Recipient Organization
UNIV OF MINNESOTA
(N/A)
ST PAUL,MN 55108
Performing Department
Applied Economics
Non Technical Summary
The challenge of a warming global climate it perhaps the greatest challenge facing the world. Failure to achieve a dramatic reduction in emissions of greenhouse gases, fairly quickly, poses a threat to the planet that is hard to exaggerate. Mitigation is the word used to describe a reduction, more or less aggressive, of emissions of greenhouse gases, with the goal of lessening the risk of dangerous warming. Estimates of the monetary benefits that will accompany successful mitigation are quite large. What about the costs? Here the estimates range widely, from something like 2-3% of global GDP annually (Pindyck 2013a) to zero (Jacobson et al. 2015). One part of the proposed research will be aimed at exploring this divergence, in particular whether the higher estimate reflects the most recent evidence regarding the cost of moving away from fossil fuels in favor of renewable sources of energy.In the event that mitigation efforts are unsuccessful, and warming continues on the trend we see today, the temptation will be great to engage in geoengineering, which the UK Royal Society (2009) has defined as "the deliberate large-scale manipulation of the planetary environment to counteract anthropogenic climate change." Geoengineering can take many forms, including the injection of reflective particles into the atmosphere, painting rooftops white, and placing large reflective shields in earth orbit. Any such approach to the climate challenge is controversial and carries a certain amount of risk. It is also relatively inexpensive, which means that an individual country, perhaps one that experiences real harm before others, could undertake one of the listed strategies on its own, thereby imposing geoengineering upon the rest of the world. The other part of the proposed project is to explore this question from scientific and economic perspectives.
Animal Health Component
50%
Research Effort Categories
Basic
50%
Applied
50%
Developmental
(N/A)
Goals / Objectives
1. Evaluate the cost of reducing emissions of greenhouse gases, in particular the share of globalGDP that must be invested in order to achieve a significant reduction.2. Estimate the optimal level of climate engineering and describe a mechanism for designing aninternational voting scheme that might be used to select this level.
Project Methods
Climate mitigation and abatement costs in Integrated Assessment Models (IAMs)The study of climate change is complex in part because of the ways in which economic activity influences and is influenced by a changing climate system. A leading tool for studying the climate-economic connection is the class of integrated assessment models or IAMs, which may usefully be divided into two categories: those developed by economists and those developed by climate scientists. The former tend to be more rudimentary in their modeling of the climate science; the latter tend to be more rudimentary in their modeling of the economy.The three most prominent economic IAMs are DICE (Nordhaus 2017), Fund (Tol 1995), and PAGE (Hope et al., 1993). These three models, modified to be more comparable in their results, formed the core analytical framework for the U.S. government's recent assessment of the Social Cost of Carbon (Greenstone et al. 2013). In broad outlines, these models are based upon economic growth models, with increased GDP over time leading to increased emissions of greenhouse gases. This in turn leads to greater warming, which in turn causes increased economic damage due to health effects, impacts on agriculture, and the like. Increased damage is manifest in a reduction in global GDP.A considerable amount of attention has been paid to the damage function in IAMs such as DICE. A critique of the DICE damage function may be found in, for example, Weitzman (2012) or Pindyck (2015). At present it does not appear that the abatement cost function in IAMs has received the same amount of attention. In the proposed research project I plan to examine the cost function, with the goal of understanding why models such as DICE yield very high estimates of the cost of reducing carbon emissions, while many other sources (Jacobson et al., 2015) argue that dramatic declines in the cost of renewables mean that over the coming decades we will be able to achieve zero emissions at little cost, perhaps even negative cost.Assumptions in DICE about abatement costs are at the core of the issue. Because DICE embeds an assumption that global GDP is growing more quickly than abatement cost is declining, DICE's dollar cost of achieving zero carbon emissions rises for many decades. What is more, the cost of abatement is higher in the 2016 version of DICE than in the previous version, from 2013. This specification of the cost of abatement appears to be at odds with a considerable amount of evidence about the trajectory of costs for alternative energy sources, including