Environmental and Energy
April 14-15, 2016
James B. Bushnell, University of California at Davis and NBER; Stephen P. Holland, University of North Carolina at Greensboro and NBER; Jonathan E. Hughes, University of Colorado at Boulder; and Christopher R. Knittel, MIT and NBER
The EPA's Clean Power Plan sets state-level 2030 goals for CO2 emission rate reductions that vary substantially across states. States can choose the regulatory mechanism they use and whether or not to join with other states in implementing their goals. Bushnell, Holland, Hughes, and Knittel analyze incentives to adopt rate standards versus cap-and-trade with theory and simulation. They show conditions where adoption of inefficient rate standards is a dominant strategy from both consumers' and generators' perspectives. Numerical simulations of the Western electricity system highlight incentives for uncoordinated policies that lower welfare and increase emissions relative to coordination.
Robin Burgess, London School of Economics; Francisco JM. Costa, Getulio Vargas Foundation (FGV/EPGE); and Benjamin A. Olken, MIT and NBER
Tropical deforestation is one of the major drivers of climate change. Much of this loss is due to illegal logging. Unlike forests in the Congo basin and South-East Asia, the world's largest tropical forest the Amazon has experienced a dramatic slowing in rates of deforestation over the last decade. The bulk of the Amazon is located in Brazil which has introduced a raft of policies to reduce illegal logging in recent years. Burgess, Costa, and Olken use Brazil's border with its neighbors to identify the impact of Brazilian policies on deforestation. Because forests are a fixed resource and geography and infrastructure vary continuously over the border, the researchers can compare annual forest loss on either side of the border to tease out the impact of national forest policies from other drivers of deforestation. To do this they employ a satellite-derived data set that measures forest cover at a 30 x 30 meter resolution for the entire Amazon area across the 2000-2014 period. Their data reveals a sharp discontinuity at the border in 2000 Amazonian pixels on the Brazilian side of the border are more likely to have been deforested and between 2001 and 2005 annual forest loss in Brazil was around four times the rate on the other side of the border. However, in 2006, just after the Brazilian government introduced a raft of policies to curtail illegal logging, these differences disappear and Brazilian rates of forest loss fall to those observed across the border. These results demonstrate the power of the state to affect whether or not natural resources are conserved or exploited even in the furthest reaches of the Amazonian jungle.
William A. Pizer, Duke University and NBER, and Brian C. Prest, Duke University
This paper considers the effect of uncertainty on the comparative advantage of price-based regulation over intertemporally tradable quantity regulation when policies are updated. In particular, Pizer and Prest examine the case where uncertainty about costs and benefits unfolds over time, and an information asymmetry implies that information known to the firm in one period does not influence government policy until the next period. Under this setup quantity regulation can achieve the first-best outcome and is always preferred to a price instrument when the government sets both policies to maximize expected net benefits. With quantity trading over time, the firm's intertemporal optimization implies current prices equal discounted expected future prices. This allows the government to circumvent the information asymmetry through policy updates that influence future expected prices and, in turn, current prices. No such opportunity exists with price regulation. The researchers also consider the possibility that policy updates are driven in part by political "noise" rather than true values of costs and benefits. Here, they find that price regulation will be preferred if the variance of shocks in the political noise process exceeds the variance of the true shocks to marginal costs and benefits. Applied to climate change or other applications where marginal benefits are flat, this condition for preferring price regulation simplifies to whether noise shocks have higher variance than benefit shocks, and cost uncertainty does not matter. All of these results sharply contrast with the Weitzman (1974) results that the relative slopes of marginal costs and benefits determine the comparative advantage of prices versus quantities and that benefit uncertainty by itself does not matter.
Jun Yang, Peking University; Antung A. Liu, Indiana University; Ping Qin, Renmin University of China; and Joshua Linn, Resources for the Future
To reduce pervasive problems of traffic congestion and air pollution, many cities in developing countries have considered restricting vehicle ownership. There is no empirical evidence on these programs' efficacy and costs, but other prior work suggests that not having a car increases the cost of commuting and limits the set of job opportunities. However, these prior studies do not address the endogeneity of car ownership. Yang, Liu, Qin, and Linn leverage a unique policy, the Beijing license plate lottery, to estimate the effect of restricting vehicles on distance traveled and commuting time, while addressing the endogeneity of car ownership. They find that adding a car has little impact on total distance traveled or time spent traveling, but a large impact on mode of travel. While reducing car ownership by 20% and car miles by 10% in Beijing, this policy has not added significantly to overall distances traveled or commute times.
