Do taxi monopolies make sense in the high-tech world?

Many cities (and airports) have official taxi monopolies. They limit the number of cabs in the city, and regulate them, typically by issuing "medallions" to cabs or drivers or licences to companies. The most famous systems are in London and New York, but they are in many other places. In New York, the medallions were created earlier in the century, and have stayed fixed in number for decades after declining from their post-creation peak. The medallion is a goldmine for its "owner." Because NY medallions can be bought and sold, recently they have changed hands at auction for around $300,000. That 300K medallion allows a cab to be painted yellow, and to pick up people hailing cabs in the street. It's illegal for ordinary cars to do this. Medallion owners lease the combination of cab and medallion for $60 to $80 for a 7-9 hour shift, I believe.

Here in San Francisco, the medallions are not transferable, and in theory are only issued (after a wait of a decade or more) to working cab drivers, who must put in about 160 4-hour shifts per year. After that, they can and do rent out their medallion to other drivers, for a more modest rental income of about $2,000 per month.

On the surface, this seems ridiculous. Why do we even need a government monopoly on taxis, and why should this monopoly just be a state-granted goldmine for those who get their hands on it? This is a complex issue, and if you search for essays on taxi medallions and monopoly systems you will find various arguments pro and con. What I want to get into here is whether some of those arguments might be ripe for change, in our new high-tech world of computer networks, GPSs and cell phones.

In most cities, there are more competitive markets for "car services" which you call for an appointment. They are not allowed to pick up hailing passengers, though a study in Manhattan found that they do -- 2 of every 5 cars responding to a hail were licenced car services doing so unlawfully.

Some of the reasons cited for the taxi monopoly are as follows:

  1. Without it, taxis would flood the streets, congesting them, wasting fuel.
  2. With too many taxis, wages for drivers would become unlivable.
  3. Regulation is necessary to control rates and assure public safety.
  4. Because you can't shop for a taxi when hailing one, there is a market failure, and rates must be controlled.

Point #2 is fairly weak. The medallion systems simply provide money to the medallion owners. They continue to hire drivers, mostly immigrants, for as little money as they can get away with. Even so, there is a shortage of drivers in many places, indicating market forces set the minimum wage, not regulation. Drivers in the competitive limo industry tend to make similar money, premium ones do better. When New York raised taxi rates, some of the new money went to drivers, but a fair bit went, for no good reason, to medallion owners, who of course increased the lease rates and resale prices of medallions as much as the market would bear.

Point #1 is best left to more detailed essays. Point #3 about public safety makes some sense -- you want to be sure your driver is safe and the vehicle is maintained. You also want to be sure your driver has not been awake forever or impaired. Right now it's hard for market forces to deal with that.

The issue I really want to deal with is the potential for moving away from the "hailing" system to something that allows markets to work better. In particular, I refer to various proposals for cell phone (and pay phone) based hailing, where one calls a Taxi number, and provides one's location and destination, along with other preferences that may stored in a profile. Your phone may transmit your location for you based on its own GPS, but saying it to a computer or human is pretty quick.

Such a system would tie to a network of GPS-linked taxis, and if it could allocate from all suitable (independently contracting) taxis in the city, should be able to assign you a cab that is very close and will come to you in seconds -- at a similar wait time to hailing on all but the busiest streets, and much faster on lesser streets. The cab driver could be provided with your location, and a picture or description of yourself -- perhaps taken by your cell phone camera to show you on the corner you're standing on -- so they can identify and pick you up quickly.

There is a potential for more serious optimization when the destination is also provided, since the system can allocate the right taxi for such a trip, or for the size of a party or how much luggage it has. The largest optimization here is the creation of "jitney" style service, where taxis take on other passengers along the route, with zero or minimal diversion from any given passenger's optimal route, yet at substantial savings. A reputation system for passengers could assure the public of the safety of this.

It could also make possible even more efficient jitney service with jitney transfers. You might find the system suggesting you step out to a different waiting taxi going along your route. Again, the savings could be substantial with minimal waits and minimal diversions from one's optimal route. Combinations of vans, buses, private cars and even public transit could arrange a smooth no-wait trip.

Key here, however, is the customer's ability to choose what taxi company to call to set up a trip. That company in turn might own cabs, or simply contract with and certify subcontractor cabs. (For all companies it remains important that they offer you the largest possible acceptable fleet to assure a vehicle will be near you when you need it.)

If you choose the company, you can competitively shop and negotiate for quality and rates. And thus, subject to minimum standards for public safety, these things would be set by a free market, as would driver wages and company profits.

