Environmental and Energy Economics
March 28 and 29, 2013
Maureen Cropper, University of Maryland and NBER; Kabir Malik, University of Maryland; Alexander Limonov, Resources for the Future; and Anoop Singh, Indian Institute of Technology (Kanpur)
Cropper, Malik, Limonov, and Singh examine the impact of electricity sector restructuring on the operating efficiency of coal-fired power plants in India. Between 1995 and 2009, 85 percent of coal-based generation capacity owned by state governments was unbundled from vertically integrated State Electricity Boards into state generating companies. The researchers find that generating units in states that unbundled before the Electricity Act of 2003 experienced reductions in forced outages of about 25 percent and improvements in availability of about 10 percent,, with the largest results occurring 3-5 years after restructuring. They find no evidence of improvements in thermal efficiency at state-owned power plants attributable to reform.
Martin Weitzman, Harvard University and NBER
The choice of an overall discount rate for climate change investments depends critically on how different components of investment payoffs are discounted at differing rates reflecting their underlying risk characteristics. Such underlying rates can vary enormously, from about 1 percent for idiosyncratic diversifiable risk to about 7 percent for systemic non-diversifiable risk. Which risk-adjusted rate is chosen can have a huge impact on cost. In this paper, Weitzman attempts to set forth in accessible language with a simple model what he thinks are some of the basic issues involved in discounting climate risks. The prototype application is calculating the social cost of carbon.
Joshua Graff Zivin, University of California, San Diego and NBER; Matthew Kotchen, Yale University and NBER; and Erin Mansur, Dartmouth College and NBER
Graff Zivin, Kotchen, and Mansur develop a methodology for estimating marginal emissions of electricity demand that vary by location and time of day across the United States. The approach takes account of the generation mix within interconnected electricity markets and shifting load profiles throughout the day. Using data available for 2007 through 2009, with a focus on carbon dioxide (CO2), they find substantial variation among locations and times of day. Marginal emission rates are more than three times as large in the upper Midwest than in the western United States, and within regions, rates for some hours of the day are more than twice those for others. They apply their results to an evaluation of plug-in electric vehicles (PEVs). The CO2 emissions per mile from driving PEVs are less than those from driving a hybrid car in the western United States and Texas. In the upper Midwest, however, charging during the recommended hours at night implies that PEVs generate more emissions per mile than the average car currently on the road. Underlying many of these results is a fundamental tension between electricity load management and environmental goals: the hours when electricity is the least expensive to produce tend to be the hours with the greatest emissions. In addition to PEVs, the authors show how their estimates are useful for evaluating the heterogeneous effects of other policies and initiatives, such as distributed solar, energy efficiency, and real-time pricing.
Timothy Beatty, University of Minnesota, and Jay Shimshack, Tulane University
Beatty and Shimshack use a large and representative database of multiple birth cohorts to study relationships between air pollution exposure and non-infant children's respiratory health outcomes. They observe several years of early-life health treatments for each of nearly 700,000 children. Three distinct research designs account for potential socioeconomic, behavioral, seasonal, and economic confounders. They find that marginal increases in carbon monoxide and ground-level ozone are associated with statistically significant increases in children's contemporaneous respiratory treatments. They also find that carbon monoxide exposure over the previous year has an effect on children's health that goes above and beyond contemporaneous exposure alone.
Cloe Garnache, University of California, Davis
Garnache examines the tradeoffs between the production of crops and the habitat for juvenile salmon, through flood events, on the Yolo Bypass floodplain. She investigates how changes in the fishery management institution affect the economic returns to fish habitat. To understand how habitat provision affects the economic surplus of the farmers and fishers, she develops a bio-economic model of Yolo Bypass agriculture, salmon population, and California ocean fishery. The results reveal large total producer surplus gains from improving habitat management and the natural resource management institution. In contrast with previous studies on open access resources, she finds that the gains from habitat management exceed those that arise from improving the management institution. Also, there are larger returns to habitat management when the management institution is suboptimal.
Christopher Knittel and Robert Pindyck, MIT and NBER
The price of crude oil in the United States had never exceeded $40 per barrel until mid-2004. By 2006 it reached $70 per barrel, and in July 2008 it reached a peak of $145. By the end of 2008 it had plummeted to about $30 before increasing again, reaching about $110 in 2011. Are "speculators" to blame for at least part of the volatility and sharp run-ups in price? Knittel and Pindyck clarify the potential and actual effects of speculators, and investors in general, on commodity prices. They focus on crude oil, but their approach can be applied to other commodities. They first address the question of what is meant by "oil price speculation," and how it relates to investments in oil reserves, oil inventories, or oil price derivatives (such as futures contracts). Next they outline the ways in which one could speculate on oil prices. Finally, they turn to the data, and calculate counterfactual prices that would have occurred from 1999 to 2012 in the absence of speculation. Their framework is based on a simple and transparent model of supply and demand in the cash and storage markets for a commodity. It lets them determine whether speculation as the driver of price changes is consistent with the data on production, consumption, inventory changes, and changes in convenience yields given reasonable elasticity assumptions. They show that speculation had little, if any, effect on prices and volatility.
