Uber Challenges Study Suggesting Its Drivers Earn $3.37 Per Hour (reuters.com)

An MIT study using data from more than 1,100 Uber and Lyft drivers concluded they're earning a median pretax profit of just $3.37 per hour. But now Reuters reports: Uber Chief Executive Dara Khosrowshahi criticized the MIT study in a tweet on Friday as "Mathematically Incompetent Theories (at least as it pertains to ride-sharing)," and linked to a response by Uber chief economist Jonathan Hall that challenged the study's methodology. Hall's rebuttal to the study said the likely misinterpretation of a survey question and the study's "inconsistent logic" produced a wage result that was below similar studies elsewhere. He said the study used a "flawed methodology" compared with a survey that found drivers' average hour earnings were $15.68. "The earnings figures suggested in the paper are less than half the hourly earnings numbers reported in the very survey the paper derives its data from," wrote Hall.

The MIT study's lead author, Stephen Zoepf, told Reuters in an email on Saturday, "I can see how the question on revenue might have been interpreted differently by respondents" and called Hall's rebuttal thoughtful. "I'm re-running the analysis this weekend using Uber's more optimistic assumptions and should have new results and a public response acknowledging the discrepancy by Monday," he wrote.
Saturday Uber's CEO tweeted a thank-you to MIT, "for listening and revisiting this study and its findings. Right thing to do."

Comments

This is the way it's supposed to work

[Solandri]

Nice, calm discussion about the merits of an argument. Listening to and understanding the points made by those with an opposing viewpoint. Assessing the points they make in an unbiased manner. And acknowledging when they may have a legitimate point, and re-doing your work to adjust for it.

Instead we mostly have people shouting at each other, refusing to listen to or even to interact with each other simply because they have different viewpoints. Because both sides "know" that their side is right and the other is wrong.

The $3.37/hr wasn't what caught my eye

[rsilvergun]

it was the part where 30% of the drivers made nothing when maintenance was factored in. I've heard Uber called a payday loan on the maintenance of your car.

I'll say this, I've taken 5 Ubers in my life and 3 of them were recently laid off folks trying to make rent with cars bought from when they were employed.

Is it just me

[quantaman]

Or is the Uber economist's more confusing than it looks at a glance?

At a high level his argument seems to be that MIT analyzed some data from this survey and came up with an inappropriately low number.

The error the MIT paper is being accused of making is the survey tries to get 3 numbers:
1) How many hours a week do you work on a ridesharing service?
2) How much money do you make in a week?
3) What percentage of your income comes from ridesharing?

So the $3.37 is basically (#2 / #3) / #1 (presumably minus expenses).

The Uber economist claims that respondents actually interpreted the first question as:
1) How many hours a week do you work total?

So if you worked 1 hour ridesharing and made $20, then you worked 40 hours at a regular job then the $20 you made ridesharing would be divided by 41 instead of 1 to give you earnings of ~$0.50/hour.

The problem is the survey questions look a bit odd and the Uber economist claims an even odder interpretation:

Q11: “How many hours per week do you work on average? Combine all of the on-demand services that you work for.”

Q14: “How much money do you make in the average month? Combine the income from all your on-demand activities.”

So the Uber economist claims that the MIT authors interpreted #11 to mean hours only from on-demand services and #14 to be money from all sources. Which is a bizarre interpretation of questions with almost identical questions wording, yet instead of pointing out how weird the interpretation is the Uber economist actually seems to imply that their interpretation is correct and it was the respondents who misunderstood. Which makes me wonder what the actual sequence and context of the questions looked like.

Unfortunately he doesn't link to the study and the survey questions seem to only be available if "If you’re a media member and would like a full question list and anonymized data/calculations used in this report, please make a request here.

The MIT authors seem to be taking the criticism seriously so maybe I just suck at reading, but given the nature of the claimed error it seems like he should be able to make a much clearer argument.

Somewhere in the middle

[dave562]

I drive for Lyft sometimes, but the pay is similar. I earn anywhere between $8 and $20 an hour. Mostly it averages in the $10-12 range. (In the Portland, Oregon area.) I have only been doing it for a few months. I have talked to some guys who were doing it when ride sharing first came to the region, and they said that they were making around $800 a night on the weekends. I pull around $200 for ~6-8 hours on Saturday. Any other night I am lucky to get between $75-100.

I do not understand how people can try to make a living doing it.

I do it for a few reasons. I like to drive and talk to people. It gets me out into the real world and off of the computer. I also appreciate that I can write off car maintenance, tires and things like that. I would also be able to write off my cell phone, but my main job already pays for that.