UNITED NATIONS CONFERENCE ON TRADE AND DEVELOPMENT COMMODITIES & DEVELOPMENT REPORT 2021 Escaping from the Commodity Dependence Trap through Technology and Innovation Geneva, 2021 © 2021, United Nations All rights reserved worldwide Requests to reproduce excerpts or to photocopy should be addressed to the Copyright Clearance Center at copyright.com. All other queries on rights and licences, including subsidiary rights, should be addressed to: United Nations Publications 405 East 42nd Street New York, New York 10017 United States of America Email: publications@un.org Website: https://shop.un.org/ The designations employed and the presentation of material on any map in this work do not imply the expression of any opinion whatsoever on the part of the United Nations concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Mention of any firm or licensed process does not imply the endorsement of the United Nations. United Nations publication issued by the United Nations Conference on Trade and Development. UNCTAD/DITC/COM/2021/1 ISBN: 978-92-1-1130188 eISBN: 978-92-1-403046-1 ISSN: 2519-8580 eISSN: 2524-2709 Sales No.: E.21.II.D.14 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Acknowledgements The Commodities and Development Report 2021: Escaping from the Commodity Dependence Trap through Technology and Innovation was prepared by Janvier D. Nkurunziza (team leader), Stefan Csordás and Marco Fugazza, from the Commodities Branch of the Division on International Trade and Commodities of the United Nations Conference on Trade and Development (UNCTAD). Clovis Freire was also part of the team and prepared the contributions from the Division on Technology and Logistics of UNCTAD, namely chapters 4 and 6. He also provided data on the technology indicators used in chapter 2. Weijing Ye provided substantive inputs and research assistance during her internship at UNCTAD. The contribution of the Division on Technology and Logistics is highly appreciated. A peer review meeting was organized on 17 February 2021 in Geneva to discuss the draft report. From UNCTAD, Junior Davis (Division for Africa, Least Developed Countries and Special Programmes) and Tansuğ Ok (Division on International Trade and Commodities) provided detailed written comments. The other participants who provided written or oral comments are the following: Ludovico Alcorta (external expert); Rachid Amui, Taro Boel, Alexandra Laurent, Claudine Sigam and Aimable Uwizeye-Mapendano (Commodities Branch of UNCTAD); and Anida Yupari (Office of the Secretary-General of UNCTAD). Graham Mott (Office of the Director, Division on International Trade and Commodities) also provided comments. At UNCTAD, the Intergovernmental Support Service provided editing of the report; Danièle Boglio and Catherine Katongola-Lindelof provided administrative support; and Nadège Hadjémian prepared the overall design and the cover. Layout of the report was undertaken by Carlos Bragunde López and Juan Carlos Korol, of the United Nations Office at Geneva. iv NOTES Notes Use of the term “dollar” ($) refers to United States dollars. The term “billion” signifies 1 000 million. The term “tons” refers to metric tons. Use of a dash between years (e.g. 2000–2001) signifies the full period involved, including the initial and final years. An oblique stroke between two years (for example, 2000/01) signifies a fiscal or crop year. References to sub-Saharan Africa in the text or tables include South Africa, unless otherwise indicated. v COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Acronyms and abbreviations ASYCUDA COVID-19 FAO GDP ICT IT MERCOSUR UNCTAD Automated System for Customs Data coronavirus disease of 2019 Food and Agriculture Organization of the United Nations gross domestic product information and communications technologies information technology Southern Common Market United Nations Conference on Trade and Development vi CONTENT Content Acknowledgements..................................................................................................................... iv Notes.......................................................................................................................................... v Acronyms and abbreviations...................................................................................................... vi Overview.................................................................................................................................... xi Chapter 1. Chapter 2. Chapter 3. Chapter 4. Chapter 5. Background....................................................................................................................... 1 1. Background................................................................................................... 2 References............................................................................................................. 8 The Commodity Dependence Trap...................................................................................9 2.1 Introduction................................................................................................. 10 2.2 The commodity dependence trap: A tale of three country trajectories.......... 10 2.3 Measuring mobility between commodity dependence States....................... 14 A brief discussion of the methodology.........................................................14 Empirical results..........................................................................................16 2.4 Correlates of commodity dependence......................................................... 18 2.4.1 Discussion of the variables............................................................... 18 2.4.2 Empirical results............................................................................... 21 2.5 Conclusion.................................................................................................. 22 References........................................................................................................... 23 Commodity Dependence, Productivity and Structural Change.................................... 25 3.1 Introduction................................................................................................. 26 3.2 Labour productivity trends........................................................................... 27 3.3 Structural change patterns.......................................................................... 29 3.4 Sectoral productivity trends and drivers....................................................... 36 3.5 Conclusion.................................................................................................. 40 References...........................................................................................................42 Appendix A. Economies included in the data set used in section 3.2.................... 43 Appendix B. Economies included in the data set used in sections 3.3 and 3.4..... 45 Structural Transformation through Technological Change and Innovation................ 47 4.1 Introduction................................................................................................. 48 4.2 Stylized facts............................................................................................... 50 4.3 Technological landscape and gaps.............................................................. 54 4.4 Conclusion.................................................................................................. 66 References...........................................................................................................67 Appendix. Technological development index, 2019..............................................68 Enabling Technological Transformation........................................................................ 73 5.1 Introduction................................................................................................. 74 5.2 Enabling technological transformation.......................................................... 74 Diversification paths....................................................................................76 5.3 Enablers of technological transformation and diversification paths............... 77 5.3.1 Horizontal enablers ......................................................................... 77 5.3.2 Vertical enablers .............................................................................. 80 5.4 Implementing technological transformation.................................................. 81 5.4.1 Illustration: Hard commodity export dependent countries................. 82 vii COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Chapter 6. Chapter 7. Forward linkages..............................................................................82 Backward linkages...........................................................................82 Intersectoral horizontal diversification................................................83 5.4.2 Illustrations: Soft commodity export dependent countries................. 86 Backward linkages...........................................................................86 Forward linkages..............................................................................88 Horizontal intersectoral diversification...............................................89 5.5 Conclusion.................................................................................................. 90 References.......................................................................................................... 92 Opportunities from Technological Revolutions............................................................. 97 6.1 Introduction................................................................................................. 98 6.2 Technological revolutions............................................................................. 99 6.3 Potential impacts of digitalization and Industry 4.0 on commodity sectors and related global value chains...................................................... 104 6.3.1 Commodity value chains................................................................105 6.3.2 Commodity trade...........................................................................106 6.3.3 Commodity sectors and climate change.........................................106 6.4 Windows of opportunity in deploying digital technology and preparing for Industry 4.0.......................................................................................... 107 6.4.1 Leapfrogging in infrastructure......................................................... 107 6.4.2 Facilitating trade............................................................................. 108 6.4.3 Challenges..................................................................................... 110 6.5 Promoting structural transformation through economic diversification and technological upgrading...................................................................... 110 6.6 Conclusion................................................................................................ 113 References......................................................................................................... 114 Conclusion and Policy Recommendations................................................................... 117 Measures at the national level............................................................................ 118 Measures at the regional level............................................................................ 119 Measures at the international level...................................................................... 120 References......................................................................................................... 122 Figures Figure 2.1 (a) Zambia: Main merchandise exports in 1965, 1985, 2005 and 2018.................... 11 Figure 2.1 (b) Nigeria: Main merchandise exports in 1965, 1985, 2005 and 2018 .................... 12 Figure 2.1 (c) Costa Rica: Main merchandise exports in 1965, 1985, 2005 and 2018 ............. 12 Figure 2.2 Commodity prices: A sixty-year perspective........................................................ 15 Figure 2.3 Long-term distribution of countries in the three states ....................................... 18 Figure 2.4 Technology level in commodity dependent developing countries and non-commodity dependent developing countries................................................ 21 Figure 3.1 Median labour productivity.................................................................................. 27 Figure 3.2 Average annual growth rate of labour productivity, 1995–2018........................... 28 Figure 3.3 Labour productivity, 1995–2018.......................................................................... 29 Figure 3.4 Commodity dependent developing countries: Average sectoral shares............... 31 viii CONTENT Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.8 Figure 3.9 Figure 3.10 Figure 3.11 Figure 3.12 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5 Figure 4.6 Figure 4.7 Figure 4.8 Figure 4.9 Figure 4.10 Figure 4.11 Figure 4.12 Figure 4.13 Figure 4.14 Figure 4.15 Figure 4.16 Figure 5.1 Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Manufacturing and output linkages, 2019........................................................... 31 Average share of manufacturing.......................................................................... 32 Share of global manufacturing employment........................................................ 33 Global manufacturing value added...................................................................... 34 Change in average sectoral employment share, 1995–2017............................... 34 Median labour productivity levels, 2017.............................................................. 35 Average aggregate labour productivity and indicators of technological development, 2015–2017................................................................................... 38 Manufacturing sector, 1995–2017....................................................................... 40 Drivers of structural transformation..................................................................... 49 Diversification and output, 2019.......................................................................... 50 Diversification and exports, 2019........................................................................ 51 Complexity of product mix of exports, 2019........................................................ 52 Viet Nam: Increasing complexity of product mix of exports................................. 52 Product space.................................................................................................... 54 Technological development index, 2019.............................................................. 58 Commodity dependent developing countries: Technological development index by type of commodity dependence.......................................................... 59 Commodity dependent developing countries: Technological development index by level of income...................................................................................... 59 Technological development index, 2019.............................................................. 60 Commodity dependent developing countries: Technological development index (median by type of commodity).................................................................. 60 Technological development index: Countries with greatest gains......................... 61 Commodity dependent developing countries: Complexity of product mix of exports by sector, 2019 ..................................................................................... 62 Evolution of distribution of product complexity, agricultural products as main commodity exports.................................................................................... 64 Evolution of distribution of product complexity, fuel-related products as main commodity exports.................................................................................... 65 Evolution of distribution of product complexity, minerals, ores and metals as main commodity exports................................................................................... 66 Diversification paths in a nutshell......................................................................... 76 Technological revolutions: Two latest waves........................................................ 98 Technological revolutions: Uneven deployment................................................. 101 Frontier technology readiness index.................................................................. 103 Population that could be served by mini-grid and off-grid solar photovoltaic solutions, to bring electricity to all by 2030........................................................ 108 Promoting structural transformation through technological transformation........ 111 ix COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Tables Table 2.1 Table 2.2 Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 4.1 Table 4.2 Table 5.1 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Commodity dependence: Mobility across three states, 1995–2018.................... 17 Descriptive statistics of the variables included in the econometric model............. 20 Sectoral disaggregation of labour productivity..................................................... 30 Selected developed countries: Greatest share of manufacturing in total employment........................................................................................................ 33 Disaggregated labour productivity growth........................................................... 37 Main variables..................................................................................................... 39 Selected indicators of technological development............................................... 55 Commodity dependent developing countries: Technological development.......... 56 Horizontal versus vertical enablers...................................................................... 77 Technological revolutions: Overview.................................................................. 100 Technological revolutions: Changes in infrastructure.......................................... 102 Gaps in access to digital infrastructure.............................................................. 103 Estimated annual demand from renewable energy and storage compared with current production..................................................................................... 