On Buying Insurance, and Ignoring Cost-Benefit Analysis

By Frank Ackerman

First in a series of posts on climate policy.  

The damages expected from climate change seem to get worse with each new study. Reports from the IPCC and the U.S. Global Change Research Project, and a multi-author review article in Science, all published in late 2018, are among the recent bearers of bad news. Even more continues to arrive in a swarm of research articles, too numerous to list here. And most of these reports are talking about not-so-long-term damages. Dramatic climate disruption and massive economic losses are coming in just a few decades, not centuries, if we continue along our present path of inaction. It’s almost enough to make you support an emergency program to reduce emissions and switch to a path of rapid decarbonization.

But wait: isn’t there something about economics we need to figure out first? Would drastic emission reductions pass a cost-benefit test? How do we know that we wouldn’t be spending too much on climate policy?

In fact, a crash program to decarbonize the economy is obviously the right answer. There are just a few things you need to know about the economics of climate policy, in order to confirm that Adam Smith and his intellectual heirs have not overturned common sense on this issue. Three key points are worth remembering.

Worst-case risks matter more than likely outcomes

For uncertain, extreme risks, policy should be based on the credible worst-case outcome, not the expected or most likely value. This is the way people think about insurance against disasters. The odds that your house won’t burn down next year are better than 99 percent – but you probably have fire insurance anyway. Likewise, young parents have more than a 99 percent chance of surviving the coming year, but often buy life insurance to protect their children.

Real uncertainty, of course, has nothing to do with the fake uncertainty of climate denial. In insurance terms, real uncertainty consists of not knowing when a house fire might occur; fake uncertainty is the (obviously wrong) claim that houses never catch fire. See my Worst-Case Economics for more detailed exploration of worst cases and (real) uncertainty, in both climate and finance.

For climate risks, worst cases are much too dreadful to ignore. What we know is that climate change could be very bad for us; but no one knows exactly how bad it will be or exactly when it will arrive. How likely are we to reach tipping points into an irreversibly worse climate, and when will these tipping points occur? As the careful qualifications in the IPCC and other reports remind us, climate change could be very bad, surprisingly soon, but almost no one is willing to put a precise number or date on the expected losses.

One group does rush in where scientists fear to tread, guessing about the precise magnitude and timing of future climate damages: economists engaged in cost-benefit analysis (CBA). Rarely used before the 1990s, CBA has become the default, “common-sense” approach to policy evaluation, particularly in environmental policy. In CBA-world you begin by measuring and monetizing the benefits, and the costs, of a policy – and then “buy” the policy if, and only if, the monetary value of the benefits exceeds the costs.

There are numerous problems with CBA, such as the need to (literally) make up monetary prices for priceless values of human life, health and the natural environment. In practice, CBA often trivializes the value of life and nature. Climate policy raises yet another problem: CBA requires a single number, such as a most likely outcome, best guess, or weighted average, for every element of costs (e.g. future costs of clean energy) and benefits (e.g. monetary value of future damages avoided by clean energy expenditures). There is no simple way to incorporate a wide range of uncertainty about such values into CBA. The second post in this series will look more deeply at economists’ misplaced precision about climate damages.

Costs of emission reduction are dropping fast

The insurance analogy is suggestive, but not a perfect fit for climate policy. There is no intergalactic insurance agency that can offer us a loaner planet to use while ours is towed back to the shop for repairs. Instead, we will have to “self-insure” against climate risks – the equivalent of spending money on fireproofing your house rather than relying on an insurance policy.

Climate self-insurance consists largely of reducing carbon emissions, in order to reduce future losses.[1] The one piece of unalloyed good news in climate policy today is the plummeting cost of clean energy. In the windiest and sunniest parts of the world (and the United States), new wind and solar power installations now produce electricity at costs equal to or lower than from fossil fuel-burning plants.

A 2017 report from the International Renewable Energy Agency (IRENA) projects that this will soon be true worldwide: global average renewable energy costs will be within the range of fossil fuel-fired costs by 2020, with on-shore wind and solar photovoltaic panels at the low end of the range. Despite low costs for clean energy, many utilities will still propose to build fossil fuel plants, reflecting the inertia of traditional energy planning and the once-prudent wisdom of the cheap-fuel, pre-climate change era.

Super-low costs for renewables, which would have seemed like fantasies 10 years ago, are now driving the economics and the feasibility of plans for decarbonization. Many progressive Democrats have endorsed a “green new deal”, calling for elimination of fossil fuels, massive investment in energy efficiency and clean energy, and fairness in the distribution of jobs and opportunities.

Robert Pollin, an economist who has studied green new deal options, estimates that annual investment of about 1.5 percent of GDP would be needed. That’s about $300 billion a year for the United States, and four times as much, $1.2 trillion a year, for the world economy. Those numbers may sound large, but so are the fossil fuel subsidies and investments that the green new deal would eliminate.

In a 2015 study, my colleagues and I calculated that 80 percent of U.S. greenhouse gas emissions could be eliminated by 2050, with no net increase in energy or transportation costs. Since that time, renewables have only gotten cheaper. (Our result does not necessarily contradict Pollin’s estimate, since the last 20 percent of emissions will be the hardest and most expensive to eliminate.)

These projections of future costs are inevitably uncertain, because the future has not happened yet. The risks, however, do not appear dangerous or burdensome. So far, the surprises on the cost side have been unexpectedly rapid decreases in renewable energy prices. These are not the risks that require rethinking our approach to climate policy.

Costs of not reducing emissions may be disastrously large

The disastrous worst-case risks are all on the benefits, or avoided climate damages, side of the ledger. The scientific uncertainties about climate change concern the timing and extent of damages. Therefore, the urgency of avoiding these damages, or conversely the cost of not avoiding them, is intrinsically uncertain, and could be disastrously large.

It has become common, among economists, to estimate the “social cost of carbon” (SCC), defined as the monetary value of the present and future climate damages per ton of carbon dioxide or equivalent. This is where the pick-a-number imperative of cost-benefit analysis introduces the greatest distortion: huge uncertainties in damages should naturally translate into huge uncertainties in the SCC, not a single point estimate.

Frank Ackerman is principal economist at Synapse Energy Economics in Cambridge, Mass., and one of the founders of Dollars & Sense, which publishes Triple Crisis. 

The next post in this series will examine the debates about the SCC, showing that there are indeed large uncertainties in its value, no matter how inconvenient that may be for economists and their models.

[1] Adaptation, or expenditure to reduce vulnerability to climate damages, is also important but may not be effective beyond the early stages of warming. And some adaptation costs are required to cope with warming that can no longer be avoided – that is, they have become sunk costs, not present or future policy choices.