting countries if the support and release prices are set too high, which can easily occur when politicians get in a bidding war for farmers’ votes in contests where farmers represent a large share of the electorate. If set too high, they cannot be easily reduced; then exports are reduced and the government needs to subsidize exports, which is costly and thus impossible for most developing countries. To have the intended price impact, the quantities purchased must be stored a considerable time, incurring storage and interest costs, and even then sale prices may still decline, as has happened in Thailand. If held too long, quality problems arise. Finally, intervention efforts can often breed corruption. The recent Thai experience with the rice pledging scheme is particularly instructive (Nipon 2014). 90. The Thai government spent large quantities of money in attempts to raise prices for farmers, but the end result was that many farmers were not paid at all for their paddy. The program failed to raise world market prices as other countries were able to increase their exports at the expense of Thailand, pushing world market prices down. As a result, Thailand’s rice ending stocks increased from 6.1 million tons (or 60 percent of domestic use) in 2009/10 to 14.4 million tons (or 134 percent of domestic use) in 2013/14 (FAS-USDA 2014). The subsidy cost the government $12.7 billion or 3.5 percent of GDP in its first year of operation, raising the country’s fiscal deficit to 4.4 percent of GDP in 2012 from 1.7 percent in 2011. The scheme cost the government another $13.9 billion or 3.6 percent of GDP in 2013 (World Bank 2014d). The 2013/14 export is projected to increase to 8.5 million tons from 7.0 million tons in 2012/13, helped by the significant decline in Thai prices, but it falls short of the 11.0 million tons exported in 2010/11 (FAS-USDA 2014). 31 91. There are many other indirect but similarly significant costs: • Thailand’s reputation of exporting high-quality rice for which a premium is paid is being undermined by farmers switching to varieties with the shortest growing seasons and highest yields. These are varieties with inferior palatability. • The long-term storage of milled rice is resulting in a deterioration of its quality and leading to reports of the presence of excess chemicals. • The already overbuilt rice milling industry is expanding, which will lead to the closure of mills not enrolled in the program. • Farmers’ costs of production are rising, reflecting a doubling of land rents and increased use of and a 20 percent increase in the prices of fertilizer and pesticides.25 The higher land rents are a direct consequence of the pledging policy. • Unable to supply their customer base of over 10 million tons with Thai rice, Thai exporters have started to trade third country rice on a large scale and invest in rice mills in competitor countries, in particular Cambodia. Both actions directly enhance the long-term competitiveness of other countries. 92. During 2004-2008, the volatility of farm-gate paddy prices was 31 percent lower than the volatility of wholesale rice prices (Table 20). Despite the above costs, the volatility of wholesale rice prices in Thailand remained high. During 2009-2013, the difference was 13 percent. It is important to understand the reasons for more stable paddy prices in Thailand and the failure of other countries such as the Philippines to achieve the same result (recall Figure 14 and see Table 20). Table 20: Volatility of paddy and rice prices in the Philippines and Thailand, % Thailand Philippines Farm-gate paddy Wholesale rice Farm-gate paddy Wholesale rice 2004-2008 5.9 7.7 6.4 4.3 2009-2013 4.4 5.0 4.1 1.7 Note: For Thailand, paddy prices are non-glutinous average national farm-gate prices and wholesale prices are 25% brokens in Bangkok. For the Philippines, paddy prices are average national farm-gate prices and wholesale prices are national average prices for regularly milled rice. Source: FAO-GIEWS (2014). 93. When a government commits to buy a large portion of production, farm-gate price stability is an achievable outcome. This has happened in Thailand. But in the Philippines the government’s paddy procurement was only 2 percent of total production during 2006-2012. Due to the lack of budget resources, an emphasis has been made on announcing a minimum farm price and requesting traders and millers to buy at the fixed price. Enforcing minimum prices in this situation is difficult, however: the actual paddy prices in the Philippines have usually been below the minimum support prices, and the volatility of paddy prices has remained high. Traders and millers pay prices to farmers based on the cost of doing business, competition, and marketing expenses (Box 5 describes a similar situation for livestock prices in Lao PDR). 94. How would Myanmar enforce minimum farm prices if they were introduced? The Farmers’ Rights Protection Act adopted in 2013 foresaw the introduction of minimum farm prices. While the design and implementation details are not yet clear, it seems that given the lack of sufficient budget resources to buy and store a large share of paddy, the Thai approach is unlikely to be used in Myanmar. If minimum farm prices are introduced, Myanmar would follow the Philippine’s approach, requesting paddy buyers to honor officially announced farm prices. If the lessons from the Philippines are instructive, the volatility of paddy prices in Myanmar will not be reduced unless the core factors of volatility (discussed in Chapter 4) are addressed. 25 “Broken Rice,” Wall Street Journal, July 26, 2013. 32 Box 5: Lao PDR: Minimum Prices for Swine Products Rice is not the only commodity in Southeast Asia for which governments have tried to set minimum farm prices. In Lao PDR, for example, the government set the prices for pigs at 18,000-20,500 kip per kg at the farm level; at 22,500-25,000 kip per kg at the slaughterhouse; and at 37,000 kip per kg at the markets. The government does not actually buy these products but sets a reference minimum price for traders. But the price compliance has proved difficult to control.26 1 Actual farm pig prices are often below the announced minimum prices for the same reason as seen with paddy prices. The private sector faces high transaction costs working in rural areas and cannot afford to pay the announced prices. Farmers prefer to obtain cash for their products rather than waiting for elusive traders to pay a high minimum price. 26 95. It is important that Myanmar does not introduce minimum farm prices but instead focuses on export promotion and public investments to address the underlying causes of price volatility (e.g., poor roads, production concentrated in time, old irrigation systems). Minimum farm prices are not suitable to stabilize rice prices in a low-income developing country aspiring to become a net rice exporter and to translate rice sector growth into poverty reduction. This policy instrument would need to be abandoned. 5.4 Safety nets 96. Although the price volatility can be reduced to some extent, rice prices in Myanmar will remain volatile. Even this reduced volatility will have a negative impact on the poor. International experience suggests that this negative impact can be mitigated by safety net programs (Grosh et al. 2008). In turn, when effective safety nets are in place, there is more popular support for open trade policy, which is vital to the successful development of the rice sector. 97. Formal safety nets in Myanmar are still absent, thus work needs to start now to establish them. There are emerging lessons from Myanmar, as several development partners have introduced safety nets (including cash and food for work, and cash transfers), predominantly to provide emergency relief to food-insecure households and other vulnerable groups in the aftermath of Cyclone Nagris and other natural disasters (see Infante-Villaroel, forthcoming). Recently, the Ministry of Education and Training introduced Myanmar’s first conditional cash transfer program – a stipend program targeted to poor students, mainly orphans to prevent early dropouts (often caused by income insecurity). There are many other lessons from around the world, as summarized below. 