105 x OVERVIEW Overview A country is commodity dependent when it derives at least 60 per cent of its merchandise export revenues from the commodity sector. In 2018–2019, about two thirds (64 per cent) of developing countries were commodity dependent compared to 13 per cent for developed countries (see chapter 2). This implies that commodity dependence is particularly a developing country phenomenon. The analysis of commodity dependence has attracted interest from development economists in view of the challenges associated with this characteristic. Indeed, commodity dependence is associated with problems such as slow growth, an undiversified economic structure, low human development, income volatility, macroeconomic instability, Dutch disease, political instability, poor political and economic governance, illicit financial flows, low social development, as well as high exposure to shocks, including those resulting from climate change and pandemics such as the coronavirus disease of 2019 (COVID-19). Commodity dependent developing countries seem to be locked into this undesirable state. The concept of a commodity dependence trap is used in this report to characterize three different outcomes. The first is a situation where a country is commodity dependent in some reference period and remains dependent over a long time. Zambia illustrates this case. The second situation, illustrated by Nigeria, relates to a country where export diversification characterizes its initial conditions but, over time, the country becomes strongly dependent on one or a few commodities. The third case is that of a country that is initially commodity dependent but, over time, diversifies its export sector and moves out of commodity dependence. Costa Rica exemplifies this case. The experience of most developing countries resembles that of Nigeria and Zambia. Indeed, once a developing country is commodity dependent, it is extremely difficult for the country to extricate itself from this state, as chapter 2 will show. However, as the experience of Costa Rica illustrates, commodity dependence can be overcome. Many illustrative examples of successful cases are presented in chapter 5. The Commodities and Development Report 2021: Escaping from the Commodity Dependence Trap through Technology and Innovation starts by exploring the extent to which commodity dependent developing countries are trapped into commodity dependence and, as a result, how their economic structures are weakened by this situation. What the role of technology could be in helping commodity dependent developing countries to diversify their economies and escape from the commodity dependence trap is then analysed. Policies are proposed to show how countries could diversify their economies, and some opportunities are highlighted to illustrate some benefits that commodity dependent developing countries could derive from digitalization and embracing the current technological revolution. The report concludes with suggestions of key measures at the national, regional and international levels that could help make this transformation possible. The Commodity Dependence Trap At any given time, each country should be in one of the following three states: not commodity dependent, commodity dependent, or strongly commodity dependent. In the short run, it is normal that countries move between these three states, depending on factors such as changes in international commodity prices; important discoveries of strategic commodities, such as oil, gold, cobalt and some other minerals; the health of the global economy; development of alternatives to traditional commodities, such as green energy sources; and other factors. When mobility is analysed empirically, the finding is that countries tend to stay in one state for long periods. Most developed countries stay in a state of non-commodity dependence, whereas most developing countries are trapped in states of commodity dependence and strong commodity dependence. xi COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Empirical data based on mobility over the period from 1995 to 2018, and covering 206 countries and territories, shows that there is indeed some mobility between all states even though, by and large, countries seem to stay within one group. On average, half the countries and territories are in a non-commodity dependent state. The other half are in a strongly dependent state (32 per cent of the sample) or in a commodity dependent state (18 per cent of the sample). This information suggests that commodity dependence – and its strong version – only affects half of the countries and territories in the sample, as discussed in chapter 2. The evidence shows limited mobility out of the non-commodity dependent and the strong commodity dependent groups. During the sample period, 95 per cent of non-commodity dependent countries remained within this group. The proportion of strongly commodity dependent countries that did not move out of the category is 92 per cent. Put differently, the risks that a non-commodity dependent country becomes commodity dependent or strongly commodity dependent are 4 per cent and 1 per cent, respectively. Similarly, the likelihood that a strongly commodity dependent country becomes non-commodity dependent over the 24-year period is very small. There is, however, a 7 per cent chance that such a country will improve, moving from strong commodity dependence to just commodity dependence. Even though this might be considered an improvement, both commodity dependent and strongly commodity dependent countries face the same challenges, only with higher severity for the latter group. Very few countries seem to escape from commodity dependence, and these results seem to be stable over time. These results suggest that, in a business-as-usual scenario, it would take the average commodity dependent country 190 years to reduce by half the difference between its current share of commodities in total merchandise exports and that of the average non-commodity dependent country. This result illustrates the challenge facing commodity dependent developing countries. Unless these countries take strong action to change the status quo, they will remain commodity dependent for the coming centuries. Doing nothing or not doing enough should not be an option, as commodity dependence will not disappear on its own. Could higher innovation and technology help commodity dependent developing countries to change their trajectory towards more diversified economies? Econometric analysis shows a strong negative correlation between the state of commodity dependence and several indicators of technology. This suggests that the odds of commodity dependence are strongly associated with low levels of technology. In other words, countries with higher technological capabilities are less likely to be commodity dependent. If the results were to be interpreted as representing causality relationships, they would suggest that, by strengthening their technological capabilities, commodity dependent developing countries may reduce their exposure to vulnerabilities associated with commodity dependence. Indeed, improving the technological ecosystem of commodity dependent developing countries would create opportunities by increasing production outside the commodity sector. Acquiring technological capabilities and adopting institutions that foster innovation and technological development could reduce the dependence of commodity dependent developing countries on commodities and the negative implications of that dependence for economic development. There is also a positive and statistically significant relationship between commodity dependence and export shares of the three types of commodities: agriculture; minerals, ores and metals; and fuels. The correlation is strongest, though, for countries dependent on exports of minerals. The implication might be that the problems associated with commodity dependence are more entrenched in mineral exporting countries and, to a high degree, countries dependent on fuel exports. One reason could be that extractives (minerals, ores and metals; and fuels) in commodity dependent developing countries are generally enclave sectors dominated by foreign firms investing in high capital activities with little incentive to diversify activities through the creation of backward xii OVERVIEW and forward domestic linkages with non-commodity sectors. For example, as value addition to primary commodities mainly takes place outside the countries where the resources are extracted, commodity dependent developing countries do not benefit from value creation and its attendant advantages, including income generation, job creation, and tax revenue, along the value chain. Commodity dependence also seems to be more prevalent in the least developed countries relative to other countries. Development of the manufacturing sector seems to be a relevant way of addressing the commodity dependence issue in commodity dependent developing countries. Indeed, industrial production, whether it uses commodities as inputs or not, contributes to product and economic diversification. The experiences of Costa Rica and other countries discussed in chapter 5 show that an economy can indeed be transformed from an extractives-based or agriculture-based to a manufacturingbased production system. Success requires a long time, strong political will and a long-term, realistic development vision, coupled with an ambitious but reasonable implementation strategy. Commodity dependence, productivity and structural change Escaping from commodity dependence implies a process of economic structural change narrowly associated with an increase in productivity. As commodity dependent developing countries exhibit lower average labour productivity growth than other country groups, improvements in labour productivity would be a key source of economic growth and overall development process. In turn, diversification and technological development play crucial roles in labour productivity growth. Labour productivity can be driven by productivity growth within individual sectors and/or by productivity-enhancing structural change, namely a reallocation of production factors from sectors with lower productivity, to sectors with higher productivity. In this context, technological upgrading and innovation can be important drivers of within-sector labour productivity growth. Structural change is particularly relevant for labour productivity growth when there are large differences in productivity levels across sectors. These intersectoral productivity differences tend to be highest in low-income countries, where agriculture is typically the least productive sector, but employs large shares of the labour force. A key question is whether commodity dependence acts as an inhibitor of the within-sector component, the structural change component or both components of labour productivity growth. This is a question of high practical relevance for policymakers in commodity dependent developing countries. For instance, if commodity dependence acts as a drag on growth-enhancing structural change, policy interventions should focus on facilitating the flow of production factors from low-productivity to higher-productivity sectors. But if commodity dependence weighs down sectoral productivity growth, policies that induce productivity growth at the sectoral level need to be strengthened. And if commodity dependence is a drag for both growth components, a policy mix would be needed. Empirical analysis shows that commodity dependence is associated with low levels of labour productivity, low productivity growth, high volatility of productivity growth and a high frequency of negative productivity shocks. The average annual growth rate of labour productivity in commodity dependent developing countries was 1.5 per cent over the period 1995–2018, lower than in developed countries (1.7 per cent), non-commodity dependent developing countries (2.3 per cent) and transition economies (4.9 per cent). Therefore, combined with a low initial level of labour productivity, slow productivity growth has been widening the productivity gap between commodity dependent developing countries and other groups of countries. Labour productivity growth is also strongly associated with technology development across sectors. Hence, technological upgrading and innovation can play important roles in an increase in productivity and economic diversification. xiii COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Commodity dependence can be overcome through the strengthening of the manufacturing sector as a driver of economic growth and productive employment. This would directly and indirectly contribute to the achievement of several Sustainable Development Goals, including Goal 1 on eradicating poverty and Goal 8 on promoting inclusive and sustainable economic growth, employment and decent work for all. The performance of the manufacturing sector is a good indicator of economic development given the strong correlation between the level of manufacturing value added per capita and average income. Commodity dependent developing countries lag substantially behind non-commodity dependent developing countries in terms of shares of global manufacturing employment, with a gap that has widened from 27.6 percentage points in 1995 to 32.4 percentage points in 2017. This points to an important policy challenge for commodity dependent developing countries: how should these countries develop their manufacturing sector? The manufacturing sector continues to expand at the global level and can thus still be an engine of growth for developing countries, including commodity dependent developing countries. Global manufacturing value added has increased both in levels and per capita from 1990 to 2019, even if China, the country with the single largest manufacturing output, is excluded. Nevertheless, commodity dependent developing countries as a group have not industrialized since 1995. Instead, manufacturing shares in employment and value added in commodity dependent developing countries have peaked at significantly lower levels than non-commodity-dependent developing countries and developed countries. Commodity dependence is found to be primarily linked to lower labour productivity growth in the manufacturing sector. Structural change in commodity dependent developing countries has been characterized by a shift of employment shares away from the agriculture sector. As labour productivity in agriculture remains low in commodity dependent developing countries, any flow out of this sector results in productivity-enhancing structural change. However, employment shares moved primarily towards non-tradable sectors at the lower end of the productivity spectrum, where the potential for future expansion is limited to domestic demand. This raises questions about the long-term viability of this structural change path. The finding that the link between technological development, human capital and investment, on the one hand, and labour productivity growth, on the other, is not homogeneous across sectors is important for policy. It suggests that, while broad-based investments in technological upgrading, education and infrastructure are likely to yield aggregate productivity gains, their impact can be maximized by considering sector-specific challenges and opportunities as part of the policy approach to addressing commodity dependence. Such targeted measures could, for instance, consist of developing specific skills required for employment in emerging manufacturing and services sectors. The main message from the analysis of productivity and structural change is that commodity dependence is an impediment to the industrialization of commodity dependent developing countries. However, a positive message for commodity dependent developing countries is that there is ample scope for growth in labour productivity through both its components. The significant distance between productivity levels in virtually all sectors of commodity dependent developing countries and the global productivity frontier represents a significant opportunity for aggregate productivity growth through intrasectoral productivity gains. Similarly, the large productivity differences between sectors in commodity dependent developing countries highlight the potential of structural change to contribute to aggregate productivity growth. Structural transformation through technological change and innovation Technological change occurs through different channels: innovation, introduction of a new product (product innovation) or the modification of production methods to increase productivity and xiv OVERVIEW reduce costs (process innovation). All forms of innovation trigger shifts in income, consumption, employment and output, resulting in economic structural change. Technological change also affects economic structure through input–output relations between sectors (e.g. change in final product prices due to change in prices of intermediate products). The process of creating new products that replace old ones, and the long-term changes in the economy and society due to the emergence of new technological–economic paradigms, also impact the structure of economies. Although both process and product innovation can result in structural transformation, in commodity dependent developing countries, process innovation (and the resulting increase in productivity) tends to result in lower prices of agricultural produce or low employment in fuel and mineral sectors. On the other hand, product innovation leads to economic diversification and the emergence of new sectors, creating new opportunities for employment and further gains of productivity through subsequent learning by doing and process innovation. Technology diffusion is strongest in countries with high technologies (the centre) and slowest in the periphery (commodity dependent developing countries), due to differences in pre-existing capabilities, including infrastructure and technological know-how. The centre–periphery differences in technological diffusion also affect structural transformation, with commodity dependent developing countries characterized by slow transformation. Innovation should be understood as a combination of existing technologies in new configurations or economic activities. Therefore, innovation is path-dependent; it depends on the set of technologies that an economy has accumulated. In turn, technology is not limited to processes within a firm or a farm; it encompasses the whole chain needed to create and bring a product to the market. It includes the following: capital-embodied technologies, such as machines, vehicles, buildings and infrastructure; and labour-embodied technologies, such as business models, operational procedures and know-how. Even though changes in technology, demand and trade patterns are intertwined in complex ways, technological change could be considered the main determinant of structural economic change. It affects demand through changes in income, input–output relations and the substitution or complementarity of products (Schumpeterian creative destruction). Technology also impacts international trade through the effects on relative prices of products in global markets. Innovation requires the exchange of knowledge among different actors, including firms, research centres, universities, Governments and consumers, the main actors of national innovation systems. Firms (and their entrepreneurs) have the critical role of taking the risk to innovate (bringing a new good or service to the market). Innovators need finance to acquire the resources to innovate. Thus, the decision to innovate depends on many factors, not only the availability of and access to technology. Among commodity dependent developing countries, countries that are more reliant on agriculture exports usually have a lower technological level, followed by countries dependent on mining and then those dependent on fuels. As argued above, this may reflect the fact that mining and energy projects are more capital-intensive than agriculture, but they are usually in enclave sectors dominated by multinational enterprises. So, they might not truly reflect domestic technological capabilities. Generally, there does not seem to be any systematic advantage or disadvantage in any type of commodity dependence. Most commodity dependent developing countries have similarly low levels of technological development. Escaping from commodity dependence implies that commodity dependent developing countries embrace new technologies and innovation that can take them into more dynamic sectors. Product space maps (see chapter 4), illustrating the path dependence of innovation, show that the production of some products, including commodities, does not connect easily with other products; they are like dead ends – once a country is in a particular product space, it is difficult to use the capabilities therein to move to another product. For example, the location of Angola on the product space xv COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 map shows that its capabilities are highly concentrated around petroleum extraction. The country’s current technological and productive capabilities may not be easily transferable to production in the digital cluster, for example. On the other hand, machinery and electronics production requires technologies that can be the building blocks of production in many other sectors. Diversification to these products can facilitate further diversification in the future. Hence, for commodity dependent developing countries, diversification into more dynamic sectors might require large “jumps” in innovation to enter clusters that are not necessarily close to countries’ positions in the product space. Indeed, some of the technologies needed are not available in an economy and should be learned or transferred from abroad. This would call for government support. Higher technology is associated with not only higher productivity but also fewer countries that can produce high-technology products. Technology also allows the production of more complex products, so higher levels of technology allow countries to produce and export products of above global average complexity. In this sense, more diversification resulting from higher technology and innovation is also associated with lower competition in export markets. Most commodity dependent developing countries export products at the lower end of the product complexity index (see chapter 4), requiring the least technological capacities. As a result, most commodity dependent developing countries have diversification levels below the global average and face competition from over 82 countries that export similar products. This helps explain why commodity dependent developing countries are stuck in the commodity sector and might need strong support to move out of commodity dependence. Moreover, the process of economic diversification is handicapped by commodity price cycles. When commodity prices are high, commodity dependent developing countries have an incentive to produce more of the same, reducing the motivation to innovate and diversify the economy. On the other hand, when commodity prices are low, the challenge for diversification relates to declining