A recent study from a team that includes Stephen Holland from the University of North Carolina makes a remarkable statement, that “electric vehicles, on average, generate greater environmental externalities than gasoline vehicles.” The study compares electric vehicles (EVs) with gasoline vehicles and finds EVs wanting.

The key problem with the paper is that it uses already-outdated data and doesn’t in its main analysis consider the ongoing transformation of our grid, which changes the chief results quite remarkably. The culprit behind the Holland study’s findings is good old coal. But King Coal is dying in the U.S.

The grid is changing rapidly around the country. In fact, we are in the middle of a massive transformation of our electricity grid, as illustrated in Figure 1. The most obvious trends are a major reduction in coal power generation and a concomitant increase in natural gas generation. A less obvious trend, but more important for the long term, is a doubling of non-hydro renewables from 3.1 percent to 6.9 percent.

Figure 1: U.S. Electric Grid Mix, 2008 and 2014

Source: EIA

None of these changes are captured in the main conclusions of the Holland study. The next few figures break out the trends captured in Figure 1 by generation source. Coal power production has fallen by almost half since 2008.

Figure 2: Coal Electricity Production in the U.S.

Source: EIA

In the same period, natural gas consumption for electricity has gone up by half.

Figure 3: Natural Gas Electricity Production in the U.S.

Source: EIA

In the same time period, renewable electricity has also climbed by half, a figure that does not even take into account the growing boom in rooftop solar because EIA’s figures only capture utility-scale renewables. Figure 4 shows data through 2014. The pace of wind and solar production is ramping up nicely in 2015, so when we compare the mid-2015 data from Figures 1 to 3, we get about the same trend in renewables as we’ve seen with natural gas: a growth of 50 percent since 2008. (This doesn’t match the numbers in Figure 1, also from EIA, but I’m not sure why.)

As we can see from Figure 4, however, some renewables are growing far faster than others. Wind’s growth has shot up by almost fourfold, or 300 percent growth. Solar is even more remarkable, though it barely registers still. It has grown from just 76 thousand megawatt-hours in 2008 to over 15,000 thousand megawatt-hours in 2015, a growth of over 20,000 percent. Even though solar is still only about one-tenth of the generation from wind, solar is growing much faster and will probably surpass wind in the next decade.

Figure 4: Renewable Electricity Production in the U.S.

Source: EIA Electric Power Monthly

The Holland study also ignores the trend toward EV owners having their own solar panels. This is arguably the perfect “PV4EV” solution and is already cost-effective today, and will become even more cost-effective in the future as solar prices and EV prices steadily come down even further. There are almost no environmental externalities in using sunshine from your own roof to charge your EV. There are some externalities involved in the manufacturing and disposal of the panels and the EV, but these externalities are generally dwarfed by comparable externalities from fossil-fuel production and consumption.

Where is the grid going?

We already know that coal is being phased out, primarily through market forces, but also through federal greenhouse gas emissions limits imposed by President Obama’s EPA, for both new power plants (the rules are already in place) and existing power plants (the rules are still being negotiated).

Natural gas is still growing due to the fracking revolution, but that trend is slowing down because well production falls very sharply after the first year, and there simply aren’t enough resources to continue to make up for these declines after a while.

And renewables will continue to grow sharply, particularly solar, because renewables are now highly cost-effective in many parts of the country with only the federal Investment Tax Credit, or even without any subsidies. We are due for an explosion in solar as prices continue to drop in alignment with a trend known as Swanson’s law, which is the observation that solar prices drop about 20 percent with every doubling of installed capacity. So the more we install, the lower prices go, and the more we install, in a beautiful virtuous circle. This observation is what recently led me to predict a “solar singularity” arriving in the next few years.

So even if the Holland study’s conclusions were accurate today (which they’re not because they rely on already-outdated data), they certainly wouldn’t continue to be accurate in the future. In sum, the study generally ignores the biggest feature of our electricity grid today: the rapid pace of the change it is undergoing.

The study does, however, include deep within its pages a sensitivity analysis and a caveat regarding the changing grid. It states (p. 26) that the main study “is based on a simple snapshot of the electricity grid in the years 2010-2012. We might expect the grid to become cleaner over time by integrating new lower-emission fuels and technologies. Of course, gasoline vehicles may become cleaner over time as well.”

Table 7 of the study (p. 46) captures the result of this sensitivity analysis and shows that the average environmental benefits of EVs are twice as good as for the Toyota Prius (the proxy for gasoline cars in this future-oriented sensitivity analysis). Why isn’t this conclusion featured prominently in the study, in the study abstract, or in news coverage of the study? That’s not at all clear, but I suspect it has something to do with a desire for publicity coming from what seems to be a highly counterintuitive result. Studies that support the conventional view don’t make headlines. 

The study implicitly acknowledges this point by recognizing, for example, that the California electricity grid is relatively clean and thus EVs provide substantial environmental benefits. The study states (p. 18) that the towns “with the highest environmental benefit are all in California because electricity generation in the West does not produce much air pollution.” The California grid is relatively clean (though still almost half natural gas, but the rest is nuclear, hydro and renewable), but not at all beyond the reach of most states. In fact, under today’s trends, most states will become like California in the next decade or two.

The time of EV charging is important

The Holland study assumes that EV charging will generally take place at night (see the page 11 figure in this appendix), which means that even in places like California with a lot of renewables, much of that charging will result in increased production of fossil-fuel-based electricity. The problem with this assumption, however, is that states with a lot of solar are already shifting toward daytime charging models for EVs, because EVs are an ideal way to absorb what may be excess solar power.

For example, in California, there are already many days when there is literally too much solar power on the grid and it has to be either curtailed or negatively priced (which means people actually get paid to use that power). This is a good problem to have. In terms of EVs, it means that EV owners may be paid to charge their cars during the day. None of this new dynamic is captured in the Holland study.

What does this mean for policy?

The Holland study discusses the policy implications of its analysis and suggests a number of times that the federal EV tax credit is unwarranted given the study’s findings that in many parts of the country, EVs don’t currently provide environmental benefits. A better approach would be to look five to 10 years ahead and make the same analysis and policy conclusions, because today’s EV tax credits aren’t about today; they’re about how our future looks. As discussed, the Holland study’s own future-oriented sensitivity analysis finds that EVs have twice the environmental benefits of future hybrid gasoline cars.

The federal EV tax credit will phase out steadily under the requirements of the existing law, which is fine. EVs shouldn’t rely on tax credits or other subsidies in perpetuity. The reality is, however, that if EVs are to realize their promise as the single best way to wean ourselves from petroleum in transportation, then we need to subsidize them in the near term, with the hope that near-term subsidies will grow the market to scale and subsidies can disappear. This is the path we’re on with current federal and state subsidies.


Tam Hunt is a lawyer and writer, owner of the renewable energy consulting company Community Renewable Solutions LLC, and author of the upcoming book The Solar Singularity: Why Our Energy Future Is So Bright.