133966 S pe n ci En o ResearchDigest al er Fo gy World Bank cu s VOLUME 13 NUMBER 2 WINTER 2019 Sustainable Energy Is Key to Achieving Global Development Goals S ustainable energy—the ability at supporting technical assistance and IN THIS ISSUE to have access to adequate and lending as well as informing public affordable supplies of safe and debate on the adoption of effective Sustainable Energy Is Key to Achieving reliable energy sources—is inextricably policies and approaches. This issue Global Development Goals … page 1 linked to nearly every major develop- features papers highlighting both chal- ment challenge the world faces today. lenges and opportunities for develop- Charting the Diffusion of Power The more than 1 billion people liv- ing countries in advancing sustainable Sector Reforms across the Developing ing without electricity confront limits energy objectives. World … page 2 in opportunities. Without electricity, One study, examining electricity How Much Do Rural Households women spend hours fetching water, sector reform in 88 developing coun- Benefit from Electrification? Evidence clinics cannot store vaccines, school- tries over a 25-year period, finds that from Bangladesh … page 3 children cannot study after dark, and in most of those countries reforms people face constraints in their choices were only partially implemented. It Forecasting Electricity Demand of business activities. The almost 3 bil- concludes that countries’ economic in Developing Countries: Is There a lion people lacking access to clean and and political characteristics have much Better Way? … page 4 affordable fuels and stoves for cooking to do with their uptake of reforms. and heating are exposed to dangerous Another study finds that rural electri- Expanding Cross-Border Electricity levels of indoor air pollution—expo- fication in Bangladesh has aided such Trade to Promote Hydropower sure that each year leads to nearly 4 key development outcomes as higher in South Asia … page 5 million premature deaths. household incomes, greater well-being Achieving sustainable energy is for women and children, and lower use The Nexus of Energy Supply and possible, but countries need to take of kerosene. Other studies highlight Human Health … page 6 concrete steps toward prioritizing en- the importance of clean, renewable ergy efficiency, reducing regulatory dis- energy sources, including solar and Can Improved Biomass Stoves Be Cost-Effective in Mitigating tortions, and investing in clean, renew- hydropower, for improving health, re- Climate Change? … page 7 able energy sources. Ensuring access ducing greenhouse gas emissions, and to affordable, reliable, sustainable, sustaining economic growth. Can Peer Pressure Encourage and modern energy for all is among Research offers ample evidence People to Adopt Energy-Saving the 17 Sustainable Development Goals that better regulatory frameworks, Technologies? … page 8 (SDGs) adopted by the United Nations more efficient energy pricing, and ef- General Assembly in 2015. Known as fective incentive structures for clean Solar Lanterns and Youth Academic SDG7, this goal interacts with many energy adoption are critical for im- Performance in Bangladesh … page 9 of the other SDGs, including improv- proving sustainable energy. Achieving ing livelihoods, providing more secure all these requires strong political will, The Complexities of Integrating Renewable living conditions, enhancing access to increased investment, and greater in- Energy into a Power System … page 10 development opportunities for women, volvement by all stakeholders. Yearly and reducing environmental harms. investment in sustainable energy Interactions of Renewable Energy The World Bank has been an im- infrastructure needs to triple, from Standards with Carbon Emission Trading portant source of financing and advice around $400 billion today to $1.25 tril- for Climate Change Mitigation … page 11 to improve the efficiency and sustain- lion by 2030. Recent Policy Research Working Papers ability of energy supplies in developing Countries must step up. Without on Energy … page 12 countries. Its research on sustainable sustainable energy, the world may fall energy, as envisaged in SDG7, is aimed short of its development goals. 2 World Bank ResearchDigest Charting the Diffusion of Power Sector Reforms across the Developing World After rapid diffusion in 1995–2005, a power sector that remains largely for power sector reform and the the spread of power sector reforms unreformed. The other 35 percent find Middle East and North Africa as well across the developing world slowed themselves at an intermediate stage, as Sub-Saharan Africa lagging behind having adopted some elements of the and introducing more limited reforms. significantly in the following decade reform package but with little sign of Moreover, the vast majority of de- S continuing reform momentum. veloping countries remaining at the ome 25 years have elapsed since In a number of developing coun- early stages of reform fall into one of international financial institu- tries published policy commitments the categories that have historically tions espoused a package of to power sector reform could not found power sector reform challenging reform measures that became known subsequently be implemented be- to implement: fragile states, small sys- as the Washington Consensus. In the cause of political opposition. As a tems, low income bracket, or weak rule power sector, similar to other utility result, reforms were only selectively of law. Many of these countries are industries, this package encompassed adopted, often according to ease of in Sub-Saharan Africa. Indeed, there the establishment of autonomous implementation. By far the most com- is not a single large, stable, middle- regulatory entities, the vertical and mon reforms were the introduction of income country in this group of coun- horizontal unbundling of integrated independent power projects and the tries at the early stages of reform. national monopoly utilities, private creation of regulatory entities; each of These patterns in the uptake of sector participation in generation these reforms was adopted by about power sector reform already tell an and distribution, and eventually the 70 percent of developing countries, illuminating story. They illustrate introduction of competition in power perhaps because neither represented that the full package of Washington generation and even retail services. a major challenge to vested interests Consensus reforms in the sector has A paper by Foster, Witte, Banerjee, in the sector. On the other hand, proved challenging to implement in and Moreno provides a much more privatization of distribution utilities its entirety in a developing country detailed characterization of the con- was undertaken by only 44 percent of environment and, though undoubtedly tent, timing, and sequencing of power developing countries and proved more much more widespread in developed sector reform measures than has difficult, with about 20 countries expe- countries, still remains far from uni- previously been possible, covering 88 riencing reversals of some 40 private versal even in that group. developing countries over the period transactions for power distribution. 1990–2015. (The paper takes a descrip- Partial reform is sometimes associ- tive approach, as other studies have ated with incoherence. For example, already undertaken broad evaluations about half the developing countries of the impact of such reform.) that created regulatory entities did so Overall, power sector reform in without introducing the private sec- the developing world lags far behind tor participation in power distribution what was implemented in the devel- that originally motivated the creation oped world during the same period. of the regulator. Unbundling of power utilities is twice Country characteristics such as as common in developed as in de- geographic region, income group, veloping countries, while creation of power system size, and attributes of wholesale power markets is 10 times the political system seem to have had as likely. Nonetheless, full power sec- a statistically significant influence on tor reform has not been universal even the uptake of reform. One of the larg- in developed countries. est influences has been system size: Following rapid diffusion during developing countries with installed ca- the decade 1995–2005, the spread of pacity of more than 10 gigawatts went power sector reforms across the de- twice as far with power sector reform veloping world slowed significantly as did those with capacity of less than in the subsequent decade (2005–15). 