SEARCH

SEARCH BY CITATION

Keywords:

  • Consumption;
  • deforestation;
  • forest policy;
  • global conservation;
  • poverty;
  • trade

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Analysis of production and trade data from 176 countries reveals that patterns of wood product consumption and harvest differ significantly across income groups. Poorer countries’ consumption consists primarily of domestic fuelwood, yet between 1972 and 2009, low-income countries harvested >171 million hectares of forest products for export. High-income countries were the only group to act as net importers, suggesting that rich countries practice preservation within borders but appropriate resources from poorer countries to sustain consumption. Harvests in poorer countries do occur at relatively low harvest efficiencies, implying that losses may be attenuated via technological improvement. However, efficiency does not mitigate the effects of high consumption. Despite exceptionally high efficiencies, high-income countries are still responsible for just as much (or more) consumption-driven forest loss as any other group. These findings highlight the importance of reducing consumption and suggest that neither technocentric solutions nor national-level conservation policies are sufficient means to preserve global forests.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Historical reconstructions suggest that anthropogenic influences have resulted in conversion of 19%-36% of global forests, with roughly 7.5 million more hectares completely deforested every year and far greater areas partially cleared (Williams 2008). Given the implications of forest conversion for both biodiversity preservation and carbon sequestration, much attention has focused on identifying the drivers of modern deforestation. There is general consensus that the primary proximate causes are conversion to agricultural land, extraction for commercial timber or domestic use (frequently as fuel), and conversion to built-up land and infrastructure (Geist & Lambin 2001; Sunderlin et al. 2005).

The ultimate drivers of deforestation are less well defined, although notable attempts have been made to categorize the indirect processes leading to forest loss. Geist & Lambin (2001), for example, establish five clusters of factors: economic, policy, technological, cultural or socio-political, and demographic. Among the indirect economic factors behind deforestation, poverty is perhaps one of the most widely implicated, showing up in 42% of reported cases of forest clearance, with all of these attributing at least some credit to poor people's lack of concern for the forest (Geist & Lambin 2001). This same hypothesis underpins economic arguments that champion economic growth (and accompanying technological advances) as the solution to environmental issues (e.g., Beckerman 1992).

Implicit in this “immiserization model” is the notion that poverty breeds environmental degradation (Rudel & Roper 1997), whereas wealth confers a degree of latitude; once a certain level of wealth is attained, people become free to choose amongst a broader range of goods and services and can therefore substitute away from extractive practices (e.g., fuelwood collection; shifting cultivation) and toward investment in the environment.

This narrative is challenged by participatory forestry projects that engage local stakeholders (including the poor) in conservation (e.g., Agarwal 2001), by recent studies (Mills & Waite 2009; Rudel et al. 2009; DeFries et al. 2010), and also by the fact that, in most cases, poverty drives deforestation to meet basic needs, not as a means of improving incomes or status (Geist & Lambin 2001; Sunderlin et al. 2008). Moreover, historical and spatial connections between receding forests and commercial activity suggest that it is not the poorest individuals who degrade forests. Rather, losses are driven by in-migrants, relative up-and-comers who seek higher incomes and have the resources and power to drive deforestation operations (Sunderlin et al. 2008).

A similar process occurs at the level of nation-states, where global trade allows wealthier countries to transfer extractive industry outside their borders. Wood production that may be seen as “surplus” within a country is actually part of the necessary exploitation of nature required to fuel the global-scale economy; trade facilitates absorption of this false surplus in directions that mirror historical patterns of exploitation (e.g., of labor; Gibson-Graham & Roelvink 2009). High-income countries can import to preserve their own forests, but do so at the expense of deforestation elsewhere, putting additional pressure on low-income countries while simultaneously displacing blame from end-consumers. Poverty does drive a portion of deforestation, particularly through the need for fuelwood. However, my examination of wood consumption across 176 countries reveals that this use has historically paled in comparison to consumption in high-income countries, much of which is supplied by wood from poor nations. These results further challenge the narrative that places poverty at the root of environmental degradation and demand a reassessment of forest policy.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Income data

