Thank you for the invitation to speak today. Many of the points I’ll be discussing today come from work I have co-authored with W. Kip Viscusi of Vanderbilt University. One paper is called “Overriding Consumer Preferences with Energy Regulations,” which was published in the Journal of Regulatory Economics. Another paper, which is currently a working paper, is called “Behavioral Public Choice: The Behavioral Paradox of Government Policy.”
In the first paper, Kip and I examine a recent wave of enacted and proposed US regulations that mandate energy efficiency standards for such things as light bulbs, appliances, and motor vehicles. Some of these regulations apply to consumer products and some to commercial products.
The purported motivation for these regulations is to reduce pollution. I am sympathetic to this motivation. Market prices for energy-intensive products can provide misleading signals to the extent that they do not account for the pollution costs stemming from the energy use.
But the best approach to addressing this problem is for the government to price these pollution costs directly. This would raise the cost of using pollution-intensive energy, which means consumers and businesses would then face the full cost of energy use, and markets would respond through some combination of new technologies, alternative fuels, and conservation.
Instead, we have these energy efficiency mandates, which are not cost-effective ways of reducing pollution. Indeed, by the agencies’ own analyses, the energy efficiency regulations have a negligible effect on reducing greenhouse gas emissions. In most instances, the agencies’ own analyses find that the environmental benefits are far outweighed by the costs of the regulations.
There are a number of explanations for why energy efficient mandates are not cost-effective: First, the one-size-fits-all energy efficiency mandates ignore the substantial diversity of preferences, financial resources, and personal situations that consumers and businesses must align in order to make decisions. Second, energy efficiency mandates do not promote conservation. Indeed, by lowering the cost of energy use, they provide an incentive to use more energy, reversing some of the desired energy savings. For example, an energy efficiency standard for air conditioners increases the incentive to run the air conditioners longer. And third, energy efficiency mandates must squeeze energy reductions out of new products only, and can even create incentives for consumers and businesses to retain older (and thus less energy-efficient) products.
Given the cost-ineffectiveness of energy efficiency mandates, how then do the agencies justify these regulations? They do so by deviating from standard economic practices for benefit-cost analyses.
Consider a simplified example of an energy-efficiency standard that removes less energy efficient products from the market place. The regulation effectively bans certain products from the marketplace. What are the consequences? On the benefit side, there is less pollution, although not as much as one would expect given the problems referred to earlier. Of course, one needs to know how these benefits of reducing the environmental externality compare to the costs in order to determine the optimal level of reduction.
Using standard benefit-cost analysis principles, one would view restricting consumer choice as a cost since less choice means less consumer surplus. However, under the agencies’ practices, restricting choice is seen as making consumers better off because it remedies irrational behavior. The agencies are presuming that consumers were causing themselves harm by buying these products. They presume that consumers are irrational and that regulators know better than consumers what they want. They provide little to no supporting evidence to substantiate this claim, and they do not document the magnitude of these behavioral anomalies. The agencies therefore can claim that the regulations are win/win: providing both an environmental benefit and a consumer protection benefit.
For the regulations that apply to commercial products (e.g., heavy duty trucks), the agencies are presuming that firms are forgoing profits, even though these are firms that operate with narrow profit margins and for which energy costs represent a substantial portion of operating expense.
The agencies are relying on engineering models that calculate the net present value of a set of possible energy-efficiency consumptions choices, which requires assumptions for such things as capital costs, current and future energy prices, duration and frequency of product use, and discount rates. Where their model indicates an optimal choice different from what consumers are choosing, the agencies do not consider the possibility that their model is wrong. Instead, they assume that their model better reflects the preferences of consumers and firms than the choices the consumers and firms would make for themselves.
There are many reasons why the problem could lie with the agencies’ models rather than with consumer and business irrationality. Net-present value studies focus on capital costs and operating costs and omit other relevant costs or benefits of the product to consumers that can drive the purchase decisions. For example, one study finds that manufacturing plants reject about half of the energy-efficiency projects recommended by engineering analyses because of unaccounted physical costs, risks, opportunity costs, lack of staff for analysis and implementation, risk of inconvenience to personnel, or suspected risk of problems with equipment. Another example: weatherization of a home can be a time-consuming and unpleasant task for the homeowner.
