In articles here at IER and at other sites, I have previously explained some of the “unintended consequences” of popular environmental regulations. A new paper from Resources for the Future (RFF) documents another quirk of popular regulations that will irk their supporters: Under reasonable assumptions, rate-based regulations such as a Renewable Portfolio Standard or a vehicle fuel economy standard can lead to an increase in pollution when there is an improvement in the “green” technology. I’ll summarize the main results to show the fun of economics but also the dangers in top-down government edicts.

Summarizing the New Perversity

The authors of the RFF study, John Horowitz and Joshua Linn, review the existing literature and do not find anyone else making the specific point of their paper. I quote liberally from their Introduction to let them explain their findings in their own words:

A large number of environmental regulations are based on proportions or averages of “clean” and “polluting” sources, a regulatory approach we call rate-based. Three prominent examples are renewable portfolio standards (RPS), a regulation many states use to determine the share of renewables in total electricity generation; emissions rate standards (ERS), which set rates of emissions per unit of electricity generation, such as the US Clean Power Plan governing greenhouse gas emissions from electricity generation; and fuel economy standards for passenger vehicles, such as the US Corporate Average Fuel Economy (CAFE) standards.

Despite the popularity of this class of policy, the economics of rate-based approaches have not been well developed. In this short paper, we show a serious potential problem with these regulations. In a nutshell, technological advances that reduce the cost of clean energy will increase use of the dirty energy source whenever the regulation is fixed and binding. Since environmental damage depends on the level of dirty energy and not its proportion, these otherwise desirable clean technology innovations result in additional pollution.

This is an unexpected outcome; improvements in clean technology should lead to pollution reduction. This result strengthens the case—as if any such strengthening were needed—for bare bones carbon taxes or cap and trade; that is, a uniform carbon price not accompanied by other inefficient regulations.

I have included the RFF authors’ advocacy of a carbon tax so as not to misrepresent their overall views. However, their study does not present a case for carbon taxes per se. Instead, they merely show that a carbon tax does not suffer from the particular perversity that they document with some other (popular) regulatory approaches.

Thus, while I agree with the RFF authors that these particular regulations contain unintended consequences, I would go further and say that the same general phenomenon applies to a carbon tax as well. (For a full critique of a U.S. carbon tax, see my recent Cato study co-authored with climate scientists Pat Michaels and Paul Knappenberger.)

Seeing the Problem in the Case of an RPS

The easiest way to see the problem uncovered by the RFF authors is to study the case of a renewable portfolio standard (RPS), in which the government mandates that a certain percentage of electricity be supplied by a qualifying source (such as wind).

The RFF study proves the result rigorously in the context of a two-fuel model, but for our purposes I’ll supply the verbal intuition. So long as the RPS regulation is “binding,” meaning that it forces electricity producers to deviate from what profit-and-loss calculations alone would lead them to do, then it must be the case that the proportion of total wind energy is constant, namely at the minimum level mandated by the RPS.

To see the problem, consider an initial equilibrium in which electricity rates paid by consumers are adequate to ration their demand to the total supply offered by wind and other producers, after taking account of the implicit taxes and subsidies resulting from the RPS credits. (In other words, wind producers receive an implicit subsidy to encourage their production, while natural gas and coal producers pay an implicit penalty.)

Now, suppose that that initial equilibrium is disturbed because of an advance in wind technology. At the existing retail price of electricity and the state of the RPS credit market, wind producers now want to expand production. Eventually things settle down into a new equilibrium, where the retail price of electricity has fallen, consumers are buying a larger quantity of electricity, and wind producers are selling more.

However—and here is the key insight—if the RPS is still binding, meaning that the wind sector is still dependent on the mandate, then the output from non-wind producers must have risen correspondingly. For example, if the RPS mandates 20 percent of the electricity be produced from wind, then any expansion of wind production must go hand-in-hand with an expansion of electricity produced from natural gas and/or coal. This must be the case because of simple math: If wind is 20 percent of the total, both before and after the innovation, then an increase in wind output (measured in MWh) must be matched by a proportional increase from non-wind output (measured in MWh).

The specific mechanism for achieving this outcome economically is that the innovation in wind technology initially causes output to exceed the RPS mandate. The price for RPS credits thus falls, reducing the implicit tax on non-wind and the implicit subsidy to wind. The lower penalty on natural gas and coal allows them to expand output, even though they didn’t directly benefit from the technological improvement. When things settle down in the new equilibrium, there is higher output from wind and non-wind electricity producers, the RPS credit price is lower (but still positive), the RPS mandate is still satisfied at the bare minimum, and consumers enjoy lower unit price for electricity at which they buy the larger quantity being produced.

