45119 The World Bank PREMnotes J U L Y 2 0 0 8 N U M B E R 1 SPECIAL SERIES ON THE ECONOMICS OF CLIMATE CHANGE Who bears the burden of environmental policies within countries? Antonio Estache and Luc Savard Climate change policies will have distributional consequences across and within countries. Most of the current environmental policy instruments tend to be regressive and impose a higher burden on the poor. Despite their limitations, more systematic incidence assessments for CC policies are needed so that adaptation and mitigation policies address their distributional effects within countries. The threat of catastrophic climate change Taxes, fees and tariffs (CC), if left unchecked, is leading govern- ments across the world to consider a wide to fund CC policies range of policy actions at the international According to the UN, lower and lower middle level. These actions often commit govern- income countries will need to commit about ments to deliver policy changes at the national 0.5% of GDP annually for the foreseeable level. While there is a lot of discussion on the future to adapt to Climate Change. The IPCC share of the global burden that the developed also estimates the cost of mitigation to achieve and developing countries will have to take a 450ppm CO2 stabilization target (i.e. the on, there is little debate on the distribution costs of financing technological change and of this burden within countries. substitutions to reduce inputs and emission Although adaptation is currently in the per every unit of output) to be about 1.6% limelight in the international community, of GDP. This means that the equivalent of an the design of mitigation policies so as to additional 2.1% of GDP will have to be funded manage demand for many basic services is a from new or existing sources. cornerstone of any CC related policy. More- Most of these costs will have to be fi- over, this design will affect the distribution of nanced through a combination of increases the sources of the greenhouse gas emissions in old taxes (such as motor vehicle taxes), that contribute to CC and the specific targets new taxes, and equivalent revenue raising these policies will have to focus on in the instruments (such as betterment taxes) or future. If OECD experience is an indicator levies to capture some of the property value of things to come within countries, the share changes resulting from the policies. In many of the residential sector as a contributor to countries, these costs will also be financed emissions will rise significantly (35-40% of through improvements in cost recovery of emissions, and one third of this from ap- public services. pliances). The impact of mitigation policies Considering that the average indirect tax aimed at switching to low carbon options on revenue is about 8-10% of GDP, depending the affordability of basic services is thus likely on the level of development, financing these to be an issue of growing concern. costs through indirect taxes or equivalent FROM THE POVERTY REDUCTION AND ECONOMIC MANAGEMENT NETWORK instruments would increase that burden by How to find out who about 20%. Unless there are explicit concerns ends up paying? for the fairness of policy instruments, for most of the poor it will result in an increase Although the literature on tax and expendi- in the cost of their current consumption ture incidence spans over 40 years, there is bundle of at least 20% (disregarding any little empirical research on the incidence of cascading effect likely to result from poorly environmental policies across income classes designed environmental taxes functioning as at the national level. Most of the analysis has the equivalent of sales taxes). been done for large countries (United States, The large adjustment in prices is likely Germany, the Nordic countries and for a few to change at least some consumption and developing countries). Most focus on partial production decisions, in addition to generat- assessments of the usual instruments con- ing the revenue to finance the investments sidered by environmental economists: price needed to address transitional adaptation based instruments such as gasoline taxes, and mitigation costs. This is consistent with other energy taxes, carbon taxes, and motor the objectives of environmental policies. Tax- vehicles taxes; cost based instruments such as es, fees, and public service tariff increases and subsidies or grants for changes in behavior or rebalancing are needed to increase aware- production technology; and pure technology ness of the risks of CC and hence change or rationing instruments, such as emission consumer and investor behavior. They also standards or permits. need to ensure that the awareness of the risks Most of this research tracks the effects of is transformed into concrete actions and not these policies through the price or cost chain just commitments. in input-output tables and on consumption patterns of various income groups. Comput- able general equilibrium (CGE) modeling is Who will pay for the sometimes used to add micro-simulation of added costs? changes in supply and demand for