Here’s a question: is it better to drive somewhere or to take a ride-hailing service like Uber or Lyft? I don’t mean better for you personally — faster or cheaper. I mean better for the world, for society, for the air and atmosphere … better, all things considered. A clever new study from researchers at Carnegie Mellon University attempts to answer that question.Ride-hailing services carry more external costs than private vehiclesIn the paper, Jacob Ward, Jeremy Michalek, and Constantine Samaras attempt to tally up the relative costs to the environment and society of taking a trip via a privately owned vehicle vs. taking the same trip in what they call a transportation network company (TNC) like Uber or Lyft, in six of the companies’ biggest markets.This involves two steps in each market. First, they add up how many miles the respective vehicles travel per trip. Then, they add up the total externalities — in vehicle emissions, congestion, crashes, and noise — represented by each mile traveled. (These externalities, notoriously, are not priced into transportation decisions; thus the name.)To understand the results, you have to understand one key fact: TNC vehicles travel more miles per trip. They have to drive between where they drop off one fare and pick up another, which sometimes involves quite a bit of just wandering around. This time spent with no passenger is called “deadheading,” and those miles must be added to their trip miles.Here’s a clear visual representation:Those dotted black lines on the bottom half? That’s deadheading. Those extra miles traveled represent more externalities — more cost to the environment and society. On average, a TNC trip carries 32 to 37 cents more in external costs than a private vehicle trip. (See also this recent MIT study, which found that TNC vehicles increase urban road congestion.)There are three countervailing factors, but only the third is big enough to flip the equation enough to give TNCs the advantage in some limited circumstances.Ride-hailing services reduce air pollutionFirst, though additional miles lead to more greenhouse gas emissions, congestion, crashes, and noise, somewhat counterintuitively, they lead to less air pollution. The secret here is that the bulk of particulate air pollution is generated from “cold starts,” i.e., engines turning over to start up. By contrast, TNC vehicles arrive “hot.” Their engines are already running, so they don’t do cold starts. In addition, TNC vehicles are, on average, newer, and thus cleaner.On average, TNC trips represent a 50 to 60 percent decline (9–13¢ per trip) in air pollutants like NOx, PM2.5, and VOCs.If TNC vehicles electrify faster than private vehicles, that advantage will grow, because EVs generate no tailpipe pollution. But if private vehicles electrify equally fast, the comparative advantage will stay the same.Regardless, the difference is not enough to overcome the other externalities. TNC trips represent a 20 percent increase in costs from greenhouse gas emissions and a 60 percent increase in costs from congestion, crashes, and noise (all told, about 45¢ more per trip). Overall, a 9–13¢ decrease and a 45¢ increase add up to 32 to 37 cents more per trip, on average.Electric ride-hailing vehicles … are still mostly worse than private vehiclesThe second countervailing factor is vehicle electrification. Doesn’t that reduce externalities? What the researchers found is that a) if the TNC car is electric, while b) the private vehicle alternative is an internal combustion engine car, and c) the TNC car charges entirely with zero-carbon electricity, then the overall environmental and social costs of a TNC trip and a personal vehicle trip are … about the same. Of course, those conditions are rarely met. On real-world grids, which are still powered overwhelmingly by fossil fuels, electrification of TNCs reduces the relative advantage of personal vehicle trips by about 16 or 17 percent. It does not eliminate the advantage. And of cour