This seems to me a big win over the monopoly system, and it might convert many users of private cars, or purchasers of second cars, to much more taxi use. That's a win for everybody as it eliminates parking load, and jitney travel actually removes vehicles from the streets. Cities with "central district congestion fees" like London might find market forces pushing for this quickly. It may also make sense to encourage most taxis to be clean vehicles, either hybrid or all-electric with some system for quick battery-swap at a taxi depot. Hybrid already makes a lot of sense, since hybrids do not have to idle while sitting waiting for a fare.

It's even possible our high tech mobile internet world can even make a driver's waiting times be productive. One could imagine a laptop/cellphone station in a taxicab that would allow waiting drivers to do certain types of short-piece jobs -- including taking locations and destinations from customers calling in -- and make money full time.

One unsolved problem is long commutes. Taxis don't do as well when everybody wants a long trip in the same direction. However, this system allows price negotiation, including having higher prices for certain times or types of trips. That's much harder in the current systems though most of them have an added night fee. Prices in this system can adjust to something reasonable for the market. If a commute ride back to the suburbs means no fare for the return trip, this can factor into the price. Trains plus taxis may be the best commute answer. An efficient taxi service solves the train commuter's problem of what to do if they have to leave outside of commute hours, where trains are infrequent or even totally absent. Instead of being faced with a $90 taxi fare, a late commuter might find a $15 or less jitney fare, with transfer right to their door.

It is true that not everybody has cell phones, including some of the poor who can't afford cars and are most in need of a taxi. I think that's changing, and low use cell phones are getting cheap enough that nobody need be without one. Taxi companies could even snap up recycled cell phones and make deals with cell carriers to allow them to be used only to call the taxi company and emergency services, billed on a per-use basis and factored into the fare. The supply of small, recycled phones is already sufficient for this. Those caught without, who can't borrow would have to find a pay phone. Those are dwindling of course but still around. Finally, there's no reason they could not still hail a cab -- but they would get no ability to shop, and have a slightly longer wait of it.

There are privacy concerns here, as there would be far less anonymous taxi travel in such a system. That's not inherent -- companies could promise to erase records and just store a good reputation beside an ID not necessarily linked to your name. One could enter such an ID when calling in, and pay in cash. However, I fear that's not how people would design it. It might however, be worth some privacy risk to avoid the problems of the monopoly. If you transmit a picture you can be identified for pickup without providing your name. The picture can be used to trace you with work, but can't be used easily for routine tracking if it's changed regularly. I would certainly encourage such privacy-friendly designs.

A cell phone hail system has another major benefit. Cabs need not roam the streets at all. Computers would forecast demand, and cabs would travel to selected spots and park. The driver could do telecommute work while parked until a hail comes in. This would vastly reduce emissions from taxis driving around looking for a fare. If somebody has figures on how much time a typical taxi spends driving empty, burning gas and congesting the streets in search of a hail, this might produce a very compelling green argument against hailing.

Update: Let me add some additional thoughts that have come from the comments

  • Computer pricing can be based on the most efficient route or shortest path route to your house, rather than miles driven. This allows the driver choice or use of faster or more efficient highways without it costing more.
  • Taxi companies can make use of GPS and other information to track live traffic patterns and route you even more efficiently.
  • Knowing destination allows mixing in a package delivery business (without the passenger even knowing this is going on) for real savings and profits.
  • Destinations will often be known in advance, allowing for sophisticated planning.

A final note: If cities still think they must have a quota on the number of cabs, instead of having medallions which bring in money for the holders, they should switch to dutch auction of the medallions on an annual basis, putting the money into city tax revenue to pay for roads and transit.


I think there are two main reasons. One is to protect the safety of passengers. Suppose you land at the
airport in Mongolia and some guy comes up and says "here's my taxi; get in; I have a good price". How do
you know he is legitimate? The other reason is to regulate the price. In Germany, for instance, prices
might vary from city to city, but are the same within a city (though they might depend on the time of
day, day of week, season etc). The alternative would be competition, of course, but in practice when hailing
a cab how can you compare prices? Talking to several cabbies lined up at the airport would cause delay in an
area where there is already delay and congestion.

So, at least in Germany, get in a licensed cab and you can be reasonably sure that you are safe and won't
be ripped off.

The whole point of the article is that in the past, we could not shop for cabs while hailing them, but the ability to do high-tech cell based hailing changes that, and lets us have a market again, and a market is, generally without exception, better.

Tourists can't shop as much -- this is a very common problem, and it's why tourists often pay vastly more than locals for all manner of products and services. But even this can improve as tourists get more information, or as the taxi companies offer "roaming" with negotiated rates and reputations.

As for safety, I noted that this could be a reason to licence the cabs, but this in no way requires a monopoly.

I would like to see a minimal version of the GPS hailing as a market experiment.

If a licensed livery company can pick up people who've made reservations but not people who hail them on the street, create a quick reservation system for all livery companies, i.e. everyone but cabs, that allows someone to reserve a car they see coming down the street as quickly as raising your hand to hail a cab.