1 gigawatt. Similarly, developing coun- By 2015 only about 25 percent of de- tries in the middle-income bracket, veloping countries had been able to and those with relatively competitive adopt the full package of Washington political systems, went much further Vivien Foster, Samantha Witte, Sudeshna Ghosh Consensus reforms in the power sec- with the reform process. Geographic Banerjee, and Alejandro Moreno. 2017. “Charting tor, often over a period of many years. region was also an important driver, the Diffusion of Power Sector Reforms across the At the same time, as many as 40 with Latin America and the Caribbean Developing World.” Policy Research Working Pa- percent of developing countries have standing out as the pioneering region per 8235, World Bank, Washington, DC. World Bank Research Digest 3 How Much Do Rural Households Benefit from Electrification? Evidence from Bangladesh The benefits of rural electrification with the selection bias, the authors Finally, using a quantile regression depend critically on the quality of use a two-stage propensity-score- approach, the study finds that higher- electricity supply—and are fully weighted fixed-effects model. They first income households benefit more estimate a household’s probability of from electrification than lower-income realized only over the long term being connected to the grid (propensi- households during the time span cov- B ty score) based on a range of house- ered by the evaluation. Richer house- ringing electricity to rural areas hold and village characteristics holds are typically early adopters of has been widely recognized as observed in the base year. They then electricity and can afford a wider range a key driver in reducing poverty assign to each household a weight of electrical appliances that could al- and boosting economic development. proportional to the propensity score so low more productive use of electricity. Not surprisingly then, rural electrifica- that households that look more similar Since the marginal return to electri- tion has ranked high on the agendas to connected households in the base fication declines over time, however, of governments, nongovernmental year receive more weight. Finally, to poorer households that are connected organizations, and international de- estimate benefits of grid electrifica- to the grid may later catch up. velopment agencies. This underscores tion, the authors compare changes Accounting for income growth the importance of accurately under- over time within households that were alone, the study estimates that the standing its benefits. Estimates vary connected to the grid in 2010 with total potential benefits from achieving widely across studies. While many changes within households that were universal access to electricity while studies find substantial benefits of not connected. also improving reliability could reach electrification, some suggest that Findings suggest that the length $2.3 billion a year in Bangladesh. It these benefits may be overstated. of daily power outages has a strong emphasizes that merely providing Using data from Bangladesh, a re- negative impact on almost all devel- connections is not enough to spur cent study by Samad and Zhang offers opment outcomes considered. For economic growth, especially where the explanations as to why the findings example, a one-hour increase in daily quality of electricity supply is poor. have not been conclusive. First, the power outages for a household is as- This issue has attracted growing in- study shows that the benefits of being sociated with a 5.9 percent increase in terest—as reflected, for example, in connected depend on the reliability of kerosene consumption and a 0.3 per- the Multi-Tier Framework, a tool for electricity supply. Poor quality of sup- cent reduction in annual income. In monitoring and evaluating energy ac- ply could constrain and even discour- addition, the benefits of electrification cess developed by the Energy Sector age the productive use of electricity by increase over time, though at a declin- Management Assistance Program households and businesses. Second, ing rate. For example, each additional (ESMAP) at the World Bank in con- the study demonstrates that how long year of being connected to the grid is sultation with multiple development a household has been connected to associated with a 7.8 percent reduc- partners. The framework goes beyond electricity matters. While some basic tion in kerosene consumption and a 1 the binary measure of connection benefits such as more efficient light- percent increase in annual income or (yes/no) to also include attributes of ing can be achieved almost immedi- expenditure. And where supply is reli- electricity service and use such as reli- ately, other development benefits ac- able, estimates show, electrification ability, quality, and capacity—issues crue only over the long run. The study brings substantial benefits, increasing that need to be addressed in any dis- is based on data from a two-round income by 17 percent and expenditure cussion of electricity access and its panel survey in rural Bangladesh, car- by 12 percent. benefits. Overall, the study’s findings ried out in 2005 and 2010 and cover- The study also finds that women suggest that achieving long-term and ing 7,018 households. and girls are likely to benefit more inclusive benefits from electrification One complication in identifying the from rural electrification than men requires sustained efforts, attention causal relationship between electrifi- and boys do. After households gain to the quality of service, and comple- cation and household welfare is that access to electricity, girls’ daily study mentary policies to encourage the pro- grid expansion and households’ time increases by about 30 minutes ductive use of electricity, particularly decision to adopt electricity may not on average, while women’s labor force among poorer households. be random. For example, the govern- participation increases by 2.3 percent- ment may target electrification age points. In contrast, no statisti- projects to areas that are more easily cally significant effects are observed accessible and have greater growth for boys’ study time or men’s labor potential. And when electricity force participation. In addition, gain- Hussain Samad and Fan Zhang. 2017. “Hetero- becomes available in a village, better- ing access to electricity improves geneous Effects of Rural Electrification: Evidence off households are more likely to women’s empowerment (as measured from Bangladesh.” Policy Research Working Paper obtain grid connections first. To deal by decision-making abilities). 8102, World Bank, Washington, DC. 4 World Bank ResearchDigest Forecasting Electricity Demand in Developing Countries: Is There a Better Way? Forecasting electricity demand grows at some predetermined constant also compares them with the predic- is particularly challenging in growth rate. And electricity demand is tions based on heuristic methods. developing countries. A recent paper proportional to the GDP growth fore- The results show that the best- cast. Unsurprisingly, these heuristic performing time-series methods yield evaluates a new approach measures often yield very inaccurate highly accurate forecasts of electricity P projections of electricity demand. production in the majority of develop- ower cannot be easily stored. The paper argues that rather than ing countries. These forecasts also So underestimating demand depending on ad hoc predictions, en- prove to have much smaller errors for electricity leads to supply ergy practitioners in developing coun- than the predictions of two heuristic shortages and forced power outages, tries can rely on time-series econo- models that assume that electric- which have detrimental effects on pro- metric forecasting models, which ity production grows at an exog- ductivity and economic growth. But yield much more