(This essay was published in Hong Kong Economic Journal on 7 September 2016.)
The taxi industry has been in decline. Over the past decade, the volume of cab usage has dropped from 1.3 million daily trips to 1 million. But the number of taxi licences has remained at 18,138 since the government stopped issuing them in 1994. The industry with few exceptions delivers a low quality commoditized service and passenger complaints are large and rising.
Most taxi drivers earn HK$15,000 each month. The industry is only able to attract those drivers whose incomes are at the low end of the income distribution. Except for a few owner-drivers, most rent their vehicles from taxi companies on a shift basis. The average age of taxi drivers is about 57 years, meaning the industry faces a greying workforce. Most young drivers prefer to work as lorry or bus drivers that offer better pay. The number of taxi drivers has shrunk from a peak of 80,000 in 2003 to the current 40,000.
One reason why driver incomes are so low is that taxi drivers with permits far outnumber licensed taxis. Another reason is that most taxi license owners are not interested in managing their fleet of cars to deliver quality service. Taxi fleets are poorly maintained and ageing – 80 per cent of the vehicles are at least five years old.
Taxi companies are happy to rent out cars on demand to any licensed driver that walks into the taxi rental office on a half-day shift basis. Without active management, there is no incentive for taxi drivers to perform. The owners regard taxi licenses as financial assets that they hold and trade. They have little interest in developing and managing the underlying taxi business.
This sorry state of affairs is completely a result of government regulatory failure. So what is to be done when passenger complaints mount, taxi rides decline, and competition from ride sharing companies enters the market?
Last June the government announced an experimental scheme under which 600 premium taxis will be introduced to the city. These taxis will be distributed among three operating franchises that will be allowed to charge higher fares but only need to pay a low franchise fee. They will operate under the same old regulatory framework that fixes both the number of cars and their tariffs, except that better service quality will be mandated through licensing requirements.
This was a positive signal that government was at least finally responding to rising market demand at the high-end, but it did not try to tackle passenger complaints about widespread bad service throughout the entire industry. Ignoring the latter is a big mistake and will only leave the industry largely in its current sorry state. The proposed move has also infuriated drivers who see premium taxis as a threat at a time when the industry is making efforts to improve.
To ensure compliance with the new requirements, the service performance of the new class of premium taxis will have to be vigilantly monitored requiring the collection of detailed passenger trip information.
The puzzle for me is how service quality compliance can be ensured now with premium taxis when it has not been done in the past with the existing taxi industry. Surely widespread passenger complaints should long have been sufficient reason for government to act.
It is worth revisiting the reasons why Hong Kong adopted the present regulatory framework. In the 1960s, passengers were being victimized by price gouging on taxi rides by unscrupulous drivers after the passengers having boarded the car and embarked on the trip. Licensing and fare fixing was necessary to address this.
The high tariffs led to the proliferation of pak pais (illegal taxis) that competed against licensed taxis. In a bid to “eliminate pak pais from masquerading as taxis” the government unveiled a requirement on September 14, 1974 that all taxis had to be painted with easily identifiable common “red and silver” colors so that they could be readily distinguished from other cars.
As a result, all taxis also became indistinguishable from each other. It became impossible for taxi companies to compete on service quality through visual signals of their brand.
So the regulatory framework that licensed taxis killed all means of competition in the industry by not allowing, first, fare competition and, second, service quality competition through visual branding that could be easily recognized by passengers. The outcome was a uniform commoditized service. But still, it is not obvious that taxi rides would in consequence unavoidably become a low quality service.
In the 1960s and 1970s, taxi rides were not an inexpensive mode of transportation and service quality was not low. Over time, the demand for taxi rides grew with rising prosperity. Since taxis were continuously cruising the roads in search of passengers, they became an increasingly important user of road space and the most important cause of traffic congestion. In 1994, the government decided to stop issuing new taxi licences to constrain supply.
This propelled taxi licenses into becoming valuable income-generating financial assets because of their scarcity value. An increasing number of these licences ended up in the hands of companies that were speculators and investors with no interest in managing their fleets. Service quality declined progressively.
If government had instead chosen to levy an annual license fee on operating taxis instead of fixing total numbers, then the industry might not have evolved into one beholden to rentiers and speculators. Unfortunately, this is what happened.
The rise of pak pais in the 1960s and 1970s stemmed from the fact that the tariffs were set too high. But a lower tariff would have fanned demand and brought more congestion on the roads. Ironically, the appearance of ride-sharing companies is rekindling a modern day pak pai service, equipped with stealth capability enabled by technology, that can trump the “red and silver” cabs.
