Tapping the Potential of Bolivia’s Agriculture and Food Systems to Support Inclusive and Sustainable Growth Tapping the Potential of Bolivia’s Agriculture and Food Systems to Support Inclusive and Sustainable Growth June 2019 Agriculture Global Practice World Bank Group © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone 202-473-1000; Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations and conclusions expressed in this work do not neces- sarily reflect the views of The World Bank, its Board of Executive Directors or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. The World Bank encourages dis- semination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licences, including subsidiary rights should be addressed to World Bank Publications: The World Bank Group, 1818 H Street NW Washington DC 20433, USA; e-mail: pubrights@worldbank.org Cover image: Programa de Alianzas Rurales (PAR), EMPODERAR. Further permission required for reuse. Cover design: Kilka diseño gráfico. Contents Acknowledgments 11 Executive Summary 13 Abbreviations and Acronyms 25 1. Introduction 29 1.1. A context of dramatic change and emerging issues 30 1.2. Objectives, scope, and audience 33 2. Bolivia’s agriculture in transition 37 2.1. Bolivia’s economic performance in the regional and structural context 38 2.2. Agriculture in a transitioning economy 40 2.3. Evolution of food demands, food system modernization and nutritional challenges in Bolivia 47 2.4. Emerging challenges and opportunities 51 3. Agricultural performance and production systems 53 3.1. Natural capital to support diversified agriculture in Bolivia 54 3.2. Patterns in crop and livestock production and value 54 3.3. Trade trends 56 3.4. Characterising Bolivia’s agriculture 58 4. Productivity and efficiency of agriculture in Bolivia 67 4.1. Bringing the productivity and efficiency picture into focus 68 4.2. Productivity and technical efficiency: main concepts 69 4.3. Methodology and variables 70 4.4. Analyzing differences in technical efficiency and meta-technical efficiency by macro-region, major crops and farm size 74 4.5. Analyzing differences in technical efficiency and meta-technical efficiency by farm size and major crop 81 4.6. Conclusions and recommendations 87 5. Competitiveness of Bolivia’s agriculture 91 5.1. Competitiveness performance: Evidence provided by global benchmarks 92 5.2. Development of competitive agricultural value chains: Approaches and examples of key value chains 95 5.3. Take-away messages 106 6. A policy framework for agricultural productivity and competitiveness 111 6.1. The policy framework for agricultural investment 112 6.2. Broad patterns of public spending on agriculture 113 6.3. Agricultural sector enablers 114 6.4. Building farm and value chain capabilities 121 7. Sustaining productivity and competitiveness gains 125 7.1. Climate risks and the future of Bolivian agriculture 126 7.2. Reducing the environmental footprint of agriculture 140 8. Conclusions 147 9. References 155 Annexes Annex 1. Irrigation and input use in Bolivia 159 Annex 2. Data requirements and methodology for TFP and technical efficiency in agriculture 160 Annex 3. Main programs/projects providing direct support to producers/ communities 165 Figures Figure 1. Rural vs urban population, Bolivia 39 Figure 2. Poverty in Bolivia 40 Figure 3. Sectoral decomposition of GDP, Bolivia, 2000-2017 (%) 41 Figure 4. Agriculture’s share of total employment, Bolivia vs. regional and structural peers (2000-2018) 43 Figure 5. Agriculture transformation in Bolivia 44 Figure 6. Growth of agricultural value added, Bolivia, 2000-2017 45 Figure 7. Composition and growth of agricultural value added, Bolivia 48 Figure 8. Dietary Composition, Bolivia versus global and regional averages 49 Figure 9. Stunting and overweight rates on children under age 5 50 Figure 10. Producer prices, US$ per ton 51 Figure 11. Harvested area shares, by department. Bolivia, 1950-2013 55 Figure 12. Agriculture exports. Bolivia, 2000-2018 57 Figure 13. Agricultural trade flows, Bolivia 2000-2018 (US$ ‘000) 58 Figure 14. Main crops by macro-region, based on the number of producers. Bolivia, 2015 61 Figure 15. Distribution of UPAs by size (%) 61 Figure 16. GDP of major crops by Macro-region (Bs$/Ha). Bolivia, 2015 62 Figure 17. Yields of main crops by Macro-regions (Quintals per Ha) 63 Figure 18. GDP (Bs$/Ha) according to farm-size, by Macro-regions 64 Figure 19. Output per worker Bs/day/UPA (annual average per UPA) 64 Figure 20. Expenditure share in inputs (Percentual distribution by Macro-region) 65 Figure 21. Agriculture TFP growth, 2001-2015 (%) 68 Figure 22. Productivity and efficiency: illustration of concepts 71 Figure 23. SFA and MFE by Macro-region 76 Figure 24. Meta-frontier production function for different production systems 78 Figure 25. Meta-frontier estimations 79 Figure 26. Technical efficiency: by Crop across Macro-Regions 82 Figure 27. Average meta-technical efficiency by land quintile by macro-region 83 Figure 28. SOYBEAN: Technical efficiency 84 Figure 29. QUINOA: Technical efficiency 85 Figure 30. POTATO: Technical efficiency 86 Figure 31. RICE: Technical efficiency 87 Figure 32. Ease of Doing Business 2016-17, selected results 94 Figure 33. Requirements associated with increasingly sophisticated agri-food markets 96 Figure 34. Public investments executed by economic sector, in USD millions 79 Figure 35. Total public expenditure and agriculture expenditure, in USD millions. 80 Figure 36. Projected climate change impacts on rainfall and temperature, 114 Figure 37. Historical crop yield losses of selected crops in Bolivia (US$ million) (1995-2017). 132 Figure 38. Value at Risk (VaR) with different exceedance probabilities in all crops and departments studied. 132 Figure 39. Agricultural underwrite premiums and loss ratio (2012-2017) 136 Figure 40. Risk layering and implementation of suitable financial instruments 138 Figure 41. Bolivian legal framework on agricultural risk management 139 Figure 42. Deforestation in Bolivia 141 Tables Table 0. Approaches to enhance the effectiveness of market-oriented agricultural development in Bolivia 24 Table 1. Average annual growth rate total and per capita GDP (%): Bolivia vs Regional and Structural Peers (Period 2003-2017) 38 Table 2. Sectoral growth in Bolivia (%) 42 Table 3. Value added of renewable primary sector and its linkages, 2012 46 Table 4. Dependence of key Economic Activities on Primary Sector 47 Table 5. National Wealth Estimates, Bolivia and Lower-Middle-Income Country Average 54 Table 6. Production volume and area by crop category. Bolivia, 2000-2017. 56 Table 7. Ranking of crops in terms of their share of total agriculture value (‘000 dollars, constant 2010). Bolivia, 2000-1016. 57 Table 8. Ranking of crops in terms of their share of total agriculture value (‘000 dollars, constant 2010). Bolivia, 2000-1016 60 Table 9. Land and labor productivity in selected LA countries 69 Table 10. Sample Size: ENA 2015 72 Table 11. Descriptive statistics of the main variables used in the analysis [Average per Agricultural Production Unit-UPA] 73 Table 12. Descriptive: Determinants of inefficiency 74 Table 13. Determinants of meta-technology efficiency 74 Table 14. Elasticities and results of the stochastic frontier estimation by region 75 Table 15. Results: Determinants of technical efficiency in Macro-Regions 77 Table 16. Characteristics of producers by Macro-Region and 80 Table 17. Global Competitiveness Index 2018, selected results for Bolivia 93 Table 18. Logistics Performance Index 2018, selected results 94 Table 19. Comparison of Potato Yields in LAC and Beyond, 2017 (MT/Hectare) 98 Table 20. Current and enhanced approaches to support smallholder competitiveness 122 Table 21. Major hazards for food commodities by region and department 129 Table 22. Gross Production Value (GPV) per department for selected crops in Bolivia (in million US$). Period 1994-2017 130 Table 23. Annual Average Losses (AAL) and loss frequency per crop (1995-2017). 131 Table 24. Simulated weighted average loss cost per crop and per department. 133 Table 25. Value at Risk with an exceedance probability of 1% for different crops in Bolivia. 133 Table 26. Evolution of SAMEP program (2013-2017) 134 Table 27. Dominant drivers of tree cover loss by 2015 141 Table 28. Government roles and instruments in environmental mitigation 146 Boxes Box 0. Enhancing productive efficiency across Bolivia’s macro-regions 15 Box 1. Bolivia’s geographical regions and their role in agriculture 31 Box 2. Re-assessing the contributions of agriculture to Bolivia’s economy 46 Box 3. Secure land tenure to promote agriculture investment 116 Box 4. Anticipated effects of climate change 127 Box 5. Agricultural insurance in Bolivia 134 Box 6. Efforts to improve land use management in Colombia: definition of the agriculture frontier as a powerful tool to guide land use planning and reach multiple goals 143 Box 7. Summary of current agricultural challenges/opportunities: 149 Acknowledgments T his report was prepared by a team that included Luz Diaz (Task Team Lead- er), Pablo Valdivia, Michael Morris, Julia Navarro, Viviana Perego, Ashwini Sebastian, and Albaro Tutasig (all of the World Bank). External contributors included Maria del Mar Polo (FAO), as well as John Horton, Jorge Caballero, Mila- gros Chiappe, Gerardo Mandieta, and Sandra Lazarte (all Consultants). The team is especially grateful to Ricardo Fort and Mauricio Espinoza (Grupo de Análisis para el Desarrollo – GRADE), authors of the productivity analysis presented in Chapter 4 of this report. Steven Jaffee, Robert Townsend, Juan Gonzalo Flores, and Francisco Obreque served as peer reviewers. Preeti Ahuja, Alberto Rodriguez, Indu John-Abraham, and Oliver Braedt supported the study and ensured that resources were available for its implementation. Executive Summary A griculture and the rural space will continue to demand the attention of policy makers in Bolivia for several reasons, even as urbanization gains momen- tum. First, agriculture is a proven engine of economic growth. Aside from showing its strength in decades past, in recent years agriculture shielded the Boliv- ian economy from the worst effects of the decline in other primary sectors, and in the future, healthy rates of agricultural growth will make the overall economy more diversified and more resilient. Second, a robust and dynamic agricultural sector will continue to curb dependence on the mining and gas sectors, while contributing sig- nificantly to inclusive growth, value addition, the creation of more and better jobs on and off of the farm, and better nutrition for all. Third, because agricultural growth in Bolivia has proven to be pro-poor, maintaining that growth is essential for continued reductions in poverty. Fourth, because climate and other shocks affecting agricul- ture can significantly disrupt steady gains in economic growth, poverty reduction, and food security, building a resilient agricultural sector is critical to sustain those gains. Finally, although policy makers will want to support agricultural growth, they will not want that growth to compromise the future for generations of Bolivians by squandering and degrading irreplaceable natural resources. Pattern of agricultural growth in Bolivia Bolivia has used its diverse natural capital to grow agriculture and generate wealth. The share of agriculture in total GDP remains substantial at around 11–12 percent, although it has declined in tandem with Bolivia’s economic transition. Since 2000, the value generated by the agricultural sector has grown at a steady rate of 3 percent an- nually, with particularly dynamic growth (at about 4.9 percent) since 2013. At the same time, agriculture has contributed to Bolivia’s tremendous gains in poverty reduction. As in other countries in the region, in Bolivia the pattern of agricultural growth has been highly influenced by urbanization, income improvements, and changing diets. Agricultural value is concentrated near the largest urban centers: 72 percent of total agricultural value added is concentrated in Santa Cruz, Cochabamba, and La Paz Departments. Agricultural growth in Bolivia has largely been based on the expansion of agricultur- al area, accompanied by input intensification. A comparison of population growth in Bolivia between 1950 and 2017 with the expansion in agricultural area that occurred 13 over the same period shows that the country has add- lenges of sustaining future growth and ensuring that ed 0.34 hectares per inhabitant. Land expansion dif- the gains from that growth translate into stronger fers markedly in scope across regions. Between 1950 economic inclusion and environmental sustainabili- and 2016, total harvested area increased fivefold in ty. For example, despite significant gains in poverty Bolivia, rising from around 654,000 hectares to more reduction, the average income of the population un- than 3.5 million hectares. Most of this expansion der the poverty line—the majority of whom depend (73.6 percent) took place in the Lowlands and Ama- on agriculture—needs to increase by 75 percent to zon Regions, considerably surpassing the expansion lift them above poverty. Furthermore, the continua- of harvested area in the Highlands (14.6 percent) and tion of current patterns of agricultural growth, which Sub-Andean Region (11.8 percent). Santa Cruz De- are based on land expansion and increased use of in- partment in the Lowlands Region had 8.9 percent of puts, are not enough to sustain long-term productivity all harvested area in 1950 and 60.7 percent by 2013. gains. They are also likely to severely compromise the future provision of key ecosystem services and will A parallel and extremely important trend is the con- exacerbate climate-related challenges. tinuing division and fragmentation of agricultural land in the Highlands, Amazon, and Sub-Andean Regions, Insights into agricultural productivity where average farm size does not exceed 3 hectares. Land fragmentation is most acute in the Highlands, Sustained gains in productivity, driven by technolog- where 60 percent of all farms are now smaller than ical change, technical efficiency, and managerial ca- 1 hectare. pacity, have been modest in Bolivia, although signs of improvement are emerging. Total Factor Productivity In today’s Bolivia, approximately 45 percent of the (TFP) growth was negative during 2001–10 but was harvested area is linked to the production of indus- estimated at 2.6 percent during 2011–15. At the same trial commodities (particularly soybeans) to satisfy time, the macro-regions of Bolivia exhibit substantial local and export demand, particularly demand from differences in the kinds of production systems they the animal feed industry (due to the increasing con- can support, the characteristics of producers, and sumption of animal protein) and increasing domes- yields of the main crops. Considering these marked tic demand for biofuels. When cereals are included, differences, more granular information on produc- this figure increases to 80 percent. The expansion of tivity and levels of technical efficiency across and land for agricultural commodities is linked to large within macro-regions, crops, and farms of different scale foreign investment (mainly from Brazil) and it sizes could provide greater insight into new sources of is part of a broader trend observed in South Amer- productivity growth in Bolivian agriculture. This report ican countries. In fact, during the period 2006–16, attempts to provide such information. Although data the land expansion for soybean production in Bolivia constraints preclude the analysis of differences in TFP has been modest in comparison with the expansion growth across macro-regions, results of a stochastic experienced in countries such as Paraguay and Uru- frontier analysis and meta-frontier estimates, which guay. In the latter, about 840,000 hectares of soybean measure technical efficiency, highlight important gaps production were added during the period 2006-2016, across and within the four macro-regions (Box 0). nearly 2.5 times the expansion experienced in Bolivia. The fact that TFP cannot be estimated separately for In the face of macro-economic constraints and chang- each macro-region is unfortunate. Filling this knowl- ing trade conditions in key agricultural export mar- edge gap would help to explain the dramatically low kets, the agricultural sector faces the important chal- 14 Box 0. Enhancing productive efficiency across Bolivia’s macro-regions Results of the Stochastic Frontier Analysis (SFA) conducted for this report indicate that investments that could effectively increase technical efficiency and bring less-efficient farmers close to the regional frontier include support for collective action, credit programs, mobile phones, and irrigation. The effi- cacy of each type of investment varies by macro-region. For example, collective action is important for reducing technical inefficiency in areas where land holdings are highly fragmented (the Highlands and Sub-Andean Regions). Programs and projects that foster collective action are more likely to improve the income of peasant families than programs that target individuals. On the other hand, improving access to credit is more likely to be an effective investment in regions with better market integration and more financial institutions (the Lowlands and Sub-Andean Regions). In the Highlands Region, where access to mobile phones and landlines is still limited, better diffusion of mobile phone technology would enable information on agricultural technology and markets to reach large numbers of producers who otherwise have very sporadic and limited access to such information. Across all four macro-regions, investments to improve access to irrigation seem likely to be effective in reducing technical inefficiency. Irrigation would enable producers to control and intensify their use of water to improve productivity and reduce uncertainty in agricultural production cycles. Although the analysis of technical efficiency across the macro-regions provides a general idea of mea- sures that could improve productivity in each region, a more targeted meta-frontier efficiency (MFE) analysis focusing on particular crops and farm-size groups helps to refine those alternatives for differ- ent regions and groups of producers. For example, the analysis shows that the technical efficiency of soybean production is higher on larger farms compared to smaller ones in the Lowlands Region, prob- ably as a result of economies of scale. In the case of potato, small-scale potato producers were found to be less efficient in the Highlands Region but not in the Sub-Andean Region. In the case of rice, the MFE found no significant differences in the efficiency of rice producers in different farm-size groups, and no differences in the technology gap. The observed differences in the technology gap and dispersion in the levels of efficiency observed across crops, within and between regions, suggest a range of differentiated strategies to improve productive efficiency. In some cases, the largest benefits would come from expanding the technical production frontier through the adoption of better agricultural technologies. Higher dispersion in efficiency levels suggests a focus on bringing less-efficient producers closer to the technical production frontier through programs that increase collective action, improve production practices, and facilitate access to credit, among others. 15 level of aggregate labor productivity in Bolivian ag- Public spending on the productive sectors increased riculture reported by some sources, which proba- steadily, especially from 2010 onward, until peaking bly conceals important variations in rural livelihood at approximately US$1.4 billion in 2015. In 2016, ex- strategies across Bolivia. For example, in the High- penditures for productive sectors decreased, despite lands, farmers mainly live in nearby urban settings, a continuing increase in overall budget execution. and agricultural work is highly seasonal, with labor These investments have been supporting common migration occurring at specific periods of the year to enablers, such as key infrastructure for connectivity. supplement agricultural incomes. For example, over the 10 years from 2008 to 2018, Aside from these productivity and efficiency consid- the public investment in the primary road network erations, Bolivian agriculture faces a competitiveness grew by 82 percent, and important efforts were made challenge. The fundamental challenge to competitive- to improve trade logistics across borders. ness is that Bolivia is a landlocked country, but it is In the agricultural sector, the share of agricultural ex- amplified by constraints in institutional capacity and penditure in total expenditure fell from 9 percent to 6 in the enabling environment for overall business dy- percent between 2006 and 2008. Between 2008 and namism and investment. As discussed below, efforts 2015, a period of strong fiscal expansion, the share of have been made to address broad competitiveness agriculture in total public expenditure remained con- constraints; at the sectoral level, the focus has been stant at around 6 percent, followed by a slight decline on creating opportunities for small-scale and vulner- in 2016. Agricultural investments have supported key able farmers and communities and expanding indus- sectoral enablers such as land titling and administra- trial agricultural production. tion, plant and animal health and food safety, and to a lesser extent research, development, and innovation. Insights on the policy framework to An important set of programs emerged to support support inclusive sectoral growth expansion of irrigation and mechanization and shift The policy framework that has supported agricul- the technological frontier outward. Still other invest- tural growth since President Morales came to pow- ments were channelled through demand-driven pro- er emphasizes achieving food self-sufficiency and grams that directly supported producers and their sovereignty at the national and household levels and organizations. For the latter, the amount invested/ redressing historical imbalances and inequalities committed is broadly estimated at US$525–540 mil- resulting from previous development models. There- lion for the period between 2006 and 2022 (excluding fore, policies and programs have heavily focused on important investment programs such as the coca pro- small-scale producers and indigenous communities, gram and investments through the Fondo Indigena). especially in more marginalized and vulnerable geo- As noted, in line with overall government policy, most graphical areas with high levels of poverty. of these programs target poor and marginalized pro- ducers and communities, often in remote areas with Revenues from the extractive industries have been high levels of poverty and vulnerability. Most of the redistributed across economic and administrative support is for production, typically through one-off units, providing resources for developing productive transfers of equipment, seed, and inputs. sectors (including agriculture) and related activi- ties. Broad estimates of public expenditure between To ensure food sovereignty, the government has also 2006 and 2016 by economic sector reveal a fivefold embraced important trade and price-related policy increase in the overall size of the executed budget. interventions as a response to price variability in in- 16 ternational markets. The policy initially consisted of to enhance opportunities for creating wealth around export bans on a set of key products in the mid and domestic value chains. The size of these opportuni- late 2000s, and during the present decade it has taken ties will largely be linked to domestic consumption the form of export quotas for key products. The export and market developments. Consequently, a first and bans were active for different products at different very important source of future agricultural growth is times; for example, beef exports were prohibited over linked to growing food demand resulting from urban 2008–12, and exports of other crops have been pro- population growth, which is projected to average 1.9 hibited intermittently. The negative effects of the beef percent annually through 2030. export ban on production and the limited impact on A second source is linked to changing diets and reducing domestic prices are well documented. Since consumer preferences for differentiated and qual- 2013, the government has adopted a more flexible sys- ity products, including increased consumption of tem of export quotas, with quotas gradually increasing vegetables, fruits, legumes, and other products in as margins in the domestic market become thinner the high-value spectrum, and further development with increased supply, and as international conditions of quality-differentiated markets through continued deteriorated for some products (such as soybeans). formalization of the retail sector, development of the In the most recent policy documents, the govern- gastronomic industry, and development of demand ment’s aspirations for agriculture are reflected in sig- for fresh and locally produced food (short-value nificant proposed production increases, to be attained chains/circuitos cortos). Increasing demand for ag- through (1) an expansion of agricultural land by more ricultural commodities and food in domestic mar- than one million hectares between 2016 and 2020 kets will provide opportunities for expanding local and (2) agricultural intensification brought about by production and substituting imports—for example, the adoption of irrigation and increased fertilizer use, some fruits and vegetables, which currently account the latter specifically in commercial agriculture com- for about 3 percent of total import value. Much of the modities. Strategies also include the creation of do- food imported into Bolivia today consists of processed mestic demand to drive the expansion in agricultural products, however, and it will be challenging for do- land (specifically, by stimulating demand for biodiesel mestic producers and processors to gain a share of crops through an ethanol mandate requiring gasoline the market for these products in the future. Signifi- to be blended with at least 10 percent ethanol begin- cant investments by the private sector will be needed ning in 2018 and at least 25 percent by 2025). An in- to build local processing capacity and overall value teresting development is the aspirations surrounding chain developments to compete with imports. the diversification and expansion of agriculture and A third emerging opportunity is linked to public pol- food exports. icies that create domestic demand in non-food sec- tors, such as demand for biofuels, which is expected Insights into opportunities and to reach 380 million liters in 2025. To fulfill this de- effective strategies for inclusive and mand, the area under biofuel crops (mainly sugarcane sustainable agricultural growth in and soybeans) is expected to reach 400,000 hectares. Bolivia As a result, the government estimates an increase in Agricultural growth opportunities in Bolivia—Given the agricultural GDP of about 3.9 percent and of total GDP large number of producers involved in supplying do- mestic food markets, a continued priority in Bolivia is 17 in about 0.2 percent, with different estimates on the In the future, a successful sustainable and inclusive number of jobs generated.1 agricultural growth strategy in Bolivia would imply targeted strategies to ensure benefits to farmers op- A fourth opportunity will come from enhancing the erating at the subsistence and commercial levels, and performance of current non-traditional export value to a set of rural and urban actors, more broadly, via chains, such as cocoa, beans, coffee, quinoa, brazil job creation opportunities. nuts, and beef, and over the longer term, from cre- ating opportunities for new export products (such as Ensuring continued gains for subsistence produc- Amazonian products). ers who are not integrated into commercial agri- culture will require productivity improvements, Looking to the future, what is our understanding of the cost-efficiencies, and the inclusion of nutrition and successful pursuit of opportunities for sustained, in- climate-smart approaches into the development of clusive agricultural growth in Bolivia? The approaches production systems. Interventions at this level would to agricultural development implemented in the past benefit from close monitoring of developments in few decades and the patterns of development provide production to support (when feasible) a graduated good lessons to inform future strategies. Support to shift from subsistence to market-oriented produc- agricultural growth has largely been based on two tion. These types of producers in particular will growth models. One model has focused on large- benefit from a combination of strategies to create scale expansion of industrial, commercial agriculture; income opportunities within (on-farm/off-farm) and incentives for private investment have taken the form outside the agricultural sector. of preferential trade arrangements with targeted re- gional markets, policies to facilitate access to land Within the context of market-oriented production, the and imported agricultural inputs (including genetical- further development and consolidation of opportuni- ly modified seed), and policies that have restricted ties for inclusive and sustainable agricultural growth exports and imports of particular products to ensure in Bolivia will require a set of targeted strategies to national self-sufficiency. The other model has focused solidify opportunities in a range of markets (Table 0). on small-scale agriculture, supporting production In the most traditional production sectors (such as improvements among mostly poor and vulnerable potato production), as well as in diversification into farmers operating at a self-sufficient level or at some high-value products for domestic markets, success limited level of market integration. will depend greatly on developing competitive advan- tages around cost reductions, quality and productiv- In both instances, drawbacks can be identified. The ity improvements, efficiencies along the value chain, industrial agricultural expansion model has relied on and market integration, accompanied by improved land expansion and given little attention to securing capacity to manage risks (price fluctuations, climate gains in long-term productivity, inclusiveness, and shocks, pest and disease outbreaks, and so on). These environmental sustainability. With respect to small- developments have potential to be largely inclusive scale agriculture, the approaches have not been and to generate gains for a sizeable number of rural sufficiently ambitious in pursuing opportunities for producers and laborers in production and, particular- market integration, value addition, and therefore for ly on value-added activities. creating quality jobs. 1 Ranging from 27,000 jobs created by the additional expansion of sugarcane production on 150,000 hectares to 100,000 jobs generated from the entire area expansion of sugarcane and soybean production. 18 In industrial commodity sectors serving domestic and/ sustainable intensification and technology adoption, or export markets (such as soybeans, sugar, sorghum, and (2) adding value by developing post-farm stages and other industrial commodities), success will de- of selected value chains. An expansion of agricultural pend on productivity improvements and cost-reduc- area will be part of the agricultural growth equation, tions (for large-scale and small-scale producers), an- but it must be anchored in proper land-use planning chored in significant improvements in environmental and management, including environmental and social and social sustainability and risk management strat- considerations. egies. The potential to generate jobs in these subsec- tors is not necessarily linked to primary production Priority actions to grow an inclusive activities but to associated activities such as logistics, and sustainable agricultural sector in storage, distribution, and domestic processing.2 Bolivia In the case of non-traditional agricultural exports, a The value chains selected for analysis in this report cost-reduction strategy will help to increase Bolivia’s span a range of subsectors, from traditional ones (po- participation in export markets that are not highly tatoes) to those oriented to target domestic/export sensitive to quality (for example, some market desti- markets (soybeans, coffee, quinoa, beef). The anal- nations for beef, although sanitary compliance would yses identify a broad range of targeted on-farm and be critical), but more integrated value chain strategies off-farm investments and services that can enhance will be needed to create and enhance opportunities the performance of each value chain in response to for products differentiated on the basis of quali- market opportunities. These investments include ty or some other trait, taking advantage of Bolivia’s improved varieties and inputs; the adoption of bet- outstanding agro-diversity. Consolidating gains in ter production and management practices; improved non-traditional export sectors has the potential to be equipment for production and processing; irrigation inclusive, benefiting large numbers of small- and me- investments; market promotion strategies; and co- dium-scale producers and other actors downstream ordination platforms, among others. Associated with in these value chains, through value-adding activities. those investments is an enabling policy and regu- latory environment and institutional capacities to The use of a combination of strategies to pursue op- create opportunities for sustained long-term agri- portunities in a range of markets will bring higher re- cultural growth. The paragraphs that follow discuss turns, but it also has several important requirements. the priority areas identified in this report that require It requires an understanding of market opportunities, strong attention to consolidate future opportunities an effort to prioritize and focus investments along for inclusive and sustainable growth across a range selected value chains to increase efficiencies and of markets. competitiveness, the proper management of risks, and steps to avoid negative social and environmental i) Supporting innovation and technology impacts. adoption Looking across all subsectors, undoubtedly the great- Agricultural growth in Bolivia has mainly come from est opportunities for agricultural growth in Bolivia area expansion and, to some extent, input intensifica- will come from: (1) using land more efficiently via tion. In the future, Bolivia will need to ensure that in- 2 Industry sources estimate that the oilseed complex contributes between 300,000 and 400,000 jobs, covering activi- ties related to the provision of seeds, other input providers, transporters, processing industries, and so on. 19 cremental advances in productivity are achieved over programs aligned with clients’ needs through pub- time, and that will be possible only through the use of lic-private partnerships, and strong international innovations and technology that leads to higher levels R&D linkages. The gains achieved by the government of efficiency in the use of production inputs. to date with multilateral and bilateral support are at risk, owing to the absence of sustainable funding and As the analysis of productivity efficiency illustrates, strategies to retain key scientists and staff. efficiency gains will come from pushing the techno- logical frontier outward, particularly in the Sub-An- In some subsectors (industrial commercial agricul- dean and Lowlands Regions, and from increasing ture), the private sector is called to take the lead in producers’ knowledge and strengthening their ca- bringing innovations and R&D initiatives, while the pacity to move them closer to the current production government can concentrate its efforts on subsec- frontier possibilities, particularly in the Highlands tors with less potential for private sector solutions. and Amazon Regions.3 To achieve this result, it will The role of the private sector in mainstreaming soft be necessary to enhance public and private capabil- and hard innovations in Bolivia’s agricultural sector ities for innovation—both “hard” and “soft” capabili- cannot be overstated, but at the same time the private ties—and for innovation and technology adoption and sector must play a more pivotal role as a provider of transfer. “Hard” innovation capabilities are related to new sustainable technologies and in developing in- research and development, and they include a range clusive business models for technology dissemina- of technological opportunities linked to production, tion and adoption. post-harvest handling, and processing, as well as Several tools are in the hands of public entities in opportunities for increasing resource and input use Bolivia to stimulate private sector innovation and efficiency to protect the natural capital for agriculture technology adoption and transfer. Examples include (land, water, and biodiversity). Examples of “soft” in- establishing public-private partnerships; facilitat- novation capabilities include the generation of new ing technology transfer (for example for genetically inclusive business models (such as value-chain fi- modified seed via the development of clear regula- nance and product differentiation, certification, and tory guidelines); expanding the market share of the traceability); strengthening public-private dialogue private sector for agricultural inputs by reducing re- and collaboration; capabilities for “sanitary diploma- strictions on participation, encouraging competition, cy”; and finding innovative and effective ways and in- and leveling the playing field; and creating incentives struments to professionalize producer organizations. for the development of inclusive business models in Bolivia’s agricultural innovation system requires ad- large-scale commercial agriculture. equate funding and the alignment of policies over the medium and long term to crowd-in private invest- ii) Value chain coordination, logistics, ment, incentivize knowledge sharing, and enhance and connectivity the capacities of service providers. The key elements An important subset of government programs directly of a successful and effective innovation system are targets poor producers with various types of support. still lacking in Bolivia: institutional autonomy, sta- Evaluations of some of these interventions/programs ble and diversified financing, incentive structures have demonstrated their success in improving pro- that reward the performance of staff and scientists, 3 Results from the analysis of crops within and across regions suggest a combination of strategies to push the techno- logical frontier outward, including knowledge dissemination and technology adoption. 20 duction, yield, and household income. As this report idating market opportunities is to enhance logistics shows, the effectiveness of these programs can be and connectivity. For Bolivia, strategic investments increased further, particularly as the government in ports, roads, and similar infrastructure are vital intensifies efforts to develop more market-oriented to competitiveness. Those investments will have to production. For example, there are opportunities to be accompanied by reductions in other post-harvest expand coordination vertically, along the full extent logistics costs, particularly those linked to trade pro- of the value chain, and horizontally, through produc- cedures, export controls, and sanitary and phytosan- er associations and producer networks. To reinforce itary inspections and certifications. coordination in value chains, it is a priority to adopt At the grassroots level, secondary and tertiary rural strategies that support and/or enhance the role of roads are integral to competitiveness for producers, farmers and other value chain actors and private but building and maintaining roads tends to be very players, such as public-private coordinating plat- expensive. There is a need to ensure that scarce pub- forms or innovation platforms that convene a broad lic resources used for this purpose are directed to set of institutions. Furthermore, efforts to support areas where they offer the potential to stimulate agri- collective action by producers would benefit from cultural growth. Along with roads, other off-farm mar- systematic knowledge sharing and initiatives that go ket infrastructure and logistics systems are needed to beyond individual producer organizations to develop match demand and supply, ensure that products meet networks of producer organizations that can be effec- standards for quality and safety, and reduce post-har- tive at the territorial/local level. vest losses. This enabling infrastructure is particular- Programs providing direct support to producers in ly critical as the country moves toward greater crop Bolivia also represent tremendous opportunities to diversification involving more perishable and quali- integrate efforts to build financial literacy, pilot finan- ty-sensitive products. Exploring partnerships with the cial innovations, mainstream climate-smart practic- private sector to build and maintain key productive es, and pilot approaches that incorporate nutritional and market infrastructure will be critical. considerations. iii) Bringing financial services close to Enhanced coordination among public agencies rep- farmers resents an important avenue for increasing the effec- tiveness of public support to agriculture. It requires The productivity analysis indicates that access to fi- policy makers and program administrators to gain a nance/credit will help to ensure productivity in the better understanding of the “big picture” and pursue Lowlands and Sub-Andean Regions and could grad- complementarities across initiatives. Initiatives to ually improve opportunities in the Highlands and Am- promote the development of small-scale agriculture azon Regions. Government regulatory interventions sometimes end up operating in isolation, oblivious to broaden access to finance for a larger number of to other active programs and projects that offer po- farmers are partially undermining credit growth dy- tential synergies. Another critical area requiring at- namics among medium and small insurance compa- tention from the government is to set up robust and nies, however. Mechanisms that facilitate the estab- integrated monitoring systems for these co-financing lishment of credit histories, support the development investments. of flexible collateral arrangements, and account for seasonality in repayment schedules, among others, Along with addressing coordination issues for value can be effective ways of tailoring financial services to chains and institutions, a critical step toward consol- 21 farmers’ needs. The many programs providing direct Reducing the environmental footprint of agriculture— support to producers in Bolivia represent tremendous In Bolivia, agricultural activities have increased water opportunities to integrate efforts to build financial lit- pollution, competition for water, and soil degradation, eracy and pilot financial innovations. undermining biodiversity, threatening a range of so- cial and economic interests, and contributing to cli- iv) Effective management of risks, mate change (through greenhouse gas emissions). including environmental and social These impacts are mainly the result of the expanding risks agricultural frontier and the intensification of agricul- tural activities. Bolivia should be commended for making an effort to develop risk transfer mechanisms to help farm- As illustrated in this report, there are important op- ers manage agricultural risks (insurance schemes). portunities to strengthen the government’s role as However, there are important opportunities to make a definer, enabler, funder, and regulator of environ- current risk management strategies more effective mental mitigation in agricultural investments. In this and to expand risk management capabilities around regard, particularly relevant steps for government integrated risk management approaches. The anal- include the alignment of growth and environmental ysis in this report finds that the agricultural insur- policies and the improvement of national, regional, ance market will not expand rapidly under the cur- and local capacity for sustainable land-use plan- rent legal framework for financial services, but other ning and management, such as participatory local financial instruments and budgeting mechanisms processes and the development of decision-support could be combined to address the government´s systems and incentives that facilitate the design of need for funds to compensate producers for losses climate-smart agricultural landscapes. triggered by different types of hazards. To this end, However, approaching social and environmental a risk financing strategy in Bolivia should be struc- challenges merely from the perspective of the gov- tured around an understanding of the probability of ernment’s roles—as a definer and enforcer of labor occurrence and the severity of hazards, and it should and environmental legislation, as a driver of land-use combine ex-ante instruments and ex-post measures planning processes at the territorial level, or as an to smooth the costs of risks and minimize cost fluc- enabler of improved environmental performance— tuations. as necessary as it may be—is still insufficient. The Along with risk transfer mechanisms, there are op- collective commitment of the private sector is also portunities to continue strengthening the capacity of required to improve the social and environment foot- INIAF and SNIAF (including public and private actors) print of agriculture (for example, by adopting and to develop, validate, and promote technologies and complying with agreed environmental and social practices that reduce agricultural risk. Continued in- principles, particularly in subsectors with a large en- vestment in irrigation and water management will be vironmental and/or social footprint such as industri- critical for increasing productivity, reducing vulnera- al commodity production, cattle ranching, and quinoa bility, and maintaining productivity gains. Improved production). water-use efficiency (and energy efficiency) and This report identifies at least four major opportunities watershed conservation and protection are require- that can prevent future agricultural expansion from ments for any such investments to be sustainable in repeating the mistakes of the past. First, Bolivia’s the long term. public and private sector can learn much from the 22 well-documented negative effects of unsustainable program evaluation; recognizing food emergencies commercial agricultural development in faster-grow- and agriculture and food security risks; and using all ing nations of Latin America and the Caribbean and of these capacities for effective policy action and pro- other regions, as well as from the emerging ap- gram implementation. Policy makers require effective proaches to reverse those effects. Equipped with that assessments of the benefits and tradeoffs of specific information, Bolivia’s public and private actors can policies (related to export and input markets, price in- make systemic changes to move agricultural indus- terventions, and the roles of state-owned enterprises tries onto a more sustainable growth path. Second, in agriculture, for instance) as well as a broad under- some farmer organizations are already leveraging standing of the evidence on policies and strategies environmentally sustainable practices that serve as that are likely to be effective in achieving agricultural proof of concept to bring those practices to scale. growth, competitiveness, and food security objectives. Third, Bolivian farmers’ ancestral knowledge of bal- Better data collection is a key requirement for better ancing nature and agriculture should be supported evidence to support better policy and planning. For and more broadly disseminated. Fourth, environmen- example, enhanced data collection is a requirement tal organizations and financial institutions have pro- to monitor gains in agricultural productivity at a more gressed substantially on sustainability, becoming vast informative scale (from the macro-region to the de- sources of knowledge and best practices from which partmental and lower levels) and develop better es- Bolivia can certainly benefit. timates of TFP. v) Strengthening public institutions vi) Public-private sector collaboration Efforts to address challenges and opportunities to A key condition for achieving sustainable and inclu- ensure an inclusive and sustainable agricultural sive agricultural growth is to make collaboration growth pattern could pay enormous dividends, but between the public and private sectors the basis for implementation will be demanding. Success in re- continued progress. Agriculture is fundamentally a sponding to Bolivia’s emerging agricultural oppor- private activity. The largest investments for support- tunities and challenges will require effective policy, ing growth of the agricultural sector into the future aligned with institutional capacity. During the last two will come from the private sector (including large, decades, Bolivia has established new institutions and medium, and small-scale producers and processors), strengthened existing ones to improve the delivery although public spending will also be essential. The of critical services to the agricultural and food sec- coordination of public and private efforts opens im- tor. These efforts must continue and even intensify portant avenues for promoting innovation and piloting as the agricultural and food sector and risks evolve comprehensive approaches that enable the conver- and opportunities emerge domestically and abroad. gence of sustainable and inclusive growth objectives. Maintaining continuity in technical staff and limiting This collaboration should be institutionalized in such political interference will be centrally important to a way that it becomes the foundation for advancing these efforts. the creation of prosperity and wealth in rural Bolivia. Aside from strengthening the capacity to provide key services to the sector, it is essential to strengthen the capacity for evidence-based policy making and investment planning; monitoring, policy analysis, and 23 Table 0. Approaches to enhance the effectiveness of market-oriented agricultural development in Bolivia Market-Oriented Production Market Levels 1-3 Level 2-3 Levels 4-6 opportunities (Products serving domestic markets Industrial commodity production for Non-traditional exports (beef, cocoa, low & high value) domestic food, feed