60073 POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise MARCH 2011 · Number 51 NE 010 · Numbe 18 The Poverty Impacts of Climate Change Emmanuel Skoufias, Mariano Rabassa, Sergio Olivieri, and Milan Brahmbhatt Over the last century, the world has seen a sustained decline in the proportion of people living in poverty. However, there is an increasing concern that climate change could slow or possibly even reverse poverty reduction progress. Given the complexi- ties involved in analyzing climate change impacts on poverty, different approaches can be helpful; this note surveys the results of recent research on climate change impacts on poverty. The continued decline in global poverty during the last 100 Climate change is a long-term problem that has been unfold- years is a remarkable achievement, a decline that has been even ing over many decades. It is likely to lead not only to changes in more significant in the last three decades. In 1981, the percent- the mean levels of temperatures and rainfall, but also to a signifi- age of the world population living below $1.25 a day was 52 cant increase in the variability of climate and in the frequency of percent. By 2005, that rate had more than halved to 25 percent extreme events. Given the complexities involved in an analysis of (Chen and Ravallion 2009). In recent years, poverty reduction the climate change impacts on poverty, different approaches may has continued in most countries, even after the financial, food, be helpful in considering these impacts. One way is to use econo- and fuel shocks of 2008­9. This trend is expected to continue, mywide growth models incorporating climate change impacts to especially if developing countries manage to sustain the rapid work out consistent scenarios for how climate change might af- per capita income growth rates they achieved over the last de- fect the path of poverty over the next decades. Another approach cade. If developing countries do maintain their income growth is to learn about sector-specific channels through which longer- rates, poverty headcounts at the $1 or $2 per day could turn term climate change affects poverty, the size of such impacts, the out to be almost obsolete as measures of well-being over the potential heterogeneity of impacts, and the types of policies that next 50 to 100 years. However, there is a growing concern that may alleviate the adverse impacts. The information generated by climate change could slow or possibly even reverse progress on this approach is useful in tackling poverty today, as well as in pre- poverty reduction. This concern is rooted in the fact that most paring for how to adapt to climate change in the future. Yet an- developing countries are more dependent on agriculture and other approach is to explore how current climate variability af- other climate-sensitive natural resources for income and well- fects poverty, and then examine the impacts of increased being, and that they also lack sufficient financial and technical variability on future poverty. This note considers some recent capacities to manage increasing climate risk. findings from these various approaches. 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Climate Change and Global Poverty from growth-poverty elasticities discussed in box 1, this contributes an Aggregate Perspective: Lessons from to more than halving the world poverty rate at the $2 a day level an Integrated Assessment Model to 14.1 percent by 2055. Under the RICE model's BAU scenar- io with climate damage, world gross domestic product (GDP) Three scenarios to model the long-term climate change impacts in 2055 would be 1.5 percent lower than in the baseline.2 In the on poverty are developed using an integrated assessment mod- BAU scenario, the estimated number of poor in 2055 would be el, the Regional Integrated Model of Climate and the Economy modestly higher by 10 million, compared to the no climate (RICE) developed by Nordhaus (2010). The first scenario change scenario, with most of the additional poor located in (baseline) simulates a world without climate change. The sec- Africa and South Asia. ond (business as usual [BAU]) reflects the impact of current It is worth stressing that this analysis focuses on the expect- trends in economic growth and greenhouse gas emissions ed or mean value of the probability distribution of damage (GHGs) on the climate, and the impact of climate change on from climate change. Obviously, looking at lower probability the overall economy without any emission abatement policies. extreme outcomes increases the estimates for GDP losses and The third (optimal abatement) is based on Nordhaus's calcula- poverty. Applying the same general methodology, Anderson tion of an emission abatement path, with full participation by (2006) reports similar results for Africa and South Asia using all countries, that maximizes global intertemporal economic