Tuesday, August 14, 2018

Economics of Climate Change: Three Recent Takes

Most economists took their last course in physical science many years ago, back in college days, and lack any particular in-depth knowledge of how to model weather or climate.  But economists can contribute usefully to the climate change debate in other ways. At least some economists do have expertise in patterns of energy use, potential for substitution, and technology, and thus have something to say about likely future paths for the emissions of carbon (and other greenhouse gases), and what it might take to change these paths. And at least some economists have expertise in thinking about how changes to climate would affect economic and human outcomes, ranging from crop yields to human mortality. Here are a few recent examples.

Richard Schmalensee takes a usefully unsentimental look at the prospects for a genuinely dramatic reduction in carbon emissions in "Handicapping the High-Stakes Race to Net-Zero," appearing in the Milken Institute Review (Third Quarter 2018). He emphasizes three main challenges:

1) Carbon emissions from emerging economies are rising rapidly, often based on building new plants that generate electricity from coal. Even if emissions in advanced economies were slashed dramatically, there won't be much progress on reducing global carbon emissions without tackling the issue in emerging economies. Schmalensee offers a scenario to illustrate this challenge:
"[S]uppose that, in the next decade or two, advanced economy emissions are cut in half, there is no population growth in emerging economies and emerging-economy emissions per dollar of GDP are cut by 31 percent (to the advanced economy average). Suppose, too, that GDP per capita in emerging economies rises to only 45 percent of the advanced economy average (roughly double what it is today). In this optimistic case, global emissions would still rise by about 1 percent."
Schmalensee also believes that the most popular form of solar energy--that is, photovoltaic technology based on crystalline silicon--is unlikely ever to become cost-competitive with fossil fuels. So the dual challenge here is to find alternatives for low-carbon or carbon-free production of electricity, and then find ways for emerging market and low-income economies to afford the switch to these new technologies.

2) Most scenarios for decarbonization of energy put a high emphasis on use of solar and wind, which raise the challenge of how to build an electricity grid that relies on a intermittent source of energy. Schmalensee writes: 
"The most mature, widely deployed carbon-free generation technologies are wind, solar, hydroelectric and nuclear. Political resistance in many nations to building more dams is substantial, as is resistance to nuclear plants using current-generation designs — though generation from both sources will no doubt expand in emerging economies. Other technologies that are potentially valuable in a carbon-constrained world — among them carbon capture and storage, biofuels, geothermal energy, nuclear fusion, waste-to-energy and wave power — are either untried, immature or only suitable for special locations.
"Accordingly, in most deep decarbonization scenarios, wind and solar play leading roles in mid-century electricity supply. ... But getting to net-zero seems likely to require going significantly beyond 50 percent wind and solar. The main problem is that wind and solar generation are intermittent, with output that is variable on time scales ranging from minutes to seasons, and imperfectly predictable. We know how to operate electric power systems with substantial intermittent generation at reasonable cost, as Germany and California have demonstrated. It is, however, almost universally agreed that we do not know how to operate systems dominated by intermittent generation at reasonable cost."
The solutions here could involve either developing cost-effective and carbon-free sources of electricity that are not intermittent (small nuclear reactors? carbon capture and storage?) or cost-effective methods for mass storage of energy (batteries?). All of these approaches need considerably more  research and development.

3) The main focus of decarbonization has been on production of electricity, but that's only one way in which humans produce and use energy. Schmalensee writes:
"While decarbonizing electricity generation is a necessary step toward net-zero, electricity generation accounts for only about one-third of human-caused CO₂ emissions. Transportation accounts for another fifth — and while road transport (about 15 percent of total emissions) could be electrified at some cost, electrification of air transport seems highly unlikely. More importantly, little attention has been paid to reducing the substantial emissions from industry and construction (about 20 percent), land use (about 13 percent) and various other sources, including cement production and building heating (about 13 percent)."
Thinking about emissions in all of these contexts, and how they occur everywhere in the world, is the actual challenge.

Schmalensee is willing to contemplate large resource expenditures to address these issues. He writes
"In 1965 and 1966, NASA accounted for more than 4 percent of federal spending, which would translate to about $160 billion today. In contrast, the U.S. Department of Energy’s budget request for clean technology development in FY2017 was a paltry $9 billion." His deeper message is that if people are actually serious about the goal of substantial decarbonization of the global economy, announcing lofty goals won't suffice, and modest subsidies for existing technologies won't be nearly enough. A genuinely enormous commitment to change is needed.