The real cause of the current malaise in the taxi industry is precisely the existing regulatory framework. The original goals were to eliminate price gouging, pak pais, and reduce traffic congestion. These objectives were mostly achieved, at least for the period before Uber and others discovered ride-sharing technology, but it also created a low service quality industry.
What did the regulators do wrong? What did they miss?
Price controls interfere with private contracting in the market place. They create conditions for profitable opportunities arising from excess demand or supply in the market, with the ownership rights to such gains unspecified and up for grabs.
When prices do not clear markets, somebody will step in to fill the vacuum. When government sets high taxi tariffs, pak pais emerge. When government fixes both the number of taxi licences and the tariff, rentiers and speculators emerge to capture the resulting excess income. When the weather is bad, drivers refuse rides to those who are unwilling to pay higher fares.
In each of the above instances, price control restricts the income terms of private contracting. The right to receive income is partly taken away from a contracting party. As a result some party (including a third party) will appear to try to capture it. Pak pais, taxi license holders, and taxi drivers each try to fill the relevant gap in the best feasible way they can.
Some taxi drivers have formed taxi clubs, each with about 300 drivers. Some clubs charge full fare, while others give 20 per cent discount. Usually, eight drivers are in charge of taking calls in one club and they transfer orders to anyone who can do it. Such private clubs have loyal customers. All this is of course illegal under the existing regulatory framework, but it exists.
Taxi drivers who do not belong to these clubs offer passengers a number of unwelcomed services that include smelly cabs, broken handles, sunken seats, a mixture of driver’s unsolicited cursing and sighing, and a ride that resembles a chase scene from a film.
The development of ride-sharing technology and the opportunity to open up competition in the industry provides a true game-changing event if government would allow a partial price decontrol of taxi tariffs.
At present consumers already have a choice of more than 10 taxi apps – including HKTaxi and Kuaidadi – and none of their creators are related to the business. The newly formed Taxi Council, comprising taxi owners, drivers and others in the industry, recently announced its own new app, called TAXI.
These apps offer different features. All of them allow customers to search online for a taxi, making it easier for passengers and drivers to connect up (with potential for reducing cruising time and reducing pollution from emissions). Taxi call centers and taxi clubs are already doing this and some are already offering opportunities to negotiate price.
Online contracting through new technology is thus able to facilitate price negotiation and contracting to clear markets during periods of demand surge and slump more easily and transparently. It also lowers the entry barrier for pak pais. The regulatory requirement of allowing only licensed taxis to carry passengers effectively becomes irrelevant even though pak pais may be borderline illegal.
The purpose of price control is to provide drivers and passengers with a good signal for private contracting to minimize price gouging and search costs, and also to transact. The problem with a controlled price is that it will fail to reflect true market conditions as demand and supply changes over time, according to the time of day, and by location. In the past it was more difficult for passengers and drivers to contract with each other before the journey began, and for drivers and companies to differentiate their product. Such activity happened somewhat clandestinely through call centers and was welcomed by passengers.
The appearance of Uber and other similar technologies is not fundamentally different from a regulatory point of view, except they provide an even better service and are therefore more disruptive. They also highlight why the existing regulatory framework is in need of change – a change that really is small and simple, even though its longer run effects will be game changing.
All the government has to do is to allow price deregulation for online contracting activities (and call centers too), but retain the controlled price for taxi rides hailed down on the streets. Since many apps already exist and taxis have a history of forming clubs on their own, it would not be at all difficult for the industry to be reorganized for competition to flourish. Uber could become part of the competition and whether it will become a leader, only the market will tell over time.
If government requires all commercial passenger trips to be undertaken by licensed taxis, then companies like Uber will still have to operate in a grey area if they continue to hire drivers through ride sharing. But if taxi tariffs are deregulated, then the entire industry could evolve and attract better drivers and Uber would have to compete for them as well.
The idea of requiring a uniform color to stamp out pak pais is only relevant for passengers who want to hail down a cruising taxi on the street. Even then, what matters is ease of identification and not uniformity of appearance. In the online world (and at call centers), identity is not established by the color of the paint on the body.
The interests of existing taxi license holders and drivers would not be harmed by deregulation. But they would come under pressure to innovate and compete. It is conceivable the widespread adoption of online contracting could reduce cruising on the streets and reduce both congestion and emissions. The number of premium taxis need not be predetermined by a regulatory body on an experimental basis for a pre-committed number of years, but rather by market competition.
There would also be no need for the regulatory body to prescribe the type and level of service to be provided in premium taxis. Indeed, the market could set how many classes of taxis should exist. Uber already is doing this. This can obviate the need for the regulatory authority to collect detailed information about passenger trips in order to monitor performance (as well as reduce the anxiety of the Privacy Commissioner).
The present regulatory framework has served its purpose. It is time for a change.