The dataset includes every country for which necessary income, trade, and population data were available, resulting in 176 countries across eight regions (Supplementary Material). For each country, I constructed a time series spanning from 1972–2009 (some countries’ series begin later; for example, former Soviet Republics have no data before independence). I used logged real GDP per capita in constant dollars (chain series, international $2005; Heston et al. 2011) as the measure of income. To make comparisons across income groups, I categorized countries using the World Bank's year 2000 classifications: low-income (per capita GNI ≤$755), lower-middle income ($756 to $2995), upper-middle income ($2996 to $9265), and high-income (≥$9266) (World Bank 2002). GNI per capita (constant $2000) was obtained from the World Development Indicators database (World Bank 2012).

Incorporating consumption

Using data from ForesSTAT (FAO 2012), I calculated two measures of forest product consumption per country: CONSNF, which excludes wood used for fuel, and CONSF, which includes fuelwood. I assumed that:

  • display math

To calculate CONSNF, I considered production to be production of industrial roundwood and imports and exports to be the sum of import and export quantities for each of the following: industrial roundwood, paper and paperboard, sawnwood, wood pulp, and wood-based panels. For CONSF, I added fuelwood quantities to each term (all refer to ForesSTAT categories; data limitations discussed in Supplementary Material).

In addition to these consumption values, I considered the excess consumption (Supplementary Material), E, for each country (for either CONSNF or CONSF) where E is defined as a country's total consumption less what they produce in-country:

  • display math

A positive E thus indicates net importers of wood products, whereas negative E indicates net exporters. While different patterns may exist for trade in particular products, the excess consumption metric provides a good sense of the overall relationship between wood consumption and harvest in each country. That is, a country may harvest just enough wood for its own use, may harvest extra wood and export it for use beyond its borders, or may harvest too little wood to sustain its usage and thus import resources, contributing to deforestation in other parts of the world.

Relating consumption to forest loss

To see how trade patterns manifest in terms of actual forest hectarage, I determined a hectare-to-timber conversion factor, H, where,

  • display math

(data: Sohngen & Tennity 2004). As such, H is essentially a measure of efficiency of forest product extraction and relates wood product consumption to the forest area that would have been required to harvest that volume of timber. Due to data availability, this metric (and subsequent operations) could only be calculated for 133 countries (Supplementary Material).

Using this conversion factor, I calculated estimates of consumption-adjusted forest loss (or gain), Haadj, in country i in year t, where:

  • display math

The resulting values of Haadj represent the adjustment to a country's standing forest hectarage that would have been realized had countries been responsible for producing (within borders) all the forest products they consume, or, in the case of positive adjustments, had they not cut down their own trees for other countries’ use (thereby leaving more acreage intact). Additionally, to test a best-case scenario for efficiency, I calculated Haadj a second time, allowing all countries to benefit from the average efficiency of high-income countries. Note that calculated values are illustrative. Different practices in terms of cutting, replanting, etc. may result in more or less apparent deforestation than predicted. Nonetheless, the method provides an approximation of the land area involved in supplying consumption, as well as potential impacts of efficiency increases.

Linear mixed effects models

I modeled excess consumption and consumption-adjusted forest gain/loss separately for each measure of consumption (with/without fuelwood). For each independent variable, I built a model in R using the nlme package for mixed-effects models (Pinheiro et al. 2012). I impose a common cross-country structure for some covariates but allow others to vary flexibly (Supplementary Material). Specifically, the models allow each country's intercept, b0,i, to vary randomly, allowing for country-specific factors of ecology, economics, and policy. In addition, each country's slope, b1,i, is allowed to vary randomly as a function of time. This accounts for temporal autocorrelation and has the added bonus of accounting for the fact that the 1972–2009 time window captures countries at different points of development. That is, there are a priori reasons to expect that countries exhibit different patterns over time, and the random slope here appropriately models that variation.

