Climate and sustainability in developing countries
Concerns of economic and social sustainability - in particular in relation to climate change - play an increasingly central role in policy discussions both internationally and in Sweden. There is also a growing consensus that the most vulnerable areas in the world are to be found in the poorest regions. Yet academic research on understanding the nature of damages in low income countries and their ways of adapting to changing climates and shortfalls of various key resources is scant and scattered. Most importantly, there is very little work on how different forms of economic policy can help both limit the fundamental challenges and adapt to them. The present program will build a coherent policy evaluation structure allowing us to address these questions.
A unique feature of the program is that it involves close cooperation between macro- and development economists. The former are specialists on building coherent theory structures aimed at concrete policy evaluations. The latter provide expertise in field research using experimental methods. A key component of the research program is to build a policy-evaluation structure at the macro level, based on empirical findings from micro studies on agricultural markets, technology adoption, migration, and estimates of economic damages due to global warming. Such a structure can then be used to examine how concrete policies aimed at mitigation and adaptation affect welfare and poverty.
A unique feature of the program is that it involves close cooperation between macro- and development economists. The former are specialists on building coherent theory structures aimed at concrete policy evaluations. The latter provide expertise in field research using experimental methods. A key component of the research program is to build a policy-evaluation structure at the macro level, based on empirical findings from micro studies on agricultural markets, technology adoption, migration, and estimates of economic damages due to global warming. Such a structure can then be used to examine how concrete policies aimed at mitigation and adaptation affect welfare and poverty.
Final report
The program stated, as an overriding goal, the aim of obtaining “a better understanding of the role of climate change and, more generally, issues of sustainability, in underdeveloped countries” and to start a close collaboration between macroeconomists and development economists. At least by our own judgment, these aims have been attained in a highly satisfactory manner.
First, purely in terms of research, we published many papers in prestigious journals and we have developed the research tools we set out to develop. We regard this research as containing important steps forward in the relevant literatures, and not just as marginal contributions. The collaboration between macroeconomists (with a climate focus) and development economists – a unique joint venture within economics, broadly defined – is in full swing and has delivered very promising insights.
Second, we have developed PhD and master’s courses in climate and economics.
Third, in its application, the program also stated that it sought to “reach policymakers with an important tool, insights, and concrete advice”. Here, we have also been very successful, at least in terms of policy impact in Sweden. The clearest expression of the impact is arguably John Hassler’s report to the government (Sveriges klimatstrategi) which not only builds on the insights generated in our research but also has had changed actual policy in Sweden, after the report was presented.
Fourth, although not a stated goal in our application, our program has also turned out to be an excellent base on which we have been able to successfully apply for research grants elsewhere and that are now helping us carry the research forward. Two ERC grants (Jakob Svensson and Ingvild Almås) as well as a Wallenberg Scholar continuation grant (Per Krusell) are now starting and other, smaller grants have been received as well.
We now give some more detail on the different subprojects and their respective research results.
The macro perspective
A core aim of the project was to construct a comprehensive integrated assessment model of climate and economics with very high regional resolution. This construction has been accomplished: relative to the early pilot version that we had at the start, a full model has now been put together. The main paper has been submitted to Econometrica (a top-5 journal) where it is now in a 2nd-round revision.
Given the particular focus on developing countries, a core avenue for the research has been to study agriculture and how the structure of agriculture, along with its interaction with the rest of the economy, changes as an economy develops. In particular, how the agricultural sector is able to adapt to climate change comes in focus. With the aim of characterizing the “aggregate agricultural production function” at each level of development, a key paper in our program has developed a method for measuring this function in a “reduced form”. A surprising finding here is that the agricultural sector appears far more able than expected to adjust to shortages in land, and more generally respond to shortages of input, by the use of other inputs. The paper has been completed and been submitted to a top-5 journal.
The paper just mentioned has also generated a very exciting project, with especially intense collaboration between macroeconomists and development economists, trying to seek the deeper, microeconomically based roots of the reduced-form production function. The idea here is to examine how severe difficulties for individual farmers to adjust to shortages can still lead to overall production output featuring significant substitutability across factors. This work will be continuing well beyond the conclusion of the program.
Finally, we have conducted a first round of policy and counterfactual experiments with the main IAM (Integrated Assessment Model), in important part motivated by the findings in the micro development literature. We describe the micro contributions of the program first, and finally the IAM findings.
The micro perspective
The microeconomic parts of the research are key in building an understanding of the basic mechanisms playing out in poor countries in response to climate change. However, building an understanding of how this works involves deepening our understanding of how poor/village economies function more generally. The sub-projects touch on a number of important aspects. The results from these studies are many and very well published. They underlie the broader, macroeconomic modeling we report here in a general sense but not in terms of specific modeling insights. The joint macro-micro work that is under way does, however, develop such specific insights; it is not fully completed research and hence not written up at this point in time.
We now very briefly describe the core micro-projects.
One project – published in the American Economic Review – studies the interaction between demand and supply factors in making high-quality produce appear as market outcomes in sub-Saharan Africa. The paper shows, using RCT methods, that demand-side constraints (“missing markets”) – which are often overlooked in the literature – play an important role in preventing more advanced products and technology from being adopted. Another project develops new methodology for comparisons between countries regarding views on inequality and fairness. One paper, published in the Journal of Political Economy, uses the new methodology to compare the U.S. and Norway, finding significant differences in attitudes toward fairness but no differences in those vis-à-vis efficiency. Another paper, in the Proceedings of the National Academy of Sciences, looks at a global evidence on people’s views on to what extent selfishness drives inequality and finds support for large differences across countries. Motivated by the concern that there are poverty traps where low incomes, via low nutrition, are self-generating, one project studies RCTs with unconditional cash transfers to low-income households in rural Kenya and measures larger effects on nutrition and food intake than in previous studies, thus strengthening the worry about traps. Another project, consisting of several papers, among them one in Econometrica and two prestigious Handbook chapters, aims to ascertain whether transfers to women are more effective in producing beneficial outcomes for children than transfers to men (or to the family). Here, it has turned out to be difficult to find significant effects in many cases, but whether this result is a reflection of mothers’ and fathers’ preferences being aligned or of mothers’ power in the household being low has not been established. Yet another project, submitted for publication, provides the first RCT-based estimates of the mitigating effects of health system strengthening, concluding that although a warming climate may amplify child mortality, health programs show great promise if implemented efficiently. However, the scaling up of such health programs proves challenging and an “adaptive” implementation approach is found to be important in a follow-up paper. Finally, one project studies the effects of heat on human behavior with identical experiments run in Berkeley and in Nairobi. The study finds no effects at all of heat on economic behavior, but it finds effects on “destructive behavior” in Nairobi but not in Berkeley, suggesting that global warming itself can give rise to an increased degree of conflict and war, even without a decrease in resources. The paper is forthcoming in Economic Journal, a top European journal.
