40 4/30/03 4:39 AM Page 1

Center for International Forestry Research Center for International Forestry Research

CIFOR Occasional Paper publishes the results of research that is particularly significant to tropical forestry. The content of each paper is peer reviewed internally and externally, and published simultaneously on the web in downloadable format (www.cifor.cgiar.org/publications/papers).

Contact publications at [email protected] to request a copy.

CIFOR Occasional Paper No. 40 Exploring the Forest—Poverty Link: Key Concepts, Issues and Research Implications

Arild Angelsen and Sven Wunder The Center for International Forestry Research (CIFOR) was established in 1993 as part of the Consultative Group on International Agricultural Research (CGIAR) in response to global concerns about the social, environmental and economic consequences of forest loss and degradation. CIFOR research produces knowledge and methods needed to improve the well-being of forest-dependent people and to help tropical countries manage their forests wisely for sustained benefits. This research is done in more than two dozen countries, in partnership with numerous partners. Since it was founded, CIFOR has also played a central role in influencing global and national forestry policies.

CIFOR is one of the 16 Future Harvest centres of the Consultative Group on International Agricultural Research (CGIAR) ISSN 0854-9818

© 2003 by Center for International Forestry Research All rights reserved. Published in May 2003

Front cover photos: Christian Cossalter, Reidar Persson Back cover photos: Christian Cossalter, Edmond Dounias, Rona Dennis

Center for International Forestry Research Mailing address: P.O. Box 6596 JKPWB, Jakarta 10065, Indonesia Office address: Jl. CIFOR, Situ Gede, Sindang Barang, Bogor Barat 16680, Indonesia Tel.: +62 (251) 622622; Fax: +62 (251) 622100 E-mail: [email protected] Web site: http://www.cifor.cgiar.org Exploring the Forest—Poverty Link: Key Concepts, Issues and Research Implications

Arild Angelsen Associate Scientist (Economist), CIFOR Bogor, Indonesia & Associate Professor, Department of Economics and Social Sciences Agricultural University of Norway, Ås, Norway

Sven Wunder Scientist (Economist), CIFOR Bogor, Indonesia ii

Contents

Acronyms and Abbreviations v Acknowledgements v Executive Summary vii Abstract 1 1. Forests and Poverty—a Controversial Link 1 2. Defining Poverty in a Research Context 3 2.1 The traditional concept 3 2.2 Extending the poverty concept 4 2.3 Poverty and subjective well-being 8 2.4 Concepts and indicators: what to choose and how to measure? 10 2.4.1 Analysis vs. measurement of poverty 10 2.4.2 Degree of poverty 11 2.4.3 Absolute vs. relative poverty 12 2.4.4 Subjective vs. objective poverty measures 12 2.4.5 Forest definitions and poverty relevance 13 3. Causes of Poverty 14 3.1 Poverty and growth 14 3.2 Inequality and growth 14 3.3 Pro-poor preconditions and growth types 16 4. Using Forests to Cushion and Reduce Poverty 18 4.1 The forest-dependent poor 18 4.2 Non-Timber Forest Products (NTFPs) 19 4.2.1 NTFPs—a heterogeneous group 19 4.2.2 ‘Poverty traps’ or ‘safety nets’? 21 4.2.3 Fuelwood—the neglected pro-poor NTFP? 24 4.2.4 A synthesis and conclusion about NTFPs 25 iii

4.3 Timber 26 4.3.1 Why have the poor not benefited from timber? 26 4.3.2 Can timber be made more pro-poor? 29 4.3.3 Forgotten indirect and off-site benefits? 32 4.3.4 Summary on pro-poor timber benefits 34 4.4 Payments for ecological services 34 4.4.1 What link to poverty? 34 4.4.2 Carbon storage and sequestration 36 4.4.3 Biodiversity conservation 37 4.4.4 Hydrological benefits 38 4.4.5 Forest based tourism 39 4.4.6 Can forest-service payments reduce poverty? 39 4.5 Summary 40 5. Future Research on Forests and Poverty 40 5.1 Research area I: Exploring the present pro-poor role of forests 41 5.2 Research area II: Emerging markets and opportunities 44 5.3 Research area III: Cross-cutting institutional and extra-sectoral issues 45 Endnotes 46 References 48 Appendix 1. Strategic Questions for Poverty-Forest Research at CIFOR 54

Figures 1. The Poverty—Well-Being Interface 4 2. Decomposing Social Capital from a ‘Five-Capital’ Perspective 6

Boxes 1. The ‘Employment of Last Resort’ Model 23

Tables 1. The Importance of Different Forest Benefits to Different Groups 19 2. Poverty-Related Forest Research—Topics, Research Methods and Priorities 42 iv

Acronyms and Abbreviations

ACM Adaptive Collaborative Management CDM Clean Development Mechanism CGIAR Consultative Group on International Agricultural Research CIAT International Center for Tropical Agriculture CIFOR Center for International Forestry Research DFID Department for International Development, UK EPMR External Programme and Management Review (CIFOR) FAO United Nations Food and Agriculture Organization GBF Global Biodiversity Forum HDI Human Development Index (UNDP) HDR Human Development Report (published annually by UNDP) ICDPs Integrated Conservation and Development Projects/Programmes ICER Internally Commissioned External Review (CIFOR) IDS Institute for Development Studies, UK IFPRI International Food Policy Research Institute JI Joint Implementation (private sector climate change initiative) LULUCF Land Use, Land Use Change and Forestry (in CDM) MDG Millennium Development Goals NARS National Agricultural Research System NFPs National Forest Programmes NGO Non-Governmental Organisation NTFPs Non-timber forest products PAFSEP Poverty Alleviation, Food Security and Environmental Protection (CGIAR) PCA Principal Component Analysis PPA Participatory Poverty Assessment PPP Purchasing Power Parity PRSPs Poverty Reduction Strategy Papers SLA Sustainable Livelihoods Approach SSE Small Scale Enterprise SWB Subjective Well-Being UNDP United Nations Development Programme UNFCCC United Nations Framework Convention on Climate Change VAT Value Added Tax WWII World War Two v

Acknowledgements

In its early stages, this document received valuable food for thought from an internal one day meeting about poverty, hosted by the Center for International Forestry Research (CIFOR) at Puncak, Indonesia, on 23 April 2001. We are grateful for the external reviews by Mike Arnold and Manfred Zeller and for extensive comments from internal CIFOR colleagues, including Brian Belcher, Bruce Campbell, Carol Colfer, David Kaimowitz and William Sunderlin. We would also like to thank the UK Department for International Development (DFID) for their financial support. Forests have both potentials and limitations in regard to poverty alleviation, with possible roles as safety nets, poverty traps and pathways out of poverty. Poor people living in and near forests (like this family in the Rio Capim area, Pará State, Brazil) are the ones who are most dependent upon forests for their livelihoods. (Photo by Sven Wunder) vii

Executive Summary

This paper provides an overview of issues, income sources only benefit the few. However, debates and research on the link from forest to our review of the macro-level literature shows poverty alleviation. Some of the key questions that economic growth more often than not does include: How do different poverty definitions trickle down to the poor—and that poverty affect the assessment of that link? What are reduction without growth is in practice extremely the current major forest contributions to poverty difficult to achieve. alleviation—distinguishing between poverty Our review of the actual and potential role prevention and poverty reduction? Why do the of forests in poverty alleviation distinguishes poor tend to depend more on forests? Does a three benefit categories: non-timber forest high level of present forest dependence products (NTFPs), timber and environmental necessarily correspond to a high potential to services. reduce poverty through forests in the future? NTFPs cover a wide range of products with Are forest products safety nets or poverty traps? different characteristics. Some of these serve Can payment for environmental services help subsistence needs, others have important gap reduce poverty, while conserving forests? In filling or ‘safety net’ functions and a few provide general, how and to what extent can forests and regular cash income. Most NTFPs are labour forestry be made more pro-poor? intensive, require little capital and skills, are Poverty definitions have gradually been openly accessible for extraction from natural broadened over time from a ‘materialistic forests and provide generally poor prospects for core’ to include an increasing number of ‘soft’ market and price growth. Unfortunately, this welfare elements, as exemplified by the so combination makes the majority of NTFPs called ‘five-capital’ approach. However, this economically inferior products, yielding low change has also come at the expense of returns for those engaging in their production tangibility, measurability and comparability. and trade. Paradoxically, the characteristics that We stress the need to distinguish between the make NTFPs important and attractive to the analysis and the measurement of poverty. It poor are the same ones that limit the potential can be useful to think about poverty causes in for increasing NTFP incomes. Yet, there are also broader terms (for example, the ‘Sustainable some NTFP exceptions that fall outside this main Livelihoods Approach’) but to measure it using pattern—these are managed more intensively, indicators closer to the original meaning of have better market prospects and/or provide the term. higher economic returns. Another key link that has a strong bearing If NTFPs have normally been the poor on the poverty-forest debate refers to the person’s lot, it is the rich who have mostly empirical macro-level relationship between captured the benefits from precious tropical economic growth and poverty reduction. Many timbers, due to a number of production and micro-level practitioners observe that new market characteristics (high capital and skill viii

intensity requirements, more specialised markets. Payments for forest based biodiversity, markets, etc.). Still, as already alluded to, this tourism, hydrological protection and carbon-sink general anti-poor verdict overlooks the fact functions all have a vast global value and the that, over time and space, some of these pro- increasing threats against them also enhance rich rents trickle down to the poor, through the users’ willingness to pay to preserve them. indirect channels (e.g. employment, profit However, the poor are often disadvantaged by reinvestment and multiplier effects). Even their insecure land tenure and the transaction when the poor are not the ‘first-round’ costs and risks that service buyers face in dealing recipients of the rents, this does not necessarily with many smallholders or communities with mean that they don’t benefit at all. In addition, internally divided interests. some current trends, such as the increased local As a result of our review, we have selected ownership of natural forests, could make what we personally see as the ten most promising certain timber sub-sectors more pro-poor in areas for future research. In our view, the highest the future. Tree commercialisation for the priority themes are: (1) Natural forest products growing markets for fibre, pulp or construction in household livelihoods (safety nets and timbers can also help the poor, for example, increased welfare); (2) Small scale wood based through creating contractual arrangements processing enterprises; (3) Globalisation, trade between companies and smallholders. Finally, liberalisation and markets; (4) Smallholder tree small scale wood processing in many areas planting and private sector partnerships; and (5) provides a significant and growing level of Payments for forest environmental services. The employment. second highest priority we assign to the following Payment for the ecological services of topics: (6) Economy wide benefits of forest based forests is an emerging area with a high growth rents; (7) On-site ecological services; (8) Local potential, but there are also uncertainties resource control and land tenure; (9) regarding how large and widespread such Decentralisation, governance and market transfers will be—and to what degree poor people deregulation; and (10) Integrating forests into will be able to enter into these emerging macroeconomic strategies. Exploring the Forest—Poverty Link 1

Abstract

This paper provides a global review of the link from forests to poverty alleviation. Definitions are clarified and the key concepts and indicators related to livelihoods and poverty reduction and prevention are explored—distinguishing between the analysis (using broader welfare elements) and the measurement of poverty (using more tangible, traditional indicators). Reviewing the macro-level literature on the relationship between economic growth, inequality and poverty, we find that economic growth usually does trickle down to the poor and that poverty reduction without growth is in practice very difficult to achieve. The potentials and limitations of forests in regard to poverty alleviation are canvassed and their possible roles as safety nets, poverty traps and pathways out of poverty are explored. A striking gap exists between, on the one hand, the current neglect of forests in many economic development and poverty reduction strategies and on the other, the high (and sometimes unrealistic) expectations regarding the role for forest products in parts of the forest literature. Both positions are critically evaluated. The core discussion addresses how forests can contribute to poverty reduction, distinguishing three main benefit categories. Firstly, non-timber forest products serve subsistence needs, may have important gap filling or safety net functions and sometimes provide regular cash income. Secondly, timber has not traditionally been very pro-poor but the current trends of increased local ownership of natural forests, growing tree commercialisation and small scale wood processing could modify that picture. Thirdly, ecological service payments are emerging rapidly but it is uncertain how much the poor will benefit. In conclusion, we outline ten promising research topics, within three broader fields: assessing current forest-based benefits to the poor; exploring emerging market opportunities; and evaluating cross- cutting institutional and extra-sectoral issues.

Keywords: Poverty Alleviation, Rural Livelihoods, Timber, Non-Timber Forest Products, Ecological Services

1. Forests and Poverty— forest loss ‘just’ by bringing development to the South was probably partly based on wishful a Controversial Link thinking (e.g. Angelsen 1997; Duraiappah 1996; The issues of poverty reduction and of Kaimowitz and Angelsen 1998; Scherr 2000). deforestation/forest degradation both rank This paper focuses on the link from forests highly on the current international agenda. to poverty. There are at least two reasons for Attempts have been made to link the two together this. Firstly, we probably know more about the in a ‘downward spiral’—poverty is seen as a cause link from poverty to forest loss than about this of forest loss and forest loss contributes to reverse link. Secondly, there is currently more maintain or even increase poverty. This implies international interest in the potential of forests that economic development and poverty reduction to alleviate poverty than in poor people should help improve forest conditions and vice threatening tropical forests. For example, the versa, that development of forest resources can Revised Forest Strategy of the World Bank, states be an important vehicle for poverty reduction. explicitly that “the strategy must give priority to This two way link has nonetheless, been poverty reduction” (World Bank 2001). Also, questioned, with both causal directions having among an increasing number of donors and some empirical evidence going against them. national governments, reducing poverty has Research findings tend to de-emphasise the become the focal point for development importance and question the general validity of interventions. In fact, the Poverty Reduction the link from poverty to resource degradation. Strategies are becoming the key framework for The rural poor’s resource management has projects and interventions in developing strikingly variable effects upon forests and their countries. The question therefore arises—what management responses also vary in relation to role can forests and forestry play in the efforts of marginal changes in welfare. The state of forests developing countries to reduce poverty? is actually as much threatened by wealth as by Prior to the analysis, a couple of definitions poverty. Hence, the belief that one could halt are required. Considerable confusion exists about 2 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

how to define and measure different adjustments making are examples of activities that we to poverty over time. This can lead to cardinal mention tangentially in the text but in principle, errors in the logic about the potential of forestry exclude from forestry proper. This is because interventions. For instance, let us assume that most of the final value in the resulting some poor people living in forest margins currently commodities is normally added further derive a high share of their income from their downstream in the secondary processing stages, forests—say, 30% (subsistence and cash income compared to the value that is generated in and combined). From this, observers often conclude near the forest. Yet, as discussed later, there are that forestry interventions are a key to reducing very important ‘forward linkages’ and ‘multiplier poverty over time. However, that extrapolation, effects’ in these downstream activities. from a static observation to a dynamic approach, Several other recent publications by CIFOR is flawed. In fact, a high dependence on certain and its associated researchers also deal with the forest products may in the first place, have been forest-poverty link. Hence, this paper serves to a cause of poverty. Or, more typically, high forest complement the existing body of CIFOR’s work dependence and poverty reflect that other on poverty and forests. Arnold (2001) focuses employment options that offer higher returns are particularly on the poverty alleviation role of not accessible to the poor. forestry aid. Wunder (2001) limits his discussion Let us begin by clarifying our terminology. of poverty reduction potentials to natural forests We use the term poverty reduction to describe a (excluding tree planting) and their products situation where people are becoming measurably (excluding forest services) but also deals with better off over time, in absolute or relative terms. the reverse link, from poverty to forest condition. That is, people are being ‘lifted out of poverty’ Byron and Arnold (1999) disaggregate the term when they climb above a pre-defined poverty line. ‘forest-dependent people’ into more functional In contrast, poverty prevention refers to the role socioeconomic sub-groups, which we will refer of forests in helping people to maintain a to later. Neumann and Hirsch (2000) review the minimum standard of living (even when it is below poverty alleviation potentials of Non-Timber a given poverty line) and helping them to survive. Forest Products (NTFPs), while Arnold et al. We are thus comparing the present to a (2003) look specifically at fuelwood. hypothetical situation where people would have A joint report by Forest Trends and CIFOR less or no access to the forest resources they use (Scherr et al. 2001) provides a comprehensive today, a threat that is very real in many specific review of markets for forest products, while situations. Poverty prevention thus refers to Smith and Scherr (2002) assess the opportunities insurance and safety net functions, it cushions for selling forest carbon under the Clean or mitigates poverty without lifting people above Development Mechanism (CDM), to improve local the poverty line or even without making them livelihoods. Kaimowitz (2002) provides an better off over time. Finally, we use the term overview and discussion of the forest benefits poverty alleviation as an inclusive term, to rural livelihoods in Africa and Sunderlin et al. encompassing both poverty reduction and poverty (2003), in a Chapter on Forests and Poverty in prevention. FAO’s “State of the World’s Forests 2003” We found that in the literature, the terms report, present a brief synopsis of all the above ‘forests’ and ‘forestry’ are not used consistently. mentioned topics. In the following, we will refer to ‘forests’ in a While clearly overlapping with most of these way that is comparable to the definition of the issues, the present document takes a slightly Food and Agriculture Organization of the United different angle. Firstly, we step back for a Nations (FAO)—i.e. 10% crown cover, including moment to look at the implications of different both open woodlands and plantations. In some key concepts related to livelihoods and to poverty cases we refer to agroforestry and trees on farms and its alleviation. Secondly, we integrate into but in principle, we count these activities as the discussion both products from planted forests part of agriculture, not forestry—in spite of the and trees and the emerging payments for forest fact that both can produce some of the same services. Thirdly, we discuss forests in the light outputs as forests. of general economic theories about the nature We interpret ‘forestry’ not only as industrial of poverty. In addition, we point to strategic timber production but also more broadly, as any implications for forest research organisations. forest related primary production activity. This Finally, we also discuss some of the ongoing includes activities in sawmills or in charcoal political processes in this field, such as the World production, while woodcarving and furniture Bank’s new Forest Strategy paper (World Bank Exploring the Forest—Poverty Link 3

2001) and FAO’s policy brief on how forests can poverty and discusses their applicability for be made more pro-poor (FAO 2001)1 and we look studying interactions with forests. Section 3 at the role of forests in Poverty Reduction provides a few insights from a long standing Strategies (Oksanen and Mersmann 2002). The development debate on the links between first two documents signal significant shifts in economic growth, inequality and poverty the positions assumed by two major actors in reduction. These sections form the basis for the the international forest development arena, core discussion in Section 4 on how forests can although the latter shows there is still some way contribute to poverty reduction within three to go in terms of implementing the changes in different fields: NTFPs, timber and national policies. environmental services. Section 5 poses some This paper will not deliver a ready made research questions and suggests promising menu of actions for forest led poverty topics. The appendix deals with some strategic alleviation. We are not able to provide definitive issues for future poverty research specifically guidelines as to which baskets should receive at CIFOR but this will also be of interest to other how many eggs, so to speak. Yet, from the institutions conducting research within this field. available evidence and according to common In a slightly more provocative tone, we also put sense, some baskets certainly appear more forward the pros and cons vis-à-vis a series of promising than others. Moreover, the choices strategic research choices. regarding both poverty and forest concepts and indicators have important implications for the basket selection. With this paper, we aim to 2. Defining Poverty in make the assumptions behind these choices more transparent. a Research Context There is a striking gap between on the one hand, the almost complete neglect of forests in 2.1 The traditional concept economic development and poverty reduction Concepts at the centre of international debates strategies and on the other, the expected role often face pressures for a broadened for forest products in parts of the forest literature interpretation. While adjustment of concepts and in the writings on NTFPs in particular. We over time may be desirable, they also generally intend to critically discuss both positions. tend to become increasingly embracing and Empirically, too little is known about the potentials inclusive and therefore face a real risk of and currently, both neglect and unrealistic overloading and ‘concept degradation’. Certainly, expectations coexist. Forests have both the term ‘poverty’ is no exception. The traditional potentials and limitations in regard to poverty definition of poverty has focused on income and alleviation (i.e. reduction and prevention). wealth—for example, as defined in Webster’s We acknowledge that there is an inherent dictionary as: “The lack or relative lack of money problem in global reviews such as this one. When or material possessions” (Webster 1993). This discussing the whole developing world at the same materialistic definition from classical economists time, few permissible generalisations exist and like Adam Smith and David Ricardo dominated almost every opinion is valid somewhere. The until the post WWII period. poor, even the so called ‘poorest of the poor’, In recent decades however, it has become tend to be a very heterogeneous group and increasingly popular to extend the definition of exhibit a range of quite different interactions poverty to other, non-material aspects of human with forest resources. Hence, proposals for well-being. Indeed, the evolution of the concept intervention must be context specific, providing of poverty reflects changes in development reference to particular forests and socioeconomic theories generally and the analysis of the causes and political conditions. The welfare status of poverty specifically. The attempts at alone of people living in and near forests tends measurement, description and analysis of to say little about their interaction with forest poverty have expanded accordingly. assets. Nevertheless, the poor do have some Figure 1 illustrates possible dimensions of characteristics in common—which served to poverty and how the perception of these make them poor in the first place: a shortage elements has evolved over time. The inner core of assets or capital (see section 2.2) and limited of the circle shows monetary income or opportunities in the labour market, etc. consumption, at different levels of aggregation This document is structured as follows. (e.g. from household surveys or national income Section 2 presents different definitions of statistics). During the post-WWII period and 4 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

up until the 1960s, the policy focus was more might call a ‘human development extension’ or less exclusively on expanding monetary of the poverty term, implying that increasing income. The accumulation of man made attention was given to indicators related to (physical) capital, through savings or aid, was health, education and nutrition (see third layer perceived to be the key to achieving economic in Figure 1). The most popular indicator has growth, which would then ‘trickle down’ and been the Human Development Index (HDI) benefit the poor. Growth and poverty reduction published annually by UNDP. HDI includes per would not always coincide perfectly but capita income but with decreasing welfare ultimately they would go hand in hand. ‘returns’ over growing income levels—an extra dollar in a poor country has a greater HDI effect than in a rich one. In addition, health (with Figure 1. The Poverty—Well-Being Interface life expectancy as the indicator) and education (with literacy and school enrolment indicators) are incorporated into the index. In regard to forest dwelling people, the HDI adds some valuable dimensions, compared to a pure income measure. People living in remote areas with abundant forest resources sometimes have good access to food consumption or even relatively high cash incomes. However, low government service levels at the forest frontier mean that they lag behind in terms of health and education indicators. Nevertheless, one could still argue that the choice of the particular HDI indicators as well as the weighting among the three indicators (income, education and health) remains arbitrary (Ravallion 1997). Attempts have thus been made to increase the number of poverty indicators and make the selection process more transparent—although this greater sophistication and complexity comes at the cost of a less intuitive interpretation (see Section 2.4). Opposition against the belief that economic growth automatically takes care of all human The second layer refers to non-monetary needs also came from the environmentalist income and consumption—the ‘hidden harvest’ corner, notably with works such as “Limits to of subsistence goods that do not enter the formal Growth” (Meadows et al. 1972). Again, this had cash based economy. In market-remote areas, implications for the poverty concept as well. these may be more important for the household During the 1970s and 1980s, emphasis gradually economy than cash income. Forest products shifted from physical, man-made capital to frequently contribute more to poor households’ human and natural capital as the foundation for non-monetary rather than monetary income. The welfare improvements. This also provided a second layer represents no fundamental shift in basis for integrating natural resource the poverty definition but points to the critical management in general and forests in particular, importance of including goods not entering the into the discussion about basic needs and marketplace. poverty alleviation. The third layer in Figure 1 illustrates this shift. Over the last decade, we have witnessed 2.2 Extending the poverty concept what one could call an empowerment and During the 1970s, the emphasis in the institutional extension of the poverty concept— development debate gradually shifted from the fourth layer in Figure 1. Some of the hardcore economics towards the ‘basic needs’ catchwords have been ‘decentralisation’, of poor target groups. This accompanied ‘devolution’ and ‘governance’. One of the changes in the measurement of welfare. The underlying reasons for this change was a growing poverty concept was subjected to what one recognition that these power related factors are Exploring the Forest—Poverty Link 5

