Mobile World Taxi

Why Barcelona has no Uber

Barcelona is one of the few major European cities where Uber, the mobile phone application-based ride sharing and taxi service, does not operate. Of comparably populous cities only Hamburg is currently Uber-less. Doubly confusing as Barcelona prides itself as host of the world’s largest gathering of the mobile technology industry, the World Mobile Congress. An irony not lost on visiting journalists as they queued at the airport for insufficient taxis.

On the face of it this is classic schizophrenic Barcelona: An Olympic image marketed to the world, which becomes fundamental to the vanity of the city’s population, even when that image brings detriment to that very same population. So deeply embedded in Barcelona’s social structure as to render a schizophrenia, not mere conflict or paradox which can at least be acknowledged and rationalised. A recurrent theme, most obvious in the inability of the Ajuntament (city government) to regulate AirBnB rentals to tourists, sold on an exported image of Barcelona, that has the consequence of forcing the actual residents of Barcelona out. The imposters from Silicon Valley are guilty of nothing more than exposing a social hypocrisy. And exploiting it ruthlessly, secure in the knowledge that the disaffected cannot turn on what they cannot acknowledge, hence only on each other. A heinous crime that cannot be fairly judged because both actions exemplify their own moral good: Silicon Valley is optimistically building a better tomorrow, its Disneyland reality disguising anything that isn’t. Much as Barcelona’s hopefulness has been wedded to becoming a global city, its schizophrenia its defense against the social threats inherent in globalisation.

Uber’s official response, which bears the same title as this article, predictably eschews all that political philosophy, opting instead to reinforce some familiar Anglo-American stereotypes about Spanish regulation, under-employment, de-emphasis on utility, and general lack of free market liberalism. Indubitable, but ostensibly fails to explain why Barcelona has no Uber when Español-centric competitor Cabify is present in the city. Which is a shame, because Uber España reveals a lot about how chaotically Spanish policymakers are managing this relatively new exposure to the external world. And for a State premised on stability, the accusation of chaos is high treason indeed. But any inquisition will have to join the queue, right behind a notoriously aggressive startup and some rather volatile Taxistas. What is it about the taxi business that fosters so much hostility?

This article describes the tech disruption of the taxi business, explores why Uber’s initial incursion into Barcelona failed, analyses the limits on VTC licensing, and finally makes time for change. Subsequent events are described up to March 2018 (when Uber returned to Barcelona) and January 2019 (when it left again).

Disruptive Taxistas

Taxis are a curious business. And curiously similar around the world, suggesting their business model is no accident. A largely commercial enterprise, yet with high levels of regulation, both formally and informally. Geographical territories which theoretically give rise to local economies of scale, but where individual owner-operators are commonplace. Surely a market ripe for disruption by a competitor with a globally-derived advantage? Not necessarily.

It is convenient to characterise the taxi business as economically inefficient. While drivers often do make poor use of time, perhaps half their working day spent without custom, customer demand is similarly peaky. Such demand requires an extremely flexible workforce, which traditional industrial Fordism cannot provide. The modern Gig economy (“economía colaborativa”) differs only in ownership and security: Unsurprisingly, it is psychologically easier to work like this when one is genuinely working for oneself, and one can trust those around for support.

So established taxi drivers faced with the threat of Uber are not over-reacting to the threat of competition or wage suppression, as is the public perception. They are fighting for their sanity, for the base psychological conditions in which their job becomes workable. That this is so readily misunderstood by those with steadier work epitomises the wider social policy struggle with the Gig economy: The solution, as practiced by the taxi industry for a century, embodies almost paradoxical extremes of both freedom and protection. If mainstream society can’t understand those extremes, let alone that paradox, then maybe the work itself is not well suited to the humans that make up society, and we should, at the very least, take greater care of those at these extremes. This is a particular concern locally, which (as discussed later) seems culturally less Hegelian than the Anglo-Germanic world – less able to deal with the perpetual change that is the Gig economy’s occupational neurosis. If Taxistas are merely the first point at which the prevailing model of stability breaks down, the state’s reaction to them could have far wider consequences than their transport niche.

Technology is not obviously a game changer for taxis, it is a marginal advantage. The majority of the commercial operating cost of road transport remains, for now, human. There is technological scope for optimising the driver’s time, but this is (as noted above) not great. Specifically not great enough to yield the 40% return on investment typically expected of a Silicon Valley startup. Nor is this technology a natural monopoly – established public transport technology integrators are just as able to develop software globally for sale locally. While technology can have a marketing advantage on the internet, this is less true in geographically discrete localities. In the locality where the journeys are actually being made, global brands like Uber can pale in comparison to historic local taxi operations. If Barcelona’s yellow-doors-on-black-body livery isn’t iconic enough, consider how New York or London cabs practically define their host cities in popular culture.

