More evidence of a distorted taxi marketplace

NYTimes: 2 Taxi Medallions Sell for $1 Million Each

Why are taxis so expensive?

I wonder this every time I’m planning a night out with friends. It’s stunning. TaxiFareFinder.com quotes a taxi ride from the University District in Seattle to Downtown Bellevue at $28.59, or about $1.70 per minute including a 15% tip. At that rate, an hour’s worth of travel would cost about $101—a rate closer to some lawyers’ fees than to other, comparable point-to-point transportation services. Zipcar, a subscription-based car rental company, charges $7.75 an hour for its basic vehicles, which includes fuel and insurance. Zipcar subscribers must drive themselves, of course, but the cost gap between a car-sharing service like Zipcar and a taxi ride—over $90 per hour, based on my calculations—seems too large to be explained solely by the driver’s wage. According the U.S. Bureau of Labor Statistics the 2010 median wage for taxi drivers in the Seattle metropolitan area was just $10.39.

It’s a shame that taxis cost so much. Cabs fill an important niche in the ecosystem of transportation choices, providing on-demand, point-to-point service, useful for time-sensitive trips, or those involving cargo. If taxis were more affordable, more people might choose to use them instead of committing to the fixed costs—and temptation to drive—presented by an extra car. Having traveled overseas, I know that paratransit (the variety of services that fall between the private automobile and the conventional bus) can represent a much bigger slice of the transportation pie than is the case here in Seattle. In Mexico City 73% of all transit trips are completed using paratransit. In the Puget Sound region, taxi, shuttles, and wheelchair-compatible vans serve just 1.5% of all trips. I am guessing that for most people, taxis are too expensive to be useful as anything beyond the occasional high-value trip.

To satisfy my curiosity about taxis’ price point in our part of the world, I started looking deeper into the issue, suspecting that our regulatory environment might contribute to their cost. Seattle, like many U.S. cities, limits the number of taxicab licenses. In 2010 there were 659 outstanding licenses, or about 1.1 per 1000 Seattle residents. Basic microeconomic theory holds that in a perfectly competitive marketplace, supply limitations distort the supply curve and lead to higher prices for consumers. In the conceptual graph below, the solid magenta line shows the how the market might respond if entry into the taxicab market was unrestricted: as the market rate for a cab ride increases more people start driving, attracted by the potential profits. Eventually supply and demand come into balance, yielding a price of P1.

If we restrict the supply of taxicabs by limiting the number of licenses, then we might experience the situation represented by the dashed portion of the magenta line. Past a certain quantity of taxi service demanded, the market cannot respond with more cabs (because the City won’t allow it). So the supply curve goes vertical, and consumers pay the higher rate represented by point P2. The horizontal difference between points 1 and 2 is what economists call a shortage.

Shortages reduce the overall well-being of consumers (their “utility” in economics-speak) since the service is now more expensive than it should be. That leaves buyers with less money for other things they might want or need. Some cab drivers will enjoy a windfall, assuming they were able to get a license cheaply. But as time goes on, drivers are no longer buy a license from the city; the city-bought licenses, available for a nominal fee, have been snapped up. The only licenses a newcomer can buy are from other cab drivers in the secondary market. The price of those licenses rises until the cost balances out the wage gains resulting from the restricted supply (currently, licenses in Seattle can cost up to $100,000). If we were to tally the utility losses suffered by consumers and the utility gains made by those lucky early cab drivers, we would find that the losses outweigh the gains, meaning that society is overall poorer as a result.

Of course, all of this depends on an assumption that I slipped into the beginning of this discussion: that the taxicab market is perfectly competitive. A perfectly competitive marketplace exists when

• There are enough buyers and sellers that no individual can affect market price
  
•  Firms (in this case license holders) are free to enter and exit the market
  
• The product is homogeneous
  
• Buyers have perfect information about the product (like the comfort and safety of the ride).

In the real world, almost no industry meets all of these conditions, although a few, like the agricultural commodity sector, come close. Market failures occur when a given marketplace does not meet one or more of these conditions, and the presence of market failures is a commonly used rationale for government intervention in a sector. If the taxicab market indeed suffers from market failures, then the regulation we see today might be justified. I was particularly interested in what market failures might explain our prohibition on new cab licenses in Seattle.

In the coming paragraphs I will show you that the taxicab market indeed suffers from market failures. And while many of the regulations currently imposed on the taxi market can be explained by these failures, I am not convinced that supply limitations—my prime suspect for the high cost of taxis—are justified.

