Socioeconomic Impacts of Global Warming are Systematically Overestimated
Part II: How Large Might be the Overestimation?
A major argument advanced for drastic GHG emission reductions is that, otherwise, we are told, global warming will exacerbate the problems that developing countries already face (e.g., low agricultural productivity, hunger, malaria, water shortage, coastal flooding). Consequently, global warming would/could/might/may swamp their meager adaptive capacity. Not only would this be a tragedy for the developing countries, it could trigger social, political and economic instability, and mass migrations which would create large negative spillover effects for the US and other industrialized countries (see, e.g., here).
It’s true, many developing countries’ adaptive capacity is relatively low today. But will it be equally low in the future when global warming, presumably, kicks in?
Figure 3 provides estimates of net GDP per capita — a determinant of adaptive capacity — in 1990 (the base year), 2100 and 2200 for four IPCC reference scenarios for areas that comprise today’s developing and industrialized countries after accounting for any losses in GDP due to future global warming. For 2100 and 2200, net GDP per capita is estimated assuming that (a) GDP per capita in the absence of global warming will grow per the IPCC SRES scenarios and (b) adjusting it downward to account for the costs of climate change per the Stern Review’s 95th percentile estimate under the “high climate change” scenario, equivalent to the IPCC’s warmest scenario (A1FI). For 1990, I use the actual GDP per capita because this is the base year from which all changes are calculated for future years.
I use the Stern Review’s estimates which, unlike most other studies, account for losses not only due to market impacts of global warming but also to non-market (i.e., environmental and public health) impacts, as well as the risk of catastrophe, despite the fact that the Stern Review is an “outlier” that many economists believe overstates losses due to global warming (Tol 2008). Its 95th percentile estimate for losses in GDP under the warmest scenario is 7.5% in 2100 and 35.2% in 2200. The precise methodology for developing this Figure 3 is provided here.
Figure 3 shows that under the warmest scenario (A1FI), the scenario that prompts much of the apocalyptic visions of global warming, net GDP per capita of inhabitants of developing countries in 2100 ($61,500) will be double that of the US in 2006 ($30,100). Therefore, by 2100, developing countries’ adaptive capacity should on average be far greater than the US’s today merely on the basis of higher GDP per capita!
[By 2200, the net GDP per capita of today’s developing countries will be almost triple the US’ in 2006 ($86,200 versus $30,100).]
Thus, the problems of poverty that warming would exacerbate (e.g., low agricultural productivity, hunger, malnutrition, malaria and other vector borne diseases) ought to be substantially reduced if not eliminated by 2100, even if one ignores any secular technological change that ought to occur in the interim. Tol and Dowlatabadi (2001), for example, show that malaria has been functionally eliminated in a society whose annual per capita income reaches $3,100. Therefore, even under the poorest scenario (A2), developing countries should be free of malaria well before 2100, even assuming no technological change in the interim. Similarly, if the average net GDP per capita in 2100 for developing countries is $10,000–$82,000, then their farmers would be able to afford technologies that are unaffordable today (e.g., precision agriculture) or new technologies that should come on line by then (e.g., drought resistant seeds). But, since impact assessments generally fail to fully factor in increases in economic development (and technological change), they substantially overestimate future net damages from global warming (see Part I).
Note that Figure 3 shows that through 2200, notwithstanding global warming, net GDP per capita will be highest under the warmest scenario, and lowest under the poorest scenario (A2). This suggests that if humanity has a choice of development paths, it ought to strive to take the path with the highest economic growth. That is, a richer-but-warmer world is better than poorer-but-cooler worlds.
The second major reason why the impacts of global warming are systematically overestimated is that few impact studies consider secular technological change and most assume that no new technologies will come on line, although some do assume that greater adoption of existing technologies with GDP per capita and, much less frequently, a modest generic improvement in productivity (see Part I).
So how much of a difference in impact would consideration of both economic development and technological change have made?
If impacts were to be estimated for 5 or so years into the future, ignoring changes in adaptive capacity between now and then probably would not be fatal. However, the time horizon of climate change impact assessments is often on the order of 50–100 years or more. The global impacts assessments discussed in Part I, for instance, use a base year of 1990 to estimate impacts for 2025, 2055 and 2085. The Stern Review’s time horizon extends out to 2100–2200 and beyond (Stern Review 2006).
It should be noted that some of the newer impacts assessments have begun to account for changes in adaptive capacity. For example, Yohe et al. (2006), in an exercise exploring the vulnerability to climate change under various climate change scenarios, allowed adaptive capacity to increase between the present and 2050 and 2100. However, they limited any increase in adaptive capacity to “either the current global mean or to a value that is 25% higher than the current value – whichever is higher” (Yohe et al. 2006, p. 4 of the full report). Such a limitation would miss most of the increase in adaptive capacity implied by Figure 3.
More recently, Tol et al. (2007) analyzed the sensitivity of deaths from malaria, diarrhea, schistosomiasis, and dengue deaths to warming, economic development and other determinants of adaptive capacity through the year 2100. Their results indicate, unsurprisingly, that consideration of economic development alone could reduce mortality substantially. For malaria, for instance, deaths would be eliminated before 2100 in a number of the more affluent Sub-Saharan countries (Tol et al. 2007, p. 702). This result is consistent with retrospective assessments which indicate that over the span of a few decades, changes in economic development and technologies can damp down various indicators of adverse environmental impacts and negative indicators of human well-being (see here). For example, due to a combination of greater wealth and secular technological change, U.S. death rates due to various climate-sensitive water-related diseases — dysentery, typhoid, paratyphoid, other gastrointestinal disease, and malaria —declined by 99.6 to 100.0 percent from 1900–1970, that is, over seventy years. See Figure 4.
Figure 4: Death rates for various water related diseases, 1900-1970. Source: Goklany (2009b), based on various issues of the Statistical Abstract.
Similarly, as shown in Figure 5, average annual global mortality and mortality rates from extreme weather events have declined by 93–98 percent since the 1920s (Goklany 2009c), a span of almost ninety years. Thus, not fully accounting for changes in the level of economic development and secular technological change would understate future adaptive capacity which then could overstate impacts by one or more orders of magnitude if the time horizon is several decades into the future.
Figure 5: Global Death and Death Rates Due to Extreme Weather Events, 1900–2008. The extreme events include the following: droughts, extreme temperatures (both extreme heat and extreme cold), floods, wet mass movement (i.e., slides, waves, and surges), wildfires, and storms (e.g., hurricanes, cyclones, tornados, typhoons, etc.). Note that data for the last period are averaged over nine years. Source: Goklany (2009c), using data from EM-DAT (2009).
In fact, it is precisely the failure to account for the combination of economic and technological development that has caused high profile prognostications such as Malthus’s original conjecture about running out of cropland, The Limits to Growth, and The Population Bomb, to fizzle spectacularly (see here).
Because they share similar methodological flaws, there is no reason to believe that the global warming impacts assessments undertaken to date will fare any better.