Also, consumers might not expect to receive as high a return in energy savings as the regulatory analyst assumes. For example, the engineering estimates of potential energy savings could misrepresent actual energy savings because they are based on highly controlled studies that do not directly apply to actual realized savings in a representative house. One study finds that the realized return to attic insulation falls short of the returns promised by engineers and product manufacturers. Finally, high discount rates can be rational in the presence of high sunk costs and uncertainty over future conservation savings, and in the presence of consumer illiquidity.
The standard approach to benefit-cost analysis focuses on the external costs of market actions and assumes that informed citizens are better able than regulators to make decisions about which products they value and which goods they should purchase given the substantial heterogeneity of preferences, financial resources, and personal situations. In their approach, the agencies have turned a cost (restricting choice) into a benefit (correcting irrationality). Their regulatory mandates are thus win/win.
What percentage of the agencies’ estimated benefits are due to this correction of presumed consumer and firm irrationality? For the fuel economy standards for passenger cars, 87 percent of the benefits stem from addressing purported consumer irrationality. Only about one percent of the benefits are for reductions of greenhouse gases for the US. Another six percent of the benefits are for reductions of greenhouse gases to other countries. The other five percent of the benefits are for energy security benefits and other environmental benefits.
For the fuel economy standards for heavy-duty trucks, 86 percent of the benefits stem from addressing purported irrationality. Only about one percent of the benefits are for reductions of greenhouse gases for the US. Another eight percent of the benefits are for reductions of greenhouse gases to other countries. The other five percent of the benefits are for energy security benefits.
For the mandates for clothes dryers, 79 percent of the benefits (assuming a three percent discount rate) stem from addressing purported consumer irrationality. Only about three percent of the benefits are for reductions of greenhouse gases for the US. Another eighteen percent of the benefits are for reductions of greenhouse gases to other countries.
For the mandates for room air conditioners, 70 percent of the benefits (assuming a three percent discount rate) stem from addressing purported consumer irrationality. Only about four percent of the benefits are for reductions of greenhouse gases for the US. Another 25 percent of the benefits are for reductions of greenhouse gases to other countries. The other one percent of the benefits is for other environmental benefits.
What are the problems with this approach of presuming consumers and firms (but not regulators) are irrational? For one, it shifts environmental regulatory policy away from the goal of mitigating the harm that people impose on others (through pollution) to a more paternalistic goal of mitigating the harm that people impose on themselves. The stated motivation for these regulations is to reduce pollution, but this is misleading because they really amount to consumer protection rather than environmental protection.
Another consequence is that we get less bang for the buck from our environmental regulations. For example, this approach puts greater weight on regulations that ban energy-inefficient products (and thus ignore the important role of heterogeneity of preferences throughout the market) than on regulations that raise the price of pollution (which account for heterogeneous tastes and more effectively reduce the environmental externality of pollution).
I think this approach also corrupts the use of behavioral economics, which is meant to find systematic deviations from conventional views of rational behavior and integrate them into economics models. Behavioral economics is not meant to assert irrationality and then justify any intervention. The name of Thaler and Sunstein’s seminal book is Nudge, not Compel, Force, or Mandate. The policy implication of behavioral economics is to use less intrusive approaches to regulation, such as information policies, to correct behavioral failings.
Indeed, EPA did try a nudge approach to fuel economy with their 2011 Motor Vehicle Fuel Economy Label rule, which mandated labels for all new cars to include overall mpg rating, a city mpg rating, a highway mpg rating, gallons/100 miles, driving range on a tank of gas, fuel costs in five years versus the average new vehicle, annual fuel costs, fuel economy and greenhouse-gas rating, and smog rating. The goal was to address the behavioral failings that people exhibit when evaluating their fuel economy choices. Yet the existence of this rule was ignored in the fuel economy mandate rule? Why? If the label rule was at all effective, then there’s less justification for the fuel economy mandate. If it was ineffective, then EPA was remiss in issuing it. Which is it?
More broadly, if the problem is consumers are making bad decisions due to behavioral biases, then the “Nudge” approach should be to provide information to help them make better decisions, not to restrict their choices. This is consistent with Executive Order 12866, which requires each agency to “identify and assess available alternatives to direct regulation…such as…providing information upon which choices can be made by the public.”