and biodiesel coffee, brazil nuts, quinoa, etc.) uses Current development pathways Expand land frontier Expand land frontier Expand land frontier Intensify input use Intensify input use Intensify input use Strategic focus Import/export restrictions in some Import/export restrictions in some Import/export restrictions in some commodities commodities commodities Price interventions Price interventions Price interventions Smallholder High Low High inclusion Job opportunities under current Low Medium Low-Medium approaches Short-term productivity gains Short-term productivity gains Short-term productivity gains Little incentives for innovation & Little incentives for innovation & Little incentives for innovation & technological change technological change technological change Tradeoffs Land & water deterioration Land grabbing Land & water deterioration Marginalization of small-scale producers Deforestation (associated mainly Large deforestation and ecosystem Deforestation (associated mainly with with cattle ranching) deterioration cattle ranching) Opportunities to enhance effectiveness Sectoral strategies Increase productivity via efficiency Increase productivity via efficiency Increase productivity via efficiency gains (technological adoption; Productivity gains (technological adoption & gains (technological adoption & technology transferring mainly for innovations) innovations) small-scale producers) Quality differentiation/cost Competitiveness Market-orientation Improved logistics/cost-efficiency efficiencies Value chain (VC) prioritization and Value chain (VC) prioritization and Approaches Landscape approaches development (territorial focus) development (territorial focus) Job creation opportunities via Job creation opportunities via investments upstream/downstream investments upstream/downstream the value chain the value chain Support to farmer collective action Support to farmer collective action Support to farmer collective action (small-scale producers) Seek opportunities for inclusion of Targeted market promotion & sanitary nutrition considerations diplomacy Seek opportunities for linking VC Seek opportunities for linking VC development with broader local development with broader economic economic development development Adoption of principles of responsible investments Knowledge and technology Inclusive business models Sustainability transferring for climate and Knowledge and technology Enhanced landscape planning and sustainable technologies transferring for climate and management sustainable technologies Risk management strategies Transversal: critical common & sectoral enablers Effective public and private sector innovation and technology transferring system Key productivity infrastructure Enabling trade policies Coordination, information and systematization of learnings Public-private sector collaboration 24 Abbreviations and Acronyms ASA Advisory Services and Analytics APS Insurance and Occupational Pension Authority (Autoridad de Fiscalización y Control de Pensiones y Seguros) BHAG Crop Water Balance Model (Balance Hídrico Agrícola) CAN Andean Community of Nations (Comunidad Andina de Naciones) CIAT International Center for Tropical Agriculture (Centro Internacional de Agricultura Tropical) CADENA Committee for the Attention of Natural Disasters and National Emergencies (Comité de Ayuda a Desastres y Emergencias Nacionales) CAF Andean Development Bank (Corporación Andina de Fomento) CAIPJ Domestic Internal Supply and Fair Price Certificate (Certificados de Abastecimiento Interno a Precio Justo) CIP International Potato Center (Centro Internacional de la Papa) CRI Climate Risk Index CRIAR Creation of Agrifood Initiatives (Creación de Iniciativas Agroalimentarias Rurales) DBI Doing Business Indicators DS Supreme Decree (Decreto Supremo) EBA Enabling the Business of Agriculture ECLAC Economic Commission for Latin America and the Caribbean ENA National Agricultural Survey (Encuesta Nacional Agropecuaria) EMAPA State Agency for Agricultural Marketing (Empresa de Apoyo a la Producción de Alimento) ENDAR National Strategy for Agricultural and Rural Development (Estrategia Nacional de Desarrollo Agropecuario y Rural) ENSO El Niño Southern Oscillation FAO Food and Agriculture Organization of the United Nations FSL Financial Services Law FDI foreign direct investment FOB free on board FONADIN National Fund of Integral Development (Fondo Nacional de Desarrollo Integral) GCI Global Competitiveness Index GDP Gross Domestic Product GHG Green House Gas GPV Gross Production Value 25 ICT Information and Communications Technology INE National Institute of Statistics (Instituto Nacional de Estadística) IMF International Monetary Fund IDB Inter-American Development Bank INIAF National Institute for Innovation in Agriculture and Forestry (Instituto Nacional de Innovación y Agropecuaria Forestal) INRA National Agrarian Reform Institute (Instituto Nacional de Reforma Agraria) INSA National Agricultural Insurance Institute (Instituto del Seguro Agrario) LAC Latin America and Caribbean LPI Logistic Performance Index MDPyEP Ministry of Production Development and the Plural Economy (Ministerio de Desarrollo Productivo y Economía Plural) MDRyT Ministry of Rural Development and Lands (Ministerio de Desarrollo Rural y Tierras) MFE Meta Frontier Efficiency MMAyA Ministry of Environment and Water (Ministerio de Medio Ambiente y Agua) MT metric ton NDP National Development Plan NGO Non-Governmental Organization OTCA Amazon Cooperation Treaty Organization (Organización del Tratado de Cooperación Amazónica) PAHO Pan American Health Organization PAR Rural Alliances Project (Proyecto de Alianzas Rurales) PASARDI Integrated Development Plan for the Agriculture and Rural Sector (Plan del Sector Agropecuario y Rural con Desarrollo Integral) PC Productive Complex (Complejo Productivo) PDES National Economic and Social Development Plan (Plan de Desarrollo Economico y Social) PISA Agricultural Innovation and Services Project (Proyecto de Innovación y Servicios Agrícolas) PLUS Land Management Plan (Plan de Uso de Suelo) PO Producer Organization POP Aligned Farm Management Plans (Plan de Ordenamiento Predial) PPP Public-Private Partnership PRONAREC National Irrigation Program with a Watershed Approach (Programa Nacional de Riego con Enfoque de Cuencas) ROE Return of Expenses 26 SAMEP Agricultural Insurance for Municipalities with High Levels of Extreme Poverty (Seguro Agrario para Municipios con mayores niveles de Extrema Pobreza) SAT Early Warning System for the Agricultural Sector (Sistema de Alerta Temprana) SENASAG National Service of Agricultural Health and Food Safety (Servicio Nacional de Sanidad Agropecuaria e Inocuidad Alimentaria) SENAMHI National Service of Meteorology and Hydrology (Servicio Nacional de Meteorología e Hidrología) SFA Stochastic Frontier Analysis SIRIC Subprogram of Investments in Inter Communal Irrigation (Subprograma de Inversiones en Riego Inter Comunal) SNIAF National Agricultural and Forestry Innovation System (Sistema Nacional de Innovación Agropecuaria y Forestal) SPF Stochastic Production Frontier TFP Total Factor Productivity TGN General Treasury of the Nation (Tesoro General de la Nación) UIPTLYM Unit for Productive Infrastructure, Local Technology and Mechanization (Unidad de Infraestructura Productiva, Tecnología Local y Mecanización) UPRA Agricultural & Rural Planning Unit (Unidad de Planificación Rural Agropecuaria) USAID United States Agency for International Development USDA United States Department of Agriculture UPA Agricultural Production Unit (Unidad de Producción Agrícola) VA value added VRHR Vice Ministry for Water Resources and Irrigation (Vice Ministerio de Recursos Hídricos y Riego) WDI World Development Indicators WRI World Resources Institute WTO World Trade Organization 27 1. Introduction Photo: Shutterstock 1.1. A context of dramatic change and food spending patterns) and the supply side (evolv- emerging issues ing agroclimatic factors, technological innovation). The dynamics of agricultural transformation and Over the past half century, agricultural production modernization have varied across Bolivia’s diverse systems have undergone sweeping change in virtu- geography, giving rise to today’s highly diversified ally every region of Bolivia. To cite just one example, agricultural sector. The sector is extremely hetero- during the 1950s in the Central Valley of Tarija De- geneous, not only in terms of structure but in terms partment, almost two-thirds of cultivated area was of performance and links to socio-cultural traditions occupied by maize, wheat, and potatoes, followed by (Box 1). barley, peaches, and peas. The push toward modern- ization in the late 1970s and early 1980s led many The transformation of the agricultural sector has oc- Central Valley farmers to move into more commer- curred in tandem with the transformation of the over- cially oriented production. By the mid-1980s, pro- all economy. As GDP per capita has risen in Bolivia, duction of many formerly minor crops had expanded, the contribution of primary agriculture to national including grapes, market vegetables, and animal feed GDP has declined in percentage terms, falling from crops such as alfalfa and oats. Maize, wheat, and po- 15.7 percent in the late 1980s to 12.2 percent in 2019. tatoes were still the most common crops, but they This decline has proceeded at a much slower pace occupied only 26 percent of cultivated area.4 Subse- than in other countries in the region and has virtually quent decades saw high levels of commodity special- ceased since 2014, following contraction of the min- ization as increasing numbers of producers took up ing/gas sector.5 commercial dairy production and horticultural crops, including grapes to supply the growing wine indus- The relevance of agriculture in a rapidly try. Not all farmers followed suit. Today in the Central urbanizing Bolivia Valley traditional subsistence agriculture persists Investing in cities is an emerging priority for policy alongside modern commercial agriculture. makers in Bolivia as urbanization continues to gain Other shifts in agricultural production patterns have momentum. A large share of Bolivia’s poor live in the occurred throughout Bolivia, but the most striking areas surrounding its three largest cities. Evidence changes occurred in the 1980s and early 1990s in also clearly shows that the economic prosperity of the lowlands. In the lowlands, particularly in eastern cities and metropolitan areas benefits small towns Bolivia, commercial agriculture expanded dramatical- and could certainly influence prosperity in nearby ly, largely based on the production of oilseeds (along rural settings. with some food crops such as beans) and commercial livestock. These shifts have been driven by multiple forces, both on the demand side (income growth, migration, ur- banization, changing dietary preferences, changing 4 Turner, K., Davidson-Hunt, I. & Desmarais, A., (2017). Agrobiodiversity, Rural Transformations and Household Experiences of Globalized Change: A Case Study from Southern Bolivia. Rural Landscapes: Society, Environment, History. 4(1), p.1. 5 The higher relative growth of other sectors, in particular, the extractives and non-tradeable sectors, and their recent contraction can in part explain these trends. 30 Box 1. Bolivia’s geographical regions and their role in agriculture Bolivia has four geographical macro-regions spread across the 1,098,580 square kilometers within its borders: the Andes and western arid highlands (Highlands Region); the semi-tropical valleys in the mid- dle third of the country (Sub-Andean Region); the eastern tropical lowlands (Lowlands Region); and the northern tropical forest, with its characteristic evergreens and rivers (Amazon Region). Highlands (Altiplano) Region. Located between the two major Andean mountain ranges—the Cordillera Occidental and the Cordillera Oriental—the altiplano (“high plateau”) covers about 28 percent of the national territory, with a large area of La Paz, Oruro, and Potosi Departments located higher than 3,000 meters above sea level (masl). Nearly 40 percent of the country’s population lives in this region, mainly around the city of La Paz. The Altiplano is arid in the south but served by Lake Titicaca in the north. Small- scale farming predominates, with a focus on raising llamas and sheep and cultivating suitable crops such as quinoa, potatoes, and maize. Farmers engage in subsistence production and some commercial production for domestic markets, with the exception of quinoa, which became a popular export product during the present decade. Although land is abundant, this region is arid, and water scarcity remains an important natural constraint on increasing agricultural productivity and expanding production. 31 Sub-Andean Region (Región Sub-Andina). The lower eastern slopes of the Cordillera Oriental, known as the Yungas, comprise the semi-tropical Sub-Andean Region, which covers 15 percent of the national territory. It has a temperate climate, many valleys (Cochabamba, Chuquisaca, and Tarija Departments), and its numerous rivers drain into the Amazon Basin. The diversity of microclimates in the valleys make them very suitable for growing less extensive crops such as fruits, vegetables, and coffee. Tarija is well known for grape and wine production. The Bolivian valleys also face natural challenges such as the er- ratic distribution of rainfall throughout the year. Lowlands Region (Llanos). In this region (spanning Pando, Beni, and Santa Cruz Departments), the north- eastern plains cover more than half of the national territory, with tropical grasslands that are well suited to extensive livestock production. The tropical lowlands in the southeast in Santa Cruz Department have seen a major expansion of agriculture. Typical crops include sugarcane, soybeans, and maize. The Gran Chaco province in Tarija Department is sometimes considered a separate geographical region; otherwise it is included within this region. Amazon Region (Región Amazonica). This region is characterized by tropical forest in the north (Pando, northern Beni, and La Paz Departments). Typical crops include yucca (cassava), sugarcane, rice, and maize (rubber was a major crop in the past). Demographic density is low compared to the other regions. Throughout Bolivia, diversity is an underused resource. Bolivia’s geographic diversity translates into significant agricultural biodiversity, and in fact Bolivia (particularly the Highlands) is the center of origin of a wide range of crop species. Much of this biodiversity has been underused or neglected. Agriculture and the farming population are concentrated in different areas. Because the lowlands possess most of the land that is easily used for agricultural activities, this region has experienced the most widespread conversion of land to agricultural uses. People have continuously migrated from oth- er regions and countries to Santa Cruz in search of economic opportunities and inexpensive land. The scope of this transformation is apparent from census data: 9 percent of Bolivia’s cropped area was in Santa Cruz Department in the 1950s; by 2013, that share had risen to 60 percent. By contrast, the farm- ing population is concentrated in the Highlands (La Paz, Oruro, and Potosi Departments), where nearly 50 percent of the population above the age of 8 formed part of an agricultural production unit in 2013, versus 13.5 percent in Santa Cruz. Economic activity in these four geographical regions is anchored around Bolivia’s three largest cit- ies. La Paz, Cochabamba, and Santa Cruz together account for 70 percent of national GDP and almost two-thirds of the national population. Agriculture in these same departments delivers approximately 72 percent of agricultural GDP, and Santa Cruz generates 50 percent of the sector’s GDP (2019). 32 Agriculture and the rural space will continue to com- one example) and unwanted impacts (such as the loss mand the attention of policy makers in Bolivia for of biodiversity). several reasons, even as urbanization gains momen- tum. First, agriculture is a proven source of economic 1.2. Objectives, scope, and audience growth. In recent years agriculture has compensated As Bolivia continues its transition to an increasingly for downturns in other primary sectors and continued urbanized economy, ensuring balanced urban/rural healthy rates of growth in agriculture will increase development is crucial to achieving an inclusive and the overall resilience of the Bolivian economy. Sec- sustainable growth pattern. The creation of oppor- ond, a diversified agricultural sector has the advan- tunities to support wealth and inclusion in the rural tages of further reducing dependency on the mining space, including opportunities in the agricultural sec- and gas sectors, contributing significantly to inclu- tor—the anchor of rural activity—will remain critical sive growth, adding value, improving the number and to Bolivia’s development agenda in the years to come. quality of jobs on and off of the farm, and address- ing nutritional challenges. Third, agricultural growth In aggregate terms the agricultural sector in Bolivia in Bolivia has proven to be pro-poor, and if it slows, has experienced rapid growth, particularly in the last an important means of reducing poverty will be lost. decade. The two important challenges for the sector Fourth, because climate and other shocks affecting are to (1) sustain growth into the future and (2) en- agriculture can significantly disrupt economic growth sure that the gains from agricultural growth translate and sustained poverty reduction, in addition to jeop- into stronger economic inclusion and environmental ardizing food security, building a resilient agricultur- sustainability. For example, despite significant gains al sector is critical to sustain gains on those fronts. in poverty reduction, the average income of the pop- Finally, although policy makers will want to support ulation under the poverty line needs to increase by agricultural growth, they will not want that growth 75 percent to lift them above poverty. Furthermore, to come at the expense of the nation’s irreplaceable the continuation of current patterns of agriculture natural resources and compromise the future of gen- growth, based on the expansion of agricultural land erations of Bolivians. and increased use of inputs, are likely to severely compromise the future provision of key ecosystem Efforts to address these issues will be rewarding but services and exacerbate climate-related challenges demanding. In the years to come, agricultural trans- if not properly guided. A better understanding of re- formation and modernization in Bolivia are expected cent sectoral performance trends and of the drivers to accelerate, creating new opportunities for farm- and disruptors of growth in the agriculture and food ers, firms, and overall sectoral growth. Those oppor- system is helpful to distil lessons and provide direc- tunities are accompanied by a range of challenges, tions for the future. including the heterogeneity of Bolivia’s geography and agroclimatic conditions, constraints to growth in With this background in mind, this report aims at productivity and competitiveness, and the need for re- analyzing the ongoing transformation and perfor- silience in a changing climate. Bolivia’s response to mance of Bolivia’s agriculture and food system and emerging agricultural opportunities and challenges highlighting opportunities and avenues for enhancing will require effective policy, aligned with institutional the sector’s productivity, competitiveness, and resil- capacity, targeted investments by the public and pri- ience. Taking into account relevant global and domes- vate sector, and the implementation of strategies for tic developments, drivers, and disruptors, the report successfully managing emerging risks (climate is just examines the constraints to, and transformational 33 opportunities for, the Bolivian agricultural sector and duction systems in Bolivia’s four macro-regions. the broader food system, looking at their potential Chapter 4 approaches the issues of productivity contribution to stronger economic inclusion and en- and efficiency through econometric analysis—Sto- vironmental sustainability. chastic Frontier Analysis (SFA) and Meta Frontier Estimates (MFE). This chapter identifies the main This report makes use of the available literature, determinants of technical efficiency in Bolivia’s econometric exercises, the analysis and synthesis of four macro-regions and highlights important dif- pertinent data, expert assessments, structured case ferences in efficiency between regions, crops, and studies, and examples of good practice that have groups of producers. These analyses, although emerged in Bolivia and elsewhere to draw lessons highly data-driven, are relevant for understand- for addressing the opportunities, challenges, and ing the current status of productivity efficiency at risks associated with the transformation of agricul- the regional and crop level and for identifying op- ture and food systems. The analyses and information portunities to address performance gaps. These presented here also adopt a territorial perspective in analyses are also highly relevant for informing assessing spatial features that contribute to (or de- policy makers of knowledge and information gaps tract from) the dynamism of agriculture in each of related to the measurement and monitoring of Bolivia’s four macro regions and in identifying policy productivity performance in Bolivia. opportunities. The report presents deep dives and in- sights into five thematic areas: • Competitiveness constraints within value chains and across macro-regions. In Bolivia only a small • Drivers and disruptors of agricultural growth number of agricultural commodities are integrat- and sectoral transformation. Chapter 2 explores ed with global value chains, and even those com- this thematic area by examining multiple dimen- modities encounter constraints that hinder their sions of agriculture’s role in Bolivia’s transitioning competitiveness. With the goal of understanding economy. It synthesizes evidence on the changing the factors that determine competitiveness, Chap- dynamics of Bolivia’s agri-food systems. With a ter 5 examines the competitiveness of agriculture view to the future, it considers the demand and in Bolivia in relation to global benchmarks and supply factors that influence the size and config- comparator countries and takes a deeper look at uration of the sector, and highlights opportunities competitiveness constraints and opportunities in for agriculture to support national development key value chains for agriculture and agribusiness. objectives more effectively. • Insights on the policy framework supporting pro- • Agricultural productivity patterns and trends ductivity and competitiveness. Chapter 6 looks at across territories. The land expansion model the current policy framework and public spend- that has dominated agricultural growth in Boliv- ing on agriculture in Bolivia. It gives particular ia in the last few decades has featured limited attention to the enablers of productivity and com- productivity gains, particularly in terms of labor petitiveness and strategies for integrating more and land productivity. Aggregate figures at the producers into value chains and improving value national level conceal important regional dispar- chain performance, including the integration of ities in productivity performance, however. Chap- nutritional considerations into value chains and ter 3 explores this thematic area by providing a into agricultural interventions more broadly. broad view of trends in agricultural performance and trade and outlining the major features of pro- 34 • Climate as a disruptor of sustained growth gains A concluding section synthesizes the findings and the and opportunities to reduce the environmental implications of the analysis. It identifies policies and footprint of the sector. Chapter 7 presents an investments that will enable agriculture to grow in- overview of the challenges to agriculture posed clusively and sustainably. by climatic factors. It highlights how climate smart In sum, this report presents a comprehensive view of agriculture (CSA) can help the sector to adapt to the agriculture and food sector in Bolivia. It combines and mitigate climate change. It focusses attention both high-level and in-depth analyses of themes of on the risk transfer instruments that are currently relevance and interest to a broad audience, including in place in Bolivia to manage climate risks. It also development practitioners, policy makers, academics, highlights opportunities to reducing the agricul- students, private actors, and all parties interested in tural sector’s environmental footprint. enhancing the contributions of the agriculture and food sector to inclusive and sustainable growth in Bolivia. 35 2. Bolivia’s Agriculture in Transition Photo: Programa de Alianzas Rurales (PAR), EMPODERAR Table 1. Average annual growth rate total and per capita GDP (%): Bolivia vs Regional and Structural Peers (Period 2003-2017) Country GDP Per Capita Total GDP Bolivia 2.8 4.5 Regional Peers & Region Brazil 1.3 2.3 Colombia 2.8 4.1 Paraguay 2.7 4.2 Peru 4.0 5.1 Latin America & the Caribbean 1.6 2.7 Structural Peers & Income Group Kazakhstan 1.7 3.4 Ecuador 4.4 5.7 Ghana 3.9 6.4 Mongolia 1.5 4.3 Cameroon 5.8 7.5 Lower Middle 4.1 5.7 Source: Calculations based on WDI data. 2.1. Bolivia’s economic performance context, when Bolivia is assessed in the context of its in the regional and structural context structural peers8 and income group, the results are somewhat less remarkable (Table 1). In less than two decades, Bolivia has emerged as a regional economic star to become one of the fast- Rapid economic growth in Bolivia has been the result est-growing economies in the Latin America and the of a favorable international environment, moderate Caribbean (LAC) Region. Real GDP growth in Bolivia macro-economic policy, public investment, and private has been among the highest in the region, growing consumption. Private consumption has been fueled by by 4.5 percent per year on average between 2003 and the expanding urban middle class and young popula- 2019.6 This rate is well above the regional average tion (two thirds of the 2.6 million Bolivians between the and surpasses several of Bolivia’s regional peers.7 ages of 16 and 29 live in urban centers). Bolivia’s rate During 2006–19, per capita GDP in Bolivia increased of urbanization is converging with that of its region- from US$1,227 to US$3,552, allowing the country to al peers. In 2017, about 68 percent of Bolivians lived achieve lower-middle-income status in 2010. While in urban areas, up from 37 percent in 1960. Between these achievements are outstanding in the regional 1950 and 2012, the country’s urban population grew 6 Recent figures published by the National Institute of Statistics (INE) estimated GDP growth in 2018 at 4.22 percent. 7 Regional peers: South American middle-income commodity exporters with more than 5 million inhabitants, excluding Venezuela. 8 Structural peers: non-fragile, middle-income oil exporters with hydrocarbon exports between 25 and 65 percent of good exports and population between one and on hundred million inhabitants. 38 Figure 1. Rural vs urban population, Bolivia a) Rural vs urban population b) Distribution of population (1960-2017) by department 100% Beni 4% Pando Oruro 1% 80% 37 40 5% 45 56 Tarija 62 66 69 5% 60% Chuquisaca La Paz 6% 25% 40% Potosí 63 60 8% 55 20% 44 38 34 Cochabamba 31 17% Santa Cruz 0% 29% 1960 1970 1980 1990 2000 2010 2017 Rural Urban Source: a) WDI, b) INE at an annual rate of 3.7 percent, almost five times as dramatically from the early 2000s to 2011 (from 0.6 fast as population growth in rural areas. As noted, ur- to 0.46), but it has changed less noticeably in recent banization in Bolivia has focused on peri-urban Santa years, reaching a value of 0.44 in 2017. Between 2000 Cruz, Cochabamba, and La Paz, the three largest cities and 2014 income growth was the most pronounced where more than half of the urban population resides determinant of poverty reduction, both in rural and (World Bank 2015).9 Between 1992 and 2012, the av- urban areas. In the years since then, as poverty has erage annual growth in the peri-urban population was declined more slowly, income redistribution has come 8.2 percent in Cochabamba, 7.4 percent in Santa Cruz, to play a more critical role in reducing poverty (World and 4.1 percent La Paz. Bank 2019). Bolivia’s positive economic performance has trans- The gains in reducing rural poverty have been partic- lated into impressive social advances since the mid- ularly remarkable; rural poverty rate was almost 90 2000s. Growth has undoubtedly been inclusive, with percent in the 1990s and had declined to 55 percent evidence of significant improvements in income re- by 2017. In urban centers, 29 percent of people are distribution and reductions in poverty and extreme poor; population wise, the largest number of poor poverty. The gains in income have allowed the emer- people in Bolivia now live in the urban centers (54 gence of a growing middle class, which official figures percent of the country’s poor). However, it is the rural show to have increased from 35 percent in 2005 to 58 areas where moderate poverty and extreme poverty percent in 2017. Poverty has fallen from 63 percent are still most highly concentrated. Sixty-two percent of the population (2003) to 36.4 percent (2017) and of the extreme poor live in rural areas. Poverty is not 34.6 percent (2018).10 The Gini coefficient of inequal- uniformly distributed across Bolivia’s geographical ity, formerly one of the highest in the region, also fell areas (Figure 2c). Poverty rates are highest in the 9 Urbanization Trends in Bolivia: Opportunities and Challenges, May 2015. The World Bank. 10 National Institute of Statistics (Instituto Nacional de Estadísticas, INE), website. 39 Figure 2. Poverty in Bolivia a) Poverty incidence (national, rural, urban; %), b) Community characterization Bolivia, 2000–17 of rural poverty in Bolivia 100 100 9090 8080 7070 6060 5050 4040 3030 2020 1010 0 0 2000 2005 2000 20052011 2013 2013 20122012 2011 2015 2015 2014 2014 2016 2016 2017 2017 Urban Rural National Urban Rural National c) Decomposition of poverty by department, 2017 (%) (urban/rural) 120% 100% 80% 60% 40% 20% 0% Urbano Rural Urbano Rural Urbano Rural Urbano Rural Urbano Rural Urbano Rural Urbano Rural Urbano Rural Urbano Rural Chuqui La Paz Cocha- Oruro Potosí Tarija Santa Beni Pando saca bamba Cruz Moderate Poverty Extreme Poverty Source: (a) and (c) calculations based on INE data; (b) Ministry of Development Planning departments of Chuquisaca and Potosi, with poverty people, who comprise 42 percent of the population rates of 52.9 and 51.2 percent, respectively in 2017; (52 percent of indigenous people live in rural areas). and they are lowest in Santa Cruz Department, with a Beyond ethnicity, gender and prevailing rural-urban rate of 25.6 percent. The Potosi, Oruro and La Paz De- gaps exacerbate poverty. partments in the Highlands Region; Pando and Beni in the Amazon Region; and Chuquisaca in the Sub-Ande- 2.2. Agriculture in a transitioning an/Lowlands Regions experienced the highest rates economy of rural poverty and extreme poverty (Figures 2b and Structural change is occurring very slowly in Boliv- 2c). Poverty disproportionately affects indigenous ia’s economy. The composition of GDP has remained 40 Figure 3. Sectoral decomposition of GDP, Bolivia, 2000-2017 (%) a) Sectoral decomposition, 2000–17 b) Structural change, 2000 vs 2017 (%) 100 26.8 2000 29.2 2017 13.8 90 15.5 80 70 12.2 60 10.5 50 40 12.8 30 18.6 11.6 18.0 20 8.8 10 2.4 9.2 2.2 0 4.6 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 3.8 Agriculture & primary activities Commerce Agriculture & Construction primary activities Mining & oil Manufacturing Transport, communications, storage Commerce Mining & Manufacturing Finance services & others oil Manufacturing Transport, communications, storage Electricity, gas & water Public administrative services Manufacturing Services Construction Other services Electricity, gas & water (financial, public adm, others) relatively unchanged since the early 2000s (Figure the value of mining and gas exports has declined sig- 3a, 3b). Primary activities such as mining/gas con- nificantly, falling from US$12.89 million at its peak in tinue to dominate economic activity; in fact, Bolivia is 2014 to US$7.8 million in 2017. still one of the most commodity-dependent countries The transformation of the agricultural sector in Boliv- in the LAC Region. In 2017, the share of GDP of the ia suggest that the country has followed the general combined mining/hydrocarbon sector reached 12.2 development pathway seen in economies worldwide. percent and contributed 62.2 percent of total export This path is described in the 2008 World Development value. The sector grew at an average of 6 percent Report: Agriculture for Development (World Bank 2008) during 2000–17, and despite the government’s desire as a series of stages in which economies transition to diversify the country’s export base, limited prog- from being agriculture-based to becoming pre-transi- ress has been achieved—the share of minerals and tion, transition, urbanizing, and ultimately developed hydrocarbons in the value of exports rose 22 percent economies. An important characteristic of this tran- points between 2003 and 2016. sition is that the share of agriculture in the economy In a context of volatility of commodity prices, the sec- and employment tends to gradually decline as an tor’s growth has shown more dynamism during the economy develops. The declining share of agriculture last four years (6.6 percent during 2016–19) than in the economy is consistent with the reallocation of other sectors, such as construction, transport, agri- labor from a highly labor intensive and low-produc- culture, and services (private and public) that regis- tivity agricultural sector to other sectors where pro- tered less dynamism (Table 2.a) and more incidence ductivity is higher, such as manufacturing or services. in the average GDP growth (Table 2.b). Furthermore, 41 Table 2. Sectoral growth in Bolivia (%) a: Sectoral GDP Annual b: Incidence of Sectors in the growth Growth Rate 2016-201911 of the average GDP (2017-2019) Agriculture, Silviculture and Fisheries 6.6% Public Administration 0.54 Public Administration 5.2% Restaurants and Hotels 0.10 Restaurants and Hotels Social Services 0.18 4.7% Social Services Financial Services 0.62 4.1% Financial Services Transport and Communications 0.43 4.8% Commerce 0.36 Transport and Communications 3.7% Construction Commerce 0.13 4.6% Eletricity, gas and wáter Construction 0.09 3.5% Manufacturing Electricity, Gas and Water 0.64 3.4% Mining -0.37 Manufacturing 4.0% Agriculture, Silviculture and Fisheries -3.5% 0.79 Mining -4.0% -2.0% 0.0% 2.0% 4.0% 6.0% 8.0% -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00 Source: Calculations based on INE data Recent work by the World Bank and others has ques- its structural peers, its share of agriculture in total tioned the magnitude of the gains in productivity and employment in 2017 is similar to that of Ecuador, al- economic growth from the reallocation of labor from though the transformation of agricultural employment agriculture to industrial sectors. The work highlights is happening more rapidly in Bolivia than in Ecuador. evidence that the potential gains may not be as large At the same time, this transformation is proceeding as they are often portrayed, and in fact they may be at a much slower pace in Bolivia than in other oil-de- lower than previously thought. The evidence suggests pendent countries (structural peers) such as Mongolia that structural transformation may proceed more and Kazakhstan. In Kazakhstan the share of agricul- rapidly through efforts to support productivity growth ture in total employment has halved since the 2000s, on and off of the farm, to shift the labor market equi- contracting from 36 percent in the early 2000s to 18 librium higher.12 11 percent in 2017. However, in comparison with other highly agriculture-based economies in the LAC region In Bolivia during the past 15 years, labor has moved such as Nicaragua and Honduras, the speed at which from agriculture to manufacturing and services (Fig- labor is moving out of agriculture has been faster in ure 4), often in the informal urban sector, yet agricul- Bolivia. In 1991, the share of agriculture in total em- ture continues to employ around one-quarter of the ployment in Bolivia, Nicaragua and Honduras was 40, population, more than agriculture in Bolivia’s regional 31, and 38 percent, respectively; in 2018, the respec- peers, except for Peru. When Bolivia is compared with tive figures were 28, 31, and 31.9 percent.13 11 Manufacturing includes categories such as: sugar & confectionary; milling & bakery products; wood & wood products; dairy products; fresh & elaborated meat; miscellaneous food products. 12 Harvesting Prosperity: Technology and Productivity Growth in Agriculture. World Bank, forthcoming. 13 Nicaragua is a country that has experienced a reversed trend since 2008, with agricultural employment increasing, after experiencing a rapid declining trend, prior to 2008. 42 Figure 4. Agriculture’s share of total employment, Bolivia vs. regional and structural peers (2000-2018) 50 70 45 60 40 35 50 30 40 25 30 20 15 20 10 10 5 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Bolivia Paraguay Bolivia Cameroon Ecuador Brazil Peru Kazakhstan Mongolia Lower middle income Colombia Latin America & Caribbean Ghana Source: WDI data The diminishing share of agricultural employment in Figure 5c presents a composite picture of where Bo- total employment over the longer term does not mean livia stands in its agricultural transition compared to that agricultural employment in Bolivia is stagnant or other countries in LAC and its structural peers, based declining. Figure 5a presents the results of a recent on changes in agriculture’s share of value-added in assessment by the Food and Agriculture Organiza- total GDP and total employment. As the figure shows, tion (FAO) of the structural transformation of agricul- although the share of agriculture value-added in total ture in LAC. The figure displays mean growth rates GDP is falling in Bolivia, it is not falling at a rate com- of employment in agricultural and non-agricultural parable to rates in urbanizing economies in the region. sectors. The red line is the iso-growth line: countries Paraguay, for example, has a slightly lower share of above it display faster growth in non-agricultural agriculture value-added in total GDP than Bolivia, but employment. Bolivia is at the top, with faster growth the share of agriculture in total employment is much in non-agricultural employment than Ecuador, Peru, lower. Furthermore, although the share of agriculture and Paraguay. Countries to the left of the blue line in total employment is similar in Bolivia and Ecuador, have experienced negative growth in agricultural em- Ecuador has more successfully diversified its econo- ployment, which emphasizes the structural transfor- my, with the share of agriculture in total GDP falling mation of their labor markets. This part of the figure to the single digits. Kazakhstan, a structural peer that shows that growth in agricultural employment has is highly dependent on oil and mining and has been been positive in Bolivia; in fact, the share of primary showing symptoms of Dutch disease since the early agriculture in total employment in rural areas is high- 2000s, has rapidly moved people out of agriculture, er in Bolivia than in all other LAC countries, signaling mainly to the services sector, significantly reducing that in Bolivia the transformation of rural labor mar- agriculture’s presence in the overall economy. kets is still pending (Figure 5b). 43 Figure 5. Agriculture transformation in Bolivia a) Growth agricultural and non-agriculture b) Rural employment by economic activity, 2016 employment in LAC, 1984-2012. .06 100 BRA BOL 90 LN .06 100 COL NIC GTM 80 .04 BRA BOL 90 LN CRI HND VEN PRY 70 COL NIC PANGTM 80 .04 DOM MEXCRI HND PER 60 VEN CHL PAN PRY 70 TTO SLV .02 ECU 50 DOMSUR MEX PER 60 ARG TTO SLV CHL 40 .02 ECU 50 JAMSUR ARG URY 30 0 CUB 40 JAM 20 URY 30 0 CUB 10 20 -.02 0 10 -.02 0 .02 .04 BOL BOL PER PER BRA BRA COL COL LAC LAC ECU ECU GTM GTM HND HND PRY PRY MEX MEX PAN PAN CHL CHL SLV SLV URY URY DOM DOM CRI CRI -.02 0 Mean Agricultural Employment Growth LA -.02 0 .02 .04 Mean Agricultural Employment Growth LA Secondary Primary non-agriculture Terciary Primary agriculture Secondary Primary non-agriculture Terciary Primary agriculture Source: (a) FAO 2016; (b) OTI 2016 c) Agricultural transformation: Bolivia and its regional and structural peers 40% 40% 35% AGRICULTURE-BASED 35% 30% AGRICULTURE-BASED 30% 25% URBANISING TRANSITION PRE-TRANSITION 25% 20% Ghana URBANISING TRANSITION PRE-TRANSITION 20% LMI 15% BOLIVIA Ghana Cameroon LMI 15% BOLIVIA 10% Paraguay Mongolia Cameroon DEVELOPED Ecuador Brazil Colombia 10% Paraguay Mongolia 5% LAC Peru DEVELOPED Kazakhstan Ecuador Brazil Colombia 5% Peru 0% LAC Kazakhstan 0% 0% 10% 20% 30% 40% 50% 60% 70% 0% 10% 20% 30% in Agriculture Employment 40% 50% 60% 70% Employment in Agriculture Source: Based on World Bank Development Report, published in 2007 44 Figure 6. Growth of agricultural value added, Bolivia, 2000-2017 a) Agriculture GDP as share of total GDP, 2000-2019 b) Ag GDP growth, Bolivia vs comparators 2000-19 Country 2000-2009 2010-2019 Bolivia 3.1 4.3 Regional Peers 600,000 16 Brasil 3.9 3.5 14 500,000 Colombia 2.5 3.2 12 Paraguay 5.4 8.0 400,000 10 Perú 3.8 3.2 300,000 8 Latin America and 6 3.1 3.0 200,000 the Caribbean 4 Structural Peers 100,000 2 Ecuador 3.9 3.5 0 0 Kazaskhstan 4.5 2.7 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Ghana n/a 3.6 Total GDP (Million Bs. Constant 1990) Agr GDP (%) Cameroon 2.5 4.8 Mongolia 5.3 6.9 Lower middle 3.9 3.4 income Source: Calculations based on: a) INE, b) WDI A declining share of agriculture in total GDP does not Those forward and backward linkages are rarely es- imply that the sector is not growing. On the contrary, timated, however, although such estimates would the volume and value of agricultural production has provide a more accurate picture of the relevance of grown steadily in Bolivia. During 2000–17, growth agriculture in an economy. To understand the “real” in agricultural GDP in Bolivia averaged 3.0 percent importance of agriculture, an “expanded agricultural per year, higher than in most of the country’s region- value added” is estimated based on the methodology al peers, and among the top rates of its structural provided by Foster and Valdes (2015). Box 2 shows peers, particularly during the present decade (Fig- the results for Bolivia. ure 6). During 2010–17, average annual growth in A distinguishing feature of Bolivia’s agricultural agriculture reached 4.3 percent, and the sector has transformation is that the growth strategy has been grown faster since 2013, reaching an average annual increasingly based on expanding the production of rate of 4.9 percent. commodities to satisfy growing domestic and interna- Another important characteristic of the general pat- tional demand. Consequently, the share of industrial tern of agricultural transformation is that forward agricultural products (particularly soybeans and soy- and backward linkages in the agricultural sector be- bean by-products) in total agricultural GDP rose from come stronger, while the share of the primary sector 7 percent to 22 percent between 1988 and 2017. Note, in total economic activity declines (World Bank 2008). however, that much of this change in the composition 45 Box 2. Re-assessing the contributions of agriculture to Bolivia’s economy (Expanded Ag GDP) Estimation of the expanded agriculture value-added used the Input-Output matrix of Bolivia for 2012. Computed and made available by the National Institute of Statistics (INE), this matrix includes 35 eco- nomic activities and the transactions between them. The first five sectors—non-industrial agricultural products, industrial agricultural products, coca, livestock products and silviculture, and hunting and fishing—are considered together as the primary sector. The estimated expanded value added for the primary sector is provided in Table 3. Note that the share of value-added increases by 5.35 percentage points, from 12.9 percent to 18.3 percent. The contribution of the forward linkages is even greater than the backward linkages, implying that other sectors such as manufacturing, and services rely much more on the primary sector for inputs, and the primary sector relies less on those sectors for its inputs. Table 3. Value added of renewable primary sector and its linkages, 2012 Activity Description Forward and back- total expanded VA Share of sector in FWD sectors (%) BWD sectors at Participation in Participation in Participation in ward sum (%) Activity code midpoint (%) total VA (%) (%) Non-Industrial Agri- 1 5.808% 1.501% 0.142% 1.643% 7.451% cultural Products Industrial Agricultur- 2 2.403% 0.679% 0.434% 1.114% 3.516% al Products 3 Coca 0.624% 0.000% 0.001% 0.001% 0.626% 4 Livestock Products 2.847% 1.462% 0.088% 1.550% 4.397% Silviculture, Hunting 5 1.250% 0.968% 0.079% 1.047% 2.297% and Fishing Total 12.932% 4.610% 0.745% 5.355% 18.287% Table 4 illustrates the dependence of certain key economic activities on the primary sector. It is inter- esting to note that dairy products have a 30 percent dependence on the primary sector for inputs and a 70 percent dependence on other sectors such as sugar, machinery, and transportation and storage. Similarly, fresh and processed meat has a 65.6 percent dependence on the primary sector and 34.4 percent on other sectors. Much of the dependence of the primary sector is related to economic activity in livestock production. 