resources, particularly the scarcity of hard currency to import capital goods. Governments’ fiscal constraints also prevent them from providing the required complementary infrastructure and quality education to increase the capacity for technological learning and innovation in the context of the economy. Hence, countercyclical fiscal policies are recommended, investing resources from the commodity sector when prices are high into non-commodity sectors, as did Indonesia, or capturing more revenue by adding value to the commodity as Oman has done (see the discussions in chapter 5). Being commodity dependent need not be fateful. Viet Nam is an example of a country that has successfully diversified its economy. Three decades ago, this country was at the same development level as the world’s least developed countries. Viet Nam has succeeded in increasing its technological and productive capacity to industrialize further and expand production from agriculture and low value added manufacturing such as garments to production in the digital cluster. Between 2005 and 2018, the country increased the share of its high-technology exports in total merchandise exports from 6 to 35 per cent, while the share of exports of primary resources fell from 52 to 22 per cent of total merchandise exports. The push for industrialization began in the 1990s, with an industrial and trade policy that merged import substitution measures and export subsidies to promote an export-driven growth strategy, supported by strong foreign direct investment. Other policies have also contributed to the country’s productive development, including the establishment of export processing and industrial zones, the development of urban infrastructure and education. As chapter 5 shows, there are several successful cases that may provide useful lessons for commodity dependent developing countries. Enabling technological transformation As discussed above, technological transformation in developing countries, including commodity dependent developing countries, goes hand in hand with economic transformation. What are the xvi OVERVIEW enablers of technological transformation, to shift from the status of being a commodity dependent developing country to having a more diversified economy? Structural change should be thought of as a meso-economic process that encompasses production composition effects, intrasectoral and intersectoral linkages, market structures, the functioning of factor markets and the underlying institutions. The set of policy interventions required to support technological transformation are determined by the mix of short- versus longer-term objectives in terms of productive capacity enhancement, as well as the diversification path chosen by a country. Market failures and government failures often act as constraints to this process. Diversification away from commodities production could follow different paths. Often recommended is a shift towards manufacturing, usually characterized by higher productivity. Such a shift may operate either by means of promotion of sectors and products unrelated to the set of commodities produced or through the exploitation of forward linkages within a process of vertical integration. Vertical integration can also operate by exploiting linkages to backward products or services. Diversification could also be through the promotion of production of other commodities. Another important source of diversification is quality upgrade of the set of commodities currently produced, as discussed in chapter 3. A diversification strategy needs to account for the type of commodity a commodity dependent developing country would be diversifying from: is it a point-source natural resource (minerals and energy commodities) or a soft commodity? Abundance in point-source natural resources is usually associated with higher rents relative to soft commodities. Such rents could provide part of the resources needed to fund a diversification strategy, as the case of Indonesia, discussed in chapters 2 and 5, illustrates. Technological transformation requires access to technology and a conducive framework for its transfer. This is particularly important for commodity dependent developing countries that acquire technologies from abroad. Accessibility refers to both cost and technical know-how. Commodity dependent developing countries generally have limited resources to access the expensive technologies needed to produce more complex products. Even when financial resources are available, as in the case of some countries endowed with strategic resources, commodity dependent developing countries may not have the skills needed to exploit those technologies. Indeed, the capacity of entrepreneurs and workers to introduce and adapt new and better production processes is the basis for technology adoption. Training is therefore an integral part of a successful technological transformation strategy. Furthermore, as firms are major actors in a successful technological transformation, they should be allowed to operate in an environment that helps the process. For example, removing excessive administrative procedures, support with filling the skills gap and provision of human and physical capital, as well as introducing relevant institutional reforms, are prerequisites for technological transformation. Encouraging foreign direct investment can help fill some of the gaps given that it is one of the channels for technology transfer and technical know-how. Effective technology transfer should lead to local innovation, at least in the medium term. This might require the creation or reinforcement of national innovation systems. The existence and effectiveness of an institutional framework able to coordinate the various actors engaged in innovation and learning – research and development centres, universities and technology schools, extension services and the innovating firms themselves – would be necessary. In addition, investments may have to be redirected over the long term towards new capabilities and to an ambitious educational strategy that supports these processes. Over time, success relies on the accumulated technological knowledge and production experience of the managers and production workers of the firms involved in a process. Infrastructure is a core enabler of technological transformation. For example, having reliable power is a basic condition for technological transformation. However, firms in many commodity dependent developing countries have only intermittent access to power, and many are forced to xvii COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 invest part of their capital in power generators, diverting resources that could have been invested in technological upgrading. Reliable access to the Internet has also become critical to unlocking the possibilities offered by digital technologies. Again, as chapters 2 and 4 show, commodity dependent developing countries lag on this metric. Where resources are limited, infrastructure development can be directed to the promotion or reinforcement of geographical clusters of firms expected to foster technological transformation. Trade integration can also significantly enable technological upgrading through an increase in productivity resulting from a better allocation of resources. Trade integration might lead to the adoption of more advanced technologies if these make firms more competitive, and if the firms benefit from wider regional markets to the extent that the benefits outweigh the costs of technology acquisition. This was the case with Argentinian firms that were able to take advantage of lower tariffs in Brazil after the establishment of the Southern Common Market (MERCOSUR). There are other enablers that depend on the type of commodity dependence and/or the diversification path followed. These are referred to as vertical enablers. Countries amply endowed with point-source natural resources should be able to mobilize public funding more easily thanks to the rents they derive from their extractive activities. In these countries, a specific challenge could be the management of natural resource windfalls. In countries depending on the agriculture sector, increasing productivity might be the most important challenge to overcome. The small size of farms in many commodity dependent developing countries dependent on agriculture may also make technology adoption very difficult. The promotion of effective technology adoption could thus be based on market and institutional features that would allow some mutualization of investment efforts. Farmer-to-farmer information diffusion can also be a cost-effective approach for improving smallholders’ practices and profits. In addition, technology adoption and eventually diversification may be facilitated by participation in global value chains. With respect to implementation of policies towards technological transformation, several examples are provided to show that, through technology and the other factors discussed above, commodity dependent developing countries can indeed diversify their economies and move away from commodity dependence. They can diversify by fostering forward linkages, as did several fuel-export dependent countries that have expanded their export baskets by moving into value added products that are energy intensive. Oman is a good example. The country expanded its production to refined fuels, such as gasoline or kerosene, and assorted petrochemicals, including alcohols, fertilizers and plastics; or products that are energy-intensive (for example, aluminium), even though most non-energy inputs are imported (for example, alumina and bauxite). Government intervention played a central role in the process. Other developed countries, such as Norway, pursued a model based on strengthening backward linkages. This led to the development of both service and industry activities with a high tradability potential. Norway set up a very innovative oil and gas industry with substantial linkages, creating a Norwegian model of petroleum exploration. At the same time, it accelerated a manufacturing industry supporting the sector. Intersectoral horizontal diversification is another approach where diversification is towards sectors that are not directly linked to the prevalent commodity, pushing an economy beyond its current comparative advantage. For example, Indonesia succeeded in reducing dependence on oil through countercyclical spending and investment into agriculture first, and later into processed and semi processed goods. In Botswana, the close relationship between the Government and the private sector in the diamond sector has also significantly contributed to the country’s success. A partnership between the Government of Botswana and a South African diamond conglomerate is an illustration of a successful private–public partnership. Through this relationship, Botswana has been able to integrate its diamond sector vertically, with the polishing and cutting of diamonds now taking place in the country. xviii OVERVIEW As an example of countries dependent on soft commodity exports, Thailand illustrates how new technologies can be used to produce higher quality and more competitive fresh organic vegetables and fruits. The so-called smart agriculture business is growing rapidly across the world. Also, the production, distribution and processing of agriculture output in producing countries are fundamental components of forward production linkages. In many commodity dependent developing countries, cotton provides a good example: edible oil is extracted from cotton seeds, while textiles and medical cotton are derived from cotton lint. There is also a list of cotton by-products, including briquettes and boards, that can be produced out of cotton stalk. All these transformations can easily take place in commodity dependent developing countries producing cotton. In all successful examples, the Government played a pivotal role by putting in place the instruments that allowed the private sector to thrive, in many cases in joint ventures with the Government. Opportunities from technological revolutions What is the role of new technologies in the structural transformation of commodity dependent developing countries? New technologies are essential for the technological upgrade of traditional production sectors in commodity dependent developing countries, as well as for diversification into other sectors. There are technologies that trigger new technological–economic paradigms – the cluster of technologies, products, industries, infrastructure and institutions that characterize a technological revolution. Arguably, the developed world now lives through the mature phase of the digital revolution’s deployment period, characterized by the Internet, mobile connectivity and the so-called Web 2.0 technologies (e.g. applications, social media, cloud computing, big data, etc.). This technological–economic paradigm has resulted in an increasing share of global value chains in global production, reduced communications and transaction costs, and the emergence of electronic commerce (e-commerce), among other changes. However, while the digital revolution has already reached a mature phase in developed countries, it is still in an installation phase in many commodity dependent developing countries. The existence of these technologies does not guarantee their applicability in the context of low-income commodity dependent developing countries. Major factors that limit the deployment of these frontier technologies include failure to build the required information and communications technology (ICT) infrastructure and skills, to implement the necessary institutional change, and a lack of investment due to the scarcity of financial resources, as discussed above. To assess commodity dependent developing countries readiness to take advantage of current revolutions, it is important to first understand where they stand in the technology landscape. Some of the elements of previous technological–economic paradigms are still being implemented in different economic activities in commodity dependent developing countries. For example, in many of these countries, mechanization (first technological revolution) has not reached most of the farms, large shares of the population lack access to electricity (third technological revolution), many production sectors have not been able to take advantage of economies of scale and become internationally competitive (fourth technological revolution), and the digital revolution (fifth technological revolution) has been limited to the use of mobile phones and digital platforms. In many commodity dependent developing countries, universal access to electricity has not yet been achieved, and the network of roads, highways and ports is still weak (which places them in the fourth technological revolution). Most commodity dependent developing countries still have a weak infrastructure of high-speed, fixed Internet connections, such as fibre optic and broadband, or high-speed mobile connections. Digital and frontier technologies also require technological literacy and skills, which may be lower in most developing countries. The development of skills to use digital technologies requires people to be exposed to these technologies and engaged actively in “learning by using”, which is challenging in low-income commodity dependent developing countries with a large share of the population that is illiterate. Hence, commodity dependent developing countries are less prepared to adopt and xix COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 adapt these technologies than developed, transition and non-commodity dependent developing countries. Commodity dependent developing countries that rely on agricultural products are less prepared than commodity dependent developing countries that depend on the energy and mining sectors. Nevertheless, the current (digitalization) and emerging (“Industry 4.0”) technological revolutions will change commodity sectors and related global value chains and will have a significant impact on commodity dependent developing countries. Even though commodity dependent developing countries are not ready to deploy Industry 4.0 technologies, there are ways of taking advantage of them. First, commodity dependent developing countries will benefit through increasing demand for their primary commodities that are fuelling digitalization and the adoption of a wide range of frontier technologies, from renewable energy to the Internet of things and big data. These commodities include lithium, cobalt, manganese, graphite, nickel, aluminium, copper, silver, bauxite, iron, lead and rare earth elements. Some of these strategic commodities, such as cobalt, lithium and copper, are found in large quantities in commodity dependent developing countries. Demand for some of those commodities could increase by 1 000 per cent by 2050. This growing demand should serve as an economic opportunity for those countries that are home to major reserves of these commodities. Another opportunity relates to possibilities offered by frontier technologies to extract new commodities that were not economically extractable before. For example, advances in biotechnology, such as biorefining techniques, have facilitated the sequential extraction of the major components of red algal biomass as commodity products, such as pigments, lipid, agar, minerals and energy-dense substrate (cellulose). The large-scale marine macroalgae production, mainly for human consumption, has given rise to their consideration as a non-lignocellulosic feedstock to produce renewable fuels. However, making biofuel from algal biomass economic requires the co-production of additional useful biochemical components unique to algae. This might form the basis for starting new ocean-based biocommodities, reducing the dependence on the terrestrial resources for food, feed, energy and chemicals. There are also new technologies for the extraction of lithium that may revolutionize the way lithium is harvested, minimizing water use and speeding the recovery process. This will significantly reduce the environmental footprint of lithium extraction as observed today. New technologies may also make some lithium deposits in countries such as the Plurinational State of Bolivia economically viable. It is also expected that frontier technologies, including drones, robots, blockchains and the Internet of things, will lead to profound transformations of global commodity chains, resulting in continuous reduction of transaction costs, increasing efficiency and profitability, and enhancing transparency, traceability and reliability. Frontier technologies can also optimize transactions’ effectiveness and transparency, minimize costs in processing data and help forecast commodity prices more accurately. Furthermore, frontier technologies can help improve the resilience of commodity sectors to climate change and strengthen their contribution to sustainable development. Smart water management, precise environmental monitoring and enforcement, and enhanced weather and disaster prediction and response are just some examples of the potential of frontier technologies to support the battle against climate change. Moreover, the adoption of cost-efficient solar photovoltaic cells may bolster energy security and support commodity sectors in remote areas that are not connected to national power grids, while reducing the traditional deleterious effect of energy production on climate change. There is also potential for blockchain to reduce the carbon footprint of commodity sectors. For example, a global low-carbon tea project in Kenya (see chapter 5) attempts to formulate a resilient and low-carbon tea value chain using blockchain technology. While increasing trust among consumers and retailers, tea promoted as a “carbon sink” could not only fetch higher prices but also give growers potential access to carbon markets, creating economic incentives for small-scale tea producers. xx OVERVIEW Frontier technologies offer economically viable alternatives to costly investment in infrastructure related to traditional technological paradigms. An example of the potential for leapfrogging frontier technologies is the development of decentralized renewable energy systems. Low-cost, highefficiency solar panels are available for household rooftop solar installations and village-level microand mini-grids. The cost of these panels has fallen by a factor of more than 100 in the last 40 years, and by 75 per cent over the past 10 years, dramatically improving their affordability and thus widening access to energy particularly in rural areas. Digitalization of trade and logisticsrelated documents, an area where firms in developed countries already have valuable experience, is another potential area of interest. Moreover, technology-enabled efficient payment systems, critical for international trade, are already benefiting from emerging technologies. Early adopters in commodity dependent developing countries can place themselves in a good position