The RFF authors walk through this example to underscore the perversity of the outcome, from the perspective of those who favored the RPS in the first place. Consider: The whole point of the RPS is to handicap carbon-dioxide-emitting sources in order to artificially boost the production of zero-emission sources, because left to its own wind cannot currently compete in the marketplace.

It thus seems to violate the spirit of the program, when an advance in wind power—which brings wind closer to competitiveness with natural gas and coal—leads to an expansion of electricity produced by fossil fuels. Because concerns over climate change and conventional air pollution are tied to the absolute level of emissions (not their proportion relative to wind production), we have the paradoxical result that an improvement in wind technology leads to higher emissions and hence a worse environmental outcome, from the perspective of the supporters of the RPS.

Seeing the Problem in the Case of Fuel Economy Standards

Now that we have seen the perverse result in the case of an RPS, we can quickly explain a similar result for fuel economy standards. Here again, the problem is that the regulation mandates a minimum average across the set of new vehicles. Therefore, so long as the regulation is binding—meaning that manufacturers just satisfy the requirement—then an innovation that lowers the production cost of high-mileage vehicles will lead to an increase in total vehicle emissions.

We can see the logic by a simplistic example (which is mine) involving only two vehicle types: The expensive Model A gets 50 miles per gallon, while the cheaper Model B gets 30 miles per gallon. The government mandates that on average the entire fleet of vehicles achieves a fuel economy of 40 miles per gallon. In this simplistic setting, manufacturers have to adjust prices such that consumers want to purchase an equal number of Models A and B. To be sure, this outcome will be a distortion of the free-market outcome: The price of Model A will be artificially reduced, while the price of Model B will be artificially increased. The mandate will force the manufacturers to implicitly subsidize their production of high-mileage vehicles on the backs of the low-mileage vehicles.

Now in this setting, suppose there is a technological innovation, so that the manufacturers figure out how to produce the high-mileage Model A vehicles more cheaply. This rightward shift in the supply curve allows them to sell more Model As, of course, and the price of Model As falls to reflect the improved technology.

However, so long as the government fuel economy mandate is still binding, it must be the case in the new equilibrium that the sale of Model B’s has also risen one-for-one with the high-mileage model. This is obvious, because in the new equilibrium it still must be the case that half of the new cars sold are the 50mpg Model A, while the other half are the 30mpg Model B. If the sales of Model As go up, then sales of Model Bs must similarly increase.

Again, once we know the result from math, we can extend it to the underlying economics. As the Model A becomes easier to build, the government mandate imposes less of a distortion, compared to the free-market outcome. After the innovation, the manufacturers are implicitly paying a lower penalty on their Model Bs (and are enjoying a lower “subsidy” for their Model As). They therefore expand their production of Model Bs, bringing it closer to what would be the case in the absence of the regulation.

Although such an outcome is good news to those who care about consumers and don’t lose sleep over carbon dioxide emissions, the regulation’s supporters will be disheartened. As in the earlier section, here too the result is perverse: An innovation that makes it cheaper to build 50mpg vehicles has paradoxically led auto manufacturers to increase their sales of 30mpg vehicles. Because prices on both models fall, consumers are buying more of both models. Absent a change in motorists’ preferences, total vehicle-miles will definitely increase. And because average fuel economy is still locked in at 40mpg, it must be that total gasoline consumption—and hence emissions—increases.

In the case of fuel economy standards, there is an added complexity because a fall in the price of new vehicles can speed the retirement of older vehicles to which the mandate does not apply. This consideration can mitigate (or possibly even reverse) the result, but nonetheless the logic above is a real factor that the proponents of a fuel economy standard have not previously considered.

To be sure, we at IER have written in the past that federal fuel economy standards inefficiently impose preferences on the car buying public, forcing them to obtain higher mileage and less of other desirables attributes (such as vehicle size and comfort, as well as affordability). Yet the RFF authors have outlined a separate problem with such top-down regulations, even from the perspective of people who are very concerned about climate change.


A new RFF study outlines a perverse result of a particular class of popular government regulations designed to promote zero-emission energy: Because these regulations mandate minimum percentages of the total mix, a technological innovation in the “green” energy source can ironically lead to an increase in output from the other sources, thus increasing total emissions. This outcome has (apparently) not been well studied before in the literature.

It is true that the RFF authors interpret their new results as yet another reason to support a straightforward carbon tax, rather than more command-and-control interventions into the energy sector. However, the RFF study doesn’t directly make the case for a carbon tax. Rather, it simply shows that this particular unintended consequence does not apply to a carbon tax.

I endorse the analysis of the RFF authors, but would place it in the broader context of the counterproductive side effects that result from government intervention into energy markets, period. There are serious problems with a carbon tax of its own. As the RFF study has shown, despite good intentions and smart policy wonks, government regulations have a history of failure and leading to perverse incentives.

Print Friendly, PDF & Email