Contestability is a term you often hear economists use to mean that a market is competitive simply because players can enter and leave freely. This concept, which developed in the USA in the early 1980s, has been used to justify deregulation policies implemented to reduce legal barriers in various sectors such as air transport and ride-hailing services. Such policies have had contrasting consequences for market contestability, as we will see when we compare passenger air transport and ride-hailing services.
The general principle: market contestability
The general principle of market contestability was put forward by Baumol in the early 1980s: to encourage businesses to operate more efficiently, the market has to be contestable, meaning that entry and exit are as free as possible. This theory of contestable markets was quite innovative: competition in a market no longer depended on the market having a large number of players (the atomicity hypothesis) as in the standard model of pure and perfect competition. Concretely, a business which was the only player in its market could find itself in a highly competitive position if there was a permanent threat of new entrants. It would have no choice but to set its prices at a competitive level. If it set its prices higher, a new entrant would immediately move into the market and “take its place”. One of the conditions for contestability is that a business entering the market needs to be able to leave easily, without incurring sunk costs.
In a contestable market, any strategy to try to exclude a competitor, such as predatory pricing, is doomed to failure. Predatory pricing is when a business cuts its prices aggressively to force a new entrant out of a market, and then increases them again later. Predatory pricing cannot work if the market is contestable, because the new entrant will just come back in again once the price war is over.
Contestability offers the main benefits of a competitive market, without the need for a large number of players. Consumers can expect to see lower prices, higher volumes (due to the price elasticity of demand), better quality and a wider variety of products. On the supply side, contestability should drive businesses working in a market to constantly improve their efficiency and productivity.
Contestability and the air transport market
After World War II, the passenger air transport market was strongly regulated in all countries. High barriers to entry were set up, the majority of them administrative. As a result, quasi-monopolies formed in each country, led by national “flag carriers”. The market had a very low degree of contestability, because legislation kept competitors out.
And yet, in economic terms, the air transport market has fairly weak barriers to entry. Fixed entry costs are low: little capital is actually needed to set up an airline, because planes and crews can be rented. In addition, fixed costs that are incurred – in particular planes – are not sunk: an airline that has purchased planes can easily sell them again on the second-hand market.
This strong potential for contestability in the air transport market, combined with a surge in liberal ideas, led the authorities to pursue a policy of deregulating the skies. Following on from the USA in the 1980s, Europe progressively liberalised its internal air transport market in the 1990s.
Looking back with hindsight, how successful were efforts to deregulate the skies in the USA and Europe? Overall, we can say that the desired effects of contestability were achieved in the internal market.
- Firstly, market contestability rapidly led new companies to enter the market and flourish. The low-cost sector grew from 9% of European air traffic in 2002 to 43% in 2017 (graph 2).
- Secondly, more competition and a dramatic fall in costs meant a sharp drop in prices. The effect on prices did not only impact those airlines in direct competition with low-cost carriers, it also spread to the entire airline sector, via what has been described as the “Southwest effect”. Morrison .
- Thirdly, demand for air transport grew sharply, both in the USA and in Europe. This is because price elasticity of demand is very high in the “tourism” market, the low-cost sector’s main target. The additional volumes do not primarily represent existing air passengers. They are essentially made up of people who would not previously have taken flights, but who decided to do so because the prices had fallen.
- Fourthly, there was strong diversification in terms of supply: low-cost carriers did not simply match the services being offered by the legacy airlines, they added capacity and frequency. Also, and most importantly, they densified route networks, in particular by linking regional cities together.
- Lastly, market contestability improved quality of service in the air transport market, and especially punctuality. Moreover, deregulation did not negatively affect flight safety.
And yet, air transport is not a perfectly contestable market
So, should we conclude that the deregulation of the skies has made passenger air transport a perfectly contestable market? In reality, there are still numerous obstacles to contestability in this market. First and foremost, there is an institutional obstacle: take-off slot allocation.
The legacy carriers have a historical advantage which cannot be reproduced by new entrants: they hold a portfolio of choice slots at congested major airports. A slot, which is allocated free of charge by an authority, gives a usage right to the holder, but does not as such constitute an ownership right over the resource.