to reap the benefits of new technologies. At the institutional level, digitalization and frontier technologies also offer Governments an opportunity to build national capacity in the provision and regulation of digital services. The UNCTAD Automated System for Customs Data – ASYCUDA – is an illustrative example. It is important to note that, at the global level, funding is available for digital and frontier technology solutions in e-commerce and global value chains. At the current stage of Web 2.0 technologies, in which the technology is more mature, finance is looking for profitable applications related to digitalization and e-commerce. These are becoming less available in developed countries. Thus, innovators in developing countries could tap into these idle resources to finance digital innovation. Most particularly, commodity dependent developing countries could access these resources to invest in digital platforms that allow them to take advantage of digitalization of commodity-based operations to become more efficient and competitive, as discussed in chapter 6. Some structural factors, such as the coming into force of the African Continental Free Trade Area, may be an incentive to attract funding for these technologies in Africa given the large size of the regional market. Moreover, the key role of China in commodity value chains and its position of leadership in many of the new technologies associated with Industry 4.0. can help in spreading them in commodity dependent developing countries and other developing countries. To benefit from these opportunities, commodity dependent developing countries will need to overcome many challenges. Among them are fast demographic growth, which might offer stronger incentives to use more labour than technology; the large technological gap that characterizes commodity dependent developing countries; the lack of economic diversification, particularly into the manufacturing sector that can absorb more sophisticated technologies relative to the commodity sector; the dearth of public and private resources to fund research and innovation; and limited access to ICT infrastructure and digital skills. In this regard, to promote structural transformation through economic diversification and technological upgrading, commodity dependent developing countries could consider pursuing a strategy of innovation in three steps: promotion of economic diversification towards more complex products, starting with those close to their position in the product space; promotion of implementation of the digital revolution (current technological–economic paradigm) to lay the ground for deeper diversification; and preparation for the implementation of Industry 4.0 and trying to enter into possible value chains related to this paradigm. This strategy should be guided by national development plans, as well as countries’ development objectives and priorities. As discussed throughout the report, taking full advantage of the opportunities offered by technology and innovation will depend on several factors. Key among them will be the level of commitment of the leadership and Governments of commodity dependent developing countries to foster technology and innovation as a way of moving out of commodity dependence. Another important factor will be the role of the international community in accompanying commodity dependent xxi COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 developing countries in this endeavour. In this regard, it will be essential that international public and private partners of commodity dependent developing countries facilitate technology transfer and participate in commodity dependent developing country efforts towards building the physical, human and institutional capabilities required for the adoption and domestication of the relevant technologies. As chapter 2 emphasizes, if nothing is done, the technological and development gap between commodity dependent developing countries and other groups of countries will only continue to widen. xxii CHAPTER 1 Background COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 1. Background A country is commodity-dependent when it derives at least 60 per cent of its merchandise export revenues from the commodity sector.1 Trade data shows that roughly 53 per cent of all UNCTAD member States were commodity dependent in 2018–2019. According to the State of Commodity Dependence 2021 (UNCTAD, 2021), commodity dependence is most widespread among developing countries. In 2018–2019, 64 per cent of developing countries were commodity dependent compared to 53 per cent for transition economies and 13 per cent for developed countries. Therefore, even though commodity dependence is found in all three country groups, the issue is primarily a developing country and, to some extent, transition economies phenomenon. Moreover, the prevalence of commodity dependence does not seem to improve over time. If anything, commodity dependence increases over time. In 2008–2009, 60 per cent of developing countries were commodity dependent, 4 percentage points lower than for the period 2018–2019. Over the same period, commodity dependence increased also in transition economies from 47 per cent to 53 per cent, and in developed countries from 10.5 per cent to 13 per cent, even though the absolute number of commodity dependent countries is much lower than in developing countries. Commodity dependence is not simply about being dependent or not. The extent to which a country is commodity dependent matters. A country deriving more than 80 per cent of its merchandise export revenues from the commodity sector is more exposed to the challenges of commodity dependence than one that derives 60 per cent. In this regard, the analysis of commodity dependence in chapter 2 distinguishes between commodity dependence, where commodity exports represent between 60 per cent and 80 per cent of total merchandise exports, and strong commodity dependence, where the share of commodities in total merchandise exports is greater than 80 per cent. The analysis of commodity dependence is important for two major reasons. First, commodity dependent developing countries seem to be in a trap: once a country is commodity dependent, it is difficult to develop a productive sector out of commodities and export non-commodity products. If countries were able to move in and out of commodity dependence seamlessly, being commodity dependent would be a less serious issue. Hence, trying to understand how countries could get out of the commodity dependence trap is relevant for development policy. The second justification for the relevance of the analysis of commodity dependence is that this status is associated with many socioeconomic challenges. As documented elsewhere (for example, UNCTAD and the Food and Agriculture Organization of the United Nations (FAO), 2017)), relative to non-commodity dependent countries, commodity dependent developing countries suffer from unpredictable export revenues due to high commodity price volatility; declining terms of trade over the long-term; macroeconomic instability due to high trade and budget deficits (van der Ploeg and Poelhekke, 2009) and unstable exchange rates. Moreover, overvaluation of the exchange rate following commodity discoveries or commodity price booms has led to Dutch disease in many commodity dependent developing countries. Dutch disease renders non-commodity exports, particularly manufacturing exports, less competitive, making the affected country even more reliant on the export of a single commodity or a limited number of commodities.2 For example, chapter 3 shows that, between 1995 and 2017, the average share of manufacturing in the total value added of commodity dependent developing countries declined from 11.5 per cent to 10.4 per cent. 1 The 60 per cent threshold was formally derived through a quantile regression by Nkurunziza, Tsowou and Cazzaniga, 2017. 2 Dutch disease is a situation whereby increasing external flows associated with a major discovery and exploitation of a new commodity, such as oil, results in the overvaluation of the domestic currency, making a country’s traditional exports less competitive. For example, the manufacturing sector in many African commodity dependent developing countries was more vibrant in the 1960s and 1970s than it has been recently, before the discovery of oil and certain minerals from the 1960s through to the 1980s. Indeed, the share of manufacturing in GDP was highest in 1990, at 15.3 per cent, and declined steadily thereafter (UNCTAD and United Nations Industrial Development Organization, 2011). 2 Chapter 1 - Background Macroeconomic challenges associated with commodity dependence have led to difficulties for households and firms. For instance, due to macroeconomic instability in commodity dependent developing countries, firms operate in a difficult economic environment, resulting in low profitability. Commodity dependent developing countries that depend on agriculture commodities suffer from low producer prices, negatively affecting household incomes and aggregate demand in countries where most of the population lives in rural areas. Commodity dependent developing countries are also less integrated into commodity value chains. In fact, the role of most commodity dependent developing countries is limited to the production of a raw commodity, with all value adding activities taking place outside.3 This may explain why commodity dependent developing countries that produce strategic commodities, such as oil and cobalt, remain some of the poorest in the world, even though the commodities they produce generate billions of dollars for other value chain participants, such as importers, refiners, retailers and so on.4 Commodity dependence has also been associated with a high level of political instability. Research has shown that the contest over the control of rents generated by natural resources has led to civil wars in many commodity dependent developing countries (for example, Collier and Hoeffler, 1998). In an econometric study of the determinants of civil wars, Collier and Hoeffler established that the probability of civil war is at its highest, 0.27, where natural resources represent 26 per cent of the gross domestic product (GDP). Beyond this threshold, the risk starts to decline as countries get more and more resources to invest in the security apparatus. There is also a nascent body of literature associating commodity dependence with high illicit financial flows (for example, Lemaître, 2019, and UNCTAD, 2016). The literature has also shown that commodity dependence is associated with poor governance and low social development. For instance, a higher share of point-source natural resources – fuels and minerals – tends to have a negative effect on the quality of institutions (Bulte et al., 2005) and on governance (Isham et al., 2005). Moreover, commodity dependence is linked to both lower social development (Carmignani and Avom, 2010) and lower human development (Nkurunziza et al., 2017). Furthermore, a higher share of commodities in exports is linked to lower non-resource export diversification (Bahar and Santos, 2018). Also, commodity dependence is associated with lower aggregate labour productivity (Csordás, 2018). Commodity dependent developing countries are also vulnerable to shocks, including shocks related to climate change. Indeed, commodity dependence has been shown to amplify the negative effects of climate change, as documented by a recent UNCTAD report. Of the 40 countries most vulnerable to climate change, 37 (92.5 per cent) are commodity dependent developing countries (UNCTAD, 2019). Moreover, most recently, the coronavirus pandemic has highlighted the vulnerability of commodity dependent developing countries to an international health shock. A simulation analysis carried out by UNCTAD and the Commonwealth Secretariat assessed the impact of the pandemic on commodity exports from Commonwealth countries, the majority of which are commodity dependent developing countries. The results of the study (Ali, Fugazza and Vickers, 2020) show that, compared with business-as-usual, commodity exports to Australia, China, the United States of America, the United Kingdom of Great Britain and Northern Ireland and the European Union5 were expected to fall by between $72 billion and $98 billion in 2020, representing an export loss of 16.5 per cent to 23.8 per cent relative to the benchmark. 3 In countries reliant on the extractive sector, even production is controlled by multinational enterprises that own the capital and technologies used to extract the commodities 4 For many commodities, including soft commodities, such as coffee and cocoa, commodity producers get a very small share of the final product’s consumer price. For example, the share of the coffee consumer price accruing to producers is less than 5 per cent (UNCTAD, 2018). 5 The European Union from February 2020, with 27 member States, after the departure of the United Kingdom. 3 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 The discussion above illustrates the vast literature analysing the deleterious effects of commodity dependence on economic and human development, the channels through which these effects are mediated and the vulnerability of commodity dependent developing countries to different types of shocks. This report builds on this literature to offer an analysis of the extent to which commodity dependent developing countries are trapped in a state of dependence and what they should consider doing to break away from it. Most particularly, the discussion in this report focuses on the potential role technology and innovation could play to extricate commodity dependent developing countries from commodity dependence. As reflected in the subtitle, “Breaking out of the commodity dependence trap through technology and innovation”, this report contributes to the understanding of economic development challenges that face commodity dependent developing countries by attempting to answer four questions. First, taking as given that commodity dependence hampers development, as explained earlier, to what extent are commodity dependent developing countries trapped in the commodity dependence state? Second, if commodity dependent developing countries are trapped, could technology and innovation help them to break out of the commodity dependence trap? The term technology here has two distinct but complementary meanings. The first meaning relates to knowledge and processes that can be used to extract, process, trade and use commodities more efficiently. The second meaning refers to the fact that appropriate technologies help to allocate resources in a way that fosters economic transformation and diversification. While technology in this report should be analysed from a positive perspective – how it could help commodity dependent developing countries to reduce their dependence on commodities – the discussion should also highlight the challenges that commodity dependent developing countries may face if they fail to adopt some of the major technological advances in the third and fourth technological revolutions. Problems could include increased inequality, falling farther behind the productivity frontier, failure to adapt to the effects of climate change, worsening governance, and security issues. The third question is, if technology could help commodity dependent developing countries to become less reliant on the commodity sector, what would be the institutional requirements that would allow this process to take place and be successful? Factors that could explain the limited use of modern technologies in commodity dependent developing countries include poor infrastructure, dearth of investment owing to scarcity of financial resources, lack of skilled workers and unfavourable institutional environment. Low productivity and high production costs, low quality and standards of production, child labour and environmental damage could be among the consequences of the technological gap in commodity dependent developing countries. Fourth, what could be the role of digitalization and new technologies associated with the fourth technological revolution in upgrading the technological landscape in commodity dependent developing countries? For example, if economically viable, adopting technologies that allow commodity dependent developing countries to internalize the commodity value chain by adding more value to natural resources within their economies would create new economic activities and generate jobs and revenues, while contributing to structural change and economic diversification. The question of how technology will affect commodity trade is also relevant. For example, wider adoption of e-commerce can help producers to sell directly to consumers, reducing the number of intermediaries, which could generate more benefits for commodity producers. Based on empirical analysis, the research in this report provides some answers to the questions above. First, transition analysis confirms that commodity dependent developing countries are indeed trapped in a commodity dependence state. Unless there is strong action at the highest political level in commodity dependent developing countries to do things differently, the analysis shows that they will remain trapped for centuries. Second, econometric analysis suggests that technology and innovation could, indeed, help commodity dependent developing countries 4 Chapter 1 - Background to diversify their economies and become less dependent on the commodity sector. This could strengthen productivity growth, which has been stunted in commodity dependent developing countries. Technology and innovation could lead the way to economic and structural transformation in commodity dependent developing countries. Third, a careful product space analysis shows some non-commodity products that commodity dependent developing countries could indeed start to produce, competitively, helping to diversify their export basket. Fourth, the analysis shows that there are available technologies from the third and fourth technological revolutions, as well as digitalization, that could help commodity dependent developing countries to move out of the commodity sector trap. Fifth, producing more technologically advanced goods would imply access and adoption of new technologies, as well as embracing innovation. This would require international cooperation. To enable commodity dependent developing countries to escape from the commodity dependence trap, there needs to be stronger cooperation between commodity dependent developing countries and their trading partners, as well as development partners, in terms of technological acquisition and domestication. Hence, a conducive framework for technology accessibility and technology transfer to commodity dependent developing countries is needed at the international level. With respect to commodity dependent developing countries, they would need to initiate or strengthen their institutional capacity to absorb and domesticate new technologies. It is, therefore, clear that finding an answer to the ills of commodity dependence afflicting commodity dependent developing countries is not just the responsibility of this group of countries. Left alone, as has been the case in the recent past, they will not succeed. Commodity dependent developing countries will succeed only if the countries that benefit from the status quo, generally from the developed world, heed and support commodity dependent developing countries’ political decisions and actions to break out of the commodity dependence trap. The analysis is carried out in five substantive chapters, in addition to an overview and a background chapter and a concluding chapter. In chapter 2, entitled “The commodity dependence trap”, the existence of a commodity dependence trap is explored and the relationship between commodity dependence and technology identified. Using empirical data, transition analysis is applied to measure the likelihood that a commodity dependent developing country breaks away from commodity dependence. The analysis particularly shows why breaking from commodity dependence is difficult. It requires strong political will and long-term commitment, with adequate human, financial and institutional resources. Using several technology indicators, suggestive evidence is provided in the chapter that the adoption of some relevant technologies, as well as innovation, may help commodity dependent developing countries to build a productive sector beside the commodity sector, diversifying the economy and reducing these countries’ strong dependence on commodities. This analysis will pave the way for discussions in the subsequent chapters. Under the title “Commodity dependence, productivity and structural change”, in chapter 3, trends are explored in labour productivity and structural change in commodity dependent developing countries. As a potential long-run driver of rising real incomes in developing countries, labour productivity growth is an important development indicator that is firmly rooted in the Sustainable Development Goal framework. Improvements in labour productivity across sectors and productivity-enhancing structural change are key determinants of economic diversification and economic growth in commodity dependent developing countries. Against this background, in this chapter, the patterns and trends of labour productivity are analysed across groups of countries and sectors, showing the difference in productivity levels and growth of commodity dependent developing countries relative to other groups of countries. Empirical results show that commodity dependence is associated with low levels of labour productivity, slow productivity growth – particularly in the manufacturing sector – and a high frequency of negative productivity shocks. 