Wednesday, May 29, 2019

Uber and wages in a free economy.

Here was a recent article about Uber and Lyft drivers in Washington, DC, colluding to game surge pricing at the airport.
Every night, several times a night, Uber and Lyft drivers at Reagan National Airport simultaneously turn off their ride share apps for a minute or two to trick the app into thinking there are no drivers available---creating a price surge. When the fare goes high enough, the drivers turn their apps back on and lock into the higher fare.
It's happening in the Uber and Lyft parking lot outside Reagan National airport. The lot fills with 120 to 150 drivers sometimes for hours, waiting for the busy evening rush. And nearly all the drivers have one complaint:
“Uber doesn’t pay us enough, what the company is doing is defrauding all these people by taking 35-40 percent,” one driver told ABC 7.
There is a lot going on here.  Really, these drivers aren't colluding against Uber and Lyft.  They are colluding against the customers, who must pay surge pricing.  Uber and Lyft must compete against each other for riders, which drives their fares down to the competitive level.  The drivers are actually colluding so that they and the firms can claim monopoly profits from airport customers.

Their complaints are against the firms, but really, the culprit is competition, which prevents both them and the firms from boosting their incomes at the expense of riders.

In fact, their complaint against the firms is even more misguided than that.  The firms are charging riders a competitive rate and they are overpaying the drivers.  This is a classic economic problem.  There is a queue at the airport.  Those drivers are choosing to go sit in line at the airport instead of driving around the rest of the city picking up riders on the go.  And the reason is that, at standard rates, airport rides are more lucrative for them.  The reason for a queue, conceptually and in this particular case, is that the price is too high.

If the price was too low, you would have a queue of customers, like during the oil shocks of the 1970s when price controls were put in place.  Here, the price Lyft and Uber pay to the drivers at the airport is too high, so the producers (the drivers) are queuing.

Paying drivers more would only make this problem worse.  If they are waiting for an hour to get a fare now, then if the typical fare doubled, drivers would wait for two hours.  Uber and Lyft aren't determining the hourly wage for these drivers.  They are determining it by deciding to wait in line.

The only other way for Uber and Lyft to solve this problem would be to ration the supply of drivers in some other way.  In a way, this is one reason drivers might want to be classified as employees instead of contractors.  If Uber and Lyft treated drivers like employees, they would manage how many drivers there were and where they drove.  They could eliminate the queuing, which would raise wages and reduce the waste of queuing, but it could only happen by being a gatekeeper.  The only way to get rid of the queue would be to tell some of the potential, qualified drivers that they aren't invited any more.  They aren't "hired".

This is similar to the issue of minimum wages.  The way this raises the wages of some is by eliminating the wages of others.

That isn't all bad.  Here, it would lead to less waste by eliminating over-long queues.  But, small scale gains due to monopoly power or economic rents don't add up to social gains.  Everyone can't earn more than the competitive income by using market power to impose exclusion.

The queue is wasteful, but I'm not sure there is a solution.  The economics of driving basically will always come down to queuing.  Whatever rate Lyft and Uber pay, whether drivers are sitting at the airport, or driving around town, the economic breakeven for the drivers will be a function of queuing in some way.  It will determine when and where they drive.  In any given part of town, how long do they need to wait to get a rider, how long do they need to drive to pick up the rider, and how long will the average ride be?  That equation comes down to how much time is a rider in the car versus how much time is the car empty.  There are several supply and demand variables that lead to an equilibrium level for any particular location, but in the end, that equilibrium will be driven by the willingness of drivers to queue in order to get a fare and it seems that some queue, such that it is, will remain wherever Uber and Lyft set their fares and their driver reimbursement levels.  Limiting the number of drivers at the airport queue, where the extra 50 or 100 cars in line has little effect on the quality of service, may seem like a no-brainer.  But, trying to reduce queuing out in the marginal markets around a city will change the supply and demand dynamic in a way that will lead to deadweight loss on the margin.  Reducing the number of drivers will necessarily increase wait times for riders, changing demand for drivers.

I am sure there are teams of economists working on this problem at Uber and Lyft.  I suspect they don't so much mind being tricked into surge pricing at the airport.  They certainly aren't going to raise driver payments in an attempt to address the issue.


  1. Excellent analysis.

    As I often say, the root cause of many problems in the United States is that we have kept labor markets artificially loose in combination with a de facto open-borders policy.

    The Fed has openly targeted a 5% unemployment rate as measured by the u3, and just recently Fed economists authored a paean to the 5% unemployment rate.

    The results of artificially loose labor markets and artificially high housing costs... well, AOC has a shot at the White House, and Trump flattened the GOP field in 2016.

    Get ready to rumble!

    1. The immigration policy doesn't have much to do with this. Labour demand is very elastic, ie in the longer run the labour market just absorps more people at nearly the same price.

      You are right that tight labour markets would be good. Apart from ensuring the right level of aggregate demand via monetary policy, the rest is mainly about supply side policies that make it easier to employ people.

      Eg lower taxes and mandatory fees on labour especially at the low end, making hiring less risky for employers (by eg making firing simple). No ban of low wage labour. Etc.

      International comparisons are useful here, and also comparisons over time.

      There's a lot of learn from how eg (West) Germany went from a heavily regulated rationed economy to the low unemployment of the Wirtschaftswunder. And then back to different kinds of heavy regulation and 10% permanent unemployment until the late 2000s, and back to full employment again.

      Similarly, how does Japan keep unemployment so low?

      Pseduoerasmus has a long and interesting article comparing textile labour in India and Japan:

      And an addendum

    2. Dude: We spell that "labor."

      "in the longer run the labour market just absorbs more people at nearly the same price."

      Well, maybe. How many decades have you got? And if the inflows of labor are continual?

      Listen, I was an employer and employee in Los Angeles. There were large inflows of low-cost labor for decades. I believe in the laws of supply and demand---more supply, and you drop the price.

      The "Americans won't do this work" proves out the canard. That means the price should be higher.

    3. Studies of the bracero programme and the aftermath of its abolition suggest otherwise.

      Lots of then time when the locals won't do the work, and you don't have foreigners, nobody will do the the work.

      (In that case the farmers switched from fruit to less labour intensive grains. Total local employment in agriculture fell as well, if memory serves right.)

      Btw, in eg Silicon Valley the prevalence of lots of other programmers is exactly what makes them productive and hence expensive. The benefits of agglomeration and networks. Similar agglomeration benefits accrue to cities in general. That applies no matter whether the additional people are born locally, move in from the same state, same country or same planet.

  2. Limiting the number of drivers would shift the queue from waiting at the airport towards waiting to be able to sign up with Uber.

    Of course, there's always an alternative to queues: actions.

    Uber could have drivers bid on trips, lowest bidder wins. No more queues at the airport.

    1. Should have been 'auctions', not 'actions'. But was probably clear from context.