The energy efficiency regulations highlight a recent change in emphasis in economics more broadly, away from the traditional focus on justifying government interventions to correct market failures such as pollution externalities, and towards justifying government interventions to prevent self-harm arising through purported cognitive limitations and psychological biases.
As a bit of review, behavioral economists generally classify the deviations from standard economic assumptions into three categories: Imperfect optimization, such as the finding that people are less likely to participate in their employer’s retirement plan as the number of investment alternatives rise, thus suggesting that a government policy of limiting options could improve welfare. Or another study that finds that the salience of a sales tax (which differs depending on whether the tax is included in the sticker price or computed at the register) influences the behavior of consumers. Bounded self-control, such as the findings of procrastination and succumbing to immediate temptation, both of which can cause self-harm. Nonstandard preferences, such as the finding that people value a good differently depending on whether they were randomly endowed with the good, and also that people do not value losses and comparable gains symmetrically.
There are reasonable critiques of these behavioral studies, both from economists and from psychologists. But, by and large, they are carefully conducted empirical studies (albeit frequently in laboratory settings rather than the marketplace). My critique today is not with the empirical findings, but rather with the cavalier way in which the findings are used to justify government interventions. In the case of energy efficiency regulations, the agencies use behavioral economics to justify as a premise—meaning without any need to substantiate—that consumers and firms are irrational and thus in need of regulations to protect themselves from self-harm.
This highlights the need for a stronger literature on “behavioral public choice,” which recognizes that government policymakers and regulators are themselves human, so are subject to psychological biases and limitations as are other people. Most behavioral economics papers focus on the biases of ordinary people, and recommend government actions to remedy these biases, while ignoring that government officials are people too and thus subject to psychological forces. One study finds that, of the behavioral economics articles proposing paternalistic policy responses, 95.5% do not contain any analysis of the cognitive ability of policymakers. Cass Sunstein, to his credit, has acknowledged that “for every bias identified for individuals, there is an accompanying bias in the public sphere.” Perhaps not surprisingly, I haven’t found examples of regulatory impact analyses in which the agencies recognize the possibility of their own psychological biases.
Behavioral public choice also recognizes that policymakers are subject to public choice incentives that could further lead to inefficient policies, and indeed could even lead to the misuse of behavioral findings by the regulator in order to enhance regulatory control or to favor the influence of powerful special interest over the interest of public welfare.
There are some types of behavioral failings that could be more common in decision-making of government officials. One example is the focusing illusion, in which people fail to consider all relevant aspects of a particular problem, restricting their thoughts to salient situational elements. This could lead government officials to fail to see past the superficial effects of their policies. Another example is the intentions heuristic, in which people tend to judge a policy based on the intentions of its advocates rather than on the policy’s actual consequences. The implicit assumption is that good results follow from good intentions. Another example is overconfidence, or the so-called Dunning-Kruger effect, which is a cognitive bias that leads people with a superficial understanding of a subject to overestimate their competence and underestimate what they do not know.
The public choice literature suggests a number of ways in which incentives could lead public officials to act against the public’s interest. First, and perhaps most obviously, is that psychological failings in citizens would suggest bad decision making in their voting practices at least as much as in their market transactions. If anything, people have less incentive to behave rationally in their capacity as voters than in their capacity as market participants. In other words, in a democratic system, theory and evidence would suggest that government policies will reflect the irrationalities of ordinary people.
Second, public choice theory suggests that private decision makers have stronger incentives to acquire information – which costs time and money – to overcome behavioral biases, since the personal costs to a citizen who makes a bad decision are higher than the personal costs to the regulator of a rule that leads to a bad outcome for that citizen. Given the evidence that people with incentives can partially reduce cognitive biases through learning, and given that the costs of cognitive biases weigh more on the citizen than on the regulator, one should expect fewer errors among private than among public decision makers.
Finally, public choice studies have also found that, where a policy has high but diffuse costs and low but concentrated benefits, the stronger incentives of the few may have greater influence than the preferences of the many, possibly leading to inefficient policies. This would suggest that government policies, based on behavioral economics, which try to deliberately manipulate choice through nudges are also prone to deliberate manipulation that can lead to suboptimal outcomes.