5 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Hence, breaking away from the commodity dependence trap can help spur economy-wide labour productivity growth, for which there remains a large potential in commodity dependent developing countries. It is argued that, to realize this potential in commodity dependent developing countries, it is necessary to strengthen broad-based drivers of productivity growth, such as technological upgrading, but also to use targeted measures to remove sector-specific obstacles for productivity growth. For example, technological upgrading and innovations that spur productivity growth need to be enabled and promoted through the development of adequate infrastructure, including digital infrastructure. In chapter 4, the discussion focuses on how technological change and innovation could foster economic diversification and structural transformation in commodity dependent developing countries. Under the title “Structural transformation through technological change and innovation”, the recent economic complexity literature is exploited to present stylized facts related to technological change, diversification and structural transformation. The distribution of the export product complexity of commodity dependent developing countries shows the technological capabilities available in those economies and finds that commodity dependent developing countries are indeed very far from the technological frontier. Given that commodity dependent developing countries have made minimal gains in terms of technological development, large jumps in product complexity (see chapter 4) are needed to close the technological gap from which they are suffering. This will require strong government intervention to build absorptive capacity and put in place the required conditions to introduce higher-technology productive systems in the economy. Seen from a different perspective, the large technological gap between commodity dependent developing countries and other groups of countries is an indication of the substantial opportunity the former could take advantage of to increase their technological capabilities. Information on countries’ product space (see chapter 4) highlights some of the products commodity dependent developing countries could produce if they adopt technologies that are within their reach. The title of chapter 5 is “Enabling technological transformation”. Two main issues are the focus of the chapter. First, what are the enablers of technological transformation in the context of a commodity dependent economy? What would a successful implementation strategy consist of? To address the first issue, in the chapter, different diversification paths are first discussed. Horizontal diversification may be undertaken within the commodity sector and by expanding into the non-commodity sector. Vertical diversification can be achieved through quality upgrading and by developing backward and forward linkages. To achieve these objectives, there are horizontal enablers or general enablers that are independent of the nature of the diversification path pursued. These include infrastructure, entrepreneurship, skills development and the capacity to fully appropriate technological innovation. Trade integration could also play an important role, as it can increase productivity through improvements in resource allocation. In addition, there are enablers that are specific to the type of commodity a country is dependent on. For example, a major constraint in countries endowed with point-source natural resources may be poor management of natural resource rents. In these countries, adopting technologies that strengthen the management of natural resource rents maybe a key enabler. In contrast, countries dependent on agriculture may need to address issues of productivity, particularly where the small size of firms prevents them from adopting technologies that address this constraint. The second focus of the chapter is to provide specific examples illustrating how several commodity dependent developing countries have been able to successfully use technology to diversify their production and escape the commodity dependence trap. In chapter 6, titled “Commodity dependent developing countries and technological revolutions”, how technological revolutions have impacted or could impact commodity dependent developing countries is discussed. Indeed, technological revolutions offer the possibility of new combinations (innovations), sometimes leading to new technological–economic paradigms. The current 6 Chapter 1 - Background (digitalization) and emerging (Industry 4.0) technological revolutions are expected to change commodity sectors and related global value chains, with a potentially significant impact on commodity dependent developing countries. Even though commodity dependent developing countries may not be ready to deploy Industry 4.0 technologies, there are ways of taking advantage of them. Harnessing these technologies could help commodity dependent developing countries to diversify and structurally transform their economies. For example, every technological revolution has been associated with specific commodities, with the current Industry 4.0 fuelling industries, such as renewable energy, robots, drones and the like, that rely on commodities including cobalt, lithium, rare earths and so on. Commodity dependent developing countries have an opportunity to gainfully play a bigger role in the value chains of these strategic commodities. Digitalization has the potential to drastically reduce transaction costs associated with commodity trade, enabling commodity dependent developing countries to become more efficient and capture more value out of their commodities. Moreover, blockchain technologies can increase transparency in commodity value chains, helping to increase product information, accountability and risk management and fostering responsible consumption and production.6 By increasing traceability, blockchain can also link consumers and producers more directly, potentially increasing the intangible value of a product. Product traceability made possible by blockchain can also help to differentiate “high quality” from “low quality” products, allowing price differentiation that benefits producers who invest in the production of high-quality commodities. Should commodity dependent developing countries miss these opportunities, they will be left behind and remain trapped in commodity dependence and underdevelopment. Chapter 7 concludes by briefly offering a summary of lessons learned and suggesting some policy actions that commodity dependent developing countries could pursue to successfully use technology to alleviate the strong dependence on commodities. 6 One example of how blockchain can increase transparency in commodity markets may be found in Pisani M, 2021, Harnessing the potential of blockchain technology for sustainability and transparency in cotton value chains, presented at the twelfth session of the Multiyear Expert Meeting on Commodities and Development, Geneva, 9 February. Available at https://unctad.org/system/files/informationdocument/cimem2_2021_9_Feb_Maria%20Teresa%20Pisani.pdf. See also United Nations, Economic and Social Council, 2021, Harnessing blockchain for sustainable development: prospects and challenges, E/CN.16/2021/3, Geneva, 4 March. 7 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 References Ali S, Fugazza M and Vickers B (2020). Assessing the impact of the COVID-19 pandemic on commodities exports from Commonwealth countries. International Trade Working Paper 2020/14, Commonwealth Secretariat, London. Bahar D and Santos MA (2018). One more resource curse: Dutch disease and export concentration. Journal of Development Economics. 132:102–114. Bulte E, Damania R and Deacon R (2005). Resource intensity, institutions, and development. World Development. 33(7):1029–1044. Carmignani F and Avom D (2010). The social development effects of primary commodity export dependence. Ecological Economics. 70(2):317–330. Collier P and Hoeffler A (1998). On economic causes of civil war. Oxford Economic Papers. 50:563–573. Csordás S (2018). Commodity exports and labour productivity in the long run. Applied Economics Letters. 25(6):362–365. Isham J, Woolcock M, Pritchett L and Busby G (2005). The varieties of resource experience: Natural resource export structures and the political economy of economic growth. World Bank Economic Review. 19(2):141–174. Lemaître S (2019). Illicit financial flows within the extractive industries sector: a glance at how legal requirements can be manipulated and diverted. Crime, Law and Social Change. 71:107–128. Nkurunziza J, Tsowou K and Cazzaniga S (2017). Commodity dependence and human development. African Development Review. 29:1–15. UNCTAD (2016). Trade Misinvoicing in Primary Commodities in Developing Countries: The Cases of Chile, Côte d’Ivoire, Nigeria, South Africa and Zambia. (United Nations publication. New York and Geneva). UNCTAD (2018). Commodities at a Glance. Special Issue on Coffee in East Africa. Issue No. 10. (United Nations publication. Geneva). UNCTAD (2019). Commodities and Development Report 2019. Commodity Dependence, Climate Change and the Paris Agreement. (United Nations publication. Sales No. E.19.II.D.18. Geneva). UNCTAD (2021). The State of Commodity Dependence 2021. (United Nations publication. Sales No. E.21.II.D.17. Geneva). UNCTAD and FAO (2017). Commodities and Development Report 2017. Commodity Markets, Economic Growth and Development. (United Nations publication. Sales No. E.17.II.D.1 New York and Geneva). UNCTAD and United Nations Industrial Development Organization (2011). Economic Development in Africa Report 2011. Fostering Industrial Development in Africa in the New Global Environment. Special Issue. (United Nations publication. Sales No. E.11.II.D.14. New York and Geneva). van der Ploeg F and Poelhekke S (2009). Volatility and the natural resource curse. Oxford Economic Papers. 61(4):727–760. 8 Chapter 1 - Background CHAPTER 2 The Commodity Dependence Trap 9 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 2.1 Introduction Whether commodity dependent developing countries are trapped in their state of commodity dependence is explored in this chapter. The persistence of dependence is documented, and some potential drivers of the phenomenon are highlighted. Identifying the correlates of commodity dependence could inform policies seeking to enable commodity dependent developing countries to break with commodity dependence. More specifically, how countries move in and out of three states is measured in this chapter. These states are: non-commodity dependence, a state that characterizes countries that derive less than 60 per cent of their merchandise exports from the commodity sector; a state of commodity dependence for countries deriving between 60 per cent and 80 per cent of merchandise export earnings from the commodity sector; and a state of strong commodity dependence for countries deriving more than 80 per cent of their merchandise export earnings from the commodity sector. As noted in chapter 1, Background, if countries moved in and out the three states seamlessly, commodity dependence would not be a serious issue. The problem arises if countries get stuck in one of the two commodity dependence states, given the negative outcomes associated with dependence. In theory, any country may be in any of the three states at any given time. However, if some countries are found to stay in a specific state over a long period of time, this might indicate that they are trapped. Therefore, for this chapter, an empirical analysis was carried out with three objectives. First, using mobility analysis, data from 1995 to 2018 and covering 206 countries and territories shows the average proportion of countries that are in each of the three groups. This highlights the level of short-term mobility. The shortcoming of short-term analysis is that countries could be in a specific group due to factors that are not necessarily associated with dependence, for example short-term shocks to export prices. Taking this into account, the second objective is to determine the distribution of countries in the three groups after all short-term movements have taken place.7 This is key to the concept of a commodity dependence trap, particularly if the ultimate objective is to assess to what extent countries are trapped in the two states of commodity dependence. The third objective is to identify some correlates of commodity dependence, highlighting technology indicators. Empirical results show that commodity dependent countries seem to be trapped in a state of dependence but the consequences of this are more important for commodity dependent developing countries, as explained throughout this report. The implication is that, if they do nothing, time by itself will not get them out of the trap. They will remain commodity dependent and continue to suffer from the negative consequences associated with it. Strong action is therefore needed to change the status quo. Most particularly, strengthening technological capabilities of commodity dependent countries is highlighted as one key avenue that could enable commodity dependent developing countries to move away from commodity dependence. In section 2, illustrative cases of countries trapped in commodity dependence are discussed by briefly presenting the examples of Zambia and Nigeria. Costa Rica is used to illustrate a country that was able to escape from commodity dependence. The methodology used to measure mobility, both in the short term and the long term, is briefly discussed in section 3. Empirical results of mobility are also presented. In section 4, the correlates of commodity dependence are identified, based on the results of an econometric probit model. A conclusion is provided in section 5. 2.2 The commodity dependence trap: A tale of three country trajectories The concept of a commodity dependence trap in this report is used to characterize three different outcomes. The first is a situation where a country is commodity dependent in some reference 7 The technical term for this is ergodic distribution. 10 Chapter 2 - The Commodity Dependence Trap period and remains dependent over a long period. Zambia illustrates this case. The second situation, illustrated by Nigeria, relates to a country where export diversification characterizes its initial conditions but, over time, the country becomes strongly dependent on one commodity. The third case is that of a country that was initially commodity dependent but, over time, diversifies its export sector and moves out of commodity export dependence. Costa Rica exemplifies this case. Data from the Atlas of Economic Complexity8 for the three countries over the period from 1965 to 2018, which spans more than half a century, reveals three different trajectories that summarize the experiences of most developing countries.9 In 1965, copper ore and concentrate, and copper alloys, represented 85 per cent of the net merchandise exports of Zambia. Twenty years later in 1985, the composition of the country’s export basket had hardly improved, with copper and copper alloys, refined or not, and unwrought representing 77 per cent of the country’s merchandise exports. By 2005, merchandise exports were still dominated by copper-based raw materials, accounting for about 60 per cent of the total. In 2018, the export concentration of Zambia around copper had worsened, increasing to almost 80 per cent of total merchandise exports (figure 2.1 (a)). Whereas Zambia has remained dependent on the same commodity for more than half a century, Nigeria was relatively diversified in 1965 but became more and more dependent on one commodity over time (figure 2.1 (b)). In 1965, even though Nigerian exports were dominated by primary commodities, the export basket was diversified with cocoa beans, groundnuts, and palm nuts and kernels, representing 15 per cent, 13 per cent and 10 per cent, respectively, of total merchandise Figure 2.1 (a) Zambia: Main merchandise exports in 1965, 1985, 2005 and 2018 (Percentage of total mechandise exports) 90 80 70 60 50 40 30 20 10 0 1965 1985 2005 2018 Copper ore and concentrates; copper matte; cement copper Copper and copper alloys, worked Copper and copper alloys, re ned or not, unwrought Other Source: UNCTAD, based on data from https://atlas.cid.harvard.edu/explore?country=247&product=undefin ed&year=2018&productClass=SITC&tradeFlow=Net&target=Product&partner=undefined&startYear =undefined (accessed 11 May 2021). 8 Available at https://atlas.cid.harvard.edu/. 9 To access disaggregated data from before 1995, the United Nations standard international trade classification, revision 4, was used. Export shares are calculated using gross trade flows. Data in figures 2.1 (a) and 2.1 (b) are derived using the 4-digit level of disaggregation. However, in the discussion, sectoral level data at 1 digit are also used to show more aggregated information. 11 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Figure 2.1 (b) Nigeria: Main merchandise exports in 1965, 1985, 2005 and 2018 (Percentage of total mechandise exports) Petroleum products, re ned Crude petroleum and oils Cocoa beans, raw, roasted Groundnuts, green Palm nuts and kernel Tin and tin alloys, unwrought Palm oil Groundnut oil 1965 1985 2005 Other 2018 Figure 2.1 (c) Costa Rica: Main merchandise exports in 1965, 1985, 2005 and 2018 (Percentage of total mechandise exports) Coffee Banana Parts of and accessories for machines Electronic microcircuits Fruit, fresh or dried Medical instruments and appliances 1965 1985 Orthopaedic appliances, 2005 hearing aids, etc. 2018 Other 0 20 40 60 80 100 Source: UNCTAD, based on data from https:// atlas.cid.harvard.edu/explore?country=15 9&product=undefined&year=2018&produ ctClass=SITC&tradeFlow=Net&target=Pro duct&partner=undefined&startYear=undefi ned (accessed 11 May 2021). 0 5 10 15 20 25 30 35 40 Source: UNCTAD, based on data from https:// atlas.cid.harvard.edu/explore?country=52 &product=undefined&year=2018&product Class=SITC&tradeFlow=Net&target=Prod uct&partner=undefined&startYear=undefin ed (accessed 11 May 2021). exports. The country also exported palm oil, groundnut oil, and tin and tin alloys, unwrought. Crude petroleum and refined petroleum accounted for 15 per cent and 10 per cent, respectively, of total merchandise exports. Twenty years later, in 1985, the country was exporting almost a single commodity, crude petroleum, which accounted for 97 per cent of total merchandise exports. In 2005, at 92 per cent of total merchandise exports, crude petroleum was still by far the major export of Nigeria. By 2018, the picture had changed only slightly, with crude petroleum still accounting for 81 per cent of total merchandise exports (petroleum gases represented an additional 12 per cent). 