Mark Jacobsen, University of California, San Diego and NBER, and Arthur van Benthem, University of Pennsylvania
Jacobsen and van Benthem estimate the sensitivity of scrap decisions to changes in used car values the "scrap elasticity" and show how it influences used car fleets under policies aimed at reducing gasoline use. Large scrap elasticities will tend to produce emissions leakage under efficiency standards as the longevity of used vehicles is increased, a process known as the Gruenspecht effect. To explore the magnitude of this leakage, they assemble a novel dataset of U.S. used vehicle registrations and prices, which they relate through time via differential effects in gasoline cost: a gasoline price increase or decrease of $1 alters the number of fuel-efficient versus fuel-inefficient vehicles scrapped by 18 percent. These relationships allow the authors to provide what they believe are the first estimates of the scrap elasticity itself, which they find to be about -0.7. When applied in a model of fuel economy standards, the elasticities they estimate suggest that 13-23 percent of the expected fuel savings will leak away through the used vehicle market. This considerably reduces the cost effectiveness of the standard, rivaling or exceeding the importance of the often-cited mileage "rebound" effect.
Eva Arceo-Gomez, CIDE; Rema Hanna, Harvard University and NBER; and Paulina Oliva, University of California, Santa Barbara and NBER
Much of what we know about the marginal effect of pollution on infant mortality is derived from developed country data. However, given the lower levels of air pollution in developed countries, these estimates may not be externally valid to the developing country context if there is a nonlinear dose relationship between pollution and mortality, or if the costs of avoidance behavior differ considerably between the two contexts. Arceo-Gomez, Hanna, and Oliva estimate the relationship between pollution and infant mortality using data from Mexico. They find that an increase of 1 part per billion in carbon monoxide (CO) over the previous week results in 0.0032 deaths per 100,000 births, while a 1 µg/m3 increase in particulate matter (PM10) results in 0.24 infant deaths per 100,000 births. The estimates for PM10 tend to be similar (or even smaller) than the U.S. estimates, while the findings on CO tend to be larger than those derived from the U.S. context. There is suggestive evidence that a non-linearity in the relationship between CO and health explains this difference.
Judson Boomhower, University of California, Berkeley, and Lucas Davis, University of California, Berkeley and NBER
Economists have long argued that many recipients of energy-efficiency subsidies may be "free riders," getting paid to do what they would have done anyway. Demonstrating this empirically has been difficult, however, because of endogeneity concerns and other challenges. Boomhower and Davis use a regression discontinuity analysis to examine participation in a large-scale appliance replacement program in Mexico. Comparing behavior just on either side of several eligibility thresholds, they find that program participation increases with larger subsidy amounts, but that the magnitude of the increase is small. For example, when an air-conditioner subsidy increases from $110 to $170, the number of participants increases by only 21 percent. The large fraction of infra-marginal households means that larger subsidy amounts are almost certainly not cost-effective. Overall, they find that accounting for free riders decreases the cost-effectiveness of the program by about 50 percent.
Emanuele Massetti and Robert Mendelsohn, Yale University, and Shun Chonabayashi, Cornell University
Massetti, Mendelsohn, and Chonabayashi test three functional forms that relate land values and temperatures in a Ricardian model of U.S. agriculture: a quadratic relationship based on average seasonal temperature and precipitations; a non-linear relationship based on degree days;and a flexible functional form in which average seasonal temperatures are interacted with dummies. Results obtained using growing season average temperature and degree days are not significantly different. They do not find evidence of a threshold if they include degree days above 34 °C. Cold degree days matter, however, and should not be omitted. Models that use a quadratic specification of average temperatures perform better than models that use degree days, which is in line with the agronomic literature. Degree days should be used only to estimate the duration of phenological events, not yields. Estimates of uniform +2 °C and +4 °C warming indicate that warming is significantly harmful for agriculture in the eastern United States. The relationship between temperatures and land values is flatter in a more flexible functional form than in the quadratic. Within and outside the growing season, seasons significantly affect land values and allow us to separate the beneficial and harmful effects of warming more effectively.
Matthew Kahn, University of California, Los Angeles and NBER, and Frank Wolak, Stanford University and NBER
Kahn and Wolak report on the results of two field experiments that examine the impact of providing current information on a household's electricity consumption on how that household responds to a nonlinear retail price schedule for electricity. Across the two utilities, over 2,000 households participated in a customized on-line interactive educational program that taught them how their monthly electricity bill was determined from a nonlinear retail pricing scheme they face. Each household also was told where their typical consumption monthly places them on this nonlinear pricing schedule. Households also were shown how changes in their major electricity-consuming activities would affect their monthly bill under the nonlinear pricing scheme. Using data from before and after this intervention for households that took the educational program (our treatment) and a randomly selected set of control households, the authors estimate the overall treatment effect associated with these educational programs, as well as a treatment effect for households on each specific pricing tier on the nonlinear price schedule during the pre-intervention period. For both utilities, they find that the overall impact of treatment is a reduction in the household's daily average consumption. In addition, the price tier-specific treatment effect results are: households that learn they face a high marginal price for consuming electricity reduce their electricity consumption; households that learn they face a low marginal price increase their electricity consumption. These results emphasize the need to provide timely and actionable information to households in order to maximize the effectiveness of nonlinear retail pricing schemes.