ASA “Support to Agricultural Modernization in Uzbekistan” Farm Restructuring in Uzbekistan: How Did It Go and What is Next? January 15, 2019 Acknowledgements This Report is prepared as a part of the Analytical and Advisory Services supporting “Agricultural Modernization in Uzbekistan.” It was financed by the Multidonor Trust Fund, with the contributions from the European Union, Switzerland, the United States of America. The Report’s preparation was led by Sergiy Zorya (Task Team Leader, Lead Agriculture Economist, World Bank), with the major contributions from Nodir Djanibekov and Martin Petrick, both from Leibniz Institute of Agricultural Development in Transition Economies (IAMO), Halle (Saale), Germany. Other task team members included Shavkat Hasanov from Samarkand Agricultural Institute, Uzbekistan, Nozilakhon Mukhamedova from IAMO, and Olivier Durand and Dilshod Khidirov, both from the World Bank. The task team is grateful for the government and donor organizations, who provided comments and suggestions to improve the Report. The task team is also grateful to the World Bank staff, who provided their inputs on the final stages of the Report’s preparation. ii List of Acronyms AGRIWANET Database on agriculture in Central Asia prepared by IAMO GAO Gross agricultural output GDP Gross domestic product Ha Hectare IAMO Leibniz Institute of Agricultural Development in Transition Economies, Halle (Saale), Germany LSU Livestock unit MAWR Former Ministry of Agriculture and Water Resources UzGosKomStat State Committee of the Republic of Uzbekistan on Statistics UZS Uzbek Som WB World Bank WDI World Development Indicators iii Executive summary 1 In January 2019, Uzbekistan started a new farm restructuring1. It is said to seek to optimize the use of farmland by increasing the size of farms producing wheat and cotton, reallocating land to more efficient farmers and even clusters, and improving crop rotation options. 2 This is not the first time that this kind of farm restructuring in Uzbekistan takes place. The country has gone through several waves of farm restructuring and land reallocations. Both these processes were administratively managed, with little reference to market or income generation opportunities. During 1992-1997, state farms were decollectivized. During 1998-2002, farm fragmentation into small production units has started, which was completed during 2003-2008. Unsatisfied with the performance of fragmented farms, the government started farm consolidation between 2008/2009 and 2016, creating a dual system when dekhkan smallholders averaging 1 ha and producing livestock and horticulture products coexisted with large individual farms, averaging 40-60 ha and producing cotton and wheat under the state order system. The 2019 restructuring seeks to double the size of cotton and wheat farms to the average of 100 ha. 3 This report takes stock of the past farm restructurings and their impacts on agricultural production outcomes and productivity of farm land and labor use to predict the outcomes of the 2019 restructuring. In addition, it provides recommendations on reforms that should complement farm restructurings to make a significant impact on agricultural development. 4 The major findings from the past restructurings are the following: a. Past farm restructurings were an integral part of the strict policy to increase production of two strategic crops, cotton and wheat. Farmers, requested to produce these crops, have received land, intermediate inputs, capital, and market; yet, outputs were procured at prices, which could rarely cover production costs. In many cases farmers made losses, especially in the areas with poor soils and insufficient water. That is why it is difficult to decouple the effects of farm restructuring from other policies on agricultural outcomes. b. Agriculture grew at the relatively high rate, driving the average productivity up. Between 1996 and 2016, the average agricultural labor productivity annually grew by 1.5 percent. Yet, over time the agricultural productivity growth has slowed down despite the farm restructuring. The most important is, however, that the growth in agricultural labor productivity was a result of a large decline in the number of agricultural workers rather than a result of the large increase in agricultural value added. Share of agriculture in total labor force declined from 43 percent in 1996 to 30 percent in 2016. This growth has been called “jobless” as it has not generated jobs, neither in primary agriculture nor in food and textile industries, which depend on raw materials from the primary agriculture, where the employment declined in both absolute and relative terms. c. The administrative increase of size of farms producing cotton in 2008 has not helped generate economies of scale and increase its yields and quality. Production of cotton has decreased over time, because of the cotton land reallocation to other crops and flattening yields. Lower yields were a result of the cotton taxation, lack of good seeds, and introduction of wheat production quota, which replaced fodder in the cotton rotation system. Sowing wheat after cotton has reduced soil fertility (and thus cotton yield), which was regenerated in the past when cotton was followed by fodder as a part of soil management practices. d. Production of wheat by large farms has been rising, resulting from the increase in land allocation and yields. But in recent years, the growth in wheat production has lagged the growth in domestic wheat consumption, which is increasingly met by imports. In 2017, 1 The Resolution of the Cabinet of Ministers No. 14 from January 11, 2019. iv wheat import accounted for 43 percent of wheat production, a large increase compared to 25 percent in 2010. The evidence is piling up that larger farms are insufficient to satisfy the requirements of even local consumers in terms of quantity and quality of wheat. e. Production of horticulture and livestock products has almost fully shifted to small dekhkan farms. The government has neither intervened into production of these commodities nor provided any substantial public services. The production of these commodities increased, and dekhkan farms are more efficient than larger (individual) farms regarding the use of land and labor in Uzbekistan partially because they produce more profitable and less water- demanding products and partially because they take care of their limited production base. They truly ‘own’ their land and invest in it. Yet, very small size of these farms seems to prevent them to generate sufficient volumes of outputs to reduce transaction costs for marketing and improve quality and consistency of their outputs. Land allocation to state- mandated cotton and wheat continues to dominate the sown area, presenting a major challenge to dekhkan farms and production of horticulture and livestock products. Dekhkan farms producing high-value crops are constrained in their access to fertilizers, fuel, machinery, credit, value chains, and export channels. Solving their land issues will be only a part of the integrated solution. f. Frequent farm restructurings and the weak land tenure security rights have curtailed farm management and investment incentives and raised issues of “fair and just” access to farmland for rural population. Land leasing and subleasing is not permitted, so farmland cannot move freely from less efficient to more efficient producers. Lessons for the 2019 farm restructuring 5 The past farm restructurings do not seem to provide an optimistic outlook for the 2019 restructuring. The results of the upcoming restructuring will be that large wheat and cotton farms will become even larger and some land will be shifted from farmers, who were not able to meet the state order targets for whatever reasons, to farmers, who were able to meet targets. There is no guarantee that the shifted land will be used more efficiently once ‘new’ land users realize that they receive low-fertile land with poor irrigation, for example. Some of the vacated land will be transferred to cotton-textile clusters, which also does not guarantee its most efficient use. But what this restructuring will surely do is to restrain farmers from making significant soil improvements and investments in innovative technologies (see paragraph 4f). Why to invest in assets, which could be expropriated by the state at any time without compensation? In addition, without making it profitable for farmers to produce cotton and wheat and without increasing public spending on agricultural programs for generating technologies, disseminating knowledge and building more human capital, reducing transaction costs, and crowding in private capital, the upcoming farm restructuring will not change the current performance of the sector. 6 Like other Central Asian governments, Uzbekistan has struggled to find a post-socialist model for its farming sector. The debates still focus on the desirable farm size where the Soviet legacy of industrialized collective farms co-existing with private household plots marked the extremes. The global experience, on the other hand, stresses the need for flexibility in farming structures. Very small plots operated by dekhkan farms in Uzbekistan are suboptimal, but once farm size reaches 30-40 ha for wheat and cotton, increasing it further does not bring much economic gain. There is generally no evidence of economy of scale in primary agriculture, and any gains from economy of scale are easily negated by extra monitoring costs of wage labor and farm managers. These factors highlight the importance of individual incentives for farm efficiency and account for the predominance of family farms in a market economy. 7 Optimal farm size globally is determined in each case by the managerial capacity of the farmer and many exogenous factors. Farm size is often growing in the countries where economic growth and rising off-farm wages lead to the outflow of agricultural labor. Thus, rather than v targeting at a type or size of farm organization, the global experience shows that it is better to focus on helping farmers: (i) improve their managerial capacity; (ii) react to market signals by removing agricultural distortions; (iii) lease or buy more land when profitable; and (iv) access a set of supporting agricultural services. Recommendations 8 The future reform agenda would need to focus on: (i) removal of the production distortions; (ii) increase of the impact of agricultural public expenditures on the sector performance; and (iii) help of smallholders to reduce transaction costs. The specific actions can be the following: Production distortions • Phase out the cotton and wheat production quotas, inside and outside of the clusters. • Align farm-gate prices for cotton and wheat with market prices. • Avoid the introduction of new mandates (such as on high value crops or diversification) that impose yet other production constraints on farmers. • Liberalize land market to allow farm restructuring to evolve responding to market signals. This does not necessarily require full privatization of farmland in the short run, but the legalization of land rentals and sub-rentals, revoking the threat of public land seizures, and formalization of labor contracts. It is important to ensure that transaction costs for moving land from less to more efficient farmers are minimized, irrespective of size or type of farms, and leasing is market oriented. Efficiency of agricultural public expenditures • Use public expenditures more effectively to help farmers increase returns to land and labor. It would require redirection of farm subsidies towards public good provision and empowering national and local governments to allow more effective service delivery such as land registries or advisory services. • Major public programs, for which the public expenditures need to be increased, include, among others: applied agricultural research and development, extension, education, soil fertility, food safety, animal disease control, veterinary services, phytosanitary services, support to smallholders (cooperatives, clusters, productive partnerships), market and statistical information, water-saving technologies, market infrastructure and logistics, environment protection, policy analysis, and sector coordination. • Ensure that an adequate amount of public expenditures is allocated to the above- mentioned programs to make a difference. Many of them were underfunded in the past. Programs for dekhkan farms • Introduce public programs targeting dekhkan farms, especially those who want to expand and grow. These programs can facilitate establishment of productive partnerships (cooperatives) among farmers, and between farm groups and agribusiness, to improve quality and safety of products and connect these farms to modern value chains. Such programs are common in many countries with small farms. The role of government would be to finance: (i) establishment of farm groups; (ii) improvement of their farming and managerial capacity; (iii) activities of anchor agribusinesses to develop vertical networks; and (iv) business arrangements to provide veterinary and advisory services to small farmers on behalf of the government. 9 How a future and still growing rural population will be employed and secure its income is one of the essential questions the government seeks to address. It would require actions beyond the farm restructuring. Farm restructuring is only a part of the puzzle to be brought together into one coherent piece, not a silver bullet as often perceived. vi Contents Acknowledgements................................................................................................................................. ii List of Acronyms ..................................................................................................................................... iii Executive summary ................................................................................................................................ iv List of Tables ........................................................................................................................................ viii List of Figures ......................................................................................................................................... ix I. Introduction .................................................................................................................................... 1 II. Major Stages of Farm Restructuring since Independence .............................................................. 1 Key agricultural indicators by region .................................................................................................. 1 Farm consolidation (2008-2015) ......................................................................................................... 5 Adjustment of cropping structure (since 2016) .................................................................................. 6 III. Restructuring Outcomes to Date .............................................................................................. 10 Number and structure of individual farms in 2016 .......................................................................... 10 IV. Evolution of Agricultural Output ............................................................................................... 15 Crop production ................................................................................................................................ 16 Livestock production ......................................................................................................................... 23 Fodder production ............................................................................................................................ 28 Productivity measures ...................................................................................................................... 30 V. Conclusions and Recommendations ............................................................................................. 38 Farm restructuring outcomes ........................................................................................................... 38 Remaining areas for improvements ................................................................................................. 39 Recommendations on the future of farming in Uzbekistan ............................................................. 39 References ............................................................................................................................................ 41 Appendix: Policies events related to restructuring of individual farms in 2006-2017 ......................... 43 vii List of Tables Table 1: Selected indicators of agriculture by provinces in 2016 ........................................................... 2 Table 2: Salient characteristics of the farm restructuring stages in Uzbekistan since 1991 .................. 