The IAM experiments
Two main IAM experiments were run, both making important use of the regional heterogeneity that the model allows. One involved setting different carbon tax rates in different regions. It is well known that an economically efficient carbon policy, from a global perspective, is to tax carbon at a common rate (or, equivalently, have a world market for emission rights so that emissions costs are identical per unit of carbon at all places of emission). And since a unit of emission causes the same damage independently of where it is emitted, the carbon tax should be the same everywhere. Thus, the use of different tax rates is inefficient. However, it is often argued that equal tax rates on carbon is not fair: poor countries should not have to pay such high taxes. Indeed, in our micro development work, we found fairness, and equality, to be important drivers of public opinion, and hence it is understandable that such proposals have been made. We thus used our IAM to examine simply how large the costs of unequal carbon taxation would be. The results were surprisingly large losses from the “fairer” policy, so that large global losses would materialize if these policies were pursued. Our conclusion is that it is better to maintain a uniform global tax and instead help the poor countries with direct transfers or technology support. In a second experiment, we examined the efficacy of replacing carbon taxes with green technology initiatives. Again, motivated by the micro studies, it is not clear that green technology, even if developed and available for free, will be adopted in developing-country contexts. This experiment also delivered surprising results. It turned out that whether the new green technology spreads across the whole world or not is not so relevant for climate change, since green technology is a poor substitute for carbon taxes. Put differently, more green technology is helpful in allowing us to use more energy, but it will not decrease the use of fossil-based energy more than marginally. Thus, the substitutability of different energy sources is sufficiently low that policies aimed at green technology development alone are not of much value for the climate in the absence of carbon taxes, and whether this green technology is then available in poor countries or not is of second-order importance. These experiments are part of the paper at Econometrica.
First, purely in terms of research, we published many papers in prestigious journals and we have developed the research tools we set out to develop. We regard this research as containing important steps forward in the relevant literatures, and not just as marginal contributions. The collaboration between macroeconomists (with a climate focus) and development economists – a unique joint venture within economics, broadly defined – is in full swing and has delivered very promising insights.
Second, we have developed PhD and master’s courses in climate and economics.
Third, in its application, the program also stated that it sought to “reach policymakers with an important tool, insights, and concrete advice”. Here, we have also been very successful, at least in terms of policy impact in Sweden. The clearest expression of the impact is arguably John Hassler’s report to the government (Sveriges klimatstrategi) which not only builds on the insights generated in our research but also has had changed actual policy in Sweden, after the report was presented.
Fourth, although not a stated goal in our application, our program has also turned out to be an excellent base on which we have been able to successfully apply for research grants elsewhere and that are now helping us carry the research forward. Two ERC grants (Jakob Svensson and Ingvild Almås) as well as a Wallenberg Scholar continuation grant (Per Krusell) are now starting and other, smaller grants have been received as well.
We now give some more detail on the different subprojects and their respective research results.
The macro perspective
A core aim of the project was to construct a comprehensive integrated assessment model of climate and economics with very high regional resolution. This construction has been accomplished: relative to the early pilot version that we had at the start, a full model has now been put together. The main paper has been submitted to Econometrica (a top-5 journal) where it is now in a 2nd-round revision.
Given the particular focus on developing countries, a core avenue for the research has been to study agriculture and how the structure of agriculture, along with its interaction with the rest of the economy, changes as an economy develops. In particular, how the agricultural sector is able to adapt to climate change comes in focus. With the aim of characterizing the “aggregate agricultural production function” at each level of development, a key paper in our program has developed a method for measuring this function in a “reduced form”. A surprising finding here is that the agricultural sector appears far more able than expected to adjust to shortages in land, and more generally respond to shortages of input, by the use of other inputs. The paper has been completed and been submitted to a top-5 journal.
The paper just mentioned has also generated a very exciting project, with especially intense collaboration between macroeconomists and development economists, trying to seek the deeper, microeconomically based roots of the reduced-form production function. The idea here is to examine how severe difficulties for individual farmers to adjust to shortages can still lead to overall production output featuring significant substitutability across factors. This work will be continuing well beyond the conclusion of the program.
Finally, we have conducted a first round of policy and counterfactual experiments with the main IAM (Integrated Assessment Model), in important part motivated by the findings in the micro development literature. We describe the micro contributions of the program first, and finally the IAM findings.
The micro perspective
The microeconomic parts of the research are key in building an understanding of the basic mechanisms playing out in poor countries in response to climate change. However, building an understanding of how this works involves deepening our understanding of how poor/village economies function more generally. The sub-projects touch on a number of important aspects. The results from these studies are many and very well published. They underlie the broader, macroeconomic modeling we report here in a general sense but not in terms of specific modeling insights. The joint macro-micro work that is under way does, however, develop such specific insights; it is not fully completed research and hence not written up at this point in time.