accurate forecasts of enous rate or is proportional to real overestimating demand may lead to electricity demand for the near decade GDP growth. overinvestment in generation capacity (that is, over a 1- to 10-year horizon). But the quality of the time-series and, ultimately, to higher electricity These models are particularly useful econometric forecasts diminishes for prices because utilities typically need for the needs of developing countries countries and regions where rapid to recover their investment costs to because they are less data-intensive economic and structural transforma- maintain financial viability. than microeconometric studies and tion makes it difficult to establish Accurate forecasts of electricity they are straightforward to imple- stable historical production trends. demand are thus critical in design- ment in most of the popular statistical These include low-income countries, ing least-cost generation plans for software packages (such as MATLAB, countries in Sub-Saharan Africa, coun- the power sector and in carrying out R, and Stata). They are also useful for tries with poor electricity access, and investment appraisals of power gen- validating long-term computational those with small electricity generation eration projects. They also matter in energy models with a forecast horizon systems. For these countries in par- designing national greenhouse gas of 30–50 years. Time-series economet- ticular, more accurate forecasts could mitigation strategies, because electric- ric models are thus important decision be produced through the collection of ity generated from fossil fuels is an support tools for investment planning higher-quality (ideally disaggregated) important source of greenhouse gas by national governments of develop- data and a more rigorous forecast- emissions. ing countries as well as by utilities, ing approach using a combination of But forecasting electricity demand international institutions, and private microeconometric and computational is difficult because it involves a range electricity generation companies. modeling methods. of uncertainties, including population To demonstrate this point, the pa- Another important direction for growth, changing technology, econom- per assesses the accuracy of a set of future research on forecasting electric- ic conditions, and prevailing weather the most popular time-series econo- ity demand in developing countries conditions (and the timing of those metric methods in forecasting electric- would involve more accurate predic- conditions). It can be particularly ity production—which, because of the tion of structural shifts. These include, challenging in developing countries, nonstorability and limited tradability for low-income countries, the transi- where data are often elusive, political of electricity, is a close and frequently tion from more primitive forms of influences are often brought to bear, the only available proxy for unob- energy (fuelwood, kerosene) toward and historical electricity demand can served electricity demand (or total fi- electricity and, for middle-income be more volatile because of macroeco- nal consumption) in many developing countries, the use of more electricity- nomic or political instability. countries. intensive appliances. As a result of these problems, in Drawing on the time-series econo- developing countries demand fore- metrics literature, the author first casting often receives less technical develops an econometric framework attention than other elements of for forecasting electricity production. power sector analysis. Indeed, policy He then obtains a number of out-of- makers and energy practitioners often sample electricity production forecasts rely on crude electricity demand fore- resulting from different econometric casts based on simple heuristic mea- methods and model specifications sures. According to a recent paper by based on historical time series for 106 Jevgenijs Steinbuks. 2017. “Assessing the Accu- Steinbuks, conversations with World developing countries over the period racy of Electricity Demand Forecasts in Developing Bank specialists suggest that among 1960–2012. Finally, the author evalu- Countries.” Policy Research Working Paper 7974, these “rules of thumb,” the two most ates the accuracy of these competing World Bank, Washington, DC. Also forthcoming popular are these: Electricity demand electricity production forecasts and in International Journal of Forecasting. World Bank Research Digest 5 Expanding Cross-Border Electricity Trade to Promote Hydropower in South Asia South Asia’s hydropower potential and Pakistan) by expanding cross- market would generate $289 billion remains mostly unexploited because border transmission capacity and over the 20-year period, assuming an of constraints on regional electricity developing regional electricity trade average price of $0.08 per kilowatt- agreements. hour (the 2018 price). This amount is trade. Relaxing these constraints In India, where domestic hydro- 10 times Nepal’s GDP in 2017. Similar would offer huge benefits power is growing more slowly than benefits would accrue to Bhutan and S total generation capacity, hydropower Afghanistan, though their electricity outh Asia has impressive poten- from neighboring countries, espe- export revenues would be smaller. tial for hydropower generation, cially Bhutan and Nepal, is needed to And thanks to imports of cheaper and more than 80 percent of it provide flexibility in the large power hydropower from their neighbors, has yet to be exploited. Utilizing this system. And the rapid increase in gen- Bangladesh, India, and Pakistan would hydropower potential would bring eration capacity based on intermittent benefit from lower power supply costs. enormous economic and environmen- resources (wind and solar) means an As noted, hydropower develop- tal benefits to the region—including even greater need for hydropower to ment would contribute to the ability growth opportunities through exports balance the power system. of Bangladesh, India, and Pakistan of surplus hydropower, greater flex- Differences across the region’s to meet targets under the Paris cli- ibility and lower supply costs in power countries in monthly or seasonal mate agreement. Under the scenario systems, and reductions in green- load profiles provide an opportunity of unconstrained regional electricity house gas emissions and local air to maximize the utilization of hydro- trade, carbon dioxide (CO2) emissions pollution. power plants, sharing reserve margins in the power sector would be lower Why does such a large share of the and thereby lowering electricity supply relative to the corresponding baseline region’s hydropower potential remain costs. In addition, regional electric- in all three countries in 2030: 43 per- unexploited? Among the main reasons ity trade would help countries with cent lower in Bangladesh, 9 percent are the limited provisions for trading “peaky” loads (such as Nepal, where lower in India, and 3 percent lower in electricity across borders. Regional peak lighting and evening cooking co- Pakistan. The corresponding reduction electricity trade is limited to small incide) either to avoid the expansion for the region as a whole would be volumes of bilateral trade between of peak load capacities or to improve 11 percent. India and its eastern and northern the capacity utilization of peak load The findings of the study should be neighbors (Bangladesh, Bhutan, and plants that have already been built in interpreted with care. The estimates Nepal). The capacity of cross-border the absence of cross-border trade. of hydropower development and the transmission interconnections is only Finally, because hydropower is corresponding economic and envi- a tenth of the potential hydropower among the economic options for re- ronmental benefits over the next two that could be traded. ducing greenhouse gas emissions, its decades indicate only the potential. In a recent study Timilsina exam- development would help Bangladesh, And this potential will not be real- ines the importance of enhancing re- India, and Pakistan meet their green- ized in the absence of strong political gional electricity trade to promote hy- house gas mitigation targets under the will and commitment and without dropower in South Asia and estimates Paris climate agreement. urgent action on concrete plans and the potential of hydropower develop- What would a scenario of un- strategies. ment and trade under different scenar- constrained cross-border electricity ios. The author also identifies several trade mean for the development of factors that are critical to promoting hydropower and its economic and hydropower in the region—and shows environmental benefits in the region? that developing this resource through The author estimates that under this the expansion of cross-border electric- scenario hydropower capacity would ity trade would benefit all countries almost quadruple—from 64 gigawatts participating in regional electricity today to 241 in 2040. The region would trade and cooperation. save almost $100 billion in electric- The author’s analysis shows that ity supply costs over the next two Afghanistan, Bhutan, and Nepal would decades through the substitution of have surplus hydropower potential hydropower for fossil fuels. even under highly optimistic scenarios The monetary value of hydropower Govinda R. Timilsina. 