12 Chapter 2 - The Commodity Dependence Trap Costa Rica followed a different, more successful trajectory. In 1965, the export base of Costa Rica was dominated by coffee and bananas, representing about 68 per cent of total net merchandise export earnings (figure 2.1 (c)). Overall, food commodities represented 83 per cent of total merchandise exports. Twenty years later, in 1985, these two commodities were still the country’s dominant exports, accounting for 61 per cent of total merchandise exports. Even though food commodities made up 76 per cent of total merchandise exports, there was a nascent manufacturing sector, which contributed about 15 per cent to total merchandise exports, against only 7 per cent 20 years earlier. Thereafter, the country embarked on a diversification drive to the extent that, by 2005, the country’s export basket had dramatically changed. In 2005, the main exports were electronic microcircuits, with a share of 26 per cent of total merchandise exports, followed by parts of and accessories for machines, with a share of 15 per cent. The share of the food sector had dropped to only 24 per cent of the total. By 2018, other sectors that had developed included the medical instruments and appliances, and orthopaedic instruments. Interestingly, the traditional food sector remained important, as banana and fruits represented an important share of exports. This contrasts with the case of Nigeria, suggesting that diversification is not about adopting new products while abandoning traditional ones. In fact, the reconfiguration of exports in 2018 shows the food sector regaining importance, mainly because there were more food commodities exported. Among the most important ones, in addition to bananas, were fruit, fresh or dried, including avocados, pineapples and mangoes; edible products or preparations; fruit and vegetable juices; fruit, prepared or preserved; and bakery products. This illustrates that diversification is not just about adding value to primary commodities or only producing more sophisticated goods. While Costa Rica diversified into more sophisticated goods, it also increased the number of products exported within the commodities sector. This highlights the point that, even when a country remains commodity dependent, it is better off relying on a larger basket of products, as Nigeria was doing in 1965. Currently, as a price taker, the total reliance of Nigeria on exports from the energy sector exposes the country much more to the vagaries of international oil markets. Zambia, Nigeria and Costa Rica illustrate three different trajectories of commodity dependence. Costa Rica illustrates a successful case of export dynamism, from a highly concentrated base to more product and sectoral diversification. The country owes its success to a combination of factors, including the adoption of a long-term plan for economic and export diversification, macroeconomic stability, openness to foreign direct investment, proximity to a large export market and health and education policies that fostered human capital development (UNCTAD and FAO, 2017). Zambia and Nigeria, in contrast, are two different illustrations of the commodity dependence trap. For more than half a century, Zambia has made limited progress in terms of economic and export diversification away from copper. Nigeria, in turn, had the opportunity to maintain a relatively diversified export sector or even develop it further. Instead, commodity dependence worsened over time. There is literature claiming that economies more reliant on point-source natural resources such as Nigeria and Zambia may be more prone to the natural resource curse than those relying on sparsely distributed agricultural commodities. One major transmission channel may be that point-source natural resources are more prone to predation by incumbent politicians and rebels (Collier and Hoeffler, 2004). The latter may capture those resources either at their extraction site or at any choke point when they are moved for export, causing instability and economic decline. This might help explain the contrast between Costa Rica and Nigeria and Zambia. The negative relationship between point-source natural resources and the resource curse might not be generalized (Alexeev and Conrad, 2011). Indeed, when revenues from natural resources are used to develop other sectors and hence contribute to diversifying the economy, commodity dependent developing countries avoid the natural resource curse. Indonesia, for example, derived 71.5 per cent of its merchandise export earnings from the oil and gas sector in 1980. Fifteen years later 13 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 in 1995, the contribution of the fossil fuels sector to total merchandise exports had been reduced to only 22.2 per cent of total merchandise exports. This was the result of massive investment of revenues from the fossil fuels sector in non-commodity sectors, diversifying the economy. By 2018, oil and gas accounted for 10 per cent of total merchandise export earnings. The exports have been diversified into products such as coal, coke and briquettes; iron and steel; road vehicles; vegetable oils and fats; metalliferous ores and metal scrap; apparel and clothing; and electric machinery.10 The discussion of the three country cases above, as well as that of Indonesia, points to one of the main issues addressed in this chapter: when observed over a long enough period, an initially commodity dependent country may move out of commodity dependence. Costa Rica, and Indonesia to some extent, illustrate this case. Zambia and Nigeria, in contrast, seem to be trapped in a state of commodity dependence. The following section briefly presents the methodology used to measure mobility, followed by empirical results. Given that most developing countries, which constitute a large share of the sample, are commodity dependent, knowledge of the likelihood that a country can escape from dependence may inform policy towards structural transformation and export diversification, as discussed in detail in chapter 5. Before calling on commodity dependent developing countries to diversify their economies, it is important to first understand the extent of the challenge that they face, helping in turn to understand why they have been stuck in commodity dependence for such a long time. 2.3 Measuring mobility between commodity dependence States The first part of the empirical analysis relies on transition analysis, using a methodology adapted from Nkurunziza (2015). The second part uses regression analysis to uncover some correlates of commodity dependence. A brief discussion of the methodology Observed at any specific time, any country can be classified in one of the three states defined above: the country may be not commodity dependent, commodity dependent or strongly commodity dependent. Allowing for a long period of time, countries may move between the three states. After all, the old debate about the need for commodity dependent developing countries to diversify their economies implies moving from a commodity-dependent to a non-commodity dependent state. Costa Rica achieved this over several decades. However, many commodity dependent developing countries remain commodity dependent even when they are observed over a period spanning half a century, as Zambia and Nigeria illustrate. These countries seem to be trapped.11 Mobility is analysed with a transition matrix. This is a tool used to determine the probability that a country in a reference period remains in the same group in the next period or moves to some other state. The unit of observation in this chapter is one year, but when results are presented, mobility is aggregated over a 24-year period, corresponding to 23 potential annual transitions, from 1995 to 2018. The 24-year period is dictated by data availability. It is long enough to study short-term mobility but not long enough to conclude that some countries may be trapped. Hence, the short-term distribution of countries in the three states is used to derive a long-term 10 Based on data from the Atlas of Economic Complexity, available at https://atlas.cid.harvard.edu/explore?country=103&product=undefin ed&year=2018&productClass=SITC&target=Product&partner=undefined&startYear=undefined (accessed on 18 January 2021). 11 Strictly speaking, the statistical concept of a “trapping” state in the analysis of dynamic systems means that once a country is in this state, it is impossible for it to move to any other state (Robert and Casella, 1999). In this chapter, unless otherwise specified, the concept is used to represent situations of very slow mobility, with probability close – but not equal – to zero. 14 Chapter 2 - The Commodity Dependence Trap or equilibrium distribution.12 Indeed, the analysis of commodity dependence dynamics requires a relatively long period of observation given that the economic transformation process leading to export diversification takes decades, as illustrated by the case of Costa Rica. Moreover, regression analysis is used to probe the potential role that technology and innovation could play to help commodity dependent developing countries move out of the commodity dependence trap. The sample period is from 1995 to 2018, with 4,944 observations or country-years (24 years for each of the 206 countries and territories). The 24-year period captures different phases of the commodity price cycle. Between 1995 and 2002, commodities prices were low, corresponding with a declining phase of the price cycle that had started in the early 1980s (figure 2.2). The period between 2003 and 2011 was characterized by a commodity price boom, with commodity prices increasing manyfold over a few years. Between 2012 and 2018, commodity prices were declining, even though they remained higher than their levels before the commodity boom of the 2000s. Indeed, before presenting empirical results of mobility, it is worth discussing first commodity price trends, considering that commodity dependence may be a function of prices, at least in the short term. For example, during the last commodity price boom, the number of commodity dependent developing countries increased from 110 in 2005 to 118 at the end of the boom in 2011. Thereafter, the number declined (Nkurunziza et al., 2017). Figure 2.2 Commodity prices: A sixty-year perspective 5 Smoothed log non-energy price 4 Log of price index (Base year = 2010) 3 2 Smoothed log energy price 1 1960m1 1980m1 2000m1 Time 2020m1 Smoothing by local polynomial; vertical line is 1995m1: start time for empirical analysis; m1 in January Source: UNCTAD, based on the World Bank commodity price index, using 2010 as the base year. Note: Data smoothing by local polynomial. Vertical line is 1995m1, start of sample used in empirical analysis. * Sixty years, from January 1960 (1960m1) through July 2020. ** For all years, m1 indicates that data covered is from month 1 (January) of the year. 12 A formal discussion may be found in a background paper prepared for this chapter: Nkurunziza JD, 2021, The commodity dependence trap. Background paper prepared for the 2021 edition of the Commodities and Development Report, UNCTAD, available at https://unctad. org/webflyer/commodities-and-development-report-2021. 15 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Decomposing commodities into fuel and non-fuel groups of commodities, figure 2.2 illustrates commonalities and differences in the behaviour of the price trends for both groups. The major common factor is that prices follow the same long-term trend. Prices have tended to increase, stabilize and decline at the same time. This finding implies that, over a long enough period, the markets for energy and non-energy commodities are fundamentally affected by the same major factors, namely supply and demand. This strong correlation suggests that, while diversification within the wider commodity sector may help, to some extent, only diversification out of the commodity sector could be an answer to the deleterious effects of commodity dependence. The manufacturing sector is identified in chapter 3 as being the sector towards which commodity dependent developing countries should strive to diversify. In other words, even though some types of commodities might face idiosyncratic challenges – for example, climate change may have a much stronger impact on agricultural commodities than on minerals, ores and metals – the major challenges facing the commodity sector are the same. Many of these challenges are highlighted in chapter 1. But there is a major difference between the two types of commodities, as illustrated by figure 2.2: the amplitude of price changes. Energy commodities experience much stronger price changes than non-energy commodities. For example, because of the commodity price boom of the 1970s, the energy price index increased by more than 1 350 per cent, from 3.4 to 49.50 between January 1973 and December 1980. Over the same period, the non-energy commodity price index increased by 114 per cent, from 27.07 to 57.87. Unequal increases in prices were again recorded during the commodity boom of the 2000s, with energy prices increasing by 552 per cent between January 2002 and July 2008, while non-energy prices increased by 189 per cent over the same period. Energy markets have also been characterized by sudden and drastic price drops, shattering exporting countries’ economies. High price volatility is indeed an intrinsic characteristic of energy markets. Between June 2014 and January 2015, oil prices dropped by more than half in just six months. In June 2014, the energy price index stood at 131.48, but it had fallen to 63.10 by January 2015, less than half its value six months earlier. Countries that had planned spending based on an oil price of $112 per barrel in June 2014 faced the challenge of substantially cutting their budgets to adapt to a price that had reached $45 by 13 January 2015. The brutal effect of commodity dependence was strongly felt by oil-export dependent countries across the world, including Angola, the Islamic Republic of Iran, Nigeria, Saudi Arabia and the Bolivarian Republic of Venezuela. Commodity dependence is harmful not only as price shocks are destabilizing, but also as the relatively short periods of high commodities prices are followed by much longer periods of depressed prices, as illustrated by figure 2.2. This information may help to understand why commodity dependence might manifest itself differently in relation to the type of commodity on which a country depends. Econometric results seem to confirm this hypothesis (UNCTAD, 2019). Empirical results The empirical transition matrix in table 2.1 summarizes aggregate mobility of 206 economies, both developed and developing, which represents almost all countries and territories in the world. Table 2.1 provides three sets of information. First, the last row is a summary measure of mobility, showing the average proportion of countries in each of the three states after mobility has taken place over a period of 24 years. On average, over the sample period, half of the countries were in the noncommodity dependent state. The other half were in the strongly dependent state (32 per cent of the sample) and the commodity dependent state (18 per cent of the sample). This summary information suggests that, while a widespread characteristic, commodity dependence – and its strong version – only affect half of the countries in the sample. Second, the fact that all elements of the table are non-zero, even though some are small, implies that there is indeed mobility across all states. 16 Chapter 2 - The Commodity Dependence Trap Table 2.1 Commodity dependence: Mobility across three states, 1995–2018 (As an average) Non-commodity dependent Commodity dependent Strongly commodity dependent Annual average proportion of countries Non-commodity dependent 0.95 0.13 0.01 0.50 Commodity dependent 0.04 0.75 0.07 0.18 Strongly commodity dependent 0.01 0.12 0.92 0.32 Source: UNCTAD, based on UNCTADstat database. Note: Values are interpreted as probabilities. The third set of information relates to values of the table, interpreted as probabilities. Starting with the two extreme values of the diagonal (in the table): there is evidence of limited mobility from the non-commodity dependent and the strongly commodity dependent groups. During the sample period, 95 per cent of non-commodity dependent countries remained within this group. The proportion of strongly commodity dependent countries that did not move out of the category is 92 per cent. Another way of interpreting these findings is that the risk that a non-commodity dependent country becomes commodity dependent or strongly commodity dependent is 4 per cent and 1 per cent, respectively. Similarly, the likelihood that a strongly commodity dependent country becomes non-commodity dependent over the 24-year period is very small. But there is a 7 per cent chance that such a country improves from strong commodity dependence to just commodity dependence. Even though this might be considered as an improvement, this information needs to be put in context as both commodity dependent and strongly commodity dependent countries face the same challenges, only with higher severity for the latter group. The value in the middle of the diagonal suggests an almost equal likelihood that a commodity dependent country becomes non-commodity dependent (probability of 13 per cent) or strongly commodity dependent (probability of 12 per cent). On average, three quarters of commoditydependent countries remain in the same state, over the sample period. This result suggests that, a priori, while some countries graduate into the non-commodity dependence category, an almost equal number fall into the worse state of strong commodity dependence. The implication is that relatively few countries can escape from commodity dependence. One question is whether a period of 24 years of observation is long enough to properly capture the transition processes occurring across all commodity dependence states. In other words, could transitions captured in table 2.1 be the result of short-term analysis of a phenomenon that requires a longer period of analysis? Indeed, it might be argued that commodity dependence is correlated with commodity price cycles that are generally longer than the 24-year sample period. Hence, it is important to analyse the evolution of commodity prices over a longer time span to ensure that what is captured by the transition matrix may be considered as representing a general pattern of mobility. To answer this question, it is important to establish that the distribution reflected in table 2.1 does not change over time. Following the methodology discussed in detail in Nkurunziza (2015) and briefly exposed in a background paper,13 the long-term distribution of countries in the three groups (see figure 2.3) is almost the same as the short-term distribution reflected in table 2.1. Moving from short-term to long-term analysis (figure 2.3), there is a small change in the proportions of countries in the non-commodity dependent and the commodity-dependent categories, from 50 per cent to 51 per cent, and from 18 per cent to 17 per cent, respectively. This finding suggests that 13 Ibid., available at https://unctad.org/webflyer/commodities-and-development-report-2021. 17 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Figure 2.3 Long-term distribution of countries in the three states (As an average) 1.0 0.8 0.51 Non-commodity dependent 0.6 0.4 0.17 Commodity dependent 0.2 0.32 Strongly commodity dependent 0.0 commodity dependence as characterized by data in the sample depicts a stable distribution of countries in the three states. This is consistent with the examples of Zambia and Nigeria which show that, for close to 60 years, the countries have remained not only commodity dependent but also dependent on the same commodity. This confirms that commodity-dependent and strongly commodity-dependent countries are indeed in a trap. Source: UNCTAD, based on table 2.1 of the present Another way of ascertaining the difficulty of report). emerging out of the dependence trap is to determine the time it would take a commodity dependent country to become non-commodity dependent, given mobility as observed between 1995 and 2018. Considering that only 1 percentage point of countries moves from commodity dependence to non-commodity dependence between the short-term and long-term analyses, and if no strong action is taken to accelerate mobility, it would take the average commodity dependent country 190 years to reduce by half the difference between its current share of commodities in total merchandise exports and that of the average non-commodity dependent country.14 This result illustrates the challenge facing commodity dependent developing countries. Unless they take strong action to change the status quo, they will remain commodity dependent for the coming centuries. This seems to be the trajectory that characterizes Zambia and Nigeria, as well as most other commodity dependent developing countries. Doing nothing or not doing enough does not seem to be an option as commodity dependence will not disappear on its own. Could technology be one of the disruptive factors that could help commodity dependent developing countries to change their trajectory towards more diversified economies? The experience of Costa Rica shows that an economy can indeed be transformed to become more diversified. As mentioned earlier, success requires time, strong political will and a long-term, realistic development vision, coupled with an ambitious but reasonable implementation strategy (UNCTAD and FAO, 2017). The remainder of the chapter identifies some correlates of commodity dependence. This information may provide potential entry points towards economic structural transformation and diversification. 2.4 Correlates of commodity dependence Correlates of commodity dependence are identified based on available data. Most particularly, given the report’s specific interest in technology, this section presents the results of a simple probit econometric model regressing commodity dependence on several indicators of technology, and other control variables. 