98. In many Asian countries, rice remains closely tied to food security objectives, particularly as an important element of a food-based social safety net program. Food-based safety nets are designed to ensure livelihoods (for example, through provision of employment in a public works program with wages paid in food), increase purchasing power (through the provision of food stamps, coupons, or vouchers), and relieve deprivation (through the direct provision of food to households or individuals). They differ from cash- based programs – public employment (with wages paid in cash) or cash transfers – in that they are tied to food as a resource. 99. Both food and cash-based safety nets range from universal subsidies to more targeted programs. The choice of instrument depends on country circumstances, such as administrative capacity to identify and deliver benefits to the targeted population. Cash-based assistance may be conditional or unconditional, depending on whether beneficiaries are required to comply with co-responsibilities, typically related to accessing health, education, and nutrition services or to being part of public work programs. Often governments choose to use food-based (in-kind) transfers because they are concerned about high food prices or because commodities markets are inadequate and therefore they have to guarantee access to food by the poorest people. But often the selection of food-based transfers is also a result of the existence of 26 “Pig Farmers Still Struggling Despite Price Rising”, Vientiane Times, May 5, 2014. 33 large government-owned grain stocks established to support farmers that need to be released and rotated. For example, India, Indonesia, and the Philippines, with their government-owned stocks to help maintain high producer prices, often release stocks at subsidized prices to consumers. Myanmar has also been considering the establishment of government-owned stocks, but so far its stock program has been limited to support for private stockholding by rice millers under the coordination of the Myanmar Rice Federation (Box 4). Implementation of such programs poses several logistical and implementation challenges related to procurement and storage of food, transportation and distribution, and waste, spoilage, and pilferage (as described in Chapter 5). International experience with food-based safety nets has very often been discouraging (Grosh et al. 2008). 100. More and more Asian countries are now moving from poorly targeted and largely ineffective in-kind food transfers to more targeted cash transfer programs, inspired by the success of other countries. This is because in most cases targeted cash-based assistance does not distort production and consumption choices and it has relatively low administrative costs, leakage, and pilferage compared to in- kind transfers. The Philippines, for example, has started to move towards cash transfers gradually replacing its subsidized rice distribution program. The global food price crisis of 2008 increased the Philippine government’s attention on the need to develop improved and more cost-effective social protection measures as an alternative to the badly targeted rice distribution program. In particular, the crisis reinforced the need to improve the targeting of social safety net spending and accelerate the roll-out of a cash transfer system. In 2008, the government launched a conditional cash transfer program that focuses on supplementing the income of the poorest households while also supporting their human capital development (Fernandes and Olfindo 2011). At the same time, the government improved the targeting of poor households by developing a Proxy Means Test methodology for identifying the poor. The early results of this conditional cash transfer program’s success in reaching the poor in a cost-efficient manner and in changing recipients’ behavior are very encouraging. The program is also credited with reducing poverty incidence among beneficiaries by about six percentage points (Velande and Fernandes 2011). 101. Cash transfers are considered to be the fastest and least costly method of reaching poor and vulnerable households if the delivery infrastructure exists and markets are functioning. Using cross-sectional household surveys from Indonesia carried out before and after the 1998/99 economic crisis, Skoufias et al. (2011) found that cash transfer programs can play an important role in helping households protect their consumption of essential nutrients during a crisis. However, to ensure that all micro-nutrients are consumed, relying entirely on cash transfers may not be sufficient. Targeted micro-nutrient supplementation programs may have to accompany cash transfer programs to ensure that key micro-nutrients are not sacrificed during crises. 102. In low-income countries, public works are often more common for social protection than direct cash transfers. Public work programs are more politically acceptable as they produce community assets, they can be quickly rolled in and out, responding to price shocks in a fiscally responsible manner, and they can contribute to peace building and social cohesion. In Ethiopia, for example, the successful Productive Safety Net Program emerged after two decades of ad hoc emergency food aid. This program uses community identification to reach poor and chronically food insecure households in need either through employment opportunities to able- workers or unconditional cash transfers to those unable to work (World Bank 2012b). In most social protection programs in Myanmar, which are mainly implemented by donors, public works (increasingly cash for work) also dominate, helping address seasonal unemployment and food insecurity (Infante-Villaroel, forthcoming) as reflected in the high seasonality of rice prices in Myanmar as discussed in Chapter 4.2. 103. It will be important for Myanmar to learn from these experiences when building its social protection system. The country does not need large government-owned rice stocks to later provide food transfers to the poor and vulnerable at times of food price spikes. Cash-based assistance, either through direct transfers to the most vulnerable or through public work programs, is more effective at achieving most social and economic objectives, and much better than food distribution from government-owned stocks.27 27 However, supplementary food-based transfers may remain relevant in isolated and thin food markets in rural areas, where cash transfers can cause additional increases in the price of food, or in emergency situations to replace nonfunctioning markets. 34 6. Conclusions and Recommendations Conclusions 104. Rice prices in Myanmar have been volatile. While some price volatility is common for agricultural markets and price volatility tends to be higher in rice-exporting countries (of which Myanmar is one) than in rice-importing countries, during 2004-2013 the volatility of domestic prices in Myanmar was above the level of its peers. This high volatility is of concern to the Myanmar government given the high importance of rice in farm incomes and consumer expenditures. 105. Pawsan prices have displayed the highest volatility. Pawsan is not widely produced and traded in Myanmar yet; as a result it is mostly exposed to local shocks. The prices of MWK have been much less volatile. 106. Export restrictions were used frequently in the past to reduce price volatility. To some extent, export restrictions might have helped reduce price volatility. The price volatility of exported Emata rice was lower than the volatility of Vietnamese and Thai rice export prices. It averaged 5.8 percent during 2004- 2013. The average price volatility was 6.2 percent in Thailand and 7.8 percent in Vietnam. 107. Yet the price volatility of exported Emata was lower during 2009-2013 than during 2004-2008. At the same time, both international and local trade has been more open in recent years than in the earlier period, implying that other factors played a larger role in defining volatility than export restrictions. 108. The lower volatility of export prices appears to have had little impact on the volatility of domestic prices. During 2004-2013, the average price volatility of MWK variety in Mandalay, Myingyan, Pathein, and Yangon markets was 60 percent higher than the price volatility of Emata, which is exported. The price volatility of high-quality, fragrant Pawsan was 90 percent higher. 