46 Table 4. Dependence of key Economic Activities on Primary Sector Sector Dependence on Primary Sector Dependence on Other Sectors Restaurants & Hotels 7.74% 92.26% Wood & Wood Products 26.68% 73.32% Dairy Products 30.14% 69.86% Milling & Bakery Products 34.46% 65.54% Sugar & Confectionery 40.96% 59.04% Fresh & Processed Meat 65.66% 34.34% Miscellaneous Food Products 68.61% 31.39% of agricultural GDP occurred in the late 1980s and Changes in dietary preferences within Bolivia, result- 1990s, with little change since then. ing from urbanization, higher incomes, and women’s increased participation in the labor force, have sig- The fastest-growing categories during 2000–17 were nificantly influenced the foods that are produced fisheries, forestry, and hunting, growing nearly 4 per- locally and imported. Changes in dietary patterns in cent during the period, albeit from a low base, followed Bolivia resemble changes at the global level, such as by livestock at 3.5 percent and industrial products at the increased consumption of animal protein, added 3.1 percent. The value of industrial crops more than caloric sweeteners (especially in beverages), refined doubled during the same period, and growth in this carbohydrates, ultra-refined highly processed foods, category has been impressive since 2013 (Figure 7d). convenience foods for snacking, and edible oils used The concentration on a few commercial crops, com- to fry foods. bined with the mechanization of production and the Meat consumption in Bolivia has increased greatly, creation of economies of scale, translated into a particularly per capita consumption of poultry, which significant expansion of agricultural area. Oilseeds reached 42 kilograms in 2016, up from 10 kilograms and cereals represented 84.5 percent of agricultural per capita 20 years earlier. Bolivia now has one of planted area in 2016. Between 1996 and 2016, the the highest rates of poultry consumption in LAC. cereal area grew by 4.1 percent on average, while Beef consumption has increased at an annual rate of oilseed area grew by 4.8 percent. The area planted to nearly 3 percent since 2010 and stands slightly above soybeans expanded by 191 percent over the last two global average consumption, but well below the re- decades, reaching 1.3 million hectares in 2016. As gional average. Recent industry reports highlight that discussed in the following paragraphs, the evolution pork consumption has increased at nearly 10 percent of food demand and overall food systems in Bolivia each year in the last four years. have had significant impacts on land use and crop/ livestock patterns. These changes in daily food consumption are reflect- ed in agricultural production developments. Poultry 2.3. Evolution of food demands, food production has exploded since 2000; it was nearly system modernization and nutritional 270 percent higher in 2017 than in 2000. The same challenges in Bolivia comparison for milk production shows a change of 75 47 Figure 7. Composition and growth of agricultural value added, Bolivia a) AG GDP decomposition, 2000-2017 b) AG GDP, 1988 vs 2017 100 6 1988 7 2017 49 90 40 80 100 70 6 1988 7 2017 49 90 60 40 30 80 50 31 70 40 60 30 30 50 20 31 8 40 10 1 7 30 0 22 1988 1989 1990 1991 1992 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 20 8 Non-Industrial Livestock products 10 1 Agricultural Products 7 Non-Industrial Agricultural Products Livestock products Forestry, Hunting 0 Industrial Agricultural and Fishing 22 Industrial Agricultural Products Forestry, Hunting and Fishing Products 1988 1989 1990 1991 1992 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Coca plantations Non-Industrial Coca plantations Livestock products Agricultural Products Non-Industrial Agricultural Products Livestock products Forestry, Hunting Industrial Agricultural and Fishing Fuente: Industrial Agricultural Products (2016); a) FAOForestry, b) Organización Hunting and Fishing InternacionalProducts del Trabajo (2016). Coca plantations Coca plantations 3,000,000 c) Growth by Ag Categories, d) Average growth rates by Ag categories 2,500,000 (Bs thousand, constant 1990) (%, in value terms) 3,000,000 2,000,000 Product 2000-2017 2016-2019 2,500,000 Category 1,500,000 Non-Industrial 2,000,000 1,000,000 Agricultural 2.5 6.6 1,500,000 500,000 Products 1,000,000 Industrial 0 Agricultural 3.1 7.9 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 500,000 Products 0 Non-industrial agricultural products Livestock Coca 1.7 -0.4 Industrial agricultural products Fisheries, forest, hunting 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Livestock 3.5 6.8 Coca Non-industrial agricultural products Livestock Fisheries, Industrial agricultural products Fisheries, forest, hunting Hunting and 3.9 2.3 Coca Forestry Source: Calculations based on INE data 48 Figure 8. Dietary Composition, Bolivia versus global and regional averages Midpoint of TMREL 2g 0%/0g of TMREL 200% of TMREL Processed meat 1.3g 22.5g Calcium Red meat 250g 2g Fruit Slat 60g 7% Legumes Saturated fat 435g 2.5g Milk Sugar-sweetened beverages 20.5g 0.5% Nuts and seeds Trans fat 0.3g 360g Omega 3 Vegetables 11% 125g Polyunsaturated fat Whole grain National Regional Global Source: Bolivia: Nutrition country profile. Global Nutrition Report, 2018 percent; for beef, it is 71 percent. Per capita sales of Changes in dietary patterns also imply changes in ultra-processed foods during 1999–2013 were low in the places where people eat and buy food. Data from Bolivia (41 kilograms per capita) compared to other the Continued Household Surveys carried out by INE countries in the region (Mexico, 160 kilograms; Chile in 2003-04 and 2015-16 show that the share of food 120 kilograms), but sales of these products have con- and drinks consumed outside of the home increased tinuously increased across all LAC countries, with Bo- from 19.4 percent to 33.4 percent during the 12-year livia experiencing the largest increase (+151 percent) period, owing to an increased number of street food (PAHO 2013). Rapid growth of the poultry and other vendors and small restaurants, and more recently of meat subsectors have fueled the large expansion in international fast food outlets. production of oilseed crops, particularly soybeans. Data published by INE provide a picture of house- Consumption of sugar-sweetened beverages is below hold food expenditure and show that Bolivians on regional standards but higher than the global average, average spend 27.06 percent of their income on food while consumption of trans fats is at par with the re- and non-alcoholic beverages, representing a slight gional average and nearly double the global average. change compared with data from 2007 (27.37 per- On the other hand, consumption of vegetables is well cent). Marked differences are observed only in the below regional and global averages and together with percentage of household income applied to food pur- milk, fruits, and legumes presents important opportu- chases outside the home, which increased from 11.07 nities for consumption increases in Bolivia (Figure 8). percent in 2007 to 13.95 percent in 2016.14 14 Critics argue that it is highly unlikely that the significant gains in average real income growth experienced, particularly during the period 2007–14, did not translate into any meaningful reduction in the share of food expenditures in total household income. They suggest that the explanation could be linked to issues with the INE data. 49 Figure 9 . Stunting and overweight rates on children under age 5 Under 5 by gender: stuning (%) Under 5 by gender: overweight(%) 33.4 12 11.1 32.4 10.1 30 9.3 31.4 9 9.2 9.2 9.1 20 16.1 6 16.1 16.1 10 3 0 0 2003 2006 2009 2012 2015 2003 2006 2009 2012 2015 Girls Boys Both Source: Bolivia: Nutrition country profile. Global Nutrition Report, 2018 Traditional markets remain the main outlet for food As these dietary shifts are taking place, Bolivia is ex- purchases in Bolivia. The modernization of tradition- periencing the spectrum of health challenges linked al markets is a recent development, but cities such to poor nutrition: hunger (and hidden hunger), malnu- as Santa Cruz and La Paz have made very import- trition, and obesity. Bolivia has significantly reduced ant investments to improve infrastructure in whole- the number of undernourished people (to 20 percent sale and large traditional retail markets (mercados in 2016 from 33 percent in 2000), but it is also see- minoristas). The penetration of supermarkets and ing rapid increases in obesity and overweight, which modern retail/wholesale outlets has moved at a are risk factors for many chronic non-communica- slower pace in Bolivia than in other LAC countries, ble diseases. The trend toward increased obesity but it is picking up quickly, particularly in Santa Cruz. and overweight in children and women that is com- Supermarket expansion is expected to accelerate in monly observed in other lower- to upper-middle-in- the next few decades, influencing food choices and come countries is also seen in Bolivia. Between 2003 creating opportunities for supply-chain consolidation and 2018, Bolivia achieved a tremendous reduction and modernization, including important investments in stunting among children under age 5—although in improved cold chain and overall agri-logistics and stunting remains very high in rural areas (24 percent) improved food safety. Along with modern food outlets, compared to urban areas (12.2 percent)—but the very traditional markets will remain an important avenue same age group now suffers from overweight, which for linking agricultural supply with consumer demand is on the rise in urban and rural settings (Figure 9). and will also continue to play an important social and The coexistence of wasting, stunting, and overweight cultural role, particularly for indigenous communities. among children under age 5 reaches 2.2 percent. Sustained income growth, infrastructure develop- Overweight/obesity rates are particularly worrisome ment, increased foreign direct investment, and policy in the next highest age bracket (ages 5–19). By 2015, factors can also significantly influence the speed of 26.2 percent of boys in this age bracket and 29.9 per- Bolivia’s agri-food transformation. 50 Figure 10. Producer prices of key agricultural products, US$ per ton 500 3500 450 3000 400 350 2500 Maize 300 Potatoes 2000 Rice, paddy 250 Soybeans 1500 200 Sugar cane 150 Quinoa 1000 100 500 50 0 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: FAOSTAT Note: Agricultural Producer Prices (APP) are prices received by farmers for their produce at the farm gate (i.e., at the point where the commodity leaves the farm). cent of girls were overweight, and 9.7 percent of boys tain dynamic growth rates in the face of international and 8.5 percent of girls were obese. pressures, such as the as the fall in international oil prices, the reduction in other export prices, and the The government is proactively addressing the chal- appreciation of the US dollar. lenge of obesity and overweight. In 2016, the govern- ment released Law N° 775 for the promotion of healthy A recent assessment by the World Bank15 found that food habits (Ley de Promocion de la Alimentacion Sa- although economic growth has remained strong, the ludable) to support emerging efforts to enhance the impacts of external shocks translated into a growth consumption of traditional and highly nutritious foods, reduction in average household income during 2014– which have been underused, sometimes stigmatized 17—a factor that drove significant poverty reduction as “poor people’s food,” and neglected, but which are in the past. The fall in labor income was not due to a integral to Bolivian culture and traditions. reduction in employment, but instead was due to an increase in employment in non-conventional jobs, par- 2.4. Emerging challenges and ticularly self-employment, with lower and decreasing opportunities incomes mainly caused by a reduction in labor income. The average household income was offset, in part, by The past decades have been a period of tremendous changes in the poverty line linked to lower costs of the economic growth and overall prosperity, yet recent basic basket of the poor, especially in rural areas. estimates suggest that gains in poverty reduction and labor markets weakened during the period from 2014 In the agricultural sector, leading traditional com- to 2017. During that period, Bolivia managed to re- modity sectors such as soybeans, and particularly 15 Evaluación de la pobreza en Bolivia: Progreso y desafíos para las políticas sociales y laborales, May, 2019. The World Bank. 51 non-traditional sectors such as quinoa (with a high sive growth. As noted here, the strong performance participation of small-scale farmers), experienced of Bolivian agriculture has been key to maintaining significant declines in income gains as a result of the momentum of economic growth over the past lower international prices. two years as exports from the mining/gas sector declined (IMF 2018). But ensuring that such growth At the same time, several products of the food bas- translates into wealth gains for hundreds of Bolivian ket (such as potatoes and rice) faced increased price farmers is critical. PDES has very ambitious goals competition from imports, resulting from the depre- for agriculture: it seeks to support economic diversi- ciation of local currencies against the US dollar in fication, increase value addition and industrialization, competing countries. and ensure food sovereignty. Its targets are equally In the face of a challenging external environment, the ambitious with respect to the expansion of cropped authorities have been implementing countercyclical area, annual increases in the volume of agricultural policies and an ambitious five-year investment plan— exports and agricultural production, and specific crop The National Economic and Social Development Plan production targets for import substitution.16 2016–2020 (PDES)—approved in early 2016. This The agriculture and food sectors can make an even plan includes a broad public investment program greater contribution to addressing Bolivia’s key pov- financed by macro-economic buffers, Central Bank erty and development challenges if they can become loans, and external financing. The PDES has been more productive, more competitive, and experience supporting growth but is also contributing to fiscal continuous, inclusive, and sustainable growth. Strat- and external current account deficits and foreign re- egies to promote these objectives will be more suc- serve losses (IMF 2018a). Experts argue that main- cessful if they are based on a more precise under- taining macro-economic stability and strengthening standing of regional opportunities and challenges competitiveness will be key for sustaining economic related to productivity, competitiveness, and agricul- and social gains. tural sector vulnerabilities. The chapters that follow The agri-food sector offers major avenues for sup- lay the groundwork to develop that understanding. porting increased economic diversification and inclu- 16 While PDES covers all sectors, the five-year Comprehensive Development Plan for the Agricultural and Rural Sector (PSARDI) adopted in 2017 details the set of policies and targets for the agricultural sector. 52 3. Agricultural performance and production systems 53 Photo: Shutterstock Table 5. National Wealth Estimates, Bolivia and Lower-Middle-Income Country Average Total wealth Average Income Total natural capital Average Income Bolivia Bolivia per capita Group* per capita Group* (2014 US dollars) 49,235 25,948 (2014 US dollars) 17,527 6,949 Subsector share % % Subsector share % % Produced capital 13.45 % 25.17 % Forest 17.79 % 3.49 % Natural capital 35.59 % 26.78 % Protected areas 19.87 % 7.19 % Human capital 50.38 % 50.55 % Crop and pasture land 43.37 % 61.37 % Net foreign assets 0.56 % -2.5% Subsoil assets 18.95 % 27.97 % Source: Lange, GM et al (2018)— The Changing Wealth of Nations 2018 3.1. Natural capital to support departments between 1950 and 2013. The most re- diversified agriculture in Bolivia markable development, apart from the increase in to- tal agricultural area reflected in the different sizes of Bolivia’s natural capital, represented by its diverse the two pies, is the striking increase in the relative im- geography and agro-ecological zones, abundant portance of Santa Cruz Department, which had 8.9 of supplies of land and water, and varied climate, pro- all harvested area in 1950 and 60.7 percent by 2013. vides the basis for its diversified agricultural sector. In 2014, Bolivia’s natural capital represented 35.6 3.2. Patterns in crop and livestock percent of the country’s total per-capita wealth (Ta- production and value ble 5),17 almost 9 percentage points above the cor- responding average share in lower-middle-income The production structure in Bolivia has not changed countries. Crop and pasture land, in turn, represent- significantly in the last 15 years. The share of the ed 44 percent of the country’s natural capital wealth, top 10 crops remains the same, estimated at about which corresponds to 13.5 percent of the total nation- 80–82 percent of harvested area and of about 87–90 al wealth per capita. percent of production. The most significant increases occurred in cereal and oilseeds/industrial crops. Oil- Bolivia’s crop and pasture land is a source of agricul- seeds and industrial crop production increases were tural growth, and Bolivia has used it. Between 1950 driven mostly by sugarcane and soybeans; in 2017, and 2016, total harvested area increased five-fold, production of these two crops surpassed 11 million from around 654,000 hectares to more than 3.5 mil- tons, roughly five times the total for cereals, which lion hectares (INE Statistics). Most of this expansion had the second-highest production volume (Table 6). took place in the lowland areas (including Amazon On the other hand, the area planted to cereals has Region), which accounted for 73.6 percent of the total doubled since the early 2000s, growing faster than expansion, followed by the highlands (14.6 percent) the area planted to oilseeds and industrial crops. The and Andean valleys (11.8 percent). Figure 11 shows recent expansion of vegetable crops is also notable; the distribution of agricultural land across Bolivian their area has nearly doubled since the early 2000s. 17 National wealth includes produced capital, natural capital, human capital, and net foreign assets. It complements GDP and other macro-economic indicators by measuring changes in a country’s underlying asset base. 54 Figure 11. Harvested area shares, by department. Bolivia, 1950-2013 1.61% 1.81% 0.48% 0.43% 8.90% 4.77% 12.60% 3.95% 10.60% 7.26% 21.04% 28.75% 60.76% 4.03% 5.79% 3.49% 4.49% 19.21% Chuquisaca Oruro Santa Cruz La Paz Potosí Beni Cochabamba Tarija Pando Source: Calculations based on INE and Census data Table 7 ranks the main crops by value. The share of Growth in the production value of high-value crops the top 10 products increased from 40 percent to 45 such as nuts (chestnuts and Brazil nuts) and quinoa percent between 2000 and 2016 (the last year for has been modest. For Brazil nuts, annual average which data are available). Soybeans, sugarcane, po- growth was nearly 2 percent during 2000–16; total tatoes, and maize continue to be the leading crops, production value of chestnuts declined during the accounting for one-third of the value of agricultural mid-2000s and has been recovering slowly in recent production in 2016. The share of soybeans in total years. For quinoa, average annual growth in the val- crop value rose from 13 percent in 2000 to 20 per- ue of production reached 16 percent during 2010–13 cent in 2016. The increasing relevance of sorghum, but has been negative since 2014, when international wheat, and bean production is notable. Sorghum, prices collapsed. Crops such as tomatoes and cassa- which ranked 26th in value of production in 2000, had va, which ranked high in terms of value in the early surged to 5th place by 2016, adding nearly 240,000 2000s, have been losing share. Area planted to to- hectares of harvested area during that period. The matoes has increased only by 1,765 hectares, while value of wheat production tripled during the same cassava (yuca) area increased by roughly 10,000 period, while the value of dried beans increased sev- hectares. Coffee area grew by approximately 3000 en-fold. Dried beans from Bolivia increasingly sup- hectares during 2000–16, but its total value declined. ply regional markets, and although the value of bean The value of livestock production has increased near- production is just 0.4 percent of the value of maize ly 70 percent since 2000 and was estimated at US$1.6 production, beans occupy only 19 percent of the land billion in 2017. The national cattle herd was about 9 area used in maize production. million head in 2017, adding about 2.6 million more animals since the early 2000s. Beef production is es- 55 Table 6. Production volume and area by crop category. Bolivia, 2000-2017. Production Area ∆ Production Annual Annual Production Area 2017 ∆ Area 2000- Crop 2000-2017 Average Average 2017 (tons) (Ha) 2017 (Ha) (Tons) Growth Rate Growth Rate (%) (%) Cereals 2,321,408 1,320,336 5.7 1,241,850 616,085 4.9 Forage 486,870 253,124 4.6 120,212 54,650 4.0 Fruits 1,486,203 619,750 3.6 147,557 50,747 2.8 Oilseeds/Industrial 11,177,679 6,386,994 4.3 1,563,302 718,884 4.2 Stimulants 27,876 5,679 1.7 34,187 6,423 1.4 Tubers and Roots 1,419,114 537,464 2.6 225,242 73,387 2.5 Horticulture/ 448,579 214,893 4.5 165,853 79,510 Vegetables 4.8 Source: Calculations based on INE data timated at 260,000 tons per year. The minor livestock agricultural exports in 2018 (excluding soybeans and species that are important in small-scale agriculture their by-products). have also increased in number. The llama herd grew Agricultural exports and soybean exports contracted by 25 percent over 2000–17 to reach nearly 2.6 mil- during 2014–16, regaining momentum in 2017 and lion head; sheep numbers increased by 18 percent 2018 (Figure 12a) but showing again a contraction in (7.4 million head in 2017) and alpacas by 23 percent 2019 (Figure 12b), reflecting changes in internation- (430,000). Poultry production expanded tremendous- al prices. For example, soybean prices dropped from ly, growing by 194 percent between 2000 and 2017. US$521 per ton in May 2014 to US$368 in December 2015 and below US$300 most recently. The price of 3.3. Trade trends quinoa, in turn, plummeted from US$6.34 per kilo- The patterns of crop expansion have supported the gram in 2014 to US$1.66 in 2017, and then improved strong export growth experienced during the past somewhat to US$2.32 in 2018. Bolivia’s main agricul- two decades, with increasing concentration of ex- tural export partners include Colombia, Peru, Ecua- ports on a few products. The share of soybeans and dor, Chile, the United States, Germany, Argentina, the soybean by-products in the total value of Bolivia’s ex- United Kingdom, Brazil and Italy. ports was estimated at 8 percent in 2007; it rose to Overall, Bolivia is a net-exporter of agricultural prod- 12 percent in 2016 and then declined to 9 percent in ucts, and its agricultural trade balance has showed 2018. The share of primary agriculture products (ex- a surplus during 2000–18 (Figure 13). The growth of cluding soybeans) in the total value of Bolivia exports agricultural imports has been very dynamic, howev- increased from 4 percent in 2007 to 4.8 percent in er, increasing on average by 8.5 percent during this 2018. Non-traditional agricultural exports are led by period, compared to growth in agricultural and food Brazil nuts, quinoa, bananas, chia, and beans, which exports of 8.7 percent. In 2018, the value of Bolivia’s together represented 87 percent of the total value of 56 Table 7. Ranking of crops in terms of their share of total agriculture value (‘000 dollars, constant 2010). Bolivia, 2000-1016. Ranking Crop 2000 Ranking Crop 2016 1 Soybeans 328,399 1 Soybeans 879,013 2 Potatoes 121,770 3 Potatoes 181,228 3 Sugar cane 118,271 2 Sugar cane 226,930 4 Maize 92,546 4 Maize 133,624 5 Rice, paddy 83,342 6 Rice, paddy 113,408 6 Plantains 78,283 7 Plantains 94,320 7 Bananas 45,603 8 Bananas 84,371 8 Tomatoes 36,117 23 Tomatoes 36,117 9 Cassava 35,754 24 Cassava 35,754 10 Brazil nuts 32,286 14 Brazil nuts 32,286 13 Chestnut 26,752 9 Chestnut 65,957 26 Beans, dry 7,999 10 Beans, dry 56,913 32 Sorghum 14,538 5 Sorghum 128,396 22 Wheat 16,016 12 Wheat 54,631 Source: FAOSTAT Figure 12. Agriculture exports. Bolivia, 2000-2018 b) Exports by main products a) Agriculture exports (‘000 US dollars) (excluding oilseed products, ‘000 US dollars) 500 1200 450 1000 400 800 350 Thousands Thousands 300 600 250 400 200 150 200 100 0 50 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Agriculture, livestock, fisheries, forestry Castaña (Nueces del Brasil) Chía Soybeans & Subproducts Quinua Bananas Beans Source: Calculations based in INE data 57 Figure 13: Agricultural trade flows, Bolivia 2000-2018 (US$ ‘000)19 $2.500 $2.000 Valor (US$ millones) $1.500 $1.000 $500 $0 -$500 -$1.000 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Imports Exports Not Trade Balance Source: Calculations based on COMTRADE data agricultural imports reached US$0.78 billion. The top exports and imports as part of its objective of achieving five leading import categories represented 62 percent food sovereignty (For more details see Chapter 6.3). of the value of imports in 2018 and included edible food preparations; products for the milling industry; 3.4. Characterising Bolivia’s cereal preparations; beverages and spirits; and ce- agriculture reals. The category of edible food preparations has doubled in value since 2010, representing 22 percent 3.4.1. Duality and spatial heterogeneity of the value of agricultural imports in 2018. Fruit Bolivian agriculture has a dual structure in which imports attained a value US$22.5 million in 2018, traditional and commercial agricultural production representing only 2.8 percent of the value of imports systems coexist. On the one hand, most agricultural that year, but this category of imports has shown very production units—unidades productivas agropecuarias, strong growth (an average annual growth rate of 8.8 referred to as “UPAs” throughout this report—and the percent since 2010). Imports of vegetables showed an bulk of the rural population form part of the peas- increasing trend during 2010–17, with average annu- ant-indigenous economy. Using traditional and only al growth of 9 percent over that period and a value of partially mechanized production methods, numerous US$4 million in 2017 (which is still less than 1 per- small-scale and medium-scale farmers produce agri- cent of the total value of imports).18 The top 10 im- cultural products mostly for their own household con- port partners are Argentina, Brazil, Chile, the United sumption; with limited access to markets and com- States, Peru, Colombia, Republic of Korea, Paraguay, mercial networks, they produce only a limited surplus Mexico, and mainland China. 19 for sale in local markets and nearby urban centers. These developments reflect the trade policy imple- A high number of producers sell their products to in- mented by the Government of Bolivia, which controls termediaries. Traditional agriculture is characterized by the low availability and quality of productive inputs 18 In 2018 the figures suggest a considerable contraction, with the value of agricultural imports estimated at US$1.7 million. 19 Exports and imports include the categories: animals, vegetables, and processed food. 58 and resources, which translates into low levels of in- however, agriculture’s largest contribution is to the vestment, automation, and innovation, and the pursuit local economy. of traditional risk-minimization strategies. Producers in different regions specialize in very differ- Commercial agroindustry, on the other hand, fea- ent sets of crops. Annex 1 displays the average area tures a large-scale, specialized production model harvested per crop in each macro-region. Table 8b that makes use of modern production techniques, a presents the cropping structure for 2014/15–2016/17, predominantly hired labor force, and capital-intensive based on the area cultivated in a representative de- production practices. Producers enjoy good levels of partment for each macro-region: Santa Cruz (Low- access to working and financial capital, as well as to lands), Cochabamba (Sub-Andean), Potosí (Highlands), technical assistance and information. Well-developed and Pando (Amazon). Even though potatoes are a major backward and forward linkages make this system crop in both Cochabamba and Potosí, and fruit produc- highly suitable for the national and export market. tion is prominent in Cochabamba and Pando, the four departments actually present very different profiles of This heterogeneity has a clear spatial component. crop specialization. For example, Santa Cruz produces Producers in the western and Andean areas of Boliv- a vast range of crops, but soybeans predominate by far, ia tend to follow a more traditional agricultural model occupying 52 percent of cultivated area. In turn, Co- dominated by subsistence farming, but moving east- chabamba (Sub-Andean Region) and Pando (Amazon ward toward the inter-Andean valleys, mixed produc- Region) show a more polarized structure; agriculture tion models are more common. In these mixed mod- in Cochabamba emphasizes fruit, potatoes, and maize els, traditional farming systems on the slopes coexist (jointly accounting for approximately 65 percent of with modest mechanization and modern technology cultivated land), and Pando emphasizes fruit, maize, in the plains, and relatively sophisticated rural societ- cassava, and rice (86 percent of the cultivated land). ies enjoy moderate market access. Eastern Bolivia is Potosí, on the other hand, reveals less emphatic crop characterized by the prevalence of modern commer- preferences, although potatoes and quinoa jointly ac- cial agribusiness and extensive production systems. count for 47 percent of the cultivated area. The segmentation of agricultural production systems Triangulating the information of Table 8b with Figure across the country calls for a regional analysis that 14, which displays the distribution of producers by brings up the diversity in Bolivia’s agriculture, ad- the main crops in the macro-regions, sheds light on dressing the key challenges and opportunities faced the scale of production for various crops. For exam- by the sector across the different macro-regions. ple, despite its unquestionable prevalence in terms of land use, soybean is grown by only 8 percent of 3.4.2. Production trends and systems in the UPAs in the Lowlands Region, which is indicative the macro-regions of large-scale, commercial agriculture. In the same Table 8a below highlights the relevance of agricultur- region, on the other end of the spectrum, maize is al activities in Bolivia’s departments to the national the most frequently harvested crop (grown by one- and local economy. Santa Cruz, Cochabamba, and La third of UPAs), but it is grown on less than 10 percent Paz are highly diversified economies; together, they of the cultivated area. In contrast, in the Sub-Andean contribute 70 percent of the value of Bolivia’s agricul- and Highlands Regions, the share of UPAs growing tural output. In departments such as Beni and Pando, a crop matches the share of cultivated area for that crop quite closely. In the Sub-Andean Region, potato 59 Table 8. Ranking of crops in terms of their share of total agriculture value (‘000 dollars, constant 2010). Bolivia, 2000-1016 a) Share of AG in total GDP and b) Percentage of planted area by major crops in select Department GDP departments, average 2014/15–2016/17 (%) Agr share Ag share of Lowlands: Sub-Andean Highlands: Amazon: Crop Department of total Department’s Santa Cruz Cochabamba Potosi Pando GDP (%) GDP (%) Barley 10.97 Beni 6.9 27.9 Broad bean 1.97 7.02 Chuquisaca 6.8 15.2 Cassava 18.12 Cochabamba 13.4 9.6 Fruit 28.22 29.15 La Paz 17.4 7.0 Maize 9.11 17.28 12.28 22.75 Oruro 2.2 5.1 Oats 2.86 Pando 2.2 26.9 Peas 2.19 Potosi 5.5 10.0 Potatoes 20.46 19.16 Santa Cruz 41.0 15.8 Quinoa 28.17 Tarija 4.7 6.5 Rice 5.11 3.42 17.60 Sorghum 11.27 Soybean 52.14 Sugarcane 5.45 1.05 Sunflower 4.70 Wheat 5.66 9.58 11.42 Total area 2,478,744 190,897 157,088 12,859 (ha) Source: Calculations based on INE data and maize production are prevalent (grown by 21 and tribution of farm size in each region (Figure 15). In 17 percent of UPAs, respectively), but the very sub- the Highlands Region, for instance, 60 percent of all stantial diversification of production is reflected in the farms are smaller than 1 hectare, and a staggering sizeable percentage of producers who grow “other” 97 percent of UPAs are smaller than 10 hectares. crops. Agriculture in the Highlands Region focuses on The situation is quite similar in the Sub-Andean and potatoes and quinoa. The crop pattern in the Amazon Amazon Regions, even though UPAs of less than 1 Region is also consistent with shares of cultivated hectare are not quite as dominant, but in any case, area; cassava and maize are among the crops grown average farm size does not exceed 3 hectares in any by large numbers of producers, and the share of pro- of these regions: it 1.9 hectares in the Highlands, 2.3 ducers growing various types of fruit is high. hectares in the Sub-Andean Region, and 2.6 hectares in the Amazon Region. The average farm size in the Differences in the scale of production across mac- Lowlands Region, at 44 hectares, reflects the more ro-regions are even more apparent from the dis- commercial nature of agriculture there. Only 27 per- 60 Figure 14. Main crops by macro-region, based on the number of producers. Bolivia, 2015 50% 45% 43% 40% 35% 34% 33% 30% 28% 25% 21% 20% 18%18%18% 18% 17% 14% 15% 15% 13%13% 11% 10% 8% 8% 8% 8% 8% 5% 5% 6% 5% 4% 4% 4% 4% 4% 3% 3% 3% 2% 0% Potato Quinoa Broad bean Barley grain Wheat Oca Green pea Others Cassava Banana (Postre) Corn Rice Banana (Banano) Other fruits Other citrics Other Corn Rice Cassava Soybean Banana (Postre) Peanut Bean Other Potato Corn Wheat Broad bean Green pea Orange Banana (Banano) Others Highlands Amazon Lowlands Sub-Andean Note: Light bars: % of total productive units that produce the crop; darker bars: % of total productive units for which the crop represents 50% or more of total GDP Source: ENA 2015 Figure 15. Distribution of UPAs by size (%) 3% 2% 2% 5% 12% 17% 20% 20% 25% 38% 20% 34% 28% 60% 43% 44% 27% H IGHLANDS AMAZON LOWLANDS SUB-ANDEAN Less than 1 ha 1 to 3 ha 3 to 10 ha 10 to 100 ha More than 100 ha Source: ENA 2015 61 Figure 16. GDP of major crops by Macro-region (Bs$/Ha). Bolivia, 2015 100,000 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 Quinoa Potato Broad bean Barley Wheat Oca Carrot Corn Green pea Cañawa Sugar cane Cassava Corn Rice Banana (Postre) Pineapple Cocoa Banana (Banano) Mango Almond Sugar cane Soybean Corn Rice Sorghum Wheat Sunflower Lemon Bean Sesame Banana (Banano) Potato Corn Banana (Postre) Peach Cocoa Wheat Orange Green pea Palmetto Highlands Amazon Lowlands Sub-Andean Source: ENA, 2015 cent of farms in the Lowlands Region are smaller be much more valuable than most other crops in the than 1 hectare; 25 percent surpass 10 hectares, and figure, its value is also clearly greater in the Low- 20 percent exceed 100 hectares. lands Region than in the Amazon Region. Similarly, the substantial diversification of production in the The East-West farm-size divide seems to extend to Sub-Andean Region seems to include the production land quality as well, with levels of soil erosion worsen- of higher-value crops such as bananas and peach- ing toward western Bolivia. Moderate levels of erosion es. In contrast, GPVs per hectare are minimal for all affect more than half of the land in the departments in crops in the Highlands Region, including quinoa. the Highlands Region like Potosí and Oruro, which also experience severe levels of erosion, in which about 10 Just as cropping patterns vary widely across mac- percent of soil is eroded. Erosion is caused mainly by ro-regions, so do crop yields (Figure 17). Although water and wind, especially in areas where slopes are yields of crops such as sugarcane and plantain ap- steep and vegetative cover is scarce. pear high compared to those of other crops grown in Bolivia, in general the yields of practically all crops Regional disparities in the value of production are grown in Bolivia are quite low compared to yields in shown in Figure 16. Although sugarcane appears to 62 Figure 17. Yields of main crops by Macro-regions (Quintals per Ha) 1200 1000 800 600 400 200 0 Sugar Corn Rice Banana Potato Wheat Banana Cassava Onion Quinoa Green Broad Barley Oca Cane (Banano) (Postre) pea bean Highlands Amazon Lowlands Sub-Andean Source: ENA, 2015 other countries. For example, FAOSTAT data show Figure 18 displays output value per hectare by farm- that the Bolivian sugarcane yield is similar to yields size class (the number of UPAs in each macro-region obtained in Argentina, Uruguay, and Paraguay but was divided into quintiles based on farm size). In the substantially lower than yields in Brazil (40 percent Amazon and Sub-Andean Regions, it is the smallest higher), Colombia (65 percent higher), or Peru (130 and largest farms that generate the highest output percent higher). Banana yields in Colombia, Ecuador, value per hectare. This result suggests what the liter- and Peru are 2–3 times higher than banana yields ature refers to as a “U-shaped relationship” between in Bolivia. Similar discrepancies are seen in other profitability and farm size. This relationship is not important crops such as potatoes and soybeans, al- seen in the Lowland and Highlands Regions, although though they are more pronounced in soybeans. Com- in the Highlands Region the smallest farms have the pared to the average potato yield in Bolivia, the potato highest output value. yield is 2 times higher in Ecuador, 2.4 times higher in Another relevant aspect of regional production pat- Peru, 3 times higher in Colombia, and almost 5 times terns to analyze is the labor productivity of the main higher in Brazil. But these are minor differences when crops. Figure 19 displays the ratio of agricultural GDP compared with the yield differentials for soybeans. to the number of days worked per year by family and Argentina and Brazil, soybean production powerhous- hired labor in each UPA. Labor productivity was high- es, have average soybean yields that are 30–40 per- est in the Lowlands Region, even when only family or cent higher than those in Bolivia. hired labor is considered. The Highlands Region has the lowest labor productivity.20 20 This higher labor productivity may reflect the fact that most crops in the Lowlands Region are less labor-intensive to produce and have higher yields. In addition to these considerations, it is possible that some of the difference in labor 63 Figure 18. GDP (Bs$/Ha) according to farm-size, by Macro-regions 25,000 20,000 15,000 10,000 5,000 0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Highlands Amazon Lowlands Sub-Andean Source: ENA 2015 Figure 19. Output per worker Bs/day/UPA (annual average per UPA) Sub-Andean Lowlands Amazon Highlands 0 20 40 60 80 100 120 140 Family Labor Hired Labor Total Source: Calculations based on ENA 2015 A more complete view of the use of factors of produc- used by UPAs, including intermediate inputs, capital tion, including labor, is presented in Figure 20. This (tractors), and hired and family labor. The Lowlands figure shows the percentage distribution of the inputs Region differs from the Amazon and Highlands Re- gion in the greater relative use of intermediate in- productivity between the Lowlands and other regions puts with respect to the other factors of production. is the result of higher levels of training, knowledge, and The greater preference for intermediate inputs in the information in the lowland workforce. However, it could also be due to the intensive use (or higher quality) of Lowlands Region could be related to the higher prof- other inputs (beyond land and labor). 64 Figure 20. Expenditure share in inputs (Percentual distribution by Macro-region) Sub-Andean Lowlands Amazon Highlands 0% 20% 40% 60% 80% 100% Tractors Intermediate Inputs Family Labour Hired Labor Source: ENA 2015 itability and yields of crops gown there. In the Amazon the country (81 percent) are the traditional flood and and Highlands Regions the main factor of production gravity type, which contributes to low agricultural is family labor. Spending on inputs is most evenly bal- productivity and progressive soil erosion. Only 5 per- anced in Sub-Andean Region. cent of all irrigation systems use modern technology. The Lowlands Region is also the national leader in Fertilizer use is extremely low in Bolivia, especially use of certified seed, with the Santa Cruz Department compared to other countries in the region (Annex 1). alone accounting for 83.4 percent of certified seed Mechanization could also improve, as suggested by use in Bolivia (Annex 1). The overwhelming majority estimates of the use of various types of farm machin- of certified seed used in Santa Cruz is soybean seed, ery and tools in 2013–14 (Annex 1). reflecting the region’s specialization in this crop. Pro- Against this review of the major distinguishing fea- ducers in Santa Cruz Department also use high level tures of agriculture in the macro-regions, the chapter of certified wheat, maize, and rice seed. Elsewhere that follows presents a detailed analysis of patterns in the country, the use of certified seed is most com- of productivity for the four macro-regions, major mon in potato (Cochabamba, La Paz, Oruro, Potosí, crops, and farm-size categories. The results offer and Tarija), rice (Beni), and maize (Chuquisaca) pro- insight into the opportunities for improving the effi- duction. ciency of agricultural production in Bolivia’s highly Irrigation is concentrated in the Sub-Andean Re- diverse agricultural landscape. gion (Annex 1). The Highlands Region has a high- er-than-average presence of small and micro irri- gation systems, while the Lowlands Region has a proportionally much higher share of medium and large irrigation systems. Most irrigation systems in 65 4. Productivity and Efficiency of Agriculture in Bolivia Photo: Shutterstock Figure 21. Agricultural TFP growth, 2001–15 (%) 6,0% 5,0% 4,0% 3,0% 2,0% 1,0% 0% -1,0% -2,0% -3,0% Paraguay Bolivia Venezuela Colombia Argentina Ecuador Chile Honduras Peru Nicaragua Brazil Uruguay 2001-2010 2011-2015 Source: USDA, 2018 4.1. Bringing the productivity and result of technological change rather than because of efficiency picture into focus increased technical efficiency or managerial capac- ity (Pfeiffer 2003; Ávila et al. 2010; Pratt et al. 2015; The agricultural sector is making and can make an Trindade and Fulginiti 2015). However, as illustrated even higher contribution to economic development in Figure 21, TPF in Bolivia displays two very distinct and poverty reduction in Bolivia if it gives high pri- patterns. The first pattern is negative growth during ority to growth in productivity. As Hayami and Ruttan the period 2001–10 (–0.8 percent), making Bolivia one (1971, 1985) point out, the success of agriculture has of the worst performers in the LAC Region. The sec- more to do with increases in productivity than with ond is a more positive growth rate during the period using more inputs or cultivating more land. 2011–15 (2.6 percent), when Bolivia was surpassed only by Colombia and Argentina. The previous chapters discussed the components of agricultural growth in Bolivia related to area ex- Despite positive TFP growth, the productivity of land pansion and input intensification. This chapter seeks and labor remains low in Bolivia compared to other to understand the levels of efficiency in the use of countries in the region (Table 9). Land productivity agriculture inputs as a key determinant of long-term is four times higher in Colombia, three times high- productivity gains. A method that is commonly used er in Ecuador, and two times higher in Peru. Labor in this kind of assessment is the estimation of To- productivity is three times higher in Mexico, Ecuador, tal Factor Productivity (TFP). Estimations of TFP are and Paraguay; it is two times higher in Colombia, and often provided at an aggregated level for the over- higher by half in Peru. all agricultural sector. Those estimates indicate that productivity growth in Bolivia has been modest over To date, the only measures of the productivity and the last two decades and that it has been mainly the technical efficiency of Bolivian agriculture come from 68 Table 9. Land and labor productivity in selected LA countries Land Productivity Labor Productivity Output per hectare Output per worker (2004- Country Av. Annual Growth Av. Annual Growth (2004-2006 USD$/Ha) 2006 USD$/worker) Year 2004-2015 (%) 2004-2015 (%) Year 2015 2015 Mexico 1435 2,11 5289 3,18 Colombia 3247 1,82 4609 2,44 Ecuador 2435 2,21 5141 2,25 Peru 1650 3,14 2648 3,40 Venezuela 1758 1,57 10 510 3,30 Brazil 1736 2,26 16 631 6,12 Argentina 1125 0,63 35 954 2,86 Chile 3841 1,63 8914 1,45 Paraguay 1133 2,14 6943 3,81 Uruguay 1552 -1,93 23 953 2,85 Honduras 1426 1,52 3302 2,01 Nicaragua 855 4,42 4882 3,26 Bolivia 751 2,64 1909 1,76 Source: Calculations based on USDA 2018 data studies at the national level. As indicated in the over- and competitiveness. The analyses also provides im- view of macro-regions in Chapter 3, however, Bolivian portant insights for policy analysists on the informa- producers operate in extremely diverse agricultural tion gaps that need to be addressed to generate esti- settings. The regions differ substantially in the kinds mates of productivity at aggregate and more granular of production systems they can support, the charac- levels and make it possible to monitor improvement teristics of producers, and yields of the main crops. over time. Considering these marked differences, more granu- 4.2. Productivity and technical lar information on productivity and efficiency across efficiency: main concepts macro-regions, farm types, and crops could provide greater insight into opportunities for tapping new The traditional measure of agricultural productivity— sources of productivity growth in Bolivian agriculture. output per hectare (volume or value per hectare)—is This chapter attempts to narrow this knowledge gap. incomplete, since it does not consider the use of oth- It describes results of a more detailed assessment of er inputs that may contribute to productivity, such as technical efficiency across Bolivian’s macro-regions, fertilizer, labor, pesticides, machinery, and so on. Total farm sizes, and crops, based on stochastic frontier factor productivity (TFP) is a more complete measure and meta-frontier analyses, and explores the impli- of productivity that accounts for all inputs used in cations for policies designed to increase productivity production and compares them with the total amount 69 of output (in this case, crop and livestock products). productivity levels among different groups and over TFP is defined as the ratio of total output (Y) to total time, while the analysis of efficiency helps to identi- inputs (X). If total output is growing faster than total fy the determinants of productivity at the farm level, inputs, TFP is said to be improving—in other words, which is useful information for defining the scope of productivity is growing. specific policies to support productivity. Since TFP is measured as a residual, it is important The relationship between productivity and efficiency to recognize that spatial and temporal differences in is depicted in Figure 22, which shows two groups of TFP can be caused by a wide range of factors. Spatial production units with different levels of productivity differences in TFP are commonly the result of struc- (TFP). In this example, group 2 (G2) has the highest tural factors, such as variability in climate, topography, productivity, and group 1 has the lowest. Productivity and soil quality. Spatial differences in TFP can also is the output/input ratio (Y/X, and technical (in)effi- result from locational differences in access to finance ciency can be understood as the difference in pro- and to the capital embodied in infrastructure invest- ductivity of one production unit (any of the blue or red ments. By contrast, temporal differences in TFP are dots) compared to the more productive units (located usually caused by other factors, such as changes oc- at the frontier), depending on the context or group in curring over time in the mix of crops, adoption of new which the operation takes place. technology, reductions in transaction costs due to the Data to carry out a disaggregated TFP analysis for ag- appearance of new information and communication riculture in Bolivia are not available, so the productiv- technology (ICT), or improved connectivity to markets ity analysis described here focuses on understanding arising from the construction of new transportation technical efficiency through stochastic frontier and infrastructure. Other factors that could contribute to meta-frontier analyses, for which data are available. changes over time in TFP could be more difficult to Annex 2 explains the data requirements and data measure, such as improved access to financial assets, sources for TFP and those used to carry out the sto- the level of entrepreneurial ability within a region, and chastic frontier analyses presented here. the attractiveness of a region to new entrepreneurs. In summary, TFP is a measure explained both by the 4.3. Methodology and variables context in which farmers operate and by their own de- cisions and productive abilities. 4.3.1. Stochastic frontier analysis and Technical efficiency is closely related to the ability of meta-frontier analysis firms to achieve the optimal combination and best Stochastic frontier analysis (SFA) is used to estimate use of inputs to produce the best results with a given a production frontier, which represents the current technology. Therefore, unlike TFP measures, techni- state of technology in a certain industry or sector. cal efficiency has to do with the productivity compo- The frontier also defines the maximum potential or nent that is largely under the farmer’s control—the optimum level of production of the industry. Firms or portion of TFP explained by decisions and productive companies operate on the frontier if they are techni- ability discussed in the previous paragraph. cally efficient, and below the frontier if they are not. Together, TFP and technical efficiency are important The frontier is considered stochastic because the de- measures for understanding the varied performance viations from the frontier include a component that of agriculture in Bolivia’s diverse agricultural land- represents random shocks affecting the production scape. TFP measures make it possible to compare 70 Figure 22. Productivity and efficiency: illustration of concepts Y (Producción) Ineficiencia G2 Ineficiencia G1 X(Insumos) Source: Author’s elaboration process and another component with a specific dis- • Stochastic frontier analysis to identify the deter- tribution function representing technical inefficiency. minants of agricultural efficiency in each mac- ro-region (Sub-Andean, Highlands, Lowlands, and One disadvantage of standard stochastic frontier Amazon). models is that it is not possible to compare the tech- nical efficiency of producers using different technol- • Stochastic meta-frontier analysis to understand ogies, since they do not operate under the same pro- the differences in efficiency levels and technology duction frontier. Spatial differences in temperature, gaps across regions. rainfall, soil characteristics, and other production fac- • Stochastic meta-frontier analysis of farms of dif- tors determine the input-output combinations that are ferent sizes (land area) producing four crops of possible for each region, and these combinations re- national importance—soybeans, maize, rice (pad- sult in different production frontiers. Bolivian produc- dy), and potatoes—to explore the relationship be- ers from different macro-regions (and even within the tween the size of the UPA and meta-efficiency in same macro-region) inevitably use different technol- the selected crops. ogies and production systems to accommodate their divergent circumstances. To get around this difficulty, • Stochastic meta-frontier analysis of producers a stochastic meta-frontier efficiency analysis (MFE) is located in different macro-regions who grow the used to calculate efficiency measures that allow com- same crop, to explore patterns of meta-efficiency. parisons between groups with different technologies (see Annex 2 for details on the methodology and data The next section describes the main variables used sources). to characterize agriculture in the macro-regions. The remainder of this chapter discusses the analysis and The overall analysis had several components: the results. 71 4.3.2. Variables used in the analysis The analysis includes a variable on the use of labor, both family and hired labor.21 Family labor accounts The sample of agricultural producers for each mac- for more than 85 percent of the value of agricultural ro-region (Table 10) comes from the National Agricul- labor in the Highlands and Amazon Regions, 75 per- tural Survey (ENA) 2015. A significant percentage of cent in the Lowlands, and 64 percent in the Sub-Ande- observations come from the Sub-Andean Region (49 an Region. Although there are observable differences percent). Table 11 presents descriptive statistics on in the value of family and hired labor, they are small the main variables used in the analysis. when compared to the observable differences in the cost of the other production inputs. Some aspects of the data should be noted. A UPA in the Lowlands Region has, on average, an annual gross The average expenditure on intermediate inputs is production value (GPV) of 106,379 Bolivian bolivianos highest in the Sub-Andean Region; it is slightly higher (Bs), which is more than twice the national average, than the average input expenditure in the Lowlands, almost four times the average of the Sub-Andean Re- even though the average GPV in the Sub-Andean Re- gion, five times the average of the Amazon Region, and gion is much lower. It may be that Sub-Andean farm- almost ten times the average of the Highlands Region. ers are producing particularly input-intensive crops These figures partly reflect differences in harvested such as potatoes, fruit, vegetables, or grapes. The area among regions—the trend for this variable re- Highlands Region has the lowest average expenditure sembles the trend in average GPV—which reflect the on intermediate inputs, corresponding to one-third of high degree of land fragmentation in the Highlands the national average. (also in the Sub-Andean and Amazon Regions) and the scope for extensive land use in the Lowlands. As expected, only 3 percent of UPAs in the Amazon Region use tractors, whereas more than 25 percent of producers in the other regions use them. The Low- Table 10. Sample Size: lands Region uses the most tractor energy.22 The rel- ENA 2015 atively high use of tractors in the Highlands probably Number of reflects the large expansion of quinoa production and Macro-Region % observations the introduction of mechanization that occurred early Highlands 1917 18 % in the 2010s. Amazon 534 5% Lowlands 2886 28 % These data alone provide relatively little insight into Sub-Andean 5088 49 % the comparative efficiency or productivity of agricul- Total 10 425 100 % ture in the four macro-regions. Clearly the average producer in the Lowlands has the highest GPV and Source: Author’s elaboration the average producer in the Highlands has the lowest, and the regions where production is highest are the 21 From the data available on the numbers and daily wages of permanent and temporary workers, it was possible to approximate the value of hired labor. The value of family labor was estimated by calculating the average daily wages at the macro-regional level and using data on the number of household members who reported agriculture-based work. 22 The indicator “intensity of tractor use,” which is used in the stochastic frontier analysis, expresses energy use in terms of horsepower. Data for this indicator are based on the type of tractor used by each UPA and the reported energy capacity (horsepower). A comparison of data from the Agricultural Census 2013 and the ENA 2008 showed consistent findings on the levels of tractor use in the different macro-regions. In all cases, tractor use is highest in the Lowlands and Highlands (40–50 percent of UPAs), followed by the Sub-Andean Region and, at a considerable distance, the Amazon Region. 