For excess consumption without fuelwood, E_CONSNF, this results in the model:

  • display math

where income group is represented with a suite of fixed-effects (IG variables); covariates for total country population (POP) and land area (Land) account for differences in size among countries (data: Heston et al. 2011; FAO 2012), and spatial autocorrelation is addressed via spatial filtering variables (S1 and S2) (Supplementary Material). Finally, dummy variables (Y) are included for each year to capture time effects that manifest across countries. Additional variables were tested and found insignificant (Supplementary Material).

Each model was assessed for goodness-of-fit using lmmfit in R (Maj 2011). Finally, I used the multcomp package (Hothorn et al. 2008) to generate Tukey's all-pairwise comparisons on the estimated means of each income group and detect significant differences amongst income levels.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Consumption and forest loss

When fuelwood is excluded, per capita wood product consumption increases with increasing per capita GDP (Fig. 1). Including fuelwood muddles this pattern (Fig. 2). Upon adding fuelwood to estimates of forest consumption, the average low-income country's total wood consumption increases 1556% (Fig. 3). This percentage is likely conservative, as fuelwood is often harvested as part of the informal economy, particularly amongst the poorest of the poor (many of whom harvest “illegally” on what were historically community lands; Robbins et al. 2006, 2009).

image

Figure 1. Per capita consumption of wood products (excluding fuelwood), CONSNF, as a function of per capita GDP. Each of 176 countries’ time series (∼1972–2009) is denoted by a unique color and symbol (N = 6,063). Consumption tallies do not account for trade in some finished wood products (e.g., furniture); see discussion in Supplementary Material.

Download figure to PowerPoint

image

Figure 2. Per capita consumption of wood products (including fuelwood), CONSF, as a function of per capita GDP. Each of 176 countries’ time series (∼1972–2009) is denoted by a unique color and symbol (N = 6,063). Consumption tallies do not account for trade in some finished wood products (e.g., furniture); see discussion in Supplementary Material.

Download figure to PowerPoint

image

Figure 3. Percent increase in total per capita wood consumption due to addition of fuelwood. Bars represent group means with error bars indicating one standard error. Points represent group medians.

Download figure to PowerPoint

The high-income group is the only group whose members (on average) act as net importers. Excess consumption is higher in this group than any other, and high-income countries have significantly higher excess consumption than either low or lower-middle income groups (Table 1). These results hold true for both consumption counts: with and without fuel.

Table 1. Income-dependent differences in excess wood product consumption, both including and excluding fuelwood. Observed means are displayed for each group, along with regression coefficients and significance as compared to high-income countries
Excess consumption (million m3)
   LowLower-MidUpper-MidHigh
  Income Group (n)(2,305)(1,469)(1,120)(1,169)
  1. **Significance at P < 0.01.

Excluding fuel Observed group mean−0.063−0.462−0.3980.528
 Wald's r2 = 0.869 Regression coefficient (high income as reference group) −1.931−0.867−0.451N/A
  Significance relative to high-income group<0.001**0.009**0.148N/A
Including fuel Observed group mean−0.062−0.472−0.4140.551
 Wald's r2 = 0.870 Regression coefficient (high income as reference group)−1.922−0.862−0.452N/A
  Significance relative to high-income group<0.001** 0.009**0.146N/A

“Consumption-adjusted forest gains/losses” relate these trends back to land cover, modeling the counterfactual of how much additional forest would have been left standing (or cut) in a given year were countries responsible for producing domestically all (and only) the forest products they consume. When these adjustments are accounted for, and each country's adjustment is calculated based on its own harvest efficiency (H), low-income countries gain significantly more forest hectarage than any other group. Differences among the other groups are non-significant (Table 2). All else being equal, had poor countries harvested only for their own needs, each low-income country would, on average, have avoided harvesting 88,290 hectares of standing forest per year. Given that there were 1,944 country-year combinations that fell into the low-income group, this amounts to a total of over 171 million hectares over the course of the study.

Table 2. Income-dependent differences in the cross-boundary effects of wood product consumption on raw hectarage of forests, both including and excluding fuelwood. Positive values indicate consumption-adjusted gains in forest hectarage; negative values indicate losses. Observed means are displayed for each group, along with regression coefficients and significance as compared to high-income countries
Raw hectarage gained/lost (thousands of hectares)
    LowLower-MidUpper-MidHigh
   Income Group (n)(1,944)(1,125)(784)(852)
  1. **Significance at P < 0.01; *significance at P <0.05.