It turns out that when we refer to the “taxi market,” we’re actually talking about three related sub-markets. The cruising or hailing market consists of taxis that drive the streets looking for customers (or fares). The fare hails the cab, which pulls to the curb and picks them up. The taxi stand market consists of rows of cabs that wait for customers, usually outside major origins like train stations or airports. By tradition, the drivers guide the fare to the first cab in the line, after which the remaining cabs pull forward. Finally there is the dispatch market, in which customers call in a request for a cab. The dispatcher sends a cab to pick up the customer. These three markets have distinct economic characteristics that prevent them from functioning as perfectly competitive markets in the absence of regulation.

In the cruising market, an individual cannot easily compare rates between taxis. Waiting for a second taxi to drive by takes time. The customer does not know whether that second cab is nearby, or how long it might take to arrive. The customer cannot keep the first cab on hand, in case it turns out that the second one costs more. Then there is the awkwardness of turning away the first driver, if the second turns out to be a better deal. Without the ability to easily compare rates, the customer has no way of knowing whether the advertised rate is reasonable. This lack of price information is a market failure known as costly search, and it violates the perfect information condition of the perfectly competitive market, described above. This can lead to monopolistic pricing practices. Indeed, at the moment that the fare is considering whether to hire the taxi, there is only one firm ready to provide the service. Not only does the unregulated taxicab operator have little incentive to lower their prices in this situation, they in fact have an incentive to increase rates late at night, for example, or during foul weather.

One might expect that the taxi stand would be free from the costly search problem, and it is true that customers in that market segments can more easily compare prices. But there are barriers to competition here, as well. At the taxi stand, customers are pressured to take the first cab in the line. As long as this tradition persists, comparison shopping is at least awkward, or at most impossible if the drivers refuse to break ranks.

Of the three market segments listed above—hailing, stand, and dispatch—the dispatch market is the one least afflicted by market failures. Assuming that there is more than one taxi company to choose from, the customer is free to search out the best deal by calling around or using the internet. The ability of the customer to comparison shop should theoretically bring down prices by rewarding those companies and individual operators who, through efficiency improvements, are able to offer the lowest rates. But the dispatch market has flaws as well, flaws that are held in common by all three markets. First, a second information problem arises from the characteristics of the service itself. A trip in a taxi is known as a “credence good”—a good the quality of which is difficult or impossible for the customer to gauge before the purchase. For example, a given cabbie might be a frightening driver, but you wouldn’t know until you began the trip. Branded fleets (Yellow Cab, Checker Cab, etc) go some way toward ameliorating this by introducing accountability—you can call the company and complain—but the root problem remains. Second, the literature suggests that in taxi markets where pricing is not regulated, operators tend to adopt complex pricing structures that make it difficult for customers to compare costs.

In addition to the issues above which interfere with competition, all three taxicab markets exhibit externalities, or impacts on society that are not reflected in the price of the good, assuming that the market is free to set the price. Externalities are another kind of market failure that governments use to justify intervention in markets; when left uncorrected by public policy they reduce the overall well-being of society by encouraging too much or too little of a good. A positive externality is present in the dispatch market: the larger the fleet, the more valuable it is to a potential customer, since there is a greater chance that a member of the fleet will be nearby when the customer calls. Without intervention—such as a requirement that all cabbies be a member of an association—these fleets will not reach the size necessary to provide the maximum benefit to consumers. Negative externalities include collisions caused by taxicab drivers and congestion, although I would argue that taxi trips tend to be high value, and that if regulation is warranted to prevent congestion it should be targeted instead toward private leisure trips. In addition, the argument that taxis cause congestion should be weighed against the increase in private auto trips that may result from a limitation on the number of taxis.

So far we’ve covered two kinds of market failures that spur regulation in the taxicab industry: competition flaws and externalities. Here are a few other issues that don’t fit neatly into those two categories.

Driving a taxicab requires relatively little specialized skill, and so incumbent drivers in an unregulated environment face stiff wage competition from entrants. Some claim that the low wages resulting from competition are unfair to incumbent drivers and are a justification for limiting the number of taxicab licenses. The idea is that each driver gets a bigger slice of the “pie,” the pie being total industry revenues. But in theory, the act of liming supply will shrink the size of the pie, even as each individual driver gets a bigger slice. This has ramifications for the public and the economy. With fewer drivers on the road, we face longer wait times and lower business tax revenues. The governmental limitation of competition in order to advance the commercial interests of a special group is termed “regulatory capture” by economists, and regarded as a form of government failure that reduces the overall economic wellbeing of a society.

As mentioned above, once a system like this is put into place, dismantling it is difficult, since new cabbies will have sunk considerable sums into buying their licenses.  As I write this, we are seeing an example of cab drivers fighting to retain such a protectionist policy: the Greek state is attempting to open its taxicab markets to competition, a condition of the bailout loans, apparently, and in response cab drivers are staging strikes and rolling protests across Athens. Most paid dearly for the licenses they own now, licenses that would be next to worthless if issued freely by the government. This issue of economic incumbency is the heart of the matter for income restrictions, and I will return to it later.

Non-market aspects of the taxicab market are also used as justification for regulation. Entry regulation is regarded as a way to limit aggressive sales behavior by competing cab drivers. Also, in theory, unregulated taxis will gravitate to high yield locations and to ignore areas with low demand, a practice known as “cream-skimming.” This practice interferes with the social goal of cross-subsidizing trips in low-demand areas with profits from high-demand areas, and can be the target of regulation on equity grounds.

As I said at the start of this post, taxis are heavily regulated in Seattle, and in many places around the world. With the economic characteristics of the industry’s three sub-markets on the table, let us now take a look at some of the most common forms of regulation, and see how well of a job they do at advancing the public interest.

First, we have safety and quality standards for taxis and taxis drivers. These include special driver safety training for cabbies, vehicle inspection regimes, and the requirement to belong to branded associations. The latter seeks to promote standards of conduct and to reduce wait times through economies of scale. In my reading I found that economists were supportive of these sorts of regulations, since they give customers a measure of certainty about the sort of ride they are able to take. Since vehicle safety and driver competency are otherwise invisible to the prospective customer, the government intervenes, and communicates the results of safety and quality testing to the customer through an official license on the dashboard.

A second common form of regulation on the taxi markets is the imposition of fare controls. Like control on supply itself, interference in suppliers’ ability to set prices can lead to surpluses or shortages: think gas lines in the 1970s. So I expected to find that economists were united against the idea of price controls in the taxi market. Not so. Some support governments’ efforts to establish uniform prices, the rationale being that in a free market, cabs tend to adopt confusing pricing structures that are difficult to compare, and that cab drivers gouge customers when they are desperate for a ride. I find the first point reasonable, if such complex structures were in fact to appear. But my personal experience abroad, where taxis and tuk-tuks competed in wild, wooly freedom, suggests that a complex fare structure would scare off prospective customers, and that flat rates prevail. Instead of setting a price cap (which dampens the market’s response to additional demand from customers), why not simply require that cabbies display their rates conspicuously, in a standard format, and let them set their own prices? Then if the going rate should rise, more drivers would serve that time or location, attracted by the potential for profits, which in turn would put downward pressure on price.

Entry restriction—limitations on the total number of licensees who can operate in a given jurisdiction—are a third type of regulation that is common within taxicab markets. Unlike the previous two regulatory frameworks, economists, as a whole, are much less favorable toward entry restrictions. The reason is that entry restrictions appear to be driven not by market failures in the taxicab submarkets, but by the desire of existing drivers to limit the competition they face. My research supported this contention; I read of several cases in which drivers and taxi associations organized to lobby for a restricted licensing regime. At this point in my reading I believed I had reached the heart of the matter. With economic theory predicting higher prices under supply limitation, and the near-consensus among economists that entry restrictions were a form of protectionism, it seemed that deregulation (at least with respect to supply) would bring prices down, reduce wait times, and make taxis a viable form of transportation for more people. Real world examples of deregulation, however, show that the issue is not so simple.

In a 1987 article for the Journal of Transport Economics & Policy, Roger Teal evaluated the impacts of cab market deregulation in six U.S. cities, including Seattle, which had experimented with loosening its entry restrictions in the years prior. Some of the results were expected, others surprising. In every city that reformed its regulations to allow free entry, the size of the taxi market in terms of the number of drivers increased substantially, between 18% and 127%. Turnover appeared to be substantial among small operators (usually individual owner-drivers), while entry and exit among larger firms was limited. There were not many new entrants into the dispatch market due to strict service standards that remained on the books, and due to the high cost of entry; it takes a lot of money and time to establish a new taxi service brand and recruit enough drivers to make your business viable. Response times improved slightly, while refusal rates (the frequency that a customer is turned down by the driver) increased. What surprised economists most, however, was the increase in rates charged. In every city studied, rates increased in real terms, with higher increases at cab stands than in the telephone market due to the former’s inelastic demand. Teal suggests that the increase in rates may have been due to pent up inflation that had been restrained by price controls in the past, and that the prices charged by cabbies were justified given their expenses.

A more recent experiment with taxicab deregulation took place in Ireland, where the country’s regulatory agencies lifted entry controls on the taxi markets following a High Court judgment in 2000. Price caps remained. Economist Sean Barrett examined the aftereffects of this reform, and wrote them up in a 2010 article for Economic Affairs. Like the cities documented by Teal, Ireland did not see major price drops. But there were a number of other positive effects. By 2008 taxicab numbers had increased five-fold compared to pre-deregulation figures, and waiting times had decreased: the percentage of customers waiting ten minutes or less for a cab increased from 58.3% pre-regulation to 85.7% in 2008. Goodbody Economic Consultants, hired to study the effects of Ireland’s deregulation, valued the time saved by Dublin cab riders at €780 million country-wide for the year 2008. The problems predicted by deregulation skeptics failed to materialize. Disorder on the roads did not occur, unless one counts rolling protests mounted by incumbent drivers. Order was also maintained at taxi stands, with no systematic fighting among drivers. Taxi drivers did not engage in “cream skimming”; outlying areas continued to be served. Vehicle standards remained high, along with customer satisfaction.

