Energy Policy Tradeoffs between Economic Efficiency and Distributional Equity

April 15-16, 2016
Tatyana Deryugina and Don Fullerton, both of University of Illinois, and Research Associate William A. Pizer of Duke University, Organizers

Chris Bruegge, Stanford University; Tatyana Deryugina; and Erica Myers, University of Illinois

The Distributional Effects of Building Codes

Building energy codes have been used for close to 40 years to improve the energy efficiency of newly constructed and renovated buildings. Over this time period, almost all U.S. states have adopted and periodically increased the stringency of statewide energy codes. However, building codes are not as efficient as pricing the externalities of energy use directly. For example, an energy tax will create incentives to not only improve the building stock but to also reduce the amount of energy people choose to consume, conditional on the energy efficiency of the building. However, such taxes are often regressive, making them politically unpopular and therefore difficult to implement. Whether building energy codes themselves are regressive or not is unclear. For example, the savings from building codes requiring greater energy efficiency may be larger for richer households, because their houses tend to be bigger and they may use more energy overall. However, even if the monetary savings are larger for richer households in absolute terms, the savings relative to income may be much larger for the poor. Bruegge, Deryugina, and Myers consider the distributional consequences of building energy codes. To do so, they will use publicly available electricity usage data from households in California, combined with housing transactions and building permit data, which will allow them to determine the building energy code regime associated with each building. The researchers will use block-group-level Census data to approximate the income of each household in their sample. Following earlier literature, they will compare the energy usage of houses built just before and just after building code changes, breaking the analysis up by income to estimate the distributional consequences of building codes. In addition, the researchers will also compare building code costs and capitalization into housing prices across income groups.


Sébastien Houde, University of Maryland, and Joseph Aldy, Harvard University and NBER

Efficiency and Distributional Effects of Energy Efficiency Subsidies for Appliances

To promote residential energy efficiency, federal, state, and local governments have relied on an array of fiscal policy instruments to subsidize energy-efficient appliances. Houde and Aldy investigate the impacts of the federal manufacturers tax credit, state and utility rebates, and sales tax holidays as well as electricity prices on consumers' decisions to purchase energy-efficient appliances. This analysis permits an evaluation of how households trade-off the purchase price of more fuel-efficient appliances with the lifetime operating costs given local electricity prices. They find significant heterogeneity in the response to the various policy instruments across the income distribution (and future work will test other socio-demographic measures). The estimated model can be used to examine various policy counterfactuals, including a comparison of a carbon tax to appliance rebates to sales tax holidays, in terms of their impacts on energy-efficient appliance purchases and expected energy consumption in aggregate and across various socio-demographic characteristics.


Mar Reguant, Northwestern University and NBER

The Distributional Impacts of Renewable Generation Policies


Stephen P. Holland, University of North Carolina at Greensboro and NBER; Erin T. Mansur, Dartmouth College and NBER; Nicholas Z. Muller, Middlebury College and NBER; and Andrew J. Yates, University of North Carolina at Chapel Hill

Egalitarian Vehicles? Distributional Effects of Electric Vehicle Driving and Purchase Subsidies


Arik Levinson, Georgetown University and NBER

Are Energy Efficiency Standards Less Regressive Than Energy Taxes?


Lucas Davis, University of California at Berkeley and NBER, and Christopher R. Knittel, MIT and NBER

How Regressive Are Fuel Economy Standards?

Don Fullerton, and Steven Sexton, Duke University

Energy Tax Rebates and Redistribution

Economists commonly assert that pollution tax burdens can be offset by transfers, and thus efficiency and equity considerations can be separated. But how well can transfers offset burdens? One concern is vertical equity, the fair distribution of burdens up and down the income scale, but another problem is horizontal equity, fair treatment of those with the same income. Within an income group, households vary in their energy use, tax liability, and transfer participation. Income-targeted transfers would undercompensate some and overcompensate others. This paper will assess the capacity of existing transfer mechanisms to achieve vertical and horizontal equity following the imposition of an energy tax.
Fullerton and Sexton use the U.S. Treasury's merged file of 300,000 tax returns plus 20,000 non-filer "information returns." First, these returns will be matched to a similar family in the Consumer Expenditure Survey (CEX), and the resulting data will be used to impute transfer program participation and receipts. These data will be augmented with existing estimates for effects of energy taxes on the market price of each consumption good. Second, these price effects will be applied to the detailed expenditures of each family in the Treasury's merged dataset to calculate the burden on each of 320,000 families to characterize both vertical and horizontal distributions of burdens. Third, the estimates will be used to assess how each energy tax burden is offset by changes to transfer programs, showing net gains or losses to each income group and within each group. Fourth, the researchers consider other proposed transfers that might better target benefits to pooamilies that bear more than average energy tax burdens. If a transfer mechanism can prevent extreme or capricious burdens, then policymakers can take advantage of the efficiency afforded by market mechanisms like taxes that minimize the cost of reducing energy use.


Carolyn Fischer, Resources for the Future, and Billy Pizer

Equity versus Efficiency in Energy Regulation

Absent transfers Pigouvian pricing policies can involve redistributions that are many times the net gain or loss to society. Burtraw and Palmer (2008) find that a pricing policy has social costs of roughly $0.5 billion annually, while consumers and producers lose more than $21 billion in payments to the government. Among firms, the aggregate loss is $3 billion, but some facilities gain $6 billion while others lose $9 billion. In contrast, performance standards involve smaller redistributions but lead to higher social costs. If distributional effects matter and are not addressed, the efficiency advantage of pricing policies is less compelling. Moreover, beliefs about individual and net social benefits could be lower still. Even if policymakers intend to address distributional concerns to avoid inequity, public support for the program could falter.
In this paper, Fischer and Pizer will construct a simple social welfare function (SWF) where society cares about both the overall cost of a policy and the inequality of its burdens. A single parameter will be used to capture the tradeoff between overall cost and inequality, measured here as inequality in the change to household incomes from the policy. The SWF will then be applied to a set of alternative policies, drawn in part from other papers in this proposal, which describe alternative cost and distributional effects. By comparing pricing policies to other forms of regulation, the study can identify the level of inequality aversion required to overturn the efficiency case for pricing policies. This comparison can be completed for price policies with and without associated redistribution through transfers. The research will highlight four points. First, social welfare is reduced by energy policies with unequal burdens across households, so the welfare cost may be larger for pricing policies. Second, efforts to correct these effects can reduce but not eliminate these costs. Third, if households do not believe the transfers will occur, their evaluation of pricing policies will differ from that of policymakers and reflect lower expected welfare. Fourth, the creation of large losers under pricing policies can lead to more vocal opposition. Uncertainty about who loses and who wins can make it harder to address the inequity and more likely that expected winners oppose the policy. Taken together, these points emphasize that without considering distributional consequences, economic analyses of various policy alternatives risk excluding important effects on overall welfare.