Country-specific Excluding fuelObserved group mean1.485−66.828−25.543−45.861
 efficiency (H) value  Wald's r2 = 0.774Regression coefficient (high income as reference group)88.290−2.341−14.678N/A
   Significance relative to high-income group<0.001**0.9980.606N/A
  Including fuelObserved group mean1.471−66.741−25.514−46.450
   Wald's r2 = 0.774Regression coefficient (high income as reference group)88.335−2.185−14.248N/A
   Significance relative to high-income group<0.001**0.9990.629N/A
Average high-income Excluding fuelObserved group mean0.2253.7373.801−3.948
  efficiency for all countries  Wald's r2 = 0.870Regression coefficient (high income as reference group)15.6426.2443.786N/A
   Significance relative to high-income group<0.001**0.039*0.207N/A
  Including fuelObserved group mean0.2233.8193.940−4.144
   Wald's r2 = 0.871Regression coefficient (high income as reference group)15.5876.2303.824N/A
   Significance relative to high-income group<0.001**0.040*0.201N/A

Harvest efficiency

The data show that rich countries do exhibit exceptionally high levels of technological efficiency, harvesting dramatically more wood per hectare than their poorer counterparts (Fig. 4). The Czech Republic, which falls within the upper-middle income group for most years, tops the efficiency list at H = 482m3/ha, followed closely by high-income countries: Belgium and Switzerland. The low-income country of Eritrea exhibits the lowest efficiency: H = 0.060 m3/ha. Comparing efficiencies across groups, high-income countries exhibit the highest average efficiency levels (164.55m3/ha), giving this group almost twice the average harvest per unit area than seen for the upper-middle income group (83.70m3/ha). The lower-middle and low-income groups are even further behind. These differences in efficiency are significant for all income- group pairs (Fig. 4).

image

Figure 4. Mean harvest efficiency (H) by income group. Error bars indicate one standard error around the mean. Differences between groups are significant for all pairs, with all P-values < 0.001.

Download figure to PowerPoint

Nevertheless, even when all countries were credited with the efficiency rate of the average high-income country, wealthy countries were still responsible for more consumption-driven forest loss than any other group. Low and lower-middle income groups each exhibited significant gains in hectarage relative to high-income countries (Table 2).

Discussion and policy conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Expansion of consumption has often been noted as a serious consequence of affluence (Myers 2000; Myers & Kent 2003; Dietz et al. 2007), and others have suggested that what wealth really bestows is the ability to externalize damages incurred by consumption, thus creating apparent improvements in environmental indicators (e.g., Rothman 1998). The consumption issue has received growing attention in case studies and discussions of forest policy (e.g., Is it better to reduce demand, substitute non-wood products, or import from other countries? Dekker-Robertson & Libby 1998; Berlik et al. 2002; Mayer et al. 2005; Meyfroidt & Lambin 2009), and recent studies report that distant (international or urban) markets dominated by wealthy consumers increasingly drive conversion to agricultural land (Rudel et al. 2009; DeFries et al. 2010). These findings contradict studies that tout domestic consumption as the primary driver of deforestation (Sierra 2001; Wunder 2005) and put the impetus on developing countries to control deforestation via “virtuous” domestic policies (Motel et al. 2009). However, with notable recent exceptions (Meyfroidt et al. 2010; Kastner et al. 2011), few have considered the impact of trade in wood products using broad-scale empirical data. Those who have advance the conclusion that forest recovery in wealthy countries is frequently made possible only via imports. Only Meyfroidt et al. (2010) explicitly examined the role of income; they found GDP to be insignificant (but included only one high-income country).