Comply with legislation by only doing pickups based on reservations, but structure the system so you could operate as if you were a cab. Before any kneejerk legislation is created to preserve the medallion monopoly, what might happen with respect to the 4 reasons cited for the monopoly?

And in fact a friend of mine and I even debated this as an early internet business back in the 90s, though nothing came of it. However, such a business has to start off competing with a the monopoly fleet which its legal advantage, so it has an initial hurdle that would not be there in a free market. But indeed, I think something like this will probably be how it starts. To compete with the hailing system, you need a large fleet to draw on so you can get somebody a cab as quickly. Ideally drivers and cab companies would transmit their location and willingness to take a fare through the network to a database, where any dispatcher they are contracted with can grab them and allocate them to the customer, transmitting a photo of the customer and their location to the screen in the cab.

Give the dispatcher your current location, where you want to go to. He allocates a cab to come and pick you up, drop you off. If you want to get fancier with gps, fine. But nothing beats the flexibility (and in several downtown locations, the speed and convenience) of hailing a cab, and knowing that you're getting a standardized metered rate.

At least in India, there isn't a cap on the number of medallions that can be issued to cabbies. And there's a substantial amount of competition from private radio cab cab operators, who have an association that fixes the price - which typically keeps in sync with "regular" cabs (painted black with yellow tops) thanks to market forces.

One alternative to an expensive and complicated taxi meter and price regulation is the "zone card" currently used (illegally) by gypsy cabs and (legally) in Washington DC. Draw a map of the city, divide it into regional neighborhood zones, and charge based on how many zones you have to pass through to get from point A to point B. If it's LEGAL to provide service on that basis, you can print the zone map on the outside of the cab (as well as printing it on the back of your business card or whatever) so people can compare fares before they get in.

Zone map for Washington, DC.

Alas, I don't think people want to shop for a cab while hailing one. It is not a very efficient market. Cabs may come by, and you can look at their prices on the door, but are you really going to do this multiple times, hailing and rejecting taxis? The drivers won't be very interested in stopping for you to get rejected either.

Now in reality, people would learn the taxi brands in a competitive market, but even if they did are they going to sit there letting Brand X taxis go by in hope of getting a Brand Y taxi? How long will they wait.

Cell phone hail should not only get you the brand you want at the rate you've negotiated, but it should tell you how long it will take, and even let you consider other options, like paying more to get one sooner, or paying less and getting one later, as well as jitney options.

You do the shopping and negotiating not on a rainy streetcorner but at home, on the web, knowing not just price, but reputation and reviews and the works. That's an efficient market.

Zone maps are simple, but of course highly unfair for certain tasks and highly biased against short rides. If you give your location and destination, this allows pricing based on advanced computer street models which can understand the value of time over distance. (For example, regulated price taxis charge by the mile, so you are often better off on slow city streets compared to the fast highways you would be driving if you were behind the wheel. And in fact for the cab, the highway might well be better, even in gas used, but the driver is not able to set that price. Computer models could give you a price based on the most efficient route to your destination but let the driver go any way that makes sense.)

Aren't most taxis hailed by phone in San Francisco anyway. I was down there a month or two ago and took at least two dozen taxis. I think one was a street hail and four were pick ups by cabs waiting outside my hotel. All the others were dispatched by phone.

Seattle is similar. I almost never hail a cab on the street. I use my cellphone, or have the restaurant make the call. Seattle has no restrictions on the number of taxis, but does have liability and other regulations. Every taxi we have seen is dispatched by one of the major taxi companies.

In effect, your call for a cab model is almost here. As for GPS, I'm not sure it's worth it. We often call from one location and walk to another location where we can wait comfortably for the cab. Besides, I'm not sure how well GPS works in the city anyway. It sure doesn't work when there are trees around, so buildings might interfere as well.

Jitneys might be a good idea for commuters, especially if you already have a transit system, but anyone who has taken one of those airport shuttles knows the problem with door to door shared drop off. Going the last few blocks from hotel to hotel often takes as long as the whole haul from the airport into town.

My proposal is to make cell phone hail be as fast as -- and sometimes faster than -- street hail. But with reputation, security and the rest.

Yes, shared airport shuttles are not a great experience. With the airport it's made worse by having to unload bags, and arrange payment. In a system like this payment would be automatic from an accout, and people not doing airport trips would not have bags, often getting out at stop lights and with no diversion from the optimal route for the other passengers. Only true computerized coordination of trips can make the last one happen.

As I see it a jitney trip might involve taking a private cab to some suitable "transfer station" where one moves into the next cab with minimal wait. Likewise the last mile of the trip might also be in a private cab, to avoid taking the other passengers out of their way. This becomes more workable if the company can negotiate fees and there is not a mandatory flag drop (which makes 1 mile trips less economical.)