2018. “How Would Cross- on the growth of their domestic elec- trade would be huge relative to GDP, Border Electricity Trade Stimulate Hydropower tricity demand. Utilizing this surplus especially for Nepal and Bhutan. Development in South Asia?” Policy Research capacity would require enhancing ac- For Nepal the unrestricted export of Working Paper 8513, World Bank, Washington, cess to markets (in Bangladesh, India, surplus hydropower to the regional DC. 6 World Bank ResearchDigest The Nexus of Energy Supply and Human Health Energy supply can lead to costly Despite these wide-ranging sources supply. This would call for fiscal and human health problems that need to of risk, however, studies investigating regulatory policy instruments that be considered in designing energy the links between the energy supply provide incentives to use energy ef- chain and human health are limited. ficiently and to switch to cleaner fuels. policies and strategies Drawing on the existing knowledge Despite being a major health concern E base, a paper by Timilsina discusses in developing countries, however, local nergy is essential to economic these links at different stages of the air pollution issues are overshadowed growth and human welfare. But energy supply chain and draws some by climate change agendas. Yet fos- through its production and use, policy implications. sil fuels and wood fuels that cause energy also has adverse effects on Air pollution gives rise to the most deforestation and forest degradation human health. This happens through prominent adverse effects of energy are key contributors to both global three channels: accidents, occupa- on human health. Two forms pose the climate change and local air pollution. tional exposure to harmful substances greatest risks: outdoor or ambient air Policies that encourage substitution of (such as coal dust in mines and power pollution and indoor or household air cleaner fuels can therefore help in ad- plants), and air pollution. pollution. While fossil fuel combus- dressing both problems. Energy mining and production tion in power plants and road trans- For accidents and occupational activities, such as coal mining and portation is the main source of ambi- exposure, the other two channels for oil and gas drilling, are a source of ent air pollution, use of biomass fuels the human health impacts of energy, accidents as well as long-term occu- and coal for cooking is the primary coal is the primary source in the en- pational exposure of miners to harm- source of household air pollution. ergy sector. Most accidents in the ful substances or pollutants. Energy Ambient and household air pollution coal sector occur in mining, mainly transformation activities, such as is estimated to cause more than 6 mil- underground explosions and flooding. power generation and charcoal pro- lion premature deaths each year. The Explosions of oil and gas wells, pipe- duction, involve health risks through most vulnerable population is in the lines, and tankers are another major all three channels—accidents, occupa- developing world, where more than 90 source of accidents in the energy sup- tional exposure, and air pollution. And percent of premature deaths related ply chain. Nuclear power, though per- energy consumption activities, such to air pollution occur, primarily among ceived as the energy source posing the as the use of biomass fuels and coal poor people. biggest threat to people, has caused for cooking, cause health risks through Women, children, and the elderly the fewest fatalities among the energy air pollution, especially for women suffer the most from air pollution— carriers. and children and mostly in developing and specifically from household air Overall in the energy sector, more countries. pollution. More than 60 percent of than 95 percent of accidents occur in The types of health impacts vary all premature deaths attributed to developing countries. Poor house- across different stages of the energy household air pollution occur among holds are particularly vulnerable to supply chain. Hazardous substances women and children, who face greater accidents as well as occupational ex- released from energy mining opera- exposure to indoor smoke from cook- posure in the energy industry because tions (lead, arsenic, cadmium, mer- ing and home heating. The risks from they provide much of the workforce cury, chromium) damage the nervous, household air pollution are again engaged in mining coal, driving oil respiratory, and immune systems greatest for people in developing tankers, drilling for oil and gas, and and cause cancers and cardiovas- countries—because of their reli- hauling coal in power plants. Reducing cular disease. Oxides of sulfur and ance on traditional biomass fuels for the human health impact from ac- nitrogen released from the burning of cooking and home heating, the main cidents and occupational exposure in coal in power plants cause asthma, source of air pollution in households. the energy supply chain will require cardiac disease, ischemic stroke, and In India alone, more than a million strengthening and rigorously enforcing chronic obstructive pulmonary dis- premature deaths occur annually as safety standards. ease. Harmful air pollutants such as a result of indoor air pollution. The suspended particulate matter, carbon impact falls mostly on low-income monoxide, polyaromatic hydrocar- households, which lack access to clean bons, and volatile organic compounds fuels for cooking and must therefore produced by the burning of fossil fuels rely on traditional biomass fuels. in power generation and transporta- The large scale of these impacts tion, and of biomass fuels and coal suggests that controlling air pollution, Govinda R. Timilsina. 2017. “The Nexus of En- in household cooking, all cause lung specifically particulate matter, should ergy Supply and Human Health.” Policy Research cancer and cardiovascular and respira- be the main policy focus in address- Working Paper 8129, World Bank, Washington, tory disease. ing the human health effects of energy DC. World Bank Research Digest 7 Can Improved Biomass Stoves Be Cost-Effective in Mitigating Climate Change? The results of a choice experiment data were collected through a survey household. At the lower price of vol- survey in Ethiopia suggest that of 504 randomly selected households untary carbon emission reductions improved biomass stoves can be in 36 randomly selected villages in the of around $6.00 a ton, these annual regional states of Amhara, Oromiya, savings would be worth about $65 mil- cost-effectively incorporated into a and SNNP (Southern Nations, lion, or about $5.70 per household. carbon payment program Nationalities, and Peoples). Survey respondents indicated a N The findings show that respon- willingness to pay about $8 more on early 40 percent of people dents prefer improved stoves that average for a stove reducing fuelwood around the world, and 68–90 have high levels of durability, reduce use by 25 percent. The price of the percent in Sub-Saharan cooking time, use less fuel, and pro- Mirt stove is about $12. Thus a carbon African countries, rely on solid fuels duce less smoke. Households