It can be useful to think about poverty in broader welfare terms, using a range of indicators of well-being (like health, education and nutrition) but to still measure it using more tangible, traditional indicators, closer to the original meaning of the term. Ghanaian immigrants in a village on the central Ivory Coast, in search of improved well-being. (Photo by Carol J.P. Colfer) welfare creating in their own right (e.g. it is more matter for poor people in their daily lives. Only satisfactory for people to live in a society with such a holistic assessment would be able to democratic rights). Additionally, they were “build on the positives which people already have assumed to have a positive impact on the creation in their grasp” (Carney 1998) and thus fully reflect of material benefits, cf. the debate on whether existing options for poverty reduction. In terms democratic institutions enhance economic of identifying constraining capital types, the growth, e.g. Rodrick (2000). Empowerment is institutional focus is summarised in what has thus both a means and an end. been termed social capital. At the same time, social capital also has an important link to environmental factors (e.g. Pretty and Ward The Sustainable Livelihoods 2001), adding to the relevance of forests. Approach (SLA) The SLA is closely linked to what has been A widely applied concept, originally developed termed the ‘five-capital approach’, cf. Scoones from work at, inter alia, the British Institute (1998), Carney (1998) and Bebbington (1999). for Development Studies (IDS), is the Natural, human, social, physical and financial Sustainable Livelihoods Approach (SLA). This capital represent the main asset categories. approach was made the cornerstone of the However, Bebbington (1999) substitutes poverty reduction strategy of the Department ‘cultural’ for financial capital as the fifth asset for International Development (DFID) in the type for his analysis of Andean peasants, thus United Kingdom (Carney 1998). The SLA has stretching the concept further away from the also been applied to the management of forest materialistic interpretation. Baumann (2000) resources (Warner 2000; Shepherd et al. further adds to the conceptual complexity by 1999). “A livelihood comprises the developing a new five-capital version with capabilities, assets and activities required for ‘political’ capital as one of the assets. a means of living” (Warner 2000). This draws Bebbington (1999) discusses how the on the ‘actor-centred’ perspective on poverty livelihood outcomes of different households are of Chambers and Conway (1992). determined by their access to capital The idea is basically that development endowments and processes of capital use, practitioners should look at all the factors that transformation and reproduction. Similarly, 6 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Carney (1998) presents a framework in which generate income)? We lean more towards a ‘flow’ assets are transformed into livelihood outcomes or ‘outcome’ oriented definition, where assets through a set of transforming structures and are a means to create welfare. This would imply processes. Her livelihood outcomes include that assets or capital are normally not ends in income, well-being, vulnerability, food security themselves. and sustainable use of natural resources. Figure 2 shows an applied example of the Recently, institutions have been emphasised five-capital approach from CIFOR’s work in more as an independent factor in the SLA Bulungan Research Forest, Kalimantan, concept, although there is a clear overlap here Indonesia. In this case, CIFOR researchers with ‘social capital’. Nobel Laureate Amartya disaggregated the social capital category into Sen has greatly influenced the empowerment and five quantitative sub-groupings: ownership, institutional extension. In his work (including security, access, conflict resolution and rule his recent book, “Development as Freedom”), enforcement. The situation in the year 2000 Sen outlines a broadened concept of poverty as (represented by the inner ‘spider web’) is then a ‘capability deprivation’. Poverty reducing compared to that of 1996 (the outer one). As development is thus viewed as a process of the 2000 representation (with less security, greater freedom of choice, in a multidimensional access and ownership and with the other way (Sen 2001). Both of these possible extensions endowments being unchanged) is fully contained are expressed in the outer circle of Figure 1. within the ‘web’ for 1996, it can be concluded Much of the discussion about these concepts that social capital has unambiguously declined depends on how one chooses to define poverty during the period under analysis. along several related dimensions: The five-capital approach uses this kind of • flows vs. stocks analysis to illustrate changing endowments of • ends vs. means natural, financial, human, social and cultural • outputs vs. inputs capital. However, it also has several problems • outcomes vs. processes as a comparative poverty descriptor. First of all, this tool cannot evaluate the phenomenon For example, are people defined as poor of ‘overlapping spider webs’—that is, of some because of their limited income or also because capital types increasing while others decrease. of their limited assets (which are used to And unfortunately, this is precisely what tends

Figure 2. Decomposing Social Capital from a ‘Five-Capital’ Perspective

An example from CIFOR research in Bulungan, East Kalimantan, Indonesia Exploring the Forest—Poverty Link 7

to happen in a society during ‘development’. progress’. There is no ‘accumulation of goods’ Structural changes occur, for example, when nor clear cut ‘investment strategies’ allowing people in forested areas reduce their natural people to actively increase the reiteration of these capital (e.g. selling a tree) to acquire more processes. We should perhaps recognise that not financial capital (e.g. accumulating savings), everything that matters for production and physical capital (e.g. buying a small truck) or welfare is a ‘capital’, nor is it easily accountable human capital (e.g. paying for children’s school in economic terms more broadly. fees). Based on these three arguments, we conclude The five-capital approach basically that the five-capital approach is instructive in ‘compares peaches to oranges’ and does not in pointing out the multidimensionality of human or of itself, provide tools to evaluate whether welfare and of the processes generating this capital substitutions make people better or worse welfare. It can be a conceptual tool at the level off. Of course, that may adequately reflect the of grasping an overview of a series of livelihood fact that human welfare is multidimensional. changes along different dimensions. However, As an analogy, we don’t live by peaches alone it cannot be used for the measurement of but also by oranges and other fruits. Yet as long poverty. Hence, it is also unsuitable in specific as there is no equivalency standard between case studies for making overall comparative these various ‘fruits’, the five-capital approach judgements about the desirability of change cannot tell us anything about the comparative processes. desirability of situations—except for the highly In summing up the points raised in this exceptional situation where, as in Figure 2, all section, the trends over the past three decades types of capital move parallel in the same can be represented by a move from the inner direction over time. materialist core towards the outer human well- A second observation is that normally, one being periphery illustrated in Figure 1. More cannot live on capital itself. It has to be used specifically, this means going: and transformed in productive ways to create • from a reductionist one-dimensional welfare. A change in stock endowments alone index to a vector with multiple indicators, can give a highly misleading picture. For • from (intended) objective measures to instance, settlers in a forest abundant area may (consciously) subjective ones or under-utilise their large stock of natural capital. • from an economistic ‘top-down’ Therefore, we can often observe that they are approach to a holistic ‘participatory’ one, eager to trade a good share of their forest for • from a pure outcome measure to a what outsiders perceive to be a negligible causality-inclusive indicator and increment in financial capital. This is because • from materialistic to ‘soft’ assets in many cases, the latter constitutes a greater (capitals) with decreasing tangibility, constraint on welfare improvements than the measurability and comparability. former. Eventually, problems of undesirable capital composition and uneven capital productivity lead us back to income, The World Bank’s three-dimensional consumption or other indexed variables related concept of poverty to the welfare outcome, rather than the capital The World Bank and other international agencies inputs producing it (see below). have also responded to the demands for a Thirdly, the definition of ‘capital’, like broader poverty definition but have taken a ‘poverty’, is originally an economic term. One different route compared to the SLA approach. might legitimately question whether it is being The three-dimensional poverty definition used stretched to an extreme where it now makes by the World Bank (2000) maintains a focus on little sense. We believe ‘cultural capital’ and the materialistic core but also tries to be more ‘political capital’ should be avoided as scientific encompassing and inclusive. The three terms and social capital is a debatable, borderline dimensions included in this definition are: case.2 This verdict in no way denies that social organisation (e.g. local institutions effectively (1) Opportunity—this includes income, managing forests) and cultural processes (e.g. education and health as measures of indigenous knowledge about forest uses) well-being. This component is contribute positively to production and welfare. therefore similar to UNDP’s Human But so do many other things that we do not label Development Index. as ‘capital’, like ‘labour’ or ‘technological 8 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

(2) Security—this refers to the risk of 2.3 Poverty and subjective experiencing periods of shortfall in well-being income (below a poverty line) or other Regarding the poverty of people living in and welfare indicators. Vulnerability is a near forests, an essential distinction is between function of both the probability and size on the one hand, people’s own subjective of an income (or health) shock and the perceptions of well-being and on the other, the ability to cope with such shocks in terms attempts of others (e.g. researchers or of safety nets, alternative income statisticians) to objectively form their conclusions sources, insurance mechanisms, etc. based on externally defined, standardised indices (3) Empowerment—this covers access and of welfare. There has been growing interest in control over local resources, public the first kind of assessment, for example, as services, influence in local decision illustrated in the work of the International Center making and so on. As pointed out by for Tropical Agriculture (CIAT) (CIAT 1999; Sen (2001), power can be seen both as Ravnborg 2002) and the World Bank’s large an end in its own right and as a means ‘Voices of the Poor’ study (Narayan et al. 2000). to create opportunities. Yet, not much literature has dealt explicitly with the relation between the two types of measures. This three-dimensional concept is an Is it best to simply ask people whether they view attractive representation of the poverty concept. themselves as poor and why, or should one come Note again though, that normally only the first in with pre-defined concepts? One might argue dimension (opportunity) is measurable, while the on philosophical grounds that only the individual’s second may be but only under certain own judgement matters—‘you are poor if you feel circumstances. The concept thus has important poor’. Further, as a research strategy, the qualitative elements that one should keep in subjective approach is valuable. What matters mind but which do not lend themselves to are people’s own perceptions and preferences and quantitative comparisons. an aim of research is to assist people in fulfilling Difficulties in measuring and comparing the their aspirations. This section will examine the ‘security’ and ‘empowerment’ dimensions are pros and cons of this approach. also obvious from the World Bank’s own use of There is a large body of literature from the poverty term (World Bank 2000).3 When developed countries on income and subjective referring to poverty reduction in other well-being (SWB), which sheds light on this issue, publications, it is almost exclusively reduction as summarised by Offer (2000). SWB is usually in income poverty (‘a dollar a day’). One measured by survey interviews on different argument for using an apparently narrow income indicators, where individuals state their degree or consumption based definition of poverty is of ‘happiness’ and satisfaction with living that, in the long run and at an aggregated scale, conditions. In most developed countries, a cross- economic growth and poverty reduction (in section pattern of ‘decreasing welfare returns terms of the income, health, life expectancy, to income’ was found: education, vulnerability and political influence • At low personal income levels, SWB rises of the poor) often go hand in hand (Ray 1998; more or less proportionally to income. World Bank 2000). • At higher income levels, SWB returns to However, there are important exceptions additional income diminish and variance to that national development pattern. In some increases (i.e. non-income factors cases, such as in Cuba, Sri Lanka or Kerala state influence SWB more). in India, long term improvements in education and health were achieved in spite of a relatively These characteristics of the relationship stagnant monetary income. Conversely, research between income and welfare are rather robust— by the International Food Policy Research they hold with regard both to different countries Institute (IFPRI) has shown that higher income and different SWB indicators. However, these does not automatically improve nutrition levels generalisations are much less valid for time- (M. Zeller, pers. e-comm., 18 Feb 2002). At the series analyses. Repeat studies in three sub-national level, many regions that are rich developed countries (France, Japan and the in forest products and become integrated with United States) after WWII—a period of steady markets may also register an abundant cash flow income growth for all three countries—show that while social services consistently lag behind. We the proportion of the population stating it was will return to these issues in Section 3. happy had not increased significantly in any of Exploring the Forest—Poverty Link 9

the three countries. Actually, differences among difficult to evaluate the results. Comparative countries are more accentuated than in-country SWB studies would thus face even larger changes over time. problems, as the poverty standards and causal This finding might support those who claim factors are likely to differ greatly across that ‘money does not make you happy’, but that locations. would be an over-simplification. Those who Participatory Poverty Assessment (PPA) is ‘can’t keep up with the Joneses’, that is, who an attempt to empirically quantify and fall significantly behind the norm of (gradually aggregate SWB. For instance, the use of PPA rising) material living standards, generally have in Uganda illustrated that income growth is not a low SWB. In developed countries, the share of necessarily increasing SWB. Uganda enjoyed people satisfied with their living conditions is the highest rate of economic growth in Africa generally in the high range of 65-85%. So, what during the 1990s and surveys also demonstrated things make them express dissatisfaction, i.e. that household consumption had been a low SWB? In developed societies with elevated increasing. Nevertheless, the PPA highlighted income levels, the following deviations from the poor people’s perceptions of deteriorating well- ‘standard norm’ are associated with feeling being over that same period. Kanbur (2001) unhappy: unemployment, low educational levels, discusses further, using examples from Ghana low income (relative to the median and in the 1980s, how such divergence in poverty irrespective of the absolute level) and belonging measures can arise: growth is unequally shared to an ethnic minority.4 As indicated above, there with some groups falling behind (at least in are also differences among countries, with some relative terms) and experiencing a cultures (e.g. Japan and Southern Europe) deterioration of public services. expressing themselves on average, to be less However, the PPA methodology has also happy than others (e.g. Scandinavia). This is in been criticised. Ruggeri-Laderchi (2001) finds spite of the North of Europe showing a much evidence of ‘cosmetic’ participatory significantly higher incidence of suicides than components in her review. Even the most the South. authentic techniques have elements of One general lesson from this dominant arbitrariness—a complex reality has to be pattern in developed countries is that absolute structured and information is simplified, which income is very important for the well-being of introduces subjectivity. Moreover, proponents the poorest segments of the population, although of participatory methods make the debatable some poor egalitarian societies can also be assumption that the poor can actually articulate negatively affected by ‘unequalising growth’. and analyse their own reality, even in rapidly As income increases, SWB generally becomes changing environments. more linked to factors of equality and relative If income and ‘development’ indicators give position in society, which include many non- a poor reflection of SWB for poor forest dwellers, income factors. Another point is that stated SWB is it preferable then to work with a predetermined gives a poor reflection of welfare in a selection of qualitative indicators? It might be comparative setting. There seem to be no in some cases, but this approach does open up objective reasons why Scandinavians should be to the possibility of arbitrariness and the biased continuously much happier than the Spaniards selection of indicators. An illustrative example or Italians. Being Scandinavians ourselves, we was given at the ‘Poverty and Biodiversity’ rather suspect that in some cultures, it is much workshop of the 15th Global Biodiversity Forum more acceptable to openly express dissatisfaction (GBF), held in May 2000, in Nairobi (which one than in others.5 of us participated in). A representative of a South What do these findings mean for forest American conservation NGO had analysed the dwelling people in poor countries? Indicators of social and environmental consequences of a new subjective well-being have still not been tested road being constructed near a village in a heavily sufficiently in developing countries to allow for forested area. Not only was much of the forest comment on the extent to which they are useful destroyed, poverty had also increased, according for applied policy research (M. Zeller, pers. e- to all of the five qualitative indices that had comm., 18 Feb 2002). For the measurement of been chosen after initial consultation with village SWB at a single site, qualitative survey methods members. That is, local self determination had may provide a good reflection. However, we know declined, cultural identity had been from our own fieldwork that culturally biased compromised, environmental quality was and strategic responses will often make it reduced, inequality had risen and the village was 10 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

exposed to higher risks. There was hence an 2.4.1. Analysis vs. measurement of unambiguous increase in poverty and according poverty to the presentation it was obvious that The above discussion suggests that one ought biodiversity destruction went hand in hand with to distinguish between the conceptual analysis diminished human well-being. and the measurement of poverty. Poverty Nevertheless, taking a step back, one contains an important quantitative dimension. wonders what happened in the same village to all A key criterion for a poverty measurement the factors in the ‘materialist core’ of Figure 1 indicator is hence to allow for a consistent (i.e. to the options for commercialising products, distinction between the poor, the not so poor to local income generation and the community’s and the non-poor. Income (monetary and non- access to health and education services, etc). monetary) and consumption are still key concepts These are precisely the factors that often make in this respect, while SWB and livelihood remote communities opt strongly in favour of indicators are poorly suited for the job, especially road construction. This example shows that the for comparative purposes. On the other hand, choice of a few qualitative indicators can the sustainable livelihoods and five-capital ultimately imply a serious risk of linking up to approaches can help us to better understand the poverty concepts that, in the worst but probably causes of poverty related processes, especially not too rare case, are normatively tailored to at individual sites. In this way, the scope and fit a specific and external agenda, rather than focus of research will also influence the choice representing local people’s own priorities. of indicator. To summarise, the key argument for using The trend over time to use more inclusive SWB is that, at the end of the day, what matters measures reflects new research about the causes is one’s own (subjective) assessment of well- of poverty, proposed solutions, political trends being. Only people themselves can judge their and pressure from NGOs and others to broaden own welfare, with their own set of indicators. previously narrow definitions. With poverty back The main limitations of this approach are on the international agenda, donors might also strategic biases and the difficulty of be using poverty alleviation as a shorthand comparisons, particularly cross-sectional ones. notation for several development objectives Cultures and ‘thresholds of happiness’ differ. (health, education, equality, equity, economic The subjective nature of the question represents growth and empowerment) in the same diffuse a significant problem for the data collection. For way as ‘sustainable development’ was/is used. instance, some people might not want to reveal The term might become too all encompassing that they are unhappy to ‘save face’, while others and therefore eventually too vague to be useful want to exaggerate their poverty out of a for analytical and practical purposes. strategic desire to increase the likelihood of In a philosophical discussion of the poverty development assistance. concept, one would like to include both its multi- dimensionality and its subjective character. In practical work and in measuring poverty, 2.4 Concepts and indicators: what however, one has to make several simplifying to choose and how to measure? assumptions. Ravallion (1997) notes the great We will now summarise the discussion from this divergence between on the one hand, the section and make recommendations for applied comprehensive definitions of poverty found in analysis regarding four essential areas: the recent literature and on the other, what it is (1) The multi-dimensional use of the possible to operationalise and measure on the poverty concept (measurement vs. ground—both in terms of methods and data causal analysis) availability. (2) Variable degrees of poverty (the Income is just one option to condense ‘poorest of the poor’ vs. the moderately ‘peaches and oranges’ into a single numeraire poor) index. Obviously, subjectivity also enters into (3) Absolute vs. relative poverty (basic income based analysis, for example, in the needs vs. inequality) choice of goods and services entering the basket (4) Subjective vs. objective poverty of minimum consumption. Strategic responses measures may apply in consumption as well and in income surveys in particular. Producing good household income surveys is a relatively expensive exercise Exploring the Forest—Poverty Link 11

and takes time. Income comparisons across comparisons across sites. For a broadened countries using exchange rates are complicated approach including non-quantifiable poverty by variations in purchasing power capabilities, causes, it is often better to talk explicitly about although economists remedy this by using ‘poverty and human well being’, rather than Purchasing Power Parities (PPP). overloading (and ultimately devaluing) the term A good complement to both subjective ‘poverty’. perceptions and externally based indicators may be found in other diagnostic measures of poverty (e.g. visual assessments of malnutrition, birth 2.4.2 Degree of poverty weight, housing, education levels and other In our context, it is important to distinguish observable indicators). These can be used to between the (moderately) poor and the very corroborate or refute claims of severe poor. The poorest of the rural poor (e.g. landless deprivation (W. Sunderlin, pers. e-comm., 10 labourers) frequently have a fundamentally Jan 2002). Recent work at the International Food different relation to the forest resource than Policy Research Institute (IFPRI) has attempted the moderately poor (e.g. a smallholder) (Byron to use quantitative methods to synthesise these and Arnold 1999). Separating the extremely poor measures into a single index of relative poverty from the moderately poor, whether by (Zeller et al. 2001). More than 200 simple, consumption based or other indicators, is interview based poverty indicators were important for the targeting of poverty screened. These included both quantitative and programmes. While many forestry projects and qualitative factors and also incorporated direct programmes have been successful in reaching welfare measures, subjective well-being factors the moderately poor, there has been some and assets that were assumed to produce welfare concern that they seldom reach the poorest (e.g. of the five-capital type). Principal group. As stated by a senior forestry official from Component Analysis was used in each case to a bilateral aid agency during FAO’s poverty forum select 10-20 indicators for the construction of in September 2001, “Forestry is probably not a an index, as an alternative to income good tool to reach the poorest of the poor”. comparisons, thus also leaving space for local- There are several distinctions to be made. specific indicators. In some cases, the moderately poor are in a This type of index is flexible, less arbitrary better position to take advantage of those and more comprehensive than an income forestry opportunities that require a minimum measure. However, combining such a large level of education or other assets. Many of the number of (absolute and relative) indicators in poorest, landless groups use the forest resource a fully empirically driven selection procedure can in a degrading way (e.g. charcoal makers - see make it difficult to interpret the results. In other section 4.2.3) as they have no self-interest and words, alternative measures can be very useful incentive to conserve it. Other groups, like some but income still has a role to play. It is an of the poorest and more isolated indigenous indicator that has been thoroughly tested (for ones, are more dependent on the conservation the monetary part within the market place), that of forests (though not necessarily on has well known pros and cons and more appeal conventional forestry) for their welfare than to policy makers than the ranking of an abstract moderately poor squatters or settlers are. index. ‘Forestry’ in the narrower sense of the word For research purposes, it may be has been less successful in reducing extreme convenient to think about poverty in terms of poverty than say, small scale agriculture or labour the livelihoods approach or the three- intensive manufacturing. Targeted forestry dimensional method sketched by the World Bank programmes might sometimes achieve that but at the same time, to measure the material purpose but even these have shown mixed indicators that are closer to the original meaning results. Impact oriented forestry research aiming of the term ‘poverty’. This would mean that to make a difference in terms of poverty measurement should be done by income, the reduction is thus more likely to succeed if it value of consumption or other multi-factor includes the moderately poor than if it, from indices taking into account human development the outset, is exclusively targeted at the poorest dimensions.6 Ultimately, the choice of indicator of the poor. The issue of target groups is in a specific project should depend on research discussed further in the Appendix. objectives, budget, duration and the need for 12 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

2.4.3 Absolute vs. relative most disadvantaged, would believe. It may be poverty considered more important to buy a television Absolute poverty is defined in terms of the like their neighbour’s than to improve their basic satisfaction of basic needs. It can be measured diet or fix their roof. Subsistence levels are often for instance, in terms of the amount of people defined in nutritional terms but what is considered living below a time-invariant poverty line a ‘minimum consumption requirement’ also has defined by income, consumption or nutrition. strong cultural and social roots. This emphasises In contrast, relative poverty implies a ranking a relative poverty dimension, sometimes labelled with respect to a particular poverty indicator, the ‘social contingency of wants’ (Brekke and be that income, consumption, land area or Howarth 2000). subjective well-being. Notably, this implies that What about the relative dimensions of non- economic growth and other progress in absolute income factors, such as ‘security’ and welfare indicators reduce absolute poverty just ‘empowerment’? The security dimension is an as long as the poor receive some benefits, even absolute concept, related primarily to basic needs. when other income groups benefit more from The power dimension is, by definition, a relative that progress. On the other hand, relative measure. To give more power to poor people poverty would increase in this situation, as the implies necessarily a redistribution of power from share of the poor in total welfare would decline. other agents. However, sometimes the term They would become more disadvantaged in empowerment is also employed more broadly in comparison to the better off income strata (e.g. ways that involve capacity building, which does Fields 1980 and Section 3). not necessarily imply a zero-sum game. How relevant are the two types of measures for subjective welfare at different income scales? In the economic history of developed 2.4.4 Subjective vs. objective countries, absolute income increases have poverty measures proven to be most welfare improving at low A subjective assessment of poverty would income levels, while relative measures gain seemingly imply that we accept our target importance at higher income levels (see above). group’s preferences and “let people The very poor care more about their basic needs, themselves decide” to what extent and in what for example, getting more and better food on respects they are ‘poor’. At first sight, this their table. However, as they become slightly looks like an honourable and irresistibly better off, they increasingly have an eye on their participatory principle but it can have some neighbour’s table. If the servings are growing unpleasant implications. Firstly, different faster on the table next door, they start to feel ‘filters’ and strategic behaviours often make poorer for failing to ‘keep up with the Joneses’, people conceal their true poverty situation and even though their individual basic needs preferences to outsiders. Secondly, local fulfilment may have actually improved. poverty perceptions may differ dramatically, Is this also a correct characterisation of even vis-à-vis those stated in the neighbouring poor forest dwelling people? Do the poorest of village. This will jeopardise any type of them care more about absolute income and basic comparative poverty assessment. needs, while the moderately poor start to Thirdly, the subjective short term compare their living standards with the people preferences of local people may be neither around them? Arguments can be raised for and consistent with their goals of long term welfare against but this would probably be a wrong improvement, nor coincide with external values. generalisation. From CIFOR’s work in East Many researchers feel uneasy, for example, when Kalimantan, as elsewhere, there are examples observing that cigarettes are the single biggest of income poor but fairly egalitarian societies cash expenditure item in rural households. where relative income is highly relevant for Sudden large cash inflows can reduce poverty people’s welfare. Typically, the horizon of remote (in terms of materialist constraints) but can also forest dwellers changes when they increase their lead to social disruptions related to alcohol contact with the market economy and their consumption, prostitution and luxury items for material consumption options are enhanced. The short lived benefits, especially when these ‘keeping up with the Joneses’ phenomenon is monetary incomes are solely appropriated by thus also present in poor societies, more than males. This ultimately raises difficult questions many Westerners who focus on basic needs as about how preferences and perceptions about being the sole key to improving the lot of the ‘the good life’ are formed—and who is to judge. Exploring the Forest—Poverty Link 13

In such a situation, it may be preferable to state higher intensity of use and average (per that “poverty was reduced” and that “other hectare) returns. This matter has become more welfare measures declined”. At the micro scale, important especially in forest-poor environments: reduced poverty (in terms of income) does not “Many ‘forest’ products are no longer drawn from always automatically result in increased human forests” (Arnold and Townson 1998). well-being. Confusion about the natural origin of products and incomes can also come to exaggerate claims about the economic 2.4.5 Forest definitions and poverty importance of forests, which would ultimately relevance misguide land use decisions. For instance, one Finally, it is not only the choices of poverty often sees household incomes derived from definition and indicators that matter. The ‘wildlands’ (including woodlands, savannahs and selection of forest and forest product definitions grasslands) generally referred to as ‘forest will also have a bearing on the poverty relevance incomes’. The renowned study by Cavendish of forest research and action. At one extreme, (1997) on drylands in Zimbabwe is often cited if a very narrow definition is used, where only for an average per household forest income share primary or ecologically intact tree vegetation of 35% but that portion actually refers to what is counted as forest, its relevance for poverty the author calls ‘environmental resources’, reduction on a grand scale will be limited. On including not only non-forest wildlands but also the other hand, if the forest definition is made certain agricultural and mining outputs. so broad that it includes plantations, areas that Excluding the latter but still including the value are fragmented or degraded and even trees of livestock grazing in the woodlands category, outside forests (‘forests and trees’), then the the income contribution shrinks to 20%.7 In other poverty relevance tends to increase. This is words, care is needed in defining explicitly what because more people usually live in and around resources are included—and in referring to the disturbed and converted ecosystems, with results of others in a correct manner.

