Would we benefit from climate insurance?

Scientists agree that climate change will have negative consequences, but there is enormous uncertainty regarding the extent of the damages. Civilization-threatening climate events such as runaway global warming or extreme sea level rise may not be likely over the next century, but they also cannot be ruled out. Yet, when economists estimate the monetary benefits of climate change policies, it is common to ignore the role that climate policy plays as insurance against the possibility of catastrophic events.

Consider the benefits an individual receives when he buys a life insurance policy. If he dies, his family is protected from financial ruin. He is likely to lose money on this policy—in fact, he is hoping to lose money—but he still finds it worthwhile to purchase the policy due to the enormous benefits it provides if he dies, as well as the comfort of knowing his family will be covered in that event.

Of course, analogous benefits exist for nearly all types of insurance policies, and the existence of billion dollar insurance industries is proof that people will pay to protect themselves against catastrophic outcomes. In economic terms, individuals who buy insurance are “risk averse.” Estimating the monetary benefits of a life insurance policy without accounting for the risk aversion of the policy holder would be plainly absurd, because that is the benefit of a life insurance policy.

Now consider the benefits to society of climate change policy. Reducing atmospheric concentrations of greenhouse gases is certain to avoid damages caused by increased temperatures and rising sea levels, but it may also avoid catastrophic events that are far worse, and could threaten human civilization in parts of the world. Placing precise probabilities on potential catastrophic climate events is not possible, but without a strong policy response, 1 to 5 percent may be as close to a consensus as is currently available for a civilization-threatening event.

There are clear parallels between an individual purchasing life insurance and a large group of individuals in a society purchasing “climate insurance” (in the form of policies and regulations). Just as individuals benefit when catastrophic risks in their own lives are reduced or eliminated, society benefits when the possibility of catastrophic climate events are reduced or eliminated. If benefits from life insurance policies are real and substantial, then similar benefits must exist for climate policies, only aggregated over millions of people.

In a way, climate policy is a life insurance policy for your whole neighborhood.

The idea that climate policies have insurance benefits is hardly novel–international agreements from the Rio Declaration to the Kyoto Protocol have referenced the “precautionary principle,” which declares that scientific uncertainty is no reason to postpone action that could avoid potentially serious or irreversible harm to society–which makes it all the more strange that when we estimate the benefits of climate policies, we typically ignore their role as insurance.

Risk Aversion and the Social Cost of Carbon

The benefits of climate change policies are estimated by multiplying emissions reductions by a social cost of carbon (SCC), which represents the damages associated with a small increase in emissions in a given year. In this article I will focus on the U.S. government-endorsed estimates of the SCC, because their use has become widespread since the first “Interagency Working Group” report in 2010.

The U.S. government developed SCC values so that government agencies would have uniform estimates of the numerical benefits of reducing carbon dioxide emissions that they could use in evaluations of regulations, such as the recent Clean Power Plan proposed by EPA. The U.S. Government SCC estimates have been used in countless other applications as well, including recently by the State of Minnesota to develop a “value of solar” tariff to be paid to utility customers that produce solar energy.

In its estimate of the SCC, the U.S. Government assumed “risk neutrality” (i.e. zero risk aversion), implying that the preference for purchasing insurance against catastrophic outcomes is not incorporated into the analysis. But if individuals prefer to insurance against catastrophic risks, and climate policies can reduce catastrophic risk, why assume risk neutrality?

One possibility is that two similar sounding (but almost completely opposite) situations are being confused: (1) a regulation has uncertain effects; and (2) a regulation reduces pre-existing uncertainty. When a regulation has uncertain effects, according to a “famous” result in economics (the Arrow-Lind Theorem), it is often acceptable for the government to ignore its constituents’ preference for risk aversion, because the risk of a government project is spread so thin among a large population that it essentially disappears. Perhaps as a consequence, both the U.S. government and U.K. government  recommend risk neutrality as a default assumption for policy evaluations.

But for pre-existing uncertainties such as the risk of civilization-threatening climate events, in no sense does the risk disappear when it is shared among a large population. Instead, all risk averse individuals benefit from reductions in pre-existing uncertainty, and these benefits should be factored into policy evaluations.

In other words, our preference for insurance does not disappear just because our neighbors are threatened as well.