with payment equal to just the carbon sav- for cooking—particularly biomass fewer than three children value each ings from a single year could more such as wood and agricultural resi- of these characteristics more than than close the gap between the mar- dues. Traditional cooking methods do households with more than three ginal willingness to pay and the price are harmful both to the environment children. And households that are of this stove. and to human health, with inefficient less than 50 percent female care more But the carbon savings would be stoves leading to greater deforestation about cooking time, fuel reduction, realized only if the stoves continue to and greenhouse gas emissions (both and smoke reduction than do those be consistently used, and the moni- contributors to climate change) as that are majority female. toring and verification required in well as to respiratory health issues. The findings also show that the coupling carbon payments to the use Without a major change in policy, biggest barrier to the adoption of of an improved stove would be chal- the number of people relying on tra- improved stoves in Ethiopia is their lenging. One feasible option could ditional biomass fuels is expected availability. In addition, those who be a program that gives part of an to remain about the same in 2030 as have seen an improved stove before annual carbon payment directly to today. Thus alongside efforts to move place a higher value on them. This households (say, $4 a year) and uses to alternative fuel sources, improving suggests the potential importance of another part (say, $2 a year) to edu- environmental and health outcomes improving the availability of improved cate households about the benefits of will also require replacing traditional stoves in rural Ethiopia as well as in- the stoves and to verify their use (per- stoves with improved ones that have creasing awareness of the stoves and haps by hiring and training community greater fuel efficiency and lower their benefits. members). At the end of three years, emissions. Finally, the study analyzed the cost- after the full value of the stoves has To effectively promote the adop- effectiveness of improved stoves for been covered, the full carbon payment tion of improved stoves, policy makers greenhouse gas mitigation under a could be provided to households as need to know what characteristics of carbon payment program—a program an added incentive to continue using improved stoves are most important providing results-based payments the stoves. to users, how much people are willing to developing countries for actions to pay for them, and what factors con- to reduce carbon emissions (such as strain the adoption of such stoves—all an REDD+ program). The analysis is issues on which there is limited infor- based on the Mirt stove, which is used mation. To fill this gap in knowledge, to cook the traditional Ethiopian flat- a team of authors conducted a choice bread. This stove leads to fuelwood experiment study in rural Ethiopia. savings in the field of 25 percent, or There, about 96 percent of the popula- 634 kilograms per household per year, tion relies on wood, charcoal, branch- which translates into a reduction in es, dung, and agricultural residues annual carbon dioxide (CO2) emis- for household energy, all of which sions of 0.94 tons per stove. produce smoke and harmful emissions On this basis the Ethiopian govern- when burned. ment’s ongoing initiative to distribute Sahan T. M. Dissanayake, Abebe Damte Beyene, The study consisted of a choice ex- 11.45 million stoves would lead to a Randall Bluffstone, Zenebe Gebreegziabher, Gilbert Kiggundu, Shannon H. Kooser, Peter Martins- periment in which respondents chose reduction in annual CO2 emissions of son, Alemu Mekonnen, and Michael Toman. between possible hypothetical stoves about 10.77 million tons. Based on the 2018. “Improved Biomass Cook Stoves for Climate based on their identified characteris- 2015 price of $13.93 a ton from carbon Change Mitigation? Evidence of Preferences, tics. The characteristics of improved allowance auctions in California, these Willingness to Pay, and Carbon Savings.” Policy stoves for the study were identified annual carbon savings would be worth Research Working Paper 8499, World Bank, through multiple focus groups. The about $150 million, or about $13 per Washington, DC. 8 World Bank ResearchDigest Can Peer Pressure Encourage People to Adopt Energy-Saving Technologies? Laboratory experiments show that a success (paying out the maximum fully aware that their peers’ informa- people may adopt innovations more reward possible). tion (and therefore actions) is irrel- quickly after seeing peers do so, Subjects receive imperfect signals evant for their own payouts. about the quality of C in every round. In looking at which technology is even if the choice of technology Thus they face a trade-off between chosen, the paper finds evidence of differs waiting—and in choosing to wait, a modest intrinsic preference for B T obtaining only a small reward while (the safe one). Beyond that, subjects he diffusion and adoption of new doing so but thereby receiving more rationally use their private and (in technologies, whether physical information for making a choice—and the treatment where relevant) public or virtual, is central to develop- reaping immediate expected benefits information, placing relatively greater ment and growth. This prompts some of upgrading but possibly making the weight on the private signals. So questions for policy makers: What wrong choice. The environment there- the second main result is that while drives people to accept or reject novel fore mimics a case in which the energy theoretically the uncorrelated treat- approaches or products? What infor- efficiency gap is due not to a lack of ment should look identical to the solo mation do they pay attention to (or information or desire but to behavior- treatment, in practice it looks almost ignore) when making these decisions? al factors such as “regret aversion” and identical to the perfectly correlated How much do they care about what hesitancy about being wrong. treatment. This turns out to be a good others are doing? Is there a trade-off The paper considers three treat- thing: even though subjects do not between taking the fully optimal ac- ments for the group dimension. In the imitate their peers’ actions, they do tion and simply taking some action? first, subjects observe nothing about imitate the simple fact that their peers What, if anything, can be done to en- their peers, so the choice becomes an have made a decision, partially over- courage socially beneficial behavior? individual decision problem. In the coming the temptation to delay. These issues come into play in second, subjects observe their peers’ The laboratory conditions, though many areas, from agriculture to edu- decisions after each round (though stylized, capture several important cation to business management. But not the private signals), but the value real-world dimensions of the diffusion one area where they are particularly of C is entirely independent (uncor- of energy technologies: uncertainty relevant, and where climate change related) from one to the other, so about benefits, observation of public suggests that much is at stake, in- conceptually this treatment should and private information, and freedom volves energy and the environment. In be no different from the first. In the to choose the timing. To the extent a paradox that scientists and natural third, they again observe the others’ that behavioral obstacles seem to be resource economists refer to as the en- decisions, but now everyone has the causing welfare-reducing delays, the ergy efficiency gap, growing evidence has same underlying value for C so there is results suggest that “nudges” may shown that both consumers and firms potential to learn useful information be able to improve outcomes more routinely fail to adopt new energy- from observation. Thus at each stage effectively—and probably less expen- conserving processes and products— in this third treatment (until making sively—than regulation or subsidies. even when those technologies yield the irrevocable investment decision) Spreading the word about early or discounted