The High Cost of Climate Uncertainty

Using nearly a century and a half of stock market data, researchers consider individual tolerance for risk to put a more accurate price on carbon and stave off the worst effects of climate change.

Print this page
Based on research by Kent Daniel, Robert Litterman, and Gernot Wagner

Less than a month into its first session, the 114th U.S. Congress has squared up for its first major confrontation with the Oval Office. The debate is over Keystone XL, a pipeline extension stretching from western Canada to the Gulf of Mexico, which has become a focal point for congressional Republicans despite public threats of a veto from the president. Despite the U.S. State Department’s assessment that the environmental effects of the pipeline are likely to be minimal, the pipeline has also become a line in the sand for many climate advocates, and this early standoff sets the stage for the central role pundits are already predicting climate change will play in the upcoming 2016 election. Much of that debate is likely to center on emissions trading systems, like the cap-and-trade scheme presently rolling out in California, and direct carbon taxes. Recent proposals for carbon taxation in the United States, however, may be as much as an order of magnitude lower than they ought to be.

This insight comes from new research by Kent Daniel that seeks to put a price on the potentially catastrophic consequences of climate change. For much of the past two decades, efforts to curb global emissions have primarily come through vague and largely unenforceable international agreements like the Kyoto Protocol, signed by the United States in 1998 but never ratified by the senate. Domestic efforts in the United States have consistently fallen shy of comprehensive reform—the last bill seriously considered by congress to reign in emissions, 2009’s American Clean Energy and Security Act, narrowly passed out of the House of Representatives before the Senate leadership declined to take up the issue. Instead, legislators have produced a patchwork of regulations targeting individual industries, like fuel efficiency mandates for automobiles.

Daniel and the majority of climate economists argue for a different approach. “Instead of, for example, telling the automakers they have to produce cars that meet a certain mileage requirement, a better way to reduce emissions is to start taxing gasoline or any activity that spits a lot of carbon into the atmosphere.” Arriving at a reasonable price for carbon and other greenhouse gas emissions has, however, remained challenging.

One of the chief challenges facing carbon pricing is lingering uncertainty about the potential scope and scale of climate change’s impact on society. While climate scientists have reached near universal consensus that climate change will have an impact on global well-being, considerable variation in the scale of damage predicted by various models persists. Estimates of the future cost of global climate change vary wildly, from as little as one quarter of one percent to as much as 20 percent of global GDP. The most devastating effects of climate change remain decades, if not centuries, off, further heightening uncertainty. As a result, some policy analysts have argued that any decision about pricing carbon should be delayed. Daniel argues, however, “delaying policy changes because of this uncertainty is like not buying automobile insurance because you don’t know if you’ll have an accident.” In the case of climate change, however, the question isn’t if, but when and how substantial the damage will be.

Rather than ignoring this persistent uncertainty, Daniel and his co-researchers Robert Litterman, of Kepos Capital and a board member of the World Wildlife Fund (WWF), and Gernot Wagner, of the Environmental Defense Fund and Columbia University’s School of International and Public Affairs, made the inherent risk associated with climate change—and individuals’ tolerance for risk at large—a central part of their model.

Using 140 years of stock and bond pricing data from the United States, Daniel, Litterman, and Wagner sought to find a proxy for the average American’s tolerance for risk. Over the period, a diversified portfolio of stocks would have earned its owner an average annual real return of 6.4 percent, while an all-bond portfolio would have earned its owner an average annual real return of just 1.6 percent. Despite the historic outperformance of stocks, few investors would hold an all-stock portfolio for a simple reason: A historically higher average rate of return offers no guarantee of a better, or even a positive return at any given point in time. Worse, the moments in which individuals are most likely to need access to their investments—after being laid off during a recession, for example—are also the moments in which their investments are likely to be least valuable.

Given the uncertainty associated with stocks, corporate and government bond purchases, with their guaranteed rates of return, act as a type of insurance against a market crash. With bonds earning, in an average year, just a quarter of the returns of stocks, Daniel and his team hit upon a rough measure of how much the average American has been willing to pay to shelter themselves from the fluctuations of the market, and the potential upside they demand to shoulder the risks associated with it.

A uniform carbon tax, Daniel argues, could similarly function as a type of insurance against the possibility of devastating effects from climate change. “Even if scientists can’t precisely estimate the consequences of higher CO2 concentrations, if there is even a possibility of a catastrophic outcome as a result of our use of fossil fuels, we should pay the cost today to bring down this risk.”

The standard metaphor that has developed around climate change policy is that of a slow-moving ocean liner, in need of early but gradual action, through a slow uptick in carbon pricing. This logic underlies the vast majority of carbon tax proposals, including the American Clean Energy and Security Act and national policies in France, Norway, and, before its repeal last summer, Australia. Daniel and his team’s model turns these assumptions on their heads. Using the historic stock market data as a proxy for the average American’s tolerance for the risks posed by global climate change, and including a probabilistic model for climate disasters—like an extended drought or a severe weather event like Hurricane Katrina—Daniel and his team found policy makers should institute a much higher initial price­­­­­­, gradually reducing it over time as uncertainty about the effects of climate change decreases.

Implementing such a tax, particularly in the new Republican-dominated congress, would be challenging, but it’s not just climate-deniers who might object to carbon pricing. As many of the most carbon-intensive human activities are also necessities—transportation to and from work, home heating and cooling—it has been argued that a carbon tax would be inherently regressive, disproportionately burdening the poor. The result, in the United States, could be an unlikely coalition of the far left and the far right of the political spectrum.

The net effect on individuals’ tax bills could, however, be minimized. A carbon tax, Daniel asserts, wouldn’t be a simple dead-weight loss. Instead the revenue generated through taxing carbon offers opportunities to lower other taxes, including federal payroll taxes and state sales taxes, minimizing changes in individuals’ overall tax burden and mitigating any negative effects of the tax on those in lower income brackets. Recent research further suggests that the United States may be an outlier in the regressivity of its present gasoline taxes, due in part to low investment in public transit. In many countries, particularly developing nations where car ownership remains out of reach for most, carbon taxation may actually be progressive. Overall, Daniel sees carbon taxes as an opportunity to better direct tax policy and monetary incentives away from potentially disastrous use of fossil fuels and towards more productive activities for society as a whole.

If implemented, Daniel and his team’s suggested carbon tax would be substantial but expressed as a percent of GDP—likely around 2 percent—it still falls far short of the largest tax hike implemented in the United States. According to, a nonpartisan project run by the University of Pennsylvania, that honor belongs to the Revenue Act of 1942, levied to finance US involvement in World War II, which raised the tax burden by 5 percent of GDP. If the more dire predictions about climate change come true, the cost of the carbon tax could turn out to be paltry by comparison.

About the researcher

Kent Daniel

Kent Daniel is a Professor in the Finance and Economics Division at the Graduate School of Business at Columbia University. From 1996 to 2006, Kent was...
Read more.
articles by Topic