Here, empirical analysis across a broad economic spectrum of countries reveals an important role for excess consumption. High-income countries are the only nations that consume, on average, more wood products than they produce, relying on imports (often from poorer nations) to fuel their consumption. This illustrates a prime example of weak (economically-based, rather than ecologically-based) sustainability. High-income countries that substitute external forest resources for internal ones create the illusion of conservation within their borders while simultaneously contributing to the drawdown of natural capital worldwide.

Poverty presently appears to drive consumption largely through the use of domestically harvested fuelwood. However, given current trends, there is little reason to expect that today's poorer countries will choose to forego increased consumption as they become more affluent. Moreover, if today's wealthy countries support wealth by consuming across borders, then that pattern will be little more than a historical artifact. Some middle-income countries may recover domestic forests by importing timber (e.g., Vietnam; Meyfroidt & Lambin 2009), but the same pathway cannot be available for all, especially if poor countries already bear the brunt of their own consumption as well as that of richer nations.

Technological efficiency as panacea?

Many propose eco-efficiency and eco-technologies as the key, both to enabling sustainable growth amongst the rich, and to allowing poverty-stricken nations to achieve development without significantly raising overall consumption (Pulliam & O'Malley 1996; Myers 2000). I tested one aspect of this notion by including in consumption calculations a measure of each country's harvest efficiency (H). Despite exceptionally high efficiencies, high-income countries are still responsible for just as much (or more) consumption-driven forest loss as any other group. Note too, that rich countries receive credit for high efficiencies but import vast amounts of wood products harvested in countries operating at lower efficiencies. Thus, while end consumers may have the ability to extract wood in an efficient manner, these advantages are not utilized to maximum effect when consumption relies instead on imports. My results are therefore conservative. Even in the “best-case” scenario, where all countries harvest at high efficiency, wealthy nations are still responsible for more forest loss than any other income group, suggesting that the patterns demonstrated here are likely to persist, even as poorer nations move towards greater efficiency.

These results point toward a somewhat counterintuitive policy conclusion. Barring rapid improvements in efficiency for poorer countries, the principles of competitive advantage, as well as the overarching goal of global forest conservation, suggest that wealthy countries with greater harvest efficiency should perhaps cut down their own forests to supply world consumption. This proposal does not preclude the need to reduce consumption (Berlik et al. 2002) (nor unambiguously support forestry in rich countries), but would seem to have doubly beneficial conservation impacts. First, fewer acres of forest would be harvested to meet demand. Second, sparing the forests of the poor to large extent corresponds with sparing tropical forests, and tropical forests contain high carbon densities and the largest portion of the world's biodiversity (Wilson 1998; Chapin et al. 2008; Lambin & Meyfroidt 2011). This strategy could therefore prove more efficient at meeting the goals of global forest conservation than have traditional (national-level) approaches to within-borders forest preservation and management. In the short-run, the economic impacts for poor countries may be of concern, but curbing exports of natural resources from low-income countries leaves more natural capital on which to build (strongly) sustainable futures.

Moreover, the likelihood of realizing rapid efficiency increases in low-income countries is unknown, and depends largely on the underlying source of current differences in harvest efficiency. While differences may reflect a need for technology transfers, they may also belie natural variation in land quality (what Geist & Lambin (2001) call “pre-disposing environmental factors”), or the fact that some wealthy countries (e.g., Sweden) have achieved efficiency increases by replacing natural forests with plantations and non-native species (Williams 2008). Efficiency increases may also carry environmental and economic disincentives, as mechanization frequently displaces laborers to other extractive practices (Lambin & Meyfroidt 2011).

Re-envisioning forest policy

National forest policies fall short, in part, because current trade regimes enable countries to export environmental damage to places outside their borders, and to export with it any awareness of, or responsibility for, the damages incurred (Dekker-Robertson & Libby 1998; Muradian & Martinez-Alier 2001; Berlik et al. 2002; Mayer et al. 2005; Ståhls et al. 2010). Economic measures imposed at the national level (e.g., Pigouvian taxes) can assist in addressing externalities, but fail to achieve increased cognizance of ecological limitations or address the distributive justice problem that arises when resources are channeled from poor nations to rich ones. This disconnect between production and consumption in turn creates a disjuncture between problem and policy. Within-country forest management is contrary to global conservation goals (even when successful at preserving within-borders forest) if it comes at the expense of forests in other parts of the world. Likewise, payment for ecosystem services schemes, such as those implicit in REDD+, may miss the point if they assume that lack of alternative income is the only thing driving deforestation in poorer nations. At the core, direct and indirect drivers of deforestation are all connected to consumption. National policies are not sufficient to address this aspect of forest loss in our globalized age, nor can we expect the same policy prescriptions to apply across the income spectrum.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