cost savings relative to the participants receive both public and lead adopters, while being somewhat status quo. A recent paper by Jamison, private information and must decide agnostic about the specific decisions, Owens, and Woroch uses laboratory how to proceed. This endogenous may encourage others to at least break experiments to further explore and choice of timing is unusual in the free of their inertia and improve on the disentangle these issues. literature on herding behavior and status quo. In the experiments each partici- signaling. pant plays the role of an agent (for The experiments lead to two main example, a citizen or a manager) and results. The first is that in all treat- is placed in a group of six. All are ini- ments subjects on average wait longer tially assigned the status quo technol- than is theoretically optimal to make a ogy (A), which pays out a small reward choice between B and C. This delayed in every round. In every round they adoption is consistent with the previ- either choose to stay with A or decide ously observed energy efficiency gap to irreversibly upgrade to technology B and may be due to regret aversion, risk or C; that is, the decision, once made, attitudes, status quo bias, or another Julian C. Jamison, David Owens, and Glenn is final. B is a safe new technology behavioral factor. Interestingly, this Woroch. 2017. “Social Learning about Environ- that pays out a larger reward. C is a delay is reduced in both the treat- mental Innovations: Experimental Analysis of novel but risky technology that will be ments with peer observation, even Adoption Timing.” Policy Research Working Paper either a failure (paying out nothing) or though in one of them subjects are 7955, World Bank, Washington, DC. World Bank Research Digest 9 Solar Lanterns and Youth Academic Performance in Bangladesh Providing solar lanterns to children smaller lanterns; others received only control groups, this potential learning in unelectrified households boosted a main high-capacity solar device; and spillover did not explain the finding of their study time in the short term— the rest served as a control group and no academic improvement. received no devices. The results of this experiment but not their academic performance Compared with the control house- might suggest that introducing simple M holds, those that received three solar solar products in other modestly dis- ost low-income countries are lanterns reduced their kerosene ex- advantaged settings cannot be expect- on track to achieve universal penditure by about 75 percent over the ed to have substantial effects on learn- access to primary education. 12 months of 2013/14, which in turn ing outcomes. The experiment was But gains in access have not necessar- reduced total household expenditure conducted in a setting where the pro- ily led to gains in learning outcomes. by 6 percent (about $63), while those vision of solar lanterns was expected This has prompted researchers to that received one reduced their kero- to have the greatest positive impacts: explore alternative interventions to sene spending by about 50 percent Access to electricity was limited in the improve student learning. Recent and their total household spending by study area. And the study distributed studies involving randomized con- 2 percent (about $21). solar lanterns directly to students, trol trials have provided valuable The provision of solar lanterns had which might have given them prior- evidence about the effectiveness of a significant effect on home-study ity over other household members in such interventions as remedial tutors, time. Between September 2013 and the use of the lanterns. Moreover, the computer-assisted learning, incen- April 2014 students receiving lanterns replacement of kerosene lamps with tives for teachers, and reductions in reduced daytime study by about 5–8 solar lanterns, by reducing indoor air class size. minutes each day and increased night- pollution, could potentially improve Empirical research has paid less time study by 20–25 minutes, a net ef- students’ health and thus their aca- attention to the after-school home- fect equivalent to an increase of 45–75 demic performance. study environment. One potentially hours of home-study time over the Yet the study area had many other critical factor in this environment is eight months. The treated students constraints on academic performance, adequate nighttime lighting, because also almost completely substituted such as in the availability of good many students in developing coun- solar lanterns for kerosene lamps for teachers and in households’ ability to tries are expected to work during the their nighttime study. The increase in afford complementary education (like day to help their family. Yet achiev- net study time was more evident in private tutoring, which can be crucial ing universal access to electricity has November 2013, just before the annual for home learning support, especially been a challenge for many low-income examination season of December, dur- for students with illiterate parents). countries. ing which home-study time increased Thus it is possible that providing so- Where access to electricity is lim- by about 20 hours. These findings hold lar lanterns for students could boost ited, can solar lanterns help improve true whether students received three learning outcomes where the students children’s academic performance? solar lanterns or only one. have better educational environments In a recent study Kudo, Shonchoy, Moreover, the school attendance than those in the study area. This may and Takahashi used a randomized rate initially increased as a result of depend both on the overall quality of control trial to evaluate this ques- the provision of solar lanterns, though educational inputs—such as teaching, tion on the river islands of northern this effect did not last long. If only the school facilities, the home learning Bangladesh. The study provided solar cost of the solar lanterns is consid- environment, and access to light—and lanterns under a free-leasing agree- ered, and with the life of a solar lan- on the complementarity across those ment to a randomly selected subset tern assumed to be five years, this in- inputs. of the target population—students tervention raised student participation enrolled in grades four through by about 0.6 years per $100 spent. eight—for 16 months (September But the increase in study time and 2013–December 2014). initial improvement in school atten- The authors selected 882 sample dance did not translate into improve- children, all from unelectrified house- ment in the academic performance holds, at 17 local schools and imple- of treated children as measured by mented within-grade randomization schools’ final examination scores in at each school. To differentiate the December 2013 and 2014. While non- Yuya Kudo, Abu S. Shonchoy, and Kazushi Taka- treatment intensity, they divided the treated students could have learned hashi. 2017. “Can Solar Lanterns Improve Youth students into three groups: some of from their treated peers, minimizing Academic Performance? Experimental Evidence the students received both a main the difference in academic perfor- from Bangladesh.” World Bank Economic Re- high-capacity solar device and two mance between the treatment and view, https://doi.org/10.1093/wber/lhw073. 