While Uber may have uncovered an entirely new kind of taxi business model or market, they would not be the first to fail to commercialise obvious gaps in transport provision. Perhaps the best example, because it was also done commercially by a corporate multinational able to draw on proper finance and managerial resource, was Stagecoach’s Yellow Taxibus in Fife, Scotland. The concept, a shared taxi, filled a price-chasm between conventional scheduled local bus or train (typically €3 fare), and private taxi (€30 fare). In spite halving the operating cost of equivalent (publicly subsidised) Demand Responsive Transport, Stagecoach could find no commercially viable price point. The explanation lay in the realm of behavioural, not rational, economics. While Stagecoach’s product differed slightly from Uber, the example counter-balances the optimistic Americanism, “if you build it, [they] will come”.

Race for the Prize

“It’s hard to fathom Uber operating so far from profitability at a time when it feels like an established mainstream brand on the global stage.”

Techcrunch (above) reflects growing concerns that, in spite of eating $8.8 billion of venture capital equity over the last 8 years, Uber is losing roughly $3 billion annually, almost half its net revenue. Uber is still behaving much like a nascent startup – still taking high risk bets, be they on China or driver-less technology – not demonstrating the maturity required for stock market ownership, which is the logical exit strategy for Uber’s Silicon Valley investors. Presuming there is a viable long term commercial core to Uber, and Uber are not simply burning cash to subsidize riders, Uber will increasingly find itself having to focus on current, not future, profitability. And as a publicly listed company, Uber’s aggressive disruption to markets would largely cease because corporate transport businesses are expected to produce low (financial) yields, but be stable – not least because those conditions describe the long run state of local passenger transport markets.

By now it should be clear that the glittering prize at the end of Uber’s rainbow is not dominance of the global taxi business, but what comes next: The taxis that drive themselves. The lay assumption (which is, for example, implicit in Mariano Rajoy’s promotion of Catalunya as a hub of both mobile and automotive technologies) is that consumer automated vehicles will be an evolution of traditional private motorcars. Even if so, factors such as technical complexity and (accident) liability may force a rental business model, that could look remarkably like a driver-less taxi. While mechanical (the natural strength of the traditional automotive industry), data (the natural strength of Google or Apple), or operational (the natural strength of the public transport industry) expertise may be important, the biggest challenges may well be legislative and societal. If decisions are decentralised (regulation of local transport at local level) then Uber, with its years of disrupting local taxi markets, would be well placed to meet those challenges.

Pop Goes Uber

Uber first arrived in Barcelona in April 2014, with a product called UberPOP. Best described as car-pooling service, not a taxi, UberPOP succeeded in antogonizing the local taxi association, while failing the basic test of legality: Under article 101 of the Ley de Ordenación de los Transportes Terrestres (LOTT), a vehicle is not private if the driver accepts remuneration from their passengers – it has become public, a taxi, and must be regulated accordingly. A caveat on “subsistence payments” gives hope of legal challenge, but in practice the more robust defense (as recently used by Blablacar in Madrid) is to claim LOTT does not apply because the business is not offering a transport service. Rather, the business is an online platform that allows communication between individuals that are arranging their own private transport. That’s evidently harder to justify if you are operating like a taxi company.

Uber were evicted from Spain at the end of 2014, amidst claims of “unfair competition“. That phraseology emphasises a misunderstanding about Spain, a country where tradition often holds that one asks permission before, not after – “visto bueno”. What’s “unfair” is not that Uber competed for the taxi market, but they were not seen to play by the rules, when everyone else was seen to do so. And when Uber continued not to play by the rules they found Spain increasingly unenthusiastic about upholding Uber’s right to play, ultimately manifest in the draconian blocking of Uber’s website. In some ways Uber got off lightly in Spain – UberPOP lead to riots on the streets of Paris before finally being abandoned.

UberPOP, far more than Uber in general, illustrates arrogance: Ultimately, the product could only succeed if long-established principles of Spanish passenger transport legislation were to crumble. To, in the meantime, act “above the law” is intensely disrespectful of the citizens that chose, “democratically” (here meaning post-Francoist, with many complex political connotations), to frame their society by that law. Silicon Valley ideologies can be autocratically imposed when a product is intrinsically of the internet, but Uber is not – its core business is physical, on local streets, where local legislatures predominate. In American plutocracy the winner of the commercial market uses their windfall to influence the law that binds them – it’s called lobbying. In contrast, the concept of visto bueno (before, not after) instead implies that incompatible law be changed prior to the winning product even existing. But of course because the product doesn’t exist yet it, the product cannot so clearly demonstrate the need for legal change. Consequently a product that engenders legal disruption cannot compete in Spain on (strength of) product alone, as perhaps it can in the United States. Instead, the question to ask is how to change Spanish policy? If there is no rational answer to that question, perhaps that is because policy is primarily a function of an inherently internalised, shared social knowledge.

Regulata España

In Uber’s defense, the rules, the pre-established options for legal taxi operation in Spain, were… limited. These are primarily laid out in the companion regulations to LOTT, Reglamento de la Ley de Ordenación de los Transportes Terrestres (ROTT). As is common in Europe, the main distinction between bus and car is number of seats: Cars have no more than 8 seats (plus driver). The conventional model of taxi operation is “transporte público de viajeros en vehículos de turismo” (in a car). This can pickup passengers anywhere (in its home territory) without pre-booking. Fares are calculated either on actual distance travelled (using a taximeter) or on a fixed rate for the route (like a bus).

ROTT (chapter IV, section 2) also defines a type of vehicle rental with a driver. Officially, “arrendamiento con conductor”, colloquially “VTC” from the French, “Voiture de Transport avec Chauffeur”. The key difference between conventional taxis and VTCs is that VTCs must be pre-booked (“abono”, pre-paid), they cannot be hailed on the street. This dual structure is also found in other parts of Europe, for example Hackney Carriage (taxi) and Private Hire (VTC) in Britain. Uber’s core (professionally driven) service typically uses VTC-like licensing, but because Uber’s technology makes pre-booking almost as simple as hailing a cab in the street, Uber can compete effectively against conventional taxis, making a mockery of the legal distinction.

Supply of taxis is necessarily suppressed: As explained in the Disruptive Taxistas section above, for taxi drivers to work (happily, or at least without going insane) an appropriate balance must be maintained between time spent driving passengers and time spent waiting for custom. Passenger demand is extremely peaky, so providing sufficient supply to match the highest peak would likely upset this balance – too much time would be spent waiting during off-peak periods. In conventional economics the fare (price) could rise, the higher bounty per journey offsetting the fact that each driver now undertakes less journeys. Aside from the elastic nature of taxi markets (where increasing the fare causes a proportionately greater reduction in patronage), this increases the variability of rewards (over the working day), encouraging a gambling mindset, the complete opposite of a professional driver mindset (which should emphasise safety and security, not chance). Our temporal model of taxi driver behavior also explains why demand and supply cannot be matched by the free market: Short run market entry and exit (of potential drivers) occurs over time, so the very process of contesting a market upsets the all-important balance of time driving to time waiting. These conditions are not unique to taxis – in the bus industry, for example, it can be psychologically difficult to drive an empty bus all day, and bus markets typically lack short run contestability, even where deregulated – but perhaps the taxi industry takes these conditions to an extreme.

All that confuses policymakers. Government is expected to intervene in what looks like a free market, with no easy rationalisation of its actions. And where wider policy is, as is common in Europe, promoting mobility through public transport, suppression of taxi supply can appear downright contradictory. For disruptive competitors, such as Uber, this is manna from heaven: Government can’t rationalise its own policy to its public, thus can’t react, except to look arbitarily protectionist. Since policy can’t be explained directly, its implementation often eschews expected models of governance, with a high incidence of informal communication, potentially verging on political corruption, and a high reliance on arcane or arbitrary “tests” to regulate the market.

VTC Limitations

VTC licences are limited by just such an arcane and arbitrary test, the now infamous 1:30 ratio of VTC licences to taxi licences. This was written into law (as an Order) from at least 1998, but it is often cited from 1987, the year of the original LOTT. Since LOTT is essentially a collation of prior law (filtered for compliance with the European Union, which Spain joined in 1986), it is quite possible the ratio has existed, in some capacity, for as long as anyone can remember. In principle (and this is all article 181 of the original 1990 ROTT actually says about the ratio), new licences should be tested against the potential users of the service (“los potenciales usuarios del servicio”), implying VTC supply is to broadly match the underlying consumer demand for VTC travel. Article 14 of the 1998 Order repeated this mantra, then promptly defined the test as a maximum of 1 VTC licence for every 30 taxi licences – which has little or no bearing on potential usage. VTC was envisaged as serving a VIP function, a limousine, not mundane public transport, so it would be quite reasonable to assess potential use as far lower than taxis. But why a fixed ratio of 1:30? The 1:30 clause was perpetuated in the replacement 2008 Order. In 2015 the clause was directly added to ROTT’s article 181, albeit with an explicit caveat that Autonomous Communities (regional governments) could modify the ratio to make it “less restrictive”.

The Spanish implementation of the European Union Services Directive in 2009 altered (chapter IV, article 21) LOTT to classify VTC as discretionary passenger transport (“transporte discrecional de viajeros”) – broadly the same category as unscheduled coach operations (typically bespoke group travel in a bus-like vehicle, especially for tourists). This was entirely logical – unscheduled coaches are to local buses, as VTC are to taxis – a luxury, bespoke journey variant. Unfortunately, entirely logical failed to account for prior regulatory differences: While Barcelona (for example) has always regulated local bus numbers via a monopoly municipal bus operator, there was no complementary stipulation of just one unscheduled coach for every thirty TMB buses. To do so would limit the city to just 35 tourist coaches, scarcely enough to disgorge the manifest of one cruise-liner onto Les Rambles. The EU Services Directive was always going to be fractious in Spain because its principle of “freedom of establishment” inverts “visto bueno” (for example, a posteriori inspection instead of a priori authorization), but the fractures created in the taxi business were particularly deep because the liberalisation had such scant regard for the quirks of the past.

In 2013, the Spanish government (now with Mariano Rajoy, stability incarnate, at the helm) re-balanced the EU requirements of the Directive against the more traditional principles of autogestión, in which workers participate in company decisions, somewhat akin to an extended family. VTC licencing largely returned to that which had gone before, and by the time Uber arrived in Spain (a year later) any VTC application would be judged against the old 1:30 ratio – which in practice would mean Uber could never operate enough vehicles using VTC licences to challenge the traditional taxi business. At least we must presume that’s what the government hoped would happen. Unfortunately in the intervening years (2009 to 2013, and in some localities, to 2015) VTC licences were being authorised without regard for the 1:30 ratio. VTC licences are held in perpetuity, even transferable as property, so unless the number of taxis dramatically increases (which is highly unlikely), in many areas the number of VTC licences will henceforth forever exceed the 1:30 ratio, and thus no new licence will ever be issued again. In “shutting the stable door after the horse has bolted”, government sought to limit the chaos. But in doing so they lost the ability to actively balance VTCs against taxis, and thus exposed the reason for the ratio, officially a tool of transport policy, as a sham.

In (the province of) Barcelona there are (as of July 2017) 768 VTC licences against 10,754 taxi, a ratio of 1:25. Nationally the ratio is 1:11, and in Madrid, 1:7. This is why Uber could re-launch in Madrid in 2016 with a VTC-based operation, but not in Barcelona: There simply aren’t enough VTC licences in Barcelona to match Uber’s market ambition, to allow them to offer sufficient service to meet expected demand. The more pragmatic Cabify is able to operate in Barcelona with just 50 VTC licences (when their cars aren’t being forced off the road by angry Taxistas). Meanwhile in Madrid, with its higher ratio of VTC licences, Uber can (presumably) find enough VTC licences prepared to work under the Uber name. Not that Madrid has been especially welcoming: Madrid’s pollution ordinance, which can limit individual vehicle use to alternate days, makes no exception for VTCs, unlike taxis. Uber called Madrid’s bluff – or for the less cynical, acknowledged Madrid’s environmental policy concerns – by deploying a fleet of electric cars. Meanwhile intending VTC applicants are read the riot (er, ROTT) act and sent away. But at least Madrid publicly acknowledges and reasons its refusal to issue new VTC licences.

Licensed Speculation

While ROTT provides the detail on the principles of LOTT, each is a separate piece of legislation. LOTT was reformed in 2013, but the complementary changes to ROTT did not happen for another two years. Why the Spanish Parliament thought this a sensible way to implement law is unclear, not least to the judiciary, who are now expected to make sense of a two year contradiction. The Tribunal Superior (High Court) of Justice of Madrid has sided with ROTT, and compelled local VTC applications made in the gap (between 2013 and 2015) to be granted. Courts in other provinces, such as Asturias and Navarre, have sided with LOTT, and rejected local VTC applications made in the gap years. The High Court of Justice of Catalunya (in Barcelona) has stalled without reaching a decision. In July 2016 this national inconsistency reached the Audiencia Nacional, the court with jurisdiction over Spain, who referred the matter to the Spanish Supreme Court. After another year of delay, the Supreme Court’s judgment is still awaited. The Spanish competition authority (CNMC) has rejected several key aspects of the 2015 ROTT changes, and presumably is waiting for the Supreme Court judgment before challenging the legislation in the courts.