109. In addition, the volatility of paddy prices was higher than the volatility of rice prices. During 2009-2012, the time period with available price data, the volatility of paddy prices was 40-110 percent higher depending on the market and rice variety. 110. While price volatility has moderated in recent years, the evolving concern is the rising price level. Emata prices increased by 41 percent in nominal US$ terms between 2009 and 2013, much more than in peer rice net-exporting countries. The increase of rice prices in Myanmar was a result of the large public wage increase in 2009, the increased cross-border trade in rice with China, and economic transition in general. By the end of 2013, Myanmar rice prices had equalized with prices in Vietnam and Thailand. These higher prices not only put competitive pressure on exporters but also put a high premium on careful management of price volatility to avoid further prices increases. 111. Why is the rice price volatility in local markets in Myanmar higher than the price volatility in other countries? Myanmar’s status as a net rice exporter does not explain much of this high domestic price volatility. The country is still weakly integrated with the world rice market; i.e., the world price’s volatility is not yet fully passed onto Myanmar. During 2004-2013, Vietnamese prices adjusted to the changes in Yangon prices but Emata prices in Yangon remained unresponsive to price changes in Vietnam, one of the major competitors of Myanmar’s rice exports. Moreover, the high market segmentation in Myanmar has weakened the link between its Emata prices and prices of domestic rice varieties. 112. Most drivers of domestic rice price volatility in Myanmar seem to be internal. Most price volatility is seasonal. The high concentration of paddy production in November and December results in sharp drops of paddy prices in December and January and sharp rises between May and October, depending on the market. Paddy prices have been more volatile than rice prices. This seasonal volatility is amplified by other 35 forces. Weak domestic market integration has prevented price signals from being transmitted quickly from surplus to deficit areas. Weak domestic integration is a result of the fragmented rice market, poor roads and low phone connectivity, the improvement of which will take time. But it is also a result of weak market transparency, which can be improved in the short run. Data on production, consumption, exports, and stocks, even if available, are not accurate. As a result, market participants and the government often overreact to small changes in prices; given the small size of private rice stocks, this ignites volatility. Weak geographic diversification of exports also contributes to price volatility in Myanmar. Most formal exports go to Africa, while China absorbs most of the cross-border exports. The latter are informal and reported to be highly unpredictable. Recommendations 113. Successfully reducing price volatility in Myanmar to achieve higher food security and lower poverty relies on addressing the structural causes of volatility. Better access to irrigation, seeds with different harvesting periods and growth duration, and advice on production technologies would allow farmers to spread production more evenly over the marketing year and would reduce the volatility of paddy prices. Investments in rural roads and telecom infrastructure would strengthen market integration: signals would pass more quickly from one market to another, including from world markets with lower price volatility than in Myanmar. More accurate and timely distributed market information would help the private sector and the government to make more rational decisions. Lower costs of doing business for rice mills and traders would reduce storage costs and increase private stocks which would buffer price fluctuations. Diversification of exports to additional markets – both geographically and by quality – supported by investments in sanitary and phytosanitary infrastructure, ports, and less costly export procedures, would further contribute to lower price volatility (see the summary of the policy recommendations in Table 21). 114. International experience suggests that safety nets can help mitigate the impact of volatile prices on the poor. Even if domestic price volatility is reduced through the policy actions and investments discussed above, many poor consumers will remain vulnerable to price spikes in the short run. International experience shows that effective safety nets can mitigate food price spikes, allowing vulnerable households to maintain their assets on which their livelihoods are based and adequately nourish and educate their children. Cash transfers, either direct or provided through public work programs, are often more successful than food subsidies at meeting the objectives of effective social safety nets. Myanmar can take advantage of the international experience to replicate best practices and avoid failures made by others. 115. A number of policy actions can be implemented quickly; they do not cost much (if anything) to the national budget, and would have a broad impact on the rice value chain’s participants. These priority actions can lead the way to the needed longer-term investments and policies, and without them it will be very difficult to significantly reduce price volatility. They include the following • Continue to promote an open trade policy. Stronger integration of Myanmar with world markets will not only lead to higher export volumes but will also lower domestic price volatility by encouraging investments in rice mills, private storage infrastructure, and export logistics. The lower volatility of the exported Emata prices is a powerful reminder to the power of open trade to reduce price volatility. • Make the rice milling industry more attractive for foreign direct investment by removing the industry from the list of strategic sectors requiring special government approval and permitting all types of investments, not just “green” ones. Myanmar needs cost-efficient and competitive mills to turn paddy into valuable rice for both domestic and foreign markets and improve rice quality, and thereby reducing price volatility. A modernization of the rice industry can trigger lasting changes at the farm level (higher productivity, better paddy quality, and more evenly spread production) and allow the industry absorbing a potentially larger paddy production, as has happened in Cambodia, Thailand, and Vietnam. Modernized rice mills will be also better positioned to increase the level of private stocks which would further reduce price volatility. 36 • Avoid exacerbating the current situation by not using distortive policy instruments. As an exporter, Myanmar is poised to have price volatility higher price volatility than seen in countries which import rice. Any direct short-term price stabilization measures need to be treated with caution so as not to undermine food security. Measures that raise domestic prices above world market prices such as large public stocks or farm minimum prices cannot be used by a country that aspires to become a large exporter and where domestic prices have already been dangerously approaching the world market levels. Measures that depress domestic prices below the world market level, such as export restrictions, would deprive farmers and hamper investments in rice mills and trade sectors. Table 21: Myanmar: Summary of policy recommendations Constraints Recommendations Implementing agencies Immediate priorities Unpredictable trade policy ignites • Commit to open trade Ministry of Commerce, in volatility and hampers investments policy, honor export license collaboration with Myanmar Rice in rice mills, storage, and trade commitments, avoid export bans, Federation infrastructure and collect more accurate market data and provide it to the public Treatment of rice mills as a strategic • Ease the rules for foreign direct Ministry of Commerce, in sector and restricting investments investment in rice mills collaboration with Myanmar Rice in existing mills