72 Table 11. Descriptive statistics of the main variables used in the analysis [Average per Agricultural Production Unit-UPA] Highlands Amazon Lowlands Sub-Andean Bolivia Gross production value (Bs-Bolivianos) 11 923 21 200 106 379 27 213 45 123 Harvested area (ha) 1,9 2,5 9,4 2,3 4,1 Labor value (Bs) 28 135 32 646 30 436 36 457 33 113 Family labor value (Bs) 24 380 27 626 22 896 23 151 23 537 Hired labor value (Bs) 3755 5020 7540 13 306 9576 Expenditure on intermediate inputs (Bs) 3883 6355 14 424 15 556 12 618 Expenditure on seeds (Bs) 1716 6242 8521 12 061 8900 Expenditure on organic fertilizer (Bs) 1726 55 397 1556 1204 Expenditure on pesticides (Bs) a 308 0 62 289 217 Expenditure on chemical fertilizer (Bs) 17 8 574 597 453 Expenditure on insecticides (Bs) 58 15 2492 470 908 Expenditure on fungicides (Bs) 26 13 523 243 265 Expenditure on herbicides (Bs) 32 22 1855 341 670 % using tractor 48 % 3% 41 % 28 % 34 % Energy used on tractor (HP) 35,8 4 42,2 24 30 Source: ENA (2015) and author’s elaboration a The category “pesticides” is understood in ENA as “organic pesticides.” regions with the highest investment in intermediate such as distance to the closest city or level of educa- inputs or the largest production area. It could well be tion were not available for the analysis. the case that a higher level of production is the result What is noteworthy about the variables that are con- of using more inputs or more land, and not necessar- sidered to contribute to inefficiency? One outstanding ily the result of any higher efficiencies in the use of feature of the data in Table 12 above is the low level these production factors. of collective action through participation in produc- An SFA and MFE can provide information to make er organizations. The Lowlands Region has the most more reliable inferences about the productive effi- producers who participate in an organization (14.4 ciency of the different macro-regions and types of percent), and the Amazon Region has the fewest (7.1 producer. The additional variables used for the SFA percent). Access to credit is also low throughout Bo- and MFE analyses are presented in Table 12. These livia, particularly in the Highlands, where only 6.3 per- variables were selected from a wider set of variables cent of producers said they had obtained some type because they showed the greatest consistency in the of credit during the last 12 months. Most producers regressions performed to develop a robust specifi- reported receiving information on climate or weath- cation of the model. Unfortunately, critical variables er events important for agriculture, with the lowest percentage in the Highlands. At the same time, access 73 Table 12. Descriptive: Determinants of inefficiency Highlands Amazon Lowlands Sub-Andean Bolivia Female household head (%) 13,20 % 10,60 % 11,60 % 12,70 % 12,40 % Membership in organization (%) 11,90 % 7,10 % 14,40 % 11,00 % 11,90 % Access to credit (%) 6,30 % 8,00 % 18,90 % 15,30 % 14,20 % Average household size 3,63 4,51 4,2 3,87 3,94 Access to telephony (mobile phone 35 % 34 % 50 % 43 % 43 % and/or landline) (%) Average agricultural landholding 1,9 2,5 9,4 2,3 4,1 Information received on climatic 65 % 72 % 72 % 73 % 71 % events in the agricultural sector (%) Access to irrigation (%) 29 % 5% 7% 33 % 24 % Source: Author’s elaboration Table 13. Determinants of meta-technology efficiency Sub- Highlands Amazon Lowlands Bolivia Andean Altitude (masl) 2855 1241 1706 2284 2184 Average slope (%) 42,2 31,2 45,4 37,5 40,2 Average precipitation (mm) 70,7 113,2 99,5 84,2 87,2 Average temperature (°C) 15,6 21,2 19,7 17,2 17,8 Source: Author’s elaboration to a telephone (landline or mobile phone) is still rela- 4.4. Analyzing differences in technical tively low at the national level (43 percent), although efficiency and meta-technical 50 percent of producers in the Lowlands Region had efficiency by macro-region, major access. The percentage of households headed by fe- crops and farm size males was slightly higher in the Highlands and the Sub-Andean Region, although it does not exceed 15 4.4.1. Stochastic frontier analysis by percent in any region. macro-region A number of geographical and climate variables in- Stochastic frontier models were estimated for each fluence the meta-frontier for technology efficiency macro-region, based on a Cobb-Douglas production (Table 13). As discussed, these variables are likely to function in which the dependent variable is the GPV explain some of the differences in use of agricultural at the producer level (in log), and the independent technology in each region. variables are harvested area and the expenditure on inputs (labor, intermediate inputs, machinery, in logs). Table 14 below shows the results (specifically, 74 Table 14. Elasticities and results of the stochastic frontier estimation by region Results of the stochastic frontier estimation (Dependent variable: Log of gross agricultural production value) Independent variables Highlands Amazon Lowlands Sub-Andean 0,2699*** 0,1741*** 0,2083*** 0,3242*** Log (Inputs) (0,0184) (0,0318) (0,0169) (0,0102) 0,0592* 0,0173 0,1148*** 0,0909*** Log (Labor) (0,0312) (0,0291) (0,0205) (0,0178) 0,5811*** 0,7168*** 0,6574*** 0,5242*** Log (Land) (0,0203) (0,0628) (0,0270) (0,0141) -0,1661* 0,7291** 0,2918*** 0,0169 Log (Capital) (0,0892) (0,3676) (0,0686) (0,0600) Returns to scale 0,74 1,64 1,27 0,96 Observations 1889 510 2698 5045 Chi-square 2274,6 334,0 1983,1 4429,2 Log Likelihood -2437 -737 -4170 -7345 σv 0,689 0,698 1,050 0,905 Prob. > chi2 0,0000 0,0000 0,0000 0,0000 Source: Authors elaboration based on ENA (2015) and other sources the average elasticities of the production factors in elasticity of production with respect to land ranges each macro-region). from 0. 52 to 0.72. The estimated elasticities of all inputs exhibit the The results reveal important differences across mac- expected sign.23 The Sub-Andean, Highlands, and ro-regions in the elasticity of the value of production Lowlands Regions exhibit the highest elasticity in the with respect to changes in farm labor. For instance, value of production with respect to changes in the the Lowlands Region exhibits an elasticity value that mix of intermediate inputs, implying that producers is more than twice the value observed in the High- in those macro-regions have greater flexibility to in- lands, indicating that an increase in the number of crease production by changing the mix of inputs. By workers would be more effective in raising the val- contrast, in the Amazon Region the value of produc- ue of production in the Lowlands Region than in the tion is much less sensitive to changes in the mix of Highlands. Farms in the Highlands are small and rely inputs. As expected, in all macro-regions the value of heavily on the use of family labor, whereas Lowland production is most sensitive to changes in land: the farms are bigger and labor use per unit of land is relatively low. 23 The only exception is capital in Highlands, which exhibits a negative sign, perhaps because of the limitations of the variable, which considers only tractor capacity (horsepower) and not the amount of energy effectively used on the farm. In addition, the capital variable does not include the use of any other type of machinery by the UPA (such infor- mation was not collected by ENA). 75 Figure 23. SFA and MFE by Macro-region a) Average and distribution of technical efficiency by b) Histogram of technical efficiency by Macro-Region Macro-Region Highlands Amazon .12 .12 .1 .1 Distribution .08 .08 Technical Efficiency (TE) .06 .06 .04 .04 Macro- Number of Stand. Average Min Max .02 .02 Region observations Dev. 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Highlands 1,889 0.56 0.17 0.03 0.99 Amazon 510 0.46 0.19 0.01 0.85 Lowlands Sub-Andean .12 .12 Lowlands 2,698 0.63 0.12 0.07 0.92 .1 .1 Distribution .08 .08 Sub- 5,045 0.57 0.12 0.02 0.85 .06 .06 Andean .04 .04 .02 .02 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Technical E ciency Technical E ciency Source: Authors’ elaboration Figure 23 below shows the results of the technical ef- cal efficiency in each macro-region, which are shown ficiency estimates for each macro-region. While these in Table 15. The results show that irrigation signifi- efficiency measures are not comparable across re- cantly reduces inefficiency in all four regions. This ef- gions, their distribution can be analyzed, as shown in fect occurs in a context in which irrigation adoption is the histogram displayed in Figure 23b. The technical quite limited, both in the Andean areas (29 percent of efficiency of UPAs is more heterogeneous in the High- Highland UPAs and 33 percent of Sub-Andean UPAs) lands and Amazon Regions, particularly the latter (as and lower areas (7 percent of Lowland UPAs and 5 shown by the higher standard deviation and longer percent of Amazon UPAs). From a policy perspective, tails in the distribution). In the Lowlands and Sub-An- and considering the technical possibilities and irri- dean Regions, a tighter distribution is observed, in- gation potential in each macro-region, investments dicating that efficiency values tend to concentrate in irrigation infrastructure appear likely to improve around the mean value. The differences in the disper- the technical efficiency of farmers, particularly those sion of technical efficiency within each macro-region with crops exposed to drought risks and crops that suggest that policies to increase agricultural produc- are seasonally grown. tivity will need to be tailored to each region. Access to credit is effective in reducing inefficiency in For the SFA, technical efficiency is considered to be a the Lowlands and the Sub-Andean Regions, which are function of a set of variables (described in the previ- already the regions with the greatest access to loans. ous section) that are related to the characteristics of Credit programs are expected to be effective in regions producers and UPAs, as well access to services such with higher economic dynamics, and among producers as credit and information. This approach makes it who are more integrated into the market. In regions possible to identify the main determinants of techni- where traditional and mainly subsistence agriculture 76 Table 15. Results: Determinants of technical efficiency in Macro-Regions Second Stage (dependent variable = technical inefficiency score) Independent variables Highlands Amazon Lowlands Sub-Andean 0,4841*** 0,1992 0,6580*** 0,1545 Female producer (1 = yes, 0 = no) (0,1388) (0,2644) (0,1990) (0,1164) -0,4303** -0,0858 -0,2265 -0,3958** Producer belongs to association (1 = yes, 0 = no) (0,2129) (0,3404) (0,2913) (0,1873) 0,1611 -0,3249 -0,4107* -0,6039*** Producer obtained credit (1 = yes, 0 = no) (0,2489) (0,3289) (0,2279) (0,1740) -0,0473* -0,0066 0,0039 -0,0091 Household size (# members) (0,0284) (0,0351) (0,0340) (0,0198) -2,5652*** -0,1903 -3,2679 -0,2984 Producer has access to telephony (1 = yes, 0 = no) (0,4385) (0,4243) (0,0000) (0,2757) -0,1545*** 0,0168 -0,0158 0,0138 Harvested area (hectares) (0,0444) (0,0185) (0,0097) (0,0085) Receives information on climatic events in the -0,2266** 0,4708** 0,2506 -0,1385 agricultural sector (1 = yes, 0 = no) (0,1102) (0,1958) (0,1600) (0,0872) -0,2493** -1,9964** -1,3202*** -0,3063*** Producer has access to irrigation (1 = yes, 0 = no) (0,1232) (0,7792) (0,4347) (0,1067) 1,2692*** 0,2728 0,8385*** 0,0913 Constant (0,1920) (0,3263) (0,2799) (0,2010) Observations 1889 510 2698 5045 Chi-square 2274,6 334,0 1983,1 4429,2 Log Likelihood -2437 -737 -4170 -7345 σv 0,689 0,698 1,050 0,905 Prob. > chi2 0,0000 0,0000 0,0000 0,0000 Source: Author’s elaboration. *** Significant at 99%, ** significant at 95%, * significant at 90%. prevails (such as the Highlands), such policies will tion to overcome barriers in obtaining credit, negotiate have limited impacts if other pressing gaps in connec- cheaper volume-based purchases, take advantage of tivity and access to basic services are not addressed. economies of scale, share risks, and more easily gain access to technical assistance and other services. This Collective action through participation in a producer result suggests that policies and programs that pro- organization reduces inefficiency, especially in High- mote organization and collective action among produc- lands and Sub-Andean Regions, where land holdings ers, especially in the Andean areas, are more likely to are highly fragmented. Producers who form an organi- improve the livelihoods of peasant families than poli- zation, association, or cooperative are in a better posi- cies and programs targeting individual farmers. 77 Figure 24. Meta-frontier production function for different production systems Output y M Metafrontier M 3 Frontier (System 3) 2 4 Frontier (System 4) 2 e-uji 2 Frontier (System 2) Y (System 2) Frontier (System 1) xi Input x Source: Melo-Becerra and Orozco-Gallo (2017) Access to a telephone (either a landline or mobile who use different technologies. In the first stage of phone) is also relevant for reducing inefficiency, spe- the analysis, standard SFA techniques are used to es- cifically in the Highlands, where access (35 percent) timate the specific frontier of each group. In the sec- is lower than in all other macro-regions. The diffusion ond stage, the meta-frontier is estimated using the of mobile phone services in particular widens access pooled estimates of the regional frontiers. Since this to information for poor producers who are far away second-stage estimation is still based on a stochastic from agricultural service centers; otherwise they frontier regression, the technology gaps represent- would obtain information only through sporadic visits ed by the one-sided term can be further specified as by extension workers and other technicians. a function of geographical/environmental variables beyond the control of firms and a vector of unknown Female-headed UPAs have lower levels of efficien- parameters to be estimated, including a set of natu- cy. It may be that these households lack training in ral and geographic factors that explain differences production techniques, encounter more severe time in technology adoption. The meta-frontier efficien- constraints, or lack access to remunerative markets. cy (MFE) is the measure of the technical efficiency of each production unit relative to the meta-frontier 4.4.2. Stochastic meta-frontier (Figure 24). efficiency by macro-region Figure 25 below shows the technology gap estimates A stochastic MFE analysis provides information on and the final meta-efficiency measures. The region differences in efficiency across regions by compar- with the highest meta-technology efficiency is the ing technical efficiency among groups of producers 78 Figure 25. Meta-frontier estimations a) Technology gap ratio and meta-efficiency b) Histogram of meta-technology efficiency by macro-region Highlands Amazon .12 .12 .1 .1 Distribution .08 .08 .06 .06 Region TE TGR MTE .04 .04 .02 .02 Highlands 0.56 0.70 0.39 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Amazon 0.46 0.94 0.44 Lowlands 0.63 0.87 0.55 Lowlands Sub-Andean .12 .12 Sub-Andean 0.57 0.87 0.50 .1 .1 Distribution .08 .08 Total 0.58 0.84 0.49 .06 .06 .04 .04 .02 .02 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Technical E ciency Technical E ciency Source: Authors’ elaboration Lowlands, followed by the Sub-Andean Region. The UPAs. To explore this possibility, differences in the Highlands have the lowest meta-technology efficiency characteristics of Highland and Amazon UPAs were values. The largest technology gap ratio is observed analyzed by meta-technology efficiency quintile (see in the Highlands. The lower technology gap ratios in Table 16). the other regions indicate that they are closer to the There are some differences between the more ef- national technological frontier. The histogram of the ficient and less efficient Highland UPAs. The most distribution of meta-technology efficiency by region productive UPAs are larger, focus more on quinoa (Figure 25b) shows that producers in the Lowlands production and less on potatoes, are less likely to be and Sub-Andean Regions display greater homoge- female-headed, have better access to credit and mo- neity in meta-technology efficiency. This finding indi- bile phones, and are more likely to belong to a pro- cates that future efficiency gains in these regions will ducer organizations. depend on technological improvements that move the frontier of these regions and at the same time In the Amazon Region, there are no significant differ- increase productivity. ences in farm size between UPAs with different levels of technology efficiency. However, the most efficient Because meta-technology efficiency within the High- UPAs devote less land to bananas and plantains and lands and Amazon Regions is so heterogeneous— have more access to permanent sources of irrigation. in other words, very efficient producers are found alongside very inefficient producers—one option to These results suggest that improvements in agricul- enhance productivity in these regions may be to en- tural productivity could be promoted by improving the able the least efficient UPAs to adopt the productive access of less efficient producers to the practices and practices and technologies used by the most efficient 79 Table 16. Characteristics of producers by Macro-Region and Meta- Technology Efficiency quintile a) Highlands ALTIPLANO/HIGHLANDS Technology Efficiency Quintile Variable 1 2 3 4 5 Average Efficiency Score 0,29 0,49 0,58 0,66 0,77 Harvested area (ha) 0,96 0,97 1,11 1,49 4,98 % growing POTATOES 73 % 74 % 73 % 73 % 52 % % growing QUINOA 21 % 20 % 21 % 32 % 51 % % growing WHEAT 12 % 20 % 18 % 10 % 3% % female producers 23 % 15 % 12 % 8% 8% % belonging to association 4% 6% 8% 17 % 24 % % accessed credit 3% 3% 4% 10 % 11 % % with access to mobile phone 28 % 33 % 35 % 37 % 42 % % received climatic information 56 % 64 % 66 % 66 % 74 % % equipped for irrigation 24 % 30 % 28 % 31 % 30 % Household size 3,17 3,55 3,68 3,94 3,81 b) Amazon AMAZON Technology Efficiency Quintile Variable 1 2 3 4 5 Average Efficiency Score 0,18 0,37 0,48 0,59 0,71 Harvested area (ha) 3,05 2,34 1,92 2,46 2,53 % growing CASSAVA 24 % 37 % 35 % 27 % 39 % % growing BANANA/PLANTAIN 47 % 49 % 33 % 39 % 20 % % growing CORN 29 % 50 % 36 % 48 % 31 % % growing RICE 18 % 28 % 33 % 34 % 27 % % female producers 13 % 9% 6% 10 % 16 % % belonging to association 6% 7% 6% 7% 10 % % accessed credit 6% 5% 11 % 10 % 9% % with access to mobile phone 35 % 31 % 34 % 34 % 34 % % received climatic information 87 % 75 % 70 % 58 % 72 % % equipped for irrigation 0% 0% 1% 3% 21 % Household size 4,45 4,49 4,36 5,09 4,17 Source: Authors’ elaboration 80 technologies used by more efficient producers. For The crop-level analysis mirrors the results of the example, irrigation in specific areas of the Amazon macro-region analysis. Figure 26 suggests that pro- Region could contribute significantly to improvements ducers of potatoes and wheat in the Sub-Andean in the agricultural productivity of less efficient UPAs. Region are more productive than producers of those In the Highlands Region, the initial SFA suggests that crops in the Highlands. The greater meta-technology a set of complementary measures is required to en- efficiency of the Sub-Andean Region would be ex- hance efficiency, such as the promotion of producer plained by the lower technology gap ratio (better pro- organizations and improvements in access to land, duction processes and/or external conditions) rather connectivity, and agricultural information. than by better use of productive resources (at least in the case of wheat). To improve productivity, pota- 4.4.3. Stochastic meta-frontier to and wheat producers in the Highlands Region will efficiency by macro-region and major need either to adopt improved technology or switch crops to higher-value crops. Although the previous analysis provides a general In the case of rice, there are no significant techno- picture of differences in average technical efficien- logical differences between the Amazon, Lowlands, cy between macro-regions, it is very likely that these and Sub-Andean Regions, although the technical differences are related to different technologies and efficiency of Lowland and Sub-Andean producers is characteristics of the mix of crops grown in each higher than the efficiency of producers in the Amazon macro-region. A more detailed analysis that explores Region. In the Amazon Region, the distribution of me- differences in efficiency levels across producers of ta-technical efficiency is flatter, and the peak of the the same crop who are located in different macro-re- distribution is located farther away from the frontier, gions could clarify the reasons for the productivity indicating that efficiency among Amazon rice farm- differentials and potentially reinforce the results of ers is more variable and that relatively large number the previous analysis. For instance, it would help to of farmers are operating at some distance from the determine whether the observed differences in pro- frontier. Policy measures to increase their efficiency ductivity of a given crop across regions are the result would need to focus on improving their crop manage- of differences in producers’ ability to make optimal ment capacity—for instance, by promoting access to use of productive resources or the result of differenc- irrigation, as suggested by the initial SFA. es in technology or context. 4.5. Analyzing differences in technical This type of detailed analysis at the crop level is pos- efficiency and meta-technical sible only for crops that are grown in more than one efficiency by farm size and major macro-region and for which there are a reasonably crop large number of observations in the ENA 2015 data. Crops that meet these conditions are potatoes and 4.5.1. Farm size and efficiency wheat (produced in the Highlands and Sub-Andean Regions) and rice (produced in the Lowlands, Amazon, After exploring the differences in technical efficiency and Sub-Andean Regions). Important crops that are across macro-regions, the analytical focus now turns not considered in the analysis, since they are most- to the relationship between farm size and meta-tech- ly produced in one macro-region, include soybeans nical efficiency. The analysis takes advantage of the (Lowlands) and quinoa (Highlands). previously estimated technical efficiency measures 81 Figure 26. Technical efficiency: by Crop across Macro-Regions a) Stochastic frontier and meta-frontier by crop and farm size WHEAT POTATO RICE Region TE TGR MTE TE TGR MTE TE TGR MTE Highlands 0.91 0.39 0.36 0.59 0.57 0.34 Amazon 0.53 0.99 0.53 Lowlands 0.76 0.97 0.74 Sub-Andean 0.74 1.00 0.74 0.55 1.00 0.55 0.89 0.97 0.85 Total 0.77 0.90 0.68 0.57 0.82 0.46 0.74 0.97 0.72 b) Histograms of meta-technology efficiency of crops and Macro-Regions WHEAT POTATO RICE Highlands Highlands Amazon .2 .2 .15 Dsitribution .16 .1 .15 .12 .05 .1 0 .08 0 .2 .4 .6 .8 1 .05 .04 Lowlands 0 0 .15 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 .1 Sub-andean Sub-andean .05 .2 .2 0 0 .2 .4 .6 .8 1 .15 .16 .12 Sub-Andean .1 .15 Dsitribution .08 .1 .05 .04 .05 0 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Source: Authors’ elaboration for the macro-regions to explore the relationship Region technical efficiency does not vary significantly between UPA size and technical efficiency. It calcu- across the farm-size quintiles. lates average efficiency levels of UPAs (grouped by While these results suggest that the relationship be- quintile of cultivated area to represent farm size) in tween farm size and meta-technical efficiency varies each macro-region (see Figure 27 below). This anal- across macro-regions, they should not be interpret- ysis shows higher efficiency among larger UPAs in ed to mean that this relationship always holds for all the Lowlands and Highlands Regions, and to some producers within each region. In fact, it is highly likely extent in the Amazon Region. It is worth mentioning, that the relationship between farm size and efficiency however, that smaller UPAs in the latter two regions is strongly influenced by the type of crop being grown, also present relatively higher technical efficiency and the next analysis focuses on this issue. than medium-sized UPAs. Finally, in the Sub-Andean 82 Figure 27. Average meta-technical efficiency by land quintile by macro- region 0.70 0.60 0.60 0.56 0.53 0.54 0.53 0.52 0.48 0.49 0.50 0.49 0.50 0.50 0.47 0.44 0.43 0.45 0.38 0.40 0.38 0.40 0.34 0.36 0.30 0.20 0.10 0.00 Highlands Amazon Lowlands Sub-Andean Q1 Q2 Q3 Q4 Q5 Source: Authors’ elaboration 4.5.2. Efficiency in relation to farm size Soybeans and crop Figure 28 below shows a positive relationship be- The efficiency analysis at the crop level was carried tween UPA size and the meta-technology efficiency out separately for producers of four crops—soybeans, of soybean producers, who are mainly located in the rice, quinoa, and potatoes. Those crops were selected Lowlands Region. Larger UPAs (40–100 hectares) dis- because they were grown by the highest number of play greater meta-efficiency in the use of productive producers or had the largest harvested area. For this resources compared to smaller UPAs. This finding three-stage analysis, the producers of each crop were is consistent with the extensive nature of soybean first classified into three UPA size ranges (small, me- production, which allows for economies of scale. In dium, and large, relative to the universe of producers contrast, UPAs of less than 20 hectares are not only of the crop).24 Second, different stochastic production the most inefficient but also display highly heteroge- frontiers were estimated for each of the size strata of neous meta-efficiency levels. the same crop. Third, a stochastic meta-frontier was These results are in line with the positive relationship estimated for each crop, including the three groups previously observed between farm size and efficiency of producers classified by UPA size. This procedure in the Lowlands Region (see Figure 18, Chapter 3). makes it possible to analyze the relationship between They also suggest ample scope to foster productivity UPA size and meta-technology efficiency (comparable through policies targeting small-scale producers in measure) for the selected crops. the Lowlands, because increasing their productivity 24 The cut-off points for each farm-size class vary by crop due to the marked differences observed in extensive land use. The goal was to obtain farm-size groups that were relatively balanced in terms of the number of observations (produc- ers) to ensure that robust stochastic frontier estimates for each group. 83 Figure 28. SOYBEAN: Technical efficiency a) Soybean: Histogram of meta-technology b) Soybean: results stochastic frontier and efficiency meta-frontier by crop and farm size Small .15 Distribution .1 .05 0 0 .2 .4 .6 .8 1 Medium UPA Size TE TGR MTE Distribution .15 0 to 20 ha 0.49 0.89 0.43 .1 20 to 40 ha 0.69 0.88 0.61 .05 40 to 100 ha 0.75 0.89 0.66 0 0 .2 .4 .6 .8 1 Total 0.67 0.89 0.59 Large .15 Distribution .1 .05 0 0 .2 .4 .6 .8 1 MTE Source: Author’s elaboration would bring higher gains in overall technical efficien- level of efficiency than large farms (2–10 hectares), cy to the region. Figure 28 shows that the great ma- although levels of efficiency within all three groups jority of small-scale soybean producers (with UPAs are highly dispersed. of less than 20 hectares) operates far from the me- The high dispersion of MTE across all farm-size ta-frontier. A strategy that increases the access of classes can be interpreted to mean that there is an low-efficiency producers to technology used by highly opportunity to reduce these efficiency gaps through efficient soybean producers, such as irrigation and policies that bring less-efficient quinoa producers credit, could be a good alternative (as suggested by closer to the meta-frontier without the need for the initial SFA). drastic technological changes. Examples of such ini- tiatives include improving access to irrigation and Quinoa communication networks (telephones and informa- For quinoa, which is almost exclusively grown in the tion) and initiatives to organize producer associations. Highlands Region, the analysis shows no significant These types of policy instruments were also suggest- technological differences between the three farm- ed in light of the results of the macro-regional SFA for size groups. However, small farms (0–1 hectare) and the Highlands Region. medium farms (1–2 hectares) show a higher average 84 Figure 29. QUINOA: Technical efficiency a) Quinoa: Histogram of meta-technology b) Quinoa: results stochastic frontier and meta- efficiency frontier by crop and farm size Small .15 Distribution .1 .05 0 0 .2 .4 .6 .8 1 Medium .15 UPA Size TE TGR MTE 0 to 1 ha 0.57 0.99 0.56 Distribution .1 1 to 3 ha 0.51 0.99 0.50 .05 3 to 10 ha 0.33 0.99 0.33 0 0 .2 .4 .6 .8 1 Total 0.62 0.97 0.60 Large .15 Distribution .1 .05 0 0 .2 .4 .6 .8 1 MTE Source: Author’s elaboration Potatoes efficient, while the meta-efficiency of producers with medium and large farms was slightly lower. Given the large number of potato producers surveyed for the ENA 2015 (2,752 producers),25 the analysis These results should be interpreted alongside the was conducted separately for the Highlands and results described in Section 4.4.3, which show that Sub-Andean Regions. The results, presented in Figure meta-technical efficiency is higher among Sub-An- 30, reveal important contrasts among macro-regions dean potato producers due to a lower technology gap in the relationship between farm size and meta-tech- ratio (better production processes and/or external nical efficiency. On the one hand, a positive relation- conditions) and lower among Highland potato pro- ship is seen between farm size and meta-efficiency ducers. These findings suggest the need for differ- in the Highlands Region, where the largest produc- entiated strategies to achieve productivity growth, ers (2–10 hectares) were the most efficient and the particularly among smallholders from the Highlands smaller UPAs (with less than 1 hectare) were the and Sub-Andean Regions. For instance, policy tools least efficient. On the other hand, differences in the to expand the technical production frontier of the average efficiency of producer groups with different Highlands Region should be considered, such as pro- farm sizes were smaller in the Sub-Andean Region. grams that promote the adoption of better agricul- Producers with the smallest holdings were the most tural technologies and machinery, as well as policies 25 1,228 producers in the Highlands and 1,524 producers in the Sub-Andean region. 85 Figure 30. POTATO: Technical efficiency a) Potato: Histogram of meta-technology b) Potato: results stochastic frontier and efficiency meta-frontier by crop and farm size Highlands Sub-andean Highlands Small Small .15 .15 UPA Size TE TGR MTE Distribution .1 .1 0 to 1 ha 0.54 0.86 0.46 .05 .05 0 0 1 to 3 ha 0.94 0.55 0.51 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 3 to 10 ha 0.76 0.87 0.66 Medium Medium .15 .15 Total 0.62 0.81 0.49 Distribution .1 .1 .05 .05 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Sub-andean Large Large UPA Size TE TGR MTE .15 .15 0 to 1 ha 0.54 0.98 0.53 Distribution .1 .1 .05 .05 1 to 3 ha 0.47 0.99 0.46 0 0 3 to 10 ha 0.48 0.99 0.48 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 MTE MTE Total 0.51 0.98 0.50 Source: Authors’ elaboration that strengthen the agricultural innovation system. nology gap ratio. In other words, the scale of produc- Although these policies would also have value in the tion does not appear to be a determinant of efficiency sub-Andean Region, the high dispersion in efficiency levels in rice production. At the same time, significant levels among small-scale potato producers presents heterogeneity is observed in farm efficiency levels in an opportunity to raise productivity without improve- all groups of producers, regardless of macro-region ments in current farm technology. In this context, pol- or farm size. For example, producers in the upper icies that push smallholders closer to the frontier can quartile for technical efficiency on average have an make a difference, such as initiatives that encourage efficiency level that is twice as high as the efficiency producers to organize while fostering better access level of the average producer in the lower quartile. to credit through innovative policy tools. Access to ir- Interventions that would enable the most inefficient rigation was also identified as an important driver of farmers to approach the efficiency levels of the most technical efficiency in the Sub-Andean Region in the efficient farmers would effectively double their effi- SFA by macro-region. ciency and result in large productivity gains to the rice sector. Rice Bringing less-efficient rice producers closer to the The analysis found no significant differences in the production frontier will require policies designed meta-efficiency of rice producers in different farm- specifically for the context and needs of each region. size groups (Figure 31) and no differences in the tech- As discussed, initiatives that encourage producers to 86 Figure 31. RICE: Technical efficiency a) Rice: Histogram of meta-technology efficiency b) Rice: results stochastic frontier and meta-frontier by crop and farm size Highlands Sub-andean Highlands Small Small .15 .15 UPA Size TE TGR MTE Distribution .1 .1 0 to 1 ha 0.54 0.86 0.46 .05 .05 0 0 1 to 3 ha 0.94 0.55 0.51 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 3 to 10 ha 0.76 0.87 0.66 Medium Medium .15 .15 Total 0.62 0.81 0.49 Distribution .1 .1 .05 .05 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Sub-andean Large Large UPA Size TE TGR MTE .15 .15 0 to 1 ha 0.54 0.98 0.53 Distribution .1 .1 .05 .05 1 to 3 ha 0.47 0.99 0.46 0 0 3 to 10 ha 0.48 0.99 0.48 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 MTE MTE Total 0.51 0.98 0.50 Source: Authors’ elaboration organize would be effective for improving efficiency reduce the productivity and efficiency gaps in each in the Sub-Andean Region, where land holdings are specific context. highly fragmented. Improving access to capital in the The econometric analysis reveals important gaps in Lowlands and Sub-Andean Regions would contribute efficiency and meta-technology efficiency, both be- to productivity growth in those contexts. Access to ir- tween and within macro-regions. For example, pro- rigation appears to be a good alternative for all three ducers in the Highlands and Amazon Regions display macro-regions to improve growth in agricultural pro- lower average levels of meta-technical efficiency and ductivity, based on the SFA at the macro-regional lev- high heterogeneity in efficiency levels. In contrast, el discussed earlier. higher levels of meta-efficiency and a tighter distri- bution are observed in the Lowlands and Sub-Andean 4.6. Conclusions and Regions, indicating that efficiency values tend to con- recommendations centrate around the mean value. By exploring the main determinants of technical ef- These results suggest several policy measures, ficiency in Bolivia’s four macro-regions, and by high- differentiated by macro-region, to enhance the effi- lighting important differences in efficiency between ciency and productivity of Bolivian agriculture. In the regions, crops, and groups of producers, the analyti- Lowlands and Sub-Andean Regions, for example, a cal work described in this chapter has helped to iden- focus on technological improvements could move the tify opportunities and inform the design of policies to 87 production frontier toward higher levels of efficien- tive in reducing technical inefficiency. Irrigation would cy and productivity, and in certain instances (small- enable producers to control and intensify their use of scale soybean producers) productivity gains can be water to improve productivity and reduce the uncer- achieved through interventions focused on enabling tainty linked to climate cycles. less-efficient farmers to approach the efficiency lev- Although the analysis of technical efficiency across el of the most productive ones. In contrast, because the macro-regions provides a general idea of mea- of the high level of heterogeneity in meta-technology sures that could improve productivity in each region, efficiency in the Highlands and Amazon Regions, a a more targeted MFE analysis focusing on particu- practical option for increasing productivity would be lar crops and farm-size groups helps to refine those to ensure that the production practices and technol- alternatives for different regions and groups of pro- ogies used by the most efficient UPAs are transferred ducers. For example, the analysis shows that the to the less-efficient ones (such as the medium-scale technical efficiency of soybean production is higher quinoa producers in the Highlands). on larger farms compared to smaller ones in the The analysis indicates that investments that could Lowlands Region, probably as a result of economies effectively increase technical efficiency and bring of scale. The high dispersion in the distribution of less-efficient farmers close to the regional frontier efficiency among small-scale soybean producers include support for collective action, credit programs, suggests that policies that encourage the transfer mobile phones, and irrigation. The efficacy of each of knowledge and technology from high-efficiency type of investment will vary by macro-region. For ex- to low-efficiency producers in this group can make ample, collective action will be important for reducing a strong contribution to productivity growth in the technical inefficiency in areas where land holdings sector. The MFE confirms the previous finding that are highly fragmented (the Highlands and Sub-Ande- improved access to credit and irrigation adoption an Region). Programs and projects that foster collec- would be effective instruments to support growth in tive action are more likely to improve the income of productivity in the Lowland context. peasant families than programs that target individu- For potatoes, which are mainly grown in the High- als. On the other hand, improving access to credit is lands and Sub-Andean Regions, the MFE points to more likely to be an effective investment in regions lower technical efficiency among potato producers with better market integration and more financial in- from the Highlands compared to producers in the stitutions (the Lowlands and Sub-Andean Regions). In Sub-Andean Region, who had a lower technology the Highlands Region, where access to mobile phones gap ratio. When results were disaggregated by farm and landlines is still limited,26 better diffusion of mo- size in each region, small-scale potato producers bile phone technology would enable information on were found to be less efficient in the Highlands Re- agricultural technology and markets to reach large gion but not in the Sub-Andean Region. From a pol- numbers of producers who otherwise have very spo- icy standpoint, these findings suggest the need for radic and limited access to this kind of information. different strategies to achieve productivity growth Finally, across all four macro-regions, investments to in these regions, particularly among smallholders. improve access to irrigation seem likely to be effec- 26 According to the 2013 agricultural census, 34 percent of the rural households reported having fixed or mobile telepho- ny in 2012. this percentage reached 82 percent in the urban areas. While the situation would have improved by 2015 (the year of the ENA), the access of more and more farmers to mobile phones provides enormous opportunities to reach them with knowledge and a range of services. 88 For instance, the efficiency of medium- and small- groups of rice producers, however, means that im- scale potato farmers in the Highlands could be im- portant productivity improvements could be achieved proved through measures to expand their technical by bringing the less-efficient rice producers closer production frontier, such as programs that promote to the production frontier. The specific policy mea- the adoption of better agricultural technologies sures will depend on the context and needs of each (peer to peer exchanges, access to new inputs and region. Initiatives to encourage producers to organize machinery), as well as programs to strengthen the are likely to be effective in improving efficiency in the agricultural innovation system. On the other hand, in Sub-Andean Region, given the highly fragmented land the Sub-Andean Region, the high dispersion in effi- holdings. Improving access to capital in the Lowlands ciency levels of small-scale potato producers can be and Sub-Andean Regions would enable less-efficient seen as an opportunity to raise productivity without producer to move closer to the production frontier. the need to improve current farm technology; in this And again, improving access to irrigation would be a case, policies should focus on bringing less-efficient good alternative for improving efficiency in all three producers closer to the technical production frontier macro-regions where rice is produced. through programs that increase collective action and Finally, it is important to acknowledge the limited access to irrigation and credit. scope of this exercise in terms of the set of variables It is worth mentioning that the MFE found no signif- that could be considered. The quality and scope of data icant differences in the efficiency of rice producers collected must improve to allow for more robust es- in different farm-size groups, and no differences in timations of productivity at disaggregated levels, in- the technology gap ratio. The substantial heteroge- cluding STF/MFE estimates, but also estimation of TFP. neity in farm efficiency levels observed among all 89 5. Competitiveness of Bolivia’s agriculture Photo: Programa de Alianzas Rurales (PAR), EMPODERAR In Bolivia only a small number of agricultural com- business dynamism, labor markets, market size, and modities are integrated with global value chains, and institutions Bolivia has notably divergent scores from even those commodities encounter constraints that the highest performer among its structural peers. hinder their competitiveness. Strong demand and Another important global indicator that highlights the supply forces have fueled the dramatic increase in influence of regulatory frameworks and administra- agricultural growth observed in the Santa Cruz re- tive procedures on business operations and private gion, but these forces have been largely absent in investments is the Doing Business Indictor (DBI). For other regions, where agriculture remains much less this indicator, Bolivia places in the bottom tercile, competitive. To understand the factors that determine ranking 156th of 190 countries, performing below or disrupt competitiveness, this chapter explores the regional and structural peers (except Cameroon) and competitiveness of agriculture in Bolivia in relation other countries in Central America, such as Guate- to global benchmarks and comparator countries. It mala, Honduras, and Nicaragua. The 2018 IDB report provides a general perspective on how to unlock the acknowledges Bolivia progress on trading across competitiveness of Bolivian agri-food systems and as borders by implementing the Sistema Único de Mod- well as specific assessments of the main constraints ernización Aduanera (SUMA), an automated customs and opportunities in key value chains for agriculture data management system. and agribusiness, including constraints related to geographical heterogeneity. To understand the influence of regulatory frame- works and administrative procedures on business 5.1. Competitiveness performance: operations and private investments in the agricul- Evidence provided by global tural sector in particular, the World Bank developed benchmarks the Enabling the Business of Agriculture (EBA) indi- cators. Insights into Bolivia’s performance from the 5.1.1. Common enablers 2017 EBA include: The Global Competitiveness Index (GCI) 2018 is cal- • Machinery: Bolivia ranks 52nd of 62 countries for culated based on 12 indicators that are considered to regulations dealing with machinery. The process influence competitiveness. Bolivia ranks in the bot- for registering a tractor is virtually nonexistent. tom third of the 140 countries compared in the GCI, holding the 105th position. Bolivia does not approach • Fertilizer: Bolivia ranks 45th, considerably below the GCI median score of 70 for any of the dozen “com- the other countries. In Bolivia, Ethiopia, and Kenya, petitiveness pillars” of the index. The country earns only the public sector is permitted by law to regis- its best marks for macro-economic stability (84) and ter new fertilizer products. financial system (85), and the worst for business • Seed: Bolivia ranks relatively well (25th) for seed dynamics (130), institutions (129), and labor market regulation, partly owing to its law requiring the (126). In the group of LAC counties, Bolivia ranks near labelling of seed containers and penalizing fraud- the bottom, outperforming only Venezuela and Haiti. ulent sales of mislabeled seed. Bolivia also has Compared to the best performer in the LAC Region, strong seed registration regulations and an effi- Bolivia must narrow important gaps, even in areas cient registration process, which is seldom seen where it ranks best (Table 17). Compared to its struc- in other countries. tural peers, Bolivia’s performance in relation to criti- cal competitiveness pillars is also disappointing. For 92 Table 17. Global Competitiveness Index 2018, selected results for Bolivia Difference Difference Difference Difference Innovation and Enabling from highest Human from highest from highest from highest Markets sophistication Environment ranked (a) Capital ranked (a) ranked (a) ranked (a) (a) LAC LAC LAC LAC LAC Product Business Institutions -22.4 Skills -13 -19.7 -19.3 Market Dynamism Innovation Infrastructure -19 Labor Market -15.7 -14.4 Capability Macroeconomic Financial -26.9 -25.5 Stability Systems Health -18.4 Market Size -16.5 b) Structural b) Structural b) Structural b) Structural Peer Peer Peer Peer Product Business Institutions Skills -14.5 -10.4 Market -8.8 Dynamism -21 (Ghana) (Kazakhstan) (Kazakhstan) (Kazakhstan) Innovation Infrastructures Labor Market -11.1 -18.3 Capability -5.8 (Kazakhstan) (Kazakhstan) (Ghana) Macroeconomic Financial Stability -7.2 Systems 2.6 (Kazakhstan) (Bolivia) Market Size Health (Bolivia) 4.2 -16.8 (Kazakhstan) Source: Global Competitiveness Index 2018 • Finance: Bolivia is doing remarkably well, ranking 5.1.2. Broad logistical constraints to 13th. The government has made a great effort to improved competitiveness implement regulations to facilitate financing for small-scale farmers over the past few years. Bo- Logistics is fundamental to value chain coordination livia is also one of only a few countries in which and the efficient movement of agricultural products financial cooperatives disclose the full cost of to market and inputs to producers. The Logistics Per- credit to loan applicants, participate in a deposit formance Index (LPI) is measures a country’s perfor- insurance system, and are allowed to merge to mance in relation to several components of trade create a new financial cooperative. logistics (Table 18). Bolivia scores below all but two of its structural peers (it is better than Cameroon in • Nondiscrimination law. Bolivia ranks among the timeliness, and better than Mongolia with regard to top-most countries for good practices in agriculture customs, infrastructure, international shipments, and related to nondiscrimination laws and regulations. tracking and tracing).27 27 Since 2007, the LPI has provided the means for benchmarking the competitiveness of logistics for 160 countries across the spectrum of country size and income level. 93 Figure 32. Ease of Doing Business 2016-17, selected results Ease of Doing Business Scores 2019 (Structural Peers) Ease of Doing Business Scores 2019 (Regional Peers) Cameroon (Rank 166) 47.78 Bolivia (Rank 156) 50.32 Ecuador (Rank 123) 57.94 Bolivia (Rank 156) 50.32 LAC Region 58.97 Ghana (Rank 114) 59.22 Paraguay (Rank 113) 59.4 Brazil (Rank 109) 60.01 Mongolia (Rank 74) 67.74 Peru (Rank 68) 68.83 Kazakhstan (Rank 28) 77.89 Colombia (Rank 65) 69.24 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 Source: EBA indicators 2018, World Bank Table 18. Logistics Performance Index 2018, selected results International Logistics Tracking and Customs Infrastructure Timeliness Shipments Competence Tracing Bolivia 2.32 2.15 2.54 2.21 2.13 2.74 Kazakhstan 2.66 2.55 2.73 2.58 2.78 3.53 Ghana 2.45 2.44 2.53 2.51 2.57 2.87 Mongolia 2.22 2.10 2.49 2.21 2.10 3.06 Cameroon 2.46 2.57 2.87 2.60 2.47 2.57 Lower Middle Income 2.37 2.37 2.57 2.50 2.59 2.96 Source: LPI World Bank, 2018 A study by the Economic Commission for Latin Amer- such as soybeans, these cost overruns weigh dispro- ica and the Caribbean (ECLAC/CEPAL) to gauge logis- portionately on competitiveness. The CEPAL study tical competitiveness of landlocked countries such as also confirms that the total distance of a shipment Bolivia highlights the effects of logistical constraints is not necessarily the best predictor of costs. The on agricultural sector performance. The study quan- substantial transaction costs incurred in transfers, tifies costs at each segment of the soybean export handling, demurrage, and delays significantly hamper chain involving logistics: pre-shipment; transfer to competitiveness.28 port; port, airport, or border entry-point; transport to Inefficiencies in logistics also affect the competitive- destination; and total inventory charges and financing ness of non-traditional exports. For the quinoa sup- costs. It finds that cost overruns can be as high as 21 ply chain, a World Bank study (2014) finds that the percent when shipping via river ports and over 23 length of time required to perform certain logistics percent for land shipments. For export commodities 28 CEPAL, 2014. Logistics inefficiencies in Latin American landlocked countries. Issue No. 335 - Number 7 / 2014. 