We now very briefly describe the core micro-projects.
One project – published in the American Economic Review – studies the interaction between demand and supply factors in making high-quality produce appear as market outcomes in sub-Saharan Africa. The paper shows, using RCT methods, that demand-side constraints (“missing markets”) – which are often overlooked in the literature – play an important role in preventing more advanced products and technology from being adopted. Another project develops new methodology for comparisons between countries regarding views on inequality and fairness. One paper, published in the Journal of Political Economy, uses the new methodology to compare the U.S. and Norway, finding significant differences in attitudes toward fairness but no differences in those vis-à-vis efficiency. Another paper, in the Proceedings of the National Academy of Sciences, looks at a global evidence on people’s views on to what extent selfishness drives inequality and finds support for large differences across countries. Motivated by the concern that there are poverty traps where low incomes, via low nutrition, are self-generating, one project studies RCTs with unconditional cash transfers to low-income households in rural Kenya and measures larger effects on nutrition and food intake than in previous studies, thus strengthening the worry about traps. Another project, consisting of several papers, among them one in Econometrica and two prestigious Handbook chapters, aims to ascertain whether transfers to women are more effective in producing beneficial outcomes for children than transfers to men (or to the family). Here, it has turned out to be difficult to find significant effects in many cases, but whether this result is a reflection of mothers’ and fathers’ preferences being aligned or of mothers’ power in the household being low has not been established. Yet another project, submitted for publication, provides the first RCT-based estimates of the mitigating effects of health system strengthening, concluding that although a warming climate may amplify child mortality, health programs show great promise if implemented efficiently. However, the scaling up of such health programs proves challenging and an “adaptive” implementation approach is found to be important in a follow-up paper. Finally, one project studies the effects of heat on human behavior with identical experiments run in Berkeley and in Nairobi. The study finds no effects at all of heat on economic behavior, but it finds effects on “destructive behavior” in Nairobi but not in Berkeley, suggesting that global warming itself can give rise to an increased degree of conflict and war, even without a decrease in resources. The paper is forthcoming in Economic Journal, a top European journal.
The IAM experiments
Two main IAM experiments were run, both making important use of the regional heterogeneity that the model allows. One involved setting different carbon tax rates in different regions. It is well known that an economically efficient carbon policy, from a global perspective, is to tax carbon at a common rate (or, equivalently, have a world market for emission rights so that emissions costs are identical per unit of carbon at all places of emission). And since a unit of emission causes the same damage independently of where it is emitted, the carbon tax should be the same everywhere. Thus, the use of different tax rates is inefficient. However, it is often argued that equal tax rates on carbon is not fair: poor countries should not have to pay such high taxes. Indeed, in our micro development work, we found fairness, and equality, to be important drivers of public opinion, and hence it is understandable that such proposals have been made. We thus used our IAM to examine simply how large the costs of unequal carbon taxation would be. The results were surprisingly large losses from the “fairer” policy, so that large global losses would materialize if these policies were pursued. Our conclusion is that it is better to maintain a uniform global tax and instead help the poor countries with direct transfers or technology support. In a second experiment, we examined the efficacy of replacing carbon taxes with green technology initiatives. Again, motivated by the micro studies, it is not clear that green technology, even if developed and available for free, will be adopted in developing-country contexts. This experiment also delivered surprising results. It turned out that whether the new green technology spreads across the whole world or not is not so relevant for climate change, since green technology is a poor substitute for carbon taxes. Put differently, more green technology is helpful in allowing us to use more energy, but it will not decrease the use of fossil-based energy more than marginally. Thus, the substitutability of different energy sources is sufficiently low that policies aimed at green technology development alone are not of much value for the climate in the absence of carbon taxes, and whether this green technology is then available in poor countries or not is of second-order importance. These experiments are part of the paper at Econometrica.