goods and Policy debates often assume that the direct factors, reflecting differences in elasticities buyers and sellers of products targeted by across producers and across consumers. new taxes or tariff adjustments driven by Policy interventions are typically mod- CC concerns are the main "victims" of these eled as an explicit or implicit product or policies. In practice, this is not necessarily emission tax on an output. Prices will change the case. The debate should be driven by a for producers and consumers. The models much deeper assessment of the direct and recognize that income also changes through indirect impacts of the policies. wage changes and other factor payments. Changes in prices driven by taxes, subsi- This allows a tracking of the effects of CC dies or brutal changes in demand or supply policies across industries and across house- have complex direct and indirect effects with hold types. For countries in which lagging unforeseen distributional effects. For instance, regions are a concern, these models can be a new tax on a product which results in a de- used to track differences across regions and cline in the consumption of that product will track the scope for trade-offs versus win-win impact various income classes through reduc- situations associated with environmentally tions in the derived demand for labor or capi- conscious, growth stimulating policies for tal in the taxed industry. If the poor represent the poorer regions. a large share of the workers in that sector, they will be hurt by the tax disproportionately. Lessons from OECD countries Consumers and investors may change their Assessments of individual CC policies reveal behavior as intended by the CC policies, but that they tend to be regressive. The result this could have socially harsh consequences. holds for a wide range of instruments, par- For any government concerned with the well ticularly for energy taxes. The evidence sug- being of the poor, or the fair distribution of gests that firms tend to respond to environ- the burden of policies, the final incidence of mental taxes by adopting less labor-intensive any CC policy must be assessed. 2 PREMNOTE JULY 2008 technologies that, in particular, reduce the The second difference is the increased demand for unskilled labor, disproportion- risks for the poor as consumers. Some of the ately penalizing the poor. specific sources of added tension include the Research also shows that CC policies' higher cost of basic nutrition sources (the regressivity can be partially, offset. Recycling debate on the use of corn for biofuels rather revenue from CC based taxes and permits than food in Mexico for instance) and the seems to be a very effective solution, but in tendency of CC policies to favor urban over order to make it progressive it would require rural consumers. a reform of the income tax system as well. Creating jobs is another way to reduce the regressivity of CC policies (a double divi- Limitations of current dend). Investing the revenue from CC tax incidence analyses instruments in infrastructure biased towards The models used to assess the incidence the needs of the poorest is a third option to of CC policies in developed countries have offset the regressivity of these taxes. Finally some limitations when applied to develop- regressivity can also be reduced by ensuring ing countries. First, standard models often that the benefits of CC policies are biased in assume competitive pricing and constant favor of the poorest. returns to scale. These are not typical in many key industries (i.e. energy) which are What we know from regulated. These models, therefore, under- estimate the progressivity of CC price-based developing countries policies intended in regulatory design but There are significantly fewer incidence analy- may also underestimate the regressivity associ- ses of CC policies for developing countries ated with rationing when service coverage than there are for developed countries. Our is an issue. review of the literature identified published Second, these models often assume case studies for Brazil, Cameroon, Chile, downward supply curves. This is not always Senegal, and Thailand. All generally confirm the case for key industries associated with CC, the regressive nature of most instruments as particularly in developing countries where observed in OECD countries. industry specific capital or dependency on When assessing the distributional out- a scarce natural resource is common. This comes of CC policies, there are two main also results in underestimates of the progressivity types of differences between developed and of CC policies since the assumption implies developing countries, and they imply a sig- lower rents rather than higher prices. nificant additional burden on the poor. Other limitations include: a) failure to The first is driven by the adjustment reflect that agriculture is a major employer, as and transition costs for the poor as workers. well as, a major supplier of local food, which As industries adopt new technologies, labor can also become a major tradable under market transitions can be