94 tasks varies considerably from 741 to 1,023 hours. integrated vision and strategy for logistics could po- This uncertainty is costly for producers and export- tentially achieve more. The increase in paved roads ers and makes it more difficult to plan an efficient in Bolivia has promoted connectivity and reduced supply chain. Delays also translate into significant- travel times for passengers and freight. Airports are ly increased costs, such as when a truck misses the being improved, and growing volumes of cargo are departure of a vessel and is required to wait up to being traded. Even so, an integrated logistics vision a week for the next departure. In the quinoa supply and strategy could be an effective way to attract in- chain, land transport has an especially pernicious ef- vestments in infrastructure and improved logistic fect on logistics; moving quinoa from the production services, promote regulatory reforms, and identify zone by truck to the Port of Arica can take from 24 to opportunities for institutional strengthening. 106 hours. The variation in transport time is main- ly the result of delays at the Tambo Quemado and 5.2. Development of competitive Chungara border crossings, caused by limited hours agricultural value chains: Approaches of operation, congestion, and system failures, among and examples of key value chains other reasons. 5.2.1. Approaches to develop One key area where logistical efficiency can improve competitive value chains is inspection and certification, overseen by SENA- SAG (Servicio Nacional de Sanidad Agropecuaria e Bolivia’s approach to developing competitive value Inocuidad Alimentaria), the national agency respon- chains has largely focused on reducing the depen- sible for the health and safety of food and agricul- dence on imports, partly because of the food crisis tural products. A 2014 World Bank analysis of trans- more than a decade ago. The limitations of focusing portation and logistics in Bolivia29 concludes that merely on internal markets are now clearly recog- routine delays associated with SENASAG inspection nized in the government plan (PDES), opening oppor- and certification create long turnaround times that tunities to work with export value chains to enhance are incompatible with business and trade realities. economic diversification and rural livelihoods. Interviews with respondents in various value chains A framework developed by the World Bank (2011) is reinforce this conclusion. They note that SENASAG is useful for understanding the strategies that Boliv- efficient at border crossings (which is corroborated ia has followed to develop value chains. Figure 33 by Bolivia’s relatively high DBI score for ease of trad- displays, in stylized form, the requirements associ- ing across borders) but that customs warehouse de- ated with a series of markets, ranked in increasing lays can last up to two weeks and that it can take as order of sophistication from Level 1 (value products long as 43 days to issue sanitary certificates. sold into informal domestic bulk markets) to Level 6 The government has made tremendous efforts to (highly perishable, high-value products sold into so- improve road infrastructure, from 1990 to date, the phisticated international markets). At one end of the nationwide road network has increased by about spectrum, Levels 5 and 6 represent highly demand- 70 percent, over the last 10 years (2008-2018), the ing markets, implying the need for strong value chain public investment in the primary road network grew coordination and large investments in infrastructure, by 82 percent. As a result, progress in creating the logistics, and coordination mechanisms. At the other conditions for better logistics is encouraging, but an end of the spectrum are the domestic markets, with 29 “Transport & Logistics: Bolivia’s Path towards Competitiveness”, June 2014 (background paper, unpublished). 95 Figure 33. Requirements associated with increasingly sophisticated agri-food markets + specifications for more advanced + specifications for and quite specific + specifications more advanced and process standards, for selected, basic often quite specific + internal quality yet implemented standards, basic process standards characteristics of in the context of + quality grades GAP, good hygiene with associated products highly integrated and varietals and approaches greater detail in + visual supply chains and preferences to safe pesticide record-keeping Upgrading steps characteristics + basic where the supplier use/storage and requirements on has a relatively + consistent quality associated record- pesticide use sophisticated and quantities keeping systems management structure for quality control and risk management + visual + visual 1st and 2nd party +2nd and 3rd +2nd and 3rd +2nd and 3rd inspections inspections inspections/testing party conformity party conformity party conformity assessment assessment assessment Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Stringency of official and buyer’s requirements, Level of sophistication of the conformity assessment systems Domestic informal Domestic High end Horticulture Horticulture Exports to High-End International Outlets markers Exports to Wholesale Markets Low risk products High risk products Source: World Bank 2011 much less stringent demands in terms of quality and (Yungas Region), and chia and Brazil nuts (also in the conformity assessment procedures. Amazon Region). In most cases, these efforts have suffered considerable setbacks from changing sup- For the most part, the development of agricultural ply and market dynamics (quinoa) and/or from the value chains in Bolivia has targeted Levels 1–3, mar- lack of coordinated public-private sector value chain kets that satisfy domestic demand or that have very strategies, with a clear shared vision and objectives. little quality differentiation, but in which improve- Such strategies, as in the case of coffee, would have ments have the potential to benefit large number of given continuity to successful early initiatives for ex- farmers. The potato chain (described in detail later in port product differentiation supported through inter- this chapter) is one example. It is important to note national cooperation. that most efforts to increase growth in these value chains have emphasized gains in production much The strategy to improve the export competitiveness less on productivity, value addition, and product dif- of commodities (oilseeds in particular) is largely ferentiation based on quality or some other trait. based on access to inexpensive land and preferential trade agreements with Andean Community markets. Perhaps the most important efforts to differentiate Those advantages are currently eroding, calling on products for export, targeting Level 3–4 markets, Bolivia to alter its approach to enhance competitive- are linked to organic quinoa (quinoa real) in the High- ness by focusing on a cost-reduction strategy based lands Region, organic coffee in the Amazon Region on the adoption of improved technology, productivity 96 improvements, improved logistics, and a policy envi- ness. The value chains include potatoes (productivity ronment that supports these goals. This cost-reduc- and market integration to benefit large numbers of tion strategy should go hand-in-hand with the imple- Bolivian producers—Levels 1 and 2), beef (building mentation and communication of commitments to a sustainable sector to supply domestic and export sustainability. demands—Levels 3 and 4), coffee (growing as a spe- cialty industry—Level 4), quinoa (enhancing position- Perhaps the most telling example of successful value ing in differentiated markets—Level 4), and soybeans chain development in Bolivia comes from the nascent (shifting the technological frontier and reducing the grape and wine value chain in the Sub-Andean Region, environmental footprint—Level 3). which chose to differentiate its wines by emphasizing that they are produced at higher altitudes (vinos de POTATOES: Productivity and market integration to altura). International cooperation played a critical role benefit large numbers of Bolivian producers in enhancing the visibility of Bolivian wines in domes- tic and international markets. Government programs While not a full-fledged value chain in many respects, have supported productivity improvements and pro- potato production and marketing may be considered motion strategies. Now that international support a traditional chain with a small but promising frontier for the industry is coming to an end, the private sec- for new chain opportunities. The potato holds a privi- tor must assume a strong leadership role, working leged place in the Bolivian diet, rural production, his- closely with the public sector to build on this early tory, and culture. While not altering its traditional role, success and address critical competitiveness issues, this traditional chain needs substantial impetus to in- such as the introduction of grape varieties specifically crease its productivity and essential competitiveness. for export markets, export promotion strategies, and quality certification. Production and producer profile. Bolivia and neigh- boring Andean are the genetic home of the potato, Some of the efforts by non-governmental organiza- which exhibits staggering biodiversity in Bolivia. Each tions to support the grape and wine value chain in- locale has its distinctive set of varieties, and 1,500 volved working with local governments to identify varieties are thought to be grown across the coun- economic opportunities that could build on grape and try. The prime growing areas are in the Sub-Ande- wine production and processing, such as tourism (la an and Highlands Regions, but potatoes—grown in Ruta del Vino) and cheese production. These territori- two-thirds of all municipalities (221 of 337)—flour- al and cross-sectoral approaches show considerable ish everywhere except for Pando and Beni Depart- potential to increase economic activity in rural Bolivia. ments. Bolivia produces over 1 million metric tons (1.18 million in 2017) across 183,000 hectares, and 5.2.2. Competitiveness constraints and Cochabamba is the single largest producer, with 28 opportunities for key agricultural value percent of national production. The crop is grown chains mostly by small-scale farmers, although larger op- Aside from the sector-wide competitiveness con- erations in Santa Cruz have shown that mechanized straints identified in the previous section, Bolivia’s production systems achieve yields that are nearly key agricultural value chains face more specific com- double the national average. To preserve its wealth petitiveness constraints and opportunities. They are of potato genetic resources, Bolivia is the rare low- discussed in the spotlight sections that follow, along er-middle-income country with a gene-bank, although with potential strategies to improve competitive- 97 Table 19. Comparison of Potato Yields in LAC and Beyond, 2017 (MT/Hectare) US Argentina Chile Brazil Bolivia World Ave Average Yield 46,4 29,3 23,8 18,4 5,5 27,38 Source: FAOSTAT, 2017. it would benefit from institutional strengthening and from Bolivia (100 tons in 2018) consisted of canned technical assistance. chuño exported to Spain. Other emerging opportuni- ties include frozen potatoes for the domestic fast- Market profile and potential. A characteristic of food trade, which are currently imported from Argen- the potato market worldwide is that only a small tina, Peru, and even Canada. Food manufacturers also fraction of potatoes is exported (3 percent average need to develop locally sourced and produced potato worldwide), even by the largest producers (China and chips to compete with expensive but popular imports. India). As a perishable crop that has a low value in A highly significant import-substitution opportunity relation to its weight, potatoes are costly to trans- is to increase national potato production to prevent port, so they tend to be grown for consumption by the current shortfall, estimated at 55,000 tons, from the farm household or sold in local markets. Bolivi- being filled by illegally imported potatoes from Peru. ans are world champion potato consumers, averaging 95–100 kilograms per person annually, led by Coch- Productivity. Just as the Bolivian value chain is tradi- abamba Department with 151 kilograms per person. tional, so too are its yields, having changed little over Although Bolivia exports very little of its production time. As depicted in Table 19 below, even neighbor- (conforming to the pattern 3 percent), significant po- ing countries have yields close to the world average, tential exists for diversification and adding value. A which is five times greater than Bolivia’s average yield, distinguishing feature of the traditional potato value while yields of leading producers are even higher. chain is that wholesalers buy the crop at dispersed Logistics and chain management. Logistical inef- traditional farmers’ markets (ferias) in the production ficiencies are primarily structural, associated with zones and transport the potatoes to wholesale stor- the rural collection and assembly (acopio) of produce age and market facilities in cities. from large numbers of dispersed producers and While fresh potatoes will remain the mainstay of the highly unusual diversity of the produce. The principal value chain, at the margin various markets present public good bearing on transaction costs is the trans- far more lucrative opportunities compared to the slim portation network. margins from selling fresh potatoes in a competitive Strategy to improve competitiveness. Without any marketplace. The first is chuño, a traditionally dried misguided notions of displacing the domestic con- product made from a particular variety of yellow po- sumption of a broad variety of traditional potato vari- tato. Chuño has the remarkable ability to be stored for eties, there is an overwhelming need to improve the up to 10 years, aiding food security and reducing risk. competitiveness of the chain. The essence of the tra- Forty-four percent of the very small potato exports ditional sector will not be compromised by increas- 98 ing productivity without displacing local diversity with BEEF: Building a sustainable sector to supply growing non-native high-yielding varieties. Increasing yields domestic and export demands significantly requires public and private investment across a range of producer support activities: exten- The cattle and beef industry in Bolivia, with its high sion of best practices and technical assistance, plant degree of informality and limited industrialization health disease control measures, soil fertility protec- and exports, does not constitute a full-fledged or tion and enhancement, small-scale irrigation systems highly articulated value chain. Eighty percent of and water-user organizations, improved post-har- transactions are informal and generate no tax reve- vest handling and storage techniques, and financial nue. Productivity is low, although good potential ex- services. An important step for the development of ists to improve it if policy and investment measures potato varieties would be to further strengthen the provide the means and incentives to establish a more genebank to preserve its valuable collection of potato competitive value chain. varieties and extend the usefulness of these genetic resources, possibly in collaboration with other gene- Production and producer profile. Census data indi- banks such as those associated with the International cate that in 2013 ranchers in Santa Cruz Department Potato Center in Peru and the International Center for grazed about 41.7 percent of the cattle produced for Tropical Agriculture in Colombia. beef, followed by the Beni Department with 35.6 per- cent. In recent years the “industry” has produced over The second leg of the strategy, which the government 260,000 tons of meat, of which only 5 percent has has been pursuing, is to focus on realizing and ex- been exported. The small export channel, entirely panding the various market opportunities outlined based in Santa Cruz, consists of some 200 producers already, such as the production-consumption gap to selling to three packing houses approved to export displace Peruvian contraband potatoes. Another sig- beef to Peru and other countries (very recently China). nificant opportunity to increase producer income is to link investments in irrigation and water management Market profile. Approximately 30 buyers truck the to reaching the large market window for local con- cattle they purchase to market, while many ranch- sumption when prices are highest. Until the typical ers deliver their cattle on the hoof, with all of the harvest period from February to June, prices are at attendant issues of weight loss, stress, and delays, their peak. With carryover from storage, prices start particularly during the rainy season. All of these in- to climb by October, peaking typically in February. termediaries are required to hold a SENASAG autho- Using small-scale irrigation, particularly pressur- rization. The buyers often purchase on credit, lack- ized systems, producers can target that window. The ing well-established financial services to support other opportunities mentioned in food processing the cattle and beef market. The outlets for the meat (frozen potatoes, chips, and other products) rely on are approximately 500 retailers around the country. private initiative, but the public sector can proactively Some wholesale function exists in El Alto near the seek to promote interest, including with international capital, La Paz. Slaughterhouse facilities are typical- partners. The experience of the development of the ly cramped and inadequate. Of the US$12 million in native potato value chain in Peru, offers important beef exported from the three approved meat pack- lessons in terms of institutional innovations that Bo- ing plants in 2017, about half targeted the market livia could pursue. for boned meat, with some value-added destined for hamburger meat sales in Peru, Ecuador, and Co- 99 lombia. The other half was exported as unprocessed ue-added provides a minimal precedent, but without meat to Peru, Vietnam, and Brazil. more critical mass, it will be difficult to incentivize the base of the value chain. Private-public collaboration Productivity. Cattle production remains essentially on export and enterprise promotion would be a pro- traditional, with few incentives or resources to raise active step to break out of the low-productivity and productivity. Ranchers in Beni Department depend low-reinvestment cycle. Eventually the formalization entirely on extensive grazing; the land has a relatively and greater integration of the chain will include in- low carrying capacity (1 head per hectare) and there creased collection of taxes and use of tax revenue to is little pressure to use land for alternative produc- invest in public goods in support of the sector. While tion. The Nelore breed is well adapted, so there are the logical candidate for increased industrialization few incentives to initiate major breeding programs. or exporting is Santa Cruz under current circum- The situation in Santa Cruz Department with Brah- stances, the competitive advantage of ranchers in man cattle is similar, although producers’ use of ra- Beni is not to be overlooked. As low-cost producers tions reflects the higher opportunity cost of land and of entirely grass-fed beef, they have the potential to incentives for intensification. Artificial insemination is capture export opportunities to satisfy high-value practiced in Bolivia, but no significant breeding pro- markets demanding grass-fed meats. If these niche grams or genetic research or sourcing are underway. markets prove feasible, then certification programs Logistics and chain management. The poor road in- and eventually traceability protocols could readi- frastructure in Beni Department is typical of extensive ly transform some of the traditional producers into areas with low economic opportunity costs or options suppliers of a differentiated, even premium product. to generate fiscal revenue. The poor condition and Such a development would justify upgrading one or maintenance of main roads and the limited network more of Beni’s 15 slaughterhouses into true packing of access roads precludes trucking in many cases. houses with SENASAG authorization, inspection, and Cattle drives are the standard practice, with associ- any other certification that may be required for export ated animal health issues and risk of delays. The fleet products. Overall, the sectoral strategy would need of trucks and vehicles used in the beef industry is to emphasize gains in productivity, higher formal- generally reported to be in poor condition. ization, sanitary compliance, and a gradual strategy for differentiation in niche markets. The opening of Strategy to improve competitiveness. The current opportunities in markets such as China, would imply low level of competitiveness reflects a domestic attention to sanitary compliance. In China relevant is- market without high-value segments and little inte- sues would include traceability, residue and pathogen gration in export markets. It is very much the result control and nonuse of anabolic hormones. of export policies that prioritized supplies for the do- mestic market, severely restricting (banning) exports Strategies for growth as a sustainable industry. between 2008 and 2012 and then establishing export The prospect that expanded livestock production quotas in 2013. The cattle and beef industry is likely will exacerbate climate change is a global concern. to operate in its current state until higher-value ex- In Bolivia, cattle are associated with deforestation port markets or domestic value-added industry are and rates of forest degradation, particularly in the developed that offer higher profitability. Few leather Chaco region, where livestock are raised in native and processed meat operations exist in Bolivia. The forest areas. Furthermore, livestock are associated very small share of production (approximately 2 per- with large greenhouse gas (GHG) emissions, which cent) that goes to an export market with some val- improved production practices could minimize. The 100 sector would benefit from strategies to support Pando (1 percent). Caranavi, a single rural municipal- sustainable expansion, including land use planning ity in the Yungas, accounts for 66 percent of national strategies and mainstreaming of sustainable and cli- production. mate-smart practices. Café de altura is arabica, predominantly of the long-established local (típica or criollo) variety, rec- COFFEE: Growing as a specialty industry ognized in international competitions as being of the Efforts initiated in the early 2000s with international highest quality. Thirty percent of graded and sorted assistance to produce high-quality coffee and present production is classified as specialty coffee, and sev- farmers with an alternative to coca production has enty percent is classified as conventional coffee of slowly yielded results. Bolivian growers received a superior quality. higher average price per kilogram than growers in any other coffee-producing country, and premium high-quality Bolivian coffees attained some of the world’s highest market prices. The premier coffee growing area of Bolivia offers optimal agroecological conditions yet underperforms, leaving limited alter- native economic activities. Low yields, weak producer organizations, poor training, degraded soils, limited input supplies, and limited infrastructure all under- mine productivity of the coffee subsector. The world coffee price fell in 2019 to its lowest point in over 13 years. Farmgate coffee prices are now below the cost of production for Bolivians. The price crisis threatens to halt exports, further weakening the industry, and the end is not in sight, as the devaluation in Brazil (the driving force) cannot be resolved easily. Both industry and the government will need to invest for the longer term to position producers for an eventual recovery and resilience in the fact of future threats. Coffee is produced by 23,000 families, and an ad- ditional 12,000 are employed elsewhere in the cof- Production and producer profile. Over ninety percent fee value chain. These smallholders operate 17,491 (91%) of the coffee crop in Bolivia is highland coffee coffee “farms” or production plots distributed over (café de altura) produced in the Yungas Region of La 22,824 hectares, for an average of slightly less than Paz Department, on the humid and milder eastern 1 hectare per producer family. Most smallholders slope of the Andes between 1,200 and 3000 masl. (an estimated 80 percent) belong to local producer Coffee beans grow more slowly at high elevations and organizations. The 90 producer organizations are develop distinct organoleptic characteristics that are affiliated with four second-tier associations or fed- prized by a segment of the high-value coffee market. erations. While the total number of coffee farmers The remaining nine percent of the coffee crop is pro- is about one-third as large as the number of quinoa duced outside of the Yungas in the Departments of farmers in Bolivia, the coffee value chain offers more Santa Cruz (5 percent), Cochabamba (3 percent), and 101 opportunities for value addition and generates more Much of this delay is attributable to storage at the employment. The benefits of growing coffee are high- buyer’s rural export warehouse (planta beneficiado- ly geographically concentrated, but the jobs creation ra) while assembling sufficient product from various downstream are dispersed. The Yungas region offers smallholders to fill a container. very few sustainable economic alternatives to coffee Strategy to improve competitiveness. Efforts should production. concentrate on the prime growing area in the Yun- Coffee yields in Bolivia average 370–700 kilograms gas to ensure the highest-quality arabica coffee per hectare, whereas the demonstrated potential possible and consolidate Bolivia’s growing strate- yield is 900–1,200 kilograms per hectare. Yield vol- gic position in the coffee specialty market. Organic umes cannot be considered apart from quality, how- and fair-trade certifications would add little benefit ever, because the quality premium can be very high. without a strong quality-cup production industry. An extreme example is a producer in the Yungas These efforts would also benefit the growing domes- growing the specialty variety called Geisha. It gar- tic demand for high-quality coffee in high-end hotels nered significant recognition in a cupping competition and restaurants, which is currently restricted due to and sells for over US$14 per kilogram. limited supply. The investments required to position Bolivia in the specialty coffee market are multi-fac- Market profile. Producer organizations currently have eted but well recognized by the industry and govern- little involvement with market segmentation. Some ment. The 2018 National Coffee Investment Program production enters domestic market channels and led by the Ministry of Rural Development and Lands some is sold directly, but most of the coffee produc- (MDR&T) provides a clear roadmap to enhance the tion goes to 28 exporters selling into the vast world coffee sector’s competitiveness. The six “strategic market for arabica coffee. The volume of coffee pro- axes” of investment are to: (1) improve coffee vari- duced fell from its previous level of 7,800–8,000 tons eties and production technology within a sustainable to around 5,400 tons in 2018, and exports, which had natural resource management approach, (2) imple- held steady around 5,000 tons annually over 2000– ment a proposed new institutional and legal frame- 09, dropped continuously from 4,600 tons in 2010 to work, (3) identify priority infrastructure investments, 1,520 tons in 2017. Even with the downward slide of (4) support the development of agro-processing that coffee prices worldwide, average prices of Bolivian improves coffee quality, (5) develop new and expand- coffee have risen. FOB prices of US$2 per kilogram ed market channels domestically and abroad, and (6) in the early 2000s surpassed US$3 per kilogram by improve government financial credit to support pri- the end of that decade, rising to values averaging over vate investments along the value chain. To succeed, US$5 per kilogram over 2011–17. The decline in ex- these efforts need to be closely implemented with port volume swamped the price effect, and Bolivia’s the private sector, including interested international export revenues have fallen from their height in 2011 buyers, particularly those linked to direct trading, and of US$26 million to US$10 million in 2018. experienced development partners. Logistics and chain management. The number of dis- tinct links in the coffee value chain means that logis- QUINOA: Enhancing Bolivia’s positioning in differen- tics and chain management strongly affect its overall tiated markets competitiveness. Currently logistical inefficiencies result in an average of 55 days from post-harvest to Quinoa is the Altiplano cash crop par excellence in the arrival of the shipping container in the exit port. several important respects. The crop originated in the 102 Highlands Region where it is grown, and the region 2012 and nearly reaching 150,000 tons in 2013. In the which remains the sole producer of the organic qui- following year, when Peru surpassed Bolivia in total noa real variety that is the most highly prized in the exports, world market prices peaked at around US$ marketplace. The same nutritional characteristics 6,000 per ton. The rising volume of exports saturated that make quinoa a high-value export (it is a high-pro- the market, depressing prices since 2015. tein grain rich in minerals and micronutrients) confer The significant production increases attained in re- a health and food security advantage to poor rural cent years were the result of an expansion in cul- households with chronic dietary deficiencies. The tivated area and not of an increase in yield. Yields low perishability of this high-value grain renders it have plateaued at around 600 kilograms per hectare relatively robust in the face of major transport costs (a recent calculation finds that average yields over and logistics hurdles that undermine Bolivia’s com- a 16-year period were 540 kilograms per hectare, petitiveness. with no significant upward trend). Experimental plots show that yields of 1 ton per hectare are feasible in Bolivia, but in farmers’ fields, depleted soil fertility and limited control over access and timing of water are clear areas for improvement, along with mech- anization; basic tillers have been shown to increase quinoa yields. An economic analysis for a represen- tative sample of growing regions should be done to determine optimal levels of inputs and public-private investment. Other agricultural research has demon- strated promising varieties of quinoa for new areas and elevations, although they would supply conven- tional quinoa markets rather than markets for pre- mium quinoa real. Quinoa generates employment and income for one out of ten Bolivian agricultural producers, dispropor- Production and producer profile. As the world mar- tionately to the benefit of small-scale farmers in the ket for this traditional grain only “discovered” quinoa Altiplano. About 70,000 small-scale highland farms and began developing major marketing channels are estimated to produce quinoa, two-thirds of which in the mid–late 2000s, Bolivia dominated the na- are centered in Potosi and one-third in Oruro and La scent export market. In 2014 Peru edged out Bolivia Paz Departments.30 Given an average family size of as the larger exporter, although Peru produces the 4–5 persons, some 300,000 of the approximately 3 conventional non-organic variety. The two countries million rural Bolivians benefit directly from the pro- continue to lead production and export, with only mi- duction and export of quinoa. Direct exports account nor exports from Ecuador or non-Andean producer for 44 percent of production, 23 percent goes to lo- countries. Bolivia’s production remained near 50,000 cal markets, and 33 percent enters informal market tons from 2001 to 2009, leaping to 120,000 tons in channels in Peru. Some of the production moving 30 This figure is repletely reported in reports, since the export boom and it is uncertain how many farmers have aban- doned quinoa production for commercial purposes as a result of the declining international prices. 103 through informal markets in Peru could possibly be Bolivian quinoa possesses fundamental advantag- retained for export from Bolivia if various measures es that must be consolidated as the foundation for to enhance competitiveness are adopted. a competitive value chain. Organic quinoa real is one such advantage. In addition to its marketing edge as Market profile. Total revenue from quinoa exports an organic product, quinoa real has demonstrated has remained relatively low since the 2014 peak, and nutritional advantages over conventional quinoa. An opportunities for a market premium will remain lim- average 20 percent price premium existed for quinoa ited until market size expands and market segmen- real as recently as 2015. This premium should be tation develops further. Over 2012–17, North America regained to the extent possible, but regardless, con- imported just over half of all quinoa in the world mar- sumer loyalty to organic quinoa must be nurtured and ket (the United States imported 39 percent and Cana- this premium product should be differentiated and da imported 17 percent). Various European countries secured throughout the value chain, via improved account for much of the rest (29 percent), followed traceability, to prevent Bolivia’s comparative advan- by Australia (4 percent) and various others (14 per- tage from being squandered. cent). At the peak price and volume in 2014, Bolivian quinoa exports reached US$200 million. With prices The Bolivian industry can work to expand market size in the US$ 4,000 per ton range the following year, and manage price risks. Bolivia, unilaterally or as part export revenue fell by half to approximately US$100 of an Andean regional effort, needs to invest in pub- million; in 2016 prices declined to US$3,000 per ton lic-private efforts to expand the total size of markets and export revenue fell to US$60 million. Prices have for quinoa. Whether expanding existing markets or not recovered, although some premium remains for developing new ones (in Asia, for instance), the in- organic quinoa real. Fair trade quinoa offers a signifi- dustry and public partners must carry out a sustained cant premium captured by Bolivia, but it only accounts campaign exposing more consumers to quinoa and for 5 percent of the export market. touting its “supergrain” or “superfood” benefits. Ex- isting links between quinoa and high-end culinary Logistics and chain management. Forty days are options can be expanded in conjunction with tourism typically required for shipments to complete SENA- promotion and point-of-sale promotional materials in SAG inspection and export certification. This process current and new markets. Bolivian industry strategy could readily be streamlined. Improved coordination and communications should always differentiate be- of shipping containers could reduce turnaround times tween organic real and conventional quinoa. from 10 days to 5. Improved rural assembly (acopio), sorting, and staging would significantly reduce The price received for exported quinoa is not entirely post-harvest shrinkage (merma), which can be as exogenous, given the small size of the world market much as 10 percent between farm and final shipment. and Bolivia’s market share. Just as Bolivia and Peru share responsibility for saturating the market when Strategy to improve competitiveness. To retain its prices peaked in 2014, so too can they avoid pitfalls comparative advantage and strengthen its position as without having to behave as a cartel. Joint Bolivia-Pe- a premier quinoa producer and prevent further loss- ru quinoa industry collaboration, or a broader Andean es of market share and profitability, Bolivia needs to dialog that includes Ecuador, would be worthwhile to resolve issues related to productivity and logistics. lower the risk of mutually damaging overproduction. At the same time, it needs to pursue market strate- In addition, the industry should focus on expanding gies to expand market size, bolster prices, or partially Fairtrade opportunities as much as possible, given hedge price risks. 104 the socio-economic profile of Andean smallholders 10 oil extraction plants linked to various internal and and the potential of Fairtrade compared to current external marketing channels. market prices. Production is centered in Santa Cruz Department but extends throughout the Lowlands Region. More than SOYBEANS: Shifting the technological frontier and 14,000 farmers grow soybeans on farms of all scales. reducing the environmental footprint Although the industry claims that an average of 85 Bolivia produces 3.2 million tons of soybeans on 1.2 hectares is planted to soybeans per farm, in fact 77 million hectares, or one-third its cultivated area. This percent of producers are small-scale producers (de- volume drives 10 processing plants to extract oil fined here as having 0–50 hectares), 21 percent are and separate the soybean cake. Exports go mainly medium-scale producers (50–1,000 hectares) and 2 to Andean neighbors, especially Colombia (42 per- percent are large-scale producers (above 1,000 hect- cent). Soybeans also hold a strategic value as a feed ares). The best yields achieved in Bolivia are 80–90 or input source for poultry, livestock, dairy and oth- percent of the yields obtained by other producers in er products, thereby generating domestic value. The the LAC Region, which limits competitiveness. Output expansion has been fueled by foreign capital (mainly volume grew by 163 percent over 2000–17. Over the Brazilian), as part of the broader expansion of oilseed same period yields increased by 54 percent to narrow commodities in the Southern cone. part of the gap with regional competitors. Currently producers are looking to biotechnology to provide the highest-yielding varieties with the most favorable re- sistance profiles. The nitrogen-fixing property of soy- beans helps to enhance yields of the crops with which it is rotated (often wheat, maize, sorghum, and rice), giving soybeans strategic importance for maintaining soil fertility in a country with the lowest rate of fertil- izer use in the region. Logistics and chain management. As a soybean ex- porter and landlocked country, Bolivia has a clear competitive disadvantage compared to neighboring countries with direct maritime access. Each link in the production and export chain is vulnerable to logisti- cal inefficiencies. The average distance of farmers to the storage silos of soy processors is 100 kilometers, and storage facilities are 90 kilometers from the pro- Production and producer profile. As the sev- cessing plants on average. Processed soybean cake is enth-largest exporter of soybeans, Bolivia captures transported by rail from Santa Cruz to the river port only a very small portion of the market. The soybean on the Brazilian border, then barged along the Para- value chain is Bolivia’s most fully articulated value guay–Paraná Waterway to ships in Rosario, Argentina. chain, encompassing a range of actors from produc- The Bolivian government’s military river port (Puerto tion to processing. Producers rely on 43 firms to pro- Busch) in the tri-border area with Paraguay and Brazil vide inputs, machinery, and services and sell to the 105 allows direct access to the Paraguay River and hence scale soybean producers, however, with much less to the Paraná River and Atlantic. monitoring of the impacts of such efforts. Strategy to improve competitiveness. Bolivia’s soy- Growing as a sustainable industry. Bolivia has im- bean industry is highly regulated. The export system portant opportunities to shift the soybean subsec- permits exports only when domestic needs are sat- tor toward a more sustainable pattern of expansion. isfied, and prices are regulated through a system of Enhancing institutional capacities for land use plan- price bands. Despite export and price controls, the ning and monitoring, as well as obtaining a stronger subsector has grown as a value-added industry, and commitment from the private sector to promote re- more than three-quarters of soybean exports are in sponsible environmental and social investments and the form of vegetable oil, flour, and soybean cake. The expansion, would deliver large dividends to valuable main competitive advantages for the industry lie in ecosystems and ensure the sustainability of the sector. producers’ access to inexpensive land, access to sub- sidized diesel, and preferential trade arrangements 5.3. Take-away messages in the Andean Community market, the main market A number of cross-cutting issues and recommenda- for Bolivian exports. These advantages are slowly tions emerge from the discussion of competitiveness eroding with the opening of the Andean Community constraints and opportunities in this chapter. They are market to competition in 2018 and the pressure to highly relevant as Bolivia takes steps to further devel- reduce diesel subsidies. op and consolidate agricultural value chains. These Export controls or other policy decisions that limit issues and recommendations are discussed in the volumes in the export chain detract from the com- paragraphs that follow. petitiveness of this subsector for several reasons. The development and/or further consolidation of val- Unlike other products that are candidates for differ- ue chains in Bolivia would need a broad set of strat- entiation (such as quinoa, coffee, beef, horticultural, egies to target opportunities in a range of markets. or food products), a true commodity such as soybeans First, given the large number of producers involved in competes based on costs and margins. Situated in supplying domestic markets (Levels 1–3), it will con- a landlocked country, in the middle of the world’s tinue to be a priority to enhance opportunities for cre- single largest soybean region, the Bolivian soybean ating wealth around domestic value chains. The size industry and the government will need to pursue ev- of these opportunities will largely be linked to domes- ery opportunity within their control to reduce costs tic consumption and market developments, particu- to become more competitive. The primary levers are larly the increased consumption of vegetables, fruits, increased productivity through genetic and techno- legumes, and other products in the high-value spec- logical improvements, water management, improve- trum, and further development of quality-differentiat- ments in soil fertility and plant health protection, ed markets through continued formalization of the re- greater efficiency of SENASAG in its inspection and tail sector, development of the gastronomic industry, certification role, and any measures identified in the and development of demand for fresh and locally pro- logistics chain to reduce costs. The analysis of pro- duced food (short-value chains/circuitos cortos). In the ductivity presented in Chapter 4 highlights important most traditional sectors (such as potato), as well as in opportunities to increase productivity of small-scale diversification into high-value products for domestic producers and large ones. Efforts have been made by markets, success will depend greatly on developing the government to level the playing field for small- competitive advantages around cost reductions, qual- 106 ity and productivity improvements, efficiencies along investments will depend greatly on the capacity to in- the value chain, and market integration, accompanied crease public and private capabilities for innovation— by improved capacity to manage risks (price fluctu- both “hard” and “soft” innovation capabilities. “Hard” ations, climate shocks, pest and disease outbreaks, innovation capabilities are related to research and de- and so on). Second, in industrial commodity sectors velopment, and they include a range of technological serving domestic and/or export markets (soybeans, opportunities linked to production, post-harvest han- sugar, sorghum, and others), a strategy based on pro- dling, and processing, as well as opportunities for in- ductivity improvements (for large-scale and small- creasing resource and input use efficiency to protect scale producers) and cost-reductions is key, anchored the natural capital for agriculture (land, water, and in improvements in environmental sustainability and biodiversity). “Soft” innovation capabilities include the risk management strategies. Lastly, in the case of generation of new inclusive business models (such non-traditional agricultural exports, a cost-reduction as value-chain finance and product differentiation, strategy will help to increase participation in mar- certification, and traceability), as well as strength- kets that are not highly quality differentiated (e.g. for ening public-private dialogue and collaboration, not beef in the Middle-East, China, but ensuring sanitary/ only within Bolivia but at the regional level, such as safety compliance would be critical). Yet, important an Andean regional business dialogue to expand mar- opportunities for Bolivia could be linked to strategies kets for quinoa or for other food and fiber products that take advantage of Bolivia’s outstanding agro-di- (potatoes, dairy, meat, leather). Public-private collab- versity to develop products differentiated by their oration should proactively reach out to those in other uniqueness and quality. countries who can assist with the transfer of technol- ogy and know-how. These innovation capabilities also Bolivia is looking at markets such as India and China, extend to institutional development at the grassroots which are highly price sensitive and competitive. The level, in terms of finding innovative ways and instru- assessment of opportunities for agricultural exports ments to professionalize producer organizations. If to these markets must be based on a clear under- these groups are not strengthened in concert with standing of comparative and competitive advantag- increased investments in irrigation and other assets es. In those markets, only a limited range of Bolivian and technologies, those investments will not achieve products may be competitive on a cost basis, and sustainable competitive levels. therefore it may be relevant for Bolivia, in parallel, to develop and consolidate opportunities in markets that Worldwide experience with developing inclusive value are more sensitive to quality. chains points to several factors that are critical for Bolivia to consider as it deepens efforts to develop its The value chains selected for analysis in this report agricultural value chains. The relevance of innovation illustrate a broad range of targeted on-farm and off- capabilities stands out among these factors, which farm investments that can enhance the performance include: of each value chain in response to market opportu- nities. These investments include improved varieties i. The most successful efforts responded to a mar- and inputs; the adoption of better production and ket opportunity that was already in place or that management practices; improved equipment for could be developed. production and processing; irrigation investments; ii. Participation of small-scale producers in dynamic and market promotion strategies and coordination value chains requires a strong level of collective platforms, among others. The effectiveness of these action and coordination to achieve logistical and 107 cost efficiencies, ensure quality and scale in vol- Furthermore, streamlining logistics for agricultural ume, and improve business and managerial ca- exports and imports is a central cross-cutting prior- pacities. ity. SENASAG is the key institution that inspects and certifies exports of agricultural produce and livestock iii. When it comes to maintaining or enhancing qual- and imports of agricultural inputs. Reducing delays ity and opening up market opportunities, produc- and enhancing SENASAG’s technical abilities to in- tion improvements are just as important as gains spect and certify products will dramatically improve downstream in the value chain. the efficiency of agricultural value chains while pro- iv. Complementary efforts by multiple actors are tecting plant and animal health. Furthermore, contin- needed, including not only value chain actors but ue strengthening of INIAF and the agricultural innova- also international agencies, NGOs, and public ac- tion system more broadly is critical. The genebanks tors, whether they are closely coordinated or act- are remarkable and underutilized assets that would ing in parallel. benefit from strengthening, most appropriately in collaboration with other genebanks in the region that v. Leadership by an entity or set of entities providing are part of the network of the Consultative Group on strategic direction, including international proj- International Agricultural Research and operate un- ects/programs, research institutions, NGO initia- der the framework of the International Treaty on Plant tives, private sector associations, and/or govern- Genetic Resources. ment programs. Important gains can also be achieved from broaden- vi. Successful efforts took years to materialize, re- ing access to credit and increasing producers’ lever- quiring persistence and a long-term vision. age through alternatives to conventional financing. It is the private sector that will move the needle to- The EBA benchmarking reveals that Bolivia’s regula- ward greater competitiveness in all of the cases an- tory framework provides relatively better access to alyzed here and throughout the agricultural sector financial services than the average for the region or more broadly. But the private sector cannot contrib- countries with comparable income levels. The World ute effectively to competitiveness without the critical Bank Enterprise Survey31 corroborates that few support of an enabling environment for public goods businesses of any scale consider access to finance and services. a major bottleneck. However, a large number of pro- ducers with market potential still have limited access As a landlocked country with a challenging topography to credit. One alternative to conventional financing is across much of its 1 million square kilometers, Boliv- value chain finance, which includes supplier credit for ia faces higher transaction costs than countries with buying inputs, to be repaid when produce is harvested maritime access and lower costs for transport infra- and sold. Buyer credit operates similarly: a processor, structure construction, operation, and maintenance. exporter, or other commercial buyer advances funds Whatever the budgetary level of infrastructure expen- tied to the delivery of a specific product. Additional- ditures may be, either in infrastructure maintenance ly, Bolivia’s introduction of the legal framework for or construction, infrastructure investments should be Moveable Collateral permits the use of warehouse prioritized within the logic of enhanced growth oppor- receipts, which enable producers or intermediaries tunities and chain coordination/connectivity. with crops in storage to leverage the value of the in- 31 Enterprise Surveys. Bolivia 2017. Country Profile. The World Bank. 108 ventory as a guarantee, which in turn increases ac- cess to credit. Buyer credit as a tool of value chain finance is an example of the potential to increase buyer-produc- er coordination, yielding cost savings and risk man- agement for both parties. Business models that en- hance opportunities for improved buyer-producer coordination should be explored and supported. The productive alliance approach, with which Bolivia has considerable experience, is an important step in this direction, but further opportunities for enhancing the role of the private sector in productive alliances should be explored. 