discourages foreign Federation direct investment Currently discussed agricultural • Avoid using export restrictions, Ministry of Commerce and Minis- policy instruments are harmful for do not build government-owned try of Agriculture and Irrigation, in food security and poverty reduction rice stocks, and do not introduce collaboration with Myanmar Rice minimum farm prices Federation Longer-term priorities The high concentration of paddy • Invest in irrigation and water Ministry of Agriculture and Irrigation production in the last quarter of the management, improve access to year increases volatility of paddy seeds with different harvesting and rice prices periods, and strengthen public farm advisory services Weak domestic market integration • Invest in rural roads, develop Ministry of Livestock and Fisheries prevents price signals from being transport services in rural areas, (Rural Development Department) transmitted quickly, leading to ex- and invest in telecommunication cessive price drops in surplus areas infrastructure and price spikes in deficit areas Ministry of Communication and Information Technology Weak geographic diversification of • Reduce energy costs to mills, Ministry of Commerce and the exports increases price volatility reduce export processing costs Port Authority, in collaboration with and invest in ports, invest in Myanmar Rice Federation sanitary and phytosanitary infrastructure, and develop rice standards aligned with international practices Ministry of Energy (for investments in energy infrastructure) The lack of safety nets does not • Study international and domestic Ministry of Social Welfare, Relief allow effective protection of the poor experiences to establish cost- and Emergency from food price spikes effective safety nets able to mitigate the impact of food price spikes Ministry of Livestock and Fisheries (Rural Development Department) 37 7. References ADB. 2011. “Food Price Inflation, Poverty, and Inclusive Growth in Asia.” Presentation of the Chief Economist of ADB at the Commodity Price Volatility and Inclusive Growth in LICs workshop in Washington, DC. ADB. 2012. “Myanmar in Transition: Opportunities and Challenges.” Asian Development Bank, Manila. Ardeni, P. 1989. “Does the Law of One Price Really Hold for Commodity Prices?” American Journal of Agricultural Economics 71: 661-669. Barrett, C. and J. Li. 2002. “Distinguishing between Equilibrium and Integration in Spatial Price Analysis.” American Journal of Agricultural Economics 84: 292–307. BAS. 2013. “Statistics on Agricultural Profitability.” Philippine Bureau of Agricultural Statistics. Manila. Cafiero, C. and J. Schmidhuber. 2011. “AMIS Expert Meeting on Defining Abnormal Conditions.” FAO, December 19-20, 2011, Rome. CSO. 2014. “Selected Monthly Statistical Indicators.” The Central Statistical Office of Myanmar, Nay Pyi Taw. Dawe, D. and S. Jaffee. 2014. “Rice and Nutritional Security: Some Connections and Disconnections.” In: Jaffee, S., D. Dawe, N. Santos, and S. Mohanty (eds.): “Rice in the Shadow of Skyscrapers: Policy Choices in a Dynamic East Asian Setting.” World Bank and FAO (forthcoming). Dawe, D. and P. Timmer. 2013. “Why Stable Food Prices are a Good Thing: Lessons from Stabilizing Rice Prices in Asia.” Global Food Security 1: 127-133. DBM. 2013. “Income Statements, Balance Sheets, and Sources and Uses of Funds of the Government Owned Corporations.” Philippine Department of Budget Management, Manila. Dickey, D. and W. Fuller. 1979. “Distribution of the Estimators for Autoregressive Time Series with a Unit Root.” Journal of the American Statistical Association 74: 427–431. Engle, R. and C. Granger. 1987. “Cointegration and Error-Correction: Representation, Estimation and Testing.” Econometrica 55: 251-276. Fackler, P. and B. Goodwin. 2001. “Spatial Price Analysis.” In: Gardner B. and Rausser, G. (eds.): “Handbook of Agriculture Economics.” Vol. 1: 971-1024. Amsterdam: Elsevier Science B.V. FAO-GIEWS. 2014. “Food Price Data and Analysis Tool.” The UN Food and Agriculture Organization, www. http://www.fao.org/giews/pricetool/. FAOSTAT. 2014. “International Agricultural Statistics.” U.N. Food and Agriculture Organization, Rome. FAS-USDA. 2014. “Production, Supply and Distribution Online Database.” Foreign Agricultural Service, the US Department of Agriculture, Washington, DC. Fernandes, L. and R. Olfindo. 2011. “Overview of the Philippines’ Conditional Cash Transfer Program: The Pantawid Pamilya Program.” Social Protection Note No. 62879, World Bank and Australian Aid, Manila. Ghoshray, A. 2011. “Underlying Trends and International Price Transmission of Agricultural Commodities.” Asian Development Bank Economic Working Paper Series 257, Manila. 38 Gilbert, C. and C. Moran. 2010. “Food Price Volatility.” Phil. Trans. R. Soc. B 2010 365: 3023-3034. Gouel, C. and S. Jean. 2012.”Optimal Food Price Stabilization in a Small Open Developing Country.” World Bank Policy Research Working Paper 5943, Washington, DC. Greb, F., N. Jamora, C. Mengel, S. von Cramon-Taubadel and N. Würriehausen. 2012. “Price Transmission from International to Domestic Markets.” Department of Agricultural Economics and Rural Development, University of Göttingen. Report prepared for the World Bank’s Rural Policies Thematic Group of ARD, Development and Research Group’s Agriculture and Rural Development, and Poverty Reduction and Equity Unit. Grosh, M., C. del Ninno, E. Tesliuc, and A. Ouerghi. 2008. “For Protection and Promotion: The Design and Implementation of Effective Safety Nets.” World Bank, Washington, DC. Gupta, S. and R. Mueller. 1982. “Analyzing the Pricing Efficiency in Spatial Markets; Concept and Applications.” European Review Agricultural Economics 9: 301-312. Hassouneh, I., S. von Cramon-Taubadel, T. Serra and J. Gil. 2012. “Recent Developments in the Econometric Analysis of Price Transmission.” TRANSFOP (Transparency of Food Pricing) Working Paper No. 2, www.transfop.eu. IMF. 2014. “Selected Indicators. Online Database.” The International Monetary Fund, Washington, DC. Jaffee, S., D. Dawe, N. Santos and S. Mohanty. 2014. “Introduction and Objectives.” In: Jaffee, S., D. Dawe, N. Santos, and S. Mohanty (eds.): “Rice in the Shadow of Skyscrapers: Policy Choices in a Dynamic East Asian Setting.” World Bank and FAO (forthcoming). Johansen, S. 1988. “Statistical Analysis of Cointegrating Vectors.” Journal of Economic Dynamics and Control 12: 231-254. Kwiatkowski, D., P. Phillips, P. Schmidt and Y. Shin. 1992. “Testing the Null Hypothesis of Stationarity against the Alternative of a Unit Root.” Journal of Econometrics 54: pp. 159-178. LIFT. 2012. “Baseline Survey Results of Myanmar Agriculture.” Livelihood and Food Security Trust Fund, Yangon. Meyer, J. and S. von Cramon-Taubadel. 2004. “Asymmetric Price Transmission: A Survey.” Journal of Agricultural Economics 55(3): 581-611. MIS/MOAI. 2014. “Wholesale Rice Prices.” Market Information System, Ministry of Agriculture and Irrigation of Myanmar, Nay Pyi Taw. MOC. 2014. “Paddy and Rice Prices.” Ministry of Commerce of Myanmar, Nay Pyi Taw. Nipon, P. 2014. “The Benefits and Costs of a Paddy Pledging Policy: Insights from a Case Study of Thailand.” In: S. Jaffee, D. Dawe, N. Santos, and S. Mohanty (eds): “Rice in the Shadow of Skyscrapers: Policy Choices in a Dynamic East Asian Setting.” World Bank, FAO, IRRI. Pardey, P., J. Alston and R. Piggott (eds.) 2006. “Agricultural R&D in the Developing World: Too Little, Too Late?” International Food Policy Research Institute, Washington, DC. Ravallion, M. 1986. “Testing Market Integration.” American Journal of Agriculture Economics 68: 102-109. Skoufias, E., S. Tiwari and H. Zaman. 2011. “Can We Rely on Cash Transfers to Protect Dietary Diversity During Crises? Estimates from Indonesia.” World Bank Policy Research Working Paper 5548, Washington, DC. 39 Slayton, T. and S. Muniroth. 2012. “Cambodia: Turning Rice into White Gold.” Report for the World Bank, August 2012. Timmer, P. 2004. “Food Security and Economic Growth: An Asian Perspective.” Canberra: Heinz W. Arndt Memorial Lecture, November 22, 2004. Velande, R. and L. Fernandes. 