The taxi trade is subject to licensing everywhere in the developed world. In the most liberal markets, however, licensing only serves to ensure certain quality standards and passenger safety, not to constrain supply. In such markets, the relevant authorities will issue as many licences as there are applicants that meet the specified criteria.

But liberal taxi markets are the exception, not the rule. In the most taxi markets, the number of taxi licenses is either explicitly or implicitly limited. If the number of licenses is smaller than the number of taxis that would result under conditions of free entry, the licenses accrue a scarcity value. There are reasons to expect the number of licenses to fall further behind the free market equilibrium supply over time:

• Taxi demand cannot be reliably estimated. There are models which try to express the required number of taxis as a function of variables such as population size, income levels, the relative price of alternative means of transport, traffic volumes etc., but these models cannot be more than complicated guesswork. In Ireland, taxi demand forecasting was inadvertently put to the test by ‘natural policy experiment’. In 1998, it was estimated that the city of Dublin would require 5,901 taxis by 2008. Two years later, in an unexpected policy U-turn, quantity restrictions were abolished. Thus, the subsequent expansion of taxi numbers can be compared to the model forecast. It turned out that actual taxi numbers reached about 12,500 in 2008 – more than double the number that were supposedly ‘needed’. If demand cannot be ‘objectively’ modelled, demand forecasts become arbitrary, and malleable by special interest groups.

• The taxi market is a classic showcase for the rent seeking mechanisms identified by the Public Choice School of economics. The gains from restricting licensing practices are highly concentrated, the losses are widely dispersed. For the majority of consumers, taxi rides only represent occasional purchases, which only account for a minor share of their household expenditure. For them, the issue is simply not important enough to get politically organised, whereas for license holders, political organisation can make all the difference. License holders are also a relatively socio-economically homogenous group, which decreases the cost of organising politically. The opposite is true for consumers. Virtually everybody has, at some point, been a consumer in this market. Taxi passengers are not a group with any identifiable common social characteristics. This leads to a political asymmetry. Producer interests will be far more likely to capture the political process, and use it for their ends, than consumers.

• When licenses are tradable, these tendencies become self-perpetuating. The value of a license reflects the net present value of the income that a driver can expect to earn in the future years. This, in turn, depends on how many licenses will be issued in the future, because, broadly speaking, the greater the number of licenses, the lower the income earned by any given license holder. So the price of a license, at any given time, reflects expectations about the future number of licenses in circulation. If a driver buys a license, and if their number is then increased at a higher rate than was anticipated at the time of the purchase, the driver may find themselves unable to recoup the initial investment. If, however, the increase in license numbers is smaller than anticipated, the driver would reap a windfall gain. So political pressure to restrain growth in the number of licenses cannot be relaxed over time, on the contrary, incentives point towards steadily increasing it.

• The number of licenses issued does not just affect a driver’s future income, but also, by the same token, the resale value of their license. From the perspective of an operator, the taxi license is an input factor like any other, just like the vehicle itself and the petrol it runs on. However, unlike other input factors, the license has no inherent productive value – it is, after all, just a piece of paper. It derives its value exclusively from legislative fiat. If quantity restrictions are abolished, the value of a license immediately drops to zero. Thus, under a system of quantity restrictions, licenses become a financial asset, rather than just a business permit. They are, however, an unusual asset class, in that their value can reach very high levels, but it can also be wiped out at the stroke of a pen any time. Political decisions matter far more than market fundamentals. License holders have a lot to gain from engaging in the political process, and a lot to lose from not doing so.

There are indeed plenty of taxi markets which behave just as we would expect, given the above points. In 2013, taxi licenses in Paris were traded for over €200,000 each. In metropolitan markets in Australia, a taxi license would fetch over AU$250,000. The extreme end was probably New York City, where taxi medallions typically sold for nearly $1m each in 2014.

Quantity restrictions can be very lucrative to those to whom the ensuing monopoly rents accrue. But any gains are achieved at the expense of other groups, especially consumers. In a quantity-restricted taxi market, fares can be expected to be higher than they would otherwise be, representing a redistribution from consumers to license holders. Waiting times for a taxi will be longer in a quantity-restricted market.

At least where licenses are tradable and lettable, it is not necessarily correct to think of the effect of quantity restrictions as a redistribution from passengers to drivers. Rather, it is a redistribution from passengers to a certain group of license holders, namely those who bought their licenses while they were still relatively cheap, and who subsequently benefited from their increasing scarcity value. Drivers who rent their license cannot be counted among the beneficiaries. According to the OECD “there is no evidence to suggest that taxi driver incomes are higher in markets with restrictive entry conditions. […] Melbourne has taxi licences valued at almost A$500,000, and driver incomes estimated at A$8 – 14 per hour”.

An extreme example of the monopoly rent effect would be the wealth of New York City’s ‘Taxi King’ Evgeny Freidman, who began buying up taxi licenses in the mid-1990s and now owns around 1,000 of them. In Freidman’s own words:

“Every day that I wake up, you know, I’m like, this is great […] You know, I live on Park Avenue, got a bunch of, like, Ferraris that I drive. I have a house in the south of France. I can have breakfast, like, at Cipriani. And it’s like, you know, pinch me. Is this real?”

Freidman is right to wonder about how ‘real’ his wealth is. If quantity restrictions were abolished, his licenses would become worthless within a split second.

Quantity restrictions also have a negative impact on service quality and the diffusion of technological and organisations innovations. This point is less measurable and therefore harder to prove, but it is striking how quickly London’s taxis came up with a series of service improvements as soon as they came under pressure from their new taxi-like competitors. For example, so far, passengers who wanted to pay by card had to pay a surcharge of 10 per cent of the fare. This surcharge is now being dropped. Previously, many London taxis did not accept card payments at all, a situation which is currently being rectified. The number of taxis offering fast WiFi access is also increasing, and there are plans to create taxi ranks outside of every Night Tube stop. Taxis are also beginning to use ‘hailing apps’ similar to those used by their new competitors, and they are making it more convenient for groups of people to split the fare.

The main effect of quantity restrictions, however, has to be the ‘deadweight loss’, a pure welfare loss that even people like Freidman cannot exploit. When taxi rides are perceived to be a luxury good and/or when average waiting times for a taxi are high, taxi rides will be reserved for special occasions, rather than an everyday mode of transport. People will make other arrangements where possible, and a ‘taxi culture’ will never develop.

Nonetheless, there are examples of places which have, against the odds, liberalised entry into the taxi market. In our new paper ‘Hire authority. Turning statutory regulation into private regulation for the UK’s taxi industry’, my colleague Diego Zuluaga and I explore how those policy experiments have worked out, and what the UK can learn from them.