109 6. A policy framework for agricultural productivity and competitiveness Photo: Programa de Alianzas Rurales (PAR), EMPODERAR 6.1. The policy framework for aranth, tarhui, chia, and so on), mainly with a focus on agricultural investment food sovereignty. The plan also emphasizes the key role of the state in integrating projects at the level of The mid-2000s marked an important shift in econom- territories and providing direct transfers to producer ic and social policy priorities in Bolivia. The 2006–10 and community organizations. Other priorities include National Development Plan (NDP) of the Ministry of expanding irrigation, mechanization, access to land, Planning and Development called for a new develop- and activities of the State Agency for Agricultural ment paradigm and envisaged three major transfor- Marketing (EMAPA) (providing seed and fertilizer to mations: (1) a new productive matrix focusing on the producers and purchasing and wholesaling agricul- redistribution of resources from the extractive indus- tural produce). The plan also envisaged the strength- tries toward other sectors of the economy to create ening of public institutions in the sector, including the wealth and economic diversification; (2) a new insti- deconcentrated and decentralized entities. tutional framework placing communities at the center of decision-making related to productive activities at Under the Government Patriotic Agenda 2015 and the the local level, redefining community relationships PDES 2016–20, the overarching development agen- with higher administrative levels (municipalities, da of the government is still, by large, based on an regions, and central government); and (3) a territo- economic model that emphasizes state-led public rial perspective, including the formation of regional investment, but it is gradually opening to private in- territories as units of territorial planning based on vestment in important areas. As discussed in Chapter associations of autonomous municipalities and the 2, the PDES sets very ambitious goals to support eco- re-ordering of administrative units. nomic diversification in agriculture and agribusiness, increase value addition and industrialization, and en- In alignment with the NDP, the National Strategy for sure food sovereignty. Its ambitious targets for the Agricultural and Rural Development (ENDAR) pri- sector expected in 2020, include: oritized family agriculture and sought to promote the social and economic inclusion of disadvantaged • Expansion of the area under production from 3.5 segments of the rural population. Increasing produc- million hectares (2014) to 4.7 million hectares by tion and productivity were part of the strategy, but 2020. competitiveness and market integration were not • Total annual production volume of 24.8 million explicitly mentioned. The 2010–15 Government Plan tons, with specific targets for wheat, soybeans, continued the general thrust of its predecessor, em- maize, quinoa, tomatoes, potatoes, and coffee that phasizing the role of the public sector in stimulating exceed domestic consumption.32 economic diversification under the new development paradigm. The corresponding plan for the agricul- • Expansion of the area under irrigation from tural and rural sector (Revolución Rural y Agraria, 362,000 ha (2006) to 700,000 ha 2010–20), reinforced the focus of policy making on • Significant yield increases (15–30 percent) in the smallholder agriculture, domestic markets, and the main product categories (cereals, fruit, vegeta- leading role of the state. It called for the development bles, oilseeds, roots and tubers, sugar, and ani- of strategic commodities, including coca, as well as mal feed). Andean crops with high nutritional value (quinoa, am- 32 Target volumes should exceed domestic consumption by ratios ranging from 8 percent (maize) to 136 percent (quinoa). 112 • Increasing the share of organic production to 10 tion infrastructure, and weak producer organizations percent. that lack entrepreneurial vision and capacity, among other factors. A further important weakness is weak • Increased participation of all producers in overall coordination among different public entities in the ag- production. ricultural sector at the central and subnational level, • Promotion of integrated production systems and and between the Ministry and subordinate entities. sustainable land management, including the res- Despite its emphasis on import substitution and food toration of degraded soil. self-sufficiency, the PSARDI recognizes the limita- tions of domestic markets for agricultural develop- The Integrated Development Plan for the Agriculture ment, given their limited size and purchasing power. and Rural Sector 2016–20 (PSARDI, its Spanish ab- It calls for increased efforts to enter export markets, breviation) operationalizes the PDES for the agricul- making use of preferential market access granted by tural sector. This plan identifies low levels of produc- a number of Bolivia’s trading partners. tivity and competitiveness as key constraints facing the sector. It recognizes that Bolivia is facing serious The PDES emphasizes productive complexes (PCs) as challenges in producing agricultural products that a key approach to diversify the structure of produc- meet national and international demand in terms of tion, promote value addition and industrialization, and costs, volumes, quality, and consistency of supply. It achieve food security and self-sufficiency as well as particularly recognizes the challenges facing small selective integration into export markets. The PCs are rural economic actors to compete successfully even conceived as economic clusters composed of differ- in national markets. Low productivity is at the heart of ent productive actors (public, private, cooperative, and the competitiveness problem, but it also affects poor other social enterprises) which bundle the provision consumers through higher food prices. The PSARDI of support services, productive infrastructure, and re- recognizes the fundamental importance of develop- search and technology.33 PDES envisages spending ing integrated policies and actions that, based on the US$ 8.4 billion on developing the PCs through 2025 identification of competitive advantages for specific (approximately US$700 million per annum). products, orchestrate and coordinate all actors along the value chain to capture market share and increase 6.2. Broad patterns of public spending value addition. on agriculture The PSARDI identifies a lack of long-term entrepre- Historically, the share of agriculture in overall public neurial vision and commitment of all actors (public spending has been low, averaging 1.5 percent be- and private, as well as producer organizations) to tween 1996 and 2008 (World Bank 2011). Since 2006, work together within a common value chain frame- revenues from the extractive industries have been re- work. The main reasons for low productivity are well distributed across economic and administrative units understood, in terms of limited technological innova- to develop productive sectors, including agriculture tion, low coverage of extension services, limited ac- and related activities. Figure 34 shows overall trends cess to financial services, difficulties in agricultural in public expenditures between 2006 and 2016 by marketing, food safety problems, inadequate produc- economic sector. It reveals a fivefold increase in the 33 PDES envisages the development of 21 PCs, of which 12 will focus on agricultural and forest products, 2 on agricultural inputs (seed and fertilizer), and another 2 will be linked to agricultural raw materials (textiles and leather). The PCs will have facilities across the national territory. 113 Figure 34. Public investments executed by economic sector, in USD millions 6.000 5.000 4.000 3.000 2.000 1.000 - 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 TOTAL Productive Infrastructure Social Multisectoral Source: Estimations based on INE data overall volume of budget executed. The largest share These statistics provide an incomplete picture of to- of the budget went to infrastructure, showing con- tal budget support to agriculture and do not include stant growth throughout the period. During the first co-financing of national programs by the subnation- five years, social spending was higher than spend- al entities, which are not reflected in the stream of ing for productive sectors (including agriculture), but budget execution. Nor do they include expenditures since 2011 expenditure levels between these catego- under the Ministry of Productive Development for the ries has levelled out. Public spending in the produc- establishment of PCs and for EMAPA. The budget of tive sectors increased at a constant rate, especially EMAPA in 2018 was US$182 million, of which US$113 from 2010 onward, to peak at approximately US$1.4 million was executed. billion in 2015. In 2016, spending for productive sec- tors decreased, despite a continued increase in over- 6.3. Agricultural sector enablers all budget execution. In recent years, the Bolivian government has made Figure 35 shows estimated overall public expendi- significant investments in the provision of public ture and the public expenditure on agriculture be- goods and services linked to food production. The tween 2006 and 2016. The share of agriculture in main programs strengthening the enabling envi- total expenditures declined slightly, falling from 9 ronment for agriculture are linked to investments percent to 6 percent between 2006 and 2008. Be- in land administration, plant and animal health and tween 2008 and 2015, a period of strong fiscal ex- food safety, agricultural innovation and technology pansion, the share of public expenditure dedicated to development, and irrigation infrastructure. Several of agriculture remained at around 6 percent, followed these programs are implemented by so-called decon- by a slight decline in 2016. centrated institutions under the MDRyT, which have a specific mandate in the respective subject area. Im- portant institutional reforms occurring over the last 114 Figure 35: Total public expenditure and agriculture expenditure, in USD millions. 6.000 10% 9% 5.000 8% 7% 4.000 6% 3.000 5% 4% 2.000 3% 2% 1.000 1% - 0% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Total absolute expenditure Agriculture absolute expenditure Agriculture expenditure (%) over total Source: Authors, based on INE data 15 years have included the creation of EMAPA (2007), percent; and medium and large entrepreneurial land the National Institute for Innovation in Agriculture and represents 10 percent. Although the government has Forestry (INIAF) (2008), and the National Agricultural been working on addressing persistent challenges Insurance Institute (INSA) (2011). with regard to the unequal distribution of land and fragmentation of arable land, the lack of consistent 6.3.1. Land rights and land reporting over the years regarding the land awaiting management “clean” titling and being titled (where, and to whom) makes it difficult to understand land distribution pat- Greater land tenure security encourages productive terns and the extent of the current conflicts in land investments on the farm and facilitates access to fi- ownership in Bolivia. Increasing concerns are also nance. It is an important element for improved pro- emerging regarding land grabbing and trafficking on ductivity and competitiveness. The information avail- fiscal lands, indigenous territories, and national re- able for Bolivia highlights the success of laws and serves, calling for stronger land registration systems policies for land reform, registration, and titling (Box and a greater overall need to continue strengthening 3). The National Agrarian Reform Institute (INRA) re- INRA’s institutional capacity. Furthermore, Bolivia fac- ports that land tenure in Bolivia was structured as fol- es a significant challenge in terms of sustainable land lows in 2017: peasant and intercultural communities use, leading to increasing deforestation and land deg- have a 21 percent share; Community Lands of Origin radation. This challenge is discussed in detail under (TCO) and Indigenous Community Lands (TIOC) have Chapter 7 (Section 7.2) of this report. a share of 23 percent; fiscal/state land represents 25 percent; land pending “clean” titling represents 21 115 Box 3. Secure land tenure to promote agriculture investment Access to agricultural land was addressed for the first time with the Agrarian Reform Act of 1953, which mainly affected properties in the Highlands and Inter-Andean valleys, where the majority of large estates were located at the time, and where the peasant population was most concentrated. The reform distrib- uted about 600,000 farms on 25–30 million hectares in the Highlands. Settlement programs between 1958 and 1985 distributed 3–5 million hectares in northern Santa Cruz, the Chapare region of Cocha- bamba, and the region of the Alto Beni to settlers, most of whom came from highland areas. Despite the achievements in breaking up the traditional haciendas, there was a resurgence of land concentration and lack of access to land in eastern Bolivia between the 1960s and 1990s. During this period, successive governments made large land grants, primarily in Santa Cruz, Beni, and Pando, of nearly 30 million hect- ares. This area has been the land base for the expanding livestock sector the boom in soybeans, wheat, sugarcane, and rice in Santa Cruz, making it the wealthiest area of Bolivia but also the most unequal.34 The second reform, the Agrarian Reform National Service Law of October 1996, sought to make land ownership more equitable, secure, and sustainable, but the results were not as expected. Projected to solve land tenure in 10 years, in practice the National Agrarian Reform Institute (INRA) was slow, cum- bersome, not very transparent, and expensive, and its results were rather poor.35 A new framework was provided by Law 3545 on Community Reconciliation of the Agrarian Reform of 2006, approved by a referendum in 2009. It opened the possibility of forming indigenous territorial au- tonomous entities and allowed communities to conserve customary land use, control the territory, and exercise self-government. Since then, significant progress has been made in regularizing land titles and property rights.36 INRA reports that during 1995–2014, 75 percent of the country’s rural properties were titled (70 million hectares and almost 650,000 titles).37 Most recent statistics by INRA indicate that in 2017 the cumulative figure reached 80 percent. The remaining area to be covered compromises 20.1 million hectares, of which one-third presents some kind of ownership conflict. About 36 percent of the more than 8.5 million hectares of state/fiscal land awaiting a “clean” title or titling has already been distributed, and the remaining 64 percent is in the processes of being distributed. A large proportion of the land awaiting a “clean” title or titling belongs to small-scale farmers (defined as farmers with less than 2 hectares in the Highlands Region and 50 hectares in the Lowlands Region). INRA also established a cadaster information system. The government has benefited from multilateral funding to undertake these reforms since 2002, with planned investments until 2022. 34353637 34 World Bank, Bolivia, Public policy options for the well-being of all, 2006. 35 Instituto Nacional de Reforma Agraria (INRA), breve historia del reparto de tierras en Bolivia, De la titulación colonial a la reconducción comunitaria de la reforma agraria: Certezas y proyecciones, 2008. 36 Under Law 3545, as of 2012, land rights had been formalized on 61 percent of the land in rural areas (24 million hect- ares) . Source. Universidad Mayor de San Andrés. Facultad de Ciencias Económicas y Financieras, Instituto de Investi- gaciones Económicas, Efecto de la implementación de la ley 3545 en el derecho propietario, La Paz, 2014. 37 International Development Bank (IDB), Bolivia, Programa de Saneamiento de Tierras II (BO-L1113) 116 6.3.2. A framework for agricultural dramatically. Even though INIAF is a strategically im- innovation portant institution for Bolivia, it has virtually no oper- ating budget to undertake agricultural research and An important institutional innovation introduced in transfer technology to producers. Under PISA the an- Bolivia was the establishment of INIAF in 2008 as nual expenditure was BS 1.5 million; now the annual the governing and coordinating body of the National budget spent by INIAF is BS 150,000. The institutional Agricultural and Forestry Innovation System (SNIAF, base can hardly be stable when 90 percent of the staff Spanish acronym). This institutional framework pro- is temporary (operating through a series of 3 3-month vided an opportunity for the government to resume contracts) and highly transient. The most significant its leadership in the development of science and agri- challenge is that technology transfer is stalled owing cultural technology and rural extension. Government to the lack of funding; virtual training and training of funding to consolidate SNIAF (including INIAF)was trainers in the departments and municipalities is be- complemented by funding from various bilateral and ing explored as a cheaper option. multilateral organizations, including the World Bank PISA project (2011–17, US$55 million, including gov- 6.3.3. Oversight for plant and animal ernment contributions). health and food safety This support made it possible to strengthen INIAF’s SENASAG is a deconcentrated public entity respon- strategic research programs substantially by upgrad- sible for plant and animal health and food safety in ing infrastructure, hiring staff, and funding a larger Bolivia. A succession of loans from the Inter-Ameri- operating budget. Infrastructure to modernize agri- can Development Bank (2010–22, US$125 million) is cultural experiment stations (upgraded to innovation enabling SENASAG to narrow its capacity gaps and centers), genebanks, seed testing laboratories, and strengthen its operations, particularly through invest- seed conditioning plants was accompanied by im- ments in laboratory facilities and the prevention of proved procedures for planning, implementing, mon- animal diseases. At the same time, this lending does itoring, and evaluating research, resulting in many not place much emphasis on narrowing capacity gaps operational efficiencies. INIAF’s technical assistance in key areas such as “sanitary diplomacy”38 and the programs were also strengthened so that INIAF could efficiency of SENASAG procedures and services. This increase its presence at the departmental level and agenda is increasingly relevant, however, to the gov- interact more closely with target groups and relevant ernment’s interest in an export strategy directed at stakeholders in the agricultural innovation system. new markets in China, India, Russia, and markets with Procedures for planning, implementing, and monitor- more stringent and complex entry requirements. ing seed certification processes were overhauled and to a large extent automated, allowing the INIAF Seed 6.3.4. Access to key infrastructure Directorate to achieve major operating efficiencies and technology: Irrigation and even as it significantly scaled up. mechanization More than 10 years on, this framework for agricul- As of October 2016, Bolivia had 7,121 irrigation sys- tural innovation—a major enabler of agricultural tems, covering about 387,377 hectares. Irrigation is productivity and competitiveness—has weakened a key determinant of agricultural productivity, resil- 38 “Sanitary diplomacy” can be described as efforts to develop consensus on food and agricultural safety and health standards that will facilitate the international movement of agricultural and food products. 117 ience, and improved income. Bolivia has made enor- on-farm irrigation systems (“technified irrigation”) mous efforts to expand irrigated area since 2006. The through programs such as Productive Alliances, PDES target is to increase the area under surface ir- with investments in on-farm irrigation estimated at rigation to 700,000 hectares, which is nearly double US$75 million, managed by EMPODERAR/MDRyT. It the current area. To support that objective, the gov- is estimated that during 2012–17 the government ernment, mainly through the Ministry of Environment invested about US$432 million (from current budget and Water (MMAyA) and its Vice Ministry for Water and through external loans) in the establishment of Resources and Irrigation (VRHR), has implemented a 83,236 hectares of newly irrigated land. series of programs with its own funding, multilateral Farm mechanization has also been a focus through- loans, and technical assistance from bilateral donors. out Bolivia since 2006. Mechanization increases pro- For example, the National Irrigation Program with a duction and labor productivity but can also enhance Watershed Approach (PRONAREC) (2009–21, total in- land productivity by ensuring that farm operations vestment of US$90 million in its phase, it is expand- are performed at the right time, which is increasingly ing considerably in its third phase (US$154.8 million), important in the context of climate change. The Unit which runs until 2021. Other programs coordinated for Productive Infrastructure, Local Technology and by VRHR through a watershed approach are the MI Mechanization (UIPTLyM, Unidad de Infraestructura RIEGO, financed by the Andean Development Bank Productiva, Tecnología Local y Mecanización), under (CAF),39 the Subprogram of Investments in Inter Com- the Vice Ministry of Rural and Agricultural Develop- munal Irrigation (SIRIC, Subprograma de Inversiones ment, is responsible for mechanization programs. en Riego Inter Comunal), funded by the German coop- Producers gain access to machinery through two eration; and the Climate Resilience-Integrated Basin channels: credit and grants. Management Project (2014-2020), funded with mul- tilateral funds from the World Bank. Legislation issued in 201641 gives the State respon- sibility for strengthening the agricultural production A program established in 2017 under MDRyT—”Our capacity of the 339 autonomous municipal govern- Well” (Nuestro Pozo)—focuses on installing wells40 ments and facilitating access to agricultural mechani- in all municipalities with agricultural production and zation to contribute to food security with sovereignty. livestock to ensure that water is available for food To achieve this objective, the Program for Municipal security. For the local contribution, the MDRyT signs Service Centers in Agricultural Mechanization42 was inter-institutional cofinancing agreements with low- created in 2018, with a budget of US$35.2 million (Bs er-level government entities. The government has 243,000,0000)43 financed through resources from the also been promoting the expansion of pressurized national treasury. 39 MI RIEGO aims to achieve sustainable increases in the agricultural incomes of rural households participating in the program by expanding the agricultural area under irrigation and improving water-use efficiency and distribution for agriculture. The program is implemented in seven departments (Cochabamba, Tarija, Chuquisaca, Oruro, Potosí, La Paz, and Santa Cruz) where community irrigation subprojects with a watershed approach will be cofinanced. 40 Nuestro Pozo, Programa Nacional de Perforación de Pozos de Aguas Subterráneas, established through Supreme Decree N° 2852. 41 Law No. 786 of March 9, 2016. 42 DS N ° 2785 of June 1, 2016 and DS N ° 3709 of November 7, 2018. 43 Additionally, Bs 38,890,080 for the payment of customs taxes for the import and storage of machinery, equipment and agricultural implements in customs areas. 118 For a five-year period starting in 2011, tariffs were re- Export licenses and permits are used to regulate ex- duced to 0 percent on imports of agricultural machin- ports of some products, such as sugar and by-prod- ery and equipment (plows, mowers, machinery for food ucts. Again, the export licenses are issued only if a preparation) and some agricultural inputs (seed, cattle surplus exists and, in the case of sugar, if the manda- feed, vaccines, and veterinary medicines). In 2016 this tory buffer stock, equivalent to two months of domes- measure was extended for another five years. tic consumption, has been built up. The government’s trade and market-related interventions are largely 6.3.5. Trade and domestic price policies the responsibility of the Ministry of Production Devel- opment and the Plural Economy (MDPyEP). Exports of The overall policy to ensure food sovereignty has em- some agricultural products benefit from the customs braced important market-related policy interventions, drawback procedure, which refunds all or part of the including trade policy, domestic price measures, and customs duty and other taxes paid by the exporter on the creation of state enterprises, as briefly discussed the inputs and other materials incorporated into the below. exported goods. Regarding agriculture and food imports, the tariff Price bands are applied to some products that are protection given to the agricultural sector is higher also subject to export restrictions, such as soybeans than that for manufacturing. In 2017, the average and rice. For soybean by-products, since 2011 do- tariff applied to agricultural products (13.3 percent) mestic prices have been set according to the require- was higher than the tariff applied to non-agricultural ments of domestic supply, and the domestic price products (10.8 percent) (WTO 2017). rises as domestic supply needs are fulfilled. Given Perhaps the most controversial measures have been the global decline in soybean prices, the industry has on the export side. As noted, the government intro- been advocating for the elimination of export quotas duced export controls in 2008, when it prohibited ex- and price-related interventions. Currently, the export ports of a range of agricultural products that were quota for soybeans is set at 300,000 tons. In May considered important to food sovereignty.44 Most of 2019, the government announced the expansion of those prohibitions were later replaced by the intro- the export quota, allowing exports of up to 60 percent duction a quota system for key exports. The govern- of total production. ment can also apply temporary tariffs and export re- An assessment of the effectiveness of the overall strictions on specific products, when relevant from trade and domestic price policies in agriculture in a food security perspective. A domestic supply and Bolivia is beyond the scope of this report. Early work fair price certificate (CAIPJ) and an export license or by FAO and others demonstrated that price stabili- permit are required prior to exporting agricultural zation efforts in several countries, including Bolivia, products. The CAIPJ verifies that domestic production through direct state interventions in the market, were meets domestic market demand and that there is an rather costly and difficult to manage, and often led to exportable surplus. adverse outcomes for agricultural producers.45 44 Live bovine animals, fresh bovine meat, chicken meats, wheat and wheat flower, corn, rice, and vegetable oil. 45 FAO, 2014. Policy responses to high food prices in Latin America and the Caribbean. Country Case Studies. Schuttel, C.; Kleinwechter, U.; Ihle, R.; Grethe H. 2011. Domestic policy responses to the food price crisis: The case of Boliv- ia. Journal of Agriculture and Rural Development in the Tropics and Subtropics. Vol. 112 No. 2 (2011) 125–139. 119 Furthermore, the most recent assessment of the enterprises operate in the sugar, dairy, seed, and impacts of Bolivia’s quantitative restrictions on ex- fertilizer industries, their share in their respective ports,46 particularly beef, show that these restrictions markets is relatively low.47 Although it is unlikely that had a negative impact not only on total production but these state enterprises in have the power to influence also on production for the domestic market. In Febru- agricultural prices and markets (WTO 2017), they can ary 2008, the government banned exports of beef as discourage private sector investment in the sector. In well as other products such as maize, sugar, poultry, the case of the production of fertilize the public sector and rice. Unlike other products, for beef the prohi- in Bolivia is expected to have an important impact on bition on exports was not relaxed until March 2012, agricultural productivity through the production and when Bolivia imposed a quota system that allowed supply of urea, and a national urea production plant producers to export beef one the domestic market established a few years back is expected to supply an cleared. Although the export restriction was expect- important share of the national demand and supply ed to reduce the domestic price of beef, the domestic regional markets. In fact, urea exports have grown, price of beef continued to rise, even faster than the and Bolivia is currently exporting product to Brazil, international beef price. Paraguay, and Uruguay. The effects of the public pro- duction/supply of fertilizer on the overall input price The Bolivian authorities are currently applying a less and productivity remain to be seen, and they need to restrictive trade policy with respect to exports. For be closely monitored to understand the real impacts. example, although export quotas persist, they have gradually increased, and the domestic supply and fair The gains achieved so far in modernizing the agricul- price certificate (CAIPJ) is no longer required for most tural trade and market policy framework in the pres- products. There are a few exceptions; for instance, ent decade (such as the increases in export quotes) exports of wheat and wheat products are banned. have being the result of pressures from the private sector. Those pressures are gradually intensifying In addition to import and export interventions, in Bo- as Bolivia encounters a more complex international livia the state is expected to play an important role in trade environment and as margins in supplying the agriculture and food processing, particularly through domestic market are increasingly narrowing. Close direct participation in production and processing and coordination and collaboration between the govern- purchasing of local output at fair prices to the produc- ment and the private sector in Bolivia will be critical ers and the sale of these products to consumers. The to ensure that the most effective policies are in place authorities consider that public enterprises comple- to support improved competitiveness of agricultur- ment private Bolivian or foreign economic actors. In al producers and enterprises, while protecting the the mining/hydrocarbon sector, public enterprises re- well-being of urban and rural consumers. tain a significant share of trade, particularly exports. In the agriculture and food sector, although state 46 (1) Garcia, E; Rossio, M; Stucchi, R. (2017). The impact of export restrictions on production: A synthetic controls ap- proach. December 2017. (2) Burgoa, R; Herrera, A. 2016. El efecto de las restricciones a la exportación de carne bovina sobre la producción y oferta doméstica. Published in the Bolivian Central Bank website. 47 Empresa Azucarera San Buenaventura has a 1.26 percent share of the sugar market, while the Empresa de Lácteos de Bolivia (LACTEOSBOL) has a 4 percent share of the dairy market. Likewise, the Empresa Pública de Abonos y Fer- tilizantes, which imports and markets fertilizer, imports a volume equivalent to 1.29 percent of demand, and the seed output of the Empresa de Semillas represents 1.88 percent of domestic production (Word Trade Organization. Trade Policy Review Bolivia. September 2017). 120 6.4. Building farm and value chain Pro-Camelidos project. A total of 461,711 benefi- capabilities ciaries are or will be reached through these invest- ments. This figure does not include past investments Complementing investments in the enabling environ- and beneficiaries of the defunct National Indigenous ment, a second set of investments directly support Fund or those expected under the New Indigenous production and productivity of rural producers and Development Fund (FDI) that runs until 2021 (and is their organizations (see Annex 3). In line with overall expected to reach nearly 300,000 beneficiaries). Nor government policy, most of these programs/invest- does it include those reached through the National ments have mainly targeted poor and marginalized Fund of Integral Development (FONADIN). producers and communities, often located in remote areas with high levels of poverty and vulnerability. The evaluations (including impact evaluations) car- Consequently, market integration and competitiveness ried out for some of these programs/projects have have not been an immediate priority. Most of this sup- reported important positive results in terms of pro- port has focused on production—the enhancement of duction and yield increases, and household impacts at on-farm assets and technologies and the provision of the income level. Important lessons can be identified seed capital and technical assistance. A small subset from Bolivia’s experience and similar experiences in of these investments has a clear market-orientation the region, which are relevant to the government as it and focuses on strengthening market linkages. explores new/complementary approaches to support small-scale producers, particularly within the context Only few projects target the more entrepreneurial of developing value chains. The following paragraphs segments of the smallholder population with a higher represent an attempt to summarize those lessons capacity for market integration. The Productive Alli- and translate those lessons into opportunities to en- ances Project stands out in this regard; for producer hance the effectiveness of the support to small-scale organizations to receive project support to finance farmers in Bolivia. their business plans, they were required to establish a link to a buyer. Even so, project support focused on Table 20 summarizes current efforts to enhance the primary production and productivity, and the produc- productive capacity of smallholders and proposes er group linked only with the immediate buyer, often some shifts and complementary approaches to en- an informal trader who already had a longstanding hance the effectiveness of this support. The discus- relationship with the producer group. No specific sup- sion that follows the table highlights these opportu- port was provided to strengthen off-takers or other nities for pursuing complementary approaches or downstream players and remove downstream bottle- strengthening current efforts. necks that could sizably affect producers. Broaden intervention entry-points to reach other val- Commitments to support smallholders through sev- ue chain actors en programs over 2006–22 (excluding investments through the Indigenous Fund and FONADIN) are ap- Most of the government programs for agriculture proximately US$525–540 million (Annex 3). Many of discussed here channel support through producer these programs have ended or are ending in 2019, organizations or other venues for collective action by however, with the exception of the Direct Support farmers and communities. As noted, however, the role for the Creation of Rural Agri-food Initiatives (CRIAR of other key actors such as aggregators, buyers, and II) project, Productive Alliances (PAR II) project, and industry associations is often overlooked. As men- 121 Table 20. Current and enhanced approaches to support smallholder competitiveness Area Features of current smallholder support Desired enhanced aspects - Create capacities a different value chain levels - Capacity strengthening of producer (organized producers and other key value Intervention target organizations (POs) to enhance market access/ chain actors: processors, aggregators, service linkages providers, input providers, buyers) - Demand driven+ identification of opportunities Approach - Demand driven (by POs) (geographical and in value chains/subsectors) - Productive & logistic efficiency + resilience + Intervention focus - Productive efficiency sustainability - Strengthening internal capacities of - Services: AT+ in very few cases financial POs but also other value chain actors, services via a combination of approaches: skills Intervention content - Grant financing: for inputs+ capital+ productive development, facilitate access to key services assets and input markets. Tranversal: use of digital - Skills & Knowledge gaps via training. technologies. Vertical linkages/ - Promoted/facilitated, active role of buyers/ - As given (producer-buyer linkages) coordination private players. - Strengthening POs, but few consolidation of - Promotion of production clusters/networks to Horizontal linkages lessons on effective approaches ensure aggregated impacts territorially. - Focus on value generation (combination of Focus of the upgrades - Largely product upgrades (focus on on-farm/ product, functional, process, and market supported POs investments) upgrades). Focus on on-farm and off-farm investments along the chain. Private sector role - Generally a passive or very limited role - More active role of the buyer (private sector) - Multi-linkages and coordination of investments Common enablers - Rarely (roads, electricity, financing) - Links to enhanced policy & regulatory Sectoral enablers - Generally not considered frameworks and institutional capacities. tioned in Chapter 5, most of the successful experienc- coordination, the risk of fragmented or overlapping es with inclusive value chain development worldwide investments is high, and there are few possibilities are anchored on integrated support—in other words, for understanding the aggregate benefits at the value they target several intervention points to address chain/subsectoral level of investments made through productive and competitiveness issues throughout grants and other services. the value chain. They also often include a strong Furthermore, most of the investments have been im- platform for leadership, coordination, and strategic plemented using a demand-driven approach, which priority setting for the value chain. Such platforms has been instrumental for generating ownership and bring a broad set of institutions and other stakehold- empowering communities and producer organiza- ers to the table (for example, public-private coordi- tions. However, this approach would need to be com- nating platforms or innovation platforms). Through plemented with targeted more strategically invest- these platforms, support currently provided through ments around subsectors (products/value chains), to grants to producer organizations can be easily linked generate aggregated impacts. to ongoing or proposed value chain strategies. With- out such platforms for strategic priority setting and 122 alone, but rather as part of a broader strategy to build Systematically build capacity among producer orga- a strong innovation system by building capacity in nizations public and private actors aside from farmers, includ- ing extension service providers. Although most investments are channeled through producer organizations, and although most programs Seek opportunities for digitizing agricultural services seek to strengthen the capacities of producer orga- nizations, little is known about which approaches are Digital solutions in the so-called “last mile” of ag- most effective. Significant gains could be generated ricultural value chains have numerous advantages by sharing experiences of working through producer and applications. They help to maximize operational organizations and building consensus on the most efficiencies, promote financial inclusion, empower effective ways to enhance their capacity. Under the individual by making transactions more transpar- Productive Alliances project (PAR II), an interesting ent, reduce travel times and transaction costs, and approach is being developed. Lessons from that ex- support information dissemination and traceability. perience should be systematized and disseminated There are important opportunities for these solutions widely. Some tools and instruments for assessing the to revitalize extension services and other services for capacity and performance of producer organizations48 farmers, increase program efficiencies, and enhance are coming into wider use throughout the world. They supply chain coordination. can be customized to enhance efforts to build capac- ity in producer organizations and can also be applied Enhance the role of the private sector and strongly more broadly to organizations and businesses across support a market orientation the agricultural sector. The production approach that has dominated de- Strengthen service providers, as part of the agricul- velopment support to smallholders in Bolivia would tural innovation framework benefit from a stronger market orientation. The de- velopment of supermarkets, agroindustry, and other As part of the support provided to farmers, de- formal market outlets may be moving more slowly in mand-driven programs often include a set of ser- Bolivia than in other countries in the region, but op- vices that can range from advice and guidance in portunities nevertheless exist to link farmers to mar- the identification, preparation, and implementation ket developments. The emerging gastronomy value of proposals, to technical assistance, extension, and chain and export markets both present options for training. It is generally assumed that capable service participation by small-scale producers. The develop- providers will always be available and that they will ment of such opportunities will entail close collabo- provide services at the right time, but often this as- ration with the private sector and enhanced coordi- sumption is incorrect. Service providers may need nation and collaboration with all participants in the training and capacity building as part of an integrated value chain. Especially in export markets, strategies approach to support the improvements demanded by to support “sanitary diplomacy” and export promotion smallholders. This should not be seen as a strategy will need to be closely aligned. to be carried out by individual programs and projects 48 An example of such is Scope Insight https://www.scopeinsight.com/, which is increasingly mainstreamed through development programs. 123 Promote linkages to financing interventions Seek convergence with common enablers in the wid- er economy An additional opportunity to explore is to improve the complementarity of grant funds with financial Better alignment of agricultural investments and services and credit schemes to expand the mid- and programs with the common enablers of growth and long-term impacts of these programs. PAR II is sup- competitiveness in the wider economy will increase porting initial efforts to enhance financial literacy their impact. A better understanding of regional and and improve producers’ links to financial services. local opportunities for agricultural investment will The lessons emerging from these efforts need to be facilitate better planning and help to tailor incentives properly systematized and disseminated. and support to make the most of those opportunities. The capacity for planning should be strengthened and Seek the convergence of programs and investments incorporate methodologies for identifying the areas that offer the greatest agricultural potential with low Initiatives to promote the development of small-scale environmental impacts, and where agriculture can agriculture sometimes end up operating in silos, make an important contribution to economic growth oblivious to other ongoing programs and projects that and poverty reduction. Through this kind of planning offer potential synergies. Policy makers and program and priority setting, efforts at the farm level can be administrators need to better understand the “big aligned with key enabling investments, such as in- picture” and pursue complementarities with other ini- vestments in infrastructure (roads, electricity, inter- tiatives. Furthermore, the setting of robust and inte- net connectivity, and so on). grated monitoring systems around these cofinancing investments is a critical area requiring government Seek opportunities to strengthen climate-smart prac- attention. Under PAR II, a first-class monitoring and tices and support better nutrition reporting system has been developed and could serve as a good foundation for a broader and more integrat- Climate change and nutrition are important cross-cut- ed monitoring and reporting system for the sector. ting aspects of efforts to build farm and value chain capabilities. As the next chapter will discuss in detail, Seek convergence with key enablers in the agricul- climate shocks (as the name implies) are major dis- tural sector ruptors of any sustained pattern of agricultural im- provement. The Government of Bolivia is proactively The combined impact of programs to improve small- supporting approaches to adapt to climate change holder productivity in Bolivia could be enhanced by and opportunities to make agriculture less vulnerable aligning the opportunities for agriculture and agri- to climate disruptors. It can achieve even more if cli- business with key enablers in the agricultural sector, mate-related investments are mainstreamed across such as research and innovation systems, sanitary the portfolio of agricultural investments. The same and phytosanitary programs, and efforts to expand mainstreaming approach is valuable for nutrition; for irrigation, mechanization, and land titling. example, it will facilitate opportunities to link agri- cultural investments with nutrition education and to pilot approaches that incorporate nutritional consid- erations in agricultural interventions. 124 7. Sustaining Productivity and Competitiveness Gains 125 Photo: Shutterstock This chapter looks to the future to understand how longed dry periods and an increase in the frequency gains in agricultural productivity and competitiveness and magnitude of floods, landslides, and other weath- might be protected from the risks that accompany cli- er-related events (World Bank 2013; WRI 2017). mate change and from cycles of worsening environ- Bolivia is already experiencing the effects of greater mental degradation. It looks at the production risks climate variability and more extreme climate events. and potential losses that Bolivian agriculture could ex- The Global Climate Risk Index (CRI) 201849 ranks Bo- perience through climate change and at government livia 25fth of 178 countries in terms of the impacts of efforts and other approaches to counter those risks. climate-related hazards during 1997–2016. In 2016, The second part of the chapter focuses on strategies Bolivia was among the top 10 countries regarded for reducing the environmental footprint of agricul- as most highly affected by climate-related events.50 ture, not only to preserve the resource base for agri- About 74 disasters have been recorded between 1900 culture but to mitigate the effects of climate change. and 2017.51 Droughts and floods jointly account for roughly 97 percent of the total number of affected 7.1. Climate risks and the future of people. Most of these events are influenced by El Bolivian agriculture Niño (more properly, the El Niño Southern Oscillation, The agricultural sector in Bolivia is exposed to a va- ENSO). riety of risks related to production, markets, or the Agriculture (including livestock production) is inextri- regulatory environment. Shocks related to price and cably linked to climate and weather, and worldwide climate are the most obvious and are often intrinsi- it is highly vulnerable to climate change. Climate cally linked. Bolivia’s geographical and socioeconomic change scenarios for Bolivia reveal patterns of in- characteristics and low institutional capacity to mit- creasingly variable rainfall and a greater incidence igate climate risks make it highly vulnerable to the of drought, which have important implications for qui- impacts of climate change. Throughout the Andean noa, maize, soybean, and potato production systems region, Bolivia is recognized as the country where the (Box 4). Scenarios predict that crop yields will decline impacts of climate change may have accelerated most (up to 40 percent for maize and soybeans) owing to rapidly. During the last 50 years, the country has lost water shortages and hot spells during critical crop approximately half of the surface area of its mountain growth stages. glaciers. Higher projected temperatures and rainfall during the rainy season will expose the country to pro- 49 The Global Climate Change Risk Index, developed by Germanwatch, analyzes the quantifiable impacts of extreme weather events, both in terms of fatalities and economic losses, based on data from the Munich Re NatCatSERVICE. 