In trying to push its case with the public, Uber decided to share its internal data with Alan Krueger, a prominent Princeton economist and former head of President Obama's Council of Economic Advisers. (Could this be part of Uber's dividend from hiring former Obama political adviser David Plouffe?) Anyhow, Kreuger finds that Uber drivers on average earn a gross premium of $6.00 an hour over the pay of drivers of traditional cabs. (He also had some rather unsurprising findings, for example that more people are now working for Uber after it expanded the number of cities in which it operates.)

The key issue here is the use of the gross premium rather than a direct earnings comparison. The difficulty, as the paper notes, is that we don't know the costs incurred by Uber drivers, who use their own car. (There is a good write-up of the study by Emily Badger in Wonkblog.) Depending on how much Uber drivers drive, they could still end up with less money than their counterparts in traditional cabs.

A useful piece of information is the cost of driving a car, which Badger's colleague, Andrea Peterson tells us is 57 cents per mile, according to the Internal Revenue Service. Well, this one seems pretty straightforward, if Uber drivers average more than 11 miles per hour, they are less well-paid than their counterparts working for traditional cab companies.

Krueger's study doesn't have data on miles traveled (this is strange, since Uber has this data, at least for the time that a paying passenger is in the car), but it does tell us that the median number of trips per hour is 1.3. We really would want the average here, since we are looking at an average wage pay difference. But if we take the 1.3 median number of trips per hour given in the study, then the average trip distance would have to be 8 miles or less for Uber drivers to come out ahead, assuming they did no unpaid miles.

This second assumption is of course obviously wrong. If an Uber driver take a rider 30 miles from downturn to a suburb, there is a good chance that they will be driving back with an empty car. Also, Uber drivers often cruise high density areas to try to be in line for a call. (This is my casual empiricism from asking the few Uber drivers I have been in contact with.) Anyhow, clearly total miles driven will exceed paid miles driven, which means that the average length of a ride would have to be considerably less than 8 miles for Uber drivers to come out ahead of drivers for traditional cabs.

There is one other item in this mix worth noting. The I.R.S figure of 57 cents a mile is a figure for a commercial driver. It assumes that this person has paid for the necessary licenses and insurance. Most Uber drivers have not paid for commercial licenses for themselves and their vehicles. Most probably also don't carry insurance that covers them for commercial driving.


In this case, their expenses might be considerably less than 57 cents a mile, which could mean that they are still coming out ahead of their counterparts in traditional cabs. Of course in this story the secret is to evade the regulations that are applied to their competitors and to impose costs on the rest of us in having uninsured and possibly unsafe vehicles on the road. (Most cities require cab drivers to meet higher safety standards for their licenses and also more closely inspect their cars.)

If that's the case, this would be a typical story of getting rich in the new economy. Find a way to get around the rules and then claim it as a great innovation. This is a trick that Washington Post owner Jeff Bezos knows well, since his company might well be out of business if it had been subject to the same rules on collecting sales tax as its brick and mortar competitors.