Please cite this paper as: Diaz-Bonilla, E., D. Orden and A. Kwieciński (2014), “Enabling Environment for Agricultural Growth and Competitiveness: Evaluation, Indicators and Indices”, OECD Food, Agriculture and Fisheries Papers, No. 67, OECD Publishing. http://dx.doi.org/10.1787/5jz48305h4vd-en OECD Food, Agriculture and Fisheries Papers No. 67 Enabling Environment for Agricultural Growth and Competitiveness EVALUATION, INDICATORS AND INDICES Eugenio Diaz-Bonilla, David Orden, Andrzej Kwieciński JEL Classification: O13, Q10, Q18 http://dx.doi.org/10.1787/5jz48305h4vd-en OECD FOOD, AGRICULTURE AND FISHERIES PAPERS This paper is published under the responsibility of the Secretary-General of the OECD. The opinions expressed and the arguments employed herein do not necessarily reflect the official views of OECD member countries. The publication of this document has been authorised by Ken Ash, Director of the Trade and Agriculture Directorate Comments are welcome and may be sent to tad.contact@oecd.org. © OECD (2014) You can copy, download or print OECD content for your own use, and you can include excerpts from OECD publications, databases and multimedia products in your own documents, presentations, blogs, websites and teaching materials, provided that suitable acknowledgment of OECD as source and copyright owner is given. All requests for commercial use and translation rights should be submitted to rights@oecd.org. tad.contact@oecd.org OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Abstract Enabling Environment for Agricultural Growth and Competitiveness: Evaluation, Indicators and Indices by Eugenio Diaz-Bonilla, International Food Policy Research Institute (IFPRI), David Orden, IFPRI and Virginia Polytechnic Institute and State University, and Andrzej Kwieciński, OECD The key contribution of this report lies in developing a typology to structure the components of the enabling environment for agricultural growth and competitiveness, and in constructing an illustrative Agricultural Growth Enabling Index (AGEI) to summarise a wide array of available information in a coherent manner. The construction of the preliminary AGEI is based on four blocks with 40% of the weight on agriculture/rural factors and 20% each on broader economy-wide governance, capital availability and market operation. The AGEI can be used to provide across-country comparisons or single-country evaluations using the index itself or its components. It allows the decomposition within each main block to show the relative strength and weaknesses of each country across various sub-indices. It has been applied here to a selected set of twenty emerging and developing countries. The preliminary results demonstrate that the AGEI brings together information relevant to the enabling environment for agricultural growth and competitiveness, and which is largely consistent with more in-depth studies of the selected countries. While constrained in some respects, the AGEI appears to be the first index completed with this objective. Further expansion and refinement of the included set of indicators to better reflect key determinants of agriculture’s enabling environment would help provide an important input into better policy decisions. The authors thank Florence Bossard for statistical assistance, and Anita Lari for formatting and preparing this report for publication. This report was declassified by Working Party for Agricultural Policies and Markets in March 2014 (TAD/CA/APM/WP(2013)32/FINAL). Keywords: Enabling environment for agriculture, agricultural growth and competitiveness, typology, agricultural indicators and indices, agricultural performance, determinants of agricultural growth, agricultural policy, agricultural productivity. JEL Classification: O13, Q10, Q18 ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 3 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Table of contents Abbreviations .......................................................................................................................................... 5 Executive summary ................................................................................................................................. 6 1. Introduction .................................................................................................................................. 9 2. Literature review ......................................................................................................................... 10 2.1. Growth and development theory and empirical studies .................................................... 10 2.2. Agricultural growth and development ............................................................................... 15 2.3. Indices of growth and competitiveness determinants ........................................................ 26 3. A typology to identify components of agriculture’s enabling environment ............................... 33 3.1. Structure of the typology ................................................................................................... 34 3.2. Flow of the typology ......................................................................................................... 38 4. An illustrative index of agriculture’s enabling environment ...................................................... 40 4.1. Structure of the preliminary AGEI .................................................................................... 41 4.2. Index analysis .................................................................................................................... 44 4.3. AGEI and agricultural growth ........................................................................................... 53 5. Summary and conclusions .......................................................................................................... 55 References ............................................................................................................................................. 59 Annex A. Enabling environment for agricultural growth: Overview of initiatives ............................... 66 Tables Table 1. Cross-classification of 36 emerging and developing countries by GCI and the World Bank ................................................................................................................... 26 Table 2. Pillars of the Global Competitiveness Index ....................................................................... 29 Table 3. Components of the Global Food Security Index .................................................................. 30 Table 4. Typology of government measures and actions across levels of the economy and selected available indicators ................................................................................................ 36 Table 5. Structure of the preliminary Agriculture Growth Enabling Index ....................................... 42 Table 6. Performance of countries on AGEI, its agriculture/rural areas block and the GCI ............. 45 Figures Figure 1. Indices of factors determining growth ................................................................................ 12 Figure 2. Decision tree for growth diagnostics .................................................................................. 13 Figure 3. Schematic of economic components of agriculture's enabling environment ...................... 34 Figure 4. AGEI and its sub-component blocks (normalised) ............................................................. 46 Figure 5. Disaggregation of governance block (normalised) ............................................................. 47 Figure 6. Disaggregation of capital block (normalised)..................................................................... 48 Figure 7. Disaggregation of markets block (normalised)................................................................... 49 Figure 8. Disaggregation of agricultural/rural blocks (normalised) ................................................... 50 Figure 9. Disaggregation of agricultural/rural blocks - Pillar A (normalised) ................................... 51 Figure 10. Disaggregation of agricultural/rural blocks - Pillar B (normalised) ................................. 52 4 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Figure 11. AGEI and agricultural value added per worker (both normalised) .................................. 55 Annex Figure 1. The structure of the OECD’s economy-wide PMR Indicator ................................. 76 Boxes Box 1. PFIA classification of policy areas ......................................................................................... 67 Box 2. MAFAP proposed classification of public expenditures ........................................................ 70 Box 3. DBA list of indicators ............................................................................................................ 72 Box 4. OECD PSE and GSSE Measures ........................................................................................... 75 ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 5 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Abbreviations ABI Agribusiness Indicators AGEI Agriculture Growth Enabling Index DB Doing Business DBA Doing Business in Agriculture EIU Economist Intelligence Unit EOS Executive Opinion Surveys ES Enterprise Surveys EU European Union FAO Food and Agriculture Organization of the United Nations GCI Global Competitiveness Index GDP Gross Domestic Product GFSI Global Food Security Index GMA Government measures and actions GSSE General Service Support Estimate HDI Human Development Index ICT Information and communications technology IFC International Finance Corporation IFPRI International Food Policy Research Institute ISI Import-substitution industrialisation LAC Latin America Countries MCC Millennium Challenge Corporation MAFAP Monitoring African Food and Agricultural Policies NRA Nominal Rate of Assistance OECD Organisation for Economic Co-operation and Development PFIA Policy Framework for Investment in Agriculture PMR Product Market Regulation PSE Producer Support Estimate R&D Research and Development RRA Relative Rate of Assistance SEA South and East Asia SSA Sub-Saharan Africa TFP Total Factor Productivity UNDP United Nations Development Program WB World Bank WDI World Development Indicators WGI World Governance Indicators 6 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Executive summary This report addresses three objectives. First, it identifies key determinants of economic growth and development, agricultural growth and competitiveness, and the existing indicators and indices by which these determinants have been measured, through an extensive literature review. Second, it proposes a new typology to structure the components of the enabling environment for agricultural growth and competitiveness and uses this typology to assess the types of indicators that would be desirable for inclusion in an index of the enabling environment across countries. Third, it constructs an illustrative, preliminary Agricultural Growth Enabling Index (AGEI) and applies this index to a selected set of 20 emerging and developing countries. While preliminary in many respects, this is, to the best knowledge of the authors, the first completed exercise of its type. Throughout the analysis, a positive enabling environment for agriculture is interpreted to comprise the following:  a multifaceted setting for the sector and economy wide of non-distorting and stable policies  adequate provision of public goods, good governance through laws and regulations that address market failures  strong and effective institutions through which government measures and activities are operationalised. The expected outcome of a positive enabling environment is enhanced agricultural growth and competitiveness driven by well-functioning markets operating in a context of stability and public sector behaviour that is supportive of a forward-looking private economy. A number of key points emerge to determine such growth and competitiveness. Economic growth and development theory and empirical studies suggest that both supply side and demand side issues need to be considered. Although extensive lists of indicators related to government measures and activities and other factors affecting growth can be identified, and may be used as check-lists of things to consider, the main challenge is to analyse the specific constraints that a country faces and work to address them. These constraints vary by country and evolve with time and changing circumstances. For agricultural growth and development, the crucial role of the rural nonfarm economy emerges as the link between agricultural supply and demand. The policy bias for or against the agricultural sector in terms of relative prices and subsidies, but also infrastructure and public services, and the overall circumstances of domestic demand, need to be taken into account. The state of the global economy will also be an important factor. There are several challenges to constructing indices of the determinants of growth or competitiveness. These include:  the choice of appropriate indicators for which to construct such an index  the availability and the quality of data or the cost of developing more adequate data for those indicators  the choice of appropriate normalisation, weights, and aggregation methods by which the various indicators are transformed into the comprehensive index. ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 7 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Within these constraints indicators and indices of the enabling environment are useful to gather quantitative information from different sources, to summarise the situation within a single country, or to allow comparisons across countries. Indices lack the depth of intensive case studies, cluster analysis or formal econometric analysis. Still, they can be utilised to increase public awareness about the current situation regarding different topics and areas and their evolution over time, and to help policy makers to focus on the issues that may require specific attention. The Global Competitiveness Index (GCI) developed for the World Economic Forum is a well-known index of global competitiveness that, acknowledging its limitations, is drawn upon in this analysis. The Global Food Security Index (GFSI) is a new index of global food security that includes a more substantial set of indicators related specifically to agriculture. There are also a number of related recent studies and on-going initiatives about agricultural growth and competitiveness that suggest alternative indicators that could be utilised. The typology developed to identify components of the enabling environment for agricultural growth and competitiveness integrates two main dimensions: the various categories of government measures and activities affecting the sector’s performance and the effects of these measures and activities across four levels of the economy: agricultural producers; the rural/regional economy, which provide the geographical and local governance settings for food and agricultural production; agricultural value chains, which are the market linkages for specific products between inputs and outputs; and the general economy, where, among other things, final demand is determined. The potential indicators of the relative performance of countries across these components are described and the limited availability of some relevant indicators on a comparable basis is appraised. Some of the cells of the typology are relatively well covered from available data and studies, particularly for the general economy and the agricultural sector at the farm level. For rural regions and agricultural value chains there are greater deficiencies of available measurements. This poses a limitation on constructing an index of the enabling environment consistent with the postulated typology of the determinants of growth and competitiveness. The preliminary AGEI is constructed taking note of these considerations. The AGEI is designed to show how an index for agriculture can be constructed to summarise a wide array of available information in a structured manner and then be used to provide across- country comparisons or single-country evaluations using either the index itself or its components. The construction of the preliminary AGEI is presented: it is comprised of four blocks with 40% of the weight on agriculture/rural factors and 20% each on broader economy-wide governance, capital availability and market operation. The countries to which the AGEI is applied were purposefully selected to include those emerging and developing countries which are the focus of OECD country analyses (including selected OECD members defined as emerging economies) supplemented by a range of other countries to provide reasonable geographic coverage. It should be noted that all countries covered are classified as factor and efficiency driven economies according to the GCI groupings. At this stage, innovation driven economies, most of OECD countries, are not included due to their different structural characteristics. The latter are covered by other OECD projects discussing various components of the agricultural enabling environment, in particular those focused on agricultural innovation systems and on green growth in agriculture. The preliminary results demonstrate that the AGEI brings together information relevant to the enabling environment for agricultural growth and competitiveness in a parsimonious 8 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 manner largely consistent with more in-depth studies of the selected countries. Brazil, Chile and China appear relatively strong across the four blocks, but also noticeable is the variability in relative scores among the main blocks for many countries. Even Brazil, Chile and China show this variation and differ in which components account for their high overall scores (governance and capital for Brazil; governance, markets and agriculture/rural for Chile; and capital for China). South Africa scores relatively lower compared to the other countries on capital and much higher on markets. India scores relatively low on the markets and agriculture/rural components compared to its scores on the other two main blocks. Among the countries with lowest AGEI scores, Ethiopia, Pakistan, Senegal and Tanzania score relatively poorly on all four blocks. Egypt scores above the average of the countries on the agriculture/rural block. For a number of other countries, there is quite a mixed set of relative scores: for example, Indonesia, Kenya, Russia and Ukraine each score on at least one main block of the AGEI well above and well below the other countries. The reasons for these results are discussed. Similar decompositions are presented for the indicators within each main block of the AGEI and the relative strengths and weaknesses described among the countries for the various sub-indices and within specific countries across the full set of blocks and their components. The key contribution of this report lies in developing a typology and constructing a preliminary index of the enabling environment for agricultural growth and competitiveness. While constrained in some respects, this initial AEGI appears to be the first index completed with this objective and, overall, provides interesting results. A purpose of this report is to stimulate discussion of the many dimensions of such an index, and of the feasibility and efficacy of how they might be approached in further research and analysis. In short, from this exercise, better indices can eventually be constructed. Expansion and refinement of the included set of indicators will allow further depth of analysis on the determinants of a positive enabling environment to promote agricultural growth and competitiveness, and provide an important input into better policy decisions. Each aspect of this study is exploratory and designed to contribute to the on-going research on agriculture’s enabling environment. On all three objectives of this report, the evaluations presented merit additional consideration. These evaluations include: i) the literature review of existing relevant indicators of the determinants of agricultural growth and competitiveness and of the recent and ongoing studies and initiatives to extend this database; ii) the elaboration of the conceptual typology for evaluating the enabling environment; iii) the matching of indicators to this typology; iv) the specification of an index characterising the enabling environment in each of its dimensions; v) the set of countries to which such an index is applied; and vi) ultimately, verification that the specified index accomplishes its intended purpose by correlating with observed growth. ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 9 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 1. Introduction In this report a positive enabling environment for agricultural growth is interpreted to comprise a multifaceted setting for the agricultural sector and economy wide of non- distorting and stable policies, adequate provision of public goods, good governance through laws and regulations that are conducive to private-sector economic activity while addressing market failures, and strong and effective institutions through which government measures and actions (herein, GMAs) are operationalised. The expected outcome of a positive enabling environment is enhanced agricultural growth and competitiveness driven by well-functioning markets operating in a context of stability and public sector behaviour that is supportive of a forward-looking private economy. Competitiveness is understood as the capacity of agriculture to grow and thrive in domestic and world markets without the support of those public policies that are considered market distorting. Although this report focuses on the enabling environment for agricultural growth and competitiveness, the analysis is placed within the context of related agricultural development objectives such as poverty alleviation, food security and nutrition, prosperity of small farmers, social equity, productivity growth and environmental sustainability in agriculture. This report is organised into five sections. The current Section 1 introduces basic definitions and provides an overview of the content of the report. In Section 2, a three-part review is provided of the literature on growth and development theory and empirical studies (Section 2.1), agricultural growth and development (Section 2.2), and indices of growth and competitiveness determinants, which includes a review of existing indices and the issues that arise in their construction, related recent studies, and ongoing or proposed initiatives to classify, select indicators and construct indices of the enabling factors for economic competitiveness, food security and agricultural growth and competitiveness (Section 2.3). Building on the literature review, Section 3 develops a typology to identify components of the enabling environment for agricultural growth and competitiveness. This typology link together two main dimensions of agriculture’s enabling environment: various categories of GMAs affecting the sector’s performance and the effects of these measures across four levels of the economy: agricultural producers; the rural/regional economy, which provide the geographical and local governance settings for food and agricultural production; agricultural value chains, which are the market linkages for specific products between inputs and outputs; and the general economy, where, among other things, final demand is determined. The potential indicators of the relative performance of countries across these components are discussed and a summary is provided of some of the existing indicators that could be used to construct an index of agriculture’s enabling environment. The limited availability of some relevant indicators on a comparable basis across countries is discussed. In Section 4, an illustrative index of agriculture’s enabling environment (a preliminary Agricultural Growth Enabling Index, AGEI) is constructed along the lines of the proposed typology and drawing on a selected subset of the available indicators both for the general economy and specific to agriculture. This preliminary index is applied to a sample of 20 emerging and developing countries, their performance on the index and its components are evaluated, and the relationship of the index to observed agricultural growth is examined. Section 5 provides a summary and conclusions from the report. The contribution of the report lies in pushing through to completion an exercise that, starting with a typology of GMAs built on the literature review (and which could also serve as a checklist for policy analysis), develops a preliminary index of the enabling environment for agricultural growth 10 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 and competitiveness. While each aspect of the study is exploratory, the purpose of the report is to stimulate discussion of the many dimensions involved in the typology and the proposed index. Annex A provides an overview of various initiatives that are trying to identify, classify and measure different GMA issues related to agricultural growth and competitiveness, beyond those initiatives which are discussed in detail in the main body of the report. 2. Literature review The literature on GMAs addressing the wide range of policies, public investments and other expenditures, laws, regulations, institutions and market performance that support economic growth and agricultural growth and competitiveness is very large. This section starts from a brief overview of literature on the determinants of economy-wide growth and development (Section 2.1) within which agricultural growth and development is discussed in a greater detail (Section 2.2). It is followed by a discussion on the Global Competitiveness Index (GCI) of the World Economic Form and the Global Food Security Index (GFSI) developed recently by the Economist Intelligence Unit (EIU) that can be drawn upon to partly characterise the enabling environment for agriculture in quantitative terms. Other studies and data available for economic enabling environment indicators are also discussed, and some recent initiatives from OECD, FAO, the World Bank and IFPRI are described that are identifying, classifying, quantifying, and ranking different GMAs related specifically to agricultural growth and competitiveness (Section 2.3). Although there are important overlaps in those exercises, they identify a multiplicity of determinants of the enabling environment for agriculture, classify them in different ways, and suggest different indicators and indices. 2.1. Growth and development theory and empirical studies Growth theory and regressions Starting with the Solow-Swan model it has been clear that the proximate causes for economic growth are factor accumulation (basically capital, depending on savings and investment, and labour, depending on demographics and human development) and productivity (a combination of available technologies and the efficient use of them) (Weil, 2005). This, however, does not answer what are the fundamental causes that lead to factor accumulation, technological development and efficiency. Adam Smith provided an early answer when he argued that “Little else is requisite to carry a state to the highest degree of opulence from the lowest barbarism but peace, easy taxes, and a tolerable administration of justice: all the rest being brought about by the natural course of things.” Other schools of thought envisioned a more active involvement of the state in the economy, a debate that still reverberates in modern analyses of growth and development policies. Quantitative efforts to test the importance of the more fundamental causes of growth started in the late 1960s and early 1970s (Robinson, 1971; Adelman and Morris, 1967), but the work on growth determinants exploded during the 1980s and 1990s (see Barro and Sala-i-Martin, 1998). The growth equations derived from the Slow-Swan model implied a relationship between the rate of growth, the (natural logarithm) of level of income per capita at the beginning of the period analysed, and the (natural logarithm) of the steady state level of income per capita as captured by a set of variables postulated as growth determinants. There has been a large literature on the growth regressions, their results, and pitfalls. Sala-i-Martin (2002) summarises the early results as follows: “(i) There is no simple ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 11 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 determinant of growth; (ii) The initial level of income is the most important and robust variable (so conditional convergence is the most robust empirical fact in the data); (iii) The size of the government does not appear to matter much. What is important is the quality of government (governments that produce hyperinflations, distortions in foreign exchange markets, extreme deficits, inefficient bureaucracies, etc. are governments that are detrimental to an economy); (iv) The relation between most measures of human capital and growth is weak. Some measures of health, however, (such as life expectancy) are robustly correlated with growth; (v) Institutions (such as free markets, property rights and the rule of law) are important for growth; and (vi) More open economies tend to grow faster.” These broad conclusions, and the policy recommendations they imply, although generally accepted, have generated some controversies as well. It has been argued that rather than distilling robust results about what constitutes good policies, the strongest conclusion of the literature has been the identification of extremely bad policies that impede growth (Easterly, 2003). Also, while early work on human resources focused on years of schooling (a quantitative variable) and did not find much correlation with economic growth (Sala-i-Martin, 1997), more recent empirical work using better proxies for the quality of education (such as comparable international tests on cognitive achievement) has found stronger links to growth (Hanushek and Wößmann, 2007; Aghion, 2009). Similarly, the adequacy and interpretation of proxies used in growth regressions for “openness,” “outward orientation,” and “globalisation,” have been questioned. For example, Birdsall and Hamoudi (2002) showed that the positive correlation reported by Dollar and Kraay (2001) between growth and globalising economies is related to the fact that the countries performing worse were commodity dependent, and the collapse in commodity prices in the 1980s and 1990s reduced both growth and the value of the variable interpreted as a proxy for openness, creating a misleading correlation. Among macroeconomic variables, while the negative impact on growth of inflation appeared weak (perhaps because of non-linearities and threshold effects), indicators of overvaluation of the exchange rate were clearly associated with low growth and economic crises (Díaz-Bonilla and Robinson [2010] present a review of the related literature). While the previous examples show some variations with the early conclusions summarised by Sala-i-Martin (2002), other aspects have been reinforced and expanded. For, instance, the importance of institutions, a theme with a large tradition in classical and development economics, has been reaffirmed by subsequent growth analysis. These institutions include political institutions (democracy, political freedom, regulation of conflict and distribution), legal institutions (property rights, the rule of law), market institutions (market structures, competition policy, international openness), governance institutions (the size of bureaucracy, government corruption), and other institutional aspects (see Adelman and Morris [1967] for an early assessment and Acemoglu and Robinson [2012] for a recent synthesis). Another topic that has received further attention is technology and innovation. While the Solow-Swan model considered technological change as an exogenous factor, more recent models within the new growth theory have proposed different approaches that look at endogenous innovation, including Schumpeterian models that involve creative destruction (Romer, 1990; Aghion and Howitt, 1998). These models suggest that innovation and productivity are fostered by “(i) better protection of (intellectual) property rights, as this will improve the extent to which successful innovators can appropriate the rents from their innovation; (ii) better financial development, as tight credit constraints will limit individuals’ ability to finance a new innovative idea; (iii) a higher stock of educated labour, as this will improve individuals’ ability to imitate more advanced technologies or to 12 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 innovate at the frontier building on giants’ shoulders; and (iv) macroeconomic stability: by ensuring low (risk‐adjusted) equilibrium interest rates, it will encourage individuals to engage in long‐term growth‐enhancing investments” (Aghion and Durlauf, 2009). Other aspects of structural conditions and GMAs, such as the size of the market and market competition, are also crucial variables. One of the implications of these theories, Aghion and Durlauf (2009) argue, is that the best policies and institutions to foster innovation, growth and productivity may take different forms depending on whether countries are catching up and are still far from the world technological frontier, or whether they need to innovate because they are close or at that frontier. For instance, countries in the latter stage may need more product market competition and entry, stronger college education, equity (as opposed to loan) finance, and more democracy and decentralisation, compared to the countries in the former stage. The list of real or postulated growth determinants and advances in data collection and processing have led to the creation of indices of different types to try to measure those determinants. For instance, the Global Competitiveness Index (GCI) (Sala-i-Martin et al., 2013) collects data on different variables under 12 pillars. These are grouped in three main blocks, which are applied to countries under three groups (factor-driven, efficiency-driven, and innovation-driven), as characterised in Figure 1. Figure 1. Indices of factors determining growth Source: Sala-i-Martin et al., 2013. Growth diagnostics A different line of empirical analysis, also emerging from a variation of the Solow- Swan model, but reacting to what is considered a “laundry-list” approach to growth, has highlighted the need to look at the specific constraints that may be affecting a country’s economic performance, in what has been called growth diagnostics (Hausmann, Rodrik, and Velasco, 2005). It is argued that growth regressions, even if they solve all the Pillar 1. Institutions Pillar 2. Infrastructure Pillar 3. Macroeconomic environment Pillar 4. Health and primary education GLOBAL COMPETITIVENESS INDEX Pillar 5. Higher education and training Pillar 6. Goods market efficiency Pillar 7. Labor market efficiency Pillar 8. Financial market ooooo development Pillar 9. Technology readiness