Take-off slots at major airports such as Roissy in Paris or Heathrow in London are a rare resource, because there are not enough slots to meet demand from airlines, in particular during the attractive morning and evening peak “hub” hours. At these saturated airports, in practical terms the rules used to allocate these take-off and landing slots favour the incumbent airlines:
– grandfather rights: a carrier that has operated a slot during one season has a right to keep it for the following season. This conservative rule was initially put in place in the USA in 1969 to encourage airlines to invest, and then mirrored in Europe;
– the 50/50 principle: slots that become available, either because they have a usage rate below the 80% threshold, because an airline goes bankrupt or because airport capacity increases, are allocated free of charge, with 50% going to new entrants and 50% going to incumbent airlines that request them. An airline is considered as a new entrant if it holds under 5% of the slots allocated. Paradoxically, this restrictive definition of “new entrant” leads slots to be allocated to multiple newcomers, whereas it is difficult for existing operators with low market shares to secure new slots. This system encourages small competitors to enter the market, but it does not help them to grow.
Access to slots is a genuine regulatory barrier. It limits market contestability by limiting the number of new entrants. The only way new entrants can potentially get around this is to buy an airline which already operates from a congested airport. A reform of the way slots are allocated would be the only solution to increase market contestability, but such an initiative is not on the cards at present.
Contestability in the individual passenger transport market is an illusion
At first glance, the revolution in the individual passenger transport market over the last decade appears to be a simple repeat of deregulation in the air transport sector: here again, the initial situation was a market with an administrative barrier to entry, which stopped new entrants coming in and so limited market contestability.
If we look at France, we had a taxi market that was highly regulated, in terms of both price (ceiling) and quantity: between 1937 and 2009 taxis benefited from a form of monopoly in that they alone were permitted to pick up passengers either at a taxi rank or in the street, and there was a strict limit on the number of licences issued. In 60 years, the number of licences increased by just 29%, whereas during the same period the population of the Paris metropolitan area increased by 84%. This resulted in a degree of short supply: in 2007, Paris had 2.8 drivers per 1,000 people, four times fewer than New York. For users, this resulted not in a direct increase in prices (because of the price ceiling), but in a fall in the quality of service, in particular in terms of waiting time. For drivers, the economic rent generated by rationing led to a sharp increase in the price of licences, which hit a peak in 2007, just before the Attali report was commissioned. This report proposed that the taxi sector should be liberalised to make the market more contestable. The price of a taxi licence, adjusted for inflation, more than doubled between 1995 and 2015.
Naturally, licence holders opposed any degree of market liberalisation, fearing that the price of licences would collapse. One solution would have been to compensate licence holders, but the cost to the authorities of doing this would have been very high (€4.5bn for the 18,000 Parisian drivers in 2010). The solution put in place in France, by way of the Novelli law for the modernisation of tourism passed in 2009, was to create an additional, and supposedly contestable, market for ride-hailing services. This market differs from the taxi market in that rides must be booked via a smartphone. But in fact, now that we all carry smartphones in our pockets, the distinction between hail-and-ride and a pre-booked service is much less significant.
As in the air transport market, the effects of deregulation for consumers have been mainly beneficial.
- Firstly, we have seen strong growth in supply, with several ride-hailing service companies entering the market. In 2018, Paris had 19,000 ride-hailing service drivers, which is approximately equal to the number of taxis: we can therefore consider that total supply has approximately doubled in the Paris area since Uber entered the market in 2012.
- Secondly, opening up the market has led to a fall in fares, as customers can choose between a ride-hailing service and a taxi. Through a knock-on effect, growing supply and falling fares have sent the value of licences tumbling. The Parisian licence market has lost over half of its value since 2013.
- Thirdly, research into Uber’s performance in the USA has confirmed that its entry into the market has led to lower fares for users, increased vehicle availability and improved service quality. As in the airline sector, it is probable that quality improvements have spread to taxis, due to comparison and incentive effects.
