The U.S. Department of Energy (DOE) is updating its social cost of carbon (SCC) estimate as part of a proposed rulemaking setting energy efficiency standards for commercial refrigerators. While this may seem like a routine bureaucratic update, it highlights a fundamental problem with how the government evaluates climate policy.
The SCC is sometimes described as a monetized estimate of the impact from releasing a metric ton of CO2 into the atmosphere. However, this characterization is misleading. The final SCC calculations aren’t measured in money, but rather in the more nebulous concept of “welfare.” This distinction is crucial, yet often goes overlooked by government agencies that misleadingly label SCC figures with dollar signs.
To illustrate this point further, consider a hypothetical scenario where a regulation generates a “$2 million benefit” today by reducing CO2 emissions, which delivers benefits to people at some point in the future. One might reasonably conclude this is twice as beneficial as a similar SCC benefit worth $1 million, but this conclusion would be incorrect.
The problem lies in the nature of the welfare measure used in the SCC calculations. Because it is a measure of economic utility, the SCC is ordinal, not cardinal, meaning the number expresses a ranking of values rather than aggregable magnitudes. In simpler terms, while we can say that a higher SCC value represents a greater benefit than a lower one, we can’t say precisely how much greater that benefit is.
Contrast this against something as simple as regulatory costs, which can and often do have a cardinal interpretation. For instance, $2 million in compliance costs for businesses is genuinely twice as much as $1 million.
To make sense of why agencies engage in this kind of sloppy reporting, one needs to dive deeper into how the SCC is calculated. Climate change has long-term consequences, so how does one weigh the relative value of benefits and costs falling on the current versus future generations? The core challenge the SCC is trying to address is to come up with a single metric that captures the impacts across all the different generations.
The approach taken by the SCC modelers is to assume that an infinitely-living “social planner” exists who considers the preferences of all generations and optimizes overall utility. The social planner weighs the relative value of benefits and costs across generations according to his utility function, called a “social welfare function.”
While some economists find this approach attractive, it is important to recognize two key points. First, the approach is normative, based on modelers’ opinions about how much to weight benefits across generations. Reasonable people can disagree about whether future people deserve more weight, equal treatment, or less weight than these modelers assume.
Second, this social welfare function approach is inconsistent with “economic efficiency,” which many economists believe should be the focus of cost-benefit analysis. Economic efficiency would add up benefits and costs in straightforward dollar terms without the use of any additional welfare weighting. Any cost-benefit analysis incorporating the SCC as it is presently calculated is not measuring economic efficiency.
This is an important finding because courts have ruled that an agency’s use of a model is “arbitrary” if it “bears no rational relationship to the reality it purports to represent.” The SCC calculations, by representing the welfare of an abstract social planner rather than real-world economic impacts, would seem to violate this principle.
To be clear, these criticisms apply to the DOE’s old estimates of the SCC, just as with the new ones it is introducing with its commercial refrigerators rulemaking. However, the DOE’s updated SCC estimates introduce additional problems as well. The new models incorporate more mortality effects of climate change than prior ones did, using metrics such as the value of a statistical life (VSL). VSL is a method economists use to assign dollar values to lives expected to be saved.
Ironically in the context of climate change, the VSL doesn’t account for intergenerational externalities—how individuals’ actions impact third parties in the future—as it is based on what specific individuals are willing to pay to reduce risk to themselves and those closest to them, not what society as a whole would pay if everyone’s preferences—including those of future generations—were respected. The bias built into both the VSL and the SCC is that these metrics encourage too much consumption at the expense of productive investment. Furthermore, investment is what stands to benefit future generations most.
These conceptual difficulties help explain the DOE and other government agencies’ confusion surrounding the SCC. As a replacement, agencies could calculate impacts on GDP, which can be derived from existing climate models or based on separate analysis and which can also be expressed in dollars easily.
However, if the government insists on using some version of the SCC, it should at least label the units accurately in its reporting. Instead of expressing figures in dollars, it should use clear labels like “welfare dollars” or, better yet, “social planner wellbeing” to reflect what is actually being measured.
The social cost of carbon’s unscientific basis in a subjective social welfare function rather than economic efficiency makes it an unreliable tool for respectable cost-benefit analysis. Policymakers and economists should reconsider this metric and develop more suitable alternatives that can truly inform efficient climate policy.