For poor rural villagers, forest products can serve as safety nets in times of need. Forest dwelling family in Cameroon, Africa. (Photo by Edmond Dounias) 14 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

3. Causes of Poverty (1) In most cases, macroeconomic growth actually raises the income of the poor Discussions on how forests can contribute to and reduces the number of people below reducing poverty are often based on vague the poverty line. Growth ‘trickles implicit assumptions about the more down’—at least in the medium and long fundamental causes of poverty and conversely, term and at an aggregated (national) about the sources of reduced poverty. As scale. indicated by Adam Smith’s classic “The Wealth (2) In a minority of deviating cases, little of Nations”, published in 1776, this topic has or no poverty reduction is achieved been central to social science for centuries. In through growth because of a skewed this section, we will refer to recent empirical initial asset distribution and/or ‘bad- work on economic growth, poverty and quality’ economic growth. This inequality. The purpose is to bring on board some development pattern is characterised basic lessons from the economic development by low labour intensity, low human- debate—lessons which despite their undeniable capital accumulation, rural neglect, relevance, are often ignored when arguments high corruption, etc. and claims about forest(ry) based development are made. This emerging consensus thus provides significant support to the developmentalist view, while it downplays the class-based approach 3.1 Poverty and growth without totally discarding it. A strong point in Historically, two major positions stand out as case is that it is hard to empirically point to extremes in the discussion about the causes of countries where poverty alleviation has been poverty: achieved without economic growth or with very (1) The developmentalist position low rates of growth. This is because the political- explains poverty in terms of lack of economy obstacles of redistribution are much economic advancement, normally larger if one needs to take away a piece of a equated with insufficient economic stagnant sized cake from the rich, rather than growth. have them accept that they will receive less of (2) The class-based (Marxian inspired) whatever increment there is in the cake. theories view poverty as a result of uneven development and exploitation, resulting in skewed asset and income 3.2 Inequality and growth distribution. Inequality is important for the poverty discussion in two major ways. As explained in Section 2, According to the first view, the problem is people living in and around forests, just like that the cake is too small, while the second other people, do not only care about their own suggests the cake has been cut unevenly. While absolute incomes. Their welfare perceptions also the first view regards poverty to be an efficiency depend on their income status vis-à-vis ‘the or growth problem, the second sees it as a Joneses’—the individuals, households or distributional one. communities of reference that they would like The two views produce fundamentally to ‘keep up with’. In other words, inequality different predictions about whether economic levels and changes induced by economic growth growth and structural change help to reduce poverty also matter for perceptions of poverty or well- or not. The ultimate Marxian view would hold that: being. “In countries at low levels of development, any In principle, continuous economic growth kind of structural change such as industrialization under a constant income distribution8 must or expanded commercialisation tends to increase necessarily reduce absolute poverty, defined as poverty among the poorest members of the the proportion of people below an income- population” (Adelman and Morris 1978). However, poverty line.9 In practice, however, economic much empirical work over the last decade, growth usually does change the income facilitated to a great extent by better income distribution—the extra pieces of the growing and consumption data at different levels of cake are not distributed in clinically equal relative aggregation, has enlightened this debate. At shares to all members of society. A number of present, there is a growing consensus (e.g. Fields processes at different scales can explain that. 2001; Ravallion 2001) that: Exploring the Forest—Poverty Link 15

Micro and macro level processes by the Gini coefficient10) in this trend scenario Most development practitioners know the will follow an inverted U: it will rise initially and following situation. When a development project then start to fall (Williamson 1985). This in a remote forest community is initiated, trying dualistic process has high relevance for many to open up new channels of income (and other developing countries. For example, relating to benefit) generation, not all community members changes in the rural-urban or formal-informal benefit equally. Some have assets or skills that sector distribution over time. Conversely, put them in a privileged position to take advantage particular developing countries without dualism, of the new opportunities. Others simply have a such as the urban enclaves of Hong Kong and more developed entrepreneurial spirit, less risk Singapore, were able to significantly reduce aversion and/or a greater willingness to invest income inequality right from the outset of their resources in a new, still experimental activity. high growth periods, without an initial rise in Consequently, they respond more quickly and reap inequality (Wunder 1987). higher benefits—or suffer higher losses if the The inverse U-curve can also be supported project fails. Although the project manager may by other societal trends over the course of try to counteract this tendency by ‘targeting’ economic development. During the early stages certain groups or by promoting a broader of development, skills and higher education are ‘participation’, one often witnesses a rise in limited to a small group of people, who benefit inequality. the most from economic growth. Over time, the This mechanism tends to be particularly greater spread of skills and secondary and pronounced in response to new market tertiary education will have an income equalising opportunities. At the beginning of an economic effect, through reducing large skill premiums. development process, higher market contact Similarly, development of insurance markets often increases inequality as individual actors tends to improve risk management. Sharing risk respond to favourable prices or other newly between economic agents becomes another emerging opportunities. This may also impact equalising force. the social coherence of communities and From a forestry point of view, Bornschier ultimately, have some negative welfare effects. (1983) has suggested an interesting hypothesis: Obviously, this is what many researchers worry the inverse U-curve may be linked to shifts in about at the micro level, justifying scepticism the society’s dominating production systems. towards economic growth. On the other hand, Inequality thus rises in the transition from hunter equalising factors may start to work over time: gatherer to subsistence agriculture and further, more people acquire the necessary skills or to commodity oriented cash crop agriculture, assets, new technologies spread to more implying larger production scales and a gradual producers, more efficient markets eliminate concentration of assets. Subsequently, in the price differentials across locations, demand for transition to industrial systems, political and unskilled labour increases and the higher income economic pressures towards income and asset has local multiplier effects (e.g. increased demand redistribution increase, which eventually for locally produced commodities). These are produces a decrease in inequality. some of the micro-level mechanisms that have justified the hypothesis that inequality follows an inverse U-curve over time—inequality rises in Empirical evidence the initial development phase but then declines. Is inequality actually following an inverted U- Similarly, there are macro-level processes curve—or is it just a theorist’s fata morgana? Do that can ‘drive’ such a development in inequality things have to get worse (more unequal) as an over time. Imagine a country with a large, low initial result of growth and structural change wage traditional sector employing 99% of the before they eventually can get better (more labour force and a tiny, high wage modern sector equal)? ‘island’, accounting for 1% of employment. In his pioneering article in the 1950s, Simon Imagine also, that the wages in each sector Kuznets—the ‘father’ of the inverse U-curve or remain fixed, so that economic growth in this Kuznets curve—observed this empirical pattern dualistic economy can only be achieved by over time in three developed economies (the modern sector enlargement: 3% of the total United States, Germany and the United labour force are transferred each year from the Kingdom). He also noted that inequality levels traditional to the modern sector. It can be shown were somewhat higher for some low income that national income inequality (e.g. measured countries (e.g. Ceylon, Puerto Rico and India) 16 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

and considerably higher for a few developing by about 186 million over the last 30 years (Sala- countries (e.g. Kenya and Rhodesia) which, at i-Martin 2002). that time, had relatively rapid growth rates Generally speaking, the income distribution (Kuznets 1955). Later, Kuznets backed up his time series data is too poor to fully resolve the conclusions with time series for 16 countries, 9 Kuznets controversy. In some cases, inequality of which were developing ones (Kuznets 1963). rises, in others it decreases (Ray 1998; Fields Over the long term, inequality seemed in most 2001). Why do the micro and macro level cases, to be reduced well below the ‘initial’ level: processes from above not consistently produce “The distribution of income tends to be more Kuznets curves at the national level? One equal the longer and more thoroughly the country possibility is that aggregated national data has been exposed to the processes of economic ‘conceal’ micro processes by adding data for and social change associated with the idea of regions at internally different development industrialization” (Kravis 1960). stages. Another is that the preconditions and The Kuznets curve has been seen by many, types of growth eventually matter more than as one of the empirical regularities in economic income levels in regard to how economic growth development. However, since the mid-1990s, changes the income distribution. This is the many scholars have questioned the existence topic of the next section. of such a pattern (Deininger and Squire 1998; Ravallion 2001; Fields 2001). Due to the lack of reliable income-distribution time series 3.3 Pro-poor preconditions and information from developing countries, there growth types are huge problems in testing any hypothesis To sum up so far, macroeconomic growth in most related to income inequality. Many researchers cases ‘trickles down’ to raise the absolute have thus tested the Kuznets hypothesis from incomes of the poor, at least over time and at cross-country data. Yet, the problem is that the aggregate scales. In general, the poverty record inter-country differences are normally much of economic growth is thus better than its larger than the variation determined by the reputation. Fields (2001) concludes from his development phase, so the cross-country pattern literature review: “It is overwhelmingly the case therefore cannot be given a time series that growth reduces poverty and recession interpretation. Specifically, some middle income increases it, though in about 10 percent of the countries in Latin America and the Middle East cases, poverty did not appear to fall when are very unequal and come to dominate the cross- growth took place”. Likewise, Ravallion (1997) country pattern—in spite of the very limited true notes on the Human Development Reports (see changes over time and income that have Section 2) that “Arguably the biggest problem occurred within these countries. Hence, the U- facing the world’s poor today is not ‘low quality curve critics argue that the income distribution growth’—in HDR terms—but too little growth of tends to be very different across countries but even quite normal quality!” fairly stable over time. Thus the structural This also means that we should not be differences across countries can create the universally prejudiced about situations where illusion of an inter-temporal Kuznets curve. ‘the size of the cake increases but the poor only Nevertheless, some maintain a general receive the crumbs’. One cannot dismiss all support for the Kuznets curve (e.g. Williamson these scenarios as undesirable. Sometimes, the 1997). China provides an affirmative case of a indirect pro-poor effects from forest derived dualism-type Kuznets curve. After 1984, income profits could be large—though not necessarily inequality rose as a result of a widening gap to the benefit of those poor who live in and near between the high growth urban/coastal areas the forests. In any case, country experiences of and the rural hinterlands (Lindert and Williamson growth cum poverty reduction are variable. Can 2001). Inequality increased in China (and in this be explained by the countries’ different Indonesia and Russia) “by differential access to preconditions and growth paths, respectively? the benefits of the new economy, not by widening gaps among those who participate in it, or among those who do not” (ibid). Although Preconditions inequality in China has increased during the Strong evidence suggests that an egalitarian recent high growth periods, trickle down effects asset distribution—in particular of land and have still made the poor better off in absolute human capital—will enhance the poverty terms. The number of Chinese who are poor by reducing effect of growth. The combination of the ‘one dollar a day’ standard has decreased land reform, together with improved Exploring the Forest—Poverty Link 17

infrastructure, education and labour intensive Indonesia’s New Order policies (Hill 1992; Warr urban growth, was a key success factor in post- 2000). Conversely, countries that try to ‘jump- WWII South Korea and Taiwan, in dramatically start’ modern sector development with strong reducing both poverty and inequality (Wunder urban biases tend to reduce poverty to a much 1987). lesser degree. Across Indian states, the variable degree of literacy is the prime factor explaining (3) Labour intensive technologies different growth-poverty outcomes (Ravallion While education is the prime pro-poor asset to and Datt 2002). In terms of the reverse causality, endow the poor with, unskilled labour is the most high asset inequality in poor countries can also important asset they already possess. Growth obstruct growth in the first place (Barro 1999). paths that favour a higher demand for unskilled Conversely, a population with widely distributed labour will tend to raise employment and wages, skills across regions and socioeconomic groups which helps to alleviate poverty. Such labour is an important growth promoting asset. Growth intensive sectors can be rural (e.g. small scale thus tends to make economies with an initially agriculture or wood processing) or urban, with equal income distribution more equal, while the latter including industries like textiles, conversely it can also reinforce large pre-existing electronics or construction and services such as inequalities (Ravallion 2001).11 tourism or commerce. Conversely, growth strategies that are highly intensive in the use of natural capital (e.g. mining or oil) or in man Pro-poor growth made capital/high skill technologies, are We know from the work in UNDP (Human unfavourable to the poor, especially in early Development Reports) and the World Bank that development phases. the type of growth path matters (i.e. that the ‘quality of growth’ (Thomas et al. 2001) has (4) Technologies for pro-poor a bearing on poverty). What types of growth consumption goods then, are most likely to be pro-poor? Obviously, The poor not only benefit from growth in their tailored approaches are always preferable to nominal incomes, but also from technologies one size fits all statements but in cutting that make the goods they consume cheaper in through the debate at a general level, it is the marketplace, thus increasing their beneficial for poverty reduction to have a purchasing power. Staple crops are a primary growth path emphasising any combination of example. The evidence suggests that urban the following five factors: consumers, not rural producers, capture the lion’s share of the benefits from new agricultural (1) More education technologies. Increased food supplies tend to Education is a direct and embodied investment lower food prices because food markets in in poor people—once the asset has been built, developing countries tend to be competitive and nobody can take it away from the poor. Education elastic demand causes prices to decline. This is and particularly improved and more extensive why the Green Revolution was good news for primary education, represents a valuable pro- poor urban consumers—their staple crops poor human capital. For instance, Sen (2001) became considerably cheaper. argues that one main difference between the recent growth paths of China and India has been (5) Good governance the larger education emphasis in the former, Clear legislation and transparent government triggering greater poverty reduction. Investment policies that recognise (the often informal) rights in education is the single most important factor of poor people over land and resources are in simultaneously promoting economic growth, important in creating benefits for the poor. In poverty reduction and a more equal income regard to forests, an important negative example distribution (Thomas et al. 2000). is the extraction of high rent products from rich natural forests, where valuable economic rents (2) Rural focus (defined here, as high profits from exploiting Rural growth tends to reduce inequality in natural resources) are often captured by elites particular, especially in the early development who manage to influence policies and rules in stages when the bulk of the poor are actually their favour (see later discussion). Specifically, rural based (Mellor 1999). Investments in rural it has been shown that a high incidence of infrastructure, technology and R&D can help corruption has a strongly anti-poor effect support this strategy, as for instance, under (Thomas et al. 2000). 18 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

4. Using Forests to Cushion supplementary way. The latter would include not only primary forest users but also those among and Reduce Poverty the poor who trade, further process and consume forest products, including those in urban areas 4.1 The forest-dependent poor (Byron and Arnold 1999; Kaimowitz 2002). “More than 1.6 billion people depend to The forest benefits to poor people can be varying degrees on forests for their categorised in a number of ways. We suggest livelihoods. About 60 million indigenous five dimensions. One approach is to look at people are almost wholly dependent on different groups of beneficiaries. A second is to forests. Some 350 million people who live evaluate the types of forest products and within or adjacent to dense forests depend services provided. A third way is to differentiate on them to a high degree for subsistence the role of forest benefits in the household and income. In developing countries about economy or livelihood strategy (e.g. subsistence 1.2 billion people rely on agroforestry use vs. cash income; gap filling vs. regular use). farming systems that help to sustain A fourth dimension refers to the extent of agricultural productivity and generate resource management (i.e. from pure natural income. Worldwide, forest industries forest extraction to lightly managed and planted provide employment for 60 million people. forests/forest products). Finally, a fifth Some 1 billion people depend on drugs distinction is between high and low rent derived from forest plants for their products. All these dimensions have important medical needs.” (World Bank 2001) bearings on the size and type of poverty alleviation potential and will be dealt with This recent quote from the World Bank throughout this section. In Table 1, we combine reflects a widespread desire to put numbers the first two dimensions (user groups and types on things that are thought to be important—in of benefits), while in Section 4.2 we discuss the this case, how important forests are to other three dimensions. different groups of people. While both We structure this section according to the environmentalists and forestry advocates may type of benefits concerned. One of the forest welcome this type of exercise, its empirical benefits in Table 1 does not involve traditional foundations are shaky. It is no accident that forest products and services—but converting these guesstimates are not backed up by forests (temporarily or permanently) to specific sources—to the best of our knowledge, agricultural land can be one of the most valued the sources simply do not exist. Is it likely that benefits locally. ‘Shifting cultivation’ represents one fourth of people on Earth (or about half of an intermediate case between permanent the global rural population) are ‘forest- conversion and conventional forest products, dependent’? Does ‘forest dependence’ mean where soil recuperation and fertilisation through everything from ‘generating most of household biomass burning and decay are an essential part income’ to ‘generating any household income’ of the process. Including such agricultural uses and to ‘using any kind of forest products’? in the debate also raises the point for discussion A recent study commissioned by the British about how much forest is ‘optimal’, but this is Department for International Development beyond the scope of this current paper.15 (Calibre Consultants and SSC 2000) set out to In the following section we concentrate on quantify the number of forest-dependent three areas: non-timber forest products people but in the end, had to give up. (NTFPs), timber and environmental services. ‘Dependence’ criteria are too variable and the In some cases, the distinction between timber statistical sources are very deficient. Byron and and everything else (NTFPs) is artificial, which Arnold (1999) came to the same conclusion. is why we distinguish various sub-categories So, just how many people depend on forests? and dimensions of NTFPs. Some of the NTFP The first answer is: that depends completely discussion is also relevant to small scale timber on how ‘dependence’ is defined. The second use by poor people. At the same time, the two answer is: due to a lack of any comprehensive debates have differed and NTFPs and timber data sources, we do not know for sure but are characterised by distinct production probably tens of millions of people depend on systems—which has a direct bearing on the forests as a dominant source of subsistence types and magnitude of benefits obtained by and cash income, while hundreds of millions the poor. depend on forest products in some Exploring the Forest—Poverty Link 19

Table 1. The Importance of Different Forest Benefits to Different Groups12

User groups13 Types of economic benefits

(A) Agricultural (B) NTFPs (C) Timber (D) On-site land & nutrients ecological services14 (1) Forest dwellers: i. Hunters and gatherers Minor benefit Main benefit Supplementary if Variable transport access exists ii. Shifting cultivators Main benefit Important supplement As above Variable

(2) Farmers living adjacent to forests: i. Smallholders Major ‘land reserve’ Supplementary Supplementary if Variable transport access exists ii. Landless Not important Important supplement As above Variable

(3) Commercial users: i. Artisans, traders, None Important Important None small entrepreneurs ii. Employees in None Supplementary Main benefit None forest industries

(4) Consumers of forest products: i. Urban poor None Some ‘pro-poor’ products Variable None

4.2 Non-Timber Forest Products and often fill income gaps, however they are (NTFPs) seldom managed more than ‘softly’ (few planted products). There is a strong correlation between 4.2.1 A heterogeneous group NTFP dependence and poverty, a fact we will NTFPs cover a wide range of products, which discuss in Section 4.2.2. are utilised in very different contexts and play different roles in household livelihood strategies. Global comparison of commercial NTFPs Four key dimensions (from above) are: Some NTFPs have a high commercial value for (1) Degree of management and those involved in their production. Over the last domestication: from pure extractivism few years, CIFOR has undertaken detailed data to intermediate systems (‘soft collection and comparative analyses of 61 cases management’) and fully domesticated involving commercial NTFP production and trade products from Asia, Africa and Latin America. These (2) Subsistence uses vs. commercialisation cases were grouped according to their role in (3) Safety net function (gap filling and the household economy, defined as: the insurance) vs. regular income source contribution of the product to total household (4) Low vs. high returns—to labour, land income (importance of product) and the and sometimes capital. proportion of total household income earned in cash (degree of market integration) (Ruiz-Pérez With this wide diversity, one might object et al. 2003)16. Using a cut-off line of 50 % for that any generalisation about NTFPs would be these two variables, three categories emerged:17 speculative. Yet, existing overviews (Byron and (1) Specialised strategies—high forest- Arnold 1999; Neumann and Hirsch 2000) suggest product contribution to household that most NTFPs produce low returns (particularly income and high household integration in terms of return per ha and often, also per into the cash economy. labour unit), are primarily used for subsistence 20 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Non-timber forest products serve subsistence needs, can have important gap filling or safety net functions and sometimes provide regular cash income. Here, women are selling plants (Gnetum africanum) collected from the forest, maize and cassava flour at the Mfoundi market in Yaoundé, Cameroon. (Photo by Michael Hailu)

(2) Diversified strategies—low forest- nuts (Bertholletia excelsa H.B.K.) in the Bolivian product share but high market Amazon, bamboo in China and cardamom integration. (Elettaria cardamomum Maton) in the Western (3) Coping strategies—low forest-product Ghats of Kerala state, India. share and low market integration. The diversified category mostly has intermediate characteristics. People produce In the specialised strategy, absolute income many other products for the market or are of forest-producing households was about one engaged in wage labour—forest products make third higher than the average (that is, forest up a lower share of their household income. Real- producers were better off than their non- world examples are found in all three tropical producing neighbours). The specialised strategy continents. Producers of benzoin (Styrax is associated with a number of favourable paralleloneurum Perk) in Tapanuli, North Sumatra, conditions: high value-weight ratio, stable Indonesia, use different management strategies, product markets, steady household involvement from forest extraction to agroforestry type in NTFP production, a low level of product techniques, depending on their orientation within adulteration, a stable resource base and the cash economy. The Panamá hat industry in interestingly, a lower incidence of customary South Ecuador has similar income-supplementary resource rules. The products tend to be produced characteristics and so too, do the woodcarving in areas where land tenure is effectively under industries in Kenya and Uganda. state administration—the producers have legal In the case of the coping strategy, household rights and do not rely on customary rules. economies are more subsistence-oriented and The specialised strategy is also associated the commercial forest products play a lesser role with more intensive management, sometimes (although the households tend to use many other by domestication (plantation) but for the NTFPs for their own consumption or sale). majority of cases, perhaps a bit surprisingly, However, this category also represented the most through ‘softly’ managed forests. Plantation dynamic group, as people use a wide variety of products are found in only one third of the products over time due to changes in the resource specialised cases, though they have higher yields base and prices, and the emergence of and fetch better prices. Examples include Brazil alternative income opportunities. Some cases Exploring the Forest—Poverty Link 21

in point include the production of the Prunus medium returns to labour; (2) Low capital and Africana bark for medicinal uses in Cameroon skill requirements; and (3) Open or semi-open and cardamom (Amomum villosum) in Vietnam. resource access. These features reflect the Note that in spite of a basically similar product rudimentary developmental characteristics of as in the specialised case of Kerala, markets forests. Most of these features favour poor were much less consolidated in the Vietnamese people, who have no access to markets for case, thus producing a different contribution to skilled labour (otherwise, they would hardly be the household economy. Thus, not only the poor in the first place) and hence have a low product as such matters—its specific role in the opportunity cost of labour. The poor also lack local economy can vary. the cash or access to credit to involve themselves In geographical terms, the African cases in capital-demanding self-employment tend to have lower trade volumes and less activities. With a lack of assets or political power income but also, growing populations in the case to get access to more valuable resources, they areas and rising market demand—which turn to a low rewarding ‘employment of last increases pressure on the resource. Africa is resort’ (see Box 1). associated with the subsistence strategies and The high level of involvement of poor people Asia with the specialised strategies. The Latin in NTFPs must also be understood in the light of American cases tend to fall somewhere in risk aversion, which induces diversification. In between the two and are mostly associated with fact, the presence of high risks impedes a diversified strategies. These research results greater degree of specialisation in the most thus substantiate a set of common, intuitive rewarding activities and can thus be a main causal perceptions about the regional specificity of factor of poverty per se. Forest products are forest-products and their differential normally not the principal cash earners of poor contribution to the household economy. households. On the other hand, they can have important ‘gap filling’ functions. Unlike regular income contributions, these benefits accrue 4.2.2 ‘Poverty traps’ or unevenly over time. “The importance of forest ‘safety nets’? product income is usually more in the way it fills There is solid empirical evidence on the positive gaps and complements other income, than in link between rural poverty and NTFP its absolute magnitude or share of overall dependence. In an excellent survey, Neumann household income” (Byron and Arnold 1999). On and Hirsch (2000) noted the “overwhelming the subsistence use side, “forest products evidence that the poorest segments of the seldom provide the staple, bulk items that societies around the world are the populations people eat” (ibid). However, NTFPs are widely principally engaged in NTFP extraction”. This used to overcome seasonal food shortfalls (e.g. correlation can be approached from two different before the main harvest) or serve as substitutes angles. If we consider poverty to be exogenous, during emergencies. the question becomes: Why are poor people Analytically, when we speak about forest forest-dependent? The poor often use forest ‘safety nets’ and ‘gap filling’, two slightly products due to the (permanent or temporary) different types of benefits seem to be involved. lack of better alternatives. We can then use the On the one hand, for example, a household correlation to argue that forest products function gathers certain forest fruits each year in the as safety nets. On the other hand, if we consider months between staple harvests. These fill a poverty to be endogenous, the relevant question gap in the sense that they provide a periodical, is: Why are forest-dependent people poor? reasonably predictable contribution to food Perhaps most forest products are bound to be security (and possibly also cash income), serving economically marginal, so they per se offer a as a seasonal buffer. Seasonality can also be poor potential for value generation. In that case, induced by employment cycles, i.e. some forest could forest products be considered a poverty collection is carried out only in seasons when trap rather than a safety net? family labour is abundant. Part of the comparative advantage of NTFPs is therefore Safety nets? that they increase the seasonal flexibility of rural Back to the aforementioned question—why livelihoods. should poor people rely relatively more on forests On the other hand, there are income and forest products? We first reiterate the basic shortfalls that are much harder to predict, such characteristics of most NTFPs: (1) Low or as family illness, political turmoil, macroeconomic 22 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Box 1. The ‘Employment of Last Resort’ Model