3 Table 3: Definition individual and dekhkan farms in Uzbekistan............................................................ 4 Table 4: Projected parameters of crop production in Uzbekistan in 2016-2020 ................................... 7 Table 5: Projected changes in sown area by province, 2015-2020, ‘000 ha........................................... 8 Table 6: Projected livestock production parameters for 2016-2020 ...................................................... 8 Table 7: Projected changes in livestock numbers by province, 2015-2020, ‘000 heads ........................ 9 Table 8: Number and average size of individual farms by province in 2016 ........................................ 10 Table 9: Share of different agricultural producer categories in sown area in 2016 ............................. 11 Table 10: GAO by producer category and province in 2016, billion UZS .............................................. 16 Table 11: Crop output by producer category and province in 2016, billion UZS .................................. 17 Table 12: Total sown area of cotton by province, ‘000 ha.................................................................... 18 Table 13: Cotton area and its share in sown area by producer categories and province in 2016 ....... 19 Table 14: Total sown area of wheat by province, ‘000 ha .................................................................... 20 Table 15: Wheat area and its share in total sown area by producer category and province in 2016 .. 21 Table 16: Total sown area of vegetables by province, ‘000 ha ............................................................. 22 Table 17: Vegetable area and its share in sown area by producer category and province in 2016 ..... 23 Table 18: Livestock output by producer category and province in 2016, billion UZS .......................... 24 Table 19: Meat production by producer category and province in 2016 ............................................. 25 Table 20: Milk production by producer category and province in 2016 .............................................. 26 Table 21: Egg production by producer category and province in 2016 ................................................ 27 Table 22: Fodder area and its share in sown area by producer category and province in 2016 .......... 29 Table 23: Fodder area per LSU by producer category and province in 2016, ha/LSU .......................... 30 Table 24: Changes in labor productivity in selected sectors, Uzbekistan, 1996-2016 ......................... 31 Table 25: Changes in employment and labor productivity of subsectors of Uzbekistan’s economy, 1996-2016 ............................................................................................................................................. 32 Table 26: Sown land productivity by producer category and province in 2016, million UZS/ha ......... 35 Table 27: Livestock output per LSU by producer category and province in 2016, million UZS/LSU .... 37 viii List of Figures Figure 1: Changes in sown area by producer category, ‘000 ha ............................................................. 5 Figure 2: Evolution of average sown area and the number of individual farms in Uzbekistan .............. 6 Figure 3: Relationship between sown area per individual farm and sown area per rural inhabitant .. 12 Figure 4: Share of sown area under cotton and wheat in individual farms, percent ........................... 12 Figure 5: Structure of individual farms according to specialization, percent of all individual farms ... 13 Figure 6: Land allocated to individual farms according to production specialization, ‘000 ha ............ 14 Figure 7: Average size of individual farms according to production specialization, ha ........................ 14 Figure 8: Evolution of GAO by producer category and sector, trillion constant 2016 UZS .................. 15 Figure 9: Evolution of gross crop output by producer category, trillion constant 2016 UZS ............... 17 Figure 10: Cotton sown area by producer category, ‘000 ha ............................................................... 18 Figure 11: Wheat sown area by producer category, ‘000 ha................................................................ 19 Figure 12: Vegetable sown area by producer category, ‘000 ha .......................................................... 21 Figure 13: Evolution of livestock output by producer category, ‘000 billion constant 2016 UZS ........ 23 Figure 14: Meat production (in light weight) by producer category, ‘000 tons ................................... 25 Figure 15: Milk production by producer category, ‘000 tons ............................................................... 26 Figure 16: Egg production by producer category, million eggs ............................................................ 27 Figure 17: Sown area of fodder crops by producer category, ‘000 ha ................................................. 28 Figure 18: Average sown area of fodder crops per LSU by producer category, ha / LSU ..................... 29 Figure 19: Agricultural output, sown area and employment in 1992-2016, 1992=100 ....................... 30 Figure 20: Agricultural output per labor, land and livestock unit, 1992=100 ....................................... 31 Figure 21: National average yields of cotton by producer category, tons/ha ...................................... 32 Figure 22: Wheat yields by producer category, tons/ha ...................................................................... 33 Figure 23: Vegetable yields by producer category, tons/ha ................................................................. 33 Figure 24: Sown land productivity by producer category, million constant 2016 UZS/ha ................... 34 Figure 25: Livestock output per LSU by producer category, million constant 2016 UZS/LSU .............. 35 Figure 26: National average milk yields by producer category, tons/cow ........................................... 36 Figure 27: National average meat production (in live weight) by producer category, kg/cattle ......... 36 ix I. Introduction 1 Since independence, Uzbekistan’s unique agricultural reform pattern involved multiple changes in farm sizes and specialization, all mandated by the government. Public ownership of land provided the legal basis for these unique restructuring experiments. Such readjustments in farm fragmentation, re-consolidation, and optimization affected the picture of an average farm and make farmers a moving-target for policy advice. 2 The aim of this report is to describe major trends of farm restructuring in Uzbekistan since independence. Special attention is paid to the evolution of individual farms since 2006, with a focus on the changes in number and average size of cotton-grain growing farms across provinces of Uzbekistan.2 3 In the first section, the major stages of farm restructuring are presented along with the overview statistics on farm numbers and size. The second section takes stock of restructuring outcomes to date by reporting detailed figures on the current structure of individual farming by region. The third section portrays major trends in land use by crop, production outcomes and productivity measures. The final section concludes with an overall assessment of the restructuring policy and implications for the 2019 restructuring. 4 The report uses official agricultural statistical data, which were collected from the State Committee of the Republic of Uzbekistan on Statistics (UzGosKomStat) and the former Ministry of Agriculture and Water Resources (MAWR). The statistical data are complemented with expert interviews and the review of local publications and national legislation. II. Major Stages of Farm Restructuring since Independence Key agricultural indicators by region 5 In 2016, rural population accounted for almost half of population in Uzbekistan with the highest share in Khorezm (68 percent), Surkhadarya (64 percent), Bukhara and Samarkand (each about 62 percent) provinces. The lowest were in Tashkent (28 percent) and Namangan (35 percent) provinces (Table 1). 6 According to official statistics, in 2016 agriculture contributed about 24 percent of gross domestic product (GDP). The lowest contribution of agriculture to domestic product was in Karakalpakstan and Navoi province. Agriculture made up almost a half of GDP in Andijan and Jizzakh provinces. 7 Cotton and wheat dominate the sown area. In 2016, almost three-fourth of sown area was allocated to these crops. In Syrdarya, Bukhara, Navoi and Surkhandarya provinces, cotton and wheat occupied over 80 percent of total sown area. Cotton and wheat were less spread in Karakalpakstan and Jizzakh, Tashkent and Khorezm provinces, occupying less than two-thirds of sown land. 8 Sown area is in relative scarcity in Uzbekistan: on average there were about 0.23 ha of sown area per rural inhabitant. The provinces with the highest density of rural population are Andijan, Samarkand, Surkhadarya and Fergana, while the lowest densities are observed in Syrdarya and Jizzakh provinces. 9 After the dissolution of the Soviet Union, the farm types in Uzbekistan were changed several times (Lerman 2008 and Djanibekov et al. 2012). 2It should be noted that in this report the terms such as ‘individual farm’ with ‘farm’, ‘dekhkan farm’ with ‘rural household’ or ‘household’, and ‘agricultural enterprise’ with ‘enterprise’ are used interchangeably. 1 10 Table 2 summarizes the five critical stages in the farm restructuring process. Each stage differs by the speed and level of regulations accompanying the transfer of state and collective property to private hands and distinguished by policy objectives. Table 1: Selected indicators of agriculture by provinces in 2016 Province Total sown Total Share of Share of Share of Sown area area, ‘000 population, rural cotton & gross agri per rural ha ‘000 population, wheat in sown output in inhabitant, persons % area, % GDP, % ha/cap Karakalpakstan 254.9 1,818 50.8 62.3 20.5 0.28 Andijan 230.1 2,963 47.6 76.6 46.5 0.16 Bukhara 240.6 1,844 62.5 83.0 40.3 0.21 Jizzakh 401.1 1,301 53.0 61.8 47.7 0.58 Kashkadarya 495.5 3,089 57.1 78.1 25.9 0.28 Navoi 103.5 943 51.1 82.6 22.1 0.21 Namangan 225.0 2,652 35.3 75.8 39.6 0.24 Samarkand 359.0 3,652 62.4 73.0 40.5 0.16 Surkhandarya 285.0 2,462 64.4 82.3 44.6 0.18 Syrdarya 232.8 803 56.9 85.3 43.3 0.51 Tashkent 353.5 5,253 27.6 61.7 29.0 0.24 Fergana 288.9 3,565 43.3 75.7 28.4 0.19 Khorezm 236.7 1,777 67.8 65.0 41.6 0.20 Uzbekistan 3,706.5 32,121 49.4 73.1 24.0 0.23 Note: Total population in Tashkent province includes also population in Tashkent city. Source: IAMO3 and the World Bank (WB), based on official statistical data from UzGosKomStat (2017). 3 IAMO is Leibniz Institute of Agricultural Development in Transition Economies from Halle (Saale), Germany. 2 Table 2: Salient characteristics of the farm restructuring stages in Uzbekistan since 1991 First stage Second stage Third stage Fourth stage Fifth stage 1992-1997 1998-2002 2003-2008 2008/09-2015 2016-present Decollectivization Partial Complete Farm Production of state farms fragmentation fragmentation consolidation specialization Main Transformation of Transformation Complete Farm Fragmentation transformation sovkhozes into of kolkhozes into transformation reconsolidation and process kolkhozes shirkats. of shirkats into (farm-size optimization Land lease to individual farms optimization) of production individual farms Dominant farm Kolkhozes, Shirkats, Shirkats, Individual Individual types sovkhozes individual farms individual farms farms, mainly farms of cotton-grain different producers specialization Main policy Expansion of Specialization of Development of Increased and Relocation of objectives wheat area & newly non- stable cotton cotton and yields, established cotton/wheat yields, wheat fields, reorganization of individual farms producing relocation of increased area state farms sectors, and cotton fields of high value livestock farms crops, multi- profile farms Source: Updated from Djanibekov (2016). Main agricultural producers 11 As will be shown in this report, across most of food products agriculture in Uzbekistan is now structured around two main producer categories, namely individual farms and dekhkan farms. Table 3 presents their main characteristics. The third category of producers is agricultural enterprises. Over the progress of farm restructuring, this category has evolved from Soviet-type of collective and state farms, to agricultural cooperatives (shirtkats) and further restructured to specialized enterprises of different forms including parastatal export-oriented producers or large poultry farms which are registered as an enterprise. These agricultural enterprises also include ‘agri-firms’ established since 2006 particularly in the horticultural sector (Larson et al. 2012). These are non- government associations and private firms with vertically-integrated organizational form as well as processing companies that also operate farm land. 3 Table 3: Definition individual and dekhkan farms in Uzbekistan Individual farm Dekhkan farm (fermer hojaligi) (dekhkan hojaligi) Basic definition Individual commercial farm organized as a Small-scale family-based farm, based legal entity operating leased land on household plot operation Utilized labor Family members, as well as permanent and Mainly family members, with option to seasonal workers hire seasonal workers Land tenure Long-term land lease (up to 50 years). The Lifetime inheritable possession. Sizes land lease duration depends on the of allocated land: 0.35 ha for irrigated fulfillment of state procurement target. land; 0.5 for rainfed land. This includes Farm size can vary with respect to also area for buildings production specialization Ownership Any adult person with sufficient agricultural Former workers of agricultural qualification enterprises, rural families Production Only agricultural produce indicated in land Any agricultural produce, mainly specialization lease contract. Mainly cotton and wheat wheat, vegetables, fruits, livestock Source: Updated from Lerman (2008). Decollectivization and fragmentation (1992-2007) 12 The first two stages (1992-2002) aimed at the development of supply and marketing infrastructure for agricultural production and organization of other types of agricultural enterprises such as corporate farms (named locally as shirkats), established on the basis of former state and collective farms. This included such measures as debt write-offs, tax exemptions and financial assistance to agricultural enterprises, and to agro industrial sector in general. This stage included first transfer of land of large inefficient farms for establishing of smaller individual farms through land lease and bidding on official tenders. Although production activities in large agricultural enterprises remained linked to the state production targets, one important change, however, relates to property rights over agricultural output. Production obligations for agricultural products were removed for all crops except strategic cotton and wheat (Pomfret 2000). 13 In 1998, the government announced the official definition and coexistence of three categories of producers: dekhkan hojalik (based on agricultural production on household plots), shirkats (agricultural enterprises) and fermer hojalik (individual farms), being distinct in land size and tenure, labor contracts and mode of production specialization. Collective farms were turned into shirkats through issuing and distributing the membership shares with private ownership of assets among its workers and allocating fields to their shareholders for agricultural production contracts. The definition of individual farms changed fundamentally providing them sovereignty from shirkats in their production decisions, both in terms of input and output allocation (Bloch 2002). 