10World Bank ResearchDigest The Complexities of Integrating Renewable Energy into a Power System Integrating wind and solar into a These more technically focused stud- cheap domestic gas and relies increas- power system requires a planning ies are devoid of economics. They are ingly on imports for both primary fuels approach that fully considers the essentially “feasibility checks” of par- and electricity. Solar could help allevi- ticular power system plans, including ate the impending domestic energy variability of these resources those with a high share of renewable crisis, especially because its cost is M resources. They provide no insight dropping at a time that Bangladesh’s any developing countries into how much of the mix should be supply costs are on the rise. The au- have become interested in renewables, however. thors find that the economic potential integrating wind and solar Third, planning models of different of solar could be somewhere between into their national power grids. These vintages—from those of the 1970s to 1,000 and 4,200 MW (between 6 per- renewable energy technologies can modern planning tools—that con- cent and 26 percent of peak demand) reduce reliance on imported fuels sider economic and technical issues by 2025, depending on the cost of so- while avoiding emissions of carbon jointly. Least-cost planning studies, lar, the price of gas, and the cost of fi- dioxide and local air pollutants, and for example, consider renewables as nancing. If renewable generation costs their investment costs have dropped part of the portfolio of generation re- drop below 6.5 cents per kilowatt- sharply. But the variability of wind and sources. Key challenges in using these hour, renewables could account for up solar introduces complex technical models are to determine what level of to 4,200 MW of generation capacity, and economic issues into power sys- detail can realistically be modeled and though this would require massive tem planning. how to coordinate among a hierarchy improvements in operational prac- Much work on renewables has of long- and short-term models. A tices to maintain system security. A been done over the past 20 years— modern planning model that can fully more realistic target is in the range of especially since 2012, when large- capture the variability of renewables 2,400–2,900 MW (about 15–18 percent scale, grid-connected wind and solar across all time frames can help deter- of peak demand). In extreme cases, in generation began to quickly ramp up mine the optimal mix of renewables which the poor operational practices in many countries around the world. and nonrenewables—one that allows continue or gas prices remain low for To ensure optimal integration of large- renewables to fill a critical niche in the the next few years, the potential might scale wind and solar, however, the system without jeopardizing system be limited to 1,000 MW. methodology needs to be standard- security. The study demonstrated the impor- ized and rigorous case studies need to In a recent study Bankuti, tance, in developing a clear pathway be carried out, especially for develop- Chattopadhyay, and Song undertook for renewables, of bringing together ing countries. such an exercise for Bangladesh, the economic and technical issues A look at the methodology of stud- which had been contemplating a through a systems approach and un- ies on integrating renewables sug- 500-megawatt (MW) solar project dertaking rigorous analysis. It also had gests that many of those from the since 2011. The authors developed practical outcomes for Bangladesh. past two decades fall in one of three the first comprehensive renewable The study developed a sound basis categories. integration study for the country, com- for a plan to integrate renewables First, simple “back of the envelope” bining a suite of planning models that and provided inputs for the prepara- analyses that rely on comparison of deal with long-term capacity expan- tion of the country’s first major solar the levelized cost of electricity—a sion, a detailed dispatch model, and project. It also transferred its model measure averaged over the lifetime of a transmission power flow analysis. to the System Planning Division of a power generation asset that fails to The proposed solar project, estimated the Bangladesh Power Development capture the variability of renewable to cost $2.76 billion, is an expensive Board, which has since used the resources across different time frames proposition for a country struggling model to update the country’s power (from subhourly to seasonal and in- to find good baseload generation op- system master plan. terannual). Many such studies have tions. Moreover, the authors’ analysis been undertaken to assess policies or of past operational practices shows investment decisions. While they fo- that lack of economic considerations cus on the economics of renewables, in power generation scheduling during they ignore nearly all the technical most peak hours poses a great chal- aspects that are important in power lenge in accommodating large-scale, system planning. grid-connected solar generation. Miklos Bankuti, Debabrata Chattopadhyay, and Second, detailed production cost- Yet solar might play an important Chong-Suk Song. 2018. “Integrating Variable ing or transmission power flow mod- part in the country’s power system. Renewable Energy in the Bangladesh Power Sys- els that look at operational aspects of Bangladesh is undergoing a major tem: A Planning Analysis.” Policy Research Work- generation or transmission systems. change in its fuel mix as it runs out of ing Paper 8517, World Bank, Washington, DC. World Bank Research Digest11 Interactions of Renewable Energy Standards with Carbon Emission Trading for Climate Change Mitigation Different policy measures to reduce The study uses a computable gen- cost of achieving the greenhouse gas greenhouse gas emissions can eral equilibrium (CGE) model repre- mitigation targets and ultimately lead overlap in ways that may dilute the senting the behavior of all economic to higher costs for the economy and agents—households, governments, the society. efficacy of each policy industries, and international trade— Moreover, the additional RET man- C for all 30 administrative regions in date would reduce the demand for ountries around the world are China. Because the model captures emission permits in the production introducing policies for green- the links between production sectors sectors trading the permits, causing house gas mitigation. These as well as between economic agents, the carbon market to shrink. It would include renewable energy technology it can measure impacts on all sectors reduce the volume of emission per- (RET) mandates (such as renewable and agents of a policy shock anywhere mits traded by 1.5 percent and their energy requirements for electricity in the economy. price by 3.5 percent. This would result supply); energy efficiency standards The model assumes a national ETS in a decrease in the transfer of funds (such as for buildings and automo- that allows the administrative regions associated with emission trading biles); fiscal policy incentives (such as in China to trade their carbon dioxide between regions as well as between subsidies and tax exemptions); and (CO2) emission allowances, creating sectors. Indeed, the study shows that carbon pricing instruments (such as incentives for regions where reduc- the transfer of funds from the central carbon taxes, emission offset mecha- ing CO2 emissions is less costly to to the western region under the ETS nisms, and emission cap and trade produce fewer emissions than their would drop by 29 percent because of systems). allowances and then to sell the sur- the effect of the RET mandate on the The member states of the European plus emission reduction to regions size of the ETS market. Union have introduced policies of all where reducing CO2 emissions is more Combining an RET policy with an these types. With the aim of meet- costly. The ETS would thus reduce the ETS would, however, lead to a greater ing pledges under the Paris climate overall cost of meeting specified emis- reduction in fossil fuel consumption agreement, so have some developing sion reduction targets. But a separate than the ETS policy alone—and thus countries. China, for example, has policy introduced earlier in China, an to a greater reduction in local air pol- been implementing fiscal and regula- RET mandate requiring electricity sup- lution. Given the importance of this tory policies aimed at greenhouse gas ply companies to rely on renewable in China, a model accounting for the mitigation for more than a decade. energy sources for a certain fraction benefits of the lower pollution might In 2017 the country also introduced of their electricity sales, would reduce show that a combined ETS and RET a national emission trading system the effectiveness of the ETS in two policy would be more economic than (ETS). Today the ETS covers only elec- ways: It would reduce the size of the the ETS policy alone. But quantifying tricity generation, but in the future it emission trading market. And it would these benefits in each of China’s 30 is expected to also cover