Pillar 10. Market size Pillar 11. Business sophistication Pillar 12. Innovation Key for factor-driven economies Key for innovation-driven economies Key for efficiency-driven economies Basic requirements subindex Efficiency enhancers subindex Innovation and sophistication factors subindex ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 13 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 econometric specification problems, would provide a list and ranking of policies that work for the “average country;” but then, individual countries are not averages, and they suffer from specific constraints and face unique opportunities. Growth diagnostics starts from neoclassical growth models where the rate at which the economy grows is a function of the difference between the expected return to asset accumulation and the cost of those assets as perceived by the private economic agents. Higher growth depends on the incentive to accumulate, which is the difference between the proportion of social economic returns that is privately appropriable (less than gross returns if there is a tax on earnings, but also driven by many other things), and the opportunity cost of funds. Growth diagnostics builds a decision tree starting from whether the returns are high or low, whether they can be appropriated, and the cost of financing. By working down the tree it identifies a series to reasons that may be constraining growth, as shown in Figure 2. Figure 2. Decision tree for growth diagnostics Source: Hausmann, Rodrik, and Velasco, 2005. The objective of growth diagnostics is to determine for specific countries what are the main constraints utilising an explicit and structured approach (which is not based on growth regressions), and then to suggest policies and other measures aimed at lifting those constraints. However, Aghion and Durlauf (2009) have argued that growth regressions can be a better way of identifying the growth constraints than the decision tree approach utilised by Hausmann, Rodrik and Velasco, 2005. Low domestic saving Problem: Low levels of private investment and entrepreneurship Low return to economic activity High cost of finance Low social returns Low appropriability Bad international finance Bad local finance Poor geography Low human capital Bad infrastructure Government failures Micro risks: property rights, corruption, taxes Macro risks: financial, monetary, fiscal instability Market failures Information externalities: «self discovery» Coordination externalities Poor inter - mediation 14 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 Demand side issues Most of the analysis mentioned so far works from the supply side, assuming that demand is always there to absorb the products generated (a form of Say’s Law) and that factors of production are fully employed, even though there are important variations in demand conditions. Robert Solow (2005), one of the originators of the basic Solow-Swan growth model, has acknowledged the omission of demand considerations as a weakness of the growth theory based on this model. Recognising the importance of the demand side for growth and the possibility of unemployed factors, Kaldorian and Keynesian growth models (see for instance, Setterfield, 2010) consider autonomous demand (such as exports), as well as income distribution between wages and profits, as factors that determine aggregate demand and therefore growth. In demand-led endogenous growth models technological change is also linked to demand side issues. For developing countries, the state of the global economy is an important determinant of the internal growth conditions in those countries, not only because of the impact on the demand side via trade flows, but also through more complex demand and supply interactions related to capital flows, technology diffusion, real interest and exchange rates, and migration. Growth regressions that do not control for the state of the world in these terms may provide inadequate advice on policy interventions and other domestic variables considered to impact growth. Pro-poor growth In development analysis growth is often linked to the need to improve the standard of living of significant sectors of the society, particularly the poor. More than forty years ago, Little, Scitovsky, and Scott (1970) and Balassa and Associates (1971) argued, among other things, that the import-substitution industrialisation (ISI) strategy followed then by many developing countries was, due to policy distortions, excessively capital-intensive (which slowed employment growth in industry) and limited the development of agriculture. Both effects had negative implications for poverty alleviation. Chenery et al. (1974) presented the case for a growth and investment programme centred especially on accumulation of human and physical capital by the rural poor. Separately, a basic-needs approach to poverty also emerged in the late 1970, arguing that objectives such as growth, or even employment and income redistribution, were means to the more concrete objective of attending to the needs of the population (defined primarily by advances in health, education and other indicators of human development, especially for the poor and vulnerable). The basic-needs approach implied an important role for the public sector in the provision of certain public services and improvements in access so as to effectively reach the poor. It also promoted organisation of the population that was to receive the services and their participation in the decisions and actions to be implemented (Streeten and Burki, 1978). After a period during the 1980s, in which macroeconomic stabilisation and structural reforms were the focus of growth policies, in the 1990s concerns about slow or no progress in poverty reduction in many developing countries led to an emphasis on pro-poor growth, as something different from growth alone calculated using the average of per capita income (World Bank, 1990). The most common notion was that growth was pro-poor if the poor benefitted the same or more than the non-poor population. Although there have been a series of statistical difficulties and differences in how to operationalise this concept, one analytic result is that the impact of growth on the reduction of poverty depends inversely on indicators of income or asset inequality (Ravallion, 2004). While overall economic growth remains a central factor for poverty reduction, the sectoral composition of growth seems to ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 15 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 matter with agricultural growth appearing to be more pro-poor than growth in other sectors in developing countries. This reflects the reality that a large proportion of the poor depend on agriculture, thus accelerated growth in this sector “is likely to disproportionally benefit the poor” (Dollar et al., 2013). Recent synthesis The latest large-scale effort to summarise the analysis of successful growth strategies has been the Commission on Growth and Development (2008, 2010), also known as the Spence Commission. The Spence Commission argues that countries with successful growth stories have:  fully exploited the world economy  maintained macroeconomic stability  showed high levels of investment, private and public, with high rates of domestic saving  respected market signals in general (although not absolutely at times) and allowed structural changes and labour mobility, protecting laid-off workers but not maintaining unviable industries, companies or jobs  have governments committed, credible, and capable of providing a range of public goods, offering a vision of the future that justified today’s efforts, and that tried to ensure that opportunities and benefits were shared widely. The Commission (2010) classifies policies into five broad categories: accumulation, innovation, stabilisation, allocation, and inclusion. It also warns that every country must tailor policies to their respective conditions, because “a list of ingredients does not make a recipe.” 2.2. Agricultural growth and development Agricultural growth and structural change The Solow-Swan model and many variations of it analyse a one-sector economy. That is the case of most of the demand-side models as well. To analyse agricultural growth as a component of the aggregate economy, it is necessary to consider more than one sector. There are some examples of both types of growth models that disaggregate the economy into agriculture and non-agriculture. For example, Acemoglu (2009) shows in a neo- classical model with a consumption specification that follows Engel’s law that the agricultural sector grows at a lower rate than industry. Thirlwall (1986), in a Kaldorian model, shows that agriculture is an important demand factor for industry. Therefore, it is necessary to consider more than one sector to be able to analyse not only growth but also development. The latter involves, among other things, the notion of structural changes, both in the composition of employment and production. An important part of the process of development is the shift of employment and production from agriculture to manufacturing, and then from manufacturing to services (Acemoglu, 2009). While accelerations in agricultural growth appear at the beginning of most successful cases of development (starting with the Industrial Revolution in the 19 th century), later an important part of the process of development is the shift of employment out of agriculture. More specifically, Barrett, Carter, and Timmer (2010) (following Timmer, 1988), in their review of one hundred years of agricultural development literature, note that 16 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 agriculture evolves through four stages when long historical periods are considered. They label these four stages as “the “Mosher” stage when getting agriculture moving is the main policy objective (Mosher, 1966); the “Johnston-Mellor” stage when agriculture contributes to economic growth in the rest of the economy through a variety of linkages, such as supplying labour and raw materials to industry, providing food for industrial workers, expanding markets for industrial production, and by the generation of foreign exchange through exports (Johnston and Mellor, 1961); the “Schultz” stage when rising agricultural incomes fall behind those in a rapidly growing nonfarm economy, inducing political tensions (Schultz, 1978); and the “Johnson” stage where labour and financial markets fully integrate the agricultural economy into the rest of the economy (Johnson, 1997). Individual countries may follow different paths, but a constant of the process of structural change is that agriculture‘s share in employment and production declines, in good measure linked to the declining percentage of food demand as incomes increase. In his context, policies to enhance an enabling environment for agricultural growth must ensure, at various stages of development, that structural change takes place without distorting incentives that may accelerate the movement of labour out of agriculture (such as the case of turning incentives against agriculture) or slowing it down (by shifting incentives in favour of the sector). Price policy biases against or supportive of agriculture Some of the ideas of linkages from agriculture to the rest of the economy were embedded in the notion of the agricultural sector as a basis for support of the strategy of inward oriented ISI that was attempted by a variety of developing countries after the end of World War II. The particular setting of incentives of the ISI led to various criticisms, including to what was considered its anti-agricultural policy bias. Several studies (Little, Scitovsky and Scott, 1970; Balassa and Associates, 1971; Krueger, 