Another likely reason for the assumption of risk neutrality is the complexity it avoids. Calculating an SCC requires estimating the damages to the entire world over hundreds of years from a small increase in carbon dioxide emissions. This is a heroic undertaking, so any opportunity to make simplifying assumptions is tempting. And there are considerable difficulties associated with an attempt to add risk aversion to an SCC estimate. There is no market data suggesting how much we are willing to pay to reduce the likelihood of various civilization-threatening events in the future, so accounting for risk aversion toward such events requires assumptions that would be somewhat arbitrary, debatable, and subject to criticism. It is far easier to avoid such complications altogether.

But simplifying assumptions are only reasonable if they are unlikely to affect the integrity of the results. Various peer reviewed studies have shown that the SCC increases dramatically when the effects of risk aversion are included (see here, here, or here)—the insurance benefits of climate policy could easily be as large as all other benefits combined.

As it turns out, people receive real and substantial benefits from insuring themselves against the annihilation of human civilization.

The government’s solution creates more problems than it solves

To be fair, the U.S. government explicitly recognizes the drawbacks of ignoring risk aversion in its SCC calculations, noting the following in its 2010 report: “These calculations do not take into account the possibility that individuals may have a higher willingness to pay to reduce the likelihood of low-probability, high-impact damages than they do to reduce the likelihood of higher-probability but lower-impact damages with the same expected cost.” The report explains that this concern motivated the publication of an additional SCC estimate for a scenario with “worse-than-expected” climate change damages. This scenario predicts a significantly higher SCC, due to the larger benefits associated with avoiding worse-than-expected damages.

But there are two major problems with this “solution.” First, despite the government’s recommendation to use its full range of SCC estimates, in practice, the tendency has been to ignore the range and focus on the value referred to as its “central” estimate. In setting its “value of solar” tariff, for example, Minnesota only used the U.S. Government’s “central” SCC estimate. The worse-than-expected damage scenario has the appearance of an outlier scenario that can be ignored, when in fact, risk aversion is a standard preference trait that should be included in all SCC estimates.

Second (and more problematic), assuming “worse-than-expected” climate damages is not a conceptually valid method of incorporating risk aversion (and thus insurance benefits) into an SCC estimate. Once any given damage outcome is assumed, uncertainty is removed from the analysis. Once uncertainty has been removed from an analysis, there is no effect of risk aversion.

To see this, consider the benefits of your health insurance policy. The U.S. government methodology is like telling you exactly what your health will be (say a worse-than-expected outcome), and then calculating the benefits of your insurance policy contingent on that outcome. These ex-post benefits are in (almost) no way related to the premiums you are willing to pay before knowing what your health outcome will be. Climate policy is analogous to these insurance premiums, not the ex-post benefits.

It is not clear whether the U.S. government’s “worse-than expected” SCC estimate will overstate or understate a calculation of the SCC when the full distribution of potential damages is considered. But the resulting SCC value is meaningless from a public policy perspective.

None of this is to say that the government SCC estimates are useless. Policy evaluations commonly find that the benefits of actions to mitigate climate change are greater than the costs, even without accounting for insurance benefits. This is persuasive evidence in favor of the policy.

But in the opposite situation, when the estimated costs of a policy are larger than the benefits, the government’s SCC estimates are less useful. Given that the omitted insurance benefits could be the largest category of benefits, the sign of the net benefits of the policy is still unknown.

Not Only “Alarmists” Prefer to Avoid Catastrophes

“Climate alarmist” is a common derogatory term used for advocates of climate change mitigation who critics see as focusing on worst-case scenarios instead of likely outcomes. Of course, the public should not be misled about the likelihood of catastrophic climate events. But these critics often miss the point that catastrophic events do not need to be certain–or even likely–to justify substantial action to mitigate climate change. As risk averse societies, it may be in our interest to reduce or eliminate the possibility of even low probability events.

We receive benefits from life insurance policies, health insurance policies, car insurance policies, etc. We should stop pretending we do not receive similar benefits from climate insurance policies.

This article is published in collaboration with The Energy Collective. Publication does not imply endorsement of views by the World Economic Forum.

To keep up with Forum:Agenda subscribe to our weekly newsletter.

Author: Noah Kaufman is a senior consultant in the Environment Economics Practice of NERA Economic Consulting

Image: Splinters of ice peel off from one of the sides of the Perito Moreno glacier in a process of a unexpected rupture during the southern hemisphere’s winter months, near the city of El Calafate in the Patagonian province of Santa Cruz, southern Argentina, July 7, 2008. REUTERS/Andres Forza.