Shaun McRae and Robyn Meeks, University of Michigan
Increasing block tariffs for electricity and water provide a subsidy for low users and a conservation incentive for high users. However, their effectiveness may depend on both how well consumers understand the nonlinear structure and how attentive they are to consumption. McRae and Meeks develop a novel price elicitation instrument to recover these components of price perceptions and apply the instrument to an electricity tariff introduced in Kyrgyzstan in late 2014. They document considerable heterogeneity in understanding of the new price structure. The households with the best understanding of the tariff had the largest reduction in their electricity consumption. However, this effect was driven by those households who were inattentive to their own position on the price schedule. The results suggest that providing consumer-specific information about the effect of new nonlinear tariffs might enhance their political acceptability.
Koichiro Ito, University of Chicago and NBER, and Shuang Zhang, University of Colorado Boulder
This paper provides among the first revealed preference estimate of willingness to pay (WTP) for clean air in developing countries. Ito and Zhang use product-by-store level transaction data on air purifier sales in Chinese cities and city-level air pollution data. Their empirical strategy leverages the Huai River heating policy, which created discontinuous quasi-experimental variation in air pollution between the north and south of the river. Using a spatial regression discontinuity design, the researchers estimate the marginal willingness to pay for removing 1 `(ug)/m^3P M_10`. Their findings provide important policy implications for optimal environmental regulation.
Joseph E. Aldy, Harvard University and NBER; Todd Gerarden, Harvard University; and Richard Sweeney, Boston College
Aldy, Gerarden, and Sweeney examine the choice between using capital and output subsidies to promote wind energy in the United States. In this sector, some subsidies support upfront investment while others reward output. The researchers exploit a natural experiment in which wind farm developers were unexpectedly given the opportunity to choose between these two options in order to estimate the differential impact of these subsidies on project productivity. Using matching and fuzzy regression discontinuity designs, they find that wind farms choosing the capital subsidy produce 8 to 14 percent less electricity per unit of capacity than wind farms selecting the output subsidy and that this effect is driven by incentives generated by these subsidies rather than selection. They then use these estimates to evaluate the public economics of U.S. wind energy subsidies. Preliminary results suggest the Federal government paid 17 to 20 percent more per unit of output from wind farms receiving capital subsidies than they would have paid under the existing output subsidy.
Kenneth Lee, University of California at Berkeley, and Edward Miguel and Catherine Wolfram, University of California at Berkeley and NBER
Lee, Miguel, and Wolfram present results from an experiment that randomized the expansion of electrical grid infrastructure in rural Kenya. Electricity distribution is the canonical example of a natural monopoly. Randomized price offers show that demand for electricity connections falls sharply with price. Experimental variation in the number of connections combined with administrative costs reveals considerable scale economies, as hypothesized. Consumer surplus, however, is less than total costs at all price levels, suggesting that household electrification may reduce social welfare. The researchers discuss how leakage, reduced demand (due to red tape, low reliability, and credit constraints), and possible spillovers may impact this conclusion.
Frank A. Wolak, Stanford University and NBER
Wolak formulates and estimates a household-level, billing-cycle water demand model under increasing block prices that accounts for the impact of monthly weather variation, the amount of vegetation on the household's property, and customer-level heterogeneity in demand due to household demographics. The model utilizes U.S. Census data on the distribution of household demographics in the utility's service territory to recover the impact of these factors on water demand. An index of the amount of vegetation on the household's property is obtained from NASA satellite data. The household-level demand models are used to compute the distribution of utility-level water demand and revenues for any possible price schedule. Knowledge of the structure of customer-level demand can be used by the utility to design nonlinear pricing plans that achieve competing revenue or water conservation goals, which is crucial for water utilities to manage increasingly uncertain water availability yet still remain financially viable. Knowledge of how these demands differ across customers based on observable household characteristics can allow the utility to reduce the utility-wide revenue or sales risk it faces for any pricing plan. Knowledge of how the structure of demand varies across customers can be used to design personalized (based on observable household demographic characteristics) increasing block price schedules to further reduce the risk the utility faces on a system-wide basis. For the utilities considered, knowledge of the customer-level demographics that predict demand differences across households reduces the uncertainty in the utility's system-wide revenues from 70 to 96 percent. Further reductions in the uncertainty in the utility's system-wide revenues, in the range of 5 to 15 percent, are possible by re-designing the utility's nonlinear price schedules to minimize the revenue risk it faces given the distribution of household-level demand in its service territory.
Solomon M. Hsiang, University of California at Berkeley and NBER, and Nitin Sekar, Princeton University
Avraham Ebenstein, Hebrew University of Jerusalem; Michael Greenstone, University of Chicago and NBER; Maoyong Fan, Ball State University; Guojun He, The Hong Kong University of Science and Technology; and Maigeng Zhou, Centers for Disease Control and Prevention