I thank E.A. Marschall, D. Martin, D. Smith, M. Busa, A. Chhatre, K. Greenwald, B. Mansfield, A. McKinney, G. White, and four anonymous reviewers for helpful comments on the manuscript, and N. Horton for statistical guidance. This work was supported through the Ohio State University Distinguished University Fellowship and the Mellon Postdoctoral Fellowship in Environmental Science & Policy at Smith College.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Agarwal, B. (2001) Participatory exclusions, community forestry, and gender: an analysis for South Asia and a conceptual framework. World Dev., 29(10), 16231648.
  • Beckerman, W. (1992) Economic growth and the environment: Whose growth? Whose environment? World Dev., 20, 481496.
  • Berlik, M.M., Kittredge, D.B. & Foster, D.R. (2002) The illusion of preservation: a global environmental argument for the local production of natural resources. J. Biogeogr., 29, 15571568.
  • Chapin, F.S., Randerson, J.T., McGuire, A.D., Foley, J.A. & Field, C.B. (2008) Changing feedbacks in the climate-biosphere system. Front. Ecol. Environ., 6, 313320.
  • DeFries, R.S., Rudel, T., Uriarte, M. & Hansen, M. (2010) Deforestation driven by urban population growth and agricultural trade in the twenty-first century. Nat. Geosci., 3, 178181.
  • Dekker-Robertson, D.L. & Libby, W.J. (1998) American forest policy—global ethical tradeoffs. BioScience, 48(6), 471477.
  • Dietz, T., Rosa, E.A. & York, R. (2007) Driving the human ecological footprint. Front. Ecol. Environ., 5, 1318.
  • FAO (2012). FAOSTAT. Available from http://faostat.fao.org. Accessed 10 January, 2012.
  • Geist, H.J. & Lambin, E.F. (2001) What drives tropical deforestation? A meta-analysis of proximate and underlying causes of deforestation based on subnational case study evidence. LUCC Report Series no. 4. LUCC International Project Office, Louvain-la-Neuve, Belgium.
  • Gibson-Graham, J.K. & Roelvink, G. (2009) An economic ethics for the Anthropocene. Antipode, 41, 320346.
  • Heston, A., Summers, R. & Aten, B. (2011) Penn World Table Version 7.0. Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania.
  • Hothorn, T., Bretz, F. & Westfall, P. (2008) Simultaneous inference in general parametric models. Biometrical J., 50(3), 346363.
  • Kastner, T., Erb, K.-H. & Nonhebel, S. (2011) International wood trade and forest change: A global analysis. Global Environ. Change, 21, 947956.
  • Lambin, E.F. & Meyfroidt, P. (2011) Global land use change, economic globalization, and the looming land scarcity. Proc. Natl. Acad. Sci. USA, 108, 34653472.
  • Maj, Aleksandra. (2011) lmmfit: goodness-of-fit measures for linear mixed models with one-level-grouping. R package version 1.0.
  • Mayer, A.L., Kauppi, P.E., Angelstam, P.K., Zhang, Y. & Tikka, P.M. (2005) Importing timber, exporting ecological impact. Science, 308, 359360.
  • Meyfroidt, P. & Lambin, E.F. (2009) Forest transition in Vietnam and displacement of deforestation abroad. Proc. Natl. Acad. Sci. USA, 106, 1613916144.
  • Meyfroidt, P., Rudel, T.K. & Lambin, E.F. (2010) Forest transitions, trade, and the global displacement of land use. Proc. Natl. Acad. Sci. USA, 107, 2091720922.
  • Mills, J.H. & Waite, T.A. (2009) Economic prosperity, biodiversity conservation, and the environmental Kuznets curve. Ecol. Econ., 68, 20872095.
  • Motel, P.C., Pirard, R. & Combes, J.-L. (2009) A methodology to estimate impacts of domestic policies on deforestation: Compensated successful efforts for “avoided deforestation” (REDD). Ecol. Econ., 68, 680691.