2011. “Welfare and Distributional Impacts of the Pantawid Pamilyang Pilipino Program.” Social Protection Note 63418, World Bank and Australian Aid, Manila. Infante-Villarroel, M. forthcoming. Cash Assistance in Myanmar: Lessons from a Social Protection and Poverty Reduction Perspective, and Public Works Programs. World Bank, Washington, DC. Warr, D. 2013. “Food Security, Agriculture, and Poverty in Asia.” ReSAKSS Asia Working Paper 1, December 2013. World Bank. 2007. “Philippines: Agriculture Public Expenditure Review.” Technical Working Paper, Manila and Washington, DC. World Bank. 2009. “Indonesia: Agriculture Public Expenditure Review.” Economic and Sector Work, Washington, DC. World Bank. 2012a. “Responding to Higher and More Volatile Global Food Prices.” Economic and Sector Work, Agriculture and Rural Development Department Report No. 68420-GLB, Washington, DC. World Bank. 2012b. “Using Public Foodgrain Stocks to Enhance Food Security.” Economic and Sector Work, Agriculture and Rural Development Department Report No. 71280-GLB, Washington, DC. World Bank. 2014a. “Cambodia: Rice Sector Monitoring in 2013.” Technical Assistance Note, Phnom Penh. World Bank. 2014b. “Myanmar: Capitalizing on Rice Export Opportunities.” Economic and Sector Work, Sustainable Development Network, East Asia and Pacific Region, Report No. 85804, Washington, DC. World Bank. 2014c. “Poverty in Myanmar: Revisiting the 2009/10 IHLCA.” Unpublished report. Washington, DC. World Bank. 2014d. “Promoting Competitiness and Sustainable Growth.” East Asia and Pacific Economic Update. April 2014, Washington, DC. 40 8. Annexes Annex 1: Rice Prices in Myanmar Nominal prices of Pawsan, Kyats per ton Year Month Yangon Pathein Mandalay Myingyan 2004 January 168,878 131,665 170,715 125,194 February 163,591 104,883 176,472 122,479 March 192,067 112,272 183,392 139,401 April 200,274 141,116 177,676 136,000 May 199,477 161,264 192,904 134,727 June 199,028 150,547 190,516 134,482 July 199,763 150,302 191,088 134,625 August 199,273 156,794 185,964 145,872 September 201,315 157,385 212,766 162,591 October 201,029 152,588 212,154 155,650 November 184,535 126,317 212,236 117,641 December 169,511 118,396 151,364 114,212 2005 January 163,489 132,686 155,650 117,764 February 163,509 127,582 170,307 194,884 March 164,877 139,218 180,167 199,865 April 169,123 151,057 192,108 197,354 May 171,634 154,630 204,336 210,929 June 175,472 164,938 205,765 226,239 July 181,187 170,450 214,277 233,057 August 185,678 173,512 217,931 232,975 September 188,046 177,860 217,870 238,160 October 223,687 195,721 238,814 244,958 November 250,286 198,824 259,288 225,014 December 175,553 173,512 246,856 212,501 2006 January 198,049 168,409 212,297 221,483 February 209,623 191,373 224,422 243,713 March 243,182 216,890 261,962 301,604 April 280,579 257,716 306,687 367,437 May 300,318 274,557 335,837 360,558 June 315,036 290,888 332,102 353,148 July 310,995 347,024 377,194 449,090 August 348,841 369,989 442,945 451,131 September 372,622 405,712 468,931 444,599 October 406,018 416,776 487,773 422,042 November 376,174 408,467 506,532 423,430 December 336,144 374,173 350,045 423,859 41 2007 January 424,165 360,292 419,143 469,156 February 434,270 384,033 460,154 483,547 March 457,867 357,230 457,622 472,912 April 466,400 363,620 467,911 511,615 May 466,563 398,057 485,752 524,496 June 491,488 422,308 498,898 523,598 July 514,024 440,924 521,903 539,234 August 517,678 445,272 544,358 566,853 September 521,291 455,009 548,073 591,124 October 521,250 393,974 553,299 591,124 November 457,255 295,991 448,559 531,559 December 433,576 286,805 383,563 530,865 2008 January 457,602 316,669 428,676 281,048 February 472,259 322,793 409,345 454,805 March 468,196 326,202 407,120 469,115 April 485,262 340,226 412,693 479,934 May 793,664 418,470 528,844 569,405 June 747,959 476,300 590,287 586,491 July 734,098 382,747 586,756 595,044 August 727,015 520,536 618,356 597,085 September 733,465 534,825 696,273 597,289 October 728,750 530,742 754,103 597,085 November 714,767 461,848 698,110 524,006 December 652,466 408,896 432,085 449,090 2009 January 652,466 292,174 401,057 458,071 February 593,574 306,197 406,283 464,971 March 595,819 315,383 408,896 454,764 April 615,294 336,817 452,641 485,833 May 563,056 353,413 515,432 504,838 June 502,674 350,086 533,886 505,226 July 473,055 385,298 553,666 505,226 August 485,425 366,416 552,156 518,637 September 457,724 357,230 557,361 536,621 October 498,489 394,239 539,949 553,605 November 453,580 389,891 519,352 556,259 December 357,230 341,920 489,916 393,729 2010 January 359,108 341,920 425,574 421,613 February 439,802 340,655 463,828 432,922 March 471,544 417,449 513,085 488,222 April 480,220 398,057 522,005 504,695 May 492,059 414,387 516,004 519,168 June 485,609 408,263 545,787 556,259 July 487,262 428,676 542,582 542,276 42 August 521,556 438,883 585,388 535,846 September 515,432 451,641 611,864 535,846 October 534,478 438,883 634,380 615,926 November 618,519 443,313 697,579 625,255 December 532,947 436,495 663,305 556,259 2011 January 483,792 392,953 551,911 547,420 February 501,919 461,848 581,632 581,571 March 502,164 474,606 643,770 617,498 April 573,773 459,296 633,665 633,951 May 571,956 481,751 634,196 658,324 June 564,220 531,518 614,661 658,488 July 577,141 603,883 632,277 658,324 August 609,537 609,843 640,545 658,324 September 615,314 601,331 663,162 688,944 October 650,302 581,775 663,550 704,254 November 666,061 582,632 677,166 704,254 December 571,569 551,666 661,672 642,586 2012 January 483,547 365,498 530,048 556,259 February 488,793 466,604 539,030 569,527 March 520,413 517,474 574,488 603,209 April 537,377 519,311 599,922 607,292 May 536,050 517,474 591,308 608,312 June 534,315 512,023 578,366 607,292 July 536,866 520,536 568,160 622,601 August 541,765 530,069 596,064 649,139 September 555,238 577,692 632,808 683,841 October 547,073 528,599 613,763 683,841 November 520,536 472,912 566,465 683,841 December 480,995 376,357 580,938 683,841 2013 January 418,470 414,734 514,595 683,841 February 435,474 430,759 509,186 500,122 March 431,228 434,943 530,742 500,122 April 456,234 438,883 527,333 357,230 May 485,017 464,604 564,771 392,953 June 502,674 474,953 589,430 548,604 July 506,246 495,019 606,822 571,569 August 551,155 525,027 663,428 571,569 September 540,949 535,846 636,217 571,569 October 530,742 536,866 611,721 571,569 November 530,742 512,881 611,721 571,569 December 477,158 581,775 557,953 571,569 Source: MIS/MOAI (2014). 43 Nominal prices of Manawthukha, Kyats per ton Year Month Yangon Pathein Mandalay Myingyan 2004 January 89,777 79,101 111,190 61,934 February 82,571 72,732 112,252 62,076 March 79,142 85,225 113,722 64,771 April 82,020 100,024 121,377 69,670 May 82,755 103,086 125,357 71,487 June 85,613 104,107 145,240 71,977 July 92,512 107,169 160,937 71,548 August 97,269 108,455 158,631 72,936 September 103,086 124,929 157,773 79,019 October 101,739 119,927 149,036 78,938 November 94,105 85,490 145,240 62,832 December 89,001 104,311 109,353 64,342 2005 January 92,267 99,514 113,946 70,670 February 108,700 103,862 128,317 112,660 March 113,579 115,743 138,095 125,071 April 116,763 128,950 150,200 140,994 May 124,684 133,461 156,691 148,077 June 126,807 142,280 169,429 159,039 July 131,665 148,771 182,453 166,000 August 142,116 144,321 186,066 156,508 September 144,505 149,016 190,863 155,365 October 161,529 148,669 200,886 170,287 November 172,573 151,874 193,578 158,427 December 134,604 133,461 183,269 136,727 2006 January 138,115 139,320 176,247 153,446 February 144,260 148,261 199,335 170,634 March 168,266 171,226 204,703 193,129 April 182,555 192,394 237,017 229,199 May 189,475 204,132 271,801 240,426 June 198,865 218,176 260,676 235,915 July 211,440 244,958 287,029 242,345 August 223,340 214,338 291,275 256,859 September 225,361 245,223 279,109 258,921 October 224,994 262,309 297,624 299,277 November 213,154 223,524 314,301 317,853 December 215,502 239,099 254,552 262,432 2007 January 269,372 261,901 304,299 285,846 February 277,272 298,298 342,492 327,264 March 287,009 290,132 344,268 328,346 April 308,443 301,094 354,025 329,183 44 May 307,524 319,670 356,230 340,594 June 329,326 339,369 364,844 345,187 July 344,676 344,574 382,563 335,756 August 350,249 348,044 394,913 322,528 September 354,944 339,369 408,263 319,731 October 352,433 321,099 390,443 327,978 November 334,204 265,371 297,297 275,741 December 326,611 267,004 