14 During the first two stages of farm restructuring (1992-2002), the speed of transfer of land from agricultural enterprises to newly-established individual farms was slow (Figure 1). In 2006, the government speeded up the fragmentation process and the re-organization of agricultural production in the form of smaller individual farms (Figure 1). The list of main policy events to the farm restructuring process in 2006-2017 is presented in the Appendix. The main result of the fragmentation was the establishment of a dual system encompassing a symbiotic relationship between individual farms and dekhkan farms (Djanibekov 2012). In 2014, there were about 4.7 million dekhkan farms that operated 20 percent of total sown land. Dekhkan farms were never part of the farm restructuring process and their land share increased along the population growth. Institutional transformations also affected the employment structure. Most of the rural population released from their work in agricultural enterprises became dehkans involved in self-sufficiency and small-scale production, while only 10 percent of households could obtain private farm land (Pugach et al. 2016). Observations by Swinkels et al. (2016) on rural income from participating in cotton 4 picking and Djanibekov et al. (2013, 2015) on labor arrangements between individual farms and households show that dekhkan farms in Uzbekistan are not necessarily subsistence oriented. They participate in market transactions and are integrated into rural labor market via on-farm employment. Figure 1: Changes in sown area by producer category, ‘000 ha 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Individual farms Agricultural enterprises Rural households Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 15 The average sown area in individual farms did not change much in 1998-2006 staying around 15 ha (Figure 2). In the last two years of the third stage of farm restructuring (2007-2008), the expansion of farmland in individual use slowed down because most shirkat land had already been redistributed. A (new) farmer could acquire additional farmland (i.e. lease new farmland) if the state allotted land that had been returned by another farmer. In 2007, the number of individual farms increased by 28,000 farms compared to 64,000 farms in 2006. By the end of the third stage (2008), the number of individual farms reached 218,600 farms with average sown area of 14 ha. Farm consolidation (2008-2015) 16 Until 2008/2009, farm restructuring did not take into account the general idea of a farm as one production unit in the sense that farm fields are located next to each other so as to form a single territory of a farm unit. Instead, an individual farm comprised fields scattered and located often at distances from each other. The demands in inputs and services of rapidly expanding individual farms soon created a number of difficulties. Among them is a mismatch of irrigation water supply infrastructure designed for large-scale farms (Djanibekov et al. 2015). A redesign of the existing irrigation system in a way to suit the many smaller farms would be costly. These along with other issues triggered decisions of the state to reduce the number of farms and increase (‘optimize’) size of their land. With the declared aim of a ‘farm-size optimization’ namely farm consolidation via merging smaller farms into larger units, the fourth stage took off in late 2008/2009 and continued till 20154 (Figure 2). Officially, farm-size optimization was reasoned by low productivity of individual farms, especially in fulfilling cotton and wheat state necessities, and the consolidation of scattered farm fields into larger production units was seen as a remedy. 4 2008 and 2010 Presidential Decree “On optimization of sown area and increase of food crop production”. 5 Figure 2: Evolution of average sown area and the number of individual farms in Uzbekistan 250 3,500 and average sown area in individual farm, ha Sown area in individual farms, 1000 ha Number of indivitual farms, 1000 ha 200 2,800 150 2,100 100 1,400 50 700 0 0 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Number of individual farms, 1000 Average sown area of individual farm, ha Total sown area in individual farms, 1000 ha Note: Average size of individual farm is given in average sown area per farm. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 17 During the first wave of farm consolidation, the number of individual farms decreased from about 218,600 in 2008 to around 103,000 in 2009. In the second consolidation wave of 2010, the number of farms was reduced further (by about 30 percent), reaching about 69,100 farms in 2011. The average sown area per individual farm increased from 13.9 ha in 2008 to 29.6 ha in 2009 and over 44 ha in 2011 (Figure 2, left Y-axe). New norms for farm restructuring came into play with the third wave of farm optimization in 2012: (i) decrease of the land size or farm liquidation became possible through voluntary or local authority initiation; and (ii) prohibition of land division was lifted conditional to condition of land size not being less than the state-set minimum according to farm specialization. Adjustment of cropping structure (since 2016) 18 The current fifth stage of farm restructuring is a five-year plan during 2016-2020 that includes a complex and province specific approach to switching optimization from farm size to production structure.5 New measures project a gradual reduction of the total area sown under cotton by 170,500 ha (13 percent less compared with 2015) and 50,000 ha for grains (3.4 percent less compared with 2015) on irrigated areas (Table 4). Although cotton and wheat remain strategic crops, the projected measures show attempts to expand and diversify agricultural production. 19 Land freed up from cotton and wheat is planned to be allocated for potato (increase by 36,000 ha), vegetables (increase of 91,000 ha), fodder crops (increase by 50,300 ha), intensive gardens, oilseeds, and other crops. By 2020, the plan is to increase vegetables and potato production area by 40 percent, while the area under oilseeds is projected to be doubled. 5 Presidential Decree No. PP-2460 “On further measures of reforms and development of agriculture for 2016-2020” (29.12.2015). 6 Table 4: Projected parameters of crop production in Uzbekistan in 2016-2020 Crop Sown area, ‘000 ha Difference in 2020 to 2015 2016 2017 2018 2019 2020 2015, % Cotton 1,285.5 1,255.0 1,221.0 1,187.0 1,151.0 1,115.0 -13.3 Wheat 1,329.5 1,329.5 1,319.5 1,304.5 1,289.5 1,279.5 -3.8 Potato 80.3 85.3 92.5 100.5 108.8 116.3 44.8 Vegetables 192.0 204.6 222.7 243.0 264.0 283.0 47.4 Fruits & melons 261.9 264.6 268.0 272.0 276.1 279.9 6.9 Fodder 309.1 316.1 326.1 337.3 348.9 359.4 16.3 Oilseeds 14.7 16.2 19.0 22.1 25.4 28.3 92.5 Grapes 144.0 145.5 147.8 150.3 152.9 155.2 7.8 Yield, tons/ha Difference in 2020 to 2015 2016 2017 2018 2019 2020 2015, % Cotton 2.61 2.62 2.63 2.65 2.67 2.69 3.1 Wheat 5.49 6.09 6.25 6.36 6.51 6.64 20.9 Potato 21.89 22.43 22.6 22.75 22.9 23.05 5.3 Vegetables 27.71 28.61 28.85 29 29.2 29.4 6.1 Fruits & melons 12.39 12.64 12.94 13.29 13.66 14.04 13.3 Fodder 22.5 22.8 23.1 23.4 23.7 24.0 6.7 Oilseeds 1.7 1.8 1.9 2 2.1 2.2 29.4 Grapes 12.67 12.79 12.99 13.21 13.46 13.71 8.2 Total harvest, ‘000 tons Difference 2015 2016 2017 2018 2019 2020 in 2020 to 2015, % Cotton 3,350.0 3,287.0 3,217.0 3,147.0 3,074.0 3,000.0 -10.4 Wheat 7,305.0 8,100.0 8,250.0 8,300.0 8,400.0 8,500.0 16.4 Potato 2,670.0 2,833.0 3,010.0 3,206.0 3,411.0 3,601.0 34.9 Vegetables 9,923.0 10,458.0 11,031.0 11,651.0 12,314.0 12,925.0 30.3 Fruits & melons 2,731.0 2,874.0 2,982.0 3,109.0 3,244.0 3,380.0 23.8 Fodder 18,725.0 18,976.0 19,303.0 19,662.0 20,039.0 20,386.0 8.9 Oilseeds 98.0 110.0 125.0 132.0 145.0 160.0 63.3 Grapes 1,556.0 1,601.0 1,651.0 1,707.0 1,769.0 1,830.0 17.6 Note: Harvest of potato, vegetables, fodder and oilseeds include second harvest and raid-fed production. Source: Presidential Decree PP-2460 (29.12.2015). 20 The projected decrease in cotton and wheat area is planned to be compensated by respective increase in their yields. As such it is projected that by 2020 the increase in cotton yield should be by about 3 percent, while wheat yield should increase by 20 percent. The latter should offset the projected decrease in wheat area and bring additional 16 percent of output. 21 By 2020, the largest relative decrease in cotton area, by 22 percent, is planned in two labor- scarce regions, namely Syrdarya and Jizzakh. These provinces will also have the highest expansion of vegetable area (Table 5). Wheat area is planned to be decreased by 3-6 percent across provinces. No reduction in wheat area is foreseen in Karakalpakstan, and Navoi and Khorezm provinces. 7 Table 5: Projected changes in sown area by province, 2015-2020, ‘000 ha Province Cotton Wheat Vegetables in Increas diff., in 2015, Increas diff., in 2015, Increas diff., 2015, e by % ‘000 ha e by % ‘000 ha e by % ‘000 2020, 2020, 2020, ha ‘000 ha ‘000 ha ‘000 ha Karakalpakstan 95.9 -7.0 -7.3 64.4 0.0 0.0 10.5 2.2 21.0 Andijan 93.4 -11.2 -12.0 84.9 -4.0 -4.7 18.7 6.4 34.2 Bukhara 110.3 -10.0 -9.1 91.7 -5.0 -5.5 8.6 7.1 82.6 Jizzakh 101.8 -22.5 -22.1 148.1 -5.0 -3.4 8.3 11.0 132.5 Kashkadarya 164.7 -18.4 -11.2 226.3 -4.0 -1.8 16.5 9.7 58.8 Navoi 35.8 -3.4 -9.5 50.2 0.0 0.0 3.8 1.4 36.8 Namangan 82.6 -10.1 -12.2 89.7 -5.0 -5.6 14.4 6.8 47.2 Samarkand 93.0 -16.2 -17.4 172.2 -6.0 -3.5 27.9 8.3 29.7 Surkhandarya 119.6 -14.3 -12.0 117.5 -5.0 -4.3 13.3 8.7 65.4 Syrdarya 106.8 -22.7 -21.2 97.1 -5.0 -5.1 4.4 10.1 229.5 Tashkent 94.4 -13.4 -14.2 129.0 -6.0 -4.7 33.3 7.5 22.5 Fergana 100.1 -12.5 -12.5 120.7 -5.0 -4.1 19.4 7.0 36.1 Khorezm 99.7 -8.8 -8.8 54.4 0.0 0.0 14.9 4.8 32.2 Uzbekistan 1,298 -171 -13.1 1,446 -50 -3.5 194 91 46.9 Source: Presidential Decree No. PP-2460 (29.12.2015). 22 The projected parameters also target the increased number of livestock and production output, and the number of livestock farms (Table 6). The number of cattle and sheep-goats is planned to be increased annually by 5 percent, reaching 20 percent increase by 2020. By 2020, the number of poultry is projected to grow by 50 percent, while production of eggs by 75 percent. Table 6: Projected livestock production parameters for 2016-2020 Number, ‘000 heads Difference in 2020 to 2015 2016 2017 2018 2019 2020 2015, % Cattle 11,637.2 12,150 12,720 13,350 14,050 14,800 27.2 Sheep & goats 19,096.1 19,600 20,380 21,240 22,170 23,187 21.4 Poultry 61,070.5 64,600 69,500 75,500 83,000 92,000 50.6 Production, ‘000 tons Difference in 2020 to 2015 2016 2017 2018 2019 2020 2015, % Meat (live weight) 2,033 2,060 2,150 2,260 2,375 2,500 23.0 Milk 9,028 9,478 10,242 11,075 11,957 13,000 44.0 Eggs (million) 5,535 6,200 6,900 7,700 8,600 9,600 73.4 Note: The numbers used for 2015 are for 01.01.2016, while PP-2460 uses the numbers as of December 2015. Source: Presidential Decree No. PP-2460 (29.12.2015). 23 Among the provinces, Karakalpakstan and Syrdarya province are projected to have the highest increase in the number of cattle heads by 2020, 42 percent and 40 percent, respectively. In terms of poultry heads, it is planned that their number will increase by three-fourth in Karakalpakstan and Namangan province, while the smallest increase is projected for Surkhandarya and Tashkent provinces. The projected relative increase in number of sheep and goats across provinces ranges from 16 percent (Bukhara province) to 32 percent (Tashkent province). 8 Table 7: Projected changes in livestock numbers by province, 2015-2020, ‘000 heads Province Cattle Sheep & goats Poultry in 2015, Increase diff, in 2015, Increase diff., in 2015, Increase diff., ‘000 by % ‘000 by % ‘000 by 2020, % 2020, 2020, ‘000 ‘000 ‘000 Karakalpakstan 955.0 396.0 41.5 901.5 225.5 25.0 3,103.6 2,296.4 74.0 Andijan 980.0 191.0 19.5 1,224.0 241.0 19.7 6,400.0 3,310.0 51.7 Bukhara 1,133.2 273.8 24.2 1,936.7 318.3 16.4 3,218.9 1,481.1 46.0 Jizzakh 825.1 257.9 31.3 1,727.4 324.6 18.8 2,395.2 1,489.8 62.2 Kashkadarya 1,405.1 366.9 26.1 4,268.2 973.8 22.8 4,403.8 2,006.2 45.6 Navoi 415.0 121.0 29.2 1,859.6 315.4 17.0 1,971.2 818.8 41.5 Namangan 640.0 230.0 35.9 696.0 179.0 25.7 3,417.2 2,542.8 74.4 Samarkand 1,425.2 234.8 16.5 2,150.7 439.3 20.4 8,603.8 3,576.2 41.6 Surkhandarya 850.1 214.9 25.3 2,024.1 477.9 23.6 3,351.7 1,258.3 37.5 Syrdarya 399.8 160.2 40.1 278.0 67.0 24.1 1,738.9 1,141.1 65.6 Tashkent 815.0 235.0 28.8 828.6 266.4 32.2 13,645.0 5,260.0 38.5 Fergana 969.4 225.6 23.3 811.5 162.5 20.0 4,000.4 2,649.6 66.2 Khorezm 824.3 256.7 31.1 389.8 100.2 25.7 4,820.8 3,099.2 64.3 Uzbekistan 11,637 3,164 27.2 19,096 4,091 21.4 61,071 30,930 50.6 Note: The numbers used for 2015 are for January 2016, while PP-2460 uses the numbers as of December 2015. Source: Presidential Decree No. PP-2460 and official data from GosKomStat (2017). 24 In addition to the PP-2460 (29.12.2015), in the beginning of 2017 the Government adopted the Strategy of Actions for the Development of Uzbekistan for 2017-2021. The Strategy of Actions outlines its political, economic, and social priorities, including measures to liberalize the economy. The National Development Strategy identifies the need for diversification out of cotton into high value-added and labor-intensive production and processing, which is expected to contribute to significant growth of rural jobs, food security and exports. Section 3.3 of the strategy on Modernization and intensive development of agriculture accents on further optimization of the sown areas, aimed at reducing the acreage of cotton and cereal crops, area expansion for potato, vegetables, forage and oilseeds, creating new intensive gardens and vineyards. In the same section the strategy sets priority of developing multi-profile farms, which could be engaged in agricultural production and processing, preparation, storage, marketing, construction and provision of services. 25 Reforms also took the direction of strengthening the value chains and giving the support for the farmers to enable their direct exports without engaging the government monopoly Uzagroeksport. Another push for facilitating agricultural exports was the reform of currency market coupled with the abolishment of the requirement that exporters surrender part of their hard- currency earnings to the state. 26 Despite these changes in farm specialization, the procurement policy in cotton and wheat directly influences the availability of inputs to high value crops. The distribution of main agricultural inputs such as fertilizers, diesel, and machinery and irrigation services are still coordinated via governmental organization. These inputs are channeled to farmers through parastatal agencies at fixed prices. Farmers are linked to certain outlets for purchasing seeds, fertilizers, fuel and machinery services and cannot access better or cheaper alternatives. The priority in input allocation is given to cotton and wheat producers. 27 The input quantities that farmer can access for each production season depends on the scientific norms of input application and planned sown area (Djanibekov et al. 2012). In calculations 9 of total fertilizer and fuel needed by a farmer for certain season only the input requirements for cotton and wheat cultivation are considered. In such prescriptions of fertilizer and diesel amounts, farmers are considered a homogenous group with only difference in production specialization, namely whether farmer cultivates cotton and wheat or not. The specific input and service needs of commercially oriented farmers, particularly of small ones and those engaged in horticulture, are neglected. Inputs can also be obtained via Birja (commodity exchange), but transaction costs can be too high for farmers. For small farms and dekhkan farms lack access to chemical fertilizers and often make use of large amounts of organic fertilizers, supplemented by smaller doze of chemical fertilizers for commercial crops. 28 Farmers responded to this by inventing an informal (black) input markets for purchasing inputs for high value crops. Farmers also divert these inputs from cotton and wheat to commercial crops or sell directly to other agricultural producers (Larson et al. 2012). Aware of this, to ensure the fulfilment of state procurement targets, the local administration monitors and controls land allocation and input application in farms by initiating frequent inspections by different state organizations (Hornidge and Shtaltovna 2014). As a result, cotton-grain producing farms often lack flexibility to make better decisions about their production and allocation of available inputs. The commercial farms involved in production of fruits and vegetables lack access to inputs and services. Their needs are neglected in years of drought and irrigation water scarcity when all efforts are mobilized to safe cotton and wheat harvest. III. Restructuring Outcomes to Date Number and structure of individual farms in 2016 29 With the increasing number of individual farms during the last five years, in 2016 there were over 132,000 registered individual farms in Uzbekistan (Table 8), with the most farms being registered in Samarkand (over 19,000 farms) and the fewest in Navoi (about 3,400 farms). Table 8: Number and average size of individual farms by province in 2016 Provinces Number Agricultural output Average area of Average Sown area in of farms per individual farm allocated land sown area total allocated per farm per farm farm land million UZS /farm ha /farm ha /farm percent Karakalpakstan 5,482 83.7 93.5 39.6 42.3 Andijan 12,293 147.7 20.1 16.0 79.6 Bukhara 7,234 172.2 109.9 27.6 25.1 Jizzakh 10,304 77.5 46.7 35.5 76.0 Kashkadarya 16,629 86.2 43.6 24.7 56.7 Navoi 3,427 168.3 62.3 23.2 37.3 Namangan 8,094 129.9 30.6 23.8 77.7 Samarkand 19,789 114.9 24.8 15.0 60.5 Surkhandarya 8,833 116.4 82.8 26.7 32.3 Syrdarya 6,116 111.3 40.3 34.5 85.6 Tashkent 14,179 144.1 33.3 20.9 63.0 Fergana 11,366 114.9 29.9 21.3 71.4 Khorezm 8,610 108.6 29.6 22.3 75.2 Uzbekistan 132,356 118.2 49.8 23.7 47.6 Source: IAMO and the WB based on UzGosKomStat (2017). 