The Luxembourg Competition Authority recently handed down a decision that found an app-based taxi booking system, Webtaxi, was not a hardcore violation of the relevant competition law banning price fixing.  The algorithm determined the precise fare the passenger would pay for a trip.  The taxis remained competitors otherwise and the cabs on the app represented only 26 percent of the relevant taxi market.  Fares were otherwise negotiable.  The Authority found the efficiency gains material and the pricing reasonably necessary to obtain the gains.   Specifically, they found the app would reduce fares, reduce empty taxis, reduce pollution, and reduce waiting times.  They also found that the collective price setting was “necessary” to achieve these goals.   Absent the price fixing, customers would not choose the nearest taxi but the one with the best price.

This “venture” would be condemned as naked price fixing in the States. The only integration the venture offers is the app.  The only thing the app does is set the fare and allocate the customer to nearest physical taxi participating in the app.  At 26 percent of the market, the conspiracy would likely fail as consumers could switch to non-participating taxis relatively easily, but that doesn’t mean it isn’t a conspiracy of “collusion.”  The US has regularly, and for over a century, condemned price fixing agreements that set “fair prices.”

Uber and Lyft are different than Webtaxi. They are more than mere aggregations of independent drivers (notwithstanding what they may claim about the independence of their drivers and the view of the lower court).  Uber sets their prices.  Uber provides them with the technology and branding as well as a widely distributed app that is easy to use.  They provide a rating system.  And there is system competition.  Incumbent taxi companies.  Lyft.

The efficiency analysis is flawed. Less pollution is a socially beneficial outcome but it is not an economic efficiency.  We would leave the decision as to what is better—a ban on price fixing versus less pollution—to the legislature, not the competition authority.  And the conclusion that the price fixing is necessary is also flawed.  Webtaxi could have just as easily included a bidding mechanism where drivers would bid for a ride and the driver would pick the best fare.  Or the drivers could set their standard fares, and the rider could pick the one that suited it the most.  The one that was near and more expensive because she had to get to her destination with dispatch; or the cheaper ride that was longer.

In fact, I would not even characterize this as algorithmic pricing. Algorithmic pricing is where a single competitor programs its pricing mechanisms to take into account several data points including, potentially, competitors’ pricing.  In certain circumstances, pricing can harmonize but only because the structure of the market is conducive to parallel pricing, that it’s sufficiently concentrated to allow for parallel pricing.  Here, there is an actual agreement between competitors to set a collective price without supporting efficiencies and integration.

American regulators looking for precedent should not rely on the logic of this decision. A classic collaboration analysis is appropriate:  is there actual integration beyond just the allocation of customers and setting of price; can the efficiencies be obtained outside of the venture; and is the price fixing actually necessary to obtain the efficiencies.