1978) pointed to the supply-side constraints generated under the ISI policies by the resulting macro structure reflected in two relative prices (the tradables/nontradables price (essentially, the real exchange rate) and the relative price of industrial products to agricultural products, reflecting tariffs and other market interventions). According to these studies the policies adopted had a triply damaging effect: i) they made the economy operate within the production possibility frontier; ii) they led to a composition of total production that did not allow the country to benefit from international trade; and iii) they slowed the outward growth of the productive possibility frontier (or productive potential) of the country. This overall critique was followed by sector-specific studies (mostly covering the period from the 1960s to the mid-1980s) that analysed the direct and indirect effects of trade, exchange rate, and other macroeconomic policies on price incentives for agriculture (Krueger, Schiff, and Valdés, 1988). The analysis focused on the production incentives provided to agricultural products by the policies implemented. This analysis found that agricultural importable goods were generally protected while exportable ones were taxed. However, once the indirect effects of overvalued exchange rates and industrial protection were considered, there was a negative price bias against agriculture that affected incentives and the performance of the sector. The policy recommendation was to eliminate inefficient industrial protection, to avoid the overvaluation of the exchange rate, and to phase out export taxes on agriculture. At the same time, it was considered that sectoral interventions that supported and subsidised agriculture should also be substantially revamped and scaled down, given that overall incentives would shift in favour of agriculture with the change in the general macroeconomic and trade framework (World Bank, 1986). ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 17 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 The elimination of a general price and macroeconomic bias against agriculture became one of the goals of policy reform strategies, including structural adjustment programmes, supported by the World Bank and others international institutions, and many countries undertook such reforms in the 1990s. This price bias, however, was different from a more general urban bias discussed by Lipton (1977), which also included the allocation of public investment and expenditures, and other policies. He argued that the poor remain poor in developing countries because public expenditures and economic policy in general (not only relative prices), benefitted urban groups who were better positioned to pressure governments to defend their interests, while rural population were short-changed. The partial-equilibrium calculations on policy bias against agriculture by Krueger, Schiff, and Valdés (1988) and others have been criticised for, among other things, the use of nominal instead of effective rates of protection and the ad-hoc estimation of the exchange rate misalignments. Jensen, Robinson, and Tarp (2002), for example, using general equilibrium models for the same countries, concluded that the partial-equilibrium measures used in earlier studies tended to overstate the price bias against agriculture. Alternatively, it could be argued that whatever the previous bias had been, it had been reduced or eliminated during the 1990s, through the changes in exchange rate, fiscal, monetary, and trade policies resulting from the structural adjustment programmes implemented by many developing countries. Furthermore, looking at levels of import tariffs since 2000, and contrary to the conventional assessment in the late 1980s, agriculture (considering both primary and processed products) seems, on average, more protected than industry in developing countries (Díaz-Bonilla and Robinson, 2010). Recent estimates of the nominal rate of assistance (NRA) for agriculture (Anderson and Valenzuela, 2008) show that such assistance has been growing in developing countries, turning positive since the mid-1990s. The increased NRAs in those countries have been both the result of more protection for importables (i.e., a growing NRA > 0) and less taxation for exportables. Likewise, the relative NRAs for agricultural and non-agricultural goods (what Anderson and Valenzuela (2008) call the relative rate of assistance, or RRA), which in their calculations showed a significant bias against agriculture during the 1960s through the early 1980s, has been moving since then in favour of the agricultural sector, turning positive in the late 1990s and early 2000s. Therefore, not only does whatever policy bias existed before seem to have been eliminated, but those indicators suggest there is often now positive relative support for the agricultural sector in many developing countries. In either case, measurement of the enabling environment for agriculture needs to consider relative price incentives for the sector using adequate indicators. It should also be recognised that analysing only relative-price biases may leave out important determinants and aspects of the performance of the agricultural sector in emerging and developing countries. Demand conditions and linkages Several studies have focused on agricultural growth multipliers for the rest of the economy, i.e. how much overall GDP was generated by expanding agricultural GDP (see for example, Haggblade and Hazell, 1989). These analyses usually found positive and large multipliers, showing the importance of agricultural development for the economic dynamism of the rest of the economy. However, considering that the focus of this report is the evaluation of the enabling environment for agricultural growth and competitiveness, the discussion that follows, while acknowledging the importance of agricultural growth and development for the whole economy, focuses on the reverse linkages from the rest of the 18 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 economy to agriculture, which are most germane for the discussion of what enables growth and competitiveness in the agricultural sector. In a recent OECD review of agricultural policies for poverty reduction, it is stressed that “many of the policies required to improve farmers’ opportunities are non-agricultural.” They include improvements in education, primary healthcare and in overall investment climate, which “depends on factors such as peace and political stability, sound macroeconomic management, developed institutions, property rights and governance” (OECD, 2012). In the FAO review based on consultations on the enabling environment for agribusiness and agro-industry development, Konig et al. (2013) argue that “policies and strategies that aim to increase agro-based investments must not only emphasise business climates, but also consider the elements that affect investment profitability and, in consequence, investors’ perceptions of risk-to-return ratios.” This recognises that the notion of business climate perceived as operating only on the supply side may be missing crucial aspects of a general enabling environment. In particular, as in growth models, demand conditions need to be considered. For example, Orden, Paarlberg and Roe (1999) and Gardner (2002) include among the underlying causes of US agricultural growth in the 20 th century, the economic growth in the non-farm economy, which provides the demand for agricultural products. Also, Barrett, Carter, and Timmer (2010), after discussing the impact of agriculture on the growth performance of the rest of the economy, refer to the reverse link when they note that “unless the non-agricultural economy grows, there is little long-run hope for agriculture.” In terms of an enabling environment for agriculture, even if overall growth occurs, a problem is that the two-way linkages between agriculture and the rest of the economy may not occur automatically without adequate governmental policies, investments and institutions. For example, in several African countries the urban demand side is not well linked to the potential supply side that exists in the domestic economy. Barrett, Carter, and Timmer (2010), when discussing those linkages, note that “the rural nonfarm sector provides the bridge between commodity-based agriculture and livelihoods earned in the modern industrial and service sectors in urban centres… The firms and activities in the rural nonfarm sector mediate many of the two-way linkages between agriculture and the macroeconomy that are at the core of the development process.” Therefore, two important enablers of growth in the agricultural sector are: i) sustained growth on the demand side (i.e. growth in the non-agriculture economy and exports) and ii) the strengthening of the rural nonfarm sector and the value chains that link agricultural supply with demand (see for instance, Haggblade, Hazell, and Reardon, 2007). Economic policies that ensure trade opportunities as part of the demand for agricultural products would include, for instance, properly managed exchange rates, which avoids overvaluation and excessive volatility. At the same time, as the larger percentage of agricultural production in many developing countries is consumed locally, that domestic demand depends on the overall functioning of the economy. In consequence, general macroeconomic and other policies that maintain sustainable growth of aggregate demand in line with potential aggregate supply, that ensures inclusive and broad-based growth, and that avoids economic crisis, would be part of the overall enabling environment for agricultural growth. It should also be noted that the composition of external and domestic demand may vary by product, and generate different growth paths. For instance, Diao, Dorosh and Rahman (2007) analyse the demand-side conditions for agricultural growth in East and Southern Africa and find that an export-led agricultural growth strategy may not generate substantial overall income growth, while increasing productivity and production of staple foods ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 19 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 supports higher growth in agriculture provided that there is rapid growth in the nonfarm economy (which creates demand) and that marketing costs are reduced (which allows African producers to supply that demand). Public spending and agricultural growth In terms of agricultural growth determinants, a literature over the past decade (building on earlier contributions) has focused on the impacts of public investments and other expenditures at relatively broad levels on the growth of agricultural output among developing countries. Benin, Fan and Johnson (2012) summarise the approach. Agricultural growth depends on advances in total factor productivity (TFP), which is heavily affected by public expenditures on agricultural research and development (R&D), and the level of utilisation of input factors, which reflects private-sector investment and allocation decisions given input and output market signals. These two growth determinants, in turn, are affected by direct, indirect and interaction effects (including the crowding-out or crowding-in effects of the public expenditures on private spending) from various forms of public expenditures on agriculture and the non-agricultural economy, and by the effects of non-expenditure factors affecting TFP and markets. In the public spending assessment studies (see, for instance, Mogues and Benin, 2012; Mogues et al., 2012; Fan, 2008), there are econometric estimates of the effects of different forms of public agricultural spending (e.g. research, irrigation, conservation) on agricultural factor use (e.g. use of labour, capital and intermediate inputs) and agricultural output and productivity (e.g. output per capita, per worker, per hectare, TFP). Likewise, econometric estimates are drawn upon to measure the effects of non-agricultural public spending (education, health, roads, electrification, rural development, telecommunications) on these growth-related variables. With these estimates, simulation analysis can be carried out to assess the impacts of alternative choices and time-paths of public outlays. Some of the results of this literature are as follows, as summarised from the extensive review of existing studies by Mogues et al. (2012):  First, aggregate public spending on agriculture seems to have moderate or even modest returns on rural welfare, agricultural growth, economic growth, or poverty reduction. This is the result of considering together public expenditures that seem to have important positive impacts with others that do not have positive impacts or even lead to negative outcomes. The implication is that it is important to analyse different types of public expenditures in and for agriculture.  