quite painful. This global CC policies (i.e. biofuels); b) not taking was the case with the initial job loss or dis- into account that the composition of supply placement associated with the introduction may change significantly as a result of global of Brazil's large-scale proalcool program that changes (such as trade) driven by carbon intended to reduce the dependence on tra- footprinting; c) overlooking the regressivity ditional fuels. Changes in the relative prices of subsidies to water, energy, and transport in of agricultural products associated with CC their current design; and d) failing to offer policies may also change comparative ad- joint assessments of the distributional effects vantages and hence production structure of benefits, income, and expenditures associ- towards more capital intensive products. ated with CC policies. Finally, changes in production location within Furthermore, in developing countries countries or across countries can be much large shares of the population are served by more hurtful in poor countries than in OECD informal markets which can distort many countries' labor markets. evaluations of the distributional effects of JULY 2008 PREMNOTE 3 CC policies. Incidence analyses are also yet the same, the adaptation of land use will be to account for dimensions such as: enforce- confronted with inertia and ultimately, this ment costs or problems, differences in de- generation's problem will simply be passed mand elasticities across income groups, and on to the next generation with even less scope the differences in expenditure and income for corrections and adjustments. structures across income groups, which can all result in misleading diagnostics. This brief draws from A. Estache and L. Savard Despite these limitation there remains (2008), "Incidence of Environmental Policies a strong analytical case for conducting more in Developing Countries: A Survey," PREM systematic incidence analysis of CC policies. Economics of Climate Change Discussion Papers. These models can help in the design of poli- The specific references and sources for the ideas cies that are both efficient and fair. summarized in this paper are available in this Discussion Paper. Why policymakers need to care about the fairness of CC policies About the authors Ultimately, it is important to recognize that Antonio Estache is with the European Center for the benchmark to compare the incidence Advanced Research in Economics and Statistics of of CC policies should be the long-run in- the Université Libre de Bruxelles (ECARES) and cidence of the failure to act. Ignoring the is also an advisor to the PREM Vice President. consequences of CC is widely expected to Luc Savard is with the Economics Department of be more regressive in developing countries the Université de Sherbrooke in Canada and the than in developed countries. For instance, Groupe de Recherche en Economie et Développe- up to 80% of the poorest rely on more dam- ment International (GREDI). aging (health wise) energy sources for their immediate needs compared to longer-term About the note series strategic concerns. Thus, policy interventions The PREM notes on the Economics of Climate seem to be the only ethical option from the Change are part of the effort conducted by the viewpoint of the poorest. Poverty Reduction and Economic Management Caring about the incidence of climate Vice Presidency of the World Bank to raise aware- change policies is not only needed for ethi- ness on poverty, distributional, financial, fiscal, cal reasons. It is also important to ensure the and trade related issues that tend to be underesti- social acceptability and political viability of mated in the more scientific and political debates reforms. Both tend to be higher with progres- surrounding Climate Change. The notes do not sive reforms than with regressive ones. necessarily reflect the view of the World Bank, its This survey has shown that most CC board or its member countries. However, they do policy reforms tend to be regressive and un- reflect the content of some of the internal debates less this is addressed, the talk about climate among economists interacting traditionally on change will continue to be just talk. No emerging or overlooked economic consequences of serious action will take place. The demand environmental policies. for energy consumption will continue to ex- For questions, please contact Milan Brahmbhatt plode, the number of private vehicles will do at mbrahmbhatt@worldbank.org. This note series is intended to summarize good practices and key policy findings on PREM-related topics. The views expressed in the notes are those of the authors and do not necessarily reflect those of the World Bank. PREMnotes are widely distributed to Bank staff and are also available on the PREM Web site (http://www.worldbank.org/prem). If you are interested in writing a PREMnote, e-mail your idea to Madjiguene Seck at mseck@worldbank.org. For additional copies of this PREMnote please contact the PREM Advisory Service at x87736. PREMnotes are edited and laid out by Grammarians, Inc. Prepared for World Bank staff