10 30 According to the official statistics, average agricultural output per individual farm in 2016 was about 118 million Uzbekistan Soms (UZS). The highest agricultural output per individual farm was observed in Bukhara province (over 172 million UZS per farm). 31 Average area of land allocated per individual farm in 2016 was about 50 ha, ranging from 110 ha in Bukhara province (the largest average size of farms) to 20 ha in Andijan province (the smallest average size of farms). Average sown area in individual farms was about 24 ha, implying that about a half of total farm land is sown, while another half was under perennial crops, gardens, vineyards, fishponds, pastures, grassland or constructions and unproductive land. 32 In 2016, individual farms operated 84 percent of sown area (Table 9), with the largest share observed in Jizzakh and Syrdarya provinces (over 90 percent of sown area). Rural households (dekhkan farms) operated 13 percent of sown area. Table 9: Share of different agricultural producer categories in sown area in 2016 Province Total sown area, from this (percent) ‘000 ha Farms Households Enterprises Karakalpakstan 254.9 85.1 12.5 2.4 Andijan 230.1 85.4 12.4 2.2 Bukhara 240.6 82.9 16.2 0.8 Jizzakh 401.1 91.2 6.1 2.7 Kashkadarya 495.5 83.0 14.1 2.9 Navoi 103.5 77.0 17.0 6.0 Namangan 225.0 85.5 13.6 1.0 Samarkand 359.0 82.9 16.0 1.1 Surkhandarya 285.0 82.9 15.6 1.5 Syrdarya 232.8 90.6 6.4 3.0 Tashkent 353.5 84.0 11.6 4.4 Fergana 288.9 83.9 14.2 1.9 Khorezm 236.7 81.0 16.7 2.3 Uzbekistan 3,706.5 84.7 13.0 2.4 Source: IAMO and the WB based on UzGosKomStat (2017). 33 There is a positive relationship between sown area per rural inhabitant (Table 1) and sown area per farm (Table 8) at the province level in 2016. This implies that individual farms on average had smaller sown area in provinces with higher density of rural population (Figure 3). On average, the sown area was 0.23 ha per a rural inhabitant with the smallest land-man ratio in Samarkand and Andijan (each about 0.16 ha), Surkhandarya (0.18 ha), and Fergana (0.19 ha) provinces. Production specialization of individual farms 34 Two main characteristics of existing farm categories marked the path of farm restructuring process since 2006, namely farm size and their specialization. Based on their production specialization, the Uzbek legislation defined four groups of individual farms. The largest and dominant farm type, cotton and grain producing farms, consisted of at least 10 ha of land, designated by the government for production of two strategic crops (Djanibekov 2012). 35 The second and third farm types, horticultural and garden farms, specialized in vegetables, melons, as well as gardening and grape production on plots of at least 1 ha in size. These farms were not part of the system of production targets. Additionally, horticultural and garden/vineyard farms could produce crops such as potato, melons and fodder crops. 11 36 The size of the fourth farm type, livestock farms, including poultry farms, was directly related to farm’s animal stock through a ratio of 0.33 ha per cattle equivalent (i.e. there should be at least 30 heads of cattle equivalents corresponding to 10 ha of land). Livestock farms are granted more decision-making freedom to produce cash crops than crop-growing farms. Figure 3: Relationship between sown area per individual farm and sown area per rural inhabitant Source: IAMO and the World Bank based on UzGosKomStat (2017). 37 Despite that the horticulture, garden and livestock farms have better decision-making freedom compared to the cotton-grain producing farms, according to Petrick and Djanibekov (2016) cotton-grain producers have better access to machinery supplied from parastatal machinery-tractor stations, as well as better access to diesel fuel and fertilizers (Larson et al. 2012). As these farms produce strategic cotton and wheat, along with access privileges to fertilizer and fuel, the state prioritizes their timely and sufficient access to irrigation water (Djanibekov et al. 2012). 38 During the farm optimization process, the minimum size of cotton and wheat farms, and gardening and horticulture farms, were redefined.6 The minimum size of cotton and wheat farms was increased from 10 ha to 30 ha, and for horticultural and gardening farms from 1 ha to 5 ha. Figure 4: Share of sown area under cotton and wheat in individual farms, percent Source: IAMO and the WB based on AGRIWANET and the UzGosKomStat (2017). 6 Law “On introduction of changes and additions to legislative acts of the Republic of Uzbekistan in connection to deepening of economic reforms in agriculture and water sector” (2009). 12 39 Figure 4 shows that with the progress of farm restructuring, production of cotton and wheat was transferred from former collective farms to individual farms. Despite the announced diversification of production activities in individual farms, cotton and wheat continue to dominate in their land use. Almost 80 percent of sown area in individual farm was under cotton and wheat production. 40 Law on Individual farms stipulated that leased land must be cultivated with due diligence to yield a certain minimum harvest of cotton and wheat. Yet, 2012 can be marked with two important recommendations provided by the state to individual farms: development of farms towards diversification apart from the main activity7; and along with strategic crops, secondary crops (such as mung beans, annual fodder crops such as maize and others) were recommended to be grown after wheat harvest (Platonov et al. 2014). 41 Between 2006 and 2015, covering the complete fragmentation and farm consolidation, the government took measures to convert cotton-grain producing farms into horticulture and garden farms in selected districts. Such examples in 2011 and 2012 cover the districts of Jamboy in Samarkand province, Asaka in Andijan, and Yangiyul in Tashkent. However, this did not lead to drastic changes in the production specialization structure of individual farms. 42 More visible change in farm structure can be observed with the start of the fifth stage in 2016. As Figure 5 shows the number of farms specializing in cotton-grain production declined, while the number of farms specializing in horticulture and gardens/vineyards increased. As a result of this change, 41 percent of all registered individual farms in 2016 were specializing in cotton-grain production, while those specializing in gardens and vineyards accounted for 33 percent of all farms. Only 10 percent of all individual farms specialized in horticulture, while another 10 percent were livestock farms. Figure 5: Structure of individual farms according to specialization, percent of all individual farms Source: IAMO and the WB based on UzGosKomStat (2017b) and MAWR (2017b). 43 In total land allocated to individual farms, cotton-grain farms operated 71 percent of total 5.56 million ha of land allocated to individual farms (Figure 6). With the progress of land consolidation, the share of cotton-grain producing farms increased up to 87 percent of total farm land in 2011. Further with the progress of production optimization in 2012-2015, their share in total 7 Presidential Decree UP-4478 “On measures of further improving the organization of activities and development of farming in Uzbekistan” (22.10.2012) 13 farm land shrank. The farms specializing in gardens and vineyards operated 16 percent of all individual farm land in 2016. Figure 6: Land allocated to individual farms according to production specialization, ‘000 ha Source: IAMO and the WB based on UzGosKomStat (2017b) and MAWR (2017b). 44 In 2016, livestock farms that require larger areas for fodder production and livestock grazing operated about 9 percent of all land allocated to individual farms. Figure 7: Average size of individual farms according to production specialization, ha Note: Average size of individual farms is given in total allocated land per farm. This land includes not only sown (or arable) area but also perennial crops, gardens, fallow, and unproductive land. Source: IAMO and the WB based on UzGosKomStat (2017) and MAWR (2017b). 45 During the farm optimization process in 2009-2012, the average area of allocated land to individual farms (farm size) was increased in all production specializations, although at a different scale (Figure 7). On the other hand, average size of individual farms gradually declined. The effect of farm optimization process in 2009-2012 is most visible for cotton-grain farms. During the first farm optimization wave, their average size increased from 40 ha to 96 ha. The farm consolidation continued with the second wave reaching 130 ha in 2010. The new state of optimization reduced average size of cotton-grain producers to 75 ha in 2016. 14 46 The first two waves of farm optimization had also affected average sizes of farm of other production specialization. The biggest increase could be observed for vegetable-melon producing farms. During the first wave of farm optimization in 2009, their average size almost three times, from 7 ha to 22 ha. IV. Evolution of Agricultural Output 47 This section presents the evolution and trends in gross output of agriculture and in the most important agricultural commodities such as cotton, wheat, vegetables, fodder crops, meat, milk and eggs across main producer categories. It is necessary to note that due to the level of aggregation, the province-level data does not show contrasting differences between districts where cotton production was phased out. Petrick and Djanibekov (2016) show that individual farms in districts released from cotton production targets have more diversified crop production portfolio, including a substantial share of fruits, melons and vegetables, than cotton-producing farms. 48 When measured in constant 2016 Uzbek UZS, there were two distinct trends in the path of the gross agricultural output (GAO). First, during the initial steps of reforms in 1992-1996, the agricultural sector experienced stagnation: GAO was mainly generated by agricultural enterprises (Figure 8). With the progress of farm restructuring the contribution of agricultural enterprises to GAO shrank from 70 percent in 1992 to 3 percent in 2007. The performance of agricultural sector recovered with the progress of farm restructuring. 49 Following the complete fragmentation of agricultural enterprises into individual farms, the GAO comprised contributions by three sectors: livestock output in dekhkan farms and crop production in individual farms and dekhkan farms (Figure 8). Figure 8: Evolution of GAO by producer category and sector, trillion constant 2016 UZS Note: Inflation, GDP deflator (annual percent) from the 2017 World Development Indicators (WDI) was used to convert GAO from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 50 One needs to consider that to convert GAO from current prices to constant 2016 UZS the GDP deflator from the WDI (2017) was used, which refers to the official statistics of Uzbekistan. The quality of the latter is often claimed to be unreliable due to the low level of transparency (see for instance, MacDonald 2012). To inform about possible issues with the use of official figures of agricultural output to derive constant values, Figure A in Appendix compares GDP deflator with annual percentage changes of unofficial (black market) exchange rates of UZS to US dollar in Figure 15 A. It suggests that the GDP deflator understates actual price inflation after 2014, so that deflated figures below are overstated for this period. The same applies for the 1990s, whereas the opposite seems to be the case during the early 2000s. 51 In 1992, rural households produced about 30 percent of GAO, while the remaining output originated in agricultural enterprises. In the following years, the share of households in GAO has been increasing reaching two-third in 2000-2001. In 2016, dekhkan farms contributed about 65 percent of GAO. 52 The contribution of individual farms in GAO has been increasing with the progress of the farm restructuring: it reached 31 percent in 2006 and then stabilized around 32-35 percent in the following years. The contribution of individual farms into GAO came almost entirely from the expansion of crop output. In 2016, the contribution of individual farms to GAO ranged from 25 percent in Navoi and Surkhandarya districts to 37 percent in Syrdarya and Kashkadarya (Table 10). Table 10: GAO by producer category and province in 2016, billion UZS Province GAO, billion UZS from which (percent) Farms House-holds Enter- Total Farms House- Enter- prises holds prises Karakalpakstan 458.7 847.5 31.1 1,337.3 34.3 63.4 2.3 Andijan 1,815.6 3,387.3 37.9 5,240.8 34.6 64.6 0.7 Bukhara 1,245.7 3,114.6 39.3 4,399.7 28.3 70.8 0.9 Jizzakh 798.6 1,519.9 21.5 2,340.0 34.1 65.0 0.9 Kashkadarya 1,433.7 2,381.2 46.7 3,861.7 37.1 61.7 1.2 Navoi 576.7 1,676.0 77.5 2,330.2 24.7 71.9 3.3 Namangan 1,051.1 2,426.0 35.0 3,512.1 29.9 69.1 1.0 Samarkand 2,273.9 3,272.6 153.9 5,700.5 39.9 57.4 2.7 Surkhandarya 1,027.7 3,019.2 42.3 4,089.3 25.1 73.8 1.0 Syrdarya 680.5 1,088.3 44.4 1,813.3 37.5 60.0 2.5 Tashkent 2,043.3 3,609.2 304.6 5,957.1 34.3 60.6 5.1 Fergana 1,305.5 2,578.5 69.7 3,953.7 33.0 65.2 1.8 Khorezm 935.0 1,983.2 32.2 2,950.4 31.7 67.2 1.1 Uzbekistan 15,646.1 30,903.7 936.3 47,486.1 32.9 65.1 2.0 Source: IAMO and the WB based on UzGosKomStat (2017). Crop production 53 The path of gross crop output can be divided into four stages (Figure 9). First stage comprises the 1990s with crop production concentrated in large agricultural enterprises and rural households. The second stage, during 2000-2006, can be characterized by increasing share of individual farms in crop output. During the third stage, in 2003-2004, before the intensified fragmentation took off, each of three producer categories contributed almost equal share to the gross crop output. Finally, the last fourth stage started with individual farms taking over the share of agricultural enterprises, turning again into the dual producer system. Despite that the new dual system was established, the previous one was dominated by large and industrialized enterprises, while the new one is based on almost equal contributions by individual farms specializing in cotton- wheat production, and rural (subsistence) households. 16 Figure 9: Evolution of gross crop output by producer category, trillion constant 2016 UZS Note: Inflation, GDP deflator (annual percent) from WDI was used to convert GAO from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 54 In 2016, individual farms, which cultivated about 85 percent of sown area, produced only one-third of crop output (Table 11). Contribution of individual farms to crop output in Surkhandarya province was the lowest among other provinces. Individual farms in Jizzakh and Kashkadarya provinces accounted for over two-thirds of the crop output. Table 11: Crop output by producer category and province in 2016, billion UZS Gross crop product, billion UZS from which (percent) Province House- Enter- House- Farms Total Farms Enterprises holds prises holds Karakalpakstan 436.4 229.6 25.9 691.9 63.1 33.2 3.7 Andijan 1,751.6 2,188.8 21.2 3,961.5 44.2 55.3 0.5 Bukhara 1,165.4 1,407.7 9.1 2,582.2 45.1 54.5 0.4 Jizzakh 777.6 352.2 9.6 1,139.4 68.2 30.9 0.8 Kashkadarya 1,380.3 663.2 18.3 2,061.8 66.9 32.2 0.9 Navoi 537.0 623.9 25.8 1,186.7 45.3 52.6 2.2 Namangan 981.9 1,336.5 17.2 2,335.5 42.0 57.2 0.7 Samarkand 2,155.7 1,522.7 33.5 3,711.9 58.1 41.0 0.9 Surkhandarya 948.9 1,428.3 21.3 2,398.5 39.6 59.5 0.9 Syrdarya 651.9 447.8 31.4 1,131.1 57.6 39.6 2.8 Tashkent 1,931.0 1,691.9 98.0 3,720.9 51.9 45.5 2.6 Fergana 1,231.4 1,316.1 30.1 2,577.6 47.8 51.1 1.2 Khorezm 869.4 662.4 11.5 1,543.3 56.3 42.9 0.7 Uzbekistan 14,818.5 13,871.0 352.9 29,042.4 51.0 47.8 1.2 Source: IAMO and the WB based on UzGosKomStat (2017). 55 At the national level, two patterns can be assigned for evolution of cotton area in 1992- 2016. First, its area has been decreasing gradually since 1991 due to decline in productive area and targeted expansion of food crops, mainly wheat and vegetables. In 2016, cotton area was 1.27 million ha or almost 25 percent lower of its 1992 level (Figure 10). The second pattern is that with 17 the progress of farm restructuring cotton production was shifted from agricultural enterprises to individual farms. With the completion of farm fragmentation, individual farms cultivate almost 99percent of cotton fields. Figure 10: Cotton sown area by producer category, ‘000 ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 56 Table 12 illustrates cotton area across different provinces in 1992 and 2012-2016. In 2016, cotton sown area across all provinces was larger than 80,000 ha (Navoi province with 35,000 ha of cotton area is the only exception). The largest cotton area was in Kashkadarya (160,500 ha) and Surkhandarya (117,100 ha) provinces. Since 1992, cotton area in Karakalpakstan was reduced by 36percent and in Syrdarya province by 33 percent, while in Bukhara, Khorezm and Surkhandarya provinces it stayed rather stable. Table 12: Total sown area of cotton by province, ‘000 ha Difference between Province 1992 2012 2013 2014 2015 2016 2016 and 1992, % Karakalpakstan 147.8 94.7 94.7 94.7 95.9 94.4 -36.1 Andijan 130.1 93.4 93.4 93.4 93.4 91.4 -29.7 Bukhara 115.1 110.1 112.3 111.3 110.3 108.0 -6.1 Jizzakh 141.2 101.5 101.8 101.7 101.8 99.7 -29.4 Kashkadarya 204.5 162.7 164.8 164.9 164.7 160.5 -21.5 Navoi 47.2 36.1 35.9 35.8 35.8 35.2 -25.4 Namangan 109.5 83.1 83.0 82.6 82.6 80.9 -26.1 Samarkand 117.5 93.7 94.8 94.7 93.0 89.8 -23.6 Surkhandarya 130.9 119.0 119.6 119.6 119.6 117.1 -10.5 Syrdarya 152.2 106.9 106.4 107.5 106.8 101.4 -33.4 Tashkent 121.9 100.4 98.2 95.6 94.4 89.1 -26.9 Fergana 141.8 100.1 100.1 100.1 100.1 98.0 -30.9 Khorezm 107.0 106.6 103.8 99.4 99.7 99.5 -7.0 Uzbekistan 1,666.7 1,308.3 1,308.8 1,301.5 1,298.1 1,265.1 -24.1 Source: IAMO and the WB based on UzGosKomStat (2017). 