264,228 269,903 2008 January 333,041 257,206 280,885 291,480 February 336,817 265,636 302,196 297,583 March 335,102 273,945 331,530 306,585 April 328,427 273,945 336,470 318,772 May 587,736 327,897 375,745 371,356 June 408,263 332,735 377,378 369,784 July 387,462 315,894 378,174 304,238 August 367,437 275,067 387,279 288,295 September 360,537 275,986 384,849 282,416 October 363,354 264,105 395,832 277,884 November 340,818 260,268 259,166 245,897 December 309,300 207,194 213,950 213,113 2009 January 309,300 197,252 228,627 224,545 February 286,295 194,190 236,160 244,958 March 280,579 219,033 260,207 250,245 April 278,578 233,731 277,905 282,661 May 261,962 251,347 297,563 306,973 June 258,227 260,268 302,013 301,094 July 288,540 263,085 315,138 301,094 August 269,454 263,330 333,163 287,479 September 267,719 270,474 339,389 284,355 October 284,764 299,829 325,120 298,420 November 273,863 283,335 320,323 324,855 December 269,229 274,822 329,489 321,936 2010 January 275,231 279,415 348,065 350,555 February 275,945 310,035 350,433 359,353 March 273,536 324,569 361,293 393,790 April 286,805 311,301 386,401 392,953 May 299,216 311,301 395,280 389,810 June 293,174 326,611 381,971 402,384 July 293,643 338,103 387,605 393,709 August 312,117 353,148 416,714 392,953 September 309,872 334,266 428,146 392,953 October 310,443 331,714 426,717 413,367 November 294,970 312,934 440,108 418,266 45 December 276,598 279,926 424,390 341,920 2011 January 267,004 278,027 353,964 333,082 February 280,436 304,932 361,333 369,274 March 292,276 305,177 396,546 414,530 April 294,480 289,520 392,096 406,283 May 287,744 280,681 382,298 392,953 June 284,641 278,640 370,703 390,912 July 279,191 274,557 360,884 382,747 August 279,987 272,618 356,230 382,747 September 284,049 270,474 357,312 394,321 October 289,336 271,760 361,272 403,160 November 280,334 251,347 368,886 387,074 December 269,597 238,324 353,964 318,098 2012 January 262,105 226,178 278,170 270,474 February 263,983 231,955 295,195 285,437 March 276,374 231,342 342,635 307,218 April 277,109 242,917 341,349 311,301 May 271,903 234,751 314,710 305,993 June 273,536 239,855 330,693 290,888 July 287,315 257,553 327,468 311,301 August 302,115 264,493 342,267 328,652 September 319,466 265,902 358,088 357,230 October 320,078 279,068 359,272 357,230 November 312,832 281,498 369,131 353,413 December 301,094 281,702 375,949 357,230 2013 January 281,967 265,371 332,571 357,230 February 306,197 292,582 359,496 362,334 March 328,142 295,991 386,156 362,334 April 323,549 299,400 398,057 500,122 May 328,244 306,197 412,489 540,949 June 316,404 338,103 431,228 399,343 July 340,900 347,024 458,337 420,511 August 387,850 358,006 484,813 418,470 September 347,024 336,817 510,329 418,470 October 371,928 355,597 515,085 418,470 November 371,928 355,597 515,085 418,470 December 275,578 285,784 432,759 418,470 Source: MIS/MOAI (2014). 46 Nominal prices of Emata, Kyats per ton Year Month Yangon 2004 January 82,673 February 78,162 March 78,938 April 78,060 May 74,508 June 79,326 July 84,633 August 87,144 September 97,269 October 86,368 November 84,613 December 84,715 2005 January 86,450 February 90,349 March 99,596 April 106,883 May 127,154 June 113,028 July 118,090 August 123,785 September 122,091 October 125,357 November 125,541 December 125,541 2006 January 129,624 February 132,624 March 138,238 April 166,367 May 182,514 June 180,126 July 203,989 August 192,272 September 189,883 October 205,030 November 227,382 December 213,522 2007 January 220,319 February 227,607 March 228,627 April 229,648 47 May 229,648 June 229,648 July 230,383 August 239,855 September 239,855 October 238,242 November 228,321 December 243,529 2008 January 266,678 February 268,433 March 268,433 April 268,433 May 319,588 June 307,892 July 301,094 August 301,094 September 301,094 October 298,032 November 298,032 December 298,032 2009 January 298,032 February 298,032 March 274,353 April 274,353 May 224,953 June 224,545 July 204,132 August 204,132 September 224,545 October 242,917 November 224,545 December 234,343 2010 January 240,875 February 255,165 March 255,165 April 255,165 May 257,859 June 276,231 July 279,660 August 279,660 September 279,660 October 279,660 November 277,619 48 December 269,454 2011 January 268,229 February 279,660 March 279,252 April 274,353 May 259,247 June 251,898 July 244,958 August 249,041 September 268,637 October 269,454 November 265,371 December 239,508 2012 January 225,035 February 231,138 March 234,751 April 234,751 May 228,627 June 231,689 July 259,247 August 273,128 September 290,888 October 295,991 November 291,908 December 281,702 2013 January 276,802 February 279,252 March 274,557 April 270,821 May 293,950 June 303,135 July 303,135 August 303,135 September 311,301 October 284,253 November 284,253 December 277,619 Source: MIS/MOAI (2014). 49 Inflation and Exchange rate in Myanmar Year Month Consumer Price Index Exchange Rate, Kyats/US$ (Jan. 2004 = 100) 2004 January 1.000 884 February 0.977 859 March 0.970 849 April 0.969 834 May 0.976 888 June 0.983 944 July 0.997 950 August 1.007 952 September 1.016 947 October 1.025 915 November 1.035 953 December 1.039 937 2005 January 1.048 911 February 1.035 913 March 1.044 914 April 1.058 948 May 1.063 985 June 1.068 1,010 July 1.078 1,068 August 1.092 1,147 September 1.117 1,234 October 1.154 1,296 November 1.172 1,202 December 1.188 1,150 2006 January 1.156 1,136 February 1.159 1,146 March 1.176 1,187 April 1.186 1,364 May 1.233 1,298 June 1.291 1,282 July 1.325 1,326 August 1.371 1,343 September 1.395 1,369 October 1.434 1,349 November 1.488 1,398 December 1.525 1,268 2007 January 1.557 1,292 February 1.593 1,272 March 1.631 1,275 50 April 1.671 1,272 May 1.711 1,268 June 1.770 1,283 July 1.806 1,295 August 1.844 1,318 September 1.872 1,373 October 1.904 1,359 November 1.932 1,294 December 1.962 1,279 2008 January 2.137 1,253 February 2.134 1,200 March 2.159 1,115 April 2.171 1,118 May 2.232 1,139 June 2.238 1,167 July 2.280 1,183 August 2.331 1,203 September 2.338 1,263 October 2.341 1,235 November 2.312 1,262 December 2.276 1,206 2009 January 2.235 1,124 February 2.215 1,041 March 2.200 1,027 April 2.235 1,060 May 2.243 1,079 June 2.262 1,113 July 2.292 1,129 August 2.312 1,118 September 2.336 1,095 October 2.346 1,083 November 2.345 1,011 December 2.325 1,002 2010 January 2.354 1,014 February 2.362 1,011 March 2.369 1,000 April 2.412 997 May 2.442 990 June 2.472 984 July 2.475 994 August 2.484 997 September 2.487 958 October 2.515 916 51 November 2.541 903 December 2.544 879 2011 January 2.552 847 February 2.561 881 March 2.579 881 April 2.615 855 May 2.625 827 June 2.606 788 July 2.580 792 August 2.566 792 September 2.562 791 October 2.558 806 November 2.569 782 December 2.561 788 2012 January 2.552 816 February 2.550 814 March 2.551 813 April 2.562 815 May 2.578 835 June 2.587 862 July 2.597 870 August 2.622 864 September 2.668 850 October 2.696 845 November 2.711 844 December 2.715 853 2013 January 2.693 855 February 2.667 860 March 2.671 869 April 2.680 881 May 2.694 905 June 2.697 943 July 2.713 980 August 2.729 975 September 2.733 970 October 2.734 971 November 2.738 983 December 2.733 985 Source: IMF and CSO (for CPI) and www.exchangerate.com (for exchange rate). 52 Annex 2: Price Volatility Price Return: Volatility of Pawsan Prices Nominal Prices Real Prices Yangon Pathein Mandalay Myingyan Average Yangon Pathein Mandalay Myingyan Average 2004-2008 10.5 11.1 12.5 14.8 12.2 10.4 10.8 12.1 15.5 12.2 2009-2013 7.8 10.4 6.7 9.8 8.7 7.6 10.3 6.5 9.6 8.5 2004-2013 9.3 10.7 10.0 12.5 10.6 9.1 10.5 9.6 12.9 10.5 Price Return: Volatility of MWK Prices Nominal Prices Real Prices Yangon Pathein Mandalay Myingyan Average Yangon Pathein Mandalay Myingyan Average 2004-2008 11.3 10.1 10.8 10.6 10.7 11.0 10.2 10.6 10.8 10.6 2009-2013 6.1 5.8 6.6 8.5 6.7 5.9 5.7 6.5 8.4 6.7 2004-2013 9.1 8.2 8.9 9.5 8.9 8.8 8.2 8.7 9.6 8.9 Price Return: Volatility of Emata Prices, Yangon Nominal Prices Real Prices 2004-2008 6.1 6.0 2009-2013 5.2 5.1 2004-2013 5.7 5.6 Coefficient of Variation: Volatility of Pawsan Prices Nominal Prices Real Prices Yangon Pathein Mandalay Myingyan Average Yangon Pathein Mandalay Myingyan Average 2004-2008 53 46 46 48 48 24 26 23 29 25 2009-2013 12 18 12 15 14 14 15 10 13 13 2004-2013 37 37 35 35 37 21 21 18 23 21 Coefficient of Variation: Volatility of MWK Prices Nominal Prices Real Prices Yangon Pathein Mandalay Myingyan Average Yangon Pathein Mandalay Myingyan Average 2004-2008 52 42 39 47 45 26 24 21 33 26 2009-2013 9 14 17 17 14 8 13 13 14 12 2004-2013 35 32 32 39 34 23 22 19 25 22 Coefficient of Variation: Volatility of Emata Prices, Yangon Nominal Prices Real Prices 2004-2008 44 19 2009-2013 10 9 2004-2013 32 17 53 Price Return: Volatility of Thai and Vietnamese Export Prices, 25% brokens Thailand Vietnam Nominal Prices Real Prices Nominal Prices Real Prices 2004-2008 8.9 8.7 8.9 8.7 2009-2013 8.1 8.2 7.0 6.9 2004-2013 8.5 8.4 7.9 7.8 Coefficient of Variation: Volatility of Thai and Vietnamese Export Prices, 25% brokens Thailand Vietnam Nominal Prices Real Prices Nominal Prices Real Prices 2004-2008 49 47 49 36 2009-2013 13 13 13 17 2004-2013 34 33 34 30 54 Annex 3: Rice Market Integration in Myanmar 3.1 Analytical framework 1. The study of price transmission (PT) for homogeneous commodities in space, or for a product as it is transformed along the stages of the marketing chain, has attracted the interest of agricultural economists for many decades (Meyer and von Cramon-Taubadel 2004). Early empirical studies of PT were based on simple correlation and regression analyses that did not account for dynamics and lead-lag relationships in price data (Fackler and Goodwin 2001). In the course of the 1980s, these methods were increasingly replaced by dynamic regression models that include lagged prices (Ravallion 1986) and studies based on the concept of Granger causality (Gupta and Mueller 1982). 2. The emerging cointegration literature highlighted several pitfalls associated with the regression analysis of price data. In particular, since price data are often non-stationary, regression can lead to spurious results (Hassouneh et al. 2012). The basic insight of the cointegration approach is that to avoid the pitfall of spurious regression, one must test whether non-stationary prices series (also referred to as “integrated” price series) are not only correlated with one another but are also “co-integrated”. Cointegrated means that there exists a linear combination of the non-stationary series that is itself stationary: in other words, the series share a common form of non-stationarity and cannot drift apart indefinitely. 3. Ardeni (1989) published the first study of PT in agricultural markets based on cointegration methods. It is fair to say that with the exception of a comparatively small literature based on so-called parity bounds models (Barrett and Li 2002), essentially the entire empirical PT literature today draws on cointegration methods and, in particular, the so-called vector error correction model (VECM). The VECM is a re-parameterization of the standard vector autoregressive model (VAR), which relates the current levels of a set of time series to lagged values of those series. A simple VECM that captures the interactions between international or world prices and domestic price takes the following form: where is the domestic price; is the world price; and ᵠ,ᵃ,ᵝ,ᵟ, and ᵖ are parameters to be estimated. 4. In matrix notation, and allowing for more than one lag of the price difference terms, this VECM can be written compactly as: 5. From the perspective of empirical PT analysis, the main advantage of the VECM over the VAR is that it separates the long-run equilibrium (or cointegrating) relationship between and – which is captured by the error correction term – from the short-run dynamics that ensure that any deviations from this long-run equilibrium are corrected and thus only temporary. The key parameters in the VECM are , which describes how one price reacts to changes in the other in the long run,28 and the so-called “adjustment” parameters, and . If and are cointegrated, then and must have negative and positive signs, respectively. If this is the case, then if for example 28 If estimation is based on prices in logarithms, then can be interpreted as the long-run elasticity of PT. 55 becomes too large relative to and the error correction term is correspondingly positive, a decrease in in the first equation of the VECM and an increase in in the second equation, will drive the prices back towards their long-run equilibrium. One-to-one PT in the long run requires that =1,while , with large (small) values of and indicating that errors are corrected rapidly (slowly).29 6. Figure 15 outlines the basic empirical strategy for estimating PT. The first step is to determine whether the individual price series and are both non-stationary (also referred to as integrated or I(1)). This is usually carried out using the ADF (Dickey and Fuller 1979) and KPSS tests (Kwiatkowski et al. 1992). If the prices are not both I(1), they cannot be cointegrated. If they are both stationary or “I(0)” they can be studied using Auto-Regressive Distributed Lag models. If the series are both I(1), the null hypothesis that they are not cointegrated can be tested using a two-step OLS procedure proposed by Engle and Granger (1987) or a maximum likelihood procedure developed by Johansen (1988). If the null of no cointegration is rejected, the VECM in equation (2) can be estimated, again using methods proposed by Engel and Granger or Johansen. Finally, the resulting estimates of β and α are interpreted. Figure 15: Conceptual framework for assessing PT and market integration Test for order of intergration No cointegration Test Ho : no cointegration Estimate ARDL Specify and estimate VECM; assess Assess overall PT and market dynamics, apeeds of adjustment integration Source: Greb et al. (2012). 3.2 Empirical analyses Integration of international prices 7. The analysis of PT in Myanmar begins with the study of international integration. The first step is to test for the stationarity properties of the univariate time series. The ADF test is used to test the order of integration with the hypothesis of H0: β = 0 vs. H1: β < 0. Optimum lag length is determined based on the Akaike Information Criterion. Monthly data on nominal prices of Vietnam (FOB, 25% brokens), Myanmar (Emata), and Myanmar (Manawthukha) denominated in US$ are used to test the hypothesis. The results of the unit root test (with intercept and trend) indicate that all three times series are I(1) implying that they can be cointegrated (Table 22) and providing a justification for using the first difference of the price series to conduct cointegration analysis. 29 The speed of error correction captured by the magnitude of an adjustment parameter must be interpreted relative to the frequency of the data used to estimate it. An α of 0.4 estimated with annual data implies that 40 percent of any deviation from the long-run equilibrium is corrected within the space of one year. An α of 0.25 estimated with monthly data is smaller in magnitude but would nevertheless lead to over 95 percent correction of any deviation from the long-run equilibrium in the course of one year. Some authors transform α’s into so-called half-lives that indicate how many units of time are required for the correction of one-half of a deviation from the long-run equilibrium. An α of 0.25 estimated with monthly data corresponds to a half-life of 2.41 months. 56 Table 22: ADF test results: Vietnam and Yangon prices Market pairs 2004-2013 2004-2008 2009-2013 p-value p-value p-value Vietnam log (p) 0.05 0.02 0.09 ∆ log (p) 0.01 0.03 0.01 Emata log (p) 0.66 0.03 0.34 ∆ log (p) 0.01 0.01 0.01 Manawthukha log (p) 0.81 0.07 0.98 ∆ log (p) 0.01 0.01 0.01 Source: Authors’ estimates based on data from MIS/MOAI (2014) and FAO-GIEWS (2014). 8. The next step is the cointegration test. Cointegration is tested using the Johansen trace test. The summary in Table 23 shows a strong cointegration between both market pairs (r^ = 1) for the period between 2004 and 2013. In both subperiods, before the global food price spike (2004-2008) and after the global food price spike (2009-2013), the cointegration is found to have been insignificant, with r^ = 0. This result is likely to occur due to the short time price series (60 observations each) when divided into two subperiods. The results of the PT analysis are presented below for both subperiods, but they need to be interpreted with great caution given the insignificant results of the Johansen trace test. Table 23: Johansen test trace results: Vietnam and Yangon prices 2004-2013 2004-2008 2009-2013 H(r) Q p-val ^r Q p-val ^r Q p-val ^r Manawthukha Vietnam - Yangon 0 23.89 0.01 1 14.32 0.27 0 12.93 0.38 0 1 4.61 0.34 Emata Vietnam - Yangon 0 21.49 0.03 1 14.62 0.25 0 13.55 0.33 0 1 4.65 0.33 Note: The hypothesis tested H(r): Number of cointegrating relationships ≤ r, where r = 0 or r = 1, against H(2); Q denotes the value of the test statistics; p-val is the p-value; and r^ is the estimated number of cointegrating relationships. Cointegrated pairs are bolded. Source: Authors’ estimates based on data from MIS/MOAI (2014) and FAO-GIEWS (2014). 9. The average PT elasticity (β) between Myanmar and Vietnam prices was 51 percent during 2004- 2013 (Table 24). Prices of Emata and MWK rice behaved similarly in relation to Vietnamese export prices, underpinning the large share of MWK rice used for Emata assembled for export. A much stronger PT elasticity was detected during 2004-2008 than 2009-2013. It is surprising that the PT abruptly decreased to 10 percent with the opening of the economy during 2009-2013. But as discussed above, the results for subperiods should not be considered as statistically significant due to the weak power estimate (see Table 23). Table 24: Estimates of PT elasticities between Myanmar and world markets 2004-2013 2004-2008 2009-2013 Manwthukha Vietnam – Yangon 0.51 0.59 0.11 Emata Vietnam-Yangon 0.52 0.73 0.10 Source: Authors’ estimates based on data from MIS/MOAI (2014) and FAO-GIEWS (2014). 57 10. The integration of the Yangon market of Emata and MWK with Vietnamese prices was apparently non-existent during 2004-2013. Table 25 shows the parameters of price adjustments: α1 for an adjustment of Vietnamese prices to changes in Yangon prices and α2 for an adjustment of Yangon prices to changes in Vietnamese prices. There was no significant adjustment in Yangon prices of both Emata and MWK to the changes in Vietnamese prices during either the whole study period or the two subperiods. On the other hand, Vietnamese prices reacted to changes in Yangon prices, increasingly so from 2009. When Yangon prices either increased or decreased, it took 6-7 months for Vietnamese prices to correct half of the price disequilibrium. But when Vietnamese prices changed, Yangon prices were not affected, pointing to the (still) weak integration of Myanmar with world markets. Table 25: Estimates of the adjustment parameters between Yangon and Vietnam prices Market pairs 2004-2013 2004-2008 2009-2013 α1 α2 α1 α2 α1 α2 Manawthukha Vietnam-Yangon -0.16*** -0.03 -0.15*** -0.03 -0.21** -0.06 Emata Vietnam-Yangon -0.14*** 0.00 -0.15*** -0.01 -0.21** -0.05 Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates based on data from MIS/MOAI (2014) and FAO-GIEWS (2014). Integration of domestic prices 11. The results of the unit root test indicate that the all times series are I(1) implying that they can be cointegrated (Table 26). This provides a justification for using the first difference of the price series to conduct cointegration analysis. Domestic prices are expressed in nominal Kyats. Table 26: ADF test results: Domestic prices in Myanmar log (p) ∆log (p) 2004-13 2004-08 2009-13 2004-13 2004-08 2009-13 Surplus markets Yangon-Pawsan 0.76 0.38 0.09 ≤0.01 ≤0.01 ≤0.01 Yangon-MWK 0.81 0.07 0.98 ≤0.01 ≤0.01 ≤0.01 Pathein-Pawsan 0.47 0.44 0.19 ≤0.01 ≤0.01 ≤0.01 Pathein-MWK 0.50 0.96 0.44 ≤0.01 ≤0.01 ≤0.01 Deficit markets Mandalay-Pawsan 0.37 0.31 0.06 ≤0.01 ≤0.01 ≤0.01 Mandalay-MWK 0.18 0.80 0.50 ≤0.01 ≤0.01 ≤0.01 Myingyan-Pawsan 0.42 0.47 0.27 ≤0.01 ≤0.01 ≤0.01 Myingyan-MWK 0.47 0.96 0.34 ≤0.01 ≤0.01 ≤0.01 Source: Authors’ estimates based on data from MIS/MOAI (2014). 58 12. The results of the Johansen trace test indicate a mixed picture of cointegration. For Pawsan, only three market pairs are found to have been cointegrated during 2004-2013 (Table 27): Pathein-Mandalay, Pathein-Myingyan, and Mandalay-Myingyan. The situation is similar for MWK. As shown in Table 28, only three markets show signs of strong cointegration: Yangon-Myingyan, Pathein-Myingyan, and Mandalay-Myingyan. There are indications of the increased cointegration of more market pairs during 2009-2013, but the small sample size for the two subperiods does not provide strong enough justification to consider these results. Table 27: Johansen test trace results: Pawsan prices 2004-2013 2004-2008 2009-2013 H(r) Q p-val ^r Q p-val ^r Q p-val ^r Yangon - Pathein 0 15.61 0.20 0 11.52 0.50 0 24.34 0.01 2 1 10.04 0.03 Yangon - Mandalay 0 15.43 0.21 0 10.71 0.58 0 35.34 0.00 2 1 9.16 0.05 Yangon - Myingyan 0 17.71 0.11 0 11.27 0.52 0 24.88 0.01 2 1 9.97 0.03 Pathein - Mandalay 0 30.33 0.00 1 24.36 0.01 1 19.29 0.07 0 1 7.34 0.11 4.96 0.30 Pathein - Myingyan 0 33.07 0.00 1 18.99 0.07 0 24.53 0.01 1 1 8.42 0.07 7.68 0.10 Mandalay - Myingyan 0 30.66 0.00 1 20.33 0.05 1 27.77 0.00 2 1 6.99 0.13 4.26 0.39 12.51 0.01 Notes: Similar explanation as for Table 23. Cointegrated pairs are in bold. Source: Authors’ estimates based on data from MIS/MOAI (2014). Table 28: Johansen test trace results: Manawthukha prices 2004-2013 2004-2008 2009-2013 H(r) Q p-val ^r Q p-val ^r Q p-val ^r Yangon - Pathein 0 25.62 0.01 2 18.23 0.09 0 21.87 0.03 1 1 10.95 0.02 6.56 0.16 Yangon - Mandalay 0 19.61 0.06 0 15.27 0.22 0 31.27 0.00 1 1 7.02 0.13 Yangon - Myingyan 0 20.22 0.05 1 15.69 0.19 0 22.04 0.03 1 1 8.06 0.08 7.08 0.13 Pathein - Mandalay 0 39.78 0.00 2 28.63 0.00 1 17.03 0.13 0 1 11.27 0.02 6.02 0.20 Pathein - Myingyan 0 28.96 0.00 1 17.36 0.12 0 24.00 0.01 2 1 7.52 0.10 9.82 0.04 Mandalay - Myingyan 0 32.26 0.00 1 17.92 0.10 0 24.81 0.01 1 1 8.33 0.07 8.58 0.06 Notes: Similar explanation as for Table 23. Cointegrated pairs are in bold. Source: Authors’ estimates based on data from MIS/MOAI (2014). 59 13. The price integration of domestic market pairs is higher than the integration with world markets, but is still relatively weak. Only Mandalay and Myingyan prices (deficit markets) adjusted to the changes in Yangon and Pathein (surplus markets), with little transmission of price changes among other markets. For many market pairs, the integration did not exist at all. Where it existed, it was slow. When Pawsan prices in Yangon market changed during 2009-2013, half of this price change was passed to Mandalay prices only after five months. Market integration is generally weaker for Pawsan (Table 29) than for MWK (Table 30). 14. Overall, longer time series and probably better quality data are required to more precisely assess market integration in Myanmar. The data used in this report points to the existence of fractured markets inside the country and weak transfer of signals from deficit markets to surplus markets. This adds to rice price volatility in Myanmar. Table 29: Estimates of the adjustment parameters of Pawsan prices 2004-2013 2004-2008 2009-2013 Market pairs α1 α2 α1 α2 α1 α2 Yangon-Pathein 0.00 0.09* -0.03 0.10 -0.15* 0.14 Yangon-Mandalay -0.07 0.11 -0.05 0.16 0.00 0.16*** Yangon-Myingyan -0.09 0.12 -0.14* -0.01 -0.22* 0.26* Pathein-Mandalay -0.17 0.27** -0.11 0.57** 0.10 0.15*** Pathein-Myingyan -0.23*** 0.24** -0.33** 0.13 0.02 0.30*** Mandalay-Myingyan -0.28*** 0.13 -0.45*** -0.08 -0.30** 0.10 Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates using the MIS/MOAI prices (2014). Table 30: Estimates of the adjustment parameters of Manawthukha prices 2004-2013 2004-2008 2009-2013 Market pairs α1 α2 α1 α2 α1 α2 Yangon-Pathein -0.09 0.11 -0.24* 0.07 -0.17 0.13 Yangon-Mandalay -0.02 0.12 -0.15 0.14 -0.13 0.42** Yangon-Myingyan -0.15*** -0.07 -0.22** -0.09 -0.13 0.36** Pathein-Mandalay -0.07 0.39*** -0.21 0.56*** 0.18 0.15* Pathein-Myingyan -0.29*** -0.08 -0.37*** -0.19 0.04 0.34*** Mandalay-Myingyan -0.34*** -0.18* -0.35*** -0.29** -0.18 0.35** Note: Asterisks indicate significance levels: *** 0.001 (strongest), ** 0.01, * 0.05, and non-existent (weakest). Source: Authors’ estimates using the MIS/MOAI prices (2014). 60 THE WORLD BANK OFFICE, BANGKOK 30th Floor, Siam Tower, 989 Rama I Road, Pathumwan Bangkok 10330 Thailand Tel: 662 686-8300 Fax: 662 686-8301 Internet: www.worldbank.org/th E-mail: thailand@worldbank.org