Why are the poor forest-dependent and the forest- What happens if the price of the forest dependent poor? Many forest activities have low product in question goes up, that is, the entry costs (open access) and few requirements downward sloping line shifts upward? Forest in terms of skills and capital. Such characteristics income goes up but less than the price increase are attractive to the poor but they also lead to because new households are attracted into the the activities becoming an ‘employment of last forest and that in itself will push the income resort’, due to their low return. down. This model can be illustrated graphically. What happens if new employment First, assume that we rank all the households in a opportunities are created, i.e. the opportunity village according to the opportunity cost of their costs curve shifts upward (at least in the range labour, that is, the best non-forest income around the equilibrium)? Some previously forest opportunity they have. This is shown by the rising dependent households will be able to earn more curve, i.e. those with the lowest opportunity cost elsewhere and will leave the forest. But also, are found to the left. The opportunity cost of the remaining poor will benefit when the number labour is strongly correlated with income, so the of households involved in forest activities drop. poorest households will tend to be to the left This model therefore demonstrates an indirect side of the figure. The declining curve shows trickle down effect. forest income per household. This income is a The model provides a simple exposition of function of many factors: the resource base, the ‘employment of last resort’ approach but as technology, market prices, etc. It also depends all models do, it simplifies reality and the factors on the number of people involved in a given forest that make the poor forest-dependent. Some of activity. When more households become engaged, the underlying assumptions behind this model at a certain point competition for the resource are: (1) All households receive an equal share of will drive down average incomes, which is why forest income; (2) Forest income is negatively the curve is downward sloping. correlated with the number of people involved, How many households will get involved in i.e. there are no net economies of scale in forest activities? Villagers will choose the activities management, processing or marketing, etc.; and that provide the highest return for their labour. (3) We do not take into account the seasonal Hence, households with the lowest market value fluctuations of labour demand. A more of labour will get involved in forest activities, i.e. sophisticated model would include these features those to the left of the point where the two (see also discussion below) but this would not lines intersect. They will earn more than their alter the underlying logic of the model: opportunity costs of labour. Those that do not households with a poor asset endowment resort get involved are better off working elsewhere. to low-return activities—from forestry or other Thus, the poorest households will tend to be ‘last resort’ activities, while those with better forest-dependent and vice versa, forest activities opportunities find employment elsewhere. only attract poor households. Exploring the Forest—Poverty Link 23

crises or ecological disasters. Forest products complicating market access. This is even more can have important roles in overcoming these problematic when the NTFP concerned is unpredictable shortfalls, either as a reservoir perishable. Usually, only a few high value of auto-consumption goods or as a source of products can thus pay their way out of the forest. ‘quick cash’ raised from collecting forest The net benefits per area unit thus tend to be products and taking them to the marketplace. too low and/or too short lived to justify the Can we simply quantify these benefits by transaction costs involved in establishing and summing up their average contribution to enforcing property rights to land and resources. subsistence and monetary incomes? No, because Yet, without secure property and access rights, the ‘safety nets’ imply more than that. They there will be little incentive to invest and can make the difference between good and bad undertake more intensive management, which nutrition, between recovered health and would potentially elevate returns. Consequently, prolonged illness or between food security and NTFPs commonly continue to be collected under starvation. Ultimately, safety nets help people de jure or de facto open-access conditions. to survive. Thirdly, some forest-product trade is Nevertheless, a warning is necessary. Any characterised by monopsonies and exploitative on-site assessment of forest based safety nets marketing chains, due to one or several also has to consider whether these are truly characteristics.18 The heterogeneity and quality irreplaceable over time. In many cases, in the differences of some products, combined with a course of reduced forest access, poor people lack of well defined standards, can lead to manage to build alternative safety nets. They intransparencies and manipulations in the may buy a couple of cattle or diversify their marketing process. Perhaps most importantly, agricultural production. They may also be able the remoteness of many forest communities to obtain seasonal off-farm employment or send reduces their level of incoming information a son to work in the city so he can send home about changing market prices and their remittances. In other words, forest based safety bargaining power vis-à-vis there only being a nets may be only one among several safety nets few traders, is low. Hence, not only do forest available to the poor household. products face constraints in their aggregate value. The remote and isolated producers also Poverty trap? tend to receive a low share of that reduced value. Turning the above question around—why are Obviously, this trinity of imminent misery forest-dependent people poor? One can similarly does not apply uniformly to all NTFPs. There are put forward several arguments to explain the examples of products that break the deadlock forest dependence-poverty link. The main and produce sizeable profits over and above the problem is the low returns from most NTFP average forest-product returns. So, what activities, as discussed above. So, why do NTFPs happens in this situation, where there is a tend to be low value resources? In spite of their sustained increase in market demand and prices? ecological richness, natural forests are frequently Many studies have pointed out that the poor do economically inferior production environments. not necessarily benefit. In the following, we One type of argument here concerns the resource outline four different reasons. per se. For instance, low per hectare densities First of all, a more valuable resource will of commercially valuable species imply that suddenly become increasingly attractive to the extraction tends to be spread over large areas, more powerful external stakeholders who triggering high costs of harvesting and low net previously ignored it. They now have an returns to extractive labour. There may be sharp incentive to actively exclude the poor from access and unpredictable year to year fluctuations in (Dove 1993). There are several ways in which the production from trees and other plants, forest the rich can capture the resource rents. Through animals may follow migration and many products their political connections and power, they can have a heterogeneous quality that it is difficult redefine the rules of access to those forests for the extractor to control. If market demand where tenure was informal or weak. They can for a forest product is increasing, there may be also control marketing and processing chains large physical obstacles relating to the increase through their networks and capital. This line of of yields in a natural forest environment. reasoning links to the more general argument Secondly, remoteness and low population that it is difficult for poor people to gain and density often mean that physical infrastructure maintain control over resources that are (roads, rail, bridges etc.) is poorly developed, producing high economic rents. 24 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Secondly, higher resource scarcity and interventions to advance the case of poor corresponding values of forest products can trigger producers of commodities from (natural) forests. internal processes of differentiation among the So, does this make non-timber forest traditional producers, resulting in economic products a ‘poverty trap’? To us, it would seem inequality (see Section 3). Scarcity promotes unfair to use this label in a general manner, given individual property rights, replacing more that NTFPs in many cases provide important inefficient yet egalitarian common property livelihood contributions to people with limited regimes (e.g. Eggertsson 1990). Paradoxically, or no alternatives—contributing to their income adding value to NTFPs might thus lead to the and providing a safety net against a series of collapse of the institutions that made them risks. Although most NTFPs are poor instruments attractive to the poor in the first place. “The for poverty reduction, some are important for increased commercialisation of NTFPs is likely poverty prevention. Preventing extreme poverty to eventually lead to the breakdown of common and helping people survive can hardly be called property systems and a trend towards individual a poverty trap. However, “care needs to be taken private property” (Neumann and Hirsch 2000). not to commit communities to institutional Even in intra-household terms, men will often arrangements which they are unlikely to be able begin to capture resources traditionally controlled to sustain once the incentive of reliance on by women when the resources become more forest products diminishes” (Arnold and Townson valuable, thus adding a gender dimension to the 1998). The ‘trap’ label would be justified in those power over resources. cases where alternative development options Thirdly, high economic rents may lead to actually exist but where policies, donor projects overexploitation and collapse of the resource or other external interventions seek to maintain base of a common property resource. As already people in their low yield forest extraction mentioned, supply from natural forests is normally activities, based on romanticised visions about inelastic—it is difficult to significantly increase the alleged ecological sustainability and large production out of a fixed size forest area. In the income generation potential of NTFPs. absence of well enforced management regimes involving all user groups (and excluding outsiders from entry), higher prices might enhance a 4.2.3 Fuelwood—the neglected pro- ‘tragedy of the commons’. poor NTFP? Finally, as a related response to scarcity, In our reading of the literature, wood based fuels higher prices can make it profitable to (firewood and charcoal) often fall into the ‘no domesticate the resource. That is, eventually man’s land’ between non-wood forest products moving the resource supply out of the forest and and timber. Thus, governments, private firms gradually replacing extraction with cultivation. and donors have mostly ignored them in recent While this may increase aggregate values and times—especially since the reappraisal of what, profit margins, it also requires land, capital and in the late 1970s, was erroneously perceived to skills, which may imply that the more well be a global fuelwood crisis. However, fuelwood endowed producers become better positioned. continues to be vital for a lot of poor producers Alternatively, sustained high price levels can also and consumers, with the bulk of rural households make it worthwhile to develop different products— in developing countries using firewood as a synthetic or natural substitutes. Both latter domestic source of energy. Yet, a recent literature options raise overall supplies and bring down survey by Arnold et al. (2003) shows that its prices but they also reduce the market shares of commercial role can also be significant. For both forest-derived products and poor forest example, in peri-urban areas of sub-Saharan based producers (Homma 1992). Africa, tens of thousands of poor farmers and In all four instances, forest products are part small traders supplement their incomes by selling of a deadlock scenario. They yield low returns fuelwood. Sometimes this activity even becomes and paradoxically, trying to raise these returns their main source of cash income. Notably, this might make the situation worse for poor group also includes the poorest of the poor. For producers, unless underlying (and often complex) instance, many rural landless people are among socioeconomic and political constraints are those specialising in fuelwood production. addressed simultaneously. Although this picture The most dynamic commodity is charcoal, represents trends rather than a deterministic the more sophisticated and almost entirely framework, it points to some of the challenges marketed cousin of firewood. With its higher facing those who would like to develop energy content per weight unit, charcoal pays Exploring the Forest—Poverty Link 25

its way much more easily into the cities than reduce petty corruption and benefit the poor. firewood and the local value added clearly Other proposals include large scale plantation benefits poor producers. Its less bulky storage projects, more energy efficient processing and many speciality uses often increase charcoal technologies and the introduction of high yielding demand over time, while firewood demand species. correspondingly becomes stagnant or goes into It is likely that most of these suggestions decline. In Africa, fuelwood demand is more only make sense in the second stage of the forest resilient than in Asia and Latin America—where transition, when wood scarcity feedback loops higher incomes are accelerating the transition have kicked in. Even then, the potential for to fossil fuels and electricity. poverty reduction needs to be carefully assessed. Firewood and charcoal markets share Yield enhancing measures would benefit important pro-poor features with other NTFPs. innovative producers, though extension services They require low technological levels, limited and communal arrangements can broaden the skills and have small market entry costs. These reach. However, widespread adoption would also factors favour a structure of many small increase supplies in markets with inelastic suppliers and (therefore) low commercial demand. This would reduce market prices, thus margins. The products are economically inferior crowding out other small suppliers or reducing goods, that is, per capita demand declines with their profits. What is gained in benefiting one increasing income, especially for firewood. group of poor producers can thus easily represent Taking a dynamic perspective, there seem a loss for others. to be two main production stages. Firstly, when forests and wood resources are still abundant, harvesting is non-sustainable, with the resource 4.2.4 A synthesis and conclusion being ‘mined’. Still, many observers exaggerate about NTFPs the destructive impact on forests of fuelwood A key insight for the forest-poverty discussion harvesting. Fuelwood producers often piggyback from the literature is the ambiguous role of on agricultural land clearing and thus use trees NTFPs. The very same characteristics that make that would have been felled anyway. them important and attractive to the poor in In the second stage, when forest conversion the first place also limit the potential for is nearing an end, wood supplies become scarcer. further income increases. By the same token, Higher prices often encourage tree growing and some of the measures designed to elevate the sometimes, a higher efficiency in the use of level of sophistication and raise economic wood. Supplies are then drawn from a larger returns may risk crowding out the poorest of number of producers, regions and ecosystems. the poor (as they may lose their comparative In particular, trees outside natural forests (e.g. advantage as suppliers). We should add small woodlots and trees on farms) become though, that a similar dilemma arises in other dominant sources of supply—to the extent that informal sectors in developing countries. it can be asked whether one necessarily needs The general asymmetry is well forests proper to maintain fuelwood supplies. summarised by Neumann and Hirsch (2000): This demand presents good tree planting “Although NTFPs are extremely critical for the opportunities for smallholders, especially in peri- rural poor as a livelihood strategy, they rarely urban zones where transport costs remain low. provide the means of socioeconomic Fuelwood is thus important for the poor at advancement”. Campbell et al. (2002), in a various stages of the forest transition. comprehensive household study of two villages What points of leverage exist for pro-poor in the drylands of Zimbabwe, found that a large policies? Regarding subsistence uses, Arnold et variety of woodland products filled important al. (2003) suggest community level interventions roles in the coping strategies of households, that secure poor people’s supplies. For example, especially the poorest ones and in particular, vis- as when newly introduced crops cause land à-vis seasonal income shortfalls. But they also scarcity that leads on to the privatisation of conclude that, “While the woodlands are communal lands or other forms of exclusion of important as a security we found no evidence the poor. For commercial uses, they recommend that they are able to lift people out of poverty” taking fuelwood more into account as an (Campbell et al. 2002). important side product in forestry interventions. NTFPs and other extractive forest products They also propose the deregulation of are currently threatened, both from government restrictions, which would serve to overexploitation and habitat destruction. “Nearly 26 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

everywhere users of forest products are faced means of transport, small business, education with a decline in the size or quality of the or migration). In this case, from a livelihood resource from which they obtain their supplies” perspective, some natural capital is thus (Byron and Arnold 1999). In regions across the deliberately sacrificed to accumulate other types developing world with dismal prospects for of capital. Although an ‘unsustainable’ strategy economic development, the challenge may be from a forestry sector perspective, forest to maintain the supply of NTFPs and their role mining can, from an individual (and sometimes as a safety net—unless other activities and also social) perspective, be a perfectly rational resources can take on that role. One might argue strategy to reduce poverty. Human history is full that forest products are more important to of examples where this strategy has been prevent worsening poverty than to promote followed. socioeconomic advancement. Stressing the safety net role becomes more critical if one focuses on the very poor. A diminishing resource 4.3 Timber base, combined with few opportunities or a low In spite of the undeniable role of NTFPs, timber ability to take advantage of other opportunities, is commercially the most important product in places them at risk of further deprivation. most forests. As much as NTFPs tend to be the The somewhat pessimistic conclusion about poor person’s lot, the benefits from timber NTFP development as a means to increase forest often seem to be captured only by the rich. If based income does not imply that this is never this is true, what are the reasons? And what possible. For instance, Fisher (2002) could possibly be done to increase the benefits demonstrates in a case study from three rural for poor people from timber? In this section, villages in Malawi that asset poor, forest- we first consider some general characteristics dependent households during her single year of timber harvesting in the tropics. We then income survey actually did at least as well in take a brief look at three different pro-poor terms of income and asset building as the better timber options for smallholders: management off households with other specialisations. of natural forests, tree growing and wood However, in general, the number of cases with processing. The section is closed with a successful pro-poor forest product development discussion of the indirect benefits for the poor is very limited (Fisher 2001). NTFPs are in the rest of the economy, from timber definitely not the ‘silver bullet’ that some were production, supplies and income. hoping for a decade ago. Silver bullets are generally rare, particularly in the field of conservation friendly poverty reduction—though 4.3.1 Why have the poor not understandably, that does not stop people from benefited from timber? searching for them. There are some fundamental characteristics of Successful pro-poor forest development timber planting, harvesting and processing (and implies both increasing the value added to some features of trees as assets) that prove to products (the size of the cake must increase) be ‘anti-poor’ in that they require capital, skills, and ensuring that the poor’s share of the cake at land tenure, technology, production systems and least does not shrink. There are many ingredients time horizons that do not favour poor people. for success, as noted by Arnold (2001): pro-poor While many of these factors relate especially to institutions, skills, marketing channels and logging from natural forests, some are equally reduced middlemen profits are among the most valid for forest plantations or trees on farms: important. As mentioned earlier, an ongoing global-comparative research project within CIFOR (1) Long time horizon is analysing what generally applicable success Tree growing (but not extracting) is a long term factors for forest-product development can be investment without significant intermediate identified. returns, often involving high risks in terms of Finally, degrading ways of using forests can price fluctuations, tenure insecurity and natural sometimes help to lift people out of poverty, hazards (like fire, diseases, etc.). especially in forest-abundant environments. People may opt to cash in forest capital and then (2) High capital intensity invest it in more profitable alternatives outside Many (though not all) timber related operations the forest sector (e.g. cattle, arable land, a require machinery and capital outlays—often to Exploring the Forest—Poverty Link 27

In most forests, timber is the most valuable commercial product extracted. Yet poor local communities tend to receive only marginal benefits. What can be done to increase the share of local people? Pará State, Brazil. (Photo by Sven Wunder) a larger extent than the smallholder agriculture more skill demanding than producing a that would be a natural comparison. This capital homogenous one. intensity often rises with value added stages, i.e. in the transport, processing and marketing (6) Trees are immobile assets phases. Trees are physically fixed assets that (normally) cannot be moved. Hence, one needs secure land (3) Technology and skill intensity tenure to plant them or at least secure usufruct At least in the processing stages and sometimes rights. Otherwise, there is a large risk of losing in the primary harvesting, skills and technology access to the trees before harvest. Yet, the are needed to an extent that is often beyond poorest people often do not have secure tenure the capabilities and possibilities of poor people. and can be pushed out of the land they occupy. Many of them also voluntarily move to other (4) Economies of scale locations to take advantage of new Related to (2) and (3), the indivisibilities of opportunities elsewhere. If they do, they cannot investments in transport, processing and take the trees with them, as they can with their marketing often mean that businesses only cow, belongings or education. become competitive above a certain minimum production level. Whereas poor producers almost (7) Trees are pure ‘cash crops’ by definition are small scale and only in When the markets of food crops fail, poor people exceptional cases are sufficiently self-organised tend to switch to subsistence uses—they need to join forces into larger cooperative units. something to eat to survive. Unfortunately, few trees are particularly nutritive and digestible. It (5) Specialised markets is harder for tree growers to switch to End user markets for wood tend to be specialised subsistence uses, though some trees produce and are becoming increasingly so. Delivering a fruits and some can be inter-planted with food specialised product to a targeted public is usually crops (see 10). 28 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Conversely, timber trees and production can We can expect the rich, those with access to obviously also have some pro-poor characteristics. capital and political power, to have the power For example: and incentive to exclude the poor from capturing the bulk of timber rents.19 The combined effect (8) Flexible harvesting time is that timber values tend to accrue in the hands Most tree management operations can be done of a few timber companies and the interest quite flexibly (e.g. during slack seasons when groups they are allied with. there is labour availability or when cash is As with other high rent natural resources, needed). This is an important pro-poor high profits from timber can also promote argument, as it allows for harmonisation of both corruption, which can jeopardise the integrity tree growing and harvesting with other of national institutions, as has occurred in productive activities. Southeast Asia (Ross 2001). In some cases, high timber profits and other elevated economic rents (9) Trees as savings are also vehicles for violent conflict and civil Trees can be a form of on-farm savings that war (Collier 2000). To the extent that corruption grow over time—in some cases without too much and conflicts come to affect whole regions or labour effort needed—although the ‘accumulated countries, other poor people can be negatively interest’ (the market prices paid for wood) affected as well. There is thus an extra-sectoral fluctuates over time. pro-poor element to good governance in high rent timber sectors. (10) Multiple use The benefits to local people from tropical Forests and tree based systems can provide logging operations have in most cases been multiple outputs both for subsistence (‘coping restricted to transitory employment, small direct strategies’) and in supplementing incomes. This compensation payments from timber firms and is obviously true of natural, biodiversity rich also, indirect benefits, such as road building, forests and of agroforestry systems but free transport, some social infrastructure and specialised plantations can also produce demand for locally produced goods (e.g. important side benefits, such as mushrooms or foodstuff). In some cases, these development game. effects are important to local communities. In Unfortunately, in many cases the obstacles addition, to the extent that these timber rents prevail. Many of the technical characteristics are large enough to stimulate significant national of timber production and the operation of economic growth, some benefits would trickle timber markets prohibit poor people from being down to some poor people elsewhere (e.g. directly involved. Hence, without being a through the reinvestment of timber rents in the deterministic feature, an economy built around urban sectors of a capital-scarce developing a leading timber producing sector is unlikely to economy). Such indirect gains would benefit poor be particularly advantageous for its poor people at the national level, not those living near people. If we recall the five point summary list the forests where the timber came from (see of common pro-poor growth features from discussion below). Section 3.3, the only relevant advantage for Still, there is also some ground for optimism timber harvesting and tree growing is that it regarding local benefits. At the margin of the gives some stimulus to the rural economy. On dominant anti-poor characteristics of timber, the four other points (education, labour several partially interrelated new trends could intensity, pro-poor consumption and good raise the share of the timber rents being governance), the scores are arguably low. The appropriated directly by the local poor.20 challenge for pro-poor timber interventions is then ‘to swim against the tide’ of (1) Devolution predominantly anti-poor preconditions. Due to a redistribution of forestland, local The political economy dimension of timber communities now own or control about one fifth is just as important as the technical features. of forests in developing countries (White and Precious timber species fetch high stumpage Martin 2002). This strengthens their options to values, which attract powerful outsiders from appropriate timber values. both the public and private sectors seeking to capture timber rents (Ross 2001). The (2) Decentralisation aforementioned ‘Dove hypothesis’ (Dove 1993) The transfer of power to local government and is even more valid for timber than for NTFPs. additional local resource control in many parts Exploring the Forest—Poverty Link 29

of the topics is serving to increase the access production supplied by small scale extractors, of the poor to forest rents—unless local elites particularly in Latin America. This sector may manage to fully take over the role that the work either independently or together with the national elites previously held. large scale sector (e.g. small scale primary operators selling timber to companies). (3) Better governance Following Arnold (2001) and the World Bank Processes of democratisation in many developing (2001), we concentrate in this section, on three countries, campaigns against corruption, a freer major direct timber benefits accruing to the press and the involvement of NGOs, all potentially poor:21 increase the bargaining power of rural (1) Local access to and management of communities and the benefits they can natural forests appropriate (see also Section 3). (2) Smallholder tree growing (3) Small scale wood processing (4) Old concessions In many tropical countries, commercial Local access to and management of natural concessionaires have over-harvested and not forests renewed their concessions, which presents an As noted by Arnold (2001), the main area for opportunity for forest communities to intercede donors’ targeting of the poor living in and near and compete for access rights. forests has been in transferring local management to rural communities. The record (5) New technologies is mixed both in terms of transferring real Emerging small scale technologies in sawnwood management authority to the communities and and plywood production can help small producers ensuring that such a transfer really benefits the to become competitive (see below). poor. A number of obstacles have been noted (most of which also affect decentralisation in (6) Rising timber demand other sectors): In many countries, demand for high volume, low • Lack of effective community level grade construction timber is growing—which the institutions, local skills and poor, especially in market accessible (e.g. peri- entrepreneurial spirit urban) areas may have a comparative advantage • Forest rent being captured by local elites to supply. • Inconsistencies in statutory laws and Some market barriers and bureaucratic regulations regulations have prevented the poor from • Bureaucratic hurdles and lack of benefiting from forest products and timber in ‘connections’ particular. Government regulations and policies • Slow institutional change in national have been geared towards extracting rents and forest agencies have created strong barriers against the poor’s market participation. Some regulations, such As already mentioned, another potential as permits, licenses and in particular, local timber benefit is direct payment from requirements for extensive forest management logging companies to forest communities. Such plans, have created barriers to entry for the poor a transfer of part of the timber rent already (say, into small scale tree plantations) and often occurs in many logged-over tropical forests, serve as de facto vehicles for corruption. Trends formally or informally, by law or negotiations. In of market deregulation, reduced government Cameroon, these arrangements are usually made interventions and campaigns against corruption prior to logging. Companies may pay because should therefore provide increased opportunities they are legally required to do so to obtain access for poor people. or simply as a means of buying local goodwill for their operations, in order to prevent violent conflict that would damage their interests. But 4.3.2 Can timber be made more there are also cases, for example, in the pro-poor? Amazon, where timber companies agree to pay Tropical timber production is normally thought post harvest compensations, only to default on of in terms of large, highly mechanised their promise.22 companies harvesting large concession areas of Given the high timber rents in many tropical natural forests. Parallel to this, there is a forests, the potential for raising payments to moderate but increasing share of primary local people is significant. Yet, as long as 30 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

