The people of Washington state have the opportunity this November to vote on a state-wide carbon tax. If it passes, Washington will be the first state in the country to enact a tax on carbon emissions. One of the consequences of the tax would be the impact on consumer costs. In Initiative 732, the revenues will be recycled to reduce sales taxes and provide a rebate for low-income households. In our research, a key finding is that the way revenues from a carbon tax are used is even more important than the level of the tax with respect to the impact on households (see our description of the proposal and a discussion of the distributional impacts).
One of the potentially important ways that a carbon tax impacts consumers is how the tax affects electricity prices. Washington has one of the cleanest electricity systems in the country, and some of the lowest electricity prices in the nation. The No on 732 campaign has released an infographic asserting that under I-732 the “Price of electricity [will increase] by 20 percent in 2020.” This figure is based on a table in a related report showing that the average price of electricity in 2020 will rise from 9 cents to 10.8 cents per kWh of consumption (20 percent).
While it is impossible to fully evaluate this claim because neither the report authors nor the No on 732 campaign have released the economic modeling underlying these results, a basic analysis of the electricity sector in Washington suggests that these estimates are implausible.
Washington’s Department of Commerce reports that the fuel mix in the state is 15 percent coal and 11 percent natural gas, with a strong trend toward reducing coal use going forward. (This energy mix accounts for the role of imported power to meet consumer demand in the state.) By 2020 the tax will not have reached $30 per ton of CO2 but we will cautiously take this value to consider the maximum possible impact the tax could have on electricity prices. Roughly speaking, I estimate that on average a carbon tax of $30 per ton of CO2 would raise the cost of electricity generation by 3 cents per kWh for coal and 1.2 cents per kWh for natural gas. Generation that has no CO2 emissions would have no increase in cost, such as the hydro, wind, solar and nuclear that currently constitute 74 percent of the generation serving consumers in Washington. Therefore, the average change in electricity generation costs from a $30 carbon tax would be 0.58 cents per kWh, compared to the average retail electricity price of 9 cents per kWh. This would result in an increase in the average cost of electricity of 6.4 percent. This is far less than (about one-third) the 1.8 cents per kWh claimed by the No on 732 campaign.
The Department of Commerce maintains a Carbon Tax Analysis Model that provides rough estimates of changes in the economy under carbon pricing. That model provides an estimate that is similar, but less than mine. That model suggests a $30 carbon tax in 2020 would increase the average retail electricity rate by 5.3 percent.
However, because of the way that electricity markets work, the change in electricity prices could be greater than this estimate based on average generation costs. This extra cost to consumers would come as a benefit to producers, as explained below.
As further background, if the electricity system in Washington were a competitive market, the marginal generation provider would determine the cost each hour of the year. If the marginal generator is a natural gas unit for all 8,760 hours of the year, then a $30 per ton CO2 tax would introduce a change in the cost of electricity of 1.2 cents per kWh, or 13 percent. This implausible outcome points to the greatest conceivable change in prices, still much less than the 20 percent increase claimed by No on 732. However in fact, Washington’s electricity system is a hybrid with cost-of-service regulation (prices based on the average cost of service rather than marginal cost, as described above) that interacts with quasi-public agencies and independent (unregulated) generators. Where exchange occurs in a power market it is characterized predominantly by bilateral exchange, not exchange within a well-organized power market. These factors imply the change in electricity cost will be well below what would be expected in a competitive market.
Further, many hours of the year fossil generation is not at the margin. In practice, natural gas generators that currently provide only 11 percent of the state’s total electricity needs are often not the marginal generator. For many hours of the year, renewables that provide 74 percent of total generation are sufficient to fulfill demand. Hydro-electric power is storable behind the dam and, subject to operating constraints, it can provide marginal generation. Also, a large portion of power does not enter the market at all but is priced through long-term power purchase agreements to cover renewable investments, or on an average cost basis for public/regulated utilities. In summary, I expect the change in electricity price to be close to the average change in generation cost reported above, that is about 6 to 7 percent of the average price of electricity.