50 The most catastrophic disasters, including floods, droughts, and landslides, accounted for US$3.1 billion in damages over a 35-year period (1982–2016) or US$91.1 million on average per year. 51 The number of natural disasters increases up to a total of 89 if earthquakes (03 disasters) and epidemics (12 events) are included. International Disaster Database (EM-DAT). 126 Box 4. Anticipated effects of climate change The direct effects of climate change in Bolivia are manifold, with substantial variation across sub-regions. In the Andes, sustained glacial recession has been documented in the last 50 years, especially since the second half of the 1980s, placing drinking water sources as well as the energy reserves for major cities like La Paz and El Alto at risk (Morales 2010). At the same time, desertification is progressing rapidly in the Altiplano, the inter-Andean valleys, and the Chaco, which are home to the bulk of Bolivia’s population. Desertification now affects 41 percent of the national territory (Mariscal 2011), with aridity indices52 pro- jected to decrease by up to 20 percent in a 2°C world and up to 40 percent in a 4°C world (World Bank 2014). In the Amazon Region, increased climate variability is inducing a higher incidence of drought as well as more frequent and intense flooding during the rainy season, leading to significant water stress: by 2050, the estimated cost of the additional water storage required to meet the monthly need for irrigation water as a result of climate change will range from US$12 million to US$60 million (World Bank 2010). Figure 36. Projected climate change impacts on rainfall and temperature, Bolivia, 2040–69 a) Rainfall b) Temperature Source: Built with World Bank Spatial Agent, using data from Climate Wizard Note: Projections under A2 Scenario, 60% ensemble model, annual season. 52 52 The Aridity Index (AI) is an indicator designed for identifying structurally arid regions, i.e. regions with a long-term av- erage precipitation deficit. Lower AIs indicate higher levels of aridity, with hyper-arid areas, generally coinciding with the great deserts, typically displaying AI values between 0 and 0.05 (World Bank, 2014). 127 Climate change scenarios reveal patterns that further affect drought and flood risks in food systems across Bolivia. An increase in the variability of rainfall and in the occurrence of periodic droughts could have important implications for quinoa, maize, soybean, and potato production systems. Under a dry scenario, crop yields are expected to decline (up to 40 percent for maize and soybean), mostly due to water shortages and hot spells during critical crop stages.53 A climate change impact risk analysis was conducted to identify priority food systems for government action in the face of climate change.54 Any disruption, not only in food production but also in food processing, marketing, and consumption, will affect food security outcomes in the most vulnerable areas, both rural and urban. Food commodities that are highly exposed and sensitive to floods and/or droughts include cereals (maize, rice, wheat, and quinoa); industrial commodities (soybeans, sunflower seed, and sugarcane); tubers (potatoes and cassava); meat, milk, and eggs (cattle, llamas, and chickens); vegetables (peanuts, green beans, green peas, beans, onions, and carrots); and fruits (bananas and plantains). All of these commodities (fresh and processed) are highly relevant for food security, as part of the Bolivian food basket. Table 21 presents a summary of relevant hazards for food commodities by region and department. The National Strategy for Agriculture Risk Management and Climate Change Adaptation (MDRyT 2018) pro- poses three macro-regions, according to their level of risk: Region I (R I) – Altiplano and Valles, Region II (R II) – Amazonía and Llanos Tropicales; and, Region III (R III) – Chaco. Regions I and III are more exposed and sensitive to drought, while Region II is more exposed and sensitive to floods. 5354 53 World Bank 2010. 54 Developed on the basis of secondary data: national census and surveys from National Statistical Institute (INE), The Agro-environmental and Productive Observatory from Ministry of Rural Development and Land (MDRyT), The Agricul- tural Risk and Climate Change Atlas for Food Sovereignty (MDRyT, 2014), and the National Strategy for Agricultural Risk Management and Adaptation to Climate Change (MDRyT, 2018). 128 Table 21. Major hazards for food commodities by region and department Region Major Hazard Food Commodities - Crops Departments Quinoa (67,000 ha), Potato (50,500 ha), Wheat (14,000 ha), Chuquisaca, Cochabamba, La Paz, Green Beans (9,000 ha), Green Peas (9,000 ha), Banana RI Drought Oruro, Potosi, Santa Cruz and (4,000 ha), Onion (3,800 ha), Carrots (3,200 ha), Plantain Tarija (3,000 ha) and Beans (2,000 ha) Soy (1,475,000 ha), Sunflower (164,200 ha), Rice (150,000 ha), Corn (130,000 ha), Sugar Cane (124,000 ha), Wheat RII Floods (92,000 ha), Beans (47,900 ha), Cassava (15,300 ha), Beni, Cochabamba y Santa Cruz Plantain (13,700 ha), Banana (11,400 ha) and Peanuts (3,600 ha) RIII Drought Corn (80,000 ha), Peanuts (9,400 ha) and Beans (6,500 ha) Chuquisaca, Santa Cruz and Tarija Region Major Hazard Food Commodities - Livestock Departments Chuquisaca, Cochabamba, La Paz, RI Drought Poultry (215,000 chickens) and Llama (1,280,000 heads) Oruro, Potosi and Tarija Poultry (176,000 chickens) and Cattle - beef and milk RII Floods Beni and Santa Cruz (3,359,000 heads) Poultry (211,000 chickens) and Cattle - beef and milk RIII Drought Santa Cruz and Tarija (373,000 heads) Source: World Bank/FAO 2019 (unpublished background report on Agriculture and Climate Change in Bolivia) The effects of unpredictable rainfall patterns on agri- production risk assessment was carried out across culture are often severe, such as the effects of erratic 9 departments for 14 crops for which 2017 wholesale rainfall during the 1997/98 El Niño. This event led to prices were available, as well as crop yield data for damages in maize and soybeans valued at around 1994–2017. The analysis aimed to: (1) provide indic- US$641 million. Similarly, the total economic impact ative figures of gross production concentration val- of the El Niño/La Niña weather pattern in 2007/08 ues, (2) survey historical in-country losses, including was estimated at US$443 million—equivalent to 62 losses recorded for selected crops, and (3) assess the percent of total economic losses or 4 percent of 2007 exceedance probability of occurrence of losses. The GDP.55 Likewise, La Niña 2017/1856 affected approx- results are discussed in the following section. imately 2.25 million hectares, many of which were planted with summer crops 7.1.1. Estimating losses from climate- related production risks If agricultural growth in Bolivia is to deliver sustained gains for producers, value chain actors, and consum- The analysis of climate-related production risk shows ers, agricultural production systems will have to be- that the exposure of selected crops, measured in come more resilient. To understand the long-term terms of GPV, was equivalent to US$7.05 billion over economic impacts of climate hazards in Bolivia, a 3.05 million hectares. Santa Cruz, Cochabamba, and 55 See https://www.gfdrr.org/bolivia 56 Presentation: Afectaciones Fenómeno Niña 2017-2018. Período: Dic. 17 – May18 (2018) Accessed: November 2018. MDRyT. 129 Table 22. Gross Production Value (GPV) per department for selected crops in Bolivia (in million US$). Period 1994-2017 Santa Grand Pecentage Crops Beni Chuquisaca Cochabamba La Paz Oruro Pando Potosi Tarija Cruz Total distribution Rice 53.99 0.35 8.10 11.28 - 2.83 - 285.16 0.60 362.30 5% Banana 2.94 - 109.72 35.56 - 0.53 - 2.33 0.17 151.26 2% Sugar cane 11.15 2.36 1.59 - - 0.80 - 1,863.41 115.84 1,995.16 28% Berley - 3.55 1.72 8.45 2.65 - 7.02 0.04 0.30 23.73 0% Bean 0.57 5.46 3.89 0.59 - 0.29 0.0 108.61 0.68 120.11 2% Maize/ 1.68 9.37 7.47 2.14 0.01 0.57 2.38 59.95 14.44 98.00 1% Corn Mandarin 0.96 0.25 33.24 19.42 - 0.13 - 73.23 1.89 129.12 2% Potatoes - 62.48 162.72 183.86 41.05 - 74.07 32.96 29.16 586.31 8% Pineapple 4.09 0.17 157.35 4.21 - 1.90 - 3.71 - 171.43 2% Plantain 98.78 0.42 490.53 175.14 - 52.51 - 78.99 0.03 896.41 13% Quinoa - 0.12 0.66 15.45 61.34 - 43.33 - 0.01 120.91 2% Sorghum - - - - - - - 208.91 1.33 210.24 3% Soybean - 0.12 - - - - - 2,008.59 14.80 2,023.51 29% Wheat - 7.98 11.90 0.86 0.38 - 9.25 131.58 0.62 162.57 2% Grand 174.17 92.64 988.88 456.95 105.43 59.56 136.08 4,857.48 179.86 7,051.05 Total Pecentage 2% 1% 14% 6% 1% 1% 2% 69% 3% distribution Source: Authors’ estimates based on the Observatorio Agroalimentario y Productivo. La Paz Departments had the largest share of GPV ex- As the data for the study period indicate, in a year posed to production risk—around 89 percent. In other when an extreme weather event occurs, crop yield words, the whole agricultural sector is likely to be losses can be almost three times as high as the av- severely impacted if any of these departments is hit erage annual loss. In 1999, for instance, crop yield by a disaster. losses amounted to US$879.7 million across all crops and departments, but three crops accounted for 84 Over the period from 1995 to 2017, total losses due percent of total loss (Figure 37). Extreme crop yield to production risks in Bolivia are estimated at US$6.4 losses also elevated losses in 2010 (US$594.0 mil- billion. Soybeans and sugarcane accounted for 66 lion) and 2007 (US$529.5 million). percent of all losses (around US$4.2 billion in total). The average annual loss across all crops and depart- Although the historical burn analysis can give a use- ments under analysis is US$277.3 million (3.93 per- ful indication of possible losses generated by past cent of total GPV). In terms of loss frequency, both events, it may not be adequate to identify extreme wheat and sorghum experienced losses in yield once losses which are likely to occur. To obtain more ac- in every two years. Yield losses occurred twice in ev- curate estimates (or forecasts) of expected crop yield ery ten years in soybeans and three times in every ten losses in Bolivia, a Monte Carlo simulation was run years in sugarcane. on 10,000 crop yield scenarios for each crop and de- partment. This analysis shows that the average loss 130 Table 23. Annual Average Losses (AAL) and loss frequency per crop (1995-2017). Crops AAL (US$ Million) AAL (%) Loss Frequency (%) Rice 15.46 4.27 36.36 Banana 1.54 1.02 33.33 Sugar cane 74.56 3.74 31.82 Barley 0.96 4.05 29.87 Bean/Poroto 1.72 1.43 38.64 Maize/Corn 3.32 3.39 27.27 Mandarin 2.52 1.95 36.36 Potatoes 8.01 1.37 38.96 Pineapple 1.68 0.98 43.94 Plantain 24.94 2.78 41.56 Quinoa 4.14 3.43 40.91 Sorghum 14.74 7.01 50.00 Soybeans 109.01 5.39 21.21 Wheat 14.65 9.01 54.55 0.00 0.00 Total 277.26 3.93 38.21 Source: Estimates based on the Observatorio Agroalimentario y Productivo (US$285.6 million) is slightly above the value obtained els. The production risk assessment also shows that from historical values (US$277.3 million), clearly re- wheat, sorghum and soybeans are the top three crops flecting a more accurate distribution of losses than with the highest loss risks, as their weighted average can be obtained through simple consideration of the losses accounted for 9.35 percent, 7.46 percent, and 24 historical loss values available for each crop and 5.61 percent, respectively. Due to the size of the GVP, department. In this respect and when considering the however, the largest losses per crop are clearly con- maximum amount expected to be lost over 1-in-100 centrated in soybeans (US$113.4 million), sugarcane year return period (or Value at Risk—VaR—with an (US$74.2 million), and plantains (US$24.8 million). exceedance probability of 1 percent), the simulated In soybeans, the VaR with an exceedance probability data exhibit a loss of around US$892.9 million (or of 1 percent is above one-quarter (US$520.1 million) 12.7 percent of the total value exposed). Agricultural of its GPV (or US$2.02 billion). At the department risk management strategies must be designed based level, the same VaR extends from a minimum of 7 on the likelihood and severity of losses being realized. percent in Chuquisaca to a maximum of 26 percent The average loss calculated from simulated crop in Santa Cruz. In contrast, the same VaR in quinoa yields shows that the departments in Bolivia with exhibits a weighted average loss equivalent to 13 the highest risk levels are Santa Cruz, followed by percent (or US$15.6 million) of its GPV (or US$120.8 Cochabamba and Tarija. In contrast, the departments million). The departments with the lowest and highest of Pando and Oruro feature the lowest hazard lev- 131 Figure 37. Historical crop yield losses of selected crops in Bolivia (US$ million) (1995-2017) 1.000,0 879,7 900,0 800,0 700,0 594,0 529,5 600,0 509,5 458,0 441,2 500,0 402,6 410,6 369,6 400,0 324,7 300,0 225,1 147,0 154,6 200,0 108,1 103,6 121,8 113,7 101,8 116,1 129,9 57,9 58,8 100,0 19,4 - 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: Estimates based on the Observatorio Agroalimentario y Productivo Figure 38. Value at Risk (VaR) with different exceedance probabilities in all crops and departments studied. $1.200,00 $1.000,00 $1.131,20 $1.067,67 $1.106,94 Value at Risk (US$) $1.010,35 $1.122,13 $800,00 $1.097,11 $1.037,32 $892,88 $804,28 $956,21 $600,00 $400,00 $200,00 $- 0 50 100 150 200 250 300 350 400 450 500 550 Return Period (years) VaR 1-in-100 year return period are Chuquisaca (6.8 nent intervention has been the effort to develop an percent of its GPV) and Oruro (21 percent of its GVP). agricultural insurance market. In 2011, the National Agricultural Insurance Institute (INSA, its Spanish ac- 7.1.2. Government efforts to manage ronym) was established, with a mandate to operate production risks and regulate the national agricultural insurance pro- gram (also known as “Pachamama”) and the insur- In response to the exposure of the agricultural sector ance premium subsidy program allocated annually to climate-related risks, the Bolivian government has for different insurance schemes (Box 5). supported a series of policy interventions. A promi- 132 Table 24. Simulated weighted average loss cost per crop and per department. W Avg Department/ Sugar GPV (US$ Rice Banana Barley Bean Maize Mandarin Potatoe Pineapple Plantain Quinoa Sorghum Soybean Wheat Loss (US$ Crops Cane Million) Million) Beni 1.8% 5.3% 2.0% 5.6% 6.6% 2.8% 6.2% 6.2% 174.06 7.89 Chuquisaca 1.1% 1.9% 0.8% 1.0% 1.2% 1.1% 1.4% 0.5% 1.3% 0.9% 0.9% 5.9% 92.58 1.63 Cochabamba 1.0% 1.2% 2.5% 2.7% 1.1% 1.1% 1.2% 1.7% 0.8% 2.4% 2.3% 4.9% 988.26 18.35 La Paz 0.5% 0.5% 6.7% 1.7% 3.9% 7.6% 0.8% 2.3% 2.0% 2.8% 5.2% 456.66 7.85 Oruro 9.9% 7.9% 1.8% 3.3% 7.6% 105.36 3.09 Pando 3.3% 4.5% 5.7% 9.9% 3.6% 2.9% 8.0% 3.4% 59.53 2.15 Potosi 0.7% 3.9% 4.3% 2.2% 3.6% 4.5% 135.99 3.71 Santa Cruz 5.5% 1.0% 3.5% 0.9% 2.2% 4.8% 0.7% 1.7% 1.7% 1.9% 7.5% 5.6% 10.4% 4,854.42 231.27 Tarija 2.2% 0.7% 7.6% 1.1% 0.8% 2.3% 1.6% 1.5% 1.6% 2.3% 1.3% 0.7% 1.7% 179.75 9.70 GPV (US$ 362.07 151.17 1,993.90 23.71 120.04 97.94 129.04 585.94 171.32 895.84 120.83 210.11 2,022.24 162.46 7,046.61 million) W Avg Loss 4.65% 1.14% 3.72% 4.04% 2.16% 3.81% 1.92% 1.43% 1.03% 2.76% 3.34% 7.46% 5.61% 9.35% (%) W Avg Loss 16.82 1.72 74.18 0.96 2.59 3.74 2.47 8.38 1.76 24.77 4.04 15.68 113.35 15.20 285.65 (US$ Million ) Table 25. Value at Risk with an exceedance probability of 1% for different crops in Bolivia. Department/ Sugar Rice Banana Barley Bean Corn Mandarin Potatoes Pineapple Plantain Quinoa Sorghum Soybean Wheat Crops Cane Beni 13.79% 27.37% 18.16% 33.64% 39.10% 11.19% 33.12% 32.51% Chuquisaca 8.11% 9.95% 5.41% 7.66% 8.60% 8.23% 8.18% 3.88% 9.19% 6.76% 6.96% 15.73% Cochabamba 7.48% 8.76% 9.78% 16.29% 11.37% 10.25% 7.28% 10.57% 5.91% 11.78% 20.79% 11.34% La Paz 4.21% 3.44% 35.89% 7.55% 23.53% 42.68% 5.87% 20.69% 10.67% 17.82% 12.22% Oruro 38.74% 45.95% 9.47% 20.93% 41.08% Pando 17.84% 27.37% 31.84% 41.88% 21.39% 11.45% 41.21% 14.95% Potosi 5.72% 34.23% 28.15% 19.78% 16.94% 13.36% Santa Cruz 46.88% 6.32% 28.23% 9.55% 20.32% 29.39% 5.76% 10.54% 7.64% 11.08% 36.99% 25.91% 64.90% Tarija 16.35% 7.57% 30.45% 7.73% 7.55% 16.40% 7.42% 14.22% 10.78% 10.39% 8.50% 7.43% 11.24% VaR 1-100 37.51% 6.65% 26.88% 14.68% 18.57% 18.77% 7.88% 4.69% 5.82% 8.71% 12.88% 36.76% 25.72% 53.54% years (%) 133 Box 5. Agricultural insurance in Bolivia SAMEP insurance program was designed by INSA in the form of a compensation payout mechanism (also known as PIRWA) to protect small-scale farmers who had historically been excluded from commercial agricultural insurance. Since 2012/13 when the SAMEP instrument became available, the cultivated area protected through this coverage increased from 91,226 hectares to 315,693 hectares. In the five years from 2013/14 to 2017/18, compensation payouts amounted to approximately Bs 88.8 million, covering 113,866 small-scale farmers. Because of El Niño, the 2016/17 crop season had both the largest area affected (exceeding 139,000 hectares) and the highest number of enrolled farmers who received payouts (more than 43,500 farmers). Table 26. Evolution of SAMEP program (2013-2017) Nº Nº Nº Farmers Covered Area Affected Area Compensation Compensated Year Municipalities Communities (covered) (ha) (ha) payouts (Bs.) farmers 2013 63 2,491 57,497 91,226 8,469 7,601,460 7,141 2014 107 4,631 106,049 175,892 19,910 18,061,748 23,425 2015 141 5,575 146,563 272,886 6,866 6,463,856 11,148 2016 142 5,379 135,456 249,893 33,992 32,111,987 43,536 2017 140 6,186 168,705 315,693 25,338 24,560,182 28,616 Total 593 24,262 614,270 1,105,590 94,575 88,799,233 113,866 Avg 119 4,852 122,854 221,118 18,915 17,759,847 22,773 Source: INSA (Seguro Agrario en Bolivia. Modalidad Catastrófica “PIRWA”) Compared to other programs developed in the region, the Bolivian program pays the highest indemnity value per hectare. This SAMEP compensation payout was established with the objective of covering farm- ers´ food expenses if insured crops suffered severe losses owing to extreme weather such as a drought, flood, frost, or hailstorm. Enrolled farmers are entitled to received Bs 1,000 per hectare (approximately US$147.05 per hectare) for up to three hectares. In comparison, Mexico’s program for responding to natural disasters (CADENA) has an insurance mechanism that protects the federal and state govern- ments from the financial losses incurred in aiding farmers who grow rainfed crops and have no access to commercial insurance. When an insured event occurs, the policyholder (the federal and state govern- ments) distributes a pre-agreed lump sum per hectare (Mx$1,500 or US$75 per hectare) of rainfed crop. In northeastern Brazil, the popular Garantia Safra program disburses R$850 (equivalent to US$216.87)57 divided into five installments to farmers whose monthly family income is less than 1.5 times the mini- mum wage and who crop 0.6–5 hectares. The Garantia Safra compensation is triggered when the yield of target crops at the municipality level is 50 percent or less than the expected yield. 57 57 Exchange rate of R$3.92 per US$1 on December 12, 2018. 134 Community leaders and community organizations play an essential role in running SAMEP. The 60 INSA technicians cannot meet the needs of the 122,800 farmers who benefit annually from the program, so work contributed by representatives of municipal administrations and community leaders is critical for supporting INSA technical staff in enrolling producers and assessing field losses. A training program to enhance efficiency, improve skills, and standardize SAMEP´s operating procedures would help to re- duce errors that could affect the frequency of payments (for instance, repeated payouts to farmers who grow the same crop within the same geographical region owing to human misjudgment or—in extreme cases—moral hazard). Currently, the Government of Bolivia absorbs all costs involved in assisting small-scale farmers who experience extreme agricultural losses. Unlike Mexico’s CADENA program, in which the government is insured against any losses resulting from this kind of assistance, SAMEP´s operational costs and finan- cial losses are not transferred to a third party. For that reason, INSA has limited budgetary scope to roll out operations to new geographical regions and benefit larger numbers of farmers. The current legal insurance framework contains provisions for subsidizing crop insurance premiums as an incentive for the insurance sector to do business and provide coverage to small-scale farmers. Information on the financial resources available for these subsidized premiums and the procedures to be followed by the insurance industry is not clear, however. For that reason, the insurance industry continues to offer crop insurance products only for a subset of clients and regions, for whom the operational costs of indemni- ty-based products are manageable. Development of an agricultural insurance market that enables commercial farmers to transfer risk to a third party is still at an embryonic stage. In 2017, direct written premiums for agricultural insurance amounted to Bs 157,000 (US$23,088) handled by three companies, La Boliviana Ciacruz de Seguros y Reaseguros (Bs 44,000), Seguros y Reaseguros Credinform International S.A. (Bs 46,000) and Alianza Compañía de Seguros y Reaseguros S.A. E.M.A. (Bs 67,000). Given the low penetration agricultural in- surance products, farmers have limited knowledge of their existence and how they work, and it is likely that only farmers who operations are exposed to serious risks are willing to pay for insurance cover- age—creating the problem known in the industry as adverse selection. 135 Figure 39. Agricultural underwrite premiums and loss ratio (2012-2017) 3,000 300% 240% Underwrite Premiums (Bs$’000) 2,500 250% 186% 2,000 2,351 200% Loss Ratio (%) 1,500 2,602 150% 1,000 83% 100% 40% 23% 500 18% 50% 293 274 438 157 0 0% 2012 2013 2014 2015 2016 2017 Underwrite Premiums (Bs$’000) Loss Ratio (%) Source: Insurance and Occupational Pension Authority (APS) Source: Authors Agricultural insurance is one of the most complex low-income populations, but some provisions of this types of business. Substantial resources are need- law have restricted the development of synergies ed to convene a multidisciplinary group of experts between the agricultural insurance and financial in- (agro-meteorologists, agronomists, legal experts, ag- dustries. The act mandates all financial institutions ricultural economists, and others) to assess whether to allocate a minimum percentage of their portfolio any type(s) of agricultural insurance instrument(s) to productive sectors (Art. 66), and it limits the fees might offer viable solutions for different stakehold- and interest rates that can be charged to clients ers. Given public budget restrictions and the incon- (Art. 59). These provisions have had the unwelcome sequential volume of premiums, the design of new spillover effect of undermining the efforts of medium insurance products depends largely on support from and small insurance companies to work with financial donors. In recent years, Swiss Cooperation—through institutions to insure agricultural loans, because the PROFIN Foundation—has made funds available to im- Financial Services Act essentially forces financial in- plement insurance products, but they never reached stitutions to focus on providing services to the clients sustainable commercial levels after the subsidy of who generate the lowest costs of banking. For that insurance premiums ended. reason, credit for productive sectors is concentrated among fewer clients, which raises the overall risk of Another role of agricultural insurance can be to min- credit default and the risk that a borrower´s credit- imize default risk for credits approved by financial worthiness will decline.58 institutions for agricultural investments. The Gov- ernment of Bolivia enacted the Financial Services The oversupply of financial services to selected cli- Act Nº 393 as part of its objectives to increase eco- ents, combined with the lack of an insurance culture, nomic growth of productive sectors, rural areas, and makes it less likely that the agricultural insurance 58 Although the default rate on the credit portfolio for productive sectors in 2017 was slightly lower than in 2016, it sur- passed 30 percent of the total default rate portfolio. 136 market will advance. Farmers do not perceive the tan- Continued strengthening of INIAF and the agricultural gible benefits of risk transfer products unless disaster innovation system (SNIAF) more broadly is critical for strikes, so they regard insurance as an additional cost effective agricultural risk management. INIAF, as the that could be eliminated. Financial institutions can be coordinating entity of SNIAF, can contribute to risk re- reluctant to promote agricultural insurance products duction through the development, validation, and pro- because they fear losing clients if they require them to motion of improved technologies (high-yielding crop purchase agricultural insurance as a prerequisite for varieties that can withstand climate stress, improved the approval of a loan, or if they require grant funding silage systems); the provision of advisory services; for the purchase of insurance products. the production and availability of certified seed, and enhanced collaboration among innovation agencies. The Government of Bolivia has also promoted the development of information systems to help deci- Today, government actions focus on generating in- sion makers monitor hazards and provide farmers formation (SAT) and implementing a compensation affected by weather or climate-related hazards with mechanism (PIRWA) for subsistence farmers. Such an early response. Since 2017, the Rural Contingen- initiatives are important and valuable and need sus- cy Unit of MDRyT has integrated state-of-the-art geo- tained financial support. Yet it is important for them to graphic information technology, meteorological data, form part of an integrated risk management agenda and forecasts with ancestral traditional knowledge that increases producers’ awareness of the potential into an Early Warning System for the Agricultural impacts of risks and helps them to adopt strategies Sector (SAT).59 This information platform monitors for adaptation, risk transfer, and coping (tailored to the evolution of weather hazards and provides in- the needs of subsistence and commercial farmers) formation that the authorities use to take action to to reduce those impacts and protect the long-term reduce or alleviate sector-wide impacts. SAT weather sustainability of their investments. forecasts are generated in close collaboration with The agricultural insurance market is not likely to the National Meteorology and Hydrology Service of expand rapidly under the current legal framework Bolivia (SENAMHI). SAT also allows end-users to com- for financial services. Nevertheless, other finan- pare SAT outputs with biotic weather forecasting indi- cial instruments and budgeting mechanisms could cators and informs farmers about seeding dates that be combined to address the government´s need to are the least exposed to agro-meteorological risks. fund compensation for farmers following a disaster (Figure 40). The government should develop a risk 7.1.3. Suggestions for reducing financing mechanism that is based on the probabil- production risks ity of occurrence and the severity of hazards in the The production risk analysis described in this chapter agricultural sector and that maintains adequate and for selected crops across all departments in Bolivia cost-effective sources of financing (ex-ante instru- shows that severe losses are likely to occur as a re- ments and ex-post measures) to smooth the cost of sult of climate variability and climate change. At the risks and minimize fluctuations in costs. The PIRWA same time, the vulnerability of the agricultural sector program is currently the vehicle established by INSA can be reduced by strengthening and/or reforming to compensate smallholders for losses caused by ad- some of the agriculture risk management initiatives verse weather, and the government has no other risk currently implemented by the government. retention or risk transfer tool to boost its financial 59 http://sat.agro.bo 137 Figure 40. Risk layering and implementation of suitable financial instruments High International donor severity assistance High risk layer Catastrophe bonds Risk Transfer Insurance and Reinsurance Medium risk layer Contingent lines of credit Risk Contingent Budget, Retention Low risk Low reserves, annual layer severity Budget allocations More Less frequent frequent Source: Adapted from Clarke and Mahul, 2011 response to losses in the agricultural sector. Figure gional and department levels. For this purpose, the 40 shows a hypothetical example of a risk financing SAT should integrate new sources of time-series structure based on the concept of risk layering. data with sufficient intervals and completeness. Decision-makers in Bolivia should consider the The SAT early warning system should meet the needs scope for integrating private weather station net- of stakeholders outside the MDRyT system. A great works into the national network and making use of effort has been made to identify the thresholds for remote sensing technology for early drought de- crop damage and yield loss, but the data and informa- tection, monitoring, and impact assessment. The tion currently available are insufficient to character- SAT should improve not only the spatial resolution ize weather hazards and assess loss estimates that of the information shown in the bulletins, but also have a specified probability of exceedance of risk at work on the generation of agro-meteorological the local level. For that reason, at present the SAT indexes that are simpler and user-friendly (such cannot be used for formulating and implementing as crop water requirement indexes and vegetative risk preparedness and mitigation plans. To harness stress indexes). SAT´s potential, the following inter-related actions are required: • Improve the communication and dissemination of information (monitoring information as well as • Increase the accuracy of hazard monitoring and warnings) on agrometeorological hazards. This forecasting to establish effective risk manage- information should reach the people who are ex- ment interventions. At present, MDRyT works in posed to hazards, and it should be adapted to their collaboration with SENAMHI to generate meteoro- level of understanding so that they can act on it. logical and hydrological bulletins at both the re- Currently MDRyT disseminates SAT-related data 138 Figure 41. Bolivian legal framework on agricultural risk management Act No. 144 Community Productive Act No. 786 National Revolution in Act No. 602 Economic and Social Constitution Agriculture Risk Management Development Plan 2009 2010 2011 2012 2014 2016 Act No. 031 Act No. 300 Act No. 777 Framework of Framework of Mother State Planning Autonomies and Earth and Integral System Decentralization Development to Live Well through the internet and field extension officers. define a methodology that local experts can use to Decision-makers should explore the dissemina- assess losses and damage at the subsector level, tion of early warnings through other channels— develop and strengthen the technical capacity of mobile phone, SMS messaging, television, radio— community leaders and members of community that are far more likely to reach stakeholders in a organizations to use it, draft contingency plans timely way and give them more lead time to take that trigger activities based on reliable informa- action. tion about potential disasters, and integrate those measures in development plans to accelerate the • Develop a clear process to alert decision-mak- response. The SAT should contain a risk module ers at all levels about impending hazards so that that allows users to estimate the magnitude and they can coordinate a response based on previ- associated annual probability of hazard (a hazard ously developed contingency plans. The govern- component), appraise the asset at risk (exposure ment has systems in place to identify disasters component), and identify the damage against the in the making, but it lacks a clear process for de- asset at risk, given the magnitude of the hazard cision-makers at more local levels to act on that (vulnerability component). information as quickly as possible. The operation of SAT and implementation of any other risk man- A risk module cannot be implemented for all hazards agement activity are enshrined in a legal frame- that are relevant for agriculture, but Bolivia can cer- work that is intended to reduce the probability that tainly draw upon the vast international experience hazards will become full-blown disasters. In prac- with drought monitoring systems. For instance, the tice, however, SAT contributes information to the Ministry of Agriculture and Livestock (MAG) of Para- National Risk Management System (SISRADE), and guay developed a Crop Water Balance Model (BHAg) to that system is used to coordinate rehabilitation, monitor crop yield reductions due to severe drought recovery, and reconstruction interventions across and estimate the impact on crop output. The BHAg institutions (in other words, it comes into play af- also helps to monitor the occurrence of extreme ter a disaster has occurred). The true challenge is rainfall events during the most critical stages of crop to implement a clear process for decision-mak- growth. Similarly, the Office of Agricultural Risk (ORA) ing at the subnational level for reducing farmers’ in Argentina estimates the probability that water defi- vulnerability to impending external shocks such cits and excess rainfall will occur in selected crops as agrometeorological risks. The MDRyT should and pasture. Such information is used by the Agri- 139 cultural Emergency Unit of the Ministry of Agriculture tropical forests and other native vegetation for agri- to identify affected areas that are likely to call for a cultural production has a role in elevated levels of CO2 state of emergency or declaration of disaster, and for and other greenhouse gases. which the federal government will have to allocate Bolivia has serious land-use challenges, including the resources to compensate affected farmers. Finally, unsustainable conversion of land to agricultural uses, under the Garantia Safra program, the Ministry of Ag- soil degradation, and conflict over land. Of Bolivia’s ricultural Development (MDA) of Brazil uses a crop 109 million hectares, 57 million hectares are forest,62 water balance model to estimate crop losses due to covering 50.6 percent of the territory. From 2001 to severe drought and determine whether participating 2011, about 2 million hectares were deforested in farmers should receive a payout. Bolivia,63 with an annual median of 181,864 hectares In alignment with risk management options to miti- per year. From 2012 to 2017 deforestation increased gate and transfer risks, an integrated framework to by 1.3 million hectares, with an annual median of risk management will imply also measures to cope 219,501 hectares per year. By 2017, total deforesta- and adapt to risks. For this, irrigation will continue to tion in Bolivia had reached 7 million hectares (includ- be critical, particularly in regions where water scarcity ing the cumulative deforestation to the year 2000, is a concern, as well as the adoption of a range of cli- which was estimated at 3.6 million hectares). This mate-smart practices for specific crops and regions. area corresponds to 12.3 percent of Bolivia’s total forested area and 6 percent of its national territory. 7.2. Reducing the environmental Rapid expansion of agriculture, and particularly live- footprint of agriculture stock production, has driven a large proportion of this While a major challenge for the agricultural sector is deforestation (Table 27). Santa Cruz Department ac- to reduce its vulnerability to climate change, a major counts for 64 percent of all tree cover lost nationally opportunity is to reduce its environmental footprint. between 2001 and 2017. The department has lost Agriculture worldwide contributes approximately 9.6 percent of its tree cover, followed by Cochabamba one-third of GHG emissions. In Bolivia, the sector is a (7.6 percent) and Beni (6.1 percent).64 major contributor. Many farming practices—such as Concerns about further deforestation have emerged burning fields and using gasoline-powered machin- in light of the recent policy around biodiesel use. The ery, cattle ranching, and the use of chemicals—add government has ambitious goals for expanding biofu- significantly to the buildup of GHGs in the atmo- el (ethanol) production to substitute for gasoline im- sphere.60 According to Friends of Nature Bolivia, 77 ports and reduce GHG emissions. In 2018, it passed percent of Bolivia’s emissions come from the conver- a mandate to begin producing ethanol from sugar- sion of forest into agricultural land.61 Destruction of 60 WWF 2018 Greenhouse gases, such as carbon dioxide, trap heat in the atmosphere and regulate our climate. These gases exist naturally but humans add more carbon dioxide by burning fossil fuels for energy (coal, oil, and natural gas) and by clearing forest. 61 Fundacion Tierra 2019, Conferencia sobre la Agenda Abandonada de la Madre Tierra 62 Wanderley Ferreira, 2018 Estado de la situación de la deforestación en Bolivia. Conference on Mother Earth. 63 Deforestación en el estado plurinacional de Bolivia Periodo 2016-2017. Ministry of Environment and Water (MMAyA, Spanish acronym), 2018. 64 Global Forest Watch, 2018 www.globalforestwatch.org/dashboards/country/BOL 140 Figure 42 . Deforestation in Bolivia 8000 7000 6000 5000 4000 3000 2000 1000 0 Accumulated 2001-2011 2012-2017 Total to 2000 deforestation Source: Ministry of Environment and Water, 2018 Table 27. Dominant drivers of tree cover loss by 2015 Driver a Area (kha) Area (%) Commodity driven 125,0 75,8 Shifting agriculture 37,8 27,9 Forestry 1,3 0,8 Wildfire 0,75 0,4 Source: Global Forest Watch a “Commodity-driven deforestation” is the long-term, permanent conversion of forest and shrubland to a non-fo- rest land use such as agriculture, mining, or energy infrastructure. “Shifting agriculture” is defined the conversion of small to medium-scale forest and shrubland for agriculture that is later abandoned and followed by subsequent forest regrowth. “Forestry” is large-scale forestry operations occurring within managed forest and tree plantations. “Wildfire” is large-scale forest loss resulting from the burning of forest vegetation with no visible human conversion or agricultural activity afterward. cane and sorghum, and in 2019 it approved the use environment and Bolivia’s key ecosystems if not prop- of genetically modified soybean varieties to produce erly guided and monitored. biodiesel, which is expected to bring an additional The impacts of biofuels on the environment and food 150,000 hectares under sugarcane production and security are still being monitored in developed coun- about 250,000 hectares under soybeans. The govern- tries. FAO has reported that the impact of biofuels on ment plans, expressed in PDES, is to expand agricul- GHG reduction is not necessarily positive, given that tural production on an additional 1.2 million hectares. in most cases new land is brought into production (for This expansion could have devastating impacts on the 141 instance, through deforestation) to produce the crops agro-chemicals as the main source of water pollu- that will meet the growing demand for biofuels.65 tion (39.1 percent), followed by industrial waste (20.9 percent) and mining (15.6 percent). Water is most Policies around land use management are critical to contaminated by agro-chemicals in Santa Cruz (63.4 ensure a more environmentally sustainable agricul- percent), Cochabamba (60.9 percent) and Tarija (56.7 tural expansion. Current regulatory framework for percent).66 Overuse of surface water and groundwater agricultural land use in Bolivia is governed by PLUS for irrigation without allowing for natural cycles of (a land management plan), which in the case of Santa replenishment can become an important challenge Cruz, was approved in 2003 (Ley el Decreto Supremo for Bolivia. The agricultural sector uses 70 percent 24124) as its instrument for land use management of all available water. Unless producers use more (ordenamiento territorial). In the case of Santa Cruz, resource-efficient technologies and practices, agri- PLUS is based on maps developed in 2009 at a scale culture consumes excessive water and energy and of 1:250,000. Under the PLUS framework, farm man- degrades water quality. There are important oppor- agement plans (Plan de Ordenamiento Predial, POP) tunities to promote and mainstream energy-efficient determine the feasibility of using land for a specific technologies, not only in relation to irrigation, but agricultural purpose and evaluate whether the soil across a range of agriculture activities. For example, is suitable for agriculture or whether forest is to be in Mexico, the Bank supported the adoption of energy cleared. A recent regulation (September 2015) allows efficient technologies in the livestock sector which up to 20 hectares of forest to be cleared for agricul- generated emission reductions estimated at 6 mil- ture or livestock production on small properties and lion tons of CO2. on community or collective land (in which case the law permits clearing up to 20 hectares of forest per The use of insecticides in Bolivia has almost quintu- family unit). The regulation does not require a POP or pled from 10,430 tons in 2004 to 50,000 tons in 2015 plan for integrated land and forest management to as more land has come under cultivation, particularly be presented to the authorities before clearing land. in the lowlands. Domestic consumption of fertilizer is expected to increase importantly in the years to Bolivia has significant opportunities to improve land come (in 2015, domestic consumption of urea was use planning and management, and it can benefit estimated at 66 MT/day, 20 MT/day more than the from regional lessons and experience with many levels achieved in 2012). Research has demonstrat- models. One example is the effort by the Agricul- ed that these chemicals significantly affect not just tural and Rural Planning Unit (UPRA) in Colombia to on the immediate natural resources such as soil and define the agricultural frontier based on land use in water, but also the biodiversity of plants and species 2010 and then optimize land use (including degraded that are beneficial to an ecosystem. Furthermore, soil land) for agriculture, based on maps moving from a degradation is a challenge in Bolivia, the MDRyT has 1:100,000 scale to a more precise 1:25,000. This de- stated that 41 percent of Bolivian territory have de- tailed planning processes should reduce deforesta- graded soils. The main causes are considered to be tion and land use conflicts in Colombia (Box 6). the loss of tree cover and unsustainable practices in Agriculture also contributes to water pollution and extensive agriculture. soil degradation. A 2014 study in Bolivia identifies 65 FAO 2018, Sustainable Bioenergy. 66 National Institute of Statistics (INRA) 2014 Bolivia: causas de contaminación del agua según departamento 142 Box 6. Efforts to improve land use management in Colombia: Definition of the agricultural frontier as a powerful tool to guide land use planning and reach multiple goals The Unit for Land Use Planning (Unidad de Planificación de Tierras Rurales, Adecuación de Tierras y Usos Agropecuarios, UPRA) was created by National Decree in 2011. Its main functions include the planning of rural land use and land improvement processes for agricultural purposes through the generation of technical, economic, social, and environmental criteria and guidelines, as a basis for the definition of pol- icies aimed at the sustainable development of natural resources and economic activities. UPRA oversees the technical studies undertaken across various departments of Colombia to assess the suitability of land and inefficiencies of current land use. UPRA also undertakes studies on the functioning of regional land markets. In 2018, UPRA, in collaboration with the Ministry of Agriculture and Rural Development and the Ministry of Environment and Sustainable Development, defined and identified the National Agricultural Frontier. The frontier has the dual purpose of ensuring that competitive agricultural entrepreneurship is aligned with sustainable development—in other words, to ensure that rural development is accompanied by the protection and conservation of key ecosystem services. The agricultural frontier in Colombia is defined based on rural land thresholds that separate areas where agricultural activities can be developed from areas where agriculture can be undertaken under certain conditions, areas that are protected, areas with special ecological importance, and areas where agricultural activities are excluded by law. According to this definition, 35 percent of the continental area of Colombia represents the land area where economic activities related to agriculture, including livestock, forestry, aquaculture, and fishing, take place. This is the area that in its entirety should be the focus for developing and managing public policy 143 for agriculture, including the efficient use of rural agricultural land and the regulation of rural property for productive and social purposes. From a land use planning perspective, delineation of the National Agricultural Frontier can guide the identification of areas suitable for agricultural production, reduces uncertainty for investors and entre- preneurs seeking to develop agricultural activities, and facilitates coordination of land management to promote agriculture. From a social perspective, the agricultural frontier can guide programs to provide access to land under the agrarian reform by identifying areas with a suitable agricultural profile. It also helps to stabilize the land market, guide land adjudication schemes, and support the productive occu- pation of rural land. Perhaps one of the greatest innovations of this tool is that it reconciles environmental and productive goals. The definition of the agricultural frontier serves as input for the design of differentiated strategies and joint work between the environmental and agricultural sectors, fishing and rural development, for the control and containment of the agricultural frontier. It will help to stabilize and reduce the loss of environmentally important ecosystems and improve their protection, which will promote the adequate supply of ecosystem services such as water and biodiversity. In addition, it will help to reduce deforesta- tion through strategies to contain the agricultural frontier. Overall, the National Agricultural Frontier in Colombia is changing the image of the agricultural sector, nationally and internationally, to the extent that it sends an exemplary message of rural agricultural development occurring alongside the conservation and protection of key ecosystems for the provision of environmental services. Source: UPRA 2018 As a definer of environmental mitigation, the govern- 70 percent of agroindustry and cattle ranching will ment has been playing an important role through its implement sustainable practices and will have also high-level political endorsement and advocacy for enhanced their resilience to climate change. The gov- climate change and green growth. Bolivia has cam- ernment has committed to achieve zero illegal defor- paigned for global sustainable practices to tackle estation by 2020. climate change, supported climate change mitiga- At the same time, it has also committed to triple the tion and adaptation agreements, and spoken open- area under surface irrigation to over 1 million hect- ly on the responsibility of countries to assume their ares by 2030, which implies that a major effort will be own climate debt. National plans propose actions for required to use water efficiently and ensure that this the agricultural sector to reduce GHG emissions and expansion does not compromise the sustainability of improve adaptation to climate change. The govern- water resources. The government has also embarked ment’s new environment-based model promotes the on a strategy to supply urea produced domestically development of environmentally sustainable produc- and support intensification via increased input use. tion systems and reduction of pollution in a climate Thus, ensuring that intensification is done sustain- change context. The PDES anticipates that by 2020, 144 able, via strategies that combine efficient input use principles of Responsible Investments in Agriculture with technologies and practices that enhance soil provide a good basis for large private investors to as- quality and soil productivity is critical. sess social and environmental risks and mitigation options, as well as for governments to screen agricul- To forestall this outcome and achieve more environ- tural investors and enhance and attract responsible mentally favorable results, Bolivia requires a deep investors.67 understanding of the related risks, opportunities, and tradeoffs; flexible policies and the resources to im- At least four major opportunities exist to prevent plement them; appropriate policy instruments to pro- agricultural expansion from repeating the mistakes mote good practices; and proper impact monitoring, of the past. First, Bolivia can learn much from the evaluation, and response. Yet in Bolivia, as in most well-documented negative effects of unsustainable countries, these measures are not well established, agricultural development in faster-growing na- and accelerated biofuel production simply increases tions of LAC as well as from emerging approaches the risk of land-use change and the associated envi- to manage those negative effects. Particularly rel- ronmental threats. evant experiences include the use of landscape ap- proaches to reduce deforestation in Brazil’s soybean There are important opportunities to strengthen the production and cattle ranching systems, as well as government’s role as the definer, enabler, funder, and approaches implemented in Peru’s cocoa and coffee regulator of environmental mitigation in agricultural value chains and sustainable approaches to cattle investments in Bolivia. Table 28 illustrates the range production in Colombia. Equipped with such infor- of instruments available to policy makers to enhance mation, Bolivia can make systemic changes to move Bolivia’s role in reducing the environmental footprint agricultural industries onto a more sustainable path. of the agricultural sector. Second, farmer organizations are already leveraging While the achievement of economic and conservation environmentally sustainable practices that serve as policy goals requires a major effort by the govern- proof of concept to bring those practices scale. Third, ment, a successful transition toward a more envi- Bolivian farmers’ ancestral knowledge of balancing ronmentally sustainable agriculture in Bolivia also nature and agriculture should be supported and more requires a strong commitment by the private sector. broadly disseminated. Fourth, environmental orga- Such a commitment would extend to undertaking to nizations and financial institutions have progressed apply agreed environmental and social principles, for substantially on sustainability, becoming vast sources example in commodity production sustainability certi- of knowledge and best practices from which Bolivia fications. It would also extend to more inclusive busi- can certainly benefit. ness models that incorporate social principles. The 67 Responsible Agricultural Investment (RAI): Knowledge into Action Notes series. https://www.worldbank.org/en/topic/ agriculture/publication/responsible-agricultural-investment. 