57 Despite the gradual decrease in sown area, cotton remains the dominant crop, occupying over one-third of total sown area in 2016 or 40 percent of sown area in individual farms (Table 13). Individual farms in Bukhara and Khorezm provinces allocated over a half of their land to cotton. 18 Table 13: Cotton area and its share in sown area by producer categories and province in 2016 Province Cotton sown area, ‘000 ha Share of cotton in sown area, percent Farms Enterprises Total Farms Enterprises Total Karakalpakstan 94.0 0.3 94.4 43.3 5.6 37.0 Andijan 90.9 0.5 91.4 46.3 9.0 39.7 Bukhara 107.8 0.3 108.0 54.0 14.0 44.9 Jizzakh 99.4 0.3 99.7 27.2 2.7 24.9 Kashkadarya 159.1 1.4 160.5 38.7 9.7 32.4 Navoi 34.8 0.4 35.2 43.7 6.1 34.0 Namangan 80.3 0.6 80.9 41.8 29.1 36.0 Samarkand 89.4 0.3 89.8 30.1 8.6 25.0 Surkhandarya 116.1 1.0 117.1 49.1 23.5 41.1 Syrdarya 99.1 2.2 101.4 47.0 32.5 43.6 Tashkent 88.7 0.5 89.1 29.9 3.0 25.2 Fergana 97.5 0.5 98.0 40.2 8.9 33.9 Khorezm 98.3 1.2 99.5 51.3 21.9 42.0 Uzbekistan 1,255.6 9.5 1,265.1 40.0 10.8 34.1 Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 58 Compared to cotton production, wheat production experienced significant growth both in sown area and yields. Among the core pillars of these changes was the national food self-sufficiency policy, which initiated the expansion of wheat production and its strategic importance. In 1992- 1997, wheat sown area in Uzbekistan more than doubled (Figure 11). In the following years, the area allocated to wheat stabilized. Further increase in wheat production can be attributed to yield improvement: from 2.1 tons/ha in 1997 to 4.8 tons/ha in 2016. Figure 11: Wheat sown area by producer category, ‘000 ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 59 Like cotton, individual farms dominate in wheat production and cultivate about 84 percent of total wheat area. Dekhkan farms cultivate the remaining 16 percent of wheat area. Households’ wheat area increased from 27,400 ha in 1992 to 170,000 ha in 2016. While the wheat cultivation in individual farms is under production target system, the expansion of wheat cultivation in households was due to the availability of new wheat varieties, machinery, and rural flour mills. Despite this, household remain net buyers of wheat. For instance, in Khorezm province households’ own 19 production of wheat can cover only one-third of their annual consumption requirements (Djanibekov 2008). 60 In 1992, wheat sown area was 627,000 ha, and mainly located in Kashkadarya, Samarkand and Jizzakh provinces (Table 14). Further expansion of wheat area was based on the introduction of winter wheat varieties that ensured high yields in irrigated lowlands. Particularly in areas that were under continuous cotton monoculture, winter wheat was preferred by farmers as it offered space for follow-up cultivation of high-value crops such as maize, rice, vegetables, and potatoes after the wheat harvest in early summer (Akramov 2011). Among the provinces, the biggest expansion of wheat sown area was in dryland area such as Karakalpakstan and Khorezm. Table 14: Total sown area of wheat by province, ‘000 ha Difference Province 1992 2012 2013 2014 2015 2016 between 2016 and 1992, % Karakalpakstan 9.0 64.2 68.5 64.4 64.3 64.4 615.4 Andijan 19.0 87.7 87.2 86.4 86.4 84.9 346.8 Bukhara 16.4 91.6 91.5 91.5 91.6 91.7 459.2 Jizzakh 92.3 156.6 141.9 143.6 139.7 148.1 60.5 Kashkadarya 183.7 203.2 217.7 223.3 226.4 226.3 23.2 Navoi 19.6 48.3 48.8 48.8 50.1 50.2 156.3 Namangan 17.4 87.5 88.9 90.1 90.0 89.7 415.3 Samarkand 121.5 150.4 180.5 180.5 174.8 172.2 41.7 Surkhandarya 56.6 114.4 117.7 118.5 118.5 117.5 107.5 Syrdarya 28.0 92.3 95.5 95.9 94.5 97.1 246.7 Tashkent 35.7 131.5 134.3 134.4 133.5 129.0 261.3 Fergana 21.5 128.3 126.0 125.1 123.8 120.7 461.4 Khorezm 6.3 48.0 51.1 52.3 52.3 54.4 762.8 Uzbekistan 627.0 1,404.0 1,449.6 1,454.8 1,445.9 1,446.1 130.6 Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 61 Table 15 illustrates the distribution of wheat area different producer categories and provinces. In 2016, wheat was the second dominant crop after cotton and occupied 39 percent of total sown area. Individual farms allocated 40 percent of their sown area to wheat cultivation. Wheat occupies almost a half of sown area in individual farms in Navoi and Samarkand provinces. 20 Table 15: Wheat area and its share in total sown area by producer category and province in 2016 Province Wheat sown area, ‘000 ha Share of wheat in sown area, percent Farms House- Enter- Total Farms House- Enter- Total holds prises holds prises Karakalpakstan 52.8 10.9 0.6 64.4 24.3 34.4 9.7 25.3 Andijan 79.5 4.5 0.9 84.9 40.4 15.8 17.9 36.9 Bukhara 65.3 25.9 0.5 91.7 32.7 66.4 25.4 38.1 Jizzakh 140.6 4.3 3.2 148.1 38.4 17.8 29.2 36.9 Kashkadarya 190.3 32.8 3.2 226.3 46.3 46.9 22.4 45.7 Navoi 40.2 8.4 1.6 50.2 50.5 47.6 26.0 48.5 Namangan 79.0 9.9 0.8 89.7 41.1 32.5 35.6 39.9 Samarkand 146.2 23.6 2.3 172.2 49.2 41.0 58.9 48.0 Surkhandarya 99.5 16.6 1.4 117.5 42.1 37.3 32.1 41.2 Syrdarya 91.2 4.3 1.6 97.1 43.3 28.5 22.6 41.7 Tashkent 125.2 0.6 3.1 129.0 42.2 1.6 20.1 36.5 Fergana 108.5 10.8 1.4 120.7 44.8 26.3 26.2 41.8 Khorezm 36.9 16.9 0.6 54.4 19.2 42.7 10.5 23.0 Uzbekistan 1,255.3 169.6 21.3 1,446.1 40.0 35.3 24.0 39.0 Source: IAMO and the WB based on UzGosKomStat (2017). 62 Dekhkan farms allocated over 35 percent of their land to wheat production. There are large differences in the share of wheat in households’ sown area across provinces: in Bukhara province households allocated two-thirds of their sown area to wheat production, while in Tashkent province – only 1.6 percent. Compared to wheat and cotton, vegetable production in Uzbekistan experienced different changes in sown area and structure of producers (Figure 12). Figure 12: Vegetable sown area by producer category, ‘000 ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 63 During the onset of independence, almost a half of total area under vegetable production was in agricultural enterprises, and another half in rural households. During the first stage of farm restructuring in 1992-1998, organizational changes going in the state and collective farms and the underdevelopment of marketing channels led to the stagnation of vegetable area. 21 64 The recovery of vegetable production can be associated with the completion of farm restructuring in 2007. In fact, the expansion of vegetable area in individual farms made the biggest contribution to the recovery and further increase in vegetable sown area after 2006. 65 Table 16 illustrates vegetables area across provinces in 1992 and 2012-2016. Vegetable sown area recovered in almost all provinces, except for Bukhara, Jizzakh and Syrdarya provinces. The largest relative increase in vegetable area compared to 1992 can be observed in Karakalpakstan, Khorezm, Namangan and Samarkand provinces. Table 16: Total sown area of vegetables by province, ‘000 ha Difference Province 1992 2012 2013 2014 2015 2016 between 2016 and 1992, % Karakalpakstan 8.0 8.5 9.8 10.5 10.5 10.9 36.1 Andijan 18.6 17.3 18.1 18.6 18.7 19.4 4.2 Bukhara 10.2 7.8 8.2 8.3 8.6 9.5 -6.4 Jizzakh 10.2 8.9 8.1 8.1 8.3 8.8 -13.4 Kashkadarya 13.7 15.6 15.8 16.1 16.5 17.2 25.3 Navoi 3.8 3.5 3.7 3.9 3.8 4.1 7.9 Namangan 10.1 13.9 14.0 14.3 14.4 15.1 49.2 Samarkand 21.8 24.9 26.8 26.9 27.9 29.9 37.1 Surkhandarya 12.1 13.5 13.2 13.3 13.3 15.1 24.7 Syrdarya 6.8 4.0 4.1 4.3 4.4 4.5 -33.6 Tashkent 36.3 33.8 34.2 33.7 33.3 36.8 1.4 Fergana 18.6 17.8 18.8 19.3 19.4 19.5 4.9 Khorezm 9.3 14.3 14.6 14.7 14.9 15.2 63.0 Uzbekistan 179.5 183.8 189.4 192.0 194.0 206.0 14.7 Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 66 At the national level, vegetables occupied 5.6 percent of sown area in 2016 (Table 17): the largest share was Tashkent province. Vegetables were mostly produced in dekhkan farms (households) which cultivated almost two-thirds of total vegetable area in 2016. Individual farms cultivated the remaining one-third of vegetable area. On average households allocated a slightly above 25 percent of their sown area to vegetables: over 40 percent of sown area was allocated to vegetables by households in Tashkent, Andijan, and Fergana provinces. Despite the increasing share of individual farms in vegetable production, they still allocate only 2.4 percent of their sown land to vegetables: this ranges from less than 1 percent of farms’ sown area in Syrdarya, Fergana, Surkhandarya, and Jizzakh provinces to as high as 5 percent in Samarkand and Tashkent provinces. 22 Table 17: Vegetable area and its share in sown area by producer category and province in 2016 Province Vegetable sown area, ‘000 ha Share of vegetables in sown area, percent Farms House- Enter- Total Farms House- Enter- Total holds prises holds prises Karakalpakstan 4.0 6.5 0.4 10.9 1.8 20.4 7.0 4.3 Andijan 6.7 12.2 0.4 19.4 3.4 43.0 8.1 8.4 Bukhara 3.4 6.0 0.1 9.5 1.7 15.4 4.5 4.0 Jizzakh 3.5 5.3 0.1 8.8 0.9 21.8 0.8 2.2 Kashkadarya 7.2 9.7 0.2 17.2 1.8 13.9 1.3 3.5 Navoi 1.1 3.0 0.1 4.1 1.3 16.8 1.2 4.0 Namangan 4.9 10.0 0.2 15.1 2.5 32.9 8.3 6.7 Samarkand 17.6 12.2 0.2 29.9 5.9 21.1 4.7 8.3 Surkhandarya 2.1 12.5 0.5 15.1 0.9 28.1 12.3 5.3 Syrdarya 0.9 3.5 0.1 4.5 0.4 23.4 1.5 1.9 Tashkent 16.5 19.4 0.8 36.8 5.6 47.6 5.3 10.4 Fergana 2.0 17.5 0.1 19.5 0.8 42.6 0.9 6.8 Khorezm 5.5 9.6 0.1 15.2 2.9 24.2 1.3 6.4 Uzbekistan 75.3 127.4 3.2 206.0 2.4 26.5 3.7 5.6 Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). Livestock production 67 Compared to crop production, livestock output is almost entirely produced by dekhkan farms: in 2016, 4.7 million households possessed 12 million cattle heads, from which 4 million were cows. The farm restructuring did not result in significant improvement in livestock production in individual farms (Figure 13). In 2016, individual farms operated 84 percent of sown area contributing about a half of total crop output but produced only 5 percent of livestock output. Figure 13: Evolution of livestock output by producer category, ‘000 billion constant 2016 UZS Note: Inflation, GDP deflator (annual percent) from WDI was used to convert GAO from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 23 68 In fact, the farm restructuring led to the concentration of livestock output in hands of dekhkan farms. In 1992, agricultural enterprises produced about 42 percent of livestock output. Yet, already by 1997 livestock production in these farms was reduced more than three-fold, and their contribution to livestock output made up less than 10 percent. In 2016, agricultural enterprises produced 3 percent of livestock output. 69 It has also led to delinking crop from livestock production. Cotton areas were left without organic manure from cattle, while wheat replaced fodder crops after harvesting cotton, leading to soil depletion and thereby cotton yield reduction and a loss of feed for cattle (AFD 2018). 70 Across the provinces, Jizzakh is characterized by almost complete concentration of livestock output in dekhkan farms; individual farms generate less than 2 percent of livestock output (Table 18). The highest contribution to the livestock output by non-household sector was in Tashkent province where agricultural enterprises, including poultry and dairy processing enterprises produced a bit less than 10 percent of livestock output. Table 18: Livestock output by producer category and province in 2016, billion UZS Province Gross livestock product, bln UZS from which (percent) Farms House- Enter- Total Farms House- Enter- holds prises holds prises Karakalpakstan 22.2 618.0 5.2 645.4 3.4 95.7 0.8 Andijan 64.1 1,198.5 16.7 1,279.3 5.0 93.7 1.3 Bukhara 80.3 1,706.9 30.2 1,817.5 4.4 93.9 1.7 Jizzakh 21.0 1,167.7 11.9 1,200.6 1.8 97.3 1.0 Kashkadarya 53.4 1,718.0 28.5 1,799.9 3.0 95.4 1.6 Navoi 39.6 1,052.1 51.7 1,143.5 3.5 92.0 4.5 Namangan 69.2 1,089.5 17.8 1,176.6 5.9 92.6 1.5 Samarkand 118.2 1,749.9 120.4 1,988.6 5.9 88.0 6.1 Surkhandarya 78.9 1,591.0 20.9 1,690.8 4.7 94.1 1.2 Syrdarya 28.7 640.5 13.0 682.2 4.2 93.9 1.9 Tashkent 112.3 1,917.2 206.7 2,236.2 5.0 85.7 9.2 Fergana 74.1 1,262.5 39.5 1,376.1 5.4 91.7 2.9 Khorezm 65.6 1,320.8 20.7 1,407.1 4.7 93.9 1.5 Uzbekistan 827.6 17,032.7 583.4 18,443.7 4.5 92.3 3.2 Source: IAMO and the WB based on UzGosKomStat (2017). 71 Meat production increased almost three times, from 777,000 tons in 1992 to 2,200,000 tons in 2016 (Figure 14). The farm restructuring did not consider transfer of livestock from former collective and state farms to newly-established individual farms. As a result, meat and milk production is constantly increasing, mainly due to expansion of livestock numbers in dekhkan farms. 24 Figure 14: Meat production (in light weight) by producer category, ‘000 tons Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 72 In 2016, almost 95 percent of meat in Uzbekistan is produced in household plots of dekhkan farms, ranging from 90 percent in Tashkent and Navoi provinces to 97 percent in Kashkadarya province (Table 19). Contribution of Individual farms into meat production ranged from 0.7 percent in Jizzakh province to 4.4 percent in Surkhandarya and Tashkent provinces. At national level, individual farms produce less than 3.0 percent of meat. Table 19: Meat production by producer category and province in 2016 Province Meat production, 1000 t Share of producer category in meat production, percent Farms House- Enter- Total Farms House- Enter- holds prises holds prises Karakalpakstan 2.0 92.1 1.0 95.2 2.1 96.8 1.1 Andijan 4.3 127.9 1.8 134.0 3.2 95.4 1.4 Bukhara 5.7 201.6 2.0 209.3 2.7 96.3 1.0 Jizzakh 1.3 187.8 1.0 190.2 0.7 98.8 0.5 Kashkadarya 4.1 258.2 4.0 266.3 1.5 97.0 1.5 Navoi 3.2 127.1 11.8 142.0 2.2 89.5 8.3 Namangan 3.4 120.8 0.9 125.0 2.7 96.6 0.7 Samarkand 8.6 237.2 12.1 258.0 3.3 92.0 4.7 Surkhandarya 7.5 159.5 1.6 168.5 4.4 94.6 0.9 Syrdarya 1.6 58.4 0.4 60.4 2.6 96.8 0.6 Tashkent 10.6 212.5 15.9 238.9 4.4 88.9 6.6 Fergana 5.8 136.5 5.7 148.0 3.9 92.2 3.9 Khorezm 4.3 131.8 0.8 136.9 3.1 96.3 0.6 Uzbekistan 62.2 2,051.3 59.1 2,172.5 2.9 94.4 2.7 Source: IAMO and the WB based on UzGosKomStat (2017). 73 Similar trend can be observed in milk production (Figure 15), where the dominance of dekhkan farms is even stronger: about 97 percent of milk in Uzbekistan is produced by smallholders. At the onset of 1990s, agricultural enterprises accounted for a quarter of all milk production. With the progress of farm restructuring by 1992, the production of milk in agricultural enterprises was almost entirely terminated. 25 Figure 15: Milk production by producer category, ‘000 tons Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 74 The concentration of milk in rural households continues to hinder the development of milk supply chains and the improvement of milk yields (UNDP 2010). 75 In 2016, Samarkand province accounts for the largest amount of milk production, almost 13percent of national milk originates in this province (Table 20). Syrdarya and Navoi provinces and Karakalpakstan had the least milk production levels. Compared to meat, contribution of individual farms and agricultural enterprises to milk production is relatively small in all provinces. Table 20: Milk production by producer category and province in 2016 Province Milk production, ‘000 tons Share of producer category in milk production, percent Farms House- Enter- Total Farms House- Enter- holds prises holds prises Karakalpakstan 14.4 328.3 4.2 347.0 4.2 94.6 1.2 Andijan 31.2 856.2 5.1 892.6 3.5 95.9 0.6 Bukhara 35.6 815.0 2.3 852.9 4.2 95.6 0.3 Jizzakh 13.2 521.9 1.5 536.7 2.5 97.3 0.3 Kashkadarya 19.8 997.3 6.0 1,023.1 1.9 97.5 0.6 Navoi 15.4 396.4 3.3 415.1 3.7 95.5 0.8 Namangan 16.3 618.8 2.3 637.4 2.6 97.1 0.4 Samarkand 54.7 1,171.7 1.8 1,228.3 4.5 95.4 0.2 Surkhandarya 17.1 760.8 3.7 781.5 2.2 97.3 0.5 Syrdarya 14.5 297.2 7.2 318.9 4.5 93.2 2.2 Tashkent 40.4 773.9 24.0 838.3 4.8 92.3 2.9 Fergana 40.2 861.1 8.2 909.6 4.4 94.7 0.9 Khorezm 39.9 880.1 2.3 922.3 4.3 95.4 0.2 Uzbekistan 352.7 9,278.8 71.9 9,703.4 3.6 95.6 0.7 Source: IAMO and the WB based on UzGosKomStat (2017). 