the mean of the predicted damage distribution. The Stern Re- welfare. Olivieri, Rabassa, and Skoufias (2010) translate the im- view (2007) reports Anderson's results using the more extreme plications of these different growth scenarios for poverty using 95th percentile of damage distribution.3 historical estimates of growth-poverty elasticities. Under the optimal abatement scenario, the extra number of Given the limitations in knowledge and large uncertainties people in poverty due to global warming in 2055 is projected to surrounding climate change and its impact on economic be only slightly smaller: 9 million. That is because the effects of growth, and the impacts of growth on poverty, this analysis abating global emissions of greenhouse gases on aggregate eco- should be viewed as indicative only of the potential conse- nomic damages necessarily accrue more to higher-income quences of climate change on global poverty (see box 1). countries. Unlike adaptation, emissions mitigation does not Table 1 summarizes the main impacts of climate change on specifically target the poor. The major gains in poverty averted global poverty under these three scenarios. In a no climate by following the optimal abatement strategy would indeed oc- change baseline, the model projects an annual, global, real per cur on a longer time horizon, by 2100 and beyond. capita output growth rate of 2.2 percent up to 2055.1 Using the Heterogeneity of Climate Change Impacts on Poverty Climate change may impact household welfare through a vari- Box 1. Sources of Uncertainty in ety of channels. For example, climate change may have a nega- Poverty Estimates tive effect on agricultural productivity, particularly in tropical A major advantage of using an integrated assessment model regions, and also affect poor people's livelihood through its ef- like the RICE Model is that climate change questions can fects on health, access to water and natural resources, homes, be evaluated within a consistent framework. However, results should be viewed with due caution given the many and infrastructure. Considering the complexities involved in uncertainties and limitations facing all analyses of future modeling some of these channels, the emerging literature, in- climate change. First, the interactions among economic cluding this note, has largely focused on the impacts on agricul- growth, emissions of GHGs, the climate system, economic tural productivity. This section reviews the quantitative esti- damages, and policies are exceedingly complex and poorly mates of climate change impacts on poverty through its effects understood. Second, there are profound uncertainties at every stage in global warming modeling: uncertainties about on agricultural productivity. future output growth; the pace and direction of technological Recent research generally agrees on the significant overall change, particularly for low carbon energy sources; negative impact of global warming on agricultural productivity migration patterns; climatic reaction to rising concentrations and household welfare, but research also finds considerable of GHGs; and about the economic and ecological responses heterogeneity in these impacts. These differences are linked to to changing climate and how impacts should be discounted. It is also hard to predict the impact of future growth on geographical location and specific household characteristics, poverty reduction over such a long time span. Olivieri, including whether households are net producers or consumers Rabassa, and Skoufias (2010) calculate regional growth- of agricultural products; income sources; types of assets owned; poverty elasticities using country-specific growth-poverty ability to adapt; and ability to access credit or safety nets. elasticity estimates from historical data and assume these elasticities will stay constant over time. They also assume Location a constant within-country distribution of per capita income Geographical location obviously plays a large part in determin- over time. These two assumptions are quite restrictive, ing exposure to climatic phenomena such as rainfall variation especially when used to project poverty in the distant future. and droughts. Location is also a key factor in access to infra- 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Table 1. Climate Change Impacts on World Poverty (at the $2 a day level) Number of poor (millions) Headcount poverty rate (%) Scenarios 2005 2055 Change 2005 2055 Change Baseline 2,069.4 1,259.1 (810.3) 32.3 14.1 (18.2) BAU 2,069.4 1,269.2 (800.2) 32.3 14.2 (18.1) Difference from baseline .. 10.1 10.1 .. 