These, and the other deregulation examples I reviewed suggest that opening the taxi market is not a sure-fire way to bring down prices, but it would still be a net gain for a city, creating hundreds of new jobs and making taxi travel faster and more convenient. It is still a mystery to me, however, why cab prices are so out of sync with comparable travel options. It could be that current prices reflect the true cost of operating a cab. Or it may be that competition between cabs is muted by customers’ inability to quickly and easily compare prices. I do not know. Perhaps you would like to offer an explanation in the comments.

If Seattle did decide to ease or eliminate its restriction of taxicab supply we would need to decide how to address the impacts to individual license holders. To those people suddenly able to find work as cabbies, of course, the financial impacts will be dramatic and positive. But the Irish example shows us that the easing of entry restrictions is ruinous to incumbent drivers who have purchased scarce, expensive licenses on the secondary market. In Ireland it was these drivers who took to the streets, outraged that the government would first erect a regulatory regime that forced them to pay a small fortune for a license, and then turn around and render the license worthless. Sean Barrett, one of the economists I mentioned earlier, is unsympathetic. He regards those purchases not as an investment, but as a bet that the government would continue to regulate the market the way it always had. If that bet failed to pay off, well, better luck next time. I see things a little differently. Driving a cab is a job that almost anyone can do, and—based on my experience—is an economic stepping stone for immigrants who do not have many employment options. I am sure that they would have preferred not to spend $100,000 on a license if they could have avoided it, and so it does not seem fair to me to pull the rug out from under them. At the same time, continuing the current structure in order to shield members of a mature industry from competition is not good public policy. It distorts the economy, eliminating jobs and reducing access to a valuable service.

Ireland offers one model for how to balance these public and private interests. In 2002 they established a Taxi Hardship Panel to distribute public funds to licensees who had suffered particularly serious financial side effects as a result of deregulation. 

Grids, novelty

Last fall I had the chance to study urban design at the Royal Danish Academy of Fine Arts in Copenhagen.  It was my first experience at an architecture school, and the contrast with the UW’s planning department was stark.  Rather than the practical, almost technocratic bent of planning school, the Academy encouraged students to approach the process of urban design like a painter or sculptor would approach their art.  This has its ups and downs.  On one hand, you can’t build delightful urban places by applying a template.  You need the right brain for that.  On the other hand, treating a neighborhood like a canvass can have unfortunate consequences.

For example, I noticed more than once that my colleagues resisted the idea of laying out streets in a grid pattern.  They considered them boring and repetitious.  Instead they gravitated toward organic forms: snaking paths, or angular, chaotic networks.

Wassily Kandinsky, Composition No. 8

I’ll admit that these plans looked nice from above, when viewed in plan.  Rather like good pieces of abstract art, which I’m sure, at some level, they were inspired by.  But I had to come to the defense of the grid.  When you’re walking down a street, you’re not comprehending the neighborhood’s entire street network at once.  From an aesthetic poi­nt of view it doesn’t matter if the next road over is parallel to the one you’re on, or veers off at an angle.  You’re looking at the shops, the plants, the pavement in front of you.  And (usually) you’re finding your way, walking with a destination in mind, which you may not have visited before.  It’s hard to find somewhere new when you keep encountering irregular intersections, and when there’s not a straight path to where you’re going.  I think the novelty of living amidst a whimsical street plan would get old after the 15^th or 20^th time you get lost, or the umpteenth call from a lost guest trying to find your house.

Piet Mondrian, Composition With
Red, Yellow and Blue

Grids have their downsides, of course.  You might find them monotonous if you’re behind the wheel of a car.  They tend not to be responsive to topography.  Assuming they’re fine-grained, they necessitate a lot of pavement, which is bad for water quality.  But there are ways to deal with these problems.  You can lay out larger blocks and section them with pedestrian and bicycle paths.  You can install green infrastructure for stormwater treatment.  And if you allow sufficient density, the water quality impacts of the grid are balanced by the mitigation of sprawl.

The main point here is not that creative street layouts are a bad thing, just that they should be considered from the point of view of the average user.  If you’re ever in a position to judge an urban design proposal and the designer tells you that a regular street layout—or any traditional feature, for that matter—would be “boring,” think carefully about the day-to-day users of the space, and about what long-term value novelty would bring to the project.

Nifty tools

ArcGIS can be slow.  Why not use these tools when you need that quick measurement?

• Distance Calculator
• Area Calculator
• Find Altitude

Way to go, Daft Logic!