These behavioral public choice issues do not mean that all behavioral justifications for government intervention are inevitably prone to misuse and will result in a reduction of social welfare. Daniel Kahneman, in many ways the father of behavioral economics, considers two modes of thinking: System 1 thinking “operates automatically and quickly, with little or no effort and no sense of voluntary control,” while System 2 thinking “allocates attention to the effortful mental activities that demand it, including complex computations.” In his excellent book, Thinking Fast and Slow, he offers many examples of the power and the pitfalls of System 1 thinking. Most of the biases found in the behavioral economics literature result from actions dominated by the “freewheeling impulses” of System 1 rather than the “conscious, reasoning self” of System 2.
The question is whether private decision makers acting in the marketplace are more or less prone to harmful biases than are the public decision makers who regulate the economy. Behavioral economists who advocate soft paternalistic policies are essentially motivated by the belief that government technocrats are, by nature, training, and employment, disposed toward System 2 thinking and can therefore design policies that overcome the problems caused by System 1 reasoning. Skeptics, like me, worry that the narrowness of the expertise of government technocrats will subject them to overconfidence, that government experts frequently will have a limited and biased understanding compared to the information provided by a more decentralized, market approach, that the decision making of government officials will be influenced by public choice factors that can limit the effectiveness of the policies, and that the use of government nudges to limit choice can reduce autonomy, dignity, and the motivation of ordinary people to engage and nurture their System 2 reasoning.
This weighing of the pros and cons of government intervention in some ways is similar to the traditional public finance calculus of weighing the inefficiencies caused by market failures against the inefficiencies caused by government failures in attempting to address the market failures through regulation.
In addition to the energy efficiency regulations I discussed earlier, Kip and I examined a number of policies regarding environmental and safety risk and uncertainty, to assess whether policies remediate well-documented perceptional biases of risk and uncertainty or whether they incorporate and institutionalize them. With respect to risk, one of the most-documented biases people exhibit is that they tend to overestimate low probability risks of death (such as the risks of botulism, lightning strikes, and natural disasters) and they tend to underestimate high probability risks of death (such as the risks of stroke, cancer, and heart disease).
We examined whether government policies tend to overcome this bias in assessing risk or whether they instead institutionalize them. Government agencies could be better suited to making more accurate risk assessments if they have additional and unbiased information about the risks that the general public may not have. Government experts who have a professional involvement in particular risk areas could have more accurate beliefs because they have obtained more information than the average citizen has about the true risks involved. Government agencies have the expertise and staff to stay informed about the evolving scientific evidence with respect to risk thus relying more on Kahneman’s System 2 thinking when evaluating these risks.
There indeed appears to be some benefits to familiarity with risks in terms of being able to make sound risk judgments. For example, survey evidence demonstrates that judges have more accurate risk assessments of various kinds of death than does the general public, as judges tend to overestimate small risks and underestimate large risks to a lesser extent than does the general public.
Unfortunately, in many instances, government policies serve to incorporate the same kinds of risk perception biases plaguing individual risk judgments. Government risk assessment practices frequently devote inordinate attention to worst-case scenarios, leading to overestimation of low-probability events.
For example, the government’s Superfund program for hazardous waste sites incorporates a series of conservatism biases that tend to lead to an overstatement of the risk level. The EPA’s assessment of the risk is a product of the level of concentration of a particular chemical, the frequency of exposure to the chemical, the amount of exposure, and the dose-response relationship linking the chemical exposure to an estimated risk, such as cancer. EPA uses an upper bound value for each of these parameter estimates, which is then compounded through each step in the risk assessment, leading to an extremely high-end risk estimate. This is known as cascading conservatism. Suppose the agency calculates cancer risk at a hazardous waste site by multiplying a series of four parameters, where for each parameter the agency uses the ninety-fifth percentile value of the parameter. If all parameters in the risk calculation are ninety-fifth percentile values, then the overall risk calculation that compounds these biases has a much lower chance than 5 percent of reflecting the actual risk. The chance that the calculated risk could be as large as the estimated risk value is only 0.0006 percent. This is extreme conservatism in risk assessment.
There are many other examples of regulatory agencies relying on estimates of risk that compound the conservatism bias. For example, in its evaluation of the risk of methyl mercury, the EPA relied on a reference does that started with a benchmark dose that is the lowest maternal blood mercury concentration expected to lead to a five percent increase in an adverse health outcome in children, then took the ninety percent lower confidence limit of this benchmark dose and then applied an additional safety factor by dividing the dose by ten.