  • Muradian, R. & Martinez-Alier, J. (2001) Trade and the environment: from a ‘Southern’ perspective. Ecol. Econ., 36, 281297.
  • Myers, N. (2000) Sustainable consumption. Science, 287, 2419.
  • Myers, N. & Kent, J. (2003) New consumers: the influence of affluence on the environment. Proc. Natl. Acad. Sci. USA, 100, 49634938.
  • Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & the R Development Core Team (2012) nlme: Linear and Nonlinear Mixed Effects Models. R package version 3. 1103.
  • Pulliam, R.H. & O'Malley, R. (1996) Economic policies should help achieve economic goals. Ecol. Appl., 6, 2122.
  • Robbins, P., McSweeney, K., Waite, T. & Rice, J. (2006) Even conservation rules are made to be broken: implications for biodiversity. Environ. Manag., 37, 162169.
  • Robbins, P., McSweeney, K., Chhangani, A.K. & Rice, J.L. (2009) Conservation as it is: Illicit resource use in a wildlife reserve in India. Hum. Ecol., 37, 559575.
  • Rothman, D.S. (1998) Environmental Kuznets curves—real progress or passing the buck? A case for consumption-based approaches. Ecol. Econ., 25, 177194.
  • Rudel, T., DeFries, R., Asner, G.P. & Laurance, W.F. (2009) Changing drivers of deforestation and new opportunities for conservation. Conserv. Biol., 23(6), 13961405.
  • Rudel, T. & Roper, J. (1997) The paths to rain forest destruction: cross national patterns of tropical deforestation, 1975–1990. World Dev., 25(1), 5365.
  • Sierra, R. (2001) The role of domestic timber markets in tropical deforestation and forest degradation in Ecuador: implications for conservation planning and policy. Ecol. Econ., 36, 327340.
  • Sohngen, B. & Tennity, C. (2004) Country Specific Global Forest Data Set V.1. Mimeo. Dept. Agr., Env., Dev. Econ., Ohio State University.
  • Ståhls, M.H., Mayer, A.L., Tikka, P.M. & Kauppi, P.E. (2010) Disparate geography of consumption, production, and environmental impacts: forest products in Finland 1991–2007. J. Ind. Ecol., 14(4), 576585.
  • Sunderlin, W.D., Angelsen, A., Belcher, B., Nasi, R., Santoso, L. & Wunder, S. (2005) Livelihoods, forests, and conservation in developing countries: an overview. World Dev., 33(9), 13831402.
  • Sunderlin, W.D., Dewi, S., Puntodewo, A., Müller, D., Angelsen, A. & Epprecht, M. (2008) Why forests are important for global poverty alleviation: a spatial explanation. Ecol. Soc., 13(2), 24.
  • Williams, M. (2008) A new look at global forest histories of land clearing. Annu. Rev. Environ. Resour., 33, 345367.
  • Wilson, E.O. (1998) Biodiversity. National Academy Press, Washington, DC.
  • The World Bank (2002) World development indicators 2002. The International Bank for Reconstruction and Development/The World Bank.
  • The World Bank (2012) World databank: World development indicators (WDI) and global development finance (GDF). Available from http://databank.worldbank.org/ddp/home.do. Accessed 20 January, 2012.
  • Wunder, S. (2005) Macroeconomic change, competitiveness and timber production: a five-country comparison. World Dev., 13(1), 6586.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and policy conclusions
  7. Acknowledgments
  8. References
  9. Supporting Information

Disclaimer: Supplementary materials have been peer-reviewed but not copyedited.

FilenameFormatSizeDescription
conl304-sup-0001-Suppmat.pdf295K

Table S1.

Table S2.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.