2.4.1 Discussion of the variables Five indicators of technology are used (see the descriptive statistics in table 2.3). First, the share of a country’s population using the Internet captures the deployment of ICTs within an economy. The Internet provides greater access to information, reduces production costs and allows far greater connectivity between people, firms and other economic agents, ultimately leading to higher productivity. 14 Ibid., available at https://unctad.org/webflyer/commodities-and-development-report-2021. 18 Chapter 2 - The Commodity Dependence Trap The second variable is the speed of the Internet. This captures the quality of ICT deployed in a country. Indeed, productivity improvements result from not only accessing the Internet, but also having Internet connections that work well and at speeds that facilitate transactions without interruption. The Internet creates so many opportunities that it is has become a vital economic resource in modern societies. The Internet permeates economic and social life to such a degree that it is now at the root of the digital divide (Aydin, 2021) between those who have access to it and those who do not. Third, high-skill employment as a percentage of the working population is an indicator that measures the quality of human resources available in a country. Indeed, measures such as the employmentto-population ratio do not account for the fact that many jobs, particularly in developing countries, are low skilled and contribute little to export and economic structural transformation. High-skill employment, on the other hand, is associated with technological innovations that commodity dependent developing countries might need to adopt to create new products and reduce their strong dependence on commodity exports. Fourth, the number of scientific publications on frontier technologies (research and development publications) is considered as a good proxy of technological activities taking place within an economy. The higher the number, the more technologically advanced an economy is. Fifth, high technology manufactures exports as a percentage share of total merchandise trade measures the share of technologically sophisticated goods that are exported by a country. As primary commodities are less sophisticated exports, moving out of the commodity dependence trap implies that commodity dependent developing countries need to upgrade their productive systems to produce manufactured goods and services that are more sophisticated. All five technology indicators show the capacity of a country to produce and export goods and services with a high technology content, unlike commodities that embed a low level of technology.15 Put differently, countries displaying high levels of technology indicators are less dependent on the commodity sector for their exports. Technology and innovation enable them to diversify exports into high-value goods and services that are less prone to the negative shocks that afflict primary commodities. Therefore, improving technology and innovation in commodity dependent developing countries is expected to help them diversify into high-value exports and increase as well as stabilize export earnings. In addition to technology and innovation indicators, other control variables, all sourced from the UNCTADstat database, are included as correlates of commodity dependence (table 2.3).16 These are the shares of fuels, minerals and agriculture exports in total exports. These are expected to capture the fact that commodity dependence may be a function of the commodity sector a country depends on, as discussed earlier (see also UNCTAD, 2019). High shares of energy and minerals in total merchandise exports tend to be associated with strong commodity dependence. In many developing countries, energy and minerals sectors concentrate the bulk of investment, particularly foreign direct investment, translating thus into highly concentrated economic activity. Sectoral allocation of employment may potentially be related to commodity dependence if a large share of employment is in a sector producing the commodity (or commodities) on which a country depends. Another way of analysing dependence is to probe the sectoral contribution to value added in an economy. High value added shares indicate the importance of a sector to an economy. For example, in commodity dependent developing countries, value added is generally generated 15 As explained earlier, extractives in commodity dependent developing countries may display high levels of technologies, but these are not embedded in the domestic economic system; they are instead controlled by multinational enterprises that operate enclave projects. 16 A detailed discussion of many of these variables and why they matter for commodity dependence and economic structural transformation is found in chapter 3. 19 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Table 2.2 Descriptive statistics of the variables included in the econometric model Variable Dummy variable for commodity dependence Share of fuels exports in total merchandise exports Share of minerals exports in total merchandise exports Share of agriculture exports in total merchandise exports Share of employment in industry Share of employment in services Share of value added in agriculture Share of value added in industry Share of value added in services Dummy variable for least developed country status Internet users (share of population) Internet speed, in megabits per second (mean download speed) High-skill employment (% working population) Research and development publications (no. of scientific publications on frontier technologies) High-technology manufacturing (% total merchandise trade) Mean 0.52 0.17 0.08 0.26 19.67 50.42 12.58 26.60 53.25 0.23 0.52 0.50 0.39 0.40 0.56 Median 1.00 0.05 0.02 0.18 20.14 52.02 8.38 24.81 53.56 0.00 0.55 0.46 0.37 0.38 0.56 Observations 4 944 4 675 4 764 4 770 4 392 4 392 4 378 4 366 4 188 4 944 474 316 474 316 474 Source: UNCTAD, based on data from the International Labour Organization, International Telecommunication Union and the UNCTADstat database. in low-technology and low-skilled primary and services sectors, suggesting that exports would also tend to be low-skilled and low-technology, which is the case for commodities. Finally, a least developed country categorical variable is introduced to proxy for three dimensions of a country’s level of development. This variable captures a country’s income level, human assets and economic vulnerability.17 The status of least developed country implies that a country faces development problems that may add to the negative effect of commodity dependence, making development even more challenging. Such countries require more attention than non-least developed countries (Gore and Kozul-Wright, 2011). In this chapter, the variable “least developed country” takes a value of 1 if a country is a least developed country and a value of 0 otherwise. Technology indicators cover only a few years, between 2015 and 2018, and slightly fewer countries, depending on which indicator is considered.18 Despite this reduction in the sample size, there is no reason to consider that the relationship between technology and commodity dependence was different during the period not covered by the data – before the period 2015–2018. It is assumed that the period covered is representative of the full sample period. Indeed, the empirical results show that these indicators reflect the expected relationship between commodity dependence and technological development. 17 See UNCTAD, 2017, The Least Developed Countries Report: Transformational Energy Access, What are the least developed countries? (United Nations publication, Sales No. E.17.II.D.6, New York and Geneva), pp. v and vi. 18 Some variables, such as Internet speed and research and development publications, cover only two years and 158 countries each. 20 Chapter 2 - The Commodity Dependence Trap 2.4.2 Empirical results Econometric results19 suggest a number of interesting findings regarding commodity dependence and its correlates. First, all the technology variables are strongly and negatively correlated with commodity dependence. This suggests that the odds of commodity dependence are strongly associated with low levels of technology. In other words, countries with higher technological capabilities are less likely to be commodity dependent. Figure 2.4 illustrates the strong and negative correlation between commodity dependence and the level of technological development, across all indicators of technology. If the results were to be interpreted as representing causality relationships, they would suggest that, by strengthening their technological capabilities, commodity dependent countries may reduce the vulnerabilities associated with commodity dependence. Indeed, improving the technological ecosystem of commodity dependent countries would create opportunities by increasing production outside the commodity sector. As chapter 4 shows, weak technologial ecosystems in commodity dependent developing countries coexist with the least sophisticated and low-value product baskets. Acquiring technological capabilities and adopting institutions that foster innovation and technological development, as argued in chapter 5, could reduce the dependence on commodities of commodity dependent developing countries and the negative implication of that dependence for economic development. There is also a positive and statistically significant relationship between commodity dependence and export shares of the three types of commodites. This means that the problem of the commodity dependence trap is not limited to some type of commodity on which a country depends. However, Figure 2.4 Technology level in commodity dependent developing countries and non-commodity dependent developing countries 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Internet users Internet speed High-skill employment Research and development publications High-technology manufacturing Digital services exports Commodity dependent developing countries Source: UNCTAD. Non-commodity dependent developing countries 19 See Nkurunziza JD, 2021, available at https://unctad.org/webflyer/commodities-and-development-report-2021. 21 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 the odds of commodity dependence are not uniform across commodity types. The correlation is strongest for countries dependent on exports of minerals. The implication might be that the issue of commodity dependence is more entrenched in mineral exporting countries and, to a great extent, fuel export-dependent countries. One reason could be that extractives in commodity dependent developing countries are generally enclave sectors dominated by foreign firms that have little incentive to create backward and forward domestic linkages with non-commodity sectors (Hansen, 2013). As value addition to primary commodities takes place outside producing countries, the latter do not benefit from value creation and its attendant advantages, including income generation, job creation and tax revenue, along the value chain. Other results suggest that development of the industrial sector would be a relevant way of addressing the commodity dependence issue. Indeed, industrial production, even when it uses commodities as inputs contributes to product and economic diversification. Finally, the least developed country variable indicates that the least developed countries are more affected by commodity dependence than other countries. If the relationship were to be interpreted as causal, that would mean that, other things being equal, the odds of being commodity dependent are between 2.7 and 5.0 times higher for a least developed country than for a non-least developed country.20 Commodity dependence seems to be positively correlated with other vulnerabilities embedded in the least developed country variable discused above. 2.5 Conclusion The main objective of this chapter was to document the level of commodity dependence and determine whether countries are trapped in a state of commodity dependence. The next step was to identify the correlates of commodity dependence in order to offer insights into possible pathways towards escaping from the dependence trap. Empirical analysis confirmed that commodity dependent countries are indeed trapped in a state of dependence. The likelihood that a (strongly) commodity dependent country becomes noncommodity dependent is very low, as shown in table 2.1. This finding seems to hold even when allowing for a long time period, implying that, if the countries concerned do not take strong action to change the status quo, they will stay commodity dependent for a very long time. Econometric results seem to confirm the hypothesis that technology and innovation could play an important role in helping countries escape from the trap. However, this process of change appears particularly challenging for countries dependent on the extractive sector. In this regard, fostering the development of a technology ecosystem in commodity dependent developing countries that encourages production of more sophisticated goods would be appropriate. Defeating commodity dependence will require that commodity dependent developing countries put in place the right physical and institutional infrastructure that allows this technology ecosystem to thrive. The message about the potential positive role that technology and innovation can play in enabling commodity dependent countries to escape from the dependence trap is key to this report. This finding lays the ground for further discussions in subsequent chapters. 20 These values are based on the smallest and the largest coefficients of the econometric models, namely the models with research and development publications and Internet use, respectively. Given that the coefficients cannot be interpreted as elasticities, the values are obtained as e1.01 ≈ 2.74 and e1.62 ≈ 5.05. 22 Chapter 2 - The Commodity Dependence Trap References Alexeev, M and Conrad R (2011). The natural resource curse and economic transition. Economic Systems. 35(4): 445–461. Aydin M (2021). Does the digital divide matter? Factors and conditions that promote ICT literacy. Telematics and Informatics. 58:2021(101536). Collier P and Hoeffler A (2004). Greed and grievance in civil war. Oxford Economic Papers. 56:563– 595. Gore C and Kozul-Wright Z (2011). An overview of UNCTAD Least Developed Countries Report 2010: Towards a new international development architecture for [least developed countries] LDCs. The European Journal of Development Research. 23:3–11. Hansen MW (2013). From enclave to linkage economies? A review of the literature on linkages between extractive multinational corporations and local industry in Africa. Danish Institute for International Studies, Working Paper 2014:02. Copenhagen Business School. Nkurunziza JD (2021). The commodity dependence trap. Background paper prepared for the 2021 edition of the Commodities and Development Report. UNCTAD. Available at https://unctad. org/webflyer/commodities-and-development-report-2021. Nkurunziza JD (2015). The distribution of firm size in Africa’s manufacturing sector and implications for industrial policy. Journal of African Development. 17(2):49–81. Nkurunziza J, Tsowou K and Cazzaniga S (2017). Commodity dependence and human development. African Development Review. 29:1–15. Robert CP and Casella G (1999). Monte Carlo Statistical Methods. Springer Verlag. New York. UNCTAD (2017). The Least Developed Countries Report: Transformational Energy Access. (United Nations publication, Sales No. E.17.II.D.6, New York and Geneva). UNCTAD (2019). Do Differences in the Types of Commodities Exported Matter for Export Concentration? (United Nations publication. Geneva). UNCTAD and FAO (2017). Commodities and Development Report 2017. Commodity Markets, Economic Growth and Development. (United Nations publication. Sales No. E.17.II.D.1 New York and Geneva). 23 CHAPTER 3 Commodity Dependence, Productivity and Structural Change COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 3.1 Introduction The connections between commodity dependence, labour productivity trends and structural change in commodity dependent developing countries are analysed in this chapter. Improvements in labour productivity are a key source of economic growth and thus closely linked to the overall development process in low-income and middle-income countries. In particular, labour productivity growth can be a long-term driver of increased real wages and improved living standards in developing countries. The importance of labour productivity growth in the development process is reflected in the inclusion of a related indicator under Millennium Development Goal 1 on eradicating extreme poverty and hunger and in its inclusion in the Sustainable Development Goals framework: target 8.2 aims to “achieve higher levels of economic productivity through diversification, technological upgrading and innovation, including through a focus on high value added and labour-intensive sectors” and indicator 8.2.1 is the annual growth rate of real GDP per employed person, as noted in resolution 71/313 of the General Assembly. Diversification and technological development play crucial roles in labour productivity growth. The growth of economy-wide labour productivity can be driven by productivity growth in individual sectors and/or by productivity-enhancing structural change, that is, a reallocation of production factors from sectors with lower levels of productivity to sectors with higher levels. Either of these components can also have negative impacts on aggregate labour productivity. In this context, technological upgrading and innovation can be important drivers of within-sector labour productivity growth. Structural change is particularly relevant to labour productivity growth if there are significant differences in productivity levels across sectors. Such differences tend to be greatest in low-income countries, in which agriculture is typically the least productive sector but employs a large share of the labour force. Starting from the observation that commodity dependent developing countries exhibit lower average levels of labour productivity growth than other country groups, a key question addressed in this chapter is whether commodity dependence acts as an inhibitor to the within-sector component, the structural change component or both components of labour productivity growth. This is a question of significant practical relevance for policymakers in commodity dependent developing countries. For example, if commodity dependence is a drag on growth-enhancing structural change, policy interventions should focus on facilitating the flow of production factors from low-productivity to higher-productivity sectors. However, if commodity dependence weighs down within-sector productivity growth, policies that induce such growth at the sectoral level need to be strengthened. Finally, if commodity dependence is a drag on both components, a policy mix will be needed. As shown in this chapter, commodity dependence is associated with low levels of labour productivity, slow productivity growth, high volatility in productivity growth and a high frequency of negative productivity shocks. The link between commodity dependence and stunted productivity growth is particularly strong in the manufacturing sector. Furthermore, there is a strong association between technology development and labour productivity growth across sectors. Overcoming commodity dependence can strengthen the role of the manufacturing sector as a driver of economic growth and productive employment, which can, directly and indirectly, contribute to the achievement of the Sustainable Development Goals. Technological upgrading and innovation can play important roles in the diversification process. The chapter has five sections, as follows: in section 3.2, labour productivity trends are analysed through the lens of commodity dependence; in section 3.3, the patterns of structural change in commodity dependent developing countries since 1995 are highlighted; in section 3.4, sectoral productivity trends and drivers and their relationship with commodity dependence and technological development are examined; and in section 3.5, a summary and conclusions are provided. 