145 Table 28. Government roles and instruments in environmental mitigation Role Instrument • High-level political endorsement and advocacy for Agriculture Green Growth. • Mechanisms for ministerial coordination across agriculture, environment, ministry of production, and related sectors. Definer • Institutional support for multistakeholder landscape dialogue, planning, and coordination. • Designation of decision-making authority across national, state, and local government levels. • Promotion of supply chain eco-standards, eco-certification, pre-competitive action. • Public procurement of sustainably sourced commodities. • Public investment in infrastructure to mitigate environmental risks. Facilitator • Research and technology innovation and information systems to support adaptation (e.g., land and soil maps, climate risks maps). • Packing environmental action within large, multisectoral, sustainable growth initiatives. • Producer subsidies to reduce costs of investments or transitioning to improved practices. • Organizational subsidies to reduce cost of collective action/investment. • Public payments to landowners/land managers for ecosystem services. Funder • Bonds and deposit refund systems (e.g., environmental performance, land reclamation, waste delivery, and environmental accident bonds). • Preferential access to bank finance for environmentally friendly producers/processors. • Improved practices adopted in state-owned enterprises. • Land use regulation • Establish norms for human and ecosystem health. • Direct regulation of private land use or management practices. • Direct regulation of processing industry. • Zoning and land use planning, including generating spatial information at granular scale. • Tenure and access rules for land and resources (including concession policy, appropria- tion of lands under production to protect nature). • Monitoring of environmental management and ecosystem health. Regulator • Financial al market regulation • Market and trade rules. • Framework for private standards. • Regulation through creation of cap and trade/floor and trade markets (e.g., carbon offset trading, nutrition trading). • Taxation of environmentally harmful practices or products. • Liability systems. • Regulatory relief for environmentally friendly companies. • Raising awareness of resource managers or users. • Public media campaigns to educate citizens/consumers/buyers/investors. Advocate • Public dissemination of monitoring data and evidence • Mobilizing and supporting allies and advocates Source: Steps Toward Green: Policy Responses to the Environmental Footprint of Commodity Agriculture in East and Southeast Asia 146 8. Conclusions Photo: Shutterstock In the coming decades, the contributions of agricul- require a set of innovations that need to be very local- ture to the Bolivian economy will remain significant. ized. Institutional reforms have focused on enhancing Further growth and modernization of the food and re- public sector capacities for R&D, but they have lacked tail sector will bring opportunities to strengthen ag- adequate funding, and they have not sought to align riculture’s backward and forward linkages. Opportu- policies over the medium and long term to crowd-in nities for agricultural growth will also come from the private investment, incentivize knowledge sharing, potential to orient food consumption toward healthier and enhance the capacities of service providers. The options such as vegetables, legumes, and fruits; con- key elements of a successful and effective innovation sumption of these foods in Bolivia remains well below system are also lacking in Bolivia: institutional auton- global and regional averages. Increased consumption omy, stable and diversified financing, incentive struc- will not only improve nutritional outcomes but help to tures that reward the performance of staff and scien- galvanize and diversify production. Directly and indi- tists, programs aligned with clients’ needs through rectly, the agricultural sector will contribute to growth public-private partnerships, and strong international and diversification, provide significant employment, R&D linkages. The gains achieved by the government and serve as a major driver of poverty reduction. to date with multilateral and bilateral support are at risk, owing to the absence of sustainable funding and Unlocking the power of agriculture and food systems strategies to retain key scientists and staff. to grow continuously, sustainably, inclusively, and competitively will require an approach that considers Along with strengthening public R&D institutions, the range of opportunities and challenges throughout Bolivia has many opportunities to strengthen and Bolivia’s macro-regions (Box 7) and then designs a crowd-in other elements of the innovation system, range of tailored policy instruments and investments such as agricultural universities, think tanks, and pri- to address them. The policy actions and correspond- vate sector capabilities. Taking advantage of these op- ing investments that this report identifies as critical portunities will require integrating non-government for ensuring a sustained trajectory of inclusive ag- actors and improving the knowledge, networking, ricultural growth in Bolivia are linked to: enhancing and coordination skills of different agents. Import- innovation capabilities; improving connectivity, logis- ant opportunities also exist to build “soft” innovation tics, and value chain coordination; bringing financial capabilities such as public-private dialog for new services close to farmers; managing climate risks; business models (value-chain finance, export mar- and strengthening public institutions. Each is dis- ket promotion, product differentiation, certification, cussed in turn next. and traceability) and “hard” innovation in the form of R&D. In Bolivia, the range of technological opportuni- i) Enhancing innovation capabilities ties extends from opportunities that are closely linked to production, post-harvest handling, and processing While studies consistently find that investment in to opportunities for increasing resource and input use agricultural R&D leads to higher productivity growth, efficiency to protect the natural capital for agriculture with social returns to public R&D averaging over 40 (land, water, and biodiversity). percent, investment in agricultural R&D is stagnant or falling in regions where agricultural growth is most Another key element of an effective innovation sys- needed.68 This is clearly the case in Bolivia. The coun- tem is strong mechanisms to disseminate knowledge try’s unique agro-diversity and climatic challenges and support extension and technology transfer. In Bo- 68 Draft May 2019—Harvesting Prosperity: Technology and Productivity Growth in Agriculture (World Bank, upcoming). 148 Box 7. Summary of current agricultural challenges/opportunities: Perspectives from Bolivia’s macro-regions Highlands and Amazon Regions (Macro-regions). A major challenge for consolidating productivity gains in the Highlands and Amazon Regions is the increasing fragmentation of agricultural land. In the two regions, average farm size does not exceed 3 hectares, but fragmentation is the most acute in the High- lands. In the latter region, agriculture is an important economic activity for rural households but rep- resents less than 10 percent of the value added in departments such as Oruro and Potosi. The Highlands also face challenges imposed by high poverty rates and a very fragile ecosystem. Unpredictable climate and weather patterns frustrate efforts by rural families to accumulate assets and improve their living conditions. The “Southern Highlands” experienced desertification following the unsustainable expansion of quinoa production, which left producers with highly degraded and eroded soils. In the Amazon Region, where a large share of Bolivia’s diversity is concentrated, most UPAs grow tradi- tional products such as cassava, bananas, and maize. The region has been seeking to expand commercial agriculture (mainly sugarcane and rice). It is already the major center for the collection of Brazil nuts for export. In Pando Department, agriculture is responsible for nearly 30 percent of GDP. A first challenge in the Highlands and Amazon Regions is to address productivity constraints. The analysis of technical efficiency shows that producers in these regions display lower average levels of meta-tech- nical efficiency and higher heterogeneity than producers in the Sub-Andean and Lowlands Regions. Approaches that support the less-efficient UPAs in the Highlands and Amazon Regions to shift their production practices and technologies toward those of the most efficient UPAs could lead to important im- provements in productivity. For some crops, such as potatoes in the Highlands, a further effort is needed to reduce technology gaps and expand the technical production frontier. Given the unique characteristics of agricultural production in the highlands, innovations will need to be highly localized. Data limitations preclude estimates of TFP by macro-region, which would provide valuable insights into the dramatically low level of aggregate labor productivity in Bolivian agriculture. This knowledge gap is important, as this low level of labor productivity probably conceals important variations in rural livelihood strategies across Bolivia. For example, in the Highlands, farmers mainly live in nearby urban settings, and agricultural work is highly seasonal, with labor migration occurring at specific periods of the year to supplement agricultural incomes. A second challenge in consolidating gains for farmers and other actors in the Highlands and Amazon Regions is linked to value addition and market integration. In the Highlands, for example, efforts to de- velop the camelid value chains have shown that without a strong market orientation and private sector involvement driving production development and quality enhancement, it will take a long time for farmers to consolidate significant gains. In the case of quinoa, the opportunity lies in the convergence of public and private efforts to agree on a sectoral strategy and move from strategy to action. The development of the native potato value chain in Peru shows the power of institutional innovations to generate value for 149 farmers, which Bolivia could replicate. The Amazon Region has vast opportunities for the bioeconomy and the sustainable use of biodiversity. A third challenge relates to reducing climate vulnerability. The Highlands Region must develop a strategy to reduce climate vulnerability. Some of the options available to mitigate the effects of climate change (crop rotations, or agroforestry with native species) could also help to reverse the soil degradation result- ing from the unsustainable expansion of quinoa production. In the Amazon Region, an essential strategy to limit climate vulnerability is to ensure that agricultural growth does not compromise critical environ- mental services and that it contributes to conservation objectives. Sub-Andean Region (Macro-region). The crop portfolio in this region is highly diversified and agricultural land is also quite fragmented. This region has been a source of important innovations, however, par- ticularly in the emerging grape and wine industry. An important challenge in this region, although not as imperative as in the Amazon and Highlands Regions, is to improve productivity. The SFA and MTE analyses suggest that the Sub-Andean Region should focus on technological improvements that move the production frontier toward higher levels of efficiency and productivity. Another important challenge for this region is value chain diversification and market integration, particularly around a broad range of high-value crops for domestic and export markets. Meeting this challenge will require investments to increase productivity and reduce vulnerability, including investments in irrigation. The region needs to grow sustainably, ensuring the protection of watersheds, achieving efficiencies in water and energy use and sustainable use of inputs. Lowlands Region. In this region commercial and extensive cattle production dominate agriculture. A major challenge is to achieve higher productivity in key crops such as soybeans and other growing commer- cially, which remain below regional standards. The ability to raise yields through intensification is largely confined to existing technology, so it will be important to focus on technological improvements that move the production frontier toward higher levels of efficiency and productivity. There is widespread consensus in Bolivia on the adoption of genetically modified seed as a strategy for improving productivity and reduc- ing vulnerability to drought in soybeans and other commercial crops, but other key elements associated with improved practices, particularly proper soil management, are often overlooked. A second major challenge is to enhance key elements of the enabling environment. Those elements in- clude common enablers such as strategic infrastructure and investments to facilitate logistics, as well as sectoral enablers related to effective trade policies and incentives, not only to attract greater private sector investment but also to ensure responsible patterns of investment (by investors large and small) aligned with a greater level of environmental accountability. A third challenge is climate resilience, par- ticularly in highly vulnerable areas such as el Gran Chaco. Finally, the diversity of the macro-regions, their producers, and their production systems is a challenge and an opportunity in itself. Efforts to support greater productivity and competitiveness will need to reflect that variation. 150 livia, the dominant focus of public programs has been reduce post-harvest losses. This enabling infrastruc- to enhance farm capabilities through the provision of ture is particularly critical as the country moves to- improved “assets” such as machinery, equipment, ward greater crop diversification involving more per- and irrigation. To be effective, however, those assets ishable and quality-sensitive products. must be accompanied by the transfer of improved Expanding coordination vertically, along the full ex- technical knowledge and efforts to build producers’ tent of the value chain, and horizontally, through pro- human capital. Bolivia has successful experience in ducer associations and producer networks, remains building farmers’ capacity to manage grant funds a priority. Opportunities to reinforce coordination in and act collectively to procure goods and inputs; it value chains are linked to the adoption of strategies is critical to enhance the power of those experiences that support and/or enhance the role of farmers and through strong knowledge and information transfer. also other value chain actors, including public-private Emerging digital technologies offer great opportuni- coordinating platforms or innovation platforms that ties to reach farmers and improve the timeliness and convene a broad set of institutions. Efforts to support quality of the information they receive. collective action by producers would benefit from systematic knowledge sharing and initiatives that go ii) Improving connectivity, logistics, and beyond individual producer organizations to develop value chain coordination networks of producer organizations that can be effec- Connectivity is a critical factor driving agricultural tive at the territorial/local level. productivity and competitiveness in landlocked Bo- livia. The country has higher transaction costs than iii) Bringing financial services close to countries with maritime access, where costs of build- farmers ing, operating, and maintaining transport infrastruc- The productivity analysis indicates that access to fi- ture are lower. For Bolivia, strategic investments in nance/credit will help to ensure productivity in the ports, roads, and similar infrastructure are vital to Lowlands and Sub-Andean Regions and could grad- competitiveness. Those investments will have to be ually improve opportunities in the Highlands and Am- accompanied by reductions in other post-harvest lo- azon Regions. Government regulatory interventions gistics costs, particularly those linked to trade pro- to broaden access to finance for a larger number of cedures, export controls, and sanitary and phytosan- farmers are partially undermining credit growth dy- itary inspections and certifications. namics among medium and small insurance compa- At the grassroots level, secondary and tertiary rural nies, however. Mechanisms that facilitate the estab- roads are integral to competitiveness for producers lishment of credit histories, support the development but building and maintaining them tends to be very of flexible collateral arrangements, and account for expensive. There is a need to ensure that scarce pub- seasonality in repayment schedules, among others, lic resources used for this purpose are directed to can be effective ways of tailoring financial services to areas where the potential to stimulate agricultural farmers’ needs. The many programs providing direct growth is highest. Along with roads, other off-farm support to producers in Bolivia represent tremendous market infrastructure and logistics systems are opportunities to integrate efforts to build financial lit- needed to match demand and supply, ensure that eracy and pilot financial innovations. products meet standards for quality and safety, and 151 iv) Managing climatic risks and In Bolivia, agricultural activities have threatened bio- reducing the environmental footprint of diversity and other social and economic interests by agriculture increasing water pollution, competition for water, and soil degradation, and by contributing to climate Bolivia should be commended for efforts to devel- change through greenhouse gas emissions. As illus- op risk transfer mechanisms to help large numbers trated in this report, there are important opportuni- of farmers to manage agricultural risks (insurance ties to strengthen the government’s role as a defin- schemes). There are important opportunities to make er, enabler, funder, and regulator of environmental current risk management strategies more effective mitigation in agricultural investments. In this regard, and to expand risk management capabilities around particularly relevant steps for government include integrated risk management approaches. The anal- the alignment of growth and environmental policies ysis in this report finds that the agricultural insur- and the improvement of national, regional, and local ance market will not expand rapidly under the cur- capacity for sustainable land-use planning and man- rent legal framework for financial services, but other agement, such as participatory local processes and financial instruments and budgeting mechanisms the development of decision-support systems and could be combined to address the government´s incentives that facilitate the design of climate-smart need for funds to compensate producers for losses agricultural landscapes. triggered by different types of hazards. To this end, a risk-financing strategy in Bolivia should be struc- Approaching social and environmental challenges tured around an understanding of the probability of merely from the perspective of the government’s occurrence and the severity of hazards, and it should roles is necessary but insufficient, however. The combine ex-ante instruments and ex-post measures collective commitment of the private sector is also to smooth the costs of risks and minimize cost fluc- required to improve the social and environment foot- tuations. The Early Warning System for the Agricul- print of the agricultural sector. tural Sector (SAT) could be enhanced to improve the accuracy of hazard monitoring and forecasting, the iv) Strengthening public institutions dissemination and communication of monitoring in- During the last two decades, Bolivia has established formation and warnings on agrometeorological haz- new institutions (INIAF, INSA, and EMAPA, among ards, and clear decision processes for feeding alerts others) and strengthened existing ones (SENASAG) into decision-making. to improve the delivery of critical services to the agri- Along with risk-transfer mechanisms, there are op- cultural and food sector. In light of current and future portunities to continue strengthening the capacity of challenges and opportunities to realize the potential INIAF and SNIAF to develop, validate, and promote of agriculture and food systems in Bolivia, these ef- technologies and practices that reduce agricultural forts must continue and even intensify as the agricul- risk. Investments in irrigation and water management tural and food sector and risks evolve. will be critical for reducing vulnerability and main- Aside from strengthening the capacity to provide key taining productivity gains. Improved water-use effi- services to the sector, it is essential to strengthen ciency (and energy efficiency) and watershed conser- the capacity for evidence-based policy making and vation and protection are requirements for any such investment planning; monitoring, policy analysis, and investments to be sustainable in the long term. program evaluation; recognizing food emergencies and agriculture and food security risks; and using all 152 of these capacities for effective policy action and pro- This clear definition of roles and enhanced capacities gram implementation. Policy makers require effective would improve the coordination and effectiveness of assessments of the benefits and tradeoffs of more policy making and investments. Systematic sharing of specific policies (related to export and input markets, lessons and experiences from investments and pro- price interventions, and the roles of state-owned en- grams would help key ministries and public agencies terprises in agriculture, for instance) as well as a to harmonize their efforts and also make investments broad understanding of the evidence on policies and more effective. strategies that are likely to be effective in achieving Finally, a key condition for achieving sustainable and agricultural growth, competitiveness, and food secu- inclusive agricultural growth is to make collaboration rity objectives. between the public and private sectors the basis for Better data collection is a key requirement for better continued progress. Agriculture is fundamentally a evidence to support better policy and planning. For private activity. The largest investments for support- example, enhanced data collection is a requirement ing growth of the agricultural sector into the future to monitor gains in agricultural productivity at a more will come from the private sector (including large, informative scale (from the macro-region to the de- medium, and small-scale producers and processors), partmental and lower levels) and develop better es- although public spending will also be essential. The timates of TFP. coordination of public and private efforts opens im- portant avenues for promoting innovation and piloting In the short term, it is critical to continue strength- comprehensive approaches that enable the conver- ening MDRyT’s capacity to perform its role critical gence of sustainable and inclusive growth objectives. role in defining, monitoring, and evaluating agricul- This collaboration should be institutionalized in such tural policy, while decentralized agencies strengthen a way that it becomes the foundation for advancing their capacity for policy action and implementation. the creation of prosperity and wealth in rural Bolivia. 153 References Photo: Programa de Alianzas Rurales (PAR), EMPODERAR Bravo-Ortega, C. and Lederman, D. (2005). “Agri- Ferreira, W. 2018. Estado de la Situación de la Defor- culture and National Welfare around the World: estación en Bolivia. Conference on Mother Earth. Causality and international heterogeneity since Foster, W. and A. Valdes. 2015. Measuring the Size 1960”. World Bank Policy Research Working Pa- of the Renewable Resource Sector: The Case of per No. 3499. Washington DC: The World Bank. Chile. World Food Policy, Vol 2, No 1. Burgoa, R; Herrera, A. 2016. El efecto de las restric- Fundación Tierra. 2019. Conferencia sobre la Agenda ciones a la exportación de carne bovina sobre la Abandonada de la Madre Tierra. Causas y Conse- producción y oferta doméstica. Published in the cuencias. La Paz, Noviembre 2018. Organized by Bolivian Central Bank website. Fundacion Tierra. CEPAL. 2014. Logistics Inefficiencies in Latin Ameri- Garcia, E; Rossio, M; Stucchi, R. (2017). The impact can Landlocked Countries. Issue No. 335. of export restrictions on production: A synthetic CEPAL. 2018. Logística y recursos naturales en los controls approach. December 2017. países sin litoral: el caso de la soya y la chía en Gauthier G., Carruthers, R., Placci F. 2016. Logística el Estado Plurinacional de Bolivia y Paraguay. de la Soya, Argentina, Uruguay, Paraguay. Serie de Comisión Económica para América Latina y el informes técnicos del Banco Mundial en Argentina, Caribe (CEPAL). Paraguay y Uruguay, Report No 4, 2016. The World Clarke, D., Mahul, O. 2011. Disaster Risk Financing Bank, 2016. and Contingent Credit: A Dynamic Analysis. Hayami, Y. and Ruttan V. (1970). “Agricultural Produc- Coelli TJ and DSP Rao (2005). “Total Factor Produc- tivity Differences Among Countries” American tivity Growth in Agriculture: A Malmquist Index Economic Review, 60, 895–911, 1970. Analysis of 93 Countries, 1980-2000.” Agricul- Hayami, Y. and Ruttan V. (1985). “Agricultural Develop- tural Economics 32: 115-134. ment: An International Perspective”. Johns Hop- Dias Avila, A. & Evenson, R. (2010). “Total Factor Pro- kins University Press, Baltimore. ductivity Growth in Agriculture: The Role of Tech- IBCE. 2019. Transformando la Industria del Trans- nological Capital,” Handbook of Agricultural Eco- porte en Bolivia por medio de la Tecnología. In- nomics, Elsevier. stituto Boliviano de Comercio Exterior (IBCE), FAO, 2014. Policy responses to high food prices in Bolivia. Latin America and the Caribbean. Country Case IBCE. 2019. Bolivia: Modos de Transporte y vías de Studies. Salida de las Exportaciones. Boletín Electróni- FAO. 2016. The Structural Transformation of Latin co Semanal Número 303. Instituto Boliviano de American Economies: A sectoral Long-Term Re- Comercio Exterior (IBCE), Bolivia. view. Rural Transformation Conference - 19-20 IICA. 2015. Sistema de Innovación del Sector Agroal- September 2016, FAO Headquarters, Rome, Italy. imentario en Bolivia. Instituto Interamericano de FAO. 2018. Sustainable Bioenergy website. Food and Cooperación para la Agricultura, IICA. Agriculture Organization, FAO. http://www.fao. IMF. 2018. Shadow Economies Around the World: org/energy/bioenergy/en What Did We Learn Over the Last 20 Years? The 156 International Monetary Fund (IMF) Working Re- culture in East and Southeast Asia. Washington, port. WP/18/17, January 2018. DC: EcoAgriculture Partners and the World Bank. IMF. 2018. Bolivia Country Report No. 18/379. Inter- Schuttel, C.; Kleinwechter, U.; Ihle, R.; Grethe H. 2011. national Monetary Fund (IMF), December 2018. Domestic policy responses to the food price cri- sis: The case of Bolivia. Journal of Agriculture and IBCE. 2018. Nuevo Record de Cultivos Genéticamente Rural Development in the Tropics and Subtropics. Mejorados en el Mundo. Instituto Boliviano de Vol. 112 No. 2 (2011) 125–139. Comercio Exterior (IBCE), Bolivia. Suarez, D. 2018. Logística y Recursos Naturales en INE. 2018. Producción de Carne y otros Subproductos los Países sin Litoral: El Caso de la Soya y la Chía de las Principales Especies Ganaderas y Avícolas en el Estado Plurinacional de Bolivia y Paraguay. a Nivel Nacional 1994-2017. Instituto Nacional de Estadística, INE. Trindade, F. and Fulginiti, L (2015). “Is there a slow- down in agricultural productivity growth in South INE. 2013. National Agricultural Census, 2013. Insti- America?” Agricultural Economics 46 (2015) sup- tuto Nacional de Estadística, INE. plement 69–81. INRA. 2008. Breve Historia del Reparto de Tierras en Turner, K., Davidson-Hunt, I. & Desmarais, A. 2017. Bolivia, de la Titulación Colonial a la Reconduc- Agrobiodiversity, Rural Transformations and ción Comunitaria de la Reforma Agraria: Certezas Household Experiences of Globalized Change: A y Proyecciones. Instituto Nacional de Reforma Case Study from Southern Bolivia. Rural Land- Agraria. scapes: Society, Environment, History. Lange, Glenn-Marie; Wodon, Quentin; Carey, Kev- USDA. 2018. Bolivia Enters Ethanol Era. GAIN Report in. 2018. The Changing Wealth of Nations 2018: Number: Global Agricultural Information Network Building a Sustainable Future. Washington, DC: (GAIN), Foreign Agricultural Service, United States World Bank. Agricultural Department (USDA), March 2018. Nin-Prat, A.; Falconi, C.; Ludena, C.. and Martel, P. World Bank. 2006. Public Policy Options for the (2015). “Productivity and the Performance of Well-being of all. The World Bank. Washington, DC. Agriculture in Latin America and the Caribbean: From the Lost Decade to the Commodity Boom,” World Bank. 2007. World Development Report 2008: Inter-American Development Bank Working Pa- Agriculture for Development. The World Bank. per No. 608 (IDB-WP-608), Washington DC Washington, DC. Pfeiffer L. (2003). “Agricultural Productivity Growth in World Bank. 2011. Making the Grade: Smallholder the Andean Community.”  American Journal of Ag- Farmers, Emerging Standards, and Development ricultural Economics, Vol. 85, No. 5, Proceedings Assistance Programs in Africa. A Research Pro- Issue (Dec., 2003), pp. 1335-1341 gram Synthesis. The World Bank. Washington, DC. Scherr, S.; Kedar, M.; Jaffee, S.; Negra, C.; Havemann, World Bank. 2014. Transport and Logistics: Bolivia’s T.; Kijtikhum, J.; Endan, U.; Nair, S.; Rosenthal, N. Path towards Competitiveness. Background paper, 2015. Steps Toward Green: Policy Responses to unpublished. The World Bank. Washington, DC. the Environmental Footprint of Commodity Agri- 157 World Bank. 2015. Urbanization Trends in Bolivia: Op- Others: portunities and Challenges. Directions in urban development. The World Bank. Washington, DC. Global Competitiveness Index, 2018. http://re- ports.weforum.org/global-competitiveness-re- World Bank. 2017. Bolivia Country Profile, Enterprise port-2018/ Surveys. The World Bank. Washington, DC. Doing Business Indicators 2018. http://www.doing- World Bank (Forthcoming). Tendencias de Pobreza business.org/en/rankings en Bolivia: Progreso y Desafíos. The World Bank. Washington, DC. Enabling the Business of Agriculture. 2019. Bolivia. http://eba.worldbank.org/en/reports World Bank (Forthcoming). Upcoming. Harvesting Prosperity: Technology and Productivity Growth Global Forest Watch. 2019. https://www.globalforest- in Agriculture. The World Bank. Washington, DC. watch.org/dashboards/country/BOL Logistic Performance Index, 2018. Https://lpi.world- WHO. 2018. Plurinational State of Bolivia: Nutrition bank.org/international/global country profile. Global Nutrition Report: Shining a light to spur action on nutrition. World Health Global Climate Change Risk Index, 2018. https://ger- Organization, 2018. manwatch.org/de/14638 WTO. 2017. Trade Policy Review Plurinational State International Disaster Database (EM-DAT). https:// of Bolivia. World Trade Organization (WTO) Report www.emdat.be/ (WTR) WT/TPR/S/363. World Development Indicators FAOSTAT National Institute of Statistics (INE) web site/data- bases International Labor Organization (OIT) databases 158 Annex 1. Irrigation and input use in Bolivia a) Area irrigated by department. b) Use of certified seed, 2015 Bolivia, 2016. 54947,0 Supervised 44178,0 National Percentage seeds, production of area sown national and of certified with quality 39793,0 imported seed (tons) seed (tons) 113806,0 Rice 3,182 42 23.40% 36896,0 Maize 2,886 3,797 65.30% Potatoes 9,289 119 2.90% 32523,0 Quinoa 22 2 2.40% 65234,0 Soybean 70,370 12,768 98.10% Wheat 6,162 275 33.70% Chuquisaca Oruro Santa Cruz Cochabamba Potosí Tarija La Paz c) Fertilizer consumption, d) Mechanisation adoption. Bolivia, average 2014-2016 2013-2014 (estimates) Fertilizer consumption Machine Quantity Country (Kg per ha of arable land) Chainsaw 28,444 Argentina 38 Electric generator 12,017 Bolivia 9 Fimigator 62,214 Brazil 173 Harvester 13,283 Chile 356 Laboratory flask 5,548 Paraguay 104 Milking machines 2,680 Peru 127 Motocultor 2,889 Uruguay 148 Peeler 19,042 Sprayer 4,485 Tractor 57,449 Water Pump 64,628 Source: a) Vice Ministry for Water Resources and Irrigation; b) IICA 2015; c) WDR; d) MDRyT, 2015 159 Annex 2. Data requirements and methodology for TFP and technical efficiency in agriculture a. Data requirements for calculating TFP Production information (quantities and prices) should and SFA in the agricultural sector be measured at crop or livestock level, while input information should be grouped at input level. It is To analyze evolution of the TFP, it is necessary to have preferable for this information to come from recent a series of production and producer prices per year, annual series and to be available in a disaggregated department/Macro-Region and crop (or livestock). form at the department/Macro-Region level for mak- Also, to count with data on the quantity and price ing the analysis at this level. of inputs (intermediate inputs, capital or machinery and hired family labor) by year and department/Mac- The following lists show the details of the data re- ro-Region. quired for analyzing the evolution of TFP for both the agricultural and livestock sectors. Lista de datos necesarios para calcular la PTF Agricultural Activity Availability of in- Indicator Level Years Source formation Production per year by crop Departments Yes 1984-2016 INE/MDRyT (total) Average price per year by crop Departments No - - Sown area per year by crop Departments Yes 1984-2016 INE/MDRyT (total) Quantity of intermediate inputs used per year (total) (seeds, Departments No - - fertilizers, herbicides, fung., etc.) Prices of inputs used per year Departments Yes 2016-2018 MDRyT Capital stock (machinery) per Departments No - - year (total) Machinery rental price per year Departments No - - Quantity of hired labor per year Departments No - - (total) Quantity of family labor per year Departments No - - (total) 160 Livestock Activity Availability of in- Indicator Level Years Source formation Producción de carne por año Departamentos No - - por especie (total) Precio promedio de la carne por Departamentos No - - año por especie Producción de leche por año por Departamentos No - - especie (total) Precio promedio de la leche por Departamentos No - - año por especie Producción de lana por año por Departamentos No - - especie (total) Precio promedio de la lana por Departamentos No - - año por especie Cantidad utilizada de insumos ganaderos importantes por año Departamentos No - - (total) Precio de los principales insumos ganaderos por año Departamentos No - - (total) Capital (maquinarias) por año Departamentos No - - (total) Precio del alquiler de maquinarias por año Cantidad de mano de obra Departamentos No - - contratada por año (total) Cantidad de mano de obra Departamentos No - - familiar por año (total) Source: Author’s elaboration Data availability to perform the analysis was a limit- ing factor. There are no available data on average out- put prices per year and by crop at the departmental level; neither for total quantities of inputs (interme- diate inputs, machinery, hired labor and family labor) at the departmental level. Without data on producer prices of crops at the departmental level, it is not pos- sible to build an output and input indices. The situa- tion is similar in the livestock sector. 161 b. Data requirements for calculating TFP and SFA in the agricultural sector List of information required for analyzing Technical Efficiency Variable Level Type Availability Source Gross Production Value (crops) APU Output Yes ENA 2015 Crop area APU Inputs Yes ENA 2015 Expenditure on intermediate inputs APU Inputs Yes ENA 2015 Expenditure on hired labor APU Inputs Yes ENA 2015 Expenditure on machinery APU Inputs Yes ENA 2015 Number of family labor employed APU Inputs Yes ENA 2015 Membership in organization APU Det. of inefficiency Yes ENA 2015 Pop. Census Electrification coverage Municipality Det. of inefficiency Yes 2012 Accessibility index Municipality Det. of inefficiency No - Access to telephony Access to technical assistance APU Det. of inefficiency Yes ENA 2015 Access to agricultural information APU Det. of inefficiency Yes ENA 2015 Non-agricultural job opportunities Municipality Det. of inefficiency Yes CENAGRO 2013 Land fragmentation index Municipality Det. of inefficiency Yes CENAGRO 2013 Crop diversification index Municipality Det. of inefficiency Yes CENAGRO 2013 Irrigation coverage Municipality Det. of inefficiency Yes CENAGRO 2013 Land titling APU Det. of inefficiency Yes MDRyT Household size APU Det. of inefficiency Yes ENA 2015 Gender of producer APU Det. of inefficiency Yes ENA 2015 Age of producer APU Det. of inefficiency Yes ENA 2015 Education of producer APU Det. of inefficiency No Slope Municipality Environmental var. Yes MDRyT Soil quality Municipality Environmental var. Yes MDRyT Altitude Municipality Environmental var. Yes MDRyT Rainfall Municipality Environmental var. Yes MDRyT Temperature Municipality Environmental var. Yes MDRyT Source: Author’s elaboration 162 Stochastic frontier and meta-frontier analyses are indicators that contain sufficient information for the generally performed using survey data at the level municipalities of the sample of producers. There is of production units. The required data/information no information either at the APU level on the level of include: (1) quantities and prices of output and inputs schooling of the agricultural producer. (it is not always necessary to have prices, the quan- Given that information is available at the producer tities or value of output and the cost of inputs may level of the elements that are part of the production be sufficient); (2) exogenous determinants of (in)effi- function (gross production value and variables of ciency; and (3) environmental variables affecting the expenditure and use of inputs) and that there is an meta-frontier. Most of this information is included in important set of variables at the municipal level of the data set of the ENA 2015 survey69, such as infor- the potential determinants of efficiency, it is feasible mation on the GPV, the area under cultivation, expen- to carry out an analysis of technical efficiency and a diture on inputs, labor use, as well as other data on meta-frontier analysis for both Macro-Regions and potential determinants of technical efficiency in each the most important crops. Macro-Region such as variables of characteristics of the producer (household size, gender and age of the c. Detailed methodological approach for producer), access to information and technical assis- STA and MFE tance, and associativity. The stochastic frontier analysis (SFA) is used to esti- Additional information is available from the 2013 Agri- mate a production frontier representing the current cultural Census and the 2012 Population Census data- state of technology in the industry. The frontier also set. This specifically concerns data on access to public defines the maximum potential or optimum level of goods and services (irrigation and electrification), crop production of the industry. Firms or companies oper- diversification, land fragmentation, non-agricultural ate on the frontier if they are technically efficient or job opportunities and data on the context in which below the frontier if they are not technically efficient. each producer operates (climate, soil, altitude, etc.). The frontier is considered stochastic because the de- The information included from both censuses is at the viations from the frontier include a component that municipal level, in order to avoid potential endogeneity represents random shocks affecting the production with the ENA variables at the producer level. process and another component with a specific dis- It would be ideal to have some indicator on acces- tribution function representing technical inefficiency. sibility to markets; however, this variable does not The problem is how to estimate this (in) efficiency exist and the limited set of information available from relative to the stochastic frontier when only the “de- the Bolivian road network is insufficient to construct terministic” part can be estimated. 69 The National Agricultural Survey is a survey designed to collect information among agricultural producers. Information is available for 2008 and 2015 (the latest version) and this survey is expected to have some continuity in the future. The ENA has departmental inference and mainly contains information on aspects related to agricultural production. Data are available on the agricultural production and the destination thereof (consumption, domestic market, foreign market, processing, storage, etc.), sown and harvested area, quantity and price of certain inputs (seed, fertilizer, pesticide, insec- ticide, fungicide and herbicide), the hiring of labor and daily wages, use of tractors and machinery and rental costs. With regard to livestock, it collects information about the production (in terms of stock variation, whether by sale, birth, ex- change, consumption, etc.) of cattle/sheep/goats/pigs, llamas and alpacas, and poultry and farmed birds; also, in the pro- duction and prices of livestock products such as milk, eggs, fiber and hides, and wool. However, there is no information on expenditures (or amounts) on livestock inputs. The ENA also contains information on the characteristics of the producer (gender, age), family labor, technical assistance, credit, associativity, access to irrigation, among other variables. 163 A linear stochastic frontier model can be expressed groups with different technologies. The measures as follows: they propose may be broken down into specific tech- nical efficiency measures of groups (or regions) and Y ji = f j ( X ji ) e v ji - u ji , technology gap ratios between different groups. This j= 1,2,…,J ; i = 1,2,…, Nj approach is complemented by Battese et al. (2004) Where Yji denotes the production of unit “i” in group and O’Donnell et al. (2008), which use a two-step or region “j”; Xji is a vector of inputs of a production procedure to estimate the meta-frontier. In the first function. vji and uji are the two residual components, stage, standard stochastic frontier techniques are where vji is a stochastic component (captures statis- used to estimate the specific frontier of each group; tical noise) and i.i. d N (0,σ2) is assumed, and where in the second stage, the data envelopment analysis uji it is a non-negative variable and represents the (DEA) is used to estimate the meta-frontier. Since technical inefficiency of unit “i” in group or region “j”. non-parametric methods such as DEA lack statisti- Following Caudill et al. (1995), uji is assumed with a cal inference and do not allow the identification of half-normal distribution N+ (0,σ ui 2 ) with a variance σ ui 2 sources of variation among groups, we prefer a new = exp (Zi , δij) that is a function of the exogenous deter- approach proposed by Huang et al. (2014) that uses minants of technical efficiency captured by vector Zji stochastic frontier techniques in the second stage and an unknown parameter vector δj to be estimated. to estimate a meta-frontier using a pool of regional frontier estimations of the first stage. The production Using this model, it is possible to predict u j i and cal- function of the meta-frontier is defined by f M ( X j i ) culate the technical efficiency of unit “i” in group or and includes the previously estimated group frontiers region “j” as TE ij = exp (- U ji) (Jondrow et al., 1982). f j ( X ji ) . It can be expressed in the following ratio: One disadvantage of standard stochastic frontier f j ( X ji ; β j ) = f M ( X ji ) e v ji - u ji M M models is that it is not possible to compare the tech- nical efficiency of producers using different tech- The technology gap ratio, representing the distance nologies, since they do not operate under the same from the production frontier to the meta-frontier, is production frontier. This complication could occur defined as follows: in the present study, because agriculture in differ- TGR ij = e - u ji M ent regions of Bolivia has different technologies and production systems in response to widely divergent u ji M represents the technology gap and is assumed to agroclimatic conditions. Spatial differences in tem- have a half-normal distribution N + (0, σ u 2 M ) with vari- ji perature, precipitation, soil characteristics and other ance σ ui = exp (Zji ∂ j), which is a function of factors of 2 M production factors determine the input-output com- the technological environment ZjiM. binations that are possible for each region and there- fore result in different production frontiers (O’Don- Efficiency measures of the producers with respect nell et al. 2008). Thus, if the technical efficiency is to the meta-frontier are obtained by multiplying the estimated independently for each region, the results technology gap ratio (hereinafter called technological cannot be compared directly due to the heterogeneity gap) by the technical efficiency measures of produc- of production technology across regions. ers. The resulting estimations of meta-technology efficiency (MTE) can be compared among producers Recognizing this problem, Battese and Rao (2002) and regions or groups. propose a meta-frontier approach to calculate effi- ciency measures that allow for comparisons between MTE ij = TGR j i x TE j i 164 Annex 3. Main programs/projects providing direct support to producers/communities70 Project/Program Period/Funding Project Objective Rural Alliances project First Phase: To improve accessibility to markets for small rural producers in the (PAR) I & II 2006-2014: US$80M selected Areas by: (a) promoting productive alliances between differ- ent small rural producer organizations and purchasers; (b) empow- ering rural producers through the establishment and strengthening Second phase: of self-managed grass-root organizations; (c) increasing access to 2013-2021: US$145M productive assets, technology and financial services; (d) promoting more effective, responsive and accountable service organizations at the local level; and (e) enhancing environmental sustainability of productive practices Community 2011-2019: To improve access to sustainable basic infrastructure70 and ser- Investment in US$100M vices for the most disadvantaged rural communities selected in Rural Areas Project some of the poorest municipalities of Bolivia. This would be ac- Additional Financing complished through a participatory process in which empowered (PICAR) rural communities identify, prepare, implement, and maintain de- velopment and productive infrastructure sub-projects that respond to their priorities, based on the Community Driven Development approach. Economic Territorial 2011-2017: To promote local productive development and adaptation to climate Development with US$25M change of small agricultural and forestry producers of indigenous Inclusion (DETI) communities and peoples, through comprehensive support for ag- ricultural, agroforestry and other non-agricultural productive initia- tives and the strengthening of institutional capacities Direct Support for the 2016-2021: To increase smallholders’ agricultural income and food security Creation of Rural US$62M through productivity improvements triggered by technological adop- Agri-food Initiatives tion (CRIAR) I & II Economic Inclusion 2013-2019: To reduce poverty by improving the living conditions of rural com- for rural families and US$59M munities through: i) strengthening capacities for sustainable man- communities agement of natural resources, ii) development of agricultural and (ACCESOS) non-agricultural economic enterprises and iii) access to financial and non-financial services, for their own development. This objective was complemented in December 2013 as an additional ASAP (Adaptation Program for Small Scale Agriculture) grant resources were approved to promote adaptation to climate change in 16 municipalities targeted by the program PRO Camelids 2016-2022: To improve living conditions of families and productive economic US$39M organizations linked to the South American camelids complex while ensuring sustainable management of natural resources. Plan VIDA 2011-2016: To improve the livelihoods of rural households residing in vulnerable US$15M municipalities in the departments of Potosí and Cochabamba. 70 Infrastructure is principally understood as productive, non-commercial investments at community and family level. 165 Project/Program Period/Funding Project Objective Development NA To improve the productive development of indigenous peoples origi- Indigenous Fund (FDI) nal, peasants and intercultural and Afro-Bolivian communities (2017 – 2020) by strengthening local production that contribute to Food Security with Sovereignty. National Fund for NA Generate sustainable and participatory Integral Development, fo- Integral Development stering community and institutional self-management capacities, (FONADIN) including private investment to eliminate the causes of poverty, exclusion and environmental deterioration. Its specific objectives are: • Diversify the productive and economic base by generating jobs. • Contribute to the improvement of the living conditions of the population of the areas covered by the National Strategy for Integral Development with Coca. • Promote the development of community and institutional parti- cipation and management capacities. 166