other sources reduce the price of tradable emis- administrative regions is an enormous of greenhouse gas emissions, such as sion permits (or reduce the emissions task and beyond the scope of the manufacturing industries. traded in the market). study. Additional policy instruments Economic theory suggests that The study shows that to achieve a can also be justified where an emis- the most efficient policy mechanism 10 percent reduction of CO2 emissions sion trading scheme does not capture for achieving greenhouse gas mitiga- in China from the base case (the situ- all potential sources of emissions tion targets is an appropriate pricing ation in the absence of the ETS), a na- (such as the household sector), which instrument. Simply getting the pricing tional ETS would cause slight losses is often the case. right encourages economic agents in GDP and welfare. If as part of the to change technologies and behav- strategy to achieve that target a sepa- iors in the desired ways, avoiding the rate RET mandate was introduced on need for separate policies to promote top of the ETS, this would exacerbate adoption of efficient and clean tech- the GDP and welfare losses. The rea- nologies. Indeed, in many instances, son is that while an ETS creates incen- introducing policies with overlapping tives for the market to implement the objectives could dilute the efficacy cheapest CO2 mitigation options, the of each policy and thus increase the RET mandate forces the power gen- Ying Fan, Jie Wu, Govinda R. Timilsina, and Yan overall cost of achieving a green- eration industry to divert resources Xia. 2017. “Understanding the Interactions be- house gas mitigation target. A recent to implementing renewable energy tween Emissions Trading Systems and Renewable study by Fan, Wu, Timilsina, and Xia technologies, no matter how costly Energy Standards Using a Multi-Regional CGE on China demonstrates this insight they are. This diversion of resources Model of China.” Policy Research Working Paper empirically. to costlier options would increase the 8159, World Bank, Washington, DC. 12World Bank ResearchDigest Recent Policy Research Working Papers on Energy 7889 Benefits of Electrification and the Role of 8130 The Impact of Electricity Shortages on Firm 8460 Taking Stock of the Impact of Power Utility Reliability: Evidence from India Productivity: Evidence from Pakistan Reform in Developing Countries: A Literature Hussain Samad and Fan Zhang Corbett A. Grainger and Fan Zhang Review 7891 Shedding Light: Understanding Energy 8131 Whose Power Gets Cut? Using High- Robert Bacon Efficiency and Electricity Reliability Frequency Satellite Images to Measure 8461 Taking Stock of Economic Regulation of Eliana Carranza and Robyn Meeks Power Supply Irregularity Power Utilities in the Developing World: A 7912 Residential Electricity Subsidies in Pakistan: Brian Min, Zachary O’Keeffe, and Fan Zhang Literature Review Targeting, Welfare Impacts, and Options for 8145 Productivity in the Non-Oil Sector in Nigeria: Martín Rodríguez Pardina and Julieta Schiro Reform Firm-Level Evidence 8468 How Much Has Nepal Lost in the Last Thomas Walker, Ezgi Canpolat, Farah Khalid Santiago Herrera and Wilfried Kouamé Decade Due to Load Shedding? An Economic Khan, and Adea Kryeziu 8147 Measuring Time Use in Development Assessment Using a CGE Model 7957 Impact of Oil Price Fluctuations on Financial Settings Govinda R. Timilsina, Prakash Sapkota, and Markets since 2014 Greg Seymour, Hazel Malapit, and Agnes Jevgenijs Steinbuks Ha Nguyen, Huong Nguyen, and Anh Pham Quisumbing 8477 Are Driving Forces of CO2 Emissions Different 7983 Special Study on Benchmarking the Quality of 8153 Natural Resources, Institutions, and across Countries? Insights from Identity and Project Economic Analysis for the South Asia Economic Growth: The Case of Nigeria Econometric Analyses Region Anna K. Raggl Kangyin Dong, Gal Hochman, and Govinda R. Kene Ezemenari and Xiao Ye 8211 Firm-Level Decomposition of Energy Timilsina 7987 Rising Incomes and Inequality of Access Consumption in Turkish Manufacturing 8493 Where Is the Carbon Tax after Thirty Years to Infrastructure among Latin American Industry of Research? Households Sebnem Sahin Govinda R. Timilsina Marianne Fay and Stephane Straub 8247 Putting Your Money Where Your Mouth Is: 8495 The Role of Major Emerging Markets in 7991 Regional Dimensions of Recent Weakness in Geographic Targeting of World Bank Projects Global Commodity Demand Investment: Drivers, Investment Needs, and to the Bottom 40 Percent John Baffes, Alain Kabundi, Peter Nagle, and Policy Responses Hannes Öhler, Mario Negre, Lodewijk Smets, Franziska Ohnsorge Ekaterine Vashakmadze, Gerard Kambou, Renzo Massari, and Željko Bogetic 8518 Taking Stock of the Political Economy Derek Chen, Boaz Nandwa, Yohei Okawa, 8280 Quasi-Fiscal Deficits in the Electricity Sector of Power Sector Reforms in Developing and Dana Vorisek of the Middle East and North Africa: Sources Countries: A Literature Review 8008 Challenges in Assessing the Costs of and Size Alan David Lee and Zainab Usman Household Cooking Energy in Lower-Income Daniel Camos, Antonio Estache, and 8519 Taking Stock of Wholesale Power Markets in Countries Mohamad M. Hamid Developing Countries: A Literature Review Michael Toman and Randall Bluffstone 8285 Public Infrastructure and Structural Hugh Rudnick and Constantin Velasquez 8070 Fuel Efficiency versus Fuel Substitution in the Transformation 8582 Electrification and Household Welfare: Transport Sector: An Econometric Analysis Fidel Perez Sebastian and Jevgenijs Evidence from Pakistan Gal Hochman and Govinda R. Timilsina Steinbuks Hussain Samad and Fan Zhang 8076 Energy Prices and International Trade: 8287 Cost Recovery and Financial Viability of the 8648 Forest Carbon Supply in Nepal: Evidence Incorporating Input-Output Linkages Power Sector in Developing Countries: A from a Choice Experiment H. Ron Chan, Edward Manderson, and Fan Literature Review Sahan Dissanayake, Randall Bluffstone, Zhang Joern Huenteler, Istvan Dobozi, Ani E. Somanathan, Harisharan Luintel, N. S. 8077 Does Energy Efficiency Promote Economic Balabanyan, and Sudeshna Ghosh Banerjee Paudel, and Michael Toman Growth? Evidence from a Multi-Country and 8385 Obstacles on the Road to Palestinian 8677 How Much Would Bangladesh Gain from Multi-Sector Panel Data Set Economic Growth the Removal of Subsidies on Electricity and Ashish Rajbhandari and Fan Zhang Roy van der Weide, Bob Rijkers, Brian Natural Gas? 8103 Scaling Up Distributed Solar in Emerging Blankespoor, and Alexei Abrahams Govinda R. Timilsina, Sheoli Pargal, Marinos Markets: The Case of the Arab Republic of 8392 Paris Climate Agreement and the Global Tsigas, and Sebnem Sahin Egypt Economy: Winners and Losers 8680 Green Sukuk, Energy Poverty, and Climate Dalia Sakr, Joern Huenteler, Tyeler Matsuo, Muthukumara Mani, Zekarias Hussein, Badri Change: A Roadmap for Sub-Saharan Africa and Ashish Khanna Narayanan Gopalakrishnan, and Deepika Dalal Aassouli, Mehmet Asutay, Mahmoud 8124 Understanding the Impact of Windstorms Wadhwa Mohieldin, and Tochukwu Chiara Nwokike on Economic Activity from Night Lights in 8415 Jobs! Electricity Shortages and Central America Unemployment in Africa Oscar A. Ishizawa, Juan Jose Miranda, and Justice Tei Mensah Hongrui Zhang Papers can be downloaded at www.worldbank.org/dec. The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the World Bank’s Editorial Committee: Shanta Devarajan, David tion disseminating findings of World Bank research. Research Committee and managed by DECSO, the Rosenblatt, Aslı Demirgüç-Kunt, Roberta Gatti, The views and interpretations in the articles are those Strategy and Operations unit of the Development Hiau Looi Kee, Chorching Goh, and Shiva S. Makki of the authors and do not necessarily represent the Economics Senior Vice Presidency (DEC). The Research (managing editor). Editor: Alison Strong; production: views of the World Bank, its Executive Directors, or the Digest is not copyrighted and may be reproduced with Roula Yazigi; workflow: Alejandra Ramon. For free countries they represent. appropriate source attribution. subscriptions email research@worldbank.org or visit http://www.worldbank.org/researchdigest. The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper