“According to an extensive study by a group of experts on behalf of the Ministry of Transport, which was published by the Ministry last year and parts of which went under academic scrutiny and were also published in international scientific journals, the liberalisation of licensing for the profession of taxi drivers in Greece will create 2,000 new jobs, will reduce average production costs in the country by 0.03%, will improve the service quality indexes that the Ministry uses to monitor the sector by 20% in the next five years, will reduce the private use of cars in the main metropolitan areas by 2% and traffic congestion by 1.2%, and it will reduce average waiting times for users by 3′ per journey.”

Of course this is not true. But as I woke up in the middle of the night yesterday I was almost convinced that I had read this somewhere. Sadly, after going through all my files and records I finally had to accept that this was only a dream. There was no study on the costs of regulation of taxis and no study on the social and economic benefits of their deregulation. The only thing that was there, exactly as it was when I went to bed last night, was an abrupt change of policy announcing the complete deregulation of the sector in a way that overturns what was announced – and agreed upon between the sector and the government – only a month ago.

Of course, one does not need an extensive study to conclude that regulations and licensing introduce distortions in the market. Scanning through the first few pages of any public economics textbook is enough to convince any sceptic that deregulation can be welfare improving as it removes monopolistic rents, lowers prices and expands output. But, as it happens, life is rarely as simple as our textbooks would have us believe. After all, they are only textbooks, they are simply there for exposition. They are not substitutes for real-life analysis. They never intended to be.

In well-defined markets, where the location of where trading takes place is known and where timing doesn’t matter (and where, I should add, marginal costs constitute the greatest element in total costs – but this is getting too deep into the technicalities), removing regulatory constraints can indeed be welfare-improving, as it reduces monopolistic rents and allows for a better equilibrium to be reached that increases the volume of output consumed and reduces its unit price.

But when it comes to the case of taxis things are markedly different. When I am on the street looking for a taxi, there is hardly a market there. All there is, is a Poisson distribution, which describes the probability of the arrival of an available taxi. If a taxi shows up, I take it. I do not negotiate the price: this remains fully regulated and non-negotiable – and for a good reason, as otherwise my taxi driver could extort a higher rent from me depending on how desperate I am to use the service (or, worse, depending on how wealthy I appear to be). And since what I am buying is a service best described as “fast transfer from point A to point B” – and since I am also competing with other potential users of the same service – there is very little room, or time, for shopping around to find the best service (a clean, well-maintained taxi, driven by a highly-qualified and customer-oriented professional – my last night’s dream again). I simply jump into the taxi and give the details of my destination; adding, as politely as I can, “…and, if possible, please try to get there fast – I am in a hurry”.

In theory (those textbooks again), if there was an infinite number of taxis waiting for my custom I would indeed have the opportunity to shop around and choose to get into the best-serviced taxi. And if I was dissatisfied with the service, I would never use the same taxi again (incidentally, I do not think I have ever used the same taxi twice ever in my life – who has? who remembers?). But in practice, if there was indeed an infinite number of taxis waiting for me, then I shouldn’t expect a good service and the best-serviced taxi wouldn’t be all that well-serviced after all. The monthly income of my infinitesimal taxi driver would be approximating zero (asymptotically, but still) and thus he or she would have very little time and money to worry about fixing those rusty breaks, updating him/herself about traffic conditions and fastest/cheapest routes, or, for that matter, maintaining a clean environment for his/her customers.

Of course, in the long-run, the market would self-regulate, my infinitesimal taxi driver would die (as a business – hopefully not literally, although you never know) and the ones who would survive would be the best-maintained and best-presented taxis providing the best service possible. But this would probably bring me back to square one, waiting for a taxi in the middle of the road and taking whichever taxi happened to show up first. At that point in time my taxi driver would still be a monopolist, as my demand would be essentially inelastic (I don’t happen to take taxis casually when I don’t really need them) and the supply of taxis, at this very time and place, would be absolutely fixed. As Cairns and Liston-Heyes say in their 1996 paper in the Journal of Public Economics (“Competition and regulation in the taxi industry” – which I am sure the Ministry of Transport has studied thoroughly), “equilibrium of a deregulated industry does not exist; price regulation is essential, and entry regulation may be useful”. (my emphasis)

Perhaps we do need reminding that this was what regulation was meant to address. By fixing the price of the supplied service and the quantity of taxis in circulation, the objective was to engineer an equilibrium that would be socially optimal. The taxi driver would have sufficient custom and income to maintain a good service (and if s/he didn’t, then the regulator would be there to price him/her off the market, through fines and penalties – am I still dreaming?) and I would be content with spending, on average, the average waiting time waiting for a taxi (you can see the sarcasm of the tautology here). If the state, or the public, were unhappy with any of these parameters, if taxi-driver incomes or average waiting times ended up being too high, a change in the policy parameters (prices and/or quantities) would be sufficient to ensure a re-engineered equilibrium with improved efficiency.

Sometimes, information problems are such that make this re-engineering very inefficient. Sometimes, letting the market determine the equilibrium is faster and less costly. But not when the traded product is locationally and temporally fixed and when the demand is inelastic. In such a case, a combination of market-and-regulation (don’t forget that the taxi driver is still free to choose which road to take when looking for customers and still has the incentive to drive through where the demand is likely to be) can produce the best outcomes.

This does not mean that we shouldn’t be concerned about how many taxi licenses are around or how distorting is the price imposed by regulation. But it is one thing to be concerned about these things and quite another to let everything to the old good invisible hands of the market. It is true that the market has this amazing ability, that regulation lacks, to match supply and demand and to produce an equilibrium at the minimum of transaction costs – we know this at least since Hayek’s critique of socialism. But unfortunately this applies only when markets are “perfect”, in other words when all parties involved have perfect information (i.e., I know where all taxis are at all times) and when demand is reasonably elastic. But as a person I rarely have perfect information, I rarely consider exhaustively all my options (I have to admit, sometimes I even act on an impulse – I blame my bounded rationality for this) and, when I take a taxi, I am rarely cavalier about whether I should do so or not. Until these good powers are instilled to me, by God or by a similarly improved-through-deregulation education system, I would very much like the regulator to regulate the taxi industry for me.

Note: This article gives the views of the author, not the position of Greece@LSE, the Hellenic Observatory or the London School of Economics.