Nevertheless, although we can see the effects of increased contestability, the ride-hailing service market has rapidly concentrated around one dominant player, Uber, which has a share of 84% of the market in France and 75% in the USA. Looking at France, if we take as a metric application audiences, measured in terms of the number of connections, we obtain an audience in 2019 of 4m people per month for Uber. The audience of Uber’s closest competitor, Kapten, is nine times smaller, and that of the next incumbent, Heetch, twelve times smaller.
Why has the market become so concentrated around a single player, and for so long?
The main cause is the nature of ride-hailing services, which are very different to the transport sector.
Ride-hailing services operate by way of a digital application used to book a vehicle via a smartphone, and so are affected by cross-side indirect network effects.
The higher the number of ride-hailing service drivers (the supply side of the market) available on a given reservation platform, the more satisfied the customers (the demand side of the market) will be: they will have to wait less time for a booking.
This indirect network effect also operates cross-side because ride-hailing service drivers are more satisfied to work via a given application if it is used by a large number of clients. They have more chance of picking up a fare. This network effect is conducive to a snowball-type phenomenon where one firm rapidly captures most of the market share: “winner takes all”, as the saying goes.
Taking this a degree further, once a ride-hailing service application has captured the market, it becomes very difficult for a new entrant to come in, even if its application is better. This phenomenon, which is well recognised in terms of technological standards, exists because even if an individual user wants to change to a new application, it is in their interest to wait for other customers to start using this new application, to avoid losing the indirect network effect.
This market’s natural tendency towards concentration is heightened by Uber’s strategy of accelerating network effects and economies of scale by setting very low prices on both sides of the market.
Uber subsidises both sides of the market – the users and the service providers – and has been loss-making since it was launched in 2011. Such practices, financed via successive investor funding rounds, raise a question: they could be described as a form of predatory pricing, with Uber making an initial loss to eliminate ride-hailing service competitors before raising prices later once it has established a non-contestable monopoly.
The situation is unusual in that this predatory pricing has been going on since Uber was launched, whereas predatory pricing is usually a very short-term strategy. It is clear that the financial markets, which in February 2021 value Uber at $110bn, still expect the company to become profitable sooner or later.
This expectation is based on three levers.
- The first is a drastic fall in costs, thanks to autonomous vehicles which would eliminate the need for drivers. This seems to be off the table, as Uber is no longer pursuing this avenue.
- A second lever would be for Uber to diversify into activities that are more profitable than the ride-hailing service market, but which make use of the same trip optimisation platform. This is what it is now trying to do with UberEats but there is no guarantee that this strategy will improve Uber’s profitability in the short term.
- The last lever, and undoubtedly the most probable and the most lucrative, is to aim for a future monopoly position in ride-hailing services through which it will be able to raise its prices, both on the client side (fares) and on the driver side (commissions).
We could argue that raising prices will make this new market more contestable again. A new competitor could try to offer lower prices than Uber, or pay drivers better. But we would be forgetting the financial cost of such a strategy and the cross-side indirect network effects that make migration to another platform very difficult due to the need to convince both users and drivers to change their behaviour simultaneously. With the technology as it stands today, the ride-hailing services market actually has a very low degree of contestability.
When you make two markets that were initially regulated contestable, you do not necessarily end up with identical market structures. In the airline sector, deregulation has generated a fairly contestable market in the long term, although the legal barrier of slots remains. The individual passenger transport market is different, because deregulation has led fairly quickly to a situation where one ride-hailing services company has a sustained dominant market position: the degree of contestability was initially high, but it has now become very low.
But even in this case, users are not necessarily losing out. Digital transaction platforms have increased transparency and ease of use, and at least while predatory pricing lasts, reduced fares. In addition, the traditional market sector has been forced to adapt to this new competition by improving its services.
Lastly, the authorities can ensure (for example, by the 2015 UberPop ban, the 2018 Grandguillaume law governing ride-hailing services and the 2019 Supreme Court of Appeal judgement reclassifying the relationship between Uber and a driver as employment) that changes in market conditions are not made to the detriment of users or drivers.