For #1 (congestion), greater availability/convenience of taxis gets other private vehicles off the road. There might even be a greater than 1::1 tradeoff -- one more operating taxi eliminates more than one private trip. (Think of the time savings from not needing to park, for one.) Also, if congestion is the issue, capping just one contributor -- taxis -- while leaving other factors uncontrolled is crazy. If congestion is the concern, target a policy at congestion -- such as congestion-zone tolls, traffic-jam-taxes, etc.

For #2 (livable income), you nail it: medallions boost the income of medallion holders, not drivers in general.

For #3 (w.r.t 'public safety'), I'm surprised no one has pointed out: you can set minimum standards of equipment, service, and driver skill *without* capping supply. That's how most other regulated industries work.

For #4 (and part of #3, markets can't set rates), there are both low-tech and high-tech solutions to allowed informed rider decisions.

Low-tech (or rather, no tech beyond what's in taxis today): You could regulate the 'unit' of service -- a pickup is 3 units, each 0.2 miles is 1 unit, each 5 minutes in traffic is 1 unit, whatever -- then let every fleet or even cabbie set their own price per 'unit', a simple single number to compare. (And, that can vary by time-of-day, weather, etc.)

Or: you could cap the number of different pricing schemes at some smallish number (3-5), but let them vary completely by operator choice at occasional intervals, and require cabs to clearly indicate which scheme they use -- ideally by their color. Regular cab users (and tourist guide books) would quickly establish some rules of thumb making choosing cabs and knowing expected fares possible, but there would still be a mechanism for price variety and competition.

(High-tech, it could all be displayed/explained/calculated/negotiated/paid by your phone.)

...for a dynamically-dispatched taxi- and jitney-centric urban transit network for around a decade, but I imagined it slightly different:

(1) You press a button on your terminal (mobile phone, 99.99% of the time) saying, 'I need to get somewhere'

(2) You pick your destination -- usually from a list of your top destinations and common endpoints, or adjusted from those as necessary.

(3) This 'want' (a request-for-bids) is posted to all alternative service systems, and they come back with options (bids), listed by price and travel time. Travel time estimates are *very* accurate, based on historical data, current fleet positions, and latest traffic info. So your list might be:

(a) $19 taxi picks you up +3 mins, reach destination +15 mins
(b) $10 walk 2 blocks east, get jitney +5 minutes, 1 stop, reach destination +22 minutes
(c) $1.50 walk 3 blocks north, wait 13 minutes for train X, 5 stops, walk 2 blocks west, reach destination +35 minutes

(4) You pick one (accept the bid). Payment is automatic with partial or full refunds if travel time promises are not met. Your handheld authenticates the vehicle that picks you up and reports back pickup/travel/dropoff data to the necessary clearinghouses.

I do believe there would be a strong market push for simplicity, and most customers would not want to perform a complex bidding operation like this. However, a free market would allow providers to experiment with it. The pricing models that customers like would quickly become the winners.

I don't actually think there would be a really serious variability in the real cost, since the base cost is fairly clear -- a low-skilled worker's time and a vehicle cost per mile, plus accounting for time and miles spent idle/driving empty. In an efficient market, profits would remain in a modest zone and the cost will get close to this. Variation in price would come instead from innovation -- such as really good jitney service, better placement of cabs to be close to expected customers, more efficient use of idle time.

Other possible ideas include merging with package delivery. In many cases you could provide door to door service for humans who don't even know you are also carrying some less-impatient packages along roughly the same path. Even if the cab makes stops for package handoff those would be brief and need never take the passenger off her shortest route. This is also greatly facilitated by planning based on knowing the destination point.

There's also the potential for good use of pre-planning. People at a theatre can reliably inform their taxi dispatcher they will want a cab from the theatre to home hours in advance. Even people in shopping and meetings can usually make predictions that they will be leaving "in 10 minutes" or even pre-enter their route and an approximate time, then update the time with quick notes.

Note there is one item of merit in the anti-congestion argument. Today's taxis spend time idle, driving around. This is not something private cars do. More idle taxis does mean more congestion and more pollution, so you don't want an oversupply. However, my plan eliminates this concern, because idle taxis -- especially if they are hybrid or electric -- would sit parked, with gasoline engine off. In fact this may be the most compelling argument to the public for ending the hailing system -- calculate just how much gasoline is burned having taxis roam the streets looking for hails.

I love the idea of selling both fleet expansion and digital hailing as an environmental boon: getting both private cars and circling taxis off the roads.

Maybe only hybrid/electric taxis would be allowed to cruise for fares. Gas taxis would have to stop and wait for a dispatch.

Add new comment