EPA’s risk assessments are frequently biased in other ways. For example, in its assessment of cancer risk from hazardous waste sites, the agency does not incorporate the number of people exposed to the risk. Instead, it treats real and hypothetical exposures equally. In his book, “Breaking the Vicious Cycle,” Justice Breyer cites his own experience presiding over a case involving a hazardous waste site: “The site was mostly cleaned up. All but one of the private parties had settled. The remaining private party litigated the cost of cleaning up the last little bit, a cost of about $9.3 million to remove a small amount of highly diluted PCBs and volatile organic compounds by incinerating the dirt. How much extra safety did this $9.3 million buy? The forty-thousand-page record of this ten-year effort indicated that, without the extra expenditure, the waste dump was clean enough for children playing on the site to eat small amounts of dirt daily for 70 days each year without significant harm. Burning the soil would have made it clean enough for the children to eat small amounts daily for 245 days per year without significant harm. But there were no dirt-eating children playing in the area, for it was a swamp. Nor were dirt-eating children likely to appear there, for future building seemed unlikely.”
There is also a tendency to institutionalize the premium on zeroing out risk, which is a consequence of the biased perception of overestimating low-probability risks. (People overestimate low-probability risk, but accurately estimate zero risk, mistakenly implying a steep drop off in risk as you go to zero.) You see this in policies that aim to achieve an “adequate margin of safety” below the safe exposure level, such as in the Clean Air Act and in FDA regulations of pharmaceuticals and in USDA’s food safety regulations.
Kip and I also examine whether government policies counter or institutionalize common psychological biases with respect to uncertainty. Uncertainty concerns the imprecision involved in assessing risk levels. Take, as an example, two automobiles. Automobile A poses a well-known defect risk of 2/1,000 over the lifetime of the vehicle. Automobile B is newer to the market, and there is a 50 percent chance that the default risk is 1/1,000 and a 50 percent chance that it is 3/1,000. Both of these cars therefore pose an average defect risk of 2/1,000 and should be viewed as posing equivalent risks. However, people generally exhibit a form of ambiguity aversion that makes the precisely known risk of automobile A less fearsome than the uncertain risk of automobile B.
Government policies frequently reflect this ambiguity aversion to novel risks. For example, court rulings tend to demonstrate a bias against innovation and the attendant uncertainties of novel drug products. In situations where there are adverse health effects from new drugs, the courts are more likely to levy sanctions against the producer. This bias is also reflected in product liability case experiments using a sample of judges participating in a legal education program.
One of the seminal findings of behavioral economics is that, contrary to the standard expected utility theory of neoclassical economics, people exhibit loss aversion. In other words, people value avoiding a loss (according to some studies, by up to 7 times) more than they value achieving an equivalent gain. From the standpoint of benefit-cost analysis, losses and gains should be treated symmetrically to achieve the most risk reduction per dollar spent. Many government policies institutionalize the loss aversion phenomenon, seeking first to “do no harm.” For example, FDA regulations take the starting point as avoiding harm, which leads to the failure to approve drugs that on balance may enhance health.
Where does this assessment of behavioral economics leave us? I believe the findings that document systematic anomalies that lead to irrational decisions are important contributions to the field of economics. But I suggest much more humility and caution, and an approach that is less dismissive of the merits of individual choice, when making policy prescriptions stemming from behavioral findings rather than traditional market failures.
What we’ve seen in the case of the energy efficiency regulations, is that the agencies assume that findings of short-sightedness in some contexts is sufficient rationale for overriding consumer preferences in other contexts. This approach establishes a dangerous precedent: If agencies can justify regulations on the unsubstantiated premise that consumers and firms (but not regulators) are irrational, then they can justify the expansive use of regulatory powers to control and constrain virtually all choices consumers and firms make.
Indeed, if anything, the energy efficiency regulations offer a nice example of how public choice incentives can lead to bad policies and the misuse of behavioral findings. In this case, the main failure of rationality is not with the energy-using consumers and firms, but instead the main failure of rationality is with the regulators themselves. Thank you, and I’m happy to take your questions.
Commentary
Energy efficiency, risk and uncertainty, and behavioral public choice
March 6, 2015