Second, a consistent result across a large number of studies is that returns to agricultural R&D expenditures are positive and substantial for agricultural productivity and growth. Four-fifth of the 120 reported internal rates of return to expenditures on agricultural research are greater than 20%, and two-fifth are between 20 and 60% (Evenson, 2001), although returns vary by regions and products (Alston, 2010). Estimates for China, for example, show that one monetary unit spent on R&D yields more than 6.5 monetary units of agricultural GDP, and in the case of India during the 1990s the result is between 9-10 monetary units of agricultural GDP. Given those results, it appears that there is a substantial underspending on agricultural R&D. 1 1. Goñi and Maloney (2014), while recognising the importance of R&D for technological catch up, argue that the rates of return to R&D follow an inverted U: they rise with distance to the technological frontier and then fall thereafter, potentially even turning negative for the poorest countries. This might be due to the weakness of such factors as education, the quality of scientific 20 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014  Third, in several countries where comparisons have been attempted, public spending on agricultural R&D outperforms other public expenditures in agriculture, such as extension, irrigation, and fertiliser subsidies, in terms of raising agricultural productivity. It also seems to outperform other spending on agriculture, such as rural road infrastructure, education, electrification, health, and telecommunications, although several of the latter spending also have positive impacts on agricultural growth and productivity. Agricultural extension expenditures show relatively high returns and irrigation appears a positive investment in some countries, but not in others. The estimates cited by Mogues et al. (2012) for China indicate that one monetary unit spent on general education yields more than 2 monetary units of agricultural GDP; for roads and telephones, between 1.5 and 1.7; and for irrigation and electricity, less than 1.5. Fertiliser subsidies, at least as measured in India, rank last out of eight different types of agricultural and non-agricultural spending in terms of their contribution to agricultural productivity. For India the cited studies indicate expenditures of one monetary unit in roads result in 7-8 units of agricultural GDP, education between 5-6 units, irrigation, 4-5 units, and credit subsidies about 4 units, whereas a monetary unit expenditure on fertiliser and power subsidies led to just 1 unit or less of agricultural GDP.  Fourth, public expenditures on agricultural R&D appear to also rank high among the most effective interventions to reduce poverty (although not always as the first option, as the case for raising agricultural productivity). Estimates summarised in Mogues et al. (2012) calculate than 1 million monetary units in China led to 1 200 fewer poor people when invested in education, while R&D was a close second, leading to a 1 000- reduction in the poverty headcount. The same calculations for India, based on one million monetary units, were 120 fewer poor people due to spending in roads, and 80 fewer due to R&D (the second largest impact); in Thailand, it was more than 250 fewer poor people for electricity and between 100-150 for R&D (also the second largest impact); and finally, in Uganda, this monetary level of R&D expenditure led to a reduction of almost 60 in the poverty headcount (the largest impact) and feeder roads ranked second with a reduction of about 30-40 person in the poverty headcount. It should be noted that 1 million monetary units may have different purchasing power in the countries mentioned, therefore comparisons of the impact of expenditures on the amount of poverty reduction across countries, as opposed to between alternative expenditures within countries, are not valid. Overall, the results suggest that there is no (or minimal) trade-off between growth and poverty alleviation policies in regards to spending on agricultural R&D.  Fifth, certain agricultural expenditure (such as biofortification) have positive impacts on health and nutritional outcomes, and are identified as highly cost-effective (Meenakshi et al., 2010).  Sixth, returns to public expenditure on and for agriculture have been declining over time, with the exception of agricultural R&D. For instance, in the studies cited by Mogues et al. (2012), the use of fertiliser subsidies in India had a larger impact in previous decades than more recently: in the 1960s-1970s 1 monetary unit spent on irrigation subsidies led to about 4-6 monetary units of agricultural GDP (lower than R&D which yielded between 8-10 units), but in the 1980s and 1990s the impact declined to between 2-3 monetary units of agricultural GDP (while the impact of infrastructure, the overall functioning of the national innovation system, the quality of the private sector, all necessary to complement R&D, but becoming “increasingly weak with distance from the frontier and the absence of which can offset the catch up effect”. ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 21 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 expenditures on R&D stayed at 8 or more). These results suggest that there may be a start-up effect, but that this impact declines over time. The consideration of the time dimension is important because the short and long term effects of public expenditures seem to be different.  Seventh, besides the type of public expenditure it may be important to consider the geographical dimension of the expenditures. Mogues et al. (2012) offer the tentative conclusion that there may be greater returns per dollar of expenditure in less-favoured areas than those in high-potential areas, both in terms of poverty reduction and, more controversially, for agricultural performance. They caution that further empirical analysis is needed on these questions. In any case, the effects of government expenditures on agricultural development are usually heterogeneous depending on geographic areas, which highlights the importance of the co-ordination between central and sub-national governments to define what expenditures are needed, where are they needed, and whether the level and composition of public resources applied in a region are adequate for the goals defined.  Eighth, the evidence is mixed on whether aggregate public expenditures on agriculture “crowd-in” private investments (Mogues et al., 2012). In any case (as shown in the country data compiled by Lowder, Carisma and Skoet, 2012) private investments are far larger than public ones, crowded-in or not.  Ninth, the findings from different studies should be interpreted as results at the margin (i.e. the impact of an additional unit of expenditures). Therefore, they cannot be utilised to justify or assess actual or proposed large changes in the structure of expenditures, such as significantly reducing funding of an activity, or allocating substantial new resources to another, which would dramatically change the impacts of those resources. Also, Mogues et al. (2012) caution about the extrapolation to other circumstances of the conclusions that emerge from certain regions or time periods in the studies they reviewed. The context and application of the conclusions of those results need to be carefully considered.  Tenth, public expenditures have opportunity costs, not only in terms of alternative uses of those funds, but related to how those expenditures are financed, such as taxes, borrowing, and money creation. Therefore, it is important to be able to justify public expenditures focused on agriculture. Usually the reasons for public interventions are related to the presence of some sort of market failure or to distributional concerns.  Eleventh, as illustrated in the preceding points, the studies reviewed in Mogues et al. (2012) show that agricultural expenditures can help with outcomes in other areas (such as health), while expenditures not directly aimed at agriculture, including energy, rural roads, education, and so on, have strong impacts on agricultural growth and productivity. Therefore, there is a need to co-ordinate across ministries and agencies, share information about the amount and characteristics of the public expenditures and their cross-sectoral effects, and improve the allocation of resources to achieve multiple development goals. The co-ordination imperative and the geographical dimension also call for a better understanding of the differential impacts of public expenditures at the national, provincial, and local level, including the impact of decentralisation in public sector expenditures and interventions in and for agriculture. Pro-poor growth, and food security Theodore Schultz (1979) famously argued that by understanding the economics of agriculture one could know much of the economics of being poor, which in turn was much 22 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 of the economics that really matters. Since then, several empirical studies have evaluated the impact of agricultural growth on poverty reduction, on income growth in different quintiles, and on similar indicators. An objective of those studies was to determine whether agricultural growth was pro-poor (Eastwood and Lipton, 2000; OECD 2006). The results in general supported the notion that agricultural growth not only is pro-poor in reducing poverty or increasing more the income of the lower quintiles of the income distribution, but also appears to have larger impacts on poverty reduction than growth in other sectors. The channels through which agricultural growth helps reduce poverty, which are also relevant for food security, are i) an increase in farms income, ii) more employment opportunities in rural areas, iii) declines in food prices for net buyers, and iv) general multiplier effects on the rest of the economy from agricultural growth and demand. The exceptions to these results appeared in developing countries with large inequalities in land holdings where agricultural growth appeared uncorrelated with poverty reduction (Eastwood and Lipton, 2000). Also, the correlation weakens with increases in a country’s income (i.e. in richer countries, agricultural growth does not have stronger impacts on poverty reduction when compared with other sectors). The issue of ensuring that agricultural growth remains pro-poor is related to the dilemma in developing countries between pursuing growth and production, usually concentrating support on larger, modern agricultural units, or emphasising poverty reduction and food security with a focus on small farmers, landless