26 Chapter 3 - Commodity Dependence, Productivity and Structural Change 3.2 Labour productivity trends Labour productivity is defined as output per unit of labour. It is therefore calculated by dividing total output by the number of workers or the number of work hours in a given period. National GDP, value added generated by an economic sector or value added generated by an individual firm can each act as a proxy for output. Aggregate labour productivity is defined as the labour productivity of the economy as a whole, that is, GDP per worker. In this chapter, the term “labour productivity” refers to aggregate labour productivity unless specified otherwise and, for country groups, medians are used when the indicator reflects a specific level (for example, labour productivity in dollars) and averages are used when the indicator reflects a percentage (for example, growth rate of labour productivity). In 1995–2018, the median labour productivity in commodity dependent developing countries was substantially below the median in non-commodity dependent developing countries and developed countries and, from 1999 onward, labour productivity in transition economies exceeded that in commodity dependent developing countries, with a rapidly widening gap (figure 3.1). The difference between the median labour productivity in commodity dependent developing countries and all other country groups was significantly greater in 2018 than in 1995, implying that while other groups significantly improved labour productivity, progress in commodity dependent developing countries was muted. Labour productivity in the latter was virtually stagnant from 1995 until the Figure 3.1 Median labour productivity (Thousands of constant 2010 dollars) 14 12 Developed countries, right axis 100 Transition economies 90 80 10 70 Non-commodity dependent developing countries 60 8 50 6 40 Commodity dependent developing countries 4 30 20 2 10 0 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Source: Notes: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. Transition economies, developing countries and developed countries are defined as in the UNCTADstat database. Commodity dependent developing countries are defined as developing countries with an average share of primary commodities in total merchandise exports greater than 60 per cent in 1995–2018. The data set covers 166 economies in 1995–2018 (see appendix, table A). 27 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 onset of the commodity price boom in 2003 and the compound annual growth rate of the median labour productivity in these countries in 1995–2002 was only 0.1 per cent. This rate increased to 4.3 per cent in the boom period in 2003–2011, after which growth levelled off, and the rate was negative in 2012–2018. During the boom period, labour productivity growth in commodity dependent developing countries was primarily fuelled by an accelerated flow of workers out of the agricultural sector towards non-farm employment in higher-productivity sectors and, to a lesser extent, by labour productivity growth within services sectors. The majority of workers exiting the agricultural sector moved to the construction sector and relatively low-productivity services sectors. In particular, the construction sector in commodity dependent developing countries benefited from increased spending on infrastructure and significant investments in mining undertaken during the boom period (World Bank, 2015). In 1995–2018, the average annual growth rate of labour productivity in commodity dependent developing countries was 1.5 per cent, lower than in developed countries, at 1.7 per cent; noncommodity dependent developing countries, at 2.3 per cent; and transition economies, at 4.9 per cent (figure 3.2). Therefore, combined with a low initial level of labour productivity, slow productivity growth has been widening the productivity gap between commodity dependent developing countries and other country groups. In addition to experiencing slower labour productivity growth, commodity dependent developing countries have also experienced negative productivity shocks at a greater frequency than other country groups. In 1995–2018, these countries experienced negative aggregate labour productivity growth on average once every three years, significantly more frequently than non-commodity Figure 3.2 Average annual growth rate of labour productivity, 1995–2018 (Percentage) 5 4 3 2 1 0 Commodity dependent developing countries Developed countries Non-commodity dependent developing countries Transition economies Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. 28 Chapter 3 - Commodity Dependence, Productivity and Structural Change dependent developing countries, at 4.3 years; developed countries, at 5.8 years; and transition economies, at 7.2 years (figure 3.3, panel (a)). Labour productivity growth in commodity dependent developing countries was also more volatile than in non-commodity dependent developing countries and in developed countries, but less volatile than in transition economies (figure 3.3, panel (b)). As shown in this section, in 1995–2018, in terms of aggregate labour productivity, commodity dependent developing countries lagged behind other country groups, including non-commodity dependent developing countries. Furthermore, commodity dependence was associated with comparatively low levels of labour productivity growth, a greater frequency of negative productivity shocks and an elevated volatility in productivity growth. 3.3 Structural change patterns Aggregate productivity trends are determined by productivity trends within individual sectors and by changes in the structural composition of an economy. To explain the aggregate productivity trends in commodity dependent developing countries highlighted in section 3.2, it is therefore necessary to examine the structures of their economies. The structure of an economy can be described by the relative weights of its individual sectors, typically expressed as the share of value added or employment but which may also be expressed as the share of final consumption; the evolution of these shares over time is referred to as structural change. Developed countries underwent profound structural change along their development paths, which featured similar patterns of industrialization followed by an expansion of the weight of services in value added and employment (Herrendorf et al., 2013). Developing countries have also experienced structural change, but its depth and contribution to economic growth has varied substantially across countries since 1990 (McMillan et al., 2017). Structural change characteristics in commodity dependent developing countries in 1995–2017 are highlighted in this section using a data set that disaggregates an economy into nine sectors (table 3.1). The empirical analyses in this section and in section 3.4 are based on a data set that incorporates data from the World Development Indicators database and the International Figure 3.3 Labour productivity, 1995–2018 (Percentage) (a) Average interval between occurrences of negative labour productivity growth (b) Average standard deviation of negative annual growth rate of labour productivity 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 Transition economies Commodity dependent developing countries Non-commodity dependent developing countries Developed countries Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. 29 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Table 3.1 Sectoral disaggregation of labour productivity Sector Agriculture Mining Manufacturing Utilities Construction Trade services Transport services Financial and business services Other services Components Agriculture, forestry and fishing Mining and quarrying Manufacturing Electricity, gas, steam and air conditioning supply Construction Wholesale and retail trade; repair of motor vehicles and motorcycles; accommodation and food service activities Transportation and storage; information and communications Financial and insurance activities; real estate activities; professional, scientific and technical activities; administrative and support service activities Public administration and defence; compulsory social security; education; human health and social work activities; arts, entertainment and recreation; other service activities; activities of households as employers; undifferentiated goods-producing and services-producing activities of households for own use; activities of extraterritorial organizations and bodies Source: UNCTAD, based on Dieppe and Matsuoka, 2020. Telecommunication Union and trade data from the UNCTADstat database, as well as sectoral labour productivity data from the World Bank and a range of indicators from the Penn World Table (Dieppe and Matsuoka, 2020; Feenstra et al., 2015). This data set covers 94 countries in 1995–2017 that represent more than 90 per cent of global GDP and more than 85 per cent of the global population, according to data on GDP in 2019 (purchasing power parity) from the International Monetary Fund and data on the global population in 2019 from the United Nations world population prospects database (see appendix, table B). In 1995–2017, structural change in commodity dependent developing countries was characterized by a steady flow of labour out of the agricultural sector and into services (figure 3.4, panel (a)). The average share of the agricultural sector in total employment decreased from 51.5 per cent in 1995 to 38.1 per cent in 2017. In the same period, the average share of services increased from 34.6 to 44.9 per cent. The average share of manufacturing remained almost constant, from 7.9 per cent in 1995 to 7.8 per cent in 2017. In commodity dependent developing countries, shares of value added showed similar trends as shares of employment (figure 3.4, panel (b)). In 1995–2017, the average share of agriculture in total value added decreased from 21.1 to 15.1 per cent. In the same period, the average share of services increased from 50.1 to 57.0 per cent and the average share of manufacturing decreased by 1.1 percentage points, from 11.5 to 10.4 per cent. These trends show that structural change in commodity dependent developing countries did not follow a path of industrialization in 1995–2017. This suggests that these countries, as a group, are not moving towards target 9.2 under the Sustainable Development Goals to “promote inclusive and sustainable industrialization and, by 2030, significantly raise industry’s share of employment and GDP, in line with national circumstances, and double its share in least developed countries”. The two indicators, as noted in resolution 71/313 of the General Assembly, are manufacturing value 30 Chapter 3 - Commodity Dependence, Productivity and Structural Change added as a proportion of GDP and per capita; and manufacturing employment as a proportion of total employment. It is important to note that the level of manufacturing value added per capita is closely linked to average income and therefore to a range of other Goals, including Goal 1 on ending poverty and Goal 8 on promoting sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all (figure 3.5). Figure 3.4 Commodity dependent developing countries: Average sectoral shares (Percentage) (a) Employment 60 50 40 30 20 10 0 (b) Value added 60 50 40 30 20 10 0 Agriculture Services Manufacturing Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. Gross domestic product per capita 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 Figure 3.5 Manufacturing and output linkages, 2019 (Constant 2015 dollars) 13 12 11 10 9 8 7 6 5 4 0 2 4 6 8 10 12 Manufacturing value added per capita Source: Notes: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. The data set includes all 208 economies in the database. The figure shows the natural logarithms of GDP per capita and value added per capita and a linear trendline. 31 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 The weight of manufacturing in employment and value added in commodity dependent developing countries has stagnated at levels far below those in non-commodity dependent developing countries (figure 3.6) and even further below the peak levels in developed countries (table 3.2). Commodity dependent developing countries also lag substantially behind non-commodity dependent developing countries in terms of the share of global manufacturing employment, with a gap that widened from 27.6 percentage points in 1995 to 32.4 percentage points in 2017 (figure 3.7). Given the crucial role of the manufacturing sector in the development process (see Haraguchi et al., 2017, Rodrik, 2013, Rodrik, 2016, and Szirmai, 2012), this indicates an important policy challenge for commodity dependent developing countries. It is important to note that the manufacturing sector continues to expand at the global level and can therefore still be a driver of growth in developing countries, including commodity dependent developing countries. Global manufacturing value added increased in terms of both level and per capita in 1990–2019, even when data for China is excluded (figure 3.8). Figure 3.6 Average share of manufacturing (Percentage) (a) Employment 18 16 14 12 10 8 6 4 2 0 Non-commodity dependent developing countries Commodity dependent developing countries 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995 (b) Value added 25 20 15 10 5 0 Non-commodity dependent developing countries Commodity dependent developing countries 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995 Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. 32 Chapter 3 - Commodity Dependence, Productivity and Structural Change Table 3.2 Selected developed countries: Greatest share of manufacturing in total employment Australia Canada France Germany Japan Republic of Korea United Kingdom United States Share of manufacturing in total employment (percentage) 24.7 22.9 26.0 35.8 26.2 28.7 30.1 22.6 Year of peak level 1971 1970 1973 1970 1973 1989 1971 1970 Source: Notes: UNCTAD, based on data from the International Labour Organization, International Telecommunication Union and the UNCTADstat database. Germany refers to the former Federal Republic of Germany. Manufacturing employment data is not available for the United Kingdom for 1970 Figure 3.7 Share of global manufacturing employment (Percentage) 45 40 35.9 35.9 37.0 38.8 35 32.3 30 25 20 15 10 4.7 5.2 5.4 5.8 6.4 5 0 1995–1999 2000–2004 2005–2009 2010–2014 2015–2017 Commodity dependent developing countries Non-commodity dependent developing countries Source: Note: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. Data for non-commodity dependent developing countries excludes China, the country with the greatest number of manufacturing jobs, since the inclusion of this data would show an even wider gap and a greater increase in the gap in 1995–2017. In commodity dependent developing countries, the majority of labour that has left the agricultural sector has moved to trade services (wholesale and retail trade; repair of motor vehicles and motorcycles; accommodation and food service activities) and to construction (figure 3.9). In 1995–2017, among all sectors, the trade services sector had the greatest increase in employment share. In 2017 in commodity dependent developing countries, among all services sectors, the trade services sector had the greatest average share of total employment, at 19.3 per cent, and of employment in services, at 43.0 per cent. 33 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 Figure 3.8 Global manufacturing value added (Constant 2015 dollars) 12 1 800 1 600 10 1 400 8 1 200 Trillions of dollars 1 000 6 800 4 600 400 2 200 0 1990 2000 0 2010 2019 Manufacturing value added Manufacturing value added per capita, right axis Source: UNCTAD calculations, based on data from the United Nations Industrial Development Organization manufacturing value added 2020 database. Note: Data for manufacturing value added excludes China, the country with the greatest manufacturing output. Figure 3.9 10 Change in average sectoral employment share, 1995–2017 (Percentage points) 5 Agriculture 0 -5 Mining Manufacturing Utilities Construction Trade services Transport services Financial and business services Other services -10 -15 Commodity dependent developing countries Non commodity-dependent developing countries Developed countries Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. 34 Chapter 3 - Commodity Dependence, Productivity and Structural Change A common feature of the construction and trade services sectors is their position at the lower end of the productivity spectrum not only in commodity dependent developing countries but also in developed countries (figure 3.10). Structural change in the former has disproportionately favoured sectors that appear to have less potential for future productivity growth compared with the manufacturing and other market services sectors. Furthermore, the difference in productivity levels between commodity dependent developing countries and developed countries is lower in trade services than in all other services sectors except other services (non-market services). This limits the potential for productivity gains through convergence effects, which help lower-productivity economies to catch up with higher-productivity economies and appear to be present in many sectors, including in services (International Monetary Fund, 2018). Furthermore, in commodity dependent developing countries, employment shares have shifted largely towards non-tradable sectors in which the potential for future expansion is limited to domestic demand. There are two additional observations with regard to sectoral labour productivity levels in commodity dependent developing countries. First, the sector with the highest median labour productivity level in commodity dependent developing countries is mining. However, the potential of this sector to contribute to aggregate labour productivity growth is limited since it generally does not employ many workers and often operates as an enclave with few linkages to other sectors. For example, in Zambia in 2017, the mining sector accounted for 80 per cent of exports but only 2.2 per cent of total employment. The employment share of mining in member States of the Organisation for Economic Co-operation and Development with large mining sectors, such as Australia and Chile, were in a similar range in 2017, at 1.8 and 2.4 per cent of total employment, respectively. In addition, sectoral differences Figure 3.10 Median labour productivity levels, 2017 (Thousands of constant 2010 dollars) 160 140 120 100 80 60 40 20 0 Agriculture Mining Manufacturing Commodity dependent developing countries Utilities Construction Trade serviceTsranFsinpaonrctisaelravnicdebsusiness services Other services Aggregate Non commodity-dependent developing countries Developed countries Source: UNCTAD calculations, based on data from Dieppe and Matsuoka, 2020, and the UNCTADstat database. 35 COMMODITIES & DEVELOPMENT Escaping from the Commodity Dependence Trap through Technology and Innovation REPORT 2021 between median labour productivity levels in commodity dependent developing countries and developed countries is lowest in the mining sector. This could perhaps be explained by the global presence of large international mining companies that apply similar, capital-intensive technologies at mining sites in different countries. Second, the sector with the second highest median labour productivity level in commodity dependent developing countries is utilities. This sector also does not have the capacity to absorb large numbers of workers. For example, in 2017, the average employment share of the utilities sector in developing countries and developed countries was 0.7 per cent and 1.4 per cent, respectively. These examples show that, while commodity dependent developing countries stand to gain from across-the-board productivity increases, not all sectors have the same potential to absorb large numbers of workers in higher productivity and better paid jobs and thereby generate broad-based development benefits. As shown in this section, commodity dependent developing countries as a group have not followed a path of industrialization since 1995. Instead, the shares of manufacturing in employment and value added have peaked at significantly lower levels than in non-commodity dependent developing countries and developed countries. Structural change in commodity dependent developing countries has been characterized by a shift of employment shares away from the agricultural sector. Since labour productivity in agriculture remains low in these countries, any flow out of this sector results in productivity-enhancing structural change. However, employment shares have moved primarily towards non-tradable sectors at the lower end of the productivity spectrum, which raises questions about the long-term viability of the structural change path. 3.4 Sectoral productivity trends and drivers The results of an empirical analysis of the links between labour productivity, commodity dependence and technological development are presented in this section. Based on the observation that aggregate productivity growth in commodity dependent developing countries is lower than that in non-commodity dependent developing countries, the focus is on the identification of the sources of productivity growth that are stunted in the former and the sectors that are most affected. This requires separating aggregate productivity growth into its two components of intrasectoral productivity growth and structural change, then examining intrasectoral productivity growth in each sector separately. There are different ways of disaggregating economy-wide productivity changes and computing average growth rates over time; the method followed here is that of Diao et al. (2017). The growth rate of economy-wide labour productivity can be disaggregated into the two ∆
关注微信