companies have many potential extraction areas access to high timber profits extracted from to choose from, forest communities would need natural forests. However, an important final to act in a concerted manner (e.g. as a cartel) caveat is: to what extent do these profits reflect to strengthen their bargaining position. true stumpage values, compared to downstream Providing improved information about logging value added or market oligopolies and policy- access fees in other locations can be a first step created monopolies? In economic terms, to towards empowering local communities. Another what extent is the high timber rent we observe factor that can significantly raise local bargaining in the marketplace a resource rent or a power is if communities have a credible monopoly rent?23 How valuable is timber in the alternative to potentially exploit the timber in hands of poor local people, without the same their own community based enterprises. External political power and access to capital and markets? technical support to producer organisations, tax These are critical questions regarding the incentives or certification schemes can be potential benefits of natural forest devolution and helpful in this respect. local management (D. Kaimowitz, pers.comm., New pro-poor options in timber production Jan 2002). need not only come from redistribution and altered institutional arrangements. Technological change Smallholder tree growing might also increase the opportunities for the poor In many parts of the tropics, particularly in Asia, to benefit from increased access to and control there has been a significant shift in focus in over natural forests, and stimulate the recent years, from natural forest extraction to involvement of small scale timber extractors. forest plantations, with the latter including For instance, small portable sawmills demand smallholder tree growing. This partly reflects an less capital outlay and should also favour poor increasing shortage of forests and tree resources people’s participation in the production of in some areas, as well as a political reluctance sawnwood. Technological changes in the plywood to create large scale industrial forestry industry allow for the use of smaller diameter plantations. The technological changes referred trees and a greater number of species. Local to above create a growing demand for short communities have, at least in the past, been rotation species (less than 10 years), which are granted control over the commercially less more attractive to smallholders. This will serve valuable forests. Thus, this technological change to increasingly extend the forest-product scope, could also raise the value of the type of forest from forest dwellers to people living in they increasingly control. established agricultural areas with trees on farm. Demand for paper and paperboard is the Significant international credits and grants (e.g. fastest growing forest product category and the from the World Bank) have gone into smallholder pulp and paper industry is rapidly expanding. tree growing schemes, including attempts to Given the scale of these operations, large utilise fragile and unproductive lands. These companies will probably continue to dominate options may be the most promising in peri-urban production. However, this may create areas, where transport costs to the market are opportunities both for poor people to grow not prohibitive. Surprisingly little, however, is trees to supply fibre (see below) and for those known about the impact, for example, of the controlling natural or degraded forests near massive smallholder tree planting efforts in India pulp mills. and China. That said, these new technological and Development of closer partnerships market opportunities can also pose threats to between on the one hand, smallholders, producer the environment. Increased demand for ‘any organisations or communities and on the other, tree of any size’ provides strong incentives for commercial companies, can present a major way over-exploitation, for large companies as well forward. As discussed above, this is particularly as small scale extractors. The danger is that relevant in those cases were timber in the hands large companies could clear-cut most of the of the poor is worth much less than it is in the natural forest and ‘rip off’ the benefits, while hands of a commercial company. An effective the poor could lose their forest based safety nets partnership between unequal partners—poor and forest-product incomes (see previous local people and commercial companies—needs section). to be based on the comparative advantages of In summary, we have claimed in this section each. The poor have cheap access to labour and that new institutional, market and technological control the land—or at least they credibly contexts might help to promote poor people’s threaten external agents’ aspirations of control Exploring the Forest—Poverty Link 31

over land. The companies have easier access to partners, producer organisations or capital, skills, technologies and markets. cooperatives (with a well defined legal status) The bargaining power of individuals or may often be more adequate counterparts for communities is often weak but mechanisms can outgrower schemes. be implemented to increase their negotiating abilities (e.g. through producer associations Small scale wood processing and alternative market outlets). A flow on Timber can benefit the poor not just through effect of this should eventually be a rise in the extraction but also through small scale incomes they receive for their trees. Then processing. Small scale enterprises can also be again, many companies have been caught out large providers of rural employment. In a six due to farmers breaking contracts and selling country study, FAO (1987) found that forest- their products to third parties. Thus, the basic based small-scale enterprises accounted for profit and risk-hedging interests of firms must between 13 and 35 % of the total employment be taken into account in order to make it generated by this group of enterprises.24 The attractive for them to enter into a partnership two main sub-sectors were carpentry/furniture (Mayers 2000). Once more, there is a certain and the weaving of baskets, mats and hats, with trade off between focussing on the size of the both mainly serving local markets. Activities like cake and on its distribution, both of which wood carving and sawmilling/pitsawing provided affect the poor. supplementary employment. The most prevalent form of such formal A more recent study (Arnold et al. 1994; partnerships has been outgrower schemes. see Arnold and Townson 1998 for a summary) Mayers (2000) and Desmond and Race (2001) in six African countries25 revealed that some summarise the lessons that have been learned. 763 000 people were employed in small scale Partnerships have generally proved to often forest enterprises. This makes it an important benefit smallholders but not necessarily the rural off-farm sector, employing on average 2.3 poorest of them. In India, where access to secure % of the total rural population in the areas land is needed to enter the contract, mainly surveyed.26 The largest sub-sectors involved medium and large scale farmers are involved in grass, cane and bamboo products (42% of total the schemes. In South Africa, the pulp and paper employment), wood carving (27 %) and forest- company Sappi has been successful in integrating product trade (20 %). Woodworking had the into their scheme more than 8000 small farmers, fastest rate of growth. These industries are with a combined plantation area of 88,000 ha characterised by many small sized activities and (Mayers et al. 2001). intense competition, high turnover levels within Partnership arrangements are highly firms, small barriers to entry and low levels of context-specific and must be viewed within a technology and capital inputs. Arnold and multi-stakeholder perspective. However, recent Townson (1998) extrapolate that about 15 CIFOR research has also suggested that some million people could be engaged in such common patterns exist (Nawir et al. 2002). For activities in sub-Saharan Africa. instance, limited competition for land helps The type of support and policy interventions make the schemes attractive to villagers. required depends on the specific products and Transparent negotiations and clear information contexts involved. In many areas the resource on risks can help make the schemes sustainable. base itself—or access to it—is threatened and Secure land ownership also clearly favours long needs to be better managed. Micro-credit term commitments. programmes can often provide working capital Companies often find it easier to deal with and help entrepreneurs upgrade equipment and individuals than with communities. Indeed, technology, thus increasing the value added. communities are heterogeneous groups with Marketing assistance can also be valuable. different interests and assets, which makes Notably, government deregulation would it difficult to reach and implement agreements stimulate the growth of small scale industries. that involve long term mutual commitments While these activities depend directly on (thereby increasing the companies’ risks and forest-product supply, they lie outside what we transaction costs). One conclusion has been defined as ‘forestry’ proper in the Introduction. that “communities as such are rarely suitable The lion’s share of the final price that the units of social organisation for tree consumer pays corresponds to secondary management” (Mayers 2000). As a solution in processing and much of this is urban based (e.g. between the options of private and communal Scherr et al. 2001 and Arnold 2001). These 32 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

‘downstream economic effects’ are also the key substantial, though both significant under- and topic of the following subsection. over-estimations take place when these figures are recorded in the national accounts. Under- estimation occurs because a large variety of 4.3.3 Forgotten indirect and off-site forest-product subsistence uses, local market benefits? sales and illegal transactions are not registered. The forest-poverty debate seems sometimes Over-estimation can happen when village studies overly focused on micro distributional processes of single sector activities (e.g. firewood) are at the forest level (“What share of the timber uncritically scaled up to the national level. For cake goes to poor local people?”). From a instance, FAO reported for the early 1990s, a general poverty alleviation perspective, this high 6% average forestry GDP share in Africa, focus pays insufficient attention to the indirect with 18 countries on the continent even and economy wide processes that were discussed exceeding 10% (FAO 1997). Uganda at the in Section 3. Here we will briefly look at five of extreme, was supposed to have a forestry share these factors: in GDP of 23 % in 1991. Yet, Uganda’s Ministry (1) Local indirect development effects of Finance, as reported by the World Bank, set (2) National forestry income and export the forestry GDP share at 1.7% in 1992/93—only receipts one fourteenth of the FAO figure. The latter is (3) National forestry employment probably too low because it is likely that it (4) Urban forest-product prices ignores subsistence uses, but the range of (5) Reinvestment of forestry rents outside discrepancy shows that much care is needed in the forestry sector evaluating aggregate income data.27 As mentioned earlier, international trade (1) Local indirect development effects in pulp and paper, furniture and other selected Local multiplier effects from the entrance of forest products is expanding rapidly. Large timber firms into an area can be significant—in companies are merging to raise their efficiency both positive and negative terms—and often far within an increasingly competitive exceeds the direct compensation paid to local marketplace. The increasing foreign direct dwellers that we discussed above. Though investment in developing countries indicates variable from case to case, logging firms often that such companies will continue to expand generate demand for locally produced products their market shares in the future, especially in and services (e.g. lodging, food, assistants, etc.) the plantations sector (Poschen 2001). In some while they operate within an area. Newly built countries, forestry is a treasured earner of logging roads often open up opportunities for foreign exchange. Total world exports of wood marketing locally produced goods and can also products were worth US$134.7 billion in 1996. give local people easier access to health and Three tropical developing countries have education services. Roads may be the most entered the top ten: Indonesia (US$5.2 billion considerable, durable benefit, unless they have - 3.9%), Malaysia (US$4.2 billion - 3.1%) and been constructed in a rudimentary way that only Brazil (US$3.2 billion - 2.4%) (Poschen 2001). makes them last temporarily. For Indonesia, a large illegal export makes this Conversely, the indirect local impacts of a figure very conservative. For all developing negative nature can include reduced NTFP countries, the average forestry share in GDP production from logged-over forests (e.g. fruit is around 2% (with the caveats mentioned trees, wildlife, etc.), increased fire risks and the above) and its share in developing country introduction of human diseases or social exports is almost 3% (World Bank 2001a; disruptions (including conflicts with logging Poschen 2001; FAO 1997). companies and their workers). Furthermore, most Historically, the forest industry has only benefits from logging accrue as short lived booms, reached high GDP shares and contributed heavily posing difficulties concerning adjustments when to the economic development and poverty the companies leave. reduction of a limited number of countries, such as Finland, Norway, Sweden and Canada. The (2) National forestry income and export 1960s had witnessed more optimistic views about receipts general forestry led development in developing The national income contribution of forestry in countries than were actually realised (Westoby general and of timber exports and production 1962). Westoby later came to recognize that value in particular, can sometimes be “forest industries have made little or no Exploring the Forest—Poverty Link 33

contribution to socio-economic development” generally held that most employees in logging (Westoby 1978). Instead of an ‘engine of operations are low skilled outsiders. Young men growth’, the sector sometimes became a drain in particular use this as a strategy to accumulate on domestic resources and hard currencies. capital before establishing a household. For Some researchers have looked to these young males, the savings out of logging explain the past failure in the inherent wages can be an important way to eventually characteristics of the timber industry (see top reduce poverty. of this section). Others have blamed policies, pointing out that the forest sector often suffers (4) Urban forest-product prices from adverse macroeconomic and sectoral Have poor urban consumers become better off policies. Using Peninsular Malaysia, Ghana and from cheaper forest products, helped along by Chile as examples of both successes and failures, more efficient production and higher market Vincent and Binkley (1992) argue that “if the supplies for forest products? This was the major right policies are followed, however, forest based pro-poor effect in the case of staple foods and industrialisation can provide an important the Green Revolution. Unfortunately, while a few source of employment and income”. non-timber products can be important for the poor (e.g. firewood or medicinal plants), most (3) National forestry employment timber derived products play only a minor role Employment in timber production generally tends in the consumption basket of poor urban people. to be less labour intensive than agriculture, High income elasticity makes these products manufacturing and services but significantly more important for middle income groups and more labour intensive than oil and mineral countries—people tend to consume more timber extraction. This has importance for the changes products only as they become richer (Wunder between agricultural and forestry land uses: 2001). So, if forest-product consumers have “Where forestry competes with agriculture for benefited from lower prices, this relates more land, increases in forest cover lead to large losses to the urban middle classes. of employment per unit area for all but the most extensive forms of agriculture” (Poschen (5) Reinvestment of forestry rents outside 2001).28 This would explain why forestry’s the forestry sector employment creation and general success has Do rich timbers produce large rents that are been largest in countries with low agricultural reallocated to other sectors, where they soil potentials.29 eventually create employment and income for There are no real ‘hard data’ on direct the poor? Obviously, timber barons do not forestry employment, but Poschen (1997) transfer the totality of their profits to Swiss bank attempted some guesstimates of full time job accounts. They are likely to reinvest a significant equivalents. For developing countries, he (though variable) share in promising domestic estimated 1.9 million jobs in roundwood sectors. They will also consume domestic goods harvesting, 0.8 million in reforestation/ and labour intensive services. Both spending silviculture and 13.3 million in fuelwood (the latter pathways would benefit overall incomes and is probably overestimated). Outside ‘core employment and are likely to cause some ‘trickle forestry’, a further 3.3 million would be in the down’ effect to poor people. formal, and an estimated 16 million in the Unfortunately, we are not aware of any study informal, wood based processing industries. Much that has quantified this process. Nevertheless, of this is part time work involving many it seems obvious that in a timber rich country individuals. This is often not realised in studies like Indonesia, some proportion of the billions done by “foresters, botanists or ethnologists [who of dollars of timber rents that have been cashed report their results on forestry employment] in in since the mid-1960s must have been invested terms of people [who are] ‘involved in’, ‘depend into the rapidly growing urban sectors. In rural on’ or ‘derive a substantial part of their income areas, there is evidence that some companies from’. They therefore tend to massively overstate investing in oil palm in Indonesia directly the employment provided” (Poschen 1997). depended on the rents from timber to finance There has been little research on the poverty clearing and planting. Both types of investment profile of such forestry employment. A lot of the must have stimulated national economic growth direct forestry employment is unskilled work with over the last three decades, thus probably also significant health risks in logging camps and helping to reduce the number of people living sawmills (Poschen 2001). In Indonesia, it is below the poverty line. 34 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

In terms of general linkages and ‘multiplier and with higher incomes (e.g. some wooden effects’, Poschen (2001) reviews studies from agricultural tools). Others have a positive but developed countries, showing that for every small income elasticity (e.g. furniture and forestry job normally created there, an extra woodcarvings) and a third group faces expanding 1.0-1.4 jobs outside the sector were also but fluctuating exports markets (e.g. pulp and created—the full estimate range was 0.5-3.2. paper). On the positive side, the general market Currently, no developing country studies of this outlook for these low value timbers may still be type are available. better than for most NTFPs.

4.3.4 Summary on pro-poor timber 4.4 Payments for ecological services benefits The high rents from previous timber species have 4.4.1 What link to poverty? conventionally been regarded as falling Forests can provide important services both to exclusively into the laps of the rich. Images of local on-site forest dwellers (e.g. clean drinking chainsaws and large trucks destroying the forest water) and off-site beneficiaries, such as regional and jeopardising poor forest dwellers’ livelihoods users (e.g. downstream water benefits), national are based on many real experiences. Some consumers (e.g. urban tourists) or global imminent characteristics of trees as assets and stakeholders (e.g. valuing the existence of of the timber production process, in extractive endemic biodiversity). Probably the highest or planted systems, prove to be ‘anti-poor’. poverty reduction potential is through payments Timber requires capital, skills, land tenure, for the off-site benefits enjoyed at the regional, technology, production systems and time national or global levels. Currently, most off-site horizons that poor people fail to deliver. The beneficiaries are ‘free-riding’—they don’t pay for causes behind this lie deeper than general what they get. Hence, a poor forest community injustice of power—putting the timber in the that owns a forest has no incentive to take this hands of the poor may make it worth much less. service provision actively into account in its land Yet this is only part of the story about the use decisions. This means that when land current impacts of timber extraction on poor becomes scarce and is increasingly put to the people. In some cases, there may be important locally most rewarding use, the off-site service indirect benefits from industrial operations, may be lost. When land use changes begin to particularly local development impacts (e.g. threaten these services or supplies (e.g. soil road building) and the reinvestment of timber erosion polluting drinking water), free-riding profits into other sectors of the economy. beneficiaries may eventually be willing to pay Millions of small enterprises and poor forest owners and users to protect the service. households also benefit directly from extracting By rewarding those who bear the opportunity costs timber from natural (including secondary) of not ‘developing’ their forestlands, both service forests and from planting their own trees, and providers and off-site beneficiaries can gain. It the number of smallholders deriving these seems to be one of the few widely applicable ‘win- timber benefits seems to be on the rise. A win’ options between forest conservation and number of recent changes in technology, markets poverty reduction. and institutional frameworks (including control Compensation mechanisms are relevant in over forestland) favour this trend. We have at least four forest related areas, examined in discussed three paths that appear promising not greater detail below: (1) carbon storage and only for poverty prevention but also for poverty sequestration; (2) biodiversity conservation; (3) reduction: increased local control over natural hydrological services; and (4) tourism. We also forests, smallholder tree growing (e.g. outgrower need to make another subtle distinction. In some schemes) and small scale, wood based cases, we have payments for safeguarding against enterprises. land use changes that threaten to jeopardise Note that some activities based on low rent, current service provisions (e.g. preventing high quantity timbers (i.e. ordinary, lower valued deforestation that would cause additional carbon species) face dilemmas similar to those of emissions). In other cases, users would pay for NTFPs. Again, it is necessary to distinguish the explicit creation of forest services, be that between different product types (Arnold and for recuperating a lost service (e.g. reforestation Townson 1998). Some low value wood products to sequestrate carbon) or for new services (e.g. are inferior and tend to disappear over time carbon motivated afforestation of grasslands). Exploring the Forest—Poverty Link 35

Payment for ecological services is an emerging area with high growth potential and opportunities for poor forest stewards, but uncertainties exist regarding how much the poor will benefit. Villagers in the Transkei, along South Africa’s ‘Wild Coast’, discuss forest management with CIFOR researchers. (Photo by Sven Wunder)

Payments of the avoidance type anticipate truly affects downstream water quality a counterfactual baseline scenario of forest- through their land use decisions and service deterioration—if no payment is made, that the direction and size of that things are supposed to get worse. This implies impact are clear. that service buyers often (though not always) (2) The downstream user only wants to pay service providers for preserving the status implement a payment if it is likely to be quo, rather than deteriorating the service.30 In effective in achieving land use change turn, forest-service creation payments are made that improves water quality (or for improvements against the baseline of the alternatively, avoids a land use change status quo.31 in order to maintain water quality). The following example serves to illustrate Otherwise, the money would be wasted. some general opportunities and challenges for This elevates the importance of service compensations. A downstream user of monitoring compliance. It also requires drinking water is concerned about deteriorating the upstream forest owner to effectively water quality. The user considers paying an control land use, if not through secure upstream forest owner to stop his ongoing forest private tenure then at least through the conversion of an erosion prone plot right to exclude others from affecting (safeguarding) and perhaps even to replant trees incompatible land use changes. in areas already affected by past clearing (3) The downstream user will not want to (recuperation). For a payment scheme between pay for the conservation of a forest plot the two to emerge and become effective, several that was in any case too remote, too ‘nuts and bolts’ have to be in place. Three basic unfertile or otherwise unattractive for prerequisites are essential for the incentive the upstream forest owner to clear. In mechanism to work: other words, when the opportunity cost (1) The downstream user (the potential of conservation is actually (perceived service buyer) must be aware and to be) zero, there is no basis for convinced of the existence of an payment. If there is no credible threat externality. In other words, they must to the continued provision of the believe that the potential service service, the buyer’s rationale for provider (the upstream forest owner) compensation vanishes. 36 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Specifically to the first point, there is transfers will eventually make the forest-service sometimes limited biophysical ‘proof’ of the providers better off. Obviously, the service buyer existence of positive forest externalities, will look for ‘the biggest (service) bang for their compared to alternative types of land use and buck’ so to speak, while the provider will want vegetation cover. This is particularly true for to minimise adjustments to their planned first- the ongoing controversy concerning forests and best land use without the transfer. However, this hydrological services. A key question is always: conflict of interest is no different to that between what alternative land use(s) can one compare any buyer and seller of forest products or indeed forests with? Some converted uses can perform any suppliers and consumers within any market. a range of ecological services that are similar The gains from trade should still make both to those that forests provide, (e.g. perennial service providers and users better off, unless crops or home gardens). In fact, some uses can special coercive scenarios come into play (see even improve specific desired services (e.g. discussion below). What is the alternative? creating greater water runoff). Others, such as Clearly, we cannot expect the poor to continue the conservation of endemic biodiversity, are providing these services for free when it conflicts unambiguously linked to the preservation of with their own interests. natural forest cover and quality. Finally, we should remember that not all the What poverty reduction potentials do we off-site service recipients are rich. There are expect that a spread of forest-service payments poor people among downstream farmers, urban will offer on a global scale? Firstly, one should water users and global beneficiaries. These keep in mind that the prime objective of these impoverished service recipients may not be able payment schemes will never be to make poor or willing to pay but they may free-ride on other people better off. Service markets and users’ payments and thus benefit from the compensation agreements emerge due to mitigation of environmental problems ‘gratis’. demand-side pressures, driven by growing Global warming for instance, is expected to hit shortages and predictably, by influential actors poor people in tropical countries especially hard reacting to them. because of the potentially greater water For payments to have a lasting positive shortages, declining crop yields and exposure impact on poor people, be that as a mere by- to diseases—and the limited means of the poor product to conservation-maximising schemes or to adjust to these threats (IPCC 2000). In this as an explicitly addressed secondary objective, sense, a poverty aspect is equally present on it is a sine qua non that the schemes are actually the forest-service demand side. effective in providing the forest services. A strict efficiency requirement is something that distinguishes payments through forestry aid 4.4.2 Carbon storage and projects, from other ‘soft’, altruistically inclined sequestration donor interventions. If the scheme fails to deliver Forestry projects that store or sequester carbon the services, no responsible service buyer would can receive technological and financial transfers want to continue paying for it. under the Clean Development Mechanism A concern has been raised that such payments (CDM) of the Kyoto Protocol. Carbon services are just another example of rich people in the from forests relate to the planting of new trees Northern Hemisphere setting an agenda to the to sequester additional carbon from the detriment of the legitimate development atmosphere (service creation) and to forest and aspirations of poor forest dwellers in the South. tree carbon storage in areas that would We believe that this is a wrong perception, simply otherwise have been cleared (safeguarding). because service-compensation schemes as a rule Yet, in July 2001, at the 6th session of the United are voluntary quid pro quo agreements, offering Nations Framework Convention on Climate an additional income source to poor households, Change (UNFCCC) in Bonn, forestry activities which they may or may not accept. In the simple under CDM were limited to reforestation and watershed example above, a poor forest owner afforestation and activities to reduce would only wish to enter into an agreement if the deforestation were not included. net compensation was higher than what they Still, private sector Joint Implementation (JI) would lose by foregoing the best alternative land schemes outside the CDM, as well as the newly use options. founded BioCarbon Fund,32 have not taken the The greater the willingness to pay on behalf same rigorous stand against the eligibility of of the recipients, the more likely that these ‘avoided deforestation’. Uncertainties about the Exploring the Forest—Poverty Link 37

future implementation of the protocol range from Unfortunately, these two enabling characteristics doubts about its adoption to the prospective role tend to be negatively correlated in space (that of Land Use, Land Use Change and Forestry is, low-population vs. high-population density (LULUCF) in CDM. Various methodological and areas). measurement problems relate to the effects of Amongst different types of projects, Smith forestry activities on the carbon balance, including and Scherr note that assisted natural regeneration additionality (establishment of a quantitative tends to be cheaper than tree planting and baseline of ‘counterfactual’ carbon release without recommend its inclusion into the CDM. Averted mitigation measures), duration, project deforestation, currently not CDM-eligible, boundaries and leakage effects (see IPCC (2000) generally provides more diverse livelihood for an overview). benefits than tree planting. A major uncertainty Carbon related services and commodities is the price of carbon itself. Without the differ from the other forest services in respects participation of the United States and with surplus that also have poverty implications. Forestry emissions from Eastern Europe being put on the related carbon storage and sequestration market, the carbon price could be very low and competes with a number of alternative could make many pro-poor projects uneconomical. mitigation mechanisms. Actually, it has been It would take specific interventions to reduce the argued that the alternatives are likely to be high transaction costs of implementing projects more cost efficient than forestry (Smith et al. with smallholders. 2000) but this remains controversial and awaits better empirical analysis. Carbon storage is also not site-specific—it does not matter where on 4.4.3 Biodiversity conservation the globe mitigation takes place. There are no Since the 1970s, many new protected areas have threshold effects or inevitable requirements for been established in tropical countries. Still, as collective coordination (see below). This high land becomes scarce, there is often resistance flexibility, homogeneity and substitutability will against areas being set aside purely for eventually favour the emergence of more conservation, restricting previous uses and/or sophisticated global markets, with many agents relocating people. This resistance can come both and a high degree of competition. from local residents and national governments. Arguably, this growing maturity of carbon Local people want alternative income sources markets is unfavourable to poor suppliers because in return and politicians usually want ‘expansion their comparative disadvantages will be fully zones’ into which they can channel population- exposed. On the positive side, the sheer size and led and market-led demand for new land. economic potential of the carbon market currently Integrated Conservation and Development appear larger than for any of the other three areas Projects (ICDPs) represent one attempt to for compensation mechanisms (i.e. relating to bridge these interests. Since the early 1980s, biodiversity conservation, hydrological services this has been a dominant approach to and tourism). Even a small market share for poor biodiversity conservation. Unfortunately, most people’s carbon forestry can be an important ICDPs have been relatively unsuccessful in contribution to poverty reduction. achieving conservation goals and due to The number of carbon-offset schemes is complex project structures and large growing rapidly. Bass et al. (2000) review 30 administrative costs, they also tend to be too such schemes, while Landell-Mills and Porras expensive in reaching their development (2002) found 75 forestry related carbon schemes objectives.33 There has been “growing for their review, of which 42 are in the three realisation that ICDPs have run the risk of tropical continents (Latin America 24, Asia Pacific contributing effectively neither to conservation 13 and Africa 5). Most of these projects are in nor to development. The result is a big gap their early stages, so the long term effects on between rhetoric and reality” (Sayer et al. rural livelihoods and poverty reduction are not 2000).34 The key underlying assumption has yet known. Smith and Scherr (2002) discuss been that providing alternative employment extensively the livelihood potential of forest (e.g. small scale enterprises) or raising incomes carbon projects. A dozen studies estimate the from land already in use (e.g. through improved costs of various carbon sequestration projects. agricultural technologies) should dampen the They appear to have the largest potential where pressure on forests through ‘conservation by the opportunity costs of labour and land are low distraction’. Unfortunately, these interventions and the demand for forest products is high. do not necessarily reduce either the incentives 38 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

for, nor the means of, forest encroachment. In more cumbersome and costly than paying the fact, the opposite might well be the case.35 government. On the other hand, direct payments One alternative incentive based approach to smallholders on the ground would be likely to is to encourage producers to adopt biodiversity have a larger poverty reduction impact than friendly methods (e.g. shade-grown coffee) and paying the forest agency in the capital for sell their certified products to environmentally abstaining from logging. conscious (predominantly developed country) consumers, who are willing to reward producers through a price premium. The premium can be 4.4.4 Hydrological benefits seen as a payment for biodiversity, though Hydrological forest services can be related to increased prices could in principle, also induce the quality, quantity and timing of water flows. producers to convert more forests to (albeit Although we know that forests are important biodiversity friendly) agriculture. There is some for some hydrological services, the exact evidence from Central America that this type of linkages between various tropical forest types certification arrangement can also be targeted and different hydrological services are not easy to (and benefit) poor producers (Pagiola and to document (Bruijnzel 1990). Calder (1998) Ruthenberg 2002). lists proposed forest benefits as higher A third option is to compensate people rainfall, greater runoff, regulated flow, less directly for conservation, that is, for setting aside erosion, less flood hazard and higher water areas for forest conservation that would otherwise quality. But each of these is contentious and have been cleared, logged and/or exploited in a highly site-specific and scale-specific and they way that implies biodiversity losses. depend crucially on what land use alternatives Compensations can take the form of so called one compares them with. There is evidence conservation concessions (payments made that foresters and conservationists have over- typically to governments in lieu of logging estimated the hydrological benefits of forests, concession payments) or conservation easements except perhaps for the case of drinking water (payments made directly to landowners for not supplies, where forest cover seems to be converting forests to alternative uses). In both unambiguously desirable for a high quality cases, conservation performance needs to be product (Chomitz and Kumari 1998; Bishop and monitored so that payments can be halted in the Landell-Mills 2002). case of non-compliance. These scientific uncertainties have not Direct incentives are still in a pioneer phase impeded the incipient implementation of of application in the tropics (Ferraro 2000, 2002; payment schemes to upstream forest owners, Hardner and Rice 2002; CIC 2000). Conservation mainly in Latin America. Experiments in International and other major conservation Colombia, Costa Rica and Ecuador have involved players have put a lot of effort into conservation payments from beneficiaries such as hydroelectric concessions as an alternative to timber power plants, drinking water consumers and users concessions. That is, paying an up front amount of irrigated water (Pagiola 2001). Fisheries and an annuity to a tropical country’s forest service represent another sector that may be for areas not to be logged, thus entering into detrimentally affected by the deterioration of direct competition with logging companies. water quality. For instance, in one of the first A challenge here is how to develop the total economic value studies, carried out in Korup concessions in a way that makes local people National Park in Cameroon, Ruitenbeek (1988) better off and provides payments that mimic found that forest protection of downstream other income flows. Another challenge is how to coastal fisheries was one of the largest forest promote not just concessions but also conservation economic benefits in the park, conservation easements in the tropics. In most compared to a scenario of continued places, agricultural conversion is eventually a deforestation. It is likely that growing global larger threat to tropical forests than logging. water shortages will significantly increase the The per hectare opportunity costs of scope of payment schemes in the coming decades. conservation in agricultural expansion zones are Payments for water services have often higher than for remote forests that would be been combined with biodiversity conservation logged without subsequent conversion.36 In many payments (see previous section) but as yet situations, one would need to compensate and there have been few assessments of the welfare monitor a myriad of smallholders with weak implications. As an example, Brazil has, in property rights over their land, which is much several federal states, implemented an Exploring the Forest—Poverty Link 39

Ecological VAT, where heavily forested To what extent does forest based tourism municipalities are tax-rewarded for their water help to reduce the poverty of forest dwellers? and recreation services (Grieg-Gran 2000; May Tourism differs from the three other forest- et al. 2002). The pay off has been an additional service cases in the sense that it provides not share of (Value Added Tax) VAT tax revenues, only off-site ‘externalities’, it also triggers local based on regular monitoring of forest size and cash flows related to tourists’ on-site quality. In terms of poverty, this payment often expenditure on tourism goods and services. The implies redistribution from rich and developed latter are usually also intensive in the to poor, under-developed municipalities that employment of unskilled and semi-skilled labour have not yet converted their forest areas to other (e.g. for services in hotels, restaurants and in land uses. One can only speculate to what extent the provision of transport, etc.), which can this redistribution from rich to poor also trickles favour the poor. In other words, the hedonic down to poor households. value of forests and protected areas attracts A recent global review of 61 watershed tourists and triggers other expenses that payment schemes concluded that their overall stimulate the local economy. This is an important poverty reduction impact was likely to be indirect factor. Moreover, an increasing number positive, that the effect was generally neglected of tour companies also pay direct fees to local in evaluations and that it was likely to depend communities for the right to continuously access much on the bargaining power of communities. a locally controlled natural site—which may entail Poor people were often disadvantaged in cases them giving up or restricting certain forest uses where they had to compete for the water service that are less compatible with the tourism operation provision with other users whose coordination/ (e.g. timber extraction and hunting). transaction costs were lower (Landell-Mills and Unfortunately, many case study Porras 2002). assessments of local impacts have focused one- sidedly either on perceived cultural distortions or on the relative distribution of tourism 4.4.5 Forest-based tourism income, stating the somewhat self-evident Tourism is one of the world’s largest and fastest conclusion that tourism companies tend to reap growing industries and eco-tourism and forest the bulk of profits from luxury operations. Yet, based tourism are highly dynamic segments even a minor local participation in high tourism within this sector. For international high class eco- revenues can significantly raise the absolute tourism, open areas with a high frequency and household incomes of poor people and provide good visibility of charismatic mammal species, efficient conservation incentives. such as the savannahs in Southern and Eastern Evidence from many parts of the world Africa and the Galápagos islands in Ecuador, have suggests that it is possible to design nature based done particularly well. But tropical forests have tourism operations in a way that significantly also occupied an important niche. benefits local people. This includes the classical A country like Costa Rica, in spite of a history CAMPFIRE project in Southern Africa (The of accelerated deforestation, has been able to Zimbabwe Trust et al. 1994), the Annapurna promote an eco-tourism concept based on a well Conservation Area Project in Nepal (Gurung and functioning protected area system. This has made Coursey 1994) and community benefits in Belize the tourism sector the principal source of foreign (Lindberg and Enríquez 1994). This also holds exchange in that country. Generally, tourism can for company-run international eco-tourism make use of high quality forests in a fragmented operations in Ecuador (Wunder 1999) or national landscape with good infrastructure, rather than tourism by urban backpackers into forest needing large and remote frontier forests. The recreation areas in Brazil (Wunder 2000). type of biodiversity and scenic beauty that tourists Regrettably, the economic potential for local demand is not necessarily the same as that income generation from this variety of protected in conservation concessions (see institutional arrangement is still widely ignored. above). While there is a clear upwards trend in global economic revenues from tourism, the downside is that international tourism is highly 4.4.6 Can forest-service payments sensitive to security problems (including the more reduce poverty? recent fear of international terrorism) and political In summary, there is some ground for optimism turmoil, which in politically unstable countries that forest-service payments can help to reduce causes large fluctuations in tourism incomes. rural poverty. Not only do these payments offer 40 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

an additional flow of income, such flows also necessarily have to be integrated for the tend to be more stable over time than the ones service to be provided effectively.37 they are designed to substitute (e.g. the As mentioned previously, even in those fluctuating prices of timber and cash crops) cases where poor people are completely (Hardner and Rice 2002). In addition, the bypassed as providers of remunerated introduction of payments can also help induce environmental services, there is little to sustain a series of indirect benefits, follow up the claim that they would be made worse off, investments and external assistance (such as due to the voluntary character of the schemes. training, improved community organisation, This could only occur where coercive forces better knowledge about forest management, come into play. For instance, one could imagine improved environmental quality, better cases where increased forestland values induce understanding about urban markets for other more powerful actors to expel poor people from products, etc) (Landell-Mills and Porras 2002). forests. On the other hand, there are also Finally, international markets for ecological examples of payment schemes helping to services provide a mechanism for long term consolidate the informal land rights of poor investment flows from the North to the South. people. Thus, much depends on the relative Even in the extreme case that all the direct power of the specific actors and on the selection payment recipients are larger landholders, the of the institutional set up, perhaps the main transfer could eventually provide indirect factor to keep in mind for donor interventions. benefits that trickle down to the poor (see discussion on logging rents, Section 4). A key question is: To what extent will poor 4.5 Summary people be able to compete in the emerging To what extent can forests help to alleviate service markets? The two main comparative poverty—that is, to prevent poor people from disadvantages of the poor are their insecure further deprivation or to reduce poverty? land tenure and the high transaction costs and • It is well documented that forests serve risks of interacting with them. Both may lead as safety nets for rural households, which potential service buyers to look for cheaper and include periodic food buffer functions and less risky providers of the same services. Large insurance values in emergencies. start up costs and long term, inflexible land Extensive deforestation endangers forest use commitments can be additional constraints safety nets and potentially increases for poor people (Pagiola et al. 2002). extreme poverty. Paradoxically, these disadvantages become • The static role of forest safety nets is more evident with the more mature and thus widely acknowledged but sophisticated markets. In the carbon market, controversy surrounds their dynamic for instance, buyers can choose more and more role. To what extent are other assets or flexibly between sectors, sites and types of diversified cash income sources (e.g. suppliers. This high transparency and cattle or off-farm employment) replacing substitutability make it difficult for the poor forest based safety nets over time? Are to compete. The same holds true for certain forest based safety nets only a transition ‘green’ certified products with embodied eco- stage? Are we simply failing to see the services. These can be supplied worldwide by new safety nets ‘for the trees’? At what rich and poor producers alike, unless it is an opportunity cost should forest based additional certification attribute that part of safety nets be maintained? Are there ways the extra profits go to poor people. to reduce the risks and fluctuations that On the other hand, if the buyer’s clear made rural people build safety nets in the objective is to protect a specific watershed or first place, so that they can specialise a unique high endemism area, there is no other more in those activities that yield the choice than to deal with the people who happen highest return? When forest based safety to live there—even when they are scattered, nets are truly ‘dynamically competitive’, disadvantaged and disorganised and have how could research and other external overlapping land claims. Services that depend interventions possibly help to enhance essentially on collective action (e.g. in them—other than by simply advocating watershed protection) or on size specific forest conservation? thresholds (e.g. for a biodiversity corridor) can • Efforts to increase the incomes from also have a pro-poor effect because the poor NTFPs have generally not delivered what Exploring the Forest—Poverty Link 41

many had hoped for a decade ago. In and safety nets and between the most cases, NTFPs provide an moderately and the very poor ‘employment of last resort,’ with households’ interests. marginal economic returns. Still, when • In spite of the constraints relating to no alternatives exist, NTFPs provide a the weak land tenure of poor people and vital supplementary income that helps the high transaction costs of making poor people cope, both in terms of their agreements with them, the emerging subsistence uses and potential for cash markets for forest services will provide income. Some high value NTFPs can new opportunities for poor forest contribute a significant amount of cash stewards in four areas: carbon, and households may specialise in their biodiversity, water and tourism. Explicit extraction and managed production. In efforts to design compensation other cases, demand is rising but the schemes and institutional frameworks resource is being mined and hence the in a pro-poor manner can help poor site specific benefits are transitory. people to effectively compete in these • Some wood products (e.g. firewood and markets with larger landholders. low value timbers) have characteristics that are similar to the bulk of non-wood forest products. Hence, as Arnold and 5. Future Research on Townson (1998) conclude for eastern and southern Africa, “much household Forests and Poverty and artisanal involvement in production Additional and better research can guide and trading of forest products is in decisions that help to safeguard or increase the labour-intensive, low-return, activities benefits poor people receive from forests. associated with poor, stagnant rural Research has already yielded key insights in some economic conditions.” Regarding the areas, for example, by pointing to fields where policy options for poverty reduction, in forests are not providing a significant means to these ‘last resort’ cases, “[I]t may reduce poverty and highlighting major obstacles therefore be more fruitful to help that are preventing forests from playing a bigger people move into other more rewarding role in poverty reduction. However, a large fields of endeavour rather than seeking number of questions remain unanswered. As a to raise their productivity in their general observation, there has been a lot of current line of work” (ibid). research done on NTFPs over the past 10 -15 • Logging and the timber industry have years, while studies on how to measure and traditionally yielded small direct benefits increase timber benefits to the poor remain to the poor, due both to anti-poor scarce. Research on ecological service payments policies and certain production has only recently started to build up, as markets characteristics. The indirect benefits in for these services gradually develop. terms of local employment and multiplier In this section, we discuss and suggest effects are in some cases more some topics for further work on the forest- significant but few studies have poverty link. In Table 2, we put forward ten quantified these. There is a marked topics where we believe research would be tendency in the forest literature to ignore particularly valuable—and of either a medium economy wide trickle down effects, or high priority. While the selection and including the effect of forest rents that prioritisation have been influenced by flow into other, faster growing sectors. discussions at CIFOR over the past couple of • In spite of negative preconditions, new years, we realise that such a list is ultimately trends with respect to markets, bound to be of a subjective nature. We group technologies and institutions provide a the topics into three categories and cross potential for raising small scale timber tabulate them with research methods. The first production, both from natural forests, category is all about how to better assess small scale plantations and a multitude current forest based benefits for the poor. The of intermediate systems producing trees. second category explores emerging market In some cases, these opportunities raise opportunities, while the third covers cross- trade offs between income and forest cutting institutional and extra-sectoral issues. conservation, between commercialisation 42 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Table 2. Poverty-Related Forest Research—Topics, Research Methods and Priorities Research area Topic Priority Preferred research methods Analytical Literature Secondary Primary data & conceptual review & data analysis collection models synthesis38 & analysis

I. Exploring (1) Forest products High XXXX the present (subsistence & income) pro-poor role in household livelihood of forests strategies—safety nets and increased welfare (2) Small scale wood High X X X based processing enterprises (3) Economy-wide Medium X X X benefits of forest based rents (4) On-site ecological Medium X X services from forests and trees

II. Emerging (5) Globalisation, High X X X market trends trade liberalisation and and markets opportunities (6) Smallholder tree High X X X planting and private sector partnerships (7) Payments for High X X X environmental services

III. Cross- (8) Local resource Medium X X X X cutting control and land institutional tenure and extra- (9) Decentralisation, Medium X X X sectoral issues governance and market deregulation (10) Integrating Medium X X forests into macroeconomic and poverty strategies

5.1 Research area I: Exploring values, where a better understanding of the role the present pro-poor role of forests in the household economy would have been Many readers may be surprised to see Topic 1 needed. Rigorous quantitative studies of the as a high priority item on our list. After all, type done by Cavendish (2000, 2002), Fisher numerous forest valuation studies have already (2002) or Campbell et al. (2002) are only recent been done over the last decade. Unfortunately, and still very rare.39 There is also a lack of studies their scientific quality and policy impact have shedding light on household behaviour with been variable and the de facto correlation respect to forest products (Mike Arnold, pers. between quality and impact has probably been e-comm., 6 Nov 2001). negative: Some of the most rudimentary One idea that CIFOR is currently pursuing valuation studies with heroic assumptions and is the prospect of working with a number of simplistic conclusions have actually had the universities on coordinated outlines for PhD greatest influence on the international debate theses, using an ex ante agreed upon (Sheil and Wunder 2002). Also, the existing methodology (e.g. Campbell and Luckert 2002 studies have focussed too much on per hectare and elements from the above mentioned Exploring the Forest—Poverty Link 43

Future research on both the poverty prevention/safety net role of forests and on their potential for monetary income generation and poverty reduction will help guide decisions to safeguard or increase the benefits for poor people (like these villagers in Kalimantan, Indonesia). (Photo by Christian Cossalter)

studies) to produce comparable data from material, rather than on marketing, markets and different regions and ecosystems. Forest safety matching demand and supply. The enterprise net functions should also be included in this approach is generally under-represented in the framework—preferably integrating quantitative, forest product literature. A major formalised techniques, such as in Pattanayak recommendation is to link future research to the and Sills (2001), and providing answers to some work on small scale rural enterprises, for instance, of the questions about safety net dynamics as done by the International Food Policy Research raised in Section 4. Only with solid evidence in Institute (IFPRI) (e.g. Haggblade et al. 2002). hand about how forests contribute to the The basis of Topic 3, researching the livelihoods of rural households can one convince economy wide benefits of forest based rents and policy makers and development planners that timber rents in particular, is undoubtedly a forests need to be considered in this light (see difficult task. However, for a handful of countries below). with very high timber export revenues (e.g. in The second theme in Table 2 (small scale Southeast Asia and Central Africa), some effort forest enterprises) has been studied in some in that respect would be very worthwhile. countries, especially in the dry forest regions of Research could also be done on local economies Africa (cf. Section 4). However, we currently lack (e.g. how do Mexican community forestry additional studies from humid areas to determine enterprises reinvest their profits?), including for whether employment generation from forest example, non-timber forest rents (e.g. how do SSEs is just as important in other regions and Asian eaglewood and bird’s nest collectors use ecosystems of the world. Small scale forest their proceeds?). Overriding research questions enterprises have a long tradition and provide include: How large are the forest based rents significant employment in rural areas. Yet, vis-à-vis the rest of the (local or national) attempts to expand these, including experiences economy? How much is reinvested and how much of NTFP commercialisation, have shown mixed is consumed? Approximately what proportion results. Both Arnold (2001) and Scherr et al. stays in the (local or national) economy and what (2001) indicate that policies and external proportion is taken out (e.g. to the capital or to interventions are partly to blame. For example, Swiss bank accounts)? Does the bulk of they have focused too much on access to raw reinvestment remain in forestry/forested 44 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

areas—and if not, why not? What are the welfare mainly to ecological sustainability but and development implications? sometimes also to poverty dimensions, such The subject of Topic 4, ecological on-site as ‘fair trade’ certified commodities, where benefits from forests and trees, is best seen as small primary producers receive a higher than an integral part of the safety net function of normal price. forests (see Topic 1). This can probably best be On the flip side, globalisation also brings researched in connection with the poverty new threats to smallholders. For instance, they prevention functions that products from standing may not be well equipped to compete in forests provide, with the protection benefits ‘green’ certified markets, or the import from keeping trees in increasingly converted liberalisation of forest products in previously landscapes (e.g. in agroforestry) and with the closed economies may out compete small restoration benefits from planting trees in national producers (e.g. this seems to happen degraded landscapes. Especially for natural in Mexico). Expanding sectors might be forests, we know very little about these services subject to economies of scale, so that larger and their relation to changing land uses. A caveat businesses out compete smallholders. More here relates to how much the results of such trade can also lead to over harvesting of forest studies can be generalised and what this means resources, depriving poor people of their for the cost effectiveness of research. access to subsistence uses and small scale As mentioned in Section 4.4, regarding off- commercialisation. site ecological services such as hydrological If they do succeed in competing, small protection, these forest functions can often be producers can become more vulnerable due very site specific. For example, if it takes to fluctuations in the global economy. Thus three years of solid research to find out globalisation provides both opportunities and whether a village’s ongoing forest conversion threats for the poor and their forests. Which to cropland endangers drinking water quality of these effects dominate is an empirical and quantity, if the total cost of that research question and further research can help to could pay for the purchase of drinking water enlighten policies that could influence these corresponding to 20 years of village processes in favour of the poor. consumption and if the results are irrelevant Smallholder tree planting (Topic 6) will to any other village than the one studied, then expand in the future in those areas where wood the research project was probably not justified supplies from natural forests are decreasing in the first place. and where, at the same time, smallholders control a significant share of the land. This trend is fairly advanced in Asia and much effort 5.2 Research area II: Emerging (including large credit programmes by the markets and opportunities World Bank) has gone into small scale The second research area deals with new trends plantations. Yet the most important cases of in markets and economies, which provide both small scale tree planting—in China and India— new opportunities and challenges. The fact that remain poorly documented in the literature. we rank all three topics under this heading as Learning from these experiences would be high priority reflects our belief that structural strategic. changes in the marketplace often assume a Partnerships such as outgrower schemes decisive role for livelihoods. are also set to further increase. Due to what ‘Globalisation’ (Topic 5) has been the we termed ‘anti-poor timber features’ in over-arching buzzword to describe a number Section 4, we believe that partnerships can in of worldwide processes, including trade many cases, present more realistic options liberalisation and the revolution in than the poor ‘going it alone’ in timber communication systems that has facilitated production and marketing. Well designed an extensive flow of information. These partnerships can reduce the potential for processes can certainly provide new economic conflict. As noted by Desmond and Race (2001), opportunities, in terms of improved access governments or NGOs rarely play the role of to international markets and a greater matchmakers. This suggests that researchers openness of economies. They can also open should preferably be working directly with the up certain niche markets, such as certified partners. Concepts such as Adaptive premiums for desirable production Collaborative Management (ACM) may provide characteristics of commodities. This refers useful techniques for future research. Exploring the Forest—Poverty Link 45

Topic 7, payment for environmental Improved legislation is often decisively services, is a strong research candidate with hampered on the ground by implementation similar characteristics. Service markets are obstacles, including corruption. Few studies growing rapidly, previous research has been have dealt with the financial incentives of limited, antagonistic viewpoints abound in the forest agencies—they often act as their own debate and the potential benefits of reconciling profit maximisers, which seldom leads to wise them are considerable. As for outgrower management of the resource. schemes, a main difficulty is in ‘marrying’ Despite widespread economic liberalisation poor people that are weakly organised and in general, many developing countries still have have insecure property rights, with more an amazing amount of forest sector prohibitions influential and resourceful outsiders looking and market regulations. Often these regulations for the efficient and continuous provision of deprive local people of the right to harvest the forest or tree based benefits. If the marriage trees on their own land, thus also opening up fails, both parties may suffer but poor people niches for local level corruption. In a FAO forum seem to have more to lose if they are left out studying six countries, the issues of deregulation of these emerging markets. Research can help and de-bureaucratisation came out as priority to design and recommend institutional action areas in all cases.40 arrangements that can help arrange the One example of anti-poor regulations is marriage in the first place and reduce the the technical requirement for forest probability of future divorce. management plans to be prepared by a professional forester, which is beyond the financial reach of smallholders. The policy 5.3 Research area III: Cross- barriers to participation in forest product cutting institutional and extra- markets are formidable (Scherr et al. 2001). sectoral issues Instead of creating an enabling environment It would seem common sense that full control for local forest enterprises, policies create over land and forest resources by poor local monopolies or oligopolies and protect vested people and communities must be a key interests. We need a better understanding of precondition for improving their lot. The these political processes and the points of recent literature on the devolution of property leverage for change. and use rights to local communities does not Finally, in macro-level poverty literature directly contradict common sense but it does and policy making (see Section 3), generally raise some doubts about how direct this link little recognition is given to the role of forests is. Much depends upon the extent to which (topic 10). This leads to normative statements framework conditions (internal organisation, such as: “The forests can—and must—take government policies, market access, etc.) a far greater role in meeting the UN Millennium allow poor people to make their own assets Summit target of halving extreme poverty by productive. In many cases, it takes time to 2015” (World Bank, 2001). Yet in the World judge whether these processes are successful Bank promoted Poverty Reduction Strategy or not and to define what determines their Papers (PRSPs), only limited attention is given outcome. Countries where local communities to environmental and forestry issues. This have owned their forests for a long time, such also indicates inconsistencies in the World as Mexico (75% of forests owned by the ejidos) Bank’s sector policies.41 One way of resolving or Papua New Guinea (99% of forests owned by the debate would be for the ‘forest lobby’ of communities), offer good opportunities to study foresters and conservationists (and their suite ‘what works and what doesn’t’ in the long run of NGO activists, consultants and researchers) for forests and (poor) people (Topic 8). to convince the hard core economists, policy A linked set of factors concerns makers and development planners that they institutional and policy issues, such as have come to ‘forget’ an important sector. decentralisation, governance and market The other extreme would be for the forest deregulation (Topic 9). Decentralisation should lobby to recognise, after a closer look, that provide pro-poor opportunities but some ‘their’ sector was actually just as marginal to experiences imply that when budget national poverty alleviation as the hard core allocations to local authorities are insufficient developers had implicitly stated by ignoring or when local elites come to monopolise power, forests. The latter extreme seems unlikely as the net outcome for the poor can be negative. forests probably have some role in poverty 46 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

alleviation—though often more in its prevention 7 Figures are according to a recalculation than its reduction, as we have argued in this done by Bruce Campbell (pers e-comm, 14 Oct paper. There is scope for integrating forest 2002). Cavendish’s own ‘environmental issues into poverty concerns (particularly resources’ include wild vegetables (mostly from national poverty reduction strategies) and vice old agricultural fields and as weeds), gold versa, for example, into National Forest panning, soil for pottery and natural fertilizers. Programmes (NFPs). Where exactly in between 8 This would not mean that everybody the two extremes the debate will eventually receives the same absolute amount of extra settle is an open question, depending on the income but that everybody’s income rises by the specific country situation and notably, on what same percentage. research and practical experiences the previous 9 The caveat of ‘continuous’ applies nine topics reveal. because of the depth of poverty. If a household is far below the poverty line initially, is economic growth sizeable enough to bring it above the Endnotes poverty line? 10 The Gini coefficient is a quantitative 1 Both authors participated in FAO’s forum, measure indicating what share of total income “The Role of Forestry in Poverty Alleviation”, mass would hypothetically have to be held 4-8 September 2001, in Cortevecchia, Italy redistributed in order to make the current and both contributed to FAO’s policy brief (FAO distribution fully equal (i.e. with every recipient 2001). having an equal ‘share of the cake’). 2 According to Webster’s Dictionary (1993), 11 Other hypotheses to explain why equality ‘capital’ is defined as: “A stock of accumulated is positive are based on political economy goods devoted to the production of other arguments: Governments in highly unequal goods in contrast to income received”. For countries face a stronger pressure for physical, financial, human and perhaps also redistribution yet various redistribution natural capital, that would normally make sense. measures (e.g. taxation) tend to hamper But even with a lot of good will, it seems futile growth, at least in the short to medium term. to try to construct a concept of ‘cultural capital’ 12 The sequence of user groups in Table 1 or ‘political capital’ that has any relation to the represents an important socio-evolutionary original meaning of the word. history. Many human cultures originated in 3 Even dimensions like ‘power’ can be forests and forest dependence is to a significant measured quantitatively. For example, the degree, determined by a society’s evolutionary number of non-related guests that attend a stage. family’s wedding or funeral can be taken as an 13 The classification follows Byron and Arnold indicator of that family’s power (M. Zeller, pers. (1999), except group no. 4, which has been e-comm., 18 Feb 2002). The difficulty here added. seems to be more in developing consistent and 14 This excludes payment for off-site comparable measures at larger scales. ecological services, which we discuss in Section 4 In addition, a number of more personalised 4.4. norm-deviation factors also trigger a 15 There are also other important inputs into significantly higher average probability of agriculture from forests, which are not dealt with unhappiness: divorce or separation, having in this section. These include fodder for more than three children or belonging to a livestock and green mulch, and represent key generation with ‘boom-age’ parents and thus inputs into farming systems in some areas (e.g. easily developing unrealistic expectations. Cavendish 2000,2002). 5 However, one more ‘objective’ reason 16 Other dimensions for which data were could be the relatively egalitarian societies and collected include: geographic setting; product the security provided by the Scandinavian characteristics; production system; ecological welfare states. implications of production; socioeconomic 6 Note that UNDP’s HDI index itself normally features of the raw material production area; cannot be used in sub-national level studies. processing industry and trade; institutional The life expectancy dimension requires repeated characteristics of producers; government demographic measurements, which are normally policies and type of external interventions only available at the national scale (M. Zeller, (Belcher and Ruiz-Pérez 2001). pers. e-comm., 18 Feb 2002). Exploring the Forest—Poverty Link 47