rural workers, and other vulnerable groups. This dilemma has many facets, including the possibility of complex two-way influences, such as whether more equal societies have higher and more stable rates of growth than their more unequal counterparts (Alessina and Perotti 1996; Deininger and Squire, 1997). Others have noted the positive impacts of an agrarian structure based on family farms on the emergence of democratic governance (Moore, 1967) and on the formation of larger domestic markets that allow the development of industry and other activities. The environmental sustainability of a strategy based on large commercial farms, versus a strategy focused on small-scale agriculture, has also been amply debated. Agricultural growth, and in particular pro-poor agricultural growth, is also important for food security for a number of reasons (see, for instance, Díaz-Bonilla et al., 2003; OECD, 2013a). The common definition of food security, such as the one adopted at the World Food Summit in 1996, includes four main components: availability (which depends on domestic supply and trade of food); access (which is influenced by income, employment, and poverty patterns related to economic growth and development); utilisation (which depends on the nutritious quality of food, but also on other factors such as health services, water and sanitation infrastructure, education, women empowerment, and good governance); and stability (i.e. that people should have physical and economic access to adequate food at all times). Agricultural growth and, in particular, increased food production contribute to all four aspects of food security. Food production directly ensures availability (first component). While overall economic growth is a key factor contributing to the remaining three aspects of food security, the role of agricultural and food production growth is also important. For example, it generates broad employment and income opportunities that are crucial for food access (second component). Agricultural growth, with its multiplier effects in the rest of the economy, also contributes to government revenues, which may be used to improve basic health services, water and sanitation systems and safety nets for the poor and vulnerable. These expenditures and investments, in turn, help both access and the proper utilisation of food (second and third components). Support for public goods that enable food production combined with an active role of trade policies help ensure stability of food consumption ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 23 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 (fourth component). Thus, across all four dimensions, a positive enabling environment for agricultural growth and competitiveness contributes to food security. While agriculture and food production contribute to the four components of food security, the latter is a multifaceted concept that is affected by a variety of factors. For instance, Smith and Haddad (2000) in a cross-country analysis with data from 63 developing countries over the period 1970-96, and using anthropometric measures of food insecurity (linked to child malnutrition), found that in the regions with the highest rates of food insecurity (Sub-Saharan Africa and South Asia), besides improvements in per capita food availability, women’s education also contributed to strong declines in food insecurity. In East Asia, the need to improve women’s education (and the status of women relative to men) ranked above food availability as a contributor to food security. In other regions, such as Latin America, North Africa and the Middles East, increases in food availability were not as relevant, while issues such as women’s education and status, and the provision of health services appear more important to reduce food insecurity. In summary, the enabling environment for agricultural growth and competitiveness appears crucial for poverty alleviation and food security in most developing countries. Additional considerations about agrarian structures, the role of small farmers and rural labourers, and women empowerment should also be taken into account when designing government interventions. World economic conditions The prevailing international economic conditions affect economy-wide growth in individual countries, as discussed briefly in Section 2.1, and in the agricultural sector, in particular. Domestic policies may have different effects depending on whether the world economy is growing or not; on whether world interest rates and world agricultural prices are high or low; on the evolution of the exchange rates of major global currencies; on the level, composition and direction of international capital flows; and on other similar indicators of the world business cycle. The world economy has gone through different growth cycles during the last decades, with two periods of relatively high agricultural prices (as well as other commodities): in the 1970s, and since the late-2000s. In between, the global economic recession of the early 1980s, a result of tight monetary policies in industrialised countries to control inflation, along with protectionism, production subsidies and export subsidy wars among some major agricultural producers, such as the European Union and the United States, combined with strong productivity improvements led to the collapse of agricultural prices in the mid-1980s and 1990s. This in turn appears to have discouraged investments in the rural sector of many developing countries, with negative consequences for the rural poor. The World Bank and other development banks cut the amounts of loans to agricultural and rural development projects, a decision that was influenced in part by low world agricultural prices that reduced the returns of those projects (Lipton and Paarlberg, 1990). Swings in the value of the US dollar, influenced by variations in the monetary stance of the Federal Reserve, also contributed to variations in the nominal prices of agricultural commodities (Mundell, 2002; Orden 2002; Frankel, 2006). Shifts in capital flows (with associated booms and busts in developing countries), have had strong impacts on agricultural conditions world-wide as well. Capital outflows and devaluations during the 1980s debt crises in Latin American and Caribbean (LAC) countries and the simultaneous strong decline in overall growth during what has been called the “lost decade” affected production of livestock and dairy products and of raw materials for non-food manufacturing 24 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 products, whereas food crop production (which tend to be more tradable, and benefited from exchange rate adjustments) fared relatively better (López-Cordovez, 1987). Another example is the sequence of financial crises between the mid-1990s and early 2000s, which disrupted the economies of many Asian and LAC countries. The 1997 financial crises in Asia led to the contraction of demand for agricultural products in world markets, while the economic problems in Brazil and Argentina in the late 1990s and early 2000s resulted in expanded world agricultural supplies, leading to the decline of world agricultural prices (IMF, 1999; Langley, 2000; Langley et al., 2000; Shane and Liefert, 2000). The impact was not limited to those countries. Most of the capital flowing out of crisis countries largely went to developed countries, mainly the United States. Capital inflows appreciated the US dollar. As a consequence of all those factors, nominal and real prices of commodities at the end of the 1990s and early 2000s reached some of the lowest levels recorded, contributing to lower investments in the sector that was a factor later in the price spike of 2008. The current global economic crisis has opened a new phase in world economic conditions. Policies are still trying to correct the imbalances in fiscal, financial, and external accounts, with uncertain future implications for macroeconomic developments and their effects on agriculture (Orden, 2010). In summary, while it makes sense to concentrate on the supply-side determinants of a positive enabling environment for agriculture in a sustained period of relatively high agricultural prices, discussing the domestic policy and investment conditions for agriculture in emerging and developing countries without considering the evolution of global macroeconomic conditions will miss an important component of the overall policy setting. It must also be recognised that those world conditions will not impact equally on these countries; rather those results will differ depending on different structural conditions and on the policies followed by governments. Heterogeneity among regions and countries Levels of agricultural growth and development are also related to the variety of agricultural conditions in different countries. The 2008 World Development Report (World Bank, 2008) divided developing countries into three groups, depending on the contribution of agriculture to growth and the importance of rural poverty. The groups were called “agriculture-based countries” (where agriculture contributes significantly to growth and the poor are concentrated in rural areas), “transforming countries” (where agriculture contributes less to growth but poverty is still predominantly rural), and “urbanised countries” (in which agriculture is not the main contributor to growth and poverty is mostly urban). Countries in Sub-Saharan Africa (SSA) represent the largest percentage in the first group; many countries from South and East Asia (SEA) and the Pacific and, to a lesser extent, North Africa and the Middle East belong in the second category; and LAC, but also Eastern Europe and Central Asia, are the main geographical regions for the third group. Agricultural sectors in these groups of countries show distinct characteristics, as discussed in depth by Diaz-Bonilla and Robinson (2010). Agriculture in LAC is less important as a percentage of the GDP and the rural population is smaller compared to total population than in other regions. SSA and SEA fall on the other extreme, with agriculture production and rural population having larger incidence in those regions. LAC depends more on agricultural exports, and agriculture appears more productive (per unit of labour), uses more capital (using tractors as a proxy), and, after South Asia, is the region better served by roads. SSA and LAC have more available arable land per capita than Asian developing countries, but average holdings are far larger in LAC and land appears to be ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 25 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 distributed more unequally in LAC than Asia, with Africa in between. SSA has an availability of arable land that is comparable to LAC, but average holdings are far smaller and of similar sizes to those in Asia. SSA also shows the lowest values for the capital/technology and roads indicators, highlighting some of the opportunities and constraints to expand agricultural production in that region. In terms of food insecurity indicators, SSA and SEA have a larger percentage of food insecure countries, and they are predominantly rural, while LAC has more countries in the neutral food-security category and are basically urban countries (Diaz-Bonilla et al., 2000). The same policy (such as maintaining domestic prices high to help producers or the opposite, keeping those prices low to help consumers) will have different impacts in these two types of countries. The different regions also differ in the structure of their agricultural trade, which has been changing over time. Africa exports mostly to the EU and other African countries, but, lately, trade with Asia has increased significantly, surpassing intraregional trade. The export