solar, wind, and battery storage. The International Energy Agency (2015), for example, estimates that the global capital and fuel costs of doing nothing to reduce carbon emissions will be $318 trillion out to 2050. The global cost of aggressive climate action, on the other hand, sufficient to keep warming below two degrees Celsius, is estimated to be $243 trillion. The first component of the proposed research project will attempt to reach some insights about which view is closer to being accurate: the high abatement cost projected by DICE; or the low or negative cost projected by Jacobson et al. (2015) and the IEA (2015).International decisions on climate engineeringIf we are unable to reduce GHG emissions fast enough to avert dangerous warming, the temptation to engage in climate engineering will be powerful. Weitzman (2015) has expressed a deep concern that this is a dangerous path to follow. Others, including Morton (2017) and Keith (2014) are more sanguine. A growing scientific literature is pursuing the question of how climate engineering might affect various parts of the climate system (Kravitz et al., 2014). Either way, the global community will face a formidable challenge if and when we need to decide how to approach the question of climate engineering.Weitzman has proposed a voting scheme that could potentially form the basis for an international agreement on climate engineering. He calls climate engineering a "free driver" problem, in that it is so cheap that (almost) any country can pursue it unilaterally; the results of that country's action would impact all other countries in the world. This offers an interesting dynamic that makes governance structures challenging. This is the challenge that I plan to address in this part of the research project. Working with colleagues, I will proceed in the following steps.First, a colleague at the Pacific Northwest National Laboratory has already performed a set of computer experiments using his climate-engineering model. For annual sulfate injection levels of 0, 5, 10, 20, 30, 40, and 50 Mt, in each case held constant from 2020 to 2100, he ran his model and, for each level of solar radiation management (SRM), produced a sequence of rasterized global maps of annual average temperature and precipitation. He then produced interpolated datasets between 0 and 5 Mt, 5 and 10 Mt, and so on. In the end he provided 51 datasets, for SRM levels 0, 1, ..., 50 Mt/yr. Each contains a rasterized map, including temperature and precipitation data, for each year from 2020 to 2100.Second, these 51 datasets are to be run through the model made available along with a paper by Burke et al. (2015). This model is able to take a time series of temperature and precipitation data, by country, and project GDP/capita out to 2100. We have performed this calculation and the preliminary results show how each country's real GDP/capita changes with SRM. The result, for each SRM level, is a trajectory of GDP/capita for each of 165 countries, for each year from2020 to 2100. Using France as an example, we find the preferred level of SRM to be about 16 Mt/yr. The loss to France, in GDP/capita terms, from doing more than their preferred level of geoengineering is not large, dropping from a max of about $79k to a little below $77k at 50 Mt/yr. These results appear to be new, and they give an indication of which countries prefer no SRM, a moderate amount, or the maximum amount of SRM.Third, in a separate theoretical project, working with several graduate students I have extended the Weitzman voting framework to be more general. In our version it can work even if countries differ in their losses relative to the preferred level of geoengineering and in the weighting of those losses. This is interesting because countries that prefer only a small amount of SRM may have small losses from too little but large losses from too much SRM. Countries that prefer a lot of SRM, on the other hand, might have the opposite preference. Next, the idea is to specify estimated V-shaped loss functions for each country based on the results of running the temperature and precipitation data through the Burke model. Then, using our modified Weitzman voting framework, we aim to solve for the level of SRM that an international body following that framework would select.My approach is first to fit a two-part, piecewise linear regression model to the GDP-SRM data.Once we have obtained the results for each country, it will not be difficult to use the extended Weitzman voting framework to compute the level of SRM that will be selected under that framework. It might be interesting to compare the results for different weighting schemes. Population is the weighting system that Weitzman seems to prefer. But the weights might instead be inversely related to median income by country. Or land area. It might be worthwhile comparing a few different weighting schemes, all of which will be fairly straightforward once the estimates have been found for each country.