17 The fourth cell (“high forest-product industries, which have not been measured and contribution; low market orientation”) was compared here. empty. Among the other cells, the frequency was 30 A contrasting case here would be 18 (subsistence), 28 (diversified) and 15 payments for tolerable land use changes. For (specialised). However, the distribution should example, the substitution of natural forests for not be over interpreted, since the cases were agroforests instead of monocultures, thus not selected randomly. safeguarding a range of services in a changing 18 See Scherr et al. (2001) and Neumann landscape (which would otherwise have been and Hirsch (2000) for a comprehensive heavily degraded). discussion. 31 Yet, the status quo assumption may 19 In some cases, the causation goes the sometimes be problematic. For example, when other way—access to timber concessions made reforestation of an area would have been done the rich people rich in the first place. even without carbon payments, raising the 20 We are grateful to William Sunderlin for question of additionality. suggesting four of these points. 32 The BioCarbon Fund, recently launched 21 See also Scherr et al. (2001) for a more by the World Bank, is a prototype fund for comprehensive review of opportunities. projects that sequester or retain carbon in forests 22 Own observation from Pará state, Brazil and agro-ecosystems. Its target of USD 100 (Wunder). million is meant to fund projects outside the 23 A third option here is that part of the CDM and JI mechanisms. Examples of pre- high timber price is not economic rent at all but identified projects include forest rehabilitation a reimbursement of the often under-estimated in Uganda and forest protection in the Dominican costs/value adding in transport, processing and Republic. marketing. 33 It has been shown that, under most 24 The countries included and the share of scenarios, one dollar spent by a conservation forest based SSE employment: Bangladesh (13 donor on direct payments will create significantly %), Honduras (16 %), Sierra Leone (20 %), Egypt more income than spending the same dollar on (24 %), Zambia (33 %) and Jamaica (35 %). Most ICDP projects (Ferraro and Simpson 2002). studies were done around 1980. Firewood and 34 See also Wells and Brandon (1992), charcoal were not included in the surveys. Gilmour (1994) or Brandon et al. (1998). Many 25 Botswana, Kenya, Lesotho, Malawi, conservation donors and agencies are Swaziland and Zimbabwe. increasingly adopting people centred agendas 26 2.3 % might not sound impressive but that resemble those of development one should keep in mind that in many villages organisations and include poverty alleviation only a minority of the population would be goals. An implicit but highly dubious assumption involved in off-farm work. behind this is that the interests of forest 27 These FAO data are also reported by conservation and economic development are Whiteman (2000: Table 2). Both in Whiteman synergistically linked (Wunder 2001). and FAO (1997), reference is made back to other 35 This discussion is parallel to the one on FAO documents and we were not able to trace the role of agricultural technologies and the original source. The data are particularly intensification in containing deforestation, see doubtful because the share of forestry in exports Angelsen and Kaimowitz (2001). is supposed to be less than 1% (FAO 1997), so 36 On the other hand, in those situations almost the entire 23% of GDP would therefore where agricultural expansion only occurs in be for subsistence use and the domestic market. tandem with logging penetration (e.g. after road From one of the author’s (Angelsen) knowledge building, etc.), from a conservation viewpoint of Uganda, we can safely say that the FAO figure it may be ‘sufficient’, at least in the short term, must be significantly off the mark. to impede logging because that will also 28 Poschen points to Chile as an illustrative jeopardise agricultural expansion. case study, where new plantations at the expense 37 Of course, if there are site specific of small scale cropping (especially in Region IX) substitutes, the argument will be reversed. For have led to large macroeconomic benefits but instance, that would be the case where urban have also sharply reduced on-site rural drinking water can be obtained from two employment (Poschen 2001). alternative watersheds, one dominated by 29 Both forestry and agriculture will also large landowners and the other by create downstream jobs in processing smallholders. 48 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

38This should include particularly the ‘grey’ International Forestry Research (CIFOR), literature, e.g. project reports and Master/PhD Bogor, Indonesia. theses. Arnold, J.E.M., Liedholm, C., Mead, C. and 39 While better general household data in Townson, I.M. 1994 Structure and growth many developing countries help to guide poverty of small enterprises in the forest sector in reduction policies, little such progress has been southern and eastern Africa. OFI Occasional made in marginal, forested areas. Paper No. 47. Oxford Forestry Institute, UK. 40 Country profiles for Bolivia, Honduras, Barro, R.J. 1999 Inequality, growth and Mali, Tanzania, Nepal and Viet Nam, ‘The Role investment. Working Paper No. 7038. of Forestry in Poverty Alleviation’, Cortevecchia, National Bureau of Economic Research Italy, 3-7 September 2001 (see FAO 2001). (NBER), Cambridge, MA. 41 For example, in Tanzania, forests and Bass, S.O., Dubois, P., Moura Costa, M., Pinard, woodlands are critical cash income earners for R., Tipper and Wilson, C. 2000 Rural the poor, accounting for as much as 50 % of livelihoods and carbon management. IIED the income in selected areas. As forest Natural Resources Issues Paper No. 1. degradation is widespread, one would expect International Institute for Environment and woodlands to receive high priority in poverty Development (IIED), London. reduction plans. However, the International Baumann, P. 2000 Sustainable livelihoods and Monetary Fund and World Bank’s Joint Staff political capital: arguments and evidence Assessment of the Tanzanian PRSP makes no from decentralisation and natural resource mention of the environment (including management in India. Sustainable forests) at all. It largely argues that economic Livelihoods Working Paper No. 136. and structural reforms must continue and that Overseas Development Institute (ODI), these will eventually benefit the poor. See also London. Oksanen and Mersmann (2002) for a Bebbington, A. 1999 Capitals and capabilities: pioneering review of how forests have been a framework for analyzing peasant viability, included in PRSPs. rural livelihoods and poverty. World 42 http://www.cifor.cgiar.org/aboutus/ Development 27 (12): 2021-2044. (accessed on 5 Oct 2001). Belcher, B. and Ruiz-Perez, M. 2001 An international comparison of cases of forest product development: overview, description References and data requirements. Working Paper No. 23. Center for International Forestry Adelman, I. and Morris, C.T. 1978 Growth and Research (CIFOR), Bogor, Indonesia. impoverishment in the middle of the 19th Bishop, J. and Landell-Mills, N. 2002 Forest century. World Development 6 (3): 245-73. environmental services: an overview. In: Angelsen, A. 1997 The poverty-environment Pagiola, S., Bishop, J. and Landell-Mills, N. thesis: was Brundtland wrong? Forum for (eds.) Selling Forest Environmental Development Studies 1997 (1): 135-154. Services, 15-36. Earthscan, London. Angelsen, A. and Kaimowitz, D. (eds.) 2001 Bornschier, V. 1983 World economy, level Agricultural technologies and tropical development and income distribution: an deforestation. CAB International, integration of different approaches to the Wallingford, UK. explanation of income inequality. World Arnold, J.E.M. 2001. Forestry, poverty and aid. Development 11 (1): 11-20. CIFOR Occasional Paper No. 33. Center for Brandon, K., Redford, K. and Sanderson, S. 1998 International Forestry Research (CIFOR), Parks in peril: people, politics, and protected Bogor, Indonesia. areas. Covelo, Island Press, CA. Arnold, J.E.M. and Townson, I. 1998 Assessing Brekke, K.A. and Howarth, R.B. 2000 The social the potential of forest product activities to contingency of wants. Land Economics 74 contribute to rural incomes in Africa. ODI (6): 493-503. Natural Perspectives 37: 1-4. Bruijnzeel, L.A. 1990 Hydrology of moist tropical Arnold, J.E.M., Köhlin, G., Persson, R. and forests and effects of conversion: a state of Shepherd, G. 2003 Fuelwood revisited: the knowledge review. Netherlands what has changed in the last decade? CIFOR Committee for the International Hydrological Occasional Paper No. 39. Center for Programme of UNESCO, Amsterdam. Exploring the Forest—Poverty Link 49

Byron, R.N. and Arnold, J.E.M. 1999 What futures lease. Global Environmental Change Report for the people of the tropical forests? World XII (19): 1-2. Development 27 (5): 789-805. de Janvry, A.I. and Sadoulet, E. 2000 Rural Calder, I.R. 1998 Water-resource and land-use poverty in Latin America. Determinants and issues. SWIM Paper 3., International Water exit paths. Food Policy 25: 389-409. Management Institute, Colombo, Sri Lanka. Deininger, K. and Squire, L. 1998 New ways of Campbell, B.M. and Luckert, M. (eds.) 2002 looking at old issues: inequality and growth. Uncovering the hidden harvest: valuation Journal of Development Economics 57 (2): methods for woodland and forest resources. 259-287. Earthscan, London. Desmond, H. and Race, D. 2001 Global survey Campbell, B.M., Jeffrey, S., Luckert, M., and analytical framework for forestry out- Mutamba, M. and Zindi, C. 2002 Household grower arrangements. Food and Agriculture livelihoods in semi-arid regions: options and Organization of the United Nations (FAO), constraints. Center for International Rome. Forestry Research (CIFOR), Bogor, Dollar, D. and Kraay, A. 2001 Growth is good for Indonesia, p. 106. the poor. World Bank, Washington D.C. Carney, D. (ed.) 1998 Sustainable rural Dove, M. 1993 A revisionist view of tropical livelihoods. What contributions can we deforestation and development. make? Department for International Environmental Conservation 20 (1): 17-24. Development (DFID), London. Duraiappah, A. 1996 Poverty and environmental Cavendish, W. 2000 Empirical regularities in the degradation: a literature review and poverty-environment relationship of rural analysis. CREED WP Series No 8. households: evidence from Zimbabwe. International Institute for Environment and World Development 28 (11): 1979-2000. Development (IIED), London and Institute Cavendish, W. 2002 Quantitative methods for for Environmental Studies (IES), estimating the economic value of resource Amsterdam. use to rural households. In: Campbell, B.M. EPMR 1998 Report of the first External and Luckert, M. (eds.) Uncovering the Programme and Management Review of the hidden harvest: valuation methods for Center for International Forestry Research woodland and forest resources, 17-65. (CIFOR), TAC Secretariat, Food and Earthscan, London. Agriculture Organization of the United Chambers, R. and Conway, G. 1992 Sustainable Nations. rural livelihoods: practical concepts for the FAO 1987 Small-scale forest-based processing 21st century. IDS Discussion Paper, No. 296. enterprises. FAO Forestry Paper 79. Food Institute for Development Studies (IDS), and Agriculture Organization of the United Brighton. Nations (FAO), Rome. Chomitz, K.M. and Kumari, K. 1998 The domestic FAO 1997 State of the world’s forests. Food and benefits of tropical forests: a critical review. Agriculture Organization of the United The World Bank Research Observer 13 (1): Nations (FAO), Rome. 13-35. FAO 2001 How forests can reduce poverty. Forest CIAT 1999. Pathways out of poverty. CIAT in Department Policy Brief. Food and Perspective. International Center for Tropical Agriculture Organization of the United Agriculture (CIAT), Palmira, Colombia. Nations (FAO), Rome, p. 36. CIFOR 1995 Management response to the 1995 Ferraro, P.J. 2000 Global habitat protection: Internally Commissioned External Review limitations of development interventions and (ICER). Mimeo, Center for International a role for conservation performance Forestry Research (CIFOR), Bogor, payments. Department of Applied Economics Indonesia. and Management Working Paper No. 2000- CIFOR 2001 Forest carbon, livelihoods and 03. Cornell University, Ithaca, USA. biodiversity. A report to the European Ferraro, P.J. and Simpson, R.D 2002 The cost- Commission. Unpublished report, Center for effectiveness of conservation payments. International Forestry Research (CIFOR), Land Economics 78 (3): 339-53. Bogor, Indonesia. Fields, G.S. 1980 Poverty, inequality and Cutter Information Corp. (CIC) 2000 Focus development, Cambridge University Press, report: saving the forest with a timber Cambridge, UK. 50 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Fields, G.S. 2001 Distribution and development: J.E. (eds.) Frontiers in development a new look at the developing world. economics. Oxford University Press, New Cambridge, New York and Russel Sage York. Foundation and MIT Press, London, p. 99. Kanbur, R. 2001 Economic policy, distribution and Fisher, R.J. 2001 Poverty alleviation and forests: poverty. The nature of disagreements. experiences from Asia. RECOFTC, Kasetsart World Development 29 (6): 1083-1094. University, Bangkok. www. RECOFTC.org/ Kravis, I.B. 1960 International differences in the publications_ at_recoftc. htm distribution of income. Review of Economics Fisher, M. 2002 Household welfare and forest and Statistics November: 408-16. dependence in rural Malawi. Draft submitted Kuznets, S. 1955 Economic growth and income to Environment and Development inequality. American Economic Review XLV Economics, University of York, UK. (1): 1-28. Grieg-Gran, M. 2000 Fiscal incentives for Kuznets, S. 1963 Quantitative aspects of the biodiversity conservation: the ICMS economic growth of nations: distribution Ecológico in Brazil. Discussion Paper No. 00- of income by size. Economic Development 01. International Institute for Environment and Cultural Change 11 (2) Part II: 1-80. and Development (IIED), London. Landell-Mills, N. and Porras, I.T. 2002 Silver Gurung, C.P. and Coursey, M.D. 1994 Nepal: bullet or fools’ gold? A global review of pioneering sustainable tourism. The markets for forest environmental services Annapurna Conservation Area Project: an and their impacts on the poor. Instruments applied experiment in integrated for Sustainable Private Sector Forestry conservation and development. The Rural Series. International Institute for Extension Bulletin 5 (August), University of Environment and Development (IIED), Reading, UK. London, p. 73-74, 143-145. Haggblade, S., Hazell, P. and Reardon, T. 2002 Lindberg, K. and Enríquez, J. 1994 An analysis of Strategies for stimulating poverty-alleviating ecotourism’s economic contribution to growth in the rural nonfarm economy in conservation in Belize. Volume 2: developing countries. Discussion paper No. Comprehensive Report. Ministry of Tourism 92. International Food Policy Research and the Environment, Belize and World Institute (IFPRI), Washington D.C. Wildlife Fund, Washington D.C. Hardner, J. and Rice, R. 2002 Rethinking green Lindert, P. H. and Williamson, J.G. 2001 consumerism. Scientific American May: Globalization and inequality: a long history. 89-95. Paper read at World Bank Annual Bank Hill, H. 1992 Regional development in a boom Conference on Development Economics - and bust petroleum economy: Indonesia Europe, Barcelona, 25-26 June, p. 36. since 1970. Economic Development and May, P.H., Veiga Neto, F., Denardin, V. and Cultural Change 40 (2): 351-79. Loureiro, W. 2002 Using fiscal instruments ICER 1995 Internally Commissioned External to encourage conservation: municipal Review. CIFOR, 28 September, Center for responses to the “ecological” value-added International Forestry Research (CIFOR), tax in Paraná and Minas Gerais, Brazil. In: Bogor, Indonesia. Pagiola, S., Bishop, J. and Landell-Mills, N. IPCC 2000 Land use, land-use change and forestry. (eds.) Selling forest environmental A special report of the intergovernmental services. Market-based mechanisms for panel on climate change. Cambridge conservation and development, 173-200. University Press, Cambridge, UK. Earthscan, London. Kaimowitz, D. and Angelsen, A. 1998 Economic Mayers, J. 2000 Company-community forestry models of tropical deforestation: a review. partnership: a growing phenomenon. Center for International Forestry Research Unasylva 200 (51): 33-41. (CIFOR), Bogor, Indonesia. Mayers, J., Evans, J. and Foy, T. 2001 Raising Kaimowitz, D. 2002 Not by bread alone. Forests the stakes. Impact of privatisation, and rural livelihoods in sub-Saharan Africa. certification and partnership in South Center for International Forestry Research African forestry. International Institute for (CIFOR), Bogor, Indonesia. Environment and Development (IIED), Kanbur, R. and Squire, L. 2001 The evolution of London. thinking about poverty. Exploring the Mayers, J. and Vermeulen, S. 2002 Company- interactions. In: Meier, G.M. and Stiglitz, community forestry partnership. From raw Exploring the Forest—Poverty Link 51

deals to mutual gains? International microeconomics of non-timber forest Institute for Environment and Development product collection in the Brazilian Amazon. (IIED), London. Land Economics 77 (4): 595-612. Meadows, D.H. et al. 1972 The limits to growth. Poschen, P. 1997 Forests and employment, much Universe Books, New York. more than meet the eye. World Forestry Mellor, J. 1999 Pro-poor growth - the relation Congress, Antalya, Turkey, October 13-27. between growth in agriculture and poverty Volume 4, Topic 20, Food and Agriculture reduction. Unpublished paper prepared for Organization of the United Nations (FAO), USAID. Rome. http://www.fao.org/montes/foda/ Narayan, D., Chambers, R., Shah, M.K. and wforcong/PUBLI/V4/T20E/1-5.HTM#TOP Petesch, P. 2000 Voices of the poor. Crying Poschen, P. 2001 Social and labour dimensions of out for change. Oxford University Press, the forestry and wood industries on the New York. move. TMFWI/2001. International Labour Nawir, A.A., Santoso, L. and Mudhofar, I. 2002 Organization (ILO), Geneva, p. 3-33, 70-71. Towards Mutually-beneficial partnership in Ravallion, M. 1997 Good and bad growth: the outgrower schemes: lessons learnt from Human Development Reports. World Indonesia. Center for International Forestry Development 25 (5): 631-638. Research (CIFOR), Bogor, Indonesia. Ravallion, M. 2001 Growth, inequality and Neumann, R.P. and Hirsch, E. 2000 poverty: looking beyond averages. World Commercialisation of non-timber forest Development 29 (11): 1803-1815. products: review and analysis of research. Ravallion, M. and Datt, G. 2002 Why has Center for International Forestry Research economic growth been more pro-poor in (CIFOR), Bogor, Indonesia and Food and some states of India than others? Journal Agriculture Organization of the United of Development Economics 68 (2): 381- Nations (FAO), Rome, p. 35. 400. Offer, A. 2000 Economic welfare measurements Ray, D. 1998 Development economics. Princeton and human well-being. Discussion Papers University Press, Princeton, NJ. in Economic and Social History, No. 34. Ravnborg, H.M. 2002 Poverty and soil University of Oxford, UK. management. Relationships from three Oksanen, T. and Mersmann, C. 2002 Forests in Honduran watersheds. Society and Natural poverty reduction strategies. An assessment Resources 15: 523-39. of the PRSP processes in sub-Saharan Africa. Rodrick, D. 2000 Institutions for high-quality Paper presented at the international growth: what they are and how to acquire workshop: Forests in Poverty Reduction them. NBER Working Paper W7540. National Strategies - Capturing the Potentials. Bureau of Economic Research (NBER), Helsinki, Finland, 1-2 October. Cambridge, MA. Pagiola, S. 2001 Paying for water services. Forest Ross, M. 2001 Timber booms and institutional Trends Workshop on New Markets for a breakdown in Southeast Asia. Cambridge Green Economy, Teresópolis, Brazil, 24-26 University Press, Cambridge, UK. March. http://www.forest-trends.org/ Ruggeri-Laderchi, C. 2001 Participatory methods whoweare/pdf/brazil2001/Paying%20for%20 in the analysis of poverty: a critical review. Water%20Services.pdf (18 Oct. 2001). Working Paper No. 62. University of Oxford, Pagiola, S. and Ruthenberg, I.M. 2002 Selling UK. biodiversity in a coffee cup. In: Pagiola, S., Ruitenbeek, H.J. 1988 Social cost-benefit Bishop, J. and Landell-Mills, N. (eds.) analysis of the Korup Project, Cameroon, Selling forest environmental services. World Wide Fund for Nature, London. Market-based mechanisms for conservation Ruiz-Perez, M., Belcher, B. et al. 2003 Global and development, 103-126. Earthscan, patterns and trends in non-timber forest London. product use. Manuscript. Center for Pagiola, S., Bishop, J. and Landell-Mills, N. 2002 International Forestry Research (CIFOR), (eds.) Selling forest environmental Bogor, Indonesia. services. Market-based mechanisms for Sala-i-Martin, X. 2002 The world distribution of conservation and development. Earthscan, income (estimated from individual country London, p. 282-4. distributions). NBER Working Paper 8933. Pattanayak, S.K. and Sills, E.O. 2001 Do tropical National Bureau of Economic Research forests provide natural insurance? The (NBER), Cambridge, MA. 52 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Sayer, J. A. et al. 2000 World Heritage forests: Thomas, V., Aailami, M., Dhareshwar, A., the World Heritage conservation as a Kaufmann, D., Kishor, N., López, R. and mechanism for conserving tropical forest Wang, Y. 2000 The quality of growth. World biodiversity. Government of Indonesia, Bank, Washington D.C. UNESCO and Center for International Vincent, J.R., Ali, R.M. and Associates 1997 Forestry Research (CIFOR), Bogor, Indonesia. Environment and development in a resource Scherr, S.J. 2000 A downward spiral? Research rich economy. Malaysia under the new evidence on the relationship between economics order. Harvard University Press, poverty and natural resource degradation. for (Harvard Institute for International Food Policy 25: 479-98. Development) HIID, Cambridge, MA. Scherr, S.J., White, A. and Kaimowitz, D. 2001 Vincent, J.R. and Binkley, C.S. 1992 Forest-based Strategies to improve rural livelihoods industrialization: a dynamic perspective. In: through markets for forest products and Sharma, N.P. Managing the World’s Forests, services. Forest Trends, Washington D.C., 93. Kendall/Hunt Publishing Company, p. 44. Dubuque, Iowa. Scoones, I. 1998 Sustainable rural livelihoods: Warner, K. 2000 Forestry and sustainable a framework for analysis. IDS Working Paper livelihoods. Unasylva 51 (202): 3-12. No. 72. Institute for Development Studies Warr, P.G. 2000 Indonesia: Trade shocks and (IDS), Brighton. construction booms. In: Collier, P. and Sen, A. 2001 Development as freedom, Anchor Gunning, J.W. (eds.) Trade shocks in Books, New York. developing countries. Volume 2: Asia And Sheil, D. and Wunder, S. 2002 The value of Latin America. Oxford University Press, tropical forests to local communities: Oxford, pp. 326-54. complications, caveats and cautions. Webster 1993 [1961] Webster’s third new Conservation Ecology 6 (2): 9. http:// international dictionary of the English www.consecol.org/vol6/iss2/art9 language unabridged, Merriam-Webster Shepherd, G., Arnold, M. and Bass, S. 1999 Inc., Springfield, Massachusetts. Forests and sustainable livelihoods. A Wells, M.P. 1997 Economic perspectives on contribution to the World Bank Forest Policy nature tourism, conservation and review process, September. http:// development. Environmental Economics wbln0018.worldbank.org/essd/forestpole. Series, Environment Department Paper nsf/MainView No.55. The World Bank, Washington D.C. Smith, J., Mulongoy, K., Persson, R. and Sayer, Westoby, J.C. 1962 The role of forest industries J. 2000 Harnessing carbon markets for in the attack on under-development. The tropical forest conservation: towards a state of food and agriculture 1962. Food more realistic assessment. Environmental and Agriculture Organization of the United Conservation 27 (3): 300-311. Nations (FAO), Rome (reprinted In: Westoby Smith, J. and Scherr, S.J. 2002 Forest carbon J.C. The purpose of forests. Blackwell). and local livelihoods: assessment of Westoby, J.C. 1978 Forest industries for socio- opportunities and policy recommendations. economic development. Commonwealth CIFOR Occasional Paper No. 37. Center for Forestry Review 58 (2): 107-16 (reprinted International Forestry Research (CIFOR), In: Westoby J.C. The purpose of forests. Bogor, Indonesia. Blackwell). Sunderlin, W.D., Angelsen, A. and Wunder, S. White, A. and Martin, A. 2002 Who owns the 2003 Forests and poverty alleviation. In: world’s forests? Forest tenure and public FAO, State of the world’s forests, 61-73. forests in transition. Forest Trends, Food and Agriculture Organization of the Washington, D.C. United Nations (FAO), Rome. Whiteman, A. 2000 Promoting rural The Zimbabwe Trust, The Department of National development through forestry policy: some Parks and Wildlife Management and The experiences from developing countries. Campfire Association 1994 Zimbabwe: Presentation to the seminar: The Role of tourism, people and wildlife. The Rural Forests and Forestry in Rural Development— Extension Bulletin 5 (August), University of Implications for Forest Policy. July 5-7, Reading, UK. Vienna, Austria. Exploring the Forest—Poverty Link 53