0.12 0.12 Optimal abatement 2,069.4 1,268.5 (800.9) 32.3 14.2 (18.1) Difference from BAU .. (0.7) (0.7) .. (0.01) (0.01) Source: Olivieri, Rabassa, and Skoufias (2010). structure and markets, as well as in the costs of necessary house- hold characteristics; land-owning households face a smaller re- hold goods. duction in consumption relative to landless households, about The climate change impacts on agriculture can be heteroge- 5 percent, because the beneficial impact of higher prices for net neous even within a single country. In a recent study of Brazil, agricultural producers offsets the decline in land productivity. Assunção and Chein (2009) estimate that, on average, agricul- Focusing on the impact of climate change on the Indian ru- tural output per hectare could decrease by 18 percent by 2040 ral population as a whole, rural poverty could increase between as a result of climate change, but that at the municipality level, 1 and 6 percentage points by 2040, compared to the counter- impacts could range from a decrease of 40 percent to an in- factual of zero warming (Jacoby, Rabassa, and Skoufias 2011). crease of 15 percent. They predict that the poverty rate of rural Further, climate change impacts tend to be regressive, falling areas in Brazil will increase by 3.2 percentage points overall, more heavily on the poor than the rich. This result can be de- though, again, there is significant geographical variation, with composed into three parts: the effect on the returns to land and already poor regions being more affected. Jacoby, Rabassa, and labor, and the effect on cereal prices. The incidence curves de- Skoufias (2011) use household level data to show that there is picted in figure 1 show that productivity losses will negatively also significant heterogeneity in the impact of climate on per affect returns to land across the entire income distribution, but capita consumption across rural districts of India. They esti- that the rich lose proportionately more than the poor because mate that increases in mean surface temperature by 2040 they hold the lion's share of land (black line). Additionally, pro- could lead to consumption impacts ranging from no change in ductivity falls will also translate into wage reductions, but this some locations to an 11 percent decrease in others. effect is more distributionally neutral (blue line, which aggre- Structure of assets, incomes, and expenditures gates the impact of returns to land and labor). It is only after accounting for the welfare effects of rising cereal prices that the Climate change impacts on the welfare of rural households combined climate change affects the poor more than the rich depend on a number of interrelated factors, including the (gray line, which accounts for the total impact). household's assets, its potential income sources (diversifica- The discussion so far has looked at climate change impacts tion), and the structure of its expenditures. For instance, cli- on rural households. Ahmed, Diffenbaugh, and Hertel (2009) mate change might reduce physical productivity on an agri- and Hertel, Burke, and Lobell (2010) provide a broader per- cultural household's cereal land. But a general decline in spective by looking at economywide impacts using a general agricultural productivity will also raise food prices, benefit- equilibrium model of global production, trade, and income dis- ting the same household as long as it is a net producer of cere- tribution. A key finding in these studies is that the most signifi- als. Also, the extent to which a decline in agricultural produc- cant climate change impacts on poverty are likely to occur tivity translates into lower rural wages depends on the among urban wage laborers, who are the most negatively af- diversification of the local economy and the ability of labor to fected by food price increases. Agricultural, self-employed move into other occupations. households in rural areas are less affected because they benefit In the case of rural India, Jacoby, Rabassa, and Skoufias from higher prices. (2011) estimate that increased mean surface temperature could lead to a 13 percent reduction in agricultural productivity by Adaptation and risk management 2040 (after taking into account adaptation by farmers in terms The extent to which climate change impacts poverty also of changes to crop mix, production methods, and so forth). depends on the extent of adaptation by households in response This substantial decrease in output translates into a more mod- to emerging circumstances. Jacoby, Rabassa, and Skoufias est 6 percent decline in average per capita consumption for ru- (2011) calculate the welfare benefits from autonomous adapta- ral households because households derive more income from tion in agriculture in India. In this context, autonomous adap- their labor endowment. This result varies according to house- tation can be defined as market-based responses to climate 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Figure 1. Climate Change Incidence Curve crease in the amount of rainfall in the 90-day postmonsoon period is associated with a 14 -2 percent reduction in the per capita expendi- percent change in per capita expenditure