76 In contrast to meat and milk, which predominantly produced in household plots, egg production in agricultural enterprises and individual farms has been developing rapidly (Figure 16). At the onset of reforms, egg production in agricultural enterprises shrank rapidly, leading to the increase in the share of dekhkan farms in egg production from 50 percent in 1992 to 70 percent in 1997. The progress of farm restructuring since 2001 and, particularly, the development of poultry- specialized enterprises and individual farms led to the recovery of egg production outside of 26 household plots. In 2016, dekhkan farms produced about 57 percent of all eggs, while individual farms contributed about a one third of production volumes. Since 2007, the contribution of individual farms in egg production increased from 4 percent to 11 percent. 77 There are big variations in egg production volumes across regions (Table 21). The biggest production is concentrated in the provinces around the two biggest cities, Tashkent and Samarkand. Two provinces surrounding these two cities accounted for about 40 percent of egg production. 78 The contributions of different producer categories also varied across provinces. For instance, poultry farms produced one-third of eggs in Namangan and 20 percent in Samarkand province. In fact, these two provinces accounted for 60 percent of all eggs produced in individual farms. In Bukhara and Kashkadarya contribution of individual farms in egg production was less than 2 percent. Figure 16: Egg production by producer category, million eggs Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). Table 21: Egg production by producer category and province in 2016 Province Egg production, million eggs Share of producer category in egg production, percent Farms House- Enter- Total Farms House- Enter- holds prises holds prises Karakalpakstan 12.4 218.1 5.5 236.0 5.3 92.4 2.3 Andijan 50.2 436.5 45.6 532.3 9.4 82.0 8.6 Bukhara 5.9 213.5 128.3 347.7 1.7 61.4 36.9 Jizzakh 15.0 156.1 61.6 232.6 6.5 67.1 26.5 Kashkadarya 8.9 315.0 134.0 457.9 1.9 68.8 29.3 Navoi 34.7 177.1 50.5 262.3 13.2 67.5 19.3 Namangan 138.6 185.2 99.5 423.3 32.8 43.7 23.5 Samarkand 234.5 525.6 406.9 1,167.0 20.1 45.0 34.9 Surkhandarya 13.7 212.3 85.8 311.8 4.4 68.1 27.5 Syrdarya 13.2 98.9 33.9 146.0 9.0 67.8 23.2 Tashkent 57.9 501.3 740.4 1,299.6 4.5 38.6 57.0 Fergana 48.4 253.3 48.2 349.9 13.8 72.4 13.8 Khorezm 23.3 235.5 127.4 386.1 6.0 61.0 33.0 Uzbekistan 656.7 3,528.2 1,967.5 6,152.5 10.7 57.3 32.0 Source: IAMO and the WB based on UzGosKomStat (2017). 27 79 Except for Karakalpakstan, Andijan and Fergana provinces, agricultural enterprises contributed over a quarter of eggs production at the province level. Due to the development of poultry farms, most eggs in Tashkent province were produced in agricultural enterprises. Fodder production 80 The expansion of wheat area that largely occurred in irrigated lands was at the expense of area under fodder crops, particularly of perennials such as alfalfa which were part of cotton rotation scheme in agricultural enterprises (Figure 17). In 1992, sown area under fodder crops was 1 million ha or 24 percent of total sown area. Fodder crops were mainly produced in agricultural enterprises. In 2016, only 333,000 ha were allocated under fodder crops, despite slight increase since 2008. With the fragmentation of large-farms, individual farms took over the largest share of fodder production for a crop rotation after winter wheat or as fodder in livestock farms. 81 Despite the decline in absolute numbers, in 2016 agricultural enterprises allocated the largest share of their sown area to fodder crops (Table 22), over one-third at the national level, ranging from over 60 percent in Bukhara province to just 10 percent in Samarkand province. Individual farms allocated slightly over 8 percent of their sown area to fodder crops. Despite this, over three-fourths of fodder crops area is operated by individual farms. Households which possess almost the entire cattle operate only 13 percent of fodder area, and thus rely on crop byproducts and fodder purchases for feeding their livestock. Figure 17: Sown area of fodder crops by producer category, ‘000 ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 82 Despite the growth in cattle herd, the sown area under fodder crops continued to decline. As a result, the trend in fodder production did not follow the growth in cattle herd: since 1992 fodder area per a head of livestock unit (LSU) has declined almost three times: from 0.15 ha/LSU in 1992 to 0.05 ha/LSU in 2016 (Figure 18). 83 The situation is even worse in dekhkan farms which possess over 90 percent of all cattle. The ratio between fodder area and LSU in these small-scale subsistence farms continues to decline. In 2016, compared to individual farms that had about 0.3 ha of fodder area pea LSU, the ratio of fodder area to LSU in dekhkan farms was only 32 m2 per LSU. 28 Table 22: Fodder area and its share in sown area by producer category and province in 2016 Province Fodder sown area, 1000 ha Share of fodder in sown area, percent Farms House- Enter- Total Farms House- Enter- Total holds prises holds prises Karakalpakstan 17.4 3.3 1.6 22.3 8.0 10.4 26.3 8.7 Andijan 10.9 1.1 2.3 14.3 5.5 3.9 45.2 6.2 Bukhara 21.0 1.9 1.2 24.1 10.5 4.9 60.8 10.0 Jizzakh 33.3 7.7 3.0 44.0 9.1 31.7 27.5 11.0 Kashkadarya 26.4 6.9 4.6 37.9 6.4 9.9 32.0 7.6 Navoi 2.0 2.4 1.6 6.0 2.5 13.6 25.6 5.8 Namangan 18.5 0.8 0.3 19.6 9.6 2.6 13.8 8.7 Samarkand 16.9 7.9 0.4 25.2 5.7 13.7 10.1 7.0 Surkhandarya 15.1 2.1 1.2 18.4 6.4 4.7 27.4 6.5 Syrdarya 8.2 1.3 1.6 11.1 3.9 8.7 23.2 4.8 Tashkent 34.2 6.1 8.5 48.8 11.5 14.9 54.3 13.8 Fergana 29.7 0.5 2.9 33.1 12.3 1.2 52.7 11.5 Khorezm 24.1 2.3 2.3 28.7 12.6 5.8 43.1 12.1 Uzbekistan 257.7 44.3 31.5 333.5 8.2 9.2 35.6 9.0 Source: IAMO and the WB based on UzGosKomStat (2017). 84 According to the official statistics, the general trend of shrinking area under fodder crops has not affected milk yields: 1.7 tons/cow in 1992 and 2.3 tons/cow in 2016, which is almost 40 percent increase. This can be explained by the fact that most cattle are held in household plots of dekhkan farms which rely in feeding their livestock by open grazing after harvesting and crop byproducts (UNDP 2010). Figure 18: Average sown area of fodder crops per LSU by producer category, ha / LSU Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 85 In 2016, an average ratio of fodder area to livestock units was about 0.022 ha/LSU, ranging from 0.043 ha in Tashkent province to just 0.009 ha in Navoi province (Table 23). Individual farms, which possess about 4 percent of all LSU, had on average 0.3 ha of fodder area per a head of LSU. This ratio of fodder-to-livestock is the highest in Tashkent and Jizzakh provinces and the smallest in Navoi province. 29 Table 23: Fodder area per LSU by producer category and province in 2016, ha/LSU Province Fodder sown area per livestock, ha/LSU Farms Households Enterprises Total Karakalpakstan 0.470 0.003 0.066 0.019 Andijan 0.201 0.001 0.125 0.011 Bukhara 0.248 0.001 0.028 0.017 Jizzakh 0.808 0.008 0.116 0.041 Kashkadarya 0.323 0.004 0.058 0.019 Navoi 0.045 0.005 0.019 0.009 Namangan 0.401 0.001 0.025 0.024 Samarkand 0.168 0.005 0.006 0.014 Surkhandarya 0.182 0.002 0.031 0.016 Syrdarya 0.223 0.003 0.162 0.024 Tashkent 0.382 0.006 0.074 0.043 Fergana 0.516 0.000 0.104 0.028 Khorezm 0.302 0.003 0.121 0.029 Uzbekistan 0.308 0.003 0.056 0.022 Source: IAMO and the WB based on UzGosKomStat (2017). Productivity measures 86 Five indicators were used to calculate the average agricultural productivity in this report. It included: (i) labor productivity calculated as a ratio of GAO in constant prices divided by number of agricultural workers; (ii) land productivity calculated as a ratio of GAO in constant prices divided by agricultural land area; (iii) livestock productivity calculated as a ratio of livestock GAO in constant prices divided number of LSU; (iv) cotton yields; and (v) wheat yields. 87 Figure 19 shows the trends of selected productivity indicators. The GAO and crop and livestock outputs in constant 2016 UZS were recovering since 1996. The output growth in livestock sector, measured in meat, milk, eggs, and wool production, outperformed the growth of crop sector. While the gross output of crop sector increased by more than 50 percent, the area of sown land was declining. In contrast, the number of livestock units and the gross livestock output increased after their initial stagnation in 1992-1998 (Figure 20). Figure 19: Agricultural output, sown area and employment in 1992-2016, 1992=100 Note: Inflation, GDP deflator (annual percent) from WDI was used to convert GAO, gross crop and livestock output from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 30 88 The restructuring of collective and state farms and the land transfer to individual farms resulted in the decline of agricultural employment up to 80 percent by 2008 compared to the 1992- level.8 Yet, following the optimization program in 2008/2009, employment in agriculture has been gradually increasing. Figure 20: Agricultural output per labor, land and livestock unit, 1992=100 Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 89 Agricultural labor productivity grew relatively fast. On average, it grew by 1.53 percent annually, between 1996 and 2016, which was faster than in manufacturing, construction, and trade and catering (Table 24). Yet, it has slowed down over time, declining from 2.16 percent ibn 1996- 2000 to 0.90 percent in 2012-2016. Table 24: Changes in labor productivity in selected sectors, Uzbekistan, 1996-2016 1996-2000 2001-2008 2009-2011 2012-2016 1996-2016 Change in total productivity, inc. in: 2.85 4.44 5.63 6.16 4.55 Agriculture 2.16 1.84 1.68 0.90 1.53 Manufacturing 0.27 1.29 0.20 0.83 0.92 Construction -0.10 0.22 0.16 1.11 0.41 Trade and catering 0.57 0.48 1.53 1.28 0.77 Source: World Bank (2018). 90 The growth in average agricultural productivity has not translated into better quality jobs. In primary agriculture, most jobs were in cotton and wheat sectors. They were largely seasonal/temporary and with low rates of return, and low labor intensity. While the absolute number of agricultural labor increased from 3.5 million in 1996 to 3.7 million in 2016, their share in total labor force declined from 43 percent to 30 percent, respectively (World Bank 2018). The small land area allocated for horticulture and livestock/fodder has limited job creation in the sector, although labor intensity of horticulture and livestock production per se is much higher than that of cotton and grains (World Bank 2012). On the other hand, light industry and food processing industry, which depend on the primary agriculture for supply of raw materials, have shed the labor in big numbers (Table 25). Among the reasons were the lack of volumes and quality of raw materials for processing and the poor access to foreign markets to sell Uzbek products. This labor shedding has reduced the average productivity of the economy, as labor productivity in food and light industries is pretty high. Most labor, who left agriculture and these industrial jobs, moved to much less 8 Official figures for agricultural employment also include self-employed members of dekhkan farms. 31 productive construction and trade & catering jobs or out-migrated. In other words, agriculture has not served the food systems well in terms of job creation. Table 25: Changes in employment and labor productivity of subsectors of Uzbekistan’s economy, 1996-2016 Employment, million Labor productivity (output per worker), ‘000 Som 1996 2016 Change, % 1996 2016 Change, % Agriculture 3.51 3.69 5.1 738 2,275 208.2 Food industry 0.82 0.71 -13.1 1,698 10,650 527.4 Light industry 0.21 0.14 -33.9 1,812 6,031 233.0 Construction 0.54 1.26 134.4 957 2,730 185.2 Trade and catering 0.71 1.52 112.6 1,068 3,751 251.1 TOTAL 8.21 12.28 49.6 1,158 2,811 142.6 Source: World Bank (2018). 91 The livestock productivity trend differs from those of sown land and labor productivity. The livestock productivity was improving in 1997-2002 recovering after the initial stagnation. In 2003- 2009 it experienced a decline as the growth of livestock numbers outperformed the growth in their output. Since 2009 the livestock productivity has been around 93-97 percent of its 1992-level. 92 During the early stages of farm restructuring (1992-2008) cotton yields at the national level were declining and had higher yearly fluctuations (Figure 21). Following the program of farm size and location optimization, the cotton yields on individual farms first increased and fluctuated less but then flattened and stopped growing. One of the reasons for this weak growth in cotton yields is a decoupling of livestock from cotton production. In the past, cattle provided manure on cotton fields, and cotton production was followed by fodder production, which also helped restore soil nutrients. With the introduction of production quota for wheat, it replaced fodder, putting pressure on soil fertility on which cotton is produced (AFD 2018). 93 Wheat yields have been going up since 1994 (Figure 22). The highest increase in wheat yields was observed during the first two stages of farm restructuring. The optimization program initiated in 2008/2009 contributed to the further increase: wheat yield reached the average of 4.8 tons/ha. Figure 21: National average yields of cotton by producer category, tons/ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 94 Dekhkan farms always achieved higher yields than other two producers, reaching unprecedented 6 tons/ha in 2016. This can be attributed to the governmental program to adopt higher yield varieties, expansion to productive irrigated land, improved mechanisms of seed distribution owing to the public maintenance of agricultural research facilities, import of modern 32 grain combines as well as continuous increase in application rates of mineral fertilizers (Pomfret 2008, Lerman et al. 2016, Bobojonov et al. 2017). This indicates that small farms can be highly productive if enabling environment is conducive for their development. Figure 22: Wheat yields by producer category, tons/ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 95 Vegetable yields, derived by dividing vegetable production to vegetable sown area, have been continuously increasing after the government initiated the farm fragmentation in 1998 (Figure 23): an increase from 19 tons/ha in 1998 to 55 tons/ha in 2016. During the first two stages of farm restructuring, vegetable yields in individual farms were about 20-25 percent lower than those in dekhkan farms. The completion of farm restructuring in 2006 narrowed the yields gap between these two producers. 96 The increase in vegetable yields in individual farms can be associated with the fact that these crops are exempted from the state production targets, and thus provide direct economic incentives to producers. Despite its economic attractiveness and increasing yields, the underdevelopment of marketing and export channels, as well as the state priority in input distribution to cotton-wheat production (Lerman et al. 2016) hinders the development of vegetable production in individual farms. In 2016, only 2.4 percent of sown area in individual farms was under vegetables (Table 17). Figure 23: Vegetable yields by producer category, tons/ha Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 97 Productivity of sown land, measured in constant 2016 UZS, has been increasing since 2005/06 (Figure 24). This can be attributed to the rapid decline of the number of loss-making 33 agricultural enterprises and completion of land transfer from of agricultural enterprises to individual farms. Dekhkan farms had much higher productivity of a hectare of sown land than individual farms and agricultural enterprises. In 2016, the total crop output per a hectare of sown area in Uzbekistan was 7.8 million UZS, ranging from 4.5 million UZS in agricultural enterprises and individual farms to almost 29 million UZS in dekhkan farm plots. The average national value of land productivity was closer to the value in individual farms, which operate about 84 percent of sown area. Figure 24: Sown land productivity by producer category, million constant 2016 UZS/ha Note: Inflation, GDP deflator (annual percent) from WDI was used to convert GAO from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 98 Before the completion of farm restructuring in 2016, the sown land productivity of individual farms and agricultural enterprises was almost the same. The completion of farm restructuring widened the gap between these two producer categories. 99 The comparison of sown land productivity across producer categories and provinces in 2016 shows that sown land productivity in individual and dekhkan farms in Karakalpakstan and in Jizzakh and Kashkadarya provinces was below the national level (Table 26). In all provinces, small-scale dekhkan farms were more productive than individual farms. Individual farms and households in Andijan province had the highest sown land productivity. In 2016, in Karakalpakstan as well as Namangan, Samarkand, Surkhandarya, Syrdarya and Fergana provinces, sown land productivity in agricultural enterprises was higher than in individual farms. 100 Livestock unit productivity in constant 2016 UZS values experienced a recovery in 1996-2002 (Figure 25) mainly due to the improvement of livestock output in dekhkan farms. Yet, with the decline of area under fodder crops, the growth of gross livestock output did not follow the speed of cattle herd growth in dekhkan farms. This can be tracked by the average meat production (in live weight) per cattle head which was increasing gradually since 1997 from 154 kg/cattle to 162 kg/cattle in 2011 (Figure 27). 101 Another reason for the stagnating livestock productivity in dekhkan farms is the scarcity of pastures and fodder sources, and dependence of households in grazing their cattle in fields after harvesting. The transfer of land to individual farms reduced opportunity for such grazing practices for dekhkan farms (UNDP 2010; ADF 2018). 34 Table 26: Sown land productivity by producer category and province in 2016, million UZS/ha Province All producers Farms Households Enterprises Karakalpakstan 2.7 2.0 7.2 4.3 Andijan 17.2 8.9 76.9 4.2 Bukhara 10.7 5.8 36.0 4.6 Jizzakh 2.8 2.1 14.5 0.9 Kashkadarya 4.2 3.4 9.5 1.3 Navoi 11.5 6.7 35.4 4.1 Namangan 10.4 5.1 43.7 7.9 Samarkand 10.3 7.2 26.4 8.5 Surkhandarya 8.4 4.0 32.2 4.9 Syrdarya 4.9 3.1 29.9 4.6 Tashkent 10.5 6.5 41.4 6.3 Fergana 8.9 5.1 32.1 5.5 Khorezm 6.5 4.5 16.7 2.2 Uzbekistan 7.8 4.7 28.9 4.0 Source: IAMO and the WB based on UzGosKomStat (2017). 102 The comparison of values of livestock unit productivity across producer categories shows that in 1995-2003 smallholders had the highest productivity. In the following years, productivity of livestock units in households and individual farms converged. In 2016, the livestock output per livestock unit was 1.22 million UZS, ranging from around 1 million UZS in agricultural enterprises and individual farms to 1.24 million in dekhkan farms. Since the latter possess over 90 percent of all LSUs, the national value of livestock unit productivity is closer to the productivity of dekhkan farms. Figure 25: Livestock output per LSU by producer category, million constant 2016 UZS/LSU Note: Inflation, GDP deflator (annual percent) from WDI was used to convert GAO from current prices to constant 2016 UZS. Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 103 Since 2010, the dekhkan farms were increasing milk yields (from 1.6 tons/cow in 2010 to 2.3 tons/cow in 2016), which stabilized the livestock unit productivity (Figure 26). Milk yields in agricultural enterprises declined rapidly during the first stage of restructuring (1992-1997) but stabilized after 1998. With the completion of farm fragmentation and after the first wave of optimization program, agricultural enterprises and individual farms experienced similar, almost parallel, increase in milk yield: milk yield per cow in individual farms doubled in 2008-2016. 35 Figure 26: National average milk yields by producer category, tons/cow Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 104 While it is difficult to associate the increasing milk yields in dekhkan farms with the expanded fodder area and improved quality of pastures, the increase can be due to the government programs to improve availability of cotton byproducts to livestock producers, import of better cow breeds, improvement of accessibility and quality of veterinary services, and better feeding practices. Understanding of the reasons of increasing milk yields requires further investigation. 105 The farm restructuring affected meat production in agricultural enterprises (Figure 27). Producing about 2 percent of total meat in Uzbekistan, agricultural enterprises had recovered meat production per a cattle head from 126 kg (in live weight) in 2002 to 409 kg in 2016. Similar recovering trend in meat production can be observed in individual farms, yet at a lower speed than in agricultural enterprises: in 2002-2016 meat production per cattle in agricultural enterprises gained additional 224 percent, in individual farms it was 77 percent. Figure 27: National average meat production (in live weight) by producer category, kg/cattle Source: IAMO and the WB based on AGRIWANET and UzGosKomStat (2017). 106 In 2002-2016, meat production per cattle in dekhkan farms was much lower than in agricultural enterprises, but higher than in individual farms. Since dekhkan farms possess most of the livestock, the general increase in meat productivity was not as rapid as in case of milk yield. The average meat production (in live weight) per cattle head increased slowly from 154 kg/cattle which in 1997 to 162 kg/cattle in 2011, or only 5.3 percent increase over 15 years. The meat production per cattle increased from 165 kg in 2002 to 178 kg in 2016. 36 107 As in the case of productivity of sown land, individual farms and dekhkan farms in Karakalpakstan had the lowest productivity of livestock units (0.6 million UZS/LSU), almost two times lower than the national average (Table 27). The highest livestock productivity can be observed in Tashkent and Navoi provinces. In most provinces households outperform individual farms and agricultural enterprises in terms of livestock productivity, except for Andijan, Namangan, Samarkand, and Fergana provinces. Table 27: Livestock output per LSU by producer category and province in 2016, million UZS/LSU Province All producers Farms Households Enterprises Karakalpakstan 0.56 0.60 0.57 0.22 Andijan 1.02 1.18 1.02 0.91 Bukhara 1.26 0.95 1.30 0.71 Jizzakh 1.11 0.51 1.15 0.46 Kashkadarya 0.90 0.65 0.94 0.36 Navoi 1.74 0.89 2.00 0.61 Namangan 1.46 1.50 1.46 1.51 Samarkand 1.10 1.18 1.07 1.94 Surkhandarya 1.46 0.95 1.54 0.53 Syrdarya 1.45 0.78 1.52 1.32 Tashkent 1.95 1.26 2.04 1.81 Fergana 1.18 1.29 1.16 1.42 Khorezm 1.45 0.82 1.51 1.09 Uzbekistan 1.22 0.99 1.24 1.04 Source: IAMO and the WB based on UzGosKomStat (2017). 37 V. Conclusions and Recommendations Farm restructuring outcomes 108 Since national independence, government-led farm restructuring in Uzbekistan has followed a non-linear path: • During the 1990s up to 2007, the former collective farms were hesitantly dismantled and split into smaller individual farms. However, land ownership remained public and farms continued to be subjected to cotton and wheat delivery quotas that persist up to date. Inputs like seed, fertilizer, and machinery have been mostly provided by government agencies, also focusing on cotton and wheat production. Farmers obtained the right to produce and sell crops other than cotton or wheat on private terms. The main policy goals were to maintain a persistent supply of raw cotton to secure government’s export revenues and of wheat to secure essential food supply. • Unsatisfied with the productivity of individual farms that resulted from the decollectivization process and aware of a mismatch between water supply infrastructure and farm sizes, the government implemented a farm consolidation policy after 2008. By government decree, smaller farms, particularly of cotton-grain production specialization, were merged into bigger ones. • After 2015, the official policy turned to the diversification of crop rotations away from cotton and wheat. In addition to productivity increases among traditional crops, the government now also aims to increase the output of fruits, vegetables and other high-value crops. Some land is being shifted from cotton and wheat to other crops, although the magnitude of this shift remains relatively small. Farmers are supposed to be integrated into multi-profile farms, which could be engaged in agricultural production and processing, preparation, storage, marketing, and provision of other services. Reforms also aim to strengthen the value chains and permit direct export without going through the government monopoly Uzagroeksport and mandatory submission of part of export revenue in foreign currency. 109 Against these declared policy goals, the major trends of farm restructuring were as follows: • The decollectivization process was formally successful and after a long delay farms were fundamentally restructured. • Cotton production was taken over completely by individual farms. The areas sown to cotton were gradually reduced. Yields stabilized but in the last ten years it stopped to grow, leading to the decline in the national cotton production. • Wheat self-sufficiency has progressed through expanding the wheat sown area and rising yields, to which farm restructuring has partially contributed. Yet, wheat import, mainly from Kazakhstan, has grown recently, accounting for 43 percent of wheat production in 2017 compared to 25 percent in 2010 (USDA 2018). Thus, the increase in demand for wheat from Uzbekistan’s consumers, which grew from 7.7 million tons in 2010 to 9.7 million tons in 2017, was largely met by imports, not domestic production, due to the mismatch in wheat quality between locally-produced wheat and higher quality import demanded by consumers. • Sown area and yields under higher-value food crops such as fruits and vegetables in individual farms have risen. The structure of individual farms according to their production specialization is currently being gradually diversified from cotton/wheat to horticulture farms, but their productivity significantly lags the productivity of dekhkan farms. 38 • Livestock on large farms mostly collapsed and shifted to dekhkan farms. 110 The official statistics aggregated at the province level does not provide full the picture to understand the changes induced through the optimization program. Such drastic changes can be observed at district level, e.g. known examples of individual farms in districts in Samarkand that were released from cotton cultivation and diversified to fruit and vegetable production. Remaining areas for improvements 111 According to the analysis in this report and the references quoted, the following problem areas persist and need to be addressed: • Over two and a half decades after independence, the government has maintained strict control of the entire production chain, with only few exceptions in horticulture and livestock products. In addition, producers of non-strategic crops greatly depend on state- mandated land and input allocations. • Strategic crops continue to dominate the sown area of individual farms and the land allocations to these crops are not driven by market signals. • Dekhkan farms are disconnected from food value chains and agribusinesses. Larger farms do not work as nuclear or anchor farms for small dekhkan farms. • Farm restructuring by decrees and weak property rights in land use curtail management and investment incentives and raise issues of the just access to farmland for the rural population. • A limited fodder base constrains livestock expansion, the sub-sector with a large potential for inclusive growth and job creation, and keeps it concentrated in dekhkan farms. Crop and livestock production are decoupled. As a result, cotton, for example, does not benefit from organic manure while livestock producers lost access to fodder, which was previously produced after harvesting cotton. Fodder in individual farms was replaced by wheat grown under production quota and procurement systems. • Individual farms producing high-value crop are constrained in their access to fertilizers, fuel, machinery, credit, value chains, and export channels. • Problems pertaining to the stability and distribution of irrigation water supply have not been resolved in a satisfactory manner. Recommendations on the future of farming in Uzbekistan 112 Like other Central Asian governments, Uzbekistan has struggled to find a post-socialist model for its farming sector. The debates still focus on the desirable farm size where the Soviet legacy of industrialized collective farms co-existing with private household plots marked the extremes. The global experience, on the other hand, stresses the need for flexibility in farming structures (Otsuka et al. 2016). Very small plots operated by dekhkan farms in Uzbekistan are suboptimal, but once farm size reaches 30-40 ha for wheat and cotton, increasing it further does not bring much economic gain. There is generally no evidence of economy of scale in primary agriculture, and any gains from economy of scale are easily negated by extra monitoring costs of wage labor and farm managers. These factors highlight the importance of individual incentives for farm efficiency and account for the predominance of family farms in a market economy (Lerman and Childress 2013). 113 Optimal farm size globally is determined in each case by the managerial capacity of the farmer and many exogenous factors. Farm size is often growing in the countries where economic growth and rising off-farm wages lead to the outflow of agricultural labor. Thus, rather than 39 targeting at a type or size of farm organization, it is better to help farmers: (i) improve their managerial capacity; (ii) react to market signals by removing agricultural distortions; (iii) lease or buy more land when profitable; and (iv) access a set of supporting agricultural services. 114 The future reform agenda would need to focus on: (i) removal of the production distortions; (ii) increase of the impact of agricultural public expenditures on the sector performance; and (iii) help of smallholders to reduce transaction costs. The specific actions can be the following: Production distortions • Phase out the cotton and wheat production quotas, inside and outside of the clusters. • Align farm-gate prices for cotton and wheat with market prices. • Avoid the introduction of new mandates (such as on high value crops or diversification) that impose yet other production constraints on farmers. • Liberalize land market to allow farm restructuring to evolve responding to market signals. This does not necessarily require full privatization of farmland in the short run, but the legalization of land rentals and sub-rentals, revoking the threat of public land seizures, and formalization of labor contracts. It is important to ensure that transaction costs for moving land from less to more efficient farmers are minimized, irrespective of size or type of farms. Efficiency of agricultural public expenditures • Use public expenditures more effectively to help farmers increase returns to land and labor. It would require redirection of farm subsidies towards public good provision and empowering national and local governments to allow more effective service delivery such as land registries or advisory services. • Increase the capacity of applied agricultural research and extension/advisory services to produce new location-specific technologies and make them available to farmers. • Ensure that an adequate amount of public expenditures is allocated to the above- mentioned programs to make a difference. Many of them were underfunded in the past. Programs for dekhkan farms • Introduce public programs targeting dekhkan farms, especially those who want to expand and grow. These programs can facilitate establishment of productive partnerships (cooperatives) among farmers, and between farm groups and agribusiness, to improve quality and safety of products and connect these farms to modern value chains. Such programs are common in many countries with small farms. The role of government would be to finance: (i) establishment of farm groups; (ii) improvement of their farming and managerial capacity; and (iii) activities of anchor agribusinesses to develop vertical networks and business arrangements and provide veterinary and advisory services to small farmers on behalf of the government. 115 How a future and still growing rural population will be employed and secure its income is one of the essential questions the government seeks to address. It would require actions beyond the farm restructuring. Farm restructuring is only a part of the puzzle to be brought together into one coherent piece, not a silver bullet as often perceived. 40 References ADF (2018): Uzbek agrarian systems: Lessons learned from 3 diagnostics done in Kachkadarya, Dzizzak and Namangan Provinces (March-December 2017). French Development Agency, Unpublished Power Point Presentation. AGRIWANET project (2017): Database on agriculture in Central Asia across producer categories and provinces. Halle (Saale), Germany. www.iamo.de/en/agriwanet Akramov, K. (2011): International food prices, agricultural transformation, and food security in Central Asia. Development in Practice 21 (4-5), 741-754. Bobojonov, I., Djanibekov, N., and P. Voigt (2017): Future perspectives on regional and international food security: Emerging players in the region: Uzbekistan. In: Gomez y Paloma, S., Mary, S., Langrell, S., Ciaian, P. (Eds.) The Eurasian Wheat Belt and Food Security: Global and Regional Aspects. Springer Cham, pp. 195-213. Djanibekov, N. (2008): A Micro-economic analysis of farm restructuring in the Khorezm Region, Uzbekistan, PhD dissertation, ZEF/Bonn University. Djanibekov, N., Van Assche, K., Bobojonov, I. and J.P.A. Lamers (2012): Farm restructuring and land consolidation in Uzbekistan: New farms with old barriers. Europe-Asia Studies 64 (6), 1101-1126. Djanibekov, N., Bobojonov, I., Van Assche, K., Rudenko, I., Nurmetov, K., and J. P.A. Lamers (2015): Farm restructuring in Uzbekistan through fragmentation to consolidation. In: Lamers, J.P.A., Khamzina, A., Rudenko, I., Vlek, P.L.G. (Eds.) Restructuring Land Allocation, Water Use and Agricultural Value Chains: Technologies, Policies and Practices for the Lower Amurdarya Region. V&R unipress, Bonn University Press, Goettingen, pp. 33-44. Djanibekov, U., Van Assche, K., Boezeman, D. and N. Djanibekov (2013): Understanding contracts in evolving agro-economies: Farmers, dekhkans and networks in Khorezm, Uzbekistan. Journal of Rural Studies 32, 137-147. Djanibekov, U., Van Assche, K., Boezeman, D., Villamor, G.B. and N. Djanibekov (2015): Revealing the role of agricultural contracts in rural livelihoods in Uzbekistan. In: Kimhi, A., Lerman. Z. (Eds.), Agricultural Transition in Post-Soviet Europe and Central Asia after 20 Years. Studies on Agricultural and Food Sector in Transition Economies, IAMO, Halle (Saale). Larson, D.L., Khidirov, D. and I. Ramniceanu (2012): Uzbekistan: Strengthening the horticulture value chain. World Bank, Washington DC. Lerman, Z. (2008): Agricultural development in Central Asia: A survey of Uzbekistan, 2007–2008. Eurasian Geography and Economics 49 (4), 481-505. Lerman, Z. and M. Childress (2013): The Evolution of Central Asia’s farming sectors: Structure, Performance, and Policy. The World Bank report. Unpublished. Lerman, Z., Sedik, D., Yusupov, Y., Stanchin, I. and I. Kazakevich (2016): Wheat production and regional food security in CIS: The case of Belarus, Turkmenistan, and Uzbekistan. FAO Policy Studies on Rural Transition No. 2016-1. MacDonald, S. (2012): Economic policy and cotton in Uzbekistan. Report from Economic Research Service CWS-12 h-01. USDA. MAWR (2017a): Official figures on number of machinery in agriculture across provinces. Tashkent. MAWR (2017b): Official figures on structure of individual farms by production specialization across provinces. Tashkent. Otsuka, K., Liu, Y. and F. Yamauchi (2016): The future of small farms in Asia. Development Policy Review 34, 441–461. Petrick, M. and N. Djanibekov (2016): Obstacles to crop diversification and cotton harvest mechanization: Farm survey evidence from two contrasting districts in Uzbekistan. IAMO Discussion Paper No. 153, Halle (Saale). http://purl.umn.edu/234226 41 Platonov, A., Wegerich, K., Kazbekov, J. and F. Kabilov (2014): Beyond the state order? Second crop production in the Ferghana Valley, Uzbekistan. International Journal of Water Governance 2, 83- 104. Pomfret, R. (2000): The Uzbek model of economic development, 1991-99. Economics of Transition 8 (3), pp. 733-748. Pomfret, R. (2008): Tajikistan, Turkmenistan, and Uzbekistan. In: Anderson K, Swinnen J (eds) Distortions to Agricultural Incentives in Europe’s Transition Economies. Washington, DC: World Bank, pp. 297-338. Pugach, I., Yusupov, Y. and Z. Berdinazarov (2016): Agricultural policy in what production and crop diversification in Uzbekistan. IAMO Discussion Paper No. 157, Halle (Saale), IAMO. Shtaltovna, A. and A.-K. Hornidge (2014) A comparative study on cotton production in Kazakhstan and Uzbekistan. ZEF, University of Bonn. Swinkels, R., Romanova, E. and E. Kochkin (2016): Assessing the social impact of cotton harvest mechanization in Uzbekistan. World Bank, Washington DC. UNDP (2010): Livestock production in Uzbekistan: Current state, challenges and prospects. Nasaf, Tashkent. USDA (2018): United States Department of Agriculture. Online database with agricultural commodity balances. Washington, DC. UzGosKomStat (2017): Agriculture of Uzbekistan, and official figures on structure of individual farms by production specialization across provinces. The State Committee of Statistics of Uzbekistan. Tashkent. World Bank (2017): GDP deflator (annual percent) for Uzbekistan 1992-2016, World Development Indicators. World Bank (2018): Growth Diagnostic for Uzbekistan. Washington, DC. 42 Appendix: Policies events related to restructuring of individual farms in 2006-2017 Date Policy event Highlight of the policy event Effect of the policy event January 9, Presidential decree UP-3709 “On UP-3709 addressed the need to transform agricultural cooperatives The law initiated the transfer of vegetable and 2006 measures of deepening economic (shirkats) specialized in vegetable, melons, fruit and grape fruit production from shirkats to individual reforms in horticulture and production into individual farms. Transferred land should be farms. The number of individual farms viniculture”. maintained for the use of gardens and wine production. The decree specialized in the production of vegetables and The decree was implemented regulates the formation of agro- firms by individual farms based on melons in 2007 increased from 11,300 to through the Presidential the use of flexible, modern mini-technologies for the processing of 13,300 (17.9 percent increase). Accordingly, in resolution PP-255 “On agricultural products. 2007 the number of individual farms organizational measures of PP-255 provided concrete organizational measures to transform specializing in horticulture and grape reforming horticulture and horticultural and vinicultural shirkats into individual farms, including production increased from 51, 300 to 59,100 viniculture” (January 11, 2006). the list of shirkats to be reorganized, indication of infrastructure (increased of 15 percent). Sown area of This included also the Cabinet facilities to be established on territory of transformed shirkats, and vegetables from 2006 to 2007 in individual Ministers resolution №42 “On investment projects of processing enterprises. farms increased from 45,000 to 53,600 ha; the approval of the regulation of The Cabinet Ministers resolution №42 determined the procedure for vegetable production in individual farms the establishment and creating and organizing the activities of horticultural and vinicultural increased from 1,305,400 tons to 1,567,800 organization of activities of agri-firms by individual farms. tons. Meanwhile, sown area of vegetables in horticultural and vinicultural shirkats dropped from 6,700 to 1,500 ha; agro-firms” (March 10, 2006). vegetable production decreased from 170,400 tons to 42,700 tons. March 23, Presidential resolution PP-308 The decree aimed at enabling legal and economic conditions for From 2006 to 2007, the number individual 2006 “On support measures designed increasing the number of dekhkan and individual farms engaged in farms specializing in raising livestock to increase the number of cattle livestock production, especially cattle, and at solving unemployment production increased from 13,700 to 16,000; in or cows in households, dekhkan issues. In particular, the procedure for allocation of preferential 2007 number of cattle in individual farms and individual farms.” targeted loans by commercial banks to citizens for purchasing cattle increased on 223,000 heads (35 percent in households, dekhkan and individual farms under guarantee of increase), in 2008 – on 135,000 heads (16 local (mahalla) committees was simplified. percent increase). Other years, annual increase was on average about 69,000 cattle heads. November Cabinet of Ministers resolution The document highlighted the priority of producing fruits, State-directed promotion of vegetable, potato, 15, 2006 №238 “On the forecast vegetables, potatoes, and grapes in specialized individual farms, fruit production in individual farms along the parameters of production and paying special attention to allocation of necessary land in individual complete fragmentation of shirkats into use of fruits and vegetables, farms in accordance with forecast production parameters. individual farms. potatoes, melons and grapes in 2007.” 43 Similar resolutions were in 2007- 2014. November Presidential resolution PP-514 The resolution provided a list of agricultural enterprises to be This regulation facilitated fragmentation of 21, 2006 “On measures of transforming reorganized in 2007 into individual farms and rural production and shirtkats into individual farms. The number of enterprises into individual farms market infrastructure. individual farms increased from 189,200 to in 2007.” 217,100. This regulation finalized the fragmentation of agricultural enterprises and formation of individual farms with average size of 26 ha. October Presidential order №3077 “On The order initiated the preparations for the farm size optimization As part of farm consolidation program, the 10, 2008 formation of a Special (consolidation) program. After a careful analysis of the sizes of land number of individual farms decreased from Commission for the elaboration plots in districts of regions, the commission will determine optimal 218,600 (2008) to 103,000 (2009). By the end of proposals on measures of sizes of land plots of individual farms for each regionalized zone of 2010, the average area of allocated land to optimization of sizes of land plots individual farm increased from 27 ha to 59 ha. under the authority of individual Presidential order №3287 specified the need for further farm The number of cotton and grain farms farms." optimization by January 1, 2010 decreased from 115,200 (of an average size of 39 ha) to 49,200 (an average size of 96 ha). Presidential order №3287 “On The number of livestock farms decreased also measures of further optimization from 16,900 to 9,900. Yet, their share in total of sizes of land plots under the number of farms increased from 8 percent to authority of individual farms" almost 10 percent. (October 22, 2009). The number of individual farms specializing in vegetables and melons decreased from 12,600 to 6,000, as their average size almost doubled (an increase from 4.5 ha to 9.7 ha). October Presidential decree UP-4041 “On The decree aimed at increasing grain area in 2009 by reducing area From 2008 to 2009, sown area 20, 2008 measures of optimization of sown under cotton. The document highlighted the need for production of of cotton decreased from 1.40 million ha to areas and increase of food crops vegetable and other food crops in areas and farms with suitable 1.33 million ha. At the same time, sown areas production.” growing conditions and available of skills and experience. for grain crops increased from 1.25 million ha to 1.30 million ha. Grain production in individual farms increased from 5.1 million tons to 5.7 million tons. March 16, Cabinet of Ministers resolution The resolution provided inventory of farm optimization process, In 2010, the number of farms decreased from 2010 №47 “On the results of rationality of their use, as well as the reasonableness, objectivity, 103,000 to 71,000 (or 31 percent). Their optimization of sizes of land plots and transparency in the provision of land to individual farms, as well average size increased from 57 ha to 80 ha. 44 under the authority of individual as financial condition of individual farms, availability and efficient farms.” use of material, labor, technical resources and capital. April 18, Presidential decree UP-4301 “On The decree stipulates that all issues related to the reorganization In 2011, the number of farms decreased to the 2011 measures of compliance with and optimization of the sizes of the land plots of individual farms are minimum of 70,700 farms, and average size legality in reorganization and decided through courts on based on claims from local authorities or was 83.1 ha, while of cotton-grain growing optimization of sizes of the land other authorized bodies strictly in accordance with the procedure farms 127 ha. The decrease in number of farms plots of individual farms.” established by law. The conditions were listed as violation of land was mainly through decreased number of rental contract, farm specialization, cultivation of crops not included horticulture, viniculture and livestock farms. into contracting arrangements, systematic (three years) crop yields below norms of cadastral valuation and failure of procurement contracts etc. October Presidential decree UP-4478 “On The decree aimed at the enhancement of the role and importance of According to the decree, a "Program of 22, 2012 measures of further improving individual farms in sustainable development of agrarian sector and measures on the further improvement of the the organization of activities and improving rural welfare. regulatory and legal framework for the development of farming in development of farming aimed at Uzbekistan.” strengthening the financial and economic independence of individual farms" was elaborated. January Cabinet of Ministers resolution The resolution determined procedures of optimizing and liquidating From 2013 to 2014 throughout the country, the 31, 2013 №22 “On the approval of the individual farms. The resolution determined that optimization is not average size of individual farm decreased from regulation on the procedure of consolidation but a change of farm size aiming at more rational use 83 ha to 79.5 ha. The number of individual optimization of land plot sizes of of land and water resources and ensuring financial sustainability of farms increased from 70,800 to 73,400, i.e. the individual farm and its individual farm. Farm can be optimized and liquidated on voluntary farm consolidation program was ceased. liquidation.” basis, but also initiated by an authority. December Cabinet of Ministers resolution The resolution provides inventory indicators of farm optimization The indicated inventory shows that between 15, 2015 362 “On measures to optimize program within the main types of production specializations for each October and December 2015 the numbers of the size of land plots provided for administrative territory. Based on the results of farm optimization, farms increased by 21 percent. In 2015-2016 individual farming.” the number of individual farms and their average sizes was provided. the number of individual farms increased from The decree introduced new changes on procedures of optimization 96,000 to 132,300, while their size declined and liquidation of individual farms (with additional functions from 60 ha to 43 ha. provided to local deputy councils), as well as indicated on the establishment of a central database on economic indicators of individual farms. December Presidential resolution PP-2460 The resolution aimed at optimization of sown area under cotton and In 2016, the number of vegetable and livestock 29, 2015 “On measures for further grains on irrigated land in 2016-2020. This implied that production of farms increased by 37 percent, while the reforming and development of two strategic crops on less productive land should be released number of horticulture farms increased by 130 45 agriculture in 2016-2020. stepwise to cultivation of potato, vegetables, and fodder crops percent. The sown area of cotton in 2016 should be increased. The resolution provided targeted parameters of declined by 33,000 ha. such optimized use of land resources, as well as increased numbers of cattle and poultry. October 9, Presidential decree UP-5199 “On The decree lists a number of problems and shortcomings in the 2017 measures of radical improvement effective protection of the rights and interests of individual farms, of the system of protection of dekhkan farms and owners of household plots. In particular, the rights and legal interests of decree indicates measures for converting individual farms into multi- individual farms, dekhkan profile farms (which includes also such activities as processing farmers, and owners of storage, service provision). The decree also indicates provision of household plots with the purpose dekhkan farms with infrastructure, inputs and extension services. In of efficient use of agricultural return, the state representatives (local deputy councils) will regularly arable land.” discuss the performance of land users and decide on their land tenure. From 1 January 2022, land lease contracts will be continued only with multi- profile farms. October Presidential resolution PP-3318 The resolution highlights the main tasks and activities of Council of 10, 2017 “On organizational measures of individual farms, dekhkan farms and owners of household plots. further development of individual farms, dekhkan farmers and owners of household plots.” Note: Presidential decree – указ президента; Presidential resolution – постановление президента; Presidential order – распоряжение президента. All policy events are of national coverage. 46 Figure A: Annual changes in GDP deflation and unofficial exchange rate of UZS to USD, percent Note: Unofficial exchange rate (annual change in percent): for 1994-2012 from MacDonald (2012), for 2012=2016 from https://uzdollar.com/ Source: GDP deflator (annual percent) from WDI (2017). 47