Williamson, J.G. 1985 Did British capitalism Ecuadorian Amazon region, CIFOR breed inequality? Routledge, London. Occasional Paper No. 21. Center for Williamson, J.G. 1997 Industrialization, International Forestry Research (CIFOR), inequality and economic growth. Edward Bogor, Indonesia. Elgar, London. Wunder, S. 2000 Big island, green forests and World Bank 2000 World development report backpackers. Land-use and development 2000/2001: attacking poverty. World Bank, options on Ilha Grande, Rio De Janeiro State, Washington D.C. Brazil. CDR Working Paper 00.3. Centre for World Bank 2001 A revised forest strategy for Development Research (CDR), Copenhagen. the World Bank group. World Bank, Wunder, S. 2001 Poverty alleviation and tropical Washington D.C., Draft 30 July, p. 14-15. forests - what scope for synergies? World Wunder, S. 1987 Økonomisk udvikling og Development 29 (11): 1817-1833. indkomstfordeling - en analytisk oversigt Zeller, M., Sharma, M., Henry, C. and Lapenu, og et cross-country studie. Master thesis, C. 2001 An operational tool for evaluating Institute of Economics, University of poverty outreach of development policies Copenhagen. and projects. IFPRI, FCND Discussion Paper, Wunder, S. 1999 Forest conservation through No.111. International Food Policy Research ecotourism income? A case study from the Institute (IFPRI), Washington D.C. 54 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

Appendix. Strategic Questions for Poverty-Forest Research at CIFOR

This appendix deals with some strategic argument is therefore less valid today. Note that questions for future poverty research the focus on “national development” implies an specifically at CIFOR but would also be of interest explicit recognition that the forestry sector needs to other institutions conducting research in this to deliver goods and services not only to field. Here, in a slightly more provocative marginalised forest dwellers but also to the manner, we put forward the pros and cons, vis-à- whole of society. vis a series of strategic research choices. We do CIFOR’s degree of poverty orientation is not always provide definitive viewpoints, not a new discussion topic—it has been a preferring to sometimes simply provide both the recurrent point of debate since CIFOR was arguments for and against so partisans in different created in 1993. CIFOR’s Internally camps can make their own judgements. Commissioned External Review (ICER) from 1995 raised two concerns. Firstly, an “ambiguous level of interest in tropical dry forests” was noted to CIFOR and poverty—in the past be poverty-relevant because “these forests are Poverty is currently a ‘hot’ topic in the able to support a higher population density”. international debate but how dominant should Plantations were also outlined as an under- it be in CIFOR research? If we compare the world prioritised area with a poverty reduction today to 1993, when CIFOR was formed, the potential. Secondly, the review team observed proportion of absolute income-poor in developing that CIFOR focused “on macro-level processes countries has on average, actually declined. rather than farmer-level processes in its Hence, it is legitimate to ask the provocative research thrusts”, implying that little research question: To what extent should changes (of was targeted to poor people. As the report unpredictable duration) in donor trends and stated: “There is a tendency for forestry political winds dictate CIFOR’s research agenda? research and the resulting interventions to offer CIFOR’s current mission statement reflects the too little to poor people unless explicit attempts institution’s double mandate, encompassing to do so have been built in”. Hence, the critical both environmental and people centred concerns assessment was that “there are potential win- and objectives: win situations.... but often [the] research is at “To contribute to the sustained well-being best irrelevant to poor people” (ICER 1995). of people in developing countries, CIFOR’s management responded to the ICER particularly in the tropics, through (CIFOR 1995) in a way that acknowledged both collaborative strategic and applied research options and limitations to pushing forward the and related activities in forest systems and poverty agenda in an international forest forestry, and by promoting the transfer of research organisation: appropriate new technologies and the • “Most of CIFOR’s forestry NARS adoption of new methods of social partners do not consider poverty organisation, for national development” alleviation to be a primary or even (our emphases).42 important focus.” • “In the future, forests in sub-humid and Rather than referring directly to poverty and drier tropics should figure more its alleviation, CIFOR’s mission speaks of the prominently in our activity portfolio.” “sustained well-being of people”—a somewhat • “Funding for CIFOR comes from broader concept, as discussed in Section 2. At conservation budgets within aid the time of writing the mission statement, agencies whose prime concern is poverty was deliberately not chosen as many felt environmental protection and not it implied a narrow income focus (G. Shepherd, poverty alleviation. There is therefore pers.comm., Nov 2001). Since then however, the an inevitable mismatch between poverty concept has been broadened in the CIFOR’s research agenda and that of international debate (cf. Section 2) and this the CGIAR.” Exploring the Forest—Poverty Link 55

Three years later, the External Programme FSEP. Unlike what was claimed above in the ICER and Management Review of CIFOR (EPMR 1998) (1995), tropical forest win-win options prove to stated that the institute’s broad well-being be quite limited (Wunder 2001). In many real approach already incorporated poverty concerns world situations, making people better off tends well and provided an appropriate bridge between to have environmental costs and protecting its environmental mandate and that of the forests tends to restrict people’s development CGIAR (Consultative Group on International and poverty reduction aspirations. Faced with Agricultural Research): these trade offs, PAFSEP might become like a • “This broader interpretation of poverty piece of wet soap in the hands of the researcher issues provides a logical, and indeed trying to apply the concept. necessary, link between CIFOR’s mission In the light of this less optimistic worldview, and the goals of CGIAR related to poverty several options emerge. First, one might take alleviation plus natural resources and the position that CIFOR should only research and environmental protection and promote actions that benefit both people and enhancement.” forests at the same time, hence confining itself • “CIFOR’s programme expands this to the ‘win-win’ sphere. That would also include interpretation to include consideration the reversal of ‘lose-lose’ policies and outcomes, of [the] means of poverty prevention— which neither benefit the poor nor the forests. poverty for future generations and It could also embrace longer term efforts to future poverty for those who currently reduce the trade offs (i.e. turning win-lose/lose- are living in adequate conditions.” win into win-win situations). However, to stay away from all development-conservation The two different reviews of CIFOR’s controversies would radically limit CIFOR’s radius poverty orientation thus also reveal how much of action. It is the trade offs, not the win-win a poverty oriented evaluation depends on the scenarios that abound in the world of tropical poverty concepts applied (see Section 2). forests. Currently, all of CIFOR’s three research Second, some might find that a poverty programmes work with issues that to some orientation should officially take priority over extent relate to poverty and poor people. In light conservation. In that case, poverty reduction of the comments in ICER in 1995, much more should be our over-arching concern, while the work is carried out at the micro level but there solutions we work for should, as far as possible, is a continued discussion about the balance not hurt the environment. Also, one might between micro-oriented case studies vs. question to what extent it is the legitimate role national and global-comparative studies. Poverty of a research organisation to advance policy issues are now much higher on the agendas of statements when they have strong normative donors, while this is still much less the case for components (e.g. regarding the redistribution the NARS (National Agricultural Research of resources). Further, environmental concern Systems). and conservationist value judgements were clearly the driving force behind CIFOR’s creation in the early 1990s. To make forest conservation How important should poverty a subordinate concern may ultimately bewilder alleviation be relative to other the institution’s core constituency and erode goals? both its image and a significant part of its CIFOR has recently attempted to think funding base. strategically about its impact in terms of the A third option is to be confined to CGIAR’s summary goal of “poverty alleviation, presenting the alternatives—make the trade offs food security and environmental protection” explicit and let others make the choices. That (PAFSEP). This consensus concept is attractive would mean that in most situations we should from an institutional planning viewpoint. not strive for impact (which presupposes a clear Everybody in the organisation is supposed to definition of desirable outcomes) but rather for work towards the same overall super goal. influence (enabling others to make more However, in applied field situations, many of informed choices on a menu of options). To us, CIFOR’s researchers have found it difficult to this sounds like the most appealing alternative. give real meaning to PAFSEP. In practice, there However, we recognise that the choices of proves to be mostly a split between the goals research topics, research design and PAFS and EP and sometimes between PA and methodologies etc., are eventually all influenced 56 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

by subjective (and normative) valuations. The generation and poverty reduction? The simple message from the positivism debate in the answer is both! But the question of relative 1970s is still valid: Try to be objective but don’t emphasis remains. This has a scientific believe that you are! dimension—what is most important to the poor Making the trade offs explicit will thus often and what are the likely potentials of forests? bring us close to judgements about the There is also a strategic dimension—how to argue desirability of an outcome. Making explicit the case of forestry to donors, governments and normative judgements about trade offs is a other policy makers. For instance, putting fourth option. Consider the specific trade off emphasis on the safety net functions of forests between poverty reduction and forest can be justified on several grounds: conservation. Most researchers would probably • Qualitative documentation (though not accept that a one dollar annual income rarely quantification) of the forest safety increase for one household is worth cutting 100 net role exists in most cases, as opposed ha of primary forest. But 100 dollars each year to the income generating function. One for each of 100 families just might be worth it, might risk over-estimating the poverty in the context of a poor developing country. Thus alleviation value of forests by putting one attempts to identify the ‘win-more—lose- too much emphasis on the latter. In most less’ (and the ‘lose-less—win-more’) situations. rural settings, sectors other than forestry This relates to the increasingly popular yet have a larger potential for stimulating controversial concept of ‘good deforestation’— pro-poor economic growth and a forest loss deemed acceptable because of the employment. high net benefits to people from its conversion • Forest degradation and deforestation to other uses. Its logical twin is even more threaten the forest based safety nets of controversial: ‘good impoverishment’—the act poor people without adequate of making people (just a little bit) worse off for alternatives or opportunities in place. the sake of safeguarding outstanding The most urgent issue in many places is environmental values. The way the winds are therefore to prevent further deprivation. blowing currently, that concept certainly does not Ultimately, this affects many times more have a taste of political correctness! people than those who will ever have the In either case, there definitely seems to be chance to use forest products as a a strong case for analysing the trade offs development pathway out of poverty. between the different objectives contained in • Poverty (and forest dependence) is not PAFSEP and making them apparent to the likely to go away. “The poor will always different stakeholders. Part of a useful myth be with us”—in spite of ambitious busting role is just to show in how few cases reduction targets such as the Millennium PAFSEP is an unambiguously defined goal. Development Goals (MDG). This is Conversely, an important influence for a research particularly the case in those regions of organisation should be to contribute to a realistic sub-Saharan Africa and South Asia where vision of these trade offs on behalf of forest economic growth is likely to be too small stakeholders, rather than advocating a to cut poverty by 50 % by the year 2015, unidirectional device. Exploring trade offs and as the MDG set forth. We need to be options to minimise them are key elements of realistic and give priority to these policy oriented research. underprivileged regions—at the end of the day, it is people’s survival that is at stake. ‘Safety nets’ or ‘engines of • The safety net functions are more likely growth’—poverty prevention or to represent win-win situations, at least reduction? for the poorest of the poor and the As argued earlier, forest products are critical as environment. Maintaining the status quo safety nets to many poor. Their role as a principal favours both the conservation of forests food or cash income earner is less significant and the livelihoods of the very poor who and the general potential of forest products to depend upon them. It is a ‘lose’ situation elevate living standards over time is for external developers but that is easier controversial. Given this picture, should future to justify in ethical terms. CIFOR research be on the poverty prevention/ safety net function or on monetary income Exploring the Forest—Poverty Link 57

Conversely, arguments for a relatively Target groups stronger emphasis on income generation Closely linked to the previous question is the include: issue of target groups. Should research focus • In the past, the potential of forests has on the very poor (the ‘poorest of the poor’) or been under-utilised. Due to the relatively on the moderately poor—what should be the recent changes in the six enabling relative emphasis? Should we exclusively be conditions mentioned in Section 4, the concerned with people at the forest margin, with world now looks different. New rural populations in general or also with the urban development thinking, policies, markets poor? and technology have created new The very poor normally lack the resources opportunities that set the stage for necessary to respond pro-actively to new substantial improvements in forest economic opportunities (see Section 3). Either based benefits for the poor. If we one needs extremely targeted measures to reach monitor, analyse and document these this group—or they are most likely to benefit processes, we can help to replicate the from general labour intensive growth, notably success stories. The potential is there— in agriculture and certain types of industries and we just need to exploit it. services. Forestry is probably not an adequate • Forest safety nets represent a development tool to reach the poorest of the temporary stage, losing their poor. However, if the main emphasis is on the importance with economic ‘defensive’ safety net function of forests, the development. When people replace primary target group would in this case, be the forests, they also replace forest safety poorest of the poor. Yet, if the main emphasis nets with other safety nets—but forest is on commercialisation, cash income generation researchers rarely take notice. To make and poverty reduction, one should realise that this transitory survival strategy (the these interventions are probably more likely in ‘employment of last resort’) assume the here and now, to benefit the moderately first priority on an international poor—and eventually also the better off people. research institute’s agenda would be Further, one needs to recognise that the like sending an army down a dead end situation and interests of ‘the moderately poor’ street. and ‘the poorest of the poor’ seldom move • The main emphasis in the international harmoniously in the same direction. Often they debate is on poverty reduction and have opposite interests and actions designed to promising the maintenance of the status favour one group may hurt the other. At the quo is too defensive a marketing strategy aforementioned FAO forum on poverty and for donors and policy makers. Poor people forests, community forestry in Nepal was want development, not just static safety mentioned as an example. Closure of forests and nets! To insist that they need to keep other restrictions on non-sustainable uses were forests only as a cushion will prepare the said to have hurt the very poor (e.g. landless ground for development strategies and charcoal makers or fuelwood gatherers), while interventions that become poverty traps. they had benefited the moderately poor Besides, safety net functions are already landowners. While advocates of community reasonably well known, offering limited forestry objected to that general characterisation, research potential. it cannot be denied that subgroups of the • Even if forest safety nets were truly ‘poorest of the poor’ often benefit from forest vital, it is hard to see what research degradation, posing an additional dilemma for really can contribute to enable mainly those who want to help both the forests and the subsistence uses. What are the points poor. of leverage? The only objective would Another dimension of targeting refers to be to mechanically document more and categories of socioeconomic groups of forest more cases and keep saying in public dependent people (cf. Table 1). How far from that forests should be conserved. The the forests should we go? Almost everybody would latter is probably better done by an include both of the first two groups (forest advocacy organisation than by an dwellers and farmers living adjacent to forests). international research institute. Many people are also inclined to target the third group, commercial users of forest products, although some of these ‘forest products’ do not 58 CIFOR Occasional Paper No. 40 Arild Angelsen and Sven Wunder

come from forests in a strict sense. However, a should remember that for the time being, CIFOR’s main question is whether poor urban consumers mandate also aims to look at national should be included too. The importance for the development, not only at rural areas or forest urban poor of fuelwood and sometimes charcoal, dwelling people. construction poles and low cost furniture, is an On the other hand, there are also arguments argument in favour. against the inclusion of the urban poor as a Observing the whole product cycle makes it target group. Research might lose focus, possible to grasp the entire economic value of an considering the large number of urban activity (i.e. not just the producer but also the consumers, the indirect and complex linkages— consumer surplus). For instance, CIFOR’s work and in most cases, there is probably only a very in Cameroon has been looking at urban marketing small weighting of forest products in the urban of forest products, revealing the importance of poor’s consumption baskets. Even when forest certain NTFPs to disadvantaged urban people. products are important to the urban poor, they Bushmeat consumption in Central and Western are not critical and we would normally expect Africa also constitutes an important urban the provision of cheap forest products to commodity, especially as a source of protein. We contribute little to urban poverty reduction. CIFOR Occasional Paper Series

1. Forestry Research within the Consultative Group on International 22. Una de Gato: Fate and Future of a Peruvian Forest Resource Agricultural Research Wil de Jong, Mary Melnyk, Luis Alfaro Lozano, Marina Rosales Jeffrey A. Sayer and Myriam García 2. Social and Economical Aspects of Miombo Woodland Management 23. Les Approches Participatives dans la Gestion des Ecosystemes in Southern Africa: Options and Opportunities for Research Forestiers d’Afrique Centrale: Revue des Initiatives Existantes Peter A. Dewees Jean-Claude Nguinguiri 3. Environment, development and poverty: A Report of the 24. Capacity for Forestry Research in Selected Countries of West and International Workshop on India’s Forest Management and Central Africa Ecological Revival Michael J. Spilsbury, Godwin S. Kowero and F. Tchala-Abina Uma Lele, Kinsuk Mitra and O.N. Kaul 25. L’Ímpact de la Crise Economique sur les Populations, les Migration 4. Science and International Nature Conservation et le Couvert Forestier du Sud-Cameroun Jeffrey A.Sayer Jacques Pokam Wadja Kemajou and William D.Sunderlin 5. Report on the Workshop on Barriers to the Application of Forestry 26. • The Impact of Sectoral Development on Natural Forest Research Results Conservation and Degradation: The Case of Timber and Tree Crop C.T.S. Nair, Thomas Enters and B. Payne Plantations in Indonesia • (Indonesian edition) Dampak Pembangunan Sektoral terhadap 6. Production and Standards for Chemical Non-Wood Forest Products Konversi dan Degradasi Hutan Alam: Kasus Pembangunan HTI dan in China Perkebunan di Indonesia Shen Zhaobang Hariadi Kartodihardjo and Agus Supriono 7. • Cattle, Broadleaf Forests and the Agricultural Modernization 27. L’Impact de la Crise Économique sur les Systèmes Agricoles et le Law of Honduras: The Case of Olancho Changement du Couvert Forestier dans la Zone Forestière Humide • (Spanish edition) Ganadería, bosques latifoliaods y Ley de du Cameroun Modernizatción en Honduras: El caso de Olancho Henriette Bikié, Ousseynou Ndoye and William D.Sunderlin William D.Sunderlin and Juan A.Rodriguez 28. • The Effect of Indonesia’s Economic Crisis on Small Farmers and 8. High quality printing stock - has research made a difference? Natural Forest Coverin the Outer Islands Francis S.P. Ng • (Indonesian Edition) Dampak Krisis Ekonomi Indonesia terhadap 9. • Rates Causes of Deforestation in Indonesia: Towards a Resolution Petani Kecil dan Tutupan Hutan Alam di Luar Jawa of the Ambiguities William D. Sunderlin, Ida Aju Pradnja Resosudarmo, Edy Rianto, • (Indonesian edition) Laju dan Penyebab Deforestasi di Indonesia: Arild Angelsen Penelaahan Kerancuan dan Penyelesaiannya William D. Sunderlin and Ida Aju Pradnja Resosudarmo 29. The Hesitant Boom: Indonesia’s Oil Palm Sub-Sector in an Era of Economic Crisis and Political Change 10. Report on Discussion Forum on Irformation Services in the Asia- Anne Casson Pacific and AGRIS/CARIS in the 21st Century and Asia-Pacific Regional Consultation 30. The Underlying Causes of Forest Decline Michael O. Ibach and Yvonne Byron Arnoldo Contreras-Hermosilla 11. Capacity for Forestry Research in the Southern African 31. ‘Wild logging’: The rise and fall of logging networks and Development Community biodiversity conservation projects on Sumatra’s rainforest frontier Godwin S. Kowero and Michael J. Spilsbury John F. McCarthy 12. Technologies for sustainable forest management: Challenges for 32. Situating Zimbabwe’s Natural Resource Governance Systems in the 21st century History Jeffrey A. Sayer, Jerome K. Vanclay and R. Neil Byron Alois Mandondo 13. Bosques secundarios como recurso para el desarrollo rural y la 33. Forestry, Poverty and Aid conservación ambiental en los trópicos de América Latina J.E. Michael Arnold Joyotee Smith, César Sabogal, Wil de Jong and David Kaimowitz 34. The Invisible Wand: Adaptive Co-management as an Emergent 14. Cameroon’s Logging Industry: Structure, Economic Importance Strategy in Complex Bio-economic systems. and Effects of Devaluation Jack Ruitenbeek and Cynthia Cartier Richard Eba’a Atyi 35. Modelling Methods for Policy Analysis in Miombo Woodlands 15. • Reduced-Impact Logging Guidelines for Lowland and Hill A. A Goal Programming Model for Planning Management of Miombo Dipterocarp Forests in Indonesia Woodlands • (Indonesian edition) Pedoman Pembalakan Berdampak Rendah I. Nhantumbo and Godwin S. Kowero untuk Hutan Dipterocarpa Lahan Rendah dan Bukit di Indonesia B. A System Dynamics Model for Management of Miombo Plinio Sist, Dennis P. Dykstra and Robert Fimbel Woodlands Ussif Rashid Sumaila, Arild Angelsen and Godwin S. Kowero 16. Site Management and Productivity in Tropical Forest Plantations A. Tiarks, E.K.S. Nambiar and Christian Cossalter 36. How to Know More about Forests? Supply and Use of Information for Forest Policy 17. Rational Exploitations: Economic Criteria and Indicators for K. Janz and R. Persson Sustainable Management of Tropical Forests Jack Ruitenbeek and Cynthia Cartier 37. Forest Carbon and Local Livelihoods: Assessment of Opportunities and Policy Recommendations 18. Tree Planting in Indonesia: Trends, Impacts and Directions Joyotee Smith and Sara J. Scherr Lesley Potter and Justin Lee 38. • Fires in Indonesia: Causes, Costs and Policy Implications 19. Le Marche des Produits Forestiers Non Ligneux de l’Afrique • (Indonesian edition) Kebakaran Hutan di Indonesia: Penyebab, Centrale en France et en Belgique: Produits, Acteurs, Circuits de Biaya dan Implikasi Kebijakan Distribution et Debouches Actuels Luca Tacconi Honoré Tabuna 39. Fuelwood Revisited: What Has Changed in the Last Decade? 20. Self-Governance and Forest Resources Michael Arnold, Gunnar Köhlin, Reidar Persson and Elinor Ostrom Gillian Shepherd 21. Promoting Forest Conservation through Ecotourism Income? A case study from the Ecuadorian Amazon region Sven Wunder 40 4/30/03 4:39 AM Page 1

Center for International Forestry Research Center for International Forestry Research

CIFOR Occasional Paper publishes the results of research that is particularly significant to tropical forestry. The content of each paper is peer reviewed internally and externally, and published simultaneously on the web in downloadable format (www.cifor.cgiar.org/publications/papers).

Contact publications at [email protected] to request a copy.

CIFOR Occasional Paper No. 40 Exploring the Forest—Poverty Link: Key Concepts, Issues and Research Implications

Arild Angelsen and Sven Wunder The Center for International Forestry Research (CIFOR) was established in 1993 as part of the Consultative Group on International Agricultural Research (CGIAR) in response to global concerns about the social, environmental and economic consequences of forest loss and degradation. CIFOR research produces knowledge and methods needed to improve the well-being of forest-dependent people and to help tropical countries manage their forests wisely for sustained benefits. This research is done in more than two dozen countries, in partnership with numerous partners. Since it was founded, CIFOR has also played a central role in influencing global and national forestry policies.

CIFOR is one of the 16 Future Harvest centres of the Consultative Group on International Agricultural Research (CGIAR)