Two recent studies by economists take a look at consequences of climate change. One group project from the Climate Impact Lab consortium, "Valuing the Global Mortality Consequences of Climate Change Accounting for Adaptation Costs and Benefits" was written by Tamma Carleton, Michael Delgado, Michael Greenstone, Trevor Houser, Solomon Hsiang, Andrew Hultgren, Amir Jina, Robert Kopp, Kelly McCusker, Ishan Nath, James Rising, Ashwin Rode, Samuel Seo, Justin Simcock, Arvid Viaene, Jiacan Yuan, and Alice Zhang (Becker Friedman Institute for Economists at the University of Chicago, Working Paper 2018-51, August 2018). They write:
"[W]e estimate the mortality-temperature relationship around the world, both today and into the future. This is accomplished by using the most exhaustive dataset ever collected on annual, subnational mortality statistics. These data cover the universe of deaths from 41 countries totaling 56% of the global population at a resolution similar to that of US counties (2nd-administrative level) for each year across multiple age categories (i.e. <5, 5-64, and >64). These data allow us to estimate the mortality-temperature relationship with substantially greater resolution and coverage of the human population than previous studies; the most comprehensive econometric analyses to date have been for a single country or individual cities from several countries. We find that in our sample an additional 35◦C day (-5◦C day), relative to a day at 20◦C, increases the annual all-age mortality rate by 0.4 (0.3) 2 deaths per 100,000."
This data allows them to look at mortality risks accounting for different age groups, different locations within countries, and different per capita income across countries. This framework also allows them to infer what kinds of adaptations that people can make to higher temperatures. They write:

The examples of Seattle, WA and Houston, TX, which have similar income levels, institutions, and other factors, but have very different climates, provide some high-level intuition for our approach. On average Seattle has just 0.001 days per year where the average temperature exceeds ≈32◦C, while Houston experiences 0.31 of these days annually. Houston has adapted to this hotter climate, evidenced by the fact that a day above 32◦C produces 1/40th of the excess mortality in Houston than it does in Seattle (Barreca et al., 2016). ... Indeed, the difference in air conditioning penetration rates, which were 27% in Washington state and 100% in Texas as of 2000-4, provide evidence that the observed differences in temperature sensitivities between these cities reflect cost-benefit decisions. 
This working paper will be hard going for those not initiated into economic research, and the results aren't simple to summarize. But the authors put it this way (citations omitted):
"Together, these two features of the analysis allow us to develop measures of the full mortality-related costs of climate change for the entire world, reflecting both the direct mortality costs (accounting for adaptation) and all adaptation costs. We find that the median estimate of the total mortality burden of climate change across 33 different climate models is projected to be worth 36 death equivalents per 100,000 at the end of the century or roughly 3.7% of global GDP when using standard assumptions about the value of a statistical life. Approximately 2/3 of the death equivalent costs are due to the costs of adaptation. Further, failing to account for income and climate adaptation as has been the norm in the literature would overstate the mortality costs of climate change by a factor of about 3.5. Finally, we note that there is evidence of substantial heterogeneity in impacts around the globe; at the end of the century we project an increase of about 3,800 death equivalents annually in Mogadishu and a decrease of about 1,100 annually in Oslo, Norway."

In another recent study,  Riccardo Colacito, Bridget Hoffmann, Toan Phan, and Tim Sablik look at "The Impact of Higher Temperatures on Economic Growth" (Federal Reserve Bank of Richmond, Economic Brief EB18-08, August 2018). A general finding in the climate change literature is that warmer temperatures would have less effect on the US economy, in part because agriculture and other obviously weather-dependent industries are a relatively small share of the US economy, and in part because the US economy has considerable resources for adaptation.

However, this paper points out that in hot summers, lots of US industries see a decline. For example, the real estate industry does less well in exceptionally hot summers--maybe because people are less enthusiastic about shopping for homes or moving when it's very hot. The insurance industry does less well in hot summers, in part because extreme heat pushes up medical costs and reduces profits for insurance firms. Other studies have found that very high summer temperatures are associated with lower production at automobile plants. In addition, these effects of  hotter summers on reduced output seem to be getting larger, rather than smaller over the last four decades.

This study is based on variation across seasons and years, and it doesn't take into account the kinds of adaptations that might occur in the longer run, so using it to project decades into the future seems like a stretch to me. But adaptations to higher temperatures often have substantial costs, too. Overall, this study, together with the previous estimates about costs of mortaility and adaptation, serve as a useful warning that higher temperatures and climate change aren't just about farming.