partners of LAC countries are mostly internal in the region, followed by the US and Canada, the EU, and Asia but with large differences from north to south on the continent (countries to the south tend to have lower levels of trade with the US than the northern ones). Developing countries in Asia sell mostly to other developing countries in the region and only after that to Japan and the EU (Diaz-Bonilla and Robinson, 2010). This heterogeneity among developing countries has implications for the impacts of the enabling determinants of agricultural growth. For instance, an improvement of the terms of trade for agricultural products (say, by a devaluation of the local currency) will have a different production response in SSA, where producers face relatively more constraints in linkages to markets, infrastructure, capital, and technology, than in the other two regions. In SSA countries, the growing urban markets appear in several cases better linked to food aid and imports than to the producers in the domestic economy. In turn, the distributive effect will be different in small-farmer agricultural economies of Asia than in many LAC countries with dualistic agrarian structures and large populations of urban poor. In the latter countries improving relative prices for agriculture, at least on impact, will help relatively more large farmers while negatively affecting poor urban consumers. In addition, changes in macroeconomic and agricultural policies in Europe, for instance, could have a relatively greater impact on Africa than in Asia, due to their greater trade and financial links. The same can be said in the case of the US and a number of LAC countries. These differences in structure, performance, production, and trade must be kept in mind in analysing possible policy and other reforms to enhance agriculture’s enabling environment. Figure 1 (from Sala-i-Martin et al., 2013) has shown the GCI that defines three types of countries, depending on whether they have factor-driven, efficiency-driven, or innovation- driven economies. Sala-i-Martin et al. (2013) also identify countries in transition from the first to the second stage, and from the second to the third one, for a total of five categories, with the stage of development determined largely by the per capita income levels of countries. Table 1 cross-tabulates those five GCI stages with the three types of agricultural economies identified by the World Bank for developing countries. Illustrative examples of 36 countries are given in the cells defined by the GCI and World Bank categorisations. The selection of countries is simply to provide examples of developing countries for most of the cells (developed countries are not included in the three World Bank groups and therefore are not considered). Table 1 offers a way of looking at the combination of factors that enable growth for the whole economy (which, as discussed, is important for the development of the agricultural sector) and the constraints and requirements posed by the type of agricultural and rural 26 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 conditions of a country. It is clear that factor-driven economies, in which the country as a whole still needs to address the four basic pillars for growth as delineated in the GCI (institutions, infrastructure, macroeconomic environment, and health and primary education) are largely those that fall under the World Bank’s category of agriculture-based economies. Only Paraguay among agriculture-based economies included in Table 1 appears in the efficiency-driven economies, that are focusing on the efficiency enhancers of growth (pillars 5-10 in Figure 1) linked to higher education, the operation of goods, labour, and financial markets, technological readiness, and that have certain market size. On the other extreme, the urbanised developing economies in the classification of the World Bank do not appear as factor-driven economies in the GCI classification. Transforming economies in terms of their transition from agriculture-based to urbanised counties are largely classified as factor-driven or efficiency-driven in the GCI classification. An implication is that for growth and development purposes, the enabling environment for agriculture is dependent on the economy-wide enabling environment. While sectoral progress can be made, one cannot expect agriculture’s enabling environment to leap far ahead of the enabling environment across the entire economy. Table 1. Cross-classification of 36 emerging and developing countries by GCI and the World Bank World Bank Classification GCI Classification Agriculture-based Transforming Urbanised Stage 1: Factor-driven Ghana, Ethiopia, Kenya, Malawi, Nepal, Nicaragua, Nigeria, Senegal, Tanzania Bangladesh, India, Pakistan, Viet Nam Transition (stage 1 to 2) Egypt Bolivia, Philippines, Venezuela Stage 2: Efficiency-driven Paraguay Albania, China, Indonesia, Thailand Bulgaria, Colombia, Ecuador, South Africa, Ukraine Transition (stage 2 to 3) Malaysia Argentina, Brazil, Chile, Hungary, Mexico, Russia, Turkey Stage 3: Innovation-driven Slovenia Source: Authors’ tabulation. 2.3. Indices of growth and competitiveness determinants As discussed above, there is a strong connection between determinants of overall economic growth and competitiveness among countries and the factors affecting agriculture. Thus, the GCI competitiveness-factors articulated in Figure 1 provide a framework that is a useful starting point in assessing components of the enabling environment for agricultural growth and competitiveness and for the selection of indicators of that environment. Greater specificity about agriculture is provided by some of the indicators of the EIU’s GFSI and there are other relevant studies, existing databases and ongoing initiatives to identify, classify, select indicators and construct indices to quantify and rank the enabling factors for agricultural innovation, growth and competitiveness. The GCI and issues arising in construction of indices The first step in constructing an index is the choice of indicators for inclusion. Further details about the 12 pillars of the GCI (shown schematically in Figure 1) are presented in Table 2. As shown in the earlier figure, four pillars make up a block of “Basic Requirements” for competitiveness. These pillars address the quality of public and private ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES – 27 OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 institutions and availability of public-good infrastructure, as well as macroeconomic policy and the quality of human capital as reflected in health conditions and primary education. The basic requirements are considered most important as determinants of competitiveness for factor-driven economies. A second set of six pillars comprise a block on “Efficiency Enhancers” considered more relevant for efficiency-driven economies. Central to this block are measurements of the adequate operation of goods, labour and financial markets, complemented by measurements of higher education and technology readiness, and conditioned by market size faced by a country. The third block of determinants of competitiveness is judged particularly relevant to innovation-driven economies. This block of “Innovation and Sophistication Factors” is comprised of two pillars: business sophistication and R&D innovation. In constructing the GCI, each of the pillars 1-10 is comprised of two or more sub-pillars of key determinants selected as important to economic competitiveness. Each of these sub- pillars, and pillars 11 and 12, are comprised of one or more categories of determinants, and each category is comprised of one or more specific indicators. The indicators of the determinants affecting competitiveness are measurements either from primary data collected specifically for the GCI or secondary data available from other sources. The primary data collected by the GCI is obtained through surveys of firms (Executive Opinion Surveys, EOS) in each country. For the 2012 report about 14 000 surveys were utilised, representing an average of 100 respondents per country. The surveys ask respondents to rank different dimensions from 1 (worse) to 7 (best) (Browne, Geiger, and Gutknecht, 2012). Although the GCI data-gathering methodology has been improved there are still some weaknesses that can be understood by comparing the GCI surveys with another potential source of information about the enabling environment for economic growth. The comparative source of information is the Enterprise Surveys (ES) conducted by the International Finance Corporation (IFC) of the World Bank. The IFC/World Bank have been collecting since 2002 firm-level data using surveys of representative samples of the private sector in developing countries. The ES are usually face-to-face interviews with owners or high-level management of firms in different sectors and of different sizes, covering a broad range of topics related to the business environment, such as access to finance, corruption, infrastructure, crime, competition, and performance measures. Since 2002, there have been ES for about 130 000 companies in 135 economies. The data, methodology, and studies conducted using the ES are accessible at www.enterprisesurveys.org. A useful feature is that the ES distinguish a Food Sector within its manufacturing coverage, which provides information otherwise not available about this component within agricultural production and marketing value chains. Comparing the ES of the IFC/WB with the EOS of the GCI there are several points to note. First, the sample of the EOS for the GCI is significantly smaller than the ES; the latter typically includes 1 200-1 800 interviews in larger economies, 360 interviews in medium- sized economies, and about 150 interviews in small economies. Second, the questions in the GCI survey ask the respondent executives for more open-ended opinions, while the ES tends to ask about specific problems faced by firms. For instance, there is a difference between asking how would the respondent rate the situation of corruption or of the electrical infrastructure in a country (from 1 to 7) versus asking what would be the amount of bribes as a percentage of sales that a firm like the one interviewed has to pay, or how many hours were lost to electric outages. The latter approach provides more comparable and objective information, while it is difficult to compare the meaning of an open-ended http://www.enterprisesurveys.org/ 28 – ENABLING ENVIRONMENT FOR AGRICULTURAL GROWTH AND COMPETITIVENESS: EVALUATION, INDICATORS AND INDICES OECD FOOD, AGRICULTURE AND FISHERIES PAPER N°67 © OECD 2014 ranking number across different countries and cultures. Third, for more open-ended, general questions, such as those asked for the GCI, there is always the question why only a specific type of agents in the economy have been asked for their views (in this case, executives from firms). Despite these considerations, the structure and competitiveness methodology of the GCI are useful, even if the specific data may provide a less reliable comparative representation of country circumstances than the ES. While the ES of the IFC/WB generate a variety of indicators, they are not aggregated into single numbers to match relevant dimensions of a growth framework, as provided by the GCI. Another methodological point to be considered relates to the interpretation of certain indicators intended to reflect the adequacy of operation or efficiency of markets. This is particularly relevant for labour markets, where some analysts consider that greater flexibility for employers is better (as seems to be the case in the GCI framework), while other studies suggest that more collaborative and shared approaches in determining wage and employment conditions ma