tures of rice farmers. Rice farm households do manage to protect their food expenditures, -4 however, but typically by reducing nonfood ex- penditures. This result is consistent with earlier -6 findings on how households cope with shocks by cutting spending on things such as health and education, which in turn has an adverse -8 long-term effect on poverty by reducing human productivity falls translate into wage reductions capital investment in children. -10 returns to land across income distribution There is a considerable literature suggesting welfare effects of rising cereal prices that, in the absence of adequate credit and in- -12 surance markets and other ex ante means for 0 20 40 60 80 100 mitigating risks, ex post risk mitigation strate- per capita expenditure percentiles gies by the poor can result in lower long-term Source: Authors' illustration. income growth and increased poverty. Skou- fias, Essama-Nssah, and Katayama (2011) con- change by individuals, households or firms, typically by adjust- firm that the availability of credit, social protection, and com- ments over time in their production and consumption pat- munity-based programs are among the strongest factors terns. Jacoby, Rabassa, and Skoufias (2011) show that adapta- 4 mitigating the impacts of extreme weather events on the well- tion (that is, changes in cropping patterns, input use, and being of rural households. technology) reduces the average long-term loss in per capita Conclusions and Some Policy Implications consumption from climate change by about half (the decline in consumption is 11 percent in the case of a weather shock, com- Climate change will slow the pace of global poverty reduction, pared to 6 percent when autonomous adaptation is factored in). but--based on the mean or expected value of climate damages Migration, the most extreme adaptation measure, can also used in mainstream analyses such as Nordhaus's (2010) RICE help reduce the potential longer-term welfare impacts of cli- model or the Stern Review (2007)--the expected poverty im- mate change. For Brazil, Assunção and Chein (2009) show that pact will be relatively modest and far from reversing the major allowing for labor mobility across sectors or across municipali- decline in poverty that is expected to occur over the next 40 ties reduces the climate-based increase in the rural poverty rate years as a result of continued economic growth. However, a from 3.2 to 2 percentage points. couple of qualifications are in order: first, much of the poverty These studies of adaptation show that the ability of house- impact is expected to be concentrated in Africa and South Asia, holds to adapt to climate change over the longer term is vital, both of which would see more substantial increases in poverty and that this ability can be strengthened by disseminating in- relative to a baseline without climate change. Second, the oc- formation about longer-term risks and anticipatory invest- currence of less probable but more extreme climate damage ments. However, longer-term impact reduction through adap- scenarios would naturally result in larger poverty increases. tation would not necessarily diminish substantial adjustments Third, aggregate projected damages are relatively low over the costs. time horizon analyzed in this note (mid-century). As climate Climate change is likely to lead not only to changes in the change continues to unfold during this and the next century, mean levels of temperatures and rainfall, but also to a significant aggregate damages could be substantial and have a larger effect increase in the variability of climate and in the frequency of ex- on poverty. treme weather-related shocks. The impact of an increase in cli- Recent empirical studies confirm that changes in climatic matic variability on household welfare will depend in part on means and variability can have substantial impacts on agricul- the ex ante and ex post risk management strategies employed by tural output, household welfare and poverty, but that there is households.5 One way to understand the impacts of these risk considerable heterogeneity in outcomes based on geographical management strategies is to examine how they are used to ad- location, the assets and income earning potential of the house- dress current climate variability. For example, Skoufias, Essama- hold, whether the household is a net agricultural producer or Nssah, and Katayama (2011) show that climatic variability-- consumer, and on the opportunities for adaptation and risk weather shocks--has significant impacts on the average management available to the household. Adaptation to climate well-being of rural households in Indonesia. In particular, a de- change can reduce poverty impacts substantially. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise The evidence also suggests there is much that policy makers 3. For example, Anderson (2006) reports that poverty in sub- can do to help the poor better adapt and cope with climate Saharan Africa in 2100 would be 0.5 percentage points higher change and extreme weather events without compromising hu- than in a no climate change baseline using the mean of the dam- man capital, which is the long-term foundation of household age distribution, but would be 2 percentage points higher using welfare. Adaptive policies can include creating well-targeted the 95th percentile. and scalable safety nets; improving the poor's access to credit 4. Autonomous adaptation is typically distinguished from and insurance markets; reducing impediments to occupational planned adaptation, which refers to policy-based actions that mobility; improving governance of common-pool natural re- are needed when market failures or other coordination prob- sources; investing in irrigation and/or improved water manage- lems hinder relevant collective responses to climate change. ment to deal with extreme precipitation events; and smoothing 5. World Bank (2010) surveys private and public risk mitiga- the food price impacts of regional or country-specific climate tion strategies in the face of natural hazards. shocks through, for example, greater openness to international References trade. Fortunately, many of the policies that can be effective for reducing the climate change impacts on poverty are also effec- Ahmed, S. A., N. Diffenbaugh, and T. Hertel. 2009. "Climate Volatility Deep- ens Poverty Vulnerability in Developing Countries." Environmental Re- tive for reducing poverty and promoting economic growth. search Letters 4 (3): 1­8. Anderson, E. 2006. "Potential Impacts of Climate Change on $2-a-Day Poverty About the Authors and Child Mortality in Sub-Saharan Africa and South Asia." Unpublished manuscript, Overseas Development Institute, UK. Emmanuel Skoufias is Lead Economist and Mariano Rabassa and Assunção, J., and F. Chein Feres. 2009. "Climate Change, Agricultural Produc- Sergio Olivieri are Consultants in the Poverty Reduction and Eq- tivity, and Poverty." Working Paper, Department of Economics, PUC-Rio uity Group at the World Bank. Milan Brahmbhatt is Senior Ad- de Janeiro, Brazil. viser in the Poverty Reduction and Economic Management (PREM) Chen, S., and M. Ravallion. 2009. "The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight Against Poverty." Policy Re- Network at the World Bank. The authors wish to thank Jaime search Working Paper 4703, World Bank, Washington, DC. Saavedra, Mike Toman, and Paula Suarez Buitron for comments. Hertel, T., M. Burke, and D. Lobell. 2010. "The Poverty Implications of Cli- mate-Induced Crop Yield Changes by 2030." Global Environmental Notes Change 20 (4): 577­585. Jacoby H., M. Rabassa, and E. Skoufias. 2011. "On the Distributional Implica- 1. Projected annual per capita growth rates by RICE are decreas- tions of Climate Change: The Case of India." Forthcoming Policy Research ing over time. The annual world output growth also masks con- Working Paper, World Bank, Washington, DC. Nordhaus, W. 2010. "Economic Aspects of Global Warming in a Post-Copen- siderable regional disparities; for example, while China and In- hagen Environment." Proceedings of the National Academy of Science 107 dia are expected to grow at a 3.6 annual per capita rate, the (26): 11721­26. European Union will grow at a 1.8 annual rate. Olivieri, S., M. Rabassa, and E. Skoufias. 2010. "Potential Impacts of Climate 2. It is useful to benchmark Nordhaus's (2010) business as usu- Change on Poverty: An Analysis Based on RICE 2010 Predictions." Un- published manuscript, PRMPR, World Bank, Washington, DC. al scenario against other well-known climate change scenarios. Skoufias, E., B. Essama-Nssah, and R. Katayama. 2011. "Too Little, Too Late: The Stern Review (2007), for example, estimates that the mean Welfare Impacts of Rainfall Shocks in Rural Indonesia." Policy Research loss in world output in 2100 would be 2.9 percent under its Working Paper, World Bank, Washington, DC (to be published also in the Bulletin of Indonesian Economic Studies). high climate scenario, inclusive of nonmarket and catastrophic Skoufias, E., M. Rabassa, and S. Olivieri. 2011. "Climate Change, Poverty, and effects. The Nordhaus business as usual scenario presumes a Economic Growth: A Review of the Evidence." Unpublished manuscript, somewhat larger 3.3 percent loss in 2105. Differences in infer- PRMPR, World Bank, Washington, DC. ences from various models depend more on whether one exam- Stern, N. 2007. The Economics of Climate Change: The Stern Review. Cambridge, UK: Cambridge University Press. ines mean impacts of uncertain climate change or the tails of World Bank. 2010. Natural Hazards and Unnatural Disasters: The Economics of the impact distribution. Effective Prevention. Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduc- tion and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise