It’s a small market, after all. Es gibt einen kleinen Markt, uber alles.

The taxi industry, berated as “Big Taxi” by Uber and its allies, is in many cities a vested cartel with medallions limiting entry. The nominal economic reason for the medallions are (1) too many taxis roaming the streets creates congestion (since roads are unpriced), and (2) too many taxis drives down wages, and (3) regulation is important for public safety. Limiting medallions of course has the happy side effect (for the taxi industry) of limiting competition.

In the early 2010s a new breed of taxi provider, ignoring the law, insurance regulations, safety, and in some cases, common decency, entered the market with a technological advantage: an app, and a new source of supply: regular cars and their otherwise underemployed drivers. Backed by oodles of venture capital money subsidizing rates, they have entered the market in many cities worldwide.

Observation 1: The new taxi providers own very few assets. They don’t own cars, for instance. Uber is under-insured. Drivers of so-called Transportation Network Companies (TNC) using their own car (UberX) almost assuredly are not paying insurance rates that would be charged for someone carrying passengers for hire, or driving so many miles per year (insurance forms typically ask for miles per year, but this is lagging and voluntarily supplied). They are operating without authorization in many places.

Observation 2: The app (and back end server) is easily replicable (evidenced by the fact that it has been replicated dozens of times: Lyft, Gett, Curb, Hailo, Blacklane, Sidecar, Zimride, iHail, Flywheel, etc.). Fixed costs are not a major issue in this market.

Observation 3: The new taxi providers are very highly valued by VCs (Uber is being valued at $40B),  who believe there is a first mover advantage not only in this taxi market, but in related markets for delivery and logistics which the new firms will capture.

Observation 4: This belief is likely due to the belief there are network externalities in two-sided “matching” markets, so that riders will go with the app with the most drivers, and drivers will go with the app with the most riders. This is the eBay logic essentially. eBay is at the time of this writing capitalized at $68B. This will drop once PayPal is split off.

Observation 5: While riders will only use one app per trip. There is nothing preventing drivers from using multiple apps.  Most of the drivers I have seen have multiple phones going, one mounted for GPS, one for other transactions.  This is fairly trivial and very low cost. Unless a driver is completely utilized by a particular service (and that is the best-paying service), the driver has no reason NOT to use multiple apps. This differs from eBay in that sellers are only selling a particular commodity on one auction at a time (otherwise risk not being able to deliver) [update: though vendors may have identical goods listed in multiple marketplaces], while buyers can search multiple sites before bidding, but will not generally bid simultaneously on the same good for fear of winning them (someone with eBay knowledge can perhaps determine how often people simultaneously bid).

Observation 6: All companies want monopolies (See Thiel, or Marx, or Pixar). No capitalist wants a free and fair and competitive market for themselves (though they do want it for their vendors and customers). If there are to be rules, they want to make sure they are not disadvantaged.

Observation 7: The reason there are not monopolies in the rail and aviation sectors in the US is because of antitrust law. The reason there are not monopolies in cell phones is because of antitrust law and initial auctions of spectrum which constructed the markets. None of these markets is a “natural monopoly” though. While there are of course economies of scale, returns also diminish with scale, so that at the margins, a larger firm does not produce at substantially less cost than current sized firms. There are fixed costs, but they are limited.

Observation 8: The basic reason there are monopolies in wired utilities (phone, cable … though note this is breaking down), electricity, natural gas is because of the high fixed cost of running infrastructure to every home, which could not be recovered if people subscribed to different utilities (reducing the base over which the fixed costs can be spread). Moreover, the public does not want duplicative services tearing up the streets, etc. Roads themselves are a natural monopoly at the local level, we don’t have competing streets.

Observation 9: Taxis are not natural monopolies, and in most markets, the places where you summon taxis (as opposed to hailing them on the street … i.e. everywhere except downtown), there are loads of local firms willing to provide the service. Look up “taxi” in your local yellow pages. (Or should I say Google it … ). Yelp gives me 22 pages of taxis in Minneapolis, of which 18 firms have at least 1 rating. Uber is at the top of the list, with 25 reviews, but only 3 stars. People seem (a) to be confused between Uber Black Cabs and Uber X and (b) not like surge pricing.

Observation 10: Most places are not downtown.

Observation 11: Most people in the US and globally don’t live in places where you hail cabs. If you want a taxi, you request it by phone or app.

Observation 12: Most people care as much about reliability as about travel time.

Observation 13: A 5 or 10 minute delay from calling the cab is not terribly important so long as the taxi in fact does arrive. The nice features of apps with built in GPS is you can see the taxi approaching you, so people have trust in the app-based system, but this is transferable between apps, not a function of the particular app vendor, since the apps are undifferentiated.

Observation 14: The number of taxis within 10 minutes is a function of 100*pi, while taxis within 5 minutes is a function of 25*pi (remember Area = pi*r^2), so there are 4 times as many taxis within 10 minutes as within 5 minutes. Very few people who are very concerned about the last 5 minutes of travel time don’t have a car already or won’t plan ahead a few minutes before departing their previous destination. Smart phones enable the last 5 minutes to be useful enough if you are not doing something at the current place. All TNC users are smart phone users. They are not sending telegrams or couriers to request hackneys.

Observation 15: Most people in the US have cars. Certainly this is peaking, and changing with all of the factors affecting peak travel.

Observation 16: Alcohol use is declining among youth. Among everyone, beer is down, wine is up (NIH: PDF). Certainly increased convenience from taxis may increase public drunkenness, but one of the best market for taxis (taking drunk people around on weekend evenings) is on the decline. Airports (another great taxi market) have peaked as well.

Observation 17: UPS and FedEx have great logistics systems and know what they are doing. Amateurs taking packages around is certainly feasible, but not terribly efficient, given economies of scale in national shipping.

Observation 18: Local purchases may be a possible source of delivery, but this is a fairly small market in most places (if it is local, aside from food, why would you order online, most online orders are going to  be national vendors due to economies of scale trumping the need for instantaneous delivery). Really, I can wait 2 days for most things. Restaurants and supermarkets have their own logistics systems that cannot be easily defeated by amateurs (restaurants are more easily switchers than supermarkets perhaps, since the organizations are much smaller, but that is what Takeout Taxi etc. are for). Really, there are special skills in transport and delivery for specialized goods, that someone who does it only once in a while won’t get. Only a macro-economist would not realize that.

Observation 19: Uber provokes hatred unlike most internet companies. See #baduber.

Observation 20: Apple  could easily place an app for an Uber competitor or competitors on the home screen, or embed it in maps, if it is concerned about Uber getting too powerful.  ApplePay is already built in. Google with Android could as well, but they won’t since they invest in Uber. There is nothing more convenient than already on the home screen.

To quote Paul Gigot “Free riding is not socialism”. Uber is exploiting the existing system as they should as putative monopolists. Lawmakers and Prosecutors are ineffective in keeping them in check.

The debate over Uber is complicated.



One thought on “It’s a small market, after all. Es gibt einen kleinen Markt, uber alles.

  1. Excellent observations.

    In capitalism free flowing capital drives excess returns on capital to zero.

    The internet serves as a disintermediary for many businesses – turning amateurs into passable service providers.

    Uber and other ride sharing/car sharing services many change the world by driving down transportation costs and increasing accessibility but investing in them looks like a hard way to make money.

    Rent seeking interests will defend their turf.


Comments are closed.