The inconsistency in the processing of VTC applications made between 2013 and 2015 also raises practical problems. As explained in the previous section, re-imposing the 1:30 ratio was highly likely to create a situation where new VTC licences would never be issued again. Anyone who knew the industry could reasonably predict the resale, and strategic, value of a VTC licence was only ever going to rise. The presence of a future Uber or Cabify would merely add to the value. Predictably, there was a rush of new applications. The number of pending applications (awaiting a Supreme Court decision) is not officially disclosed, but media reports indicate several thousand in both Barcelona and Madrid, threatening final ratios of 1:5 or closer. That should be low enough for a viable Uber operation in Barcelona. Uber’s regional director, Carles Lloret, has hinted that Uber could return to Barcelona in the second half of 2017, assuming the Supreme Court judgment is favourable.

Resold VTC licences are still worth less than a third of a taxi licence (at least in Madrid), but the Spanish government is already nervous about a “speculative bubble” and plans further regulation to ensure VTC licence holders use licences themselves. However, the problem is not so much the speculative bubble, as who is speculating: VTC applications require access to the tools of the trade, not least vehicles. Naturally, that’s easiest for established taxi drivers. Julio Sanz, president of the “Federación Profesional del Taxi de Madrid”, notoriously attempted to apply for 5000 VTC licences, supposedly to block their use by Uber and Cabify, although that explanation did not convince the rival “Asociación Gremial del Taxi de Madrid”, who saw it as a betrayal. Sanz later withdrew his application, but the case shows how the lines of battle between VTC and taxi drivers can blur: Taxistas are as much fighting the external (Silicon Valley) enemy as fighting themselves. The eventual resolution of the current licencing speculation will almost inevitably benefit some drivers at the expense of others, promoting inequality within a driver community of equals. For an industry where tribalism and relationships matter, civil war is especially brutal. Cynically, Uber need only “divide and conquer” by letting the Taxistas exhaust themselves on one other.

I Ara Què?

So, now what?

In the words of Pere Padrosa, head of transport for the Generalitat de Catalunya, the government of the autonomous community (which includes Barcelona) has taken, “a restrictive interpretation of the law” on VTC licences. Uber’s 2014 incursion into Barcelona, perhaps more than any other factor, appears to have engendered a deep hostility to the liberalisation of the taxi sector. In November 2016 the regional government decreed additional regulation of taxi intermediaries, such as Uber. That decree was subsequently overturned by the Audiencia Nacional, who disagreed that the Catalan government’s intervention was proportional in the context of competition law. Where Madrid reluctantly capitulated to the new order, Catalunya, including Barcelona, is becoming increasingly fervent in its defense of the old order, reportedly vowing to defend LOTT in front of the Spanish Ministry of Public Works (“Fomento”). Taxistas now take to the streets of Barcelona with banners emblazoned “1/30”, as if it were a defining right. Given the financial cost of a Barcelona taxi licence (€134,000 in 2016 – resold, since no new taxi licences seem to have been issued for almost 40 years), the Taxistas have a lot at stake. In the current political turment of Catalunya, it cannot be assumed that the Generalitat will yield, should the judiciary uphold pending VTC applications. Increasingly disagreements with central government serve as rallying cries, regardless of their relevance to Independència.

The Spanish competition authority (CNMC) is reportedly so split on the issue of Uber and the wider “collaborative economy” that it cannot agree to publish the result of its own policy investigation. Both the European Commission and Parliament have recently published studies of the modern taxi market: The Commission report predictably focuses on (organizational) regulation, the Parliament on (consumer) mobility. The International Transport Forum of the OECD has also reported on taxis in the age of apps. Regulation only in extremis, emphasis on providing passenger transport, all good stuff. Yet still struggling to rationalise historic interventions (as discussed in Regulata España), and hence not one iota of use to politicians and officials faced with an angry mob bearing “1/30” placards.

Time for Change

In spite of technological disruption, taxi markets remain predominantly local, so on the European principle of subsidiarity, regulatory emphasis should also be local. Spanish regulation can look excessively recalcitrant to northern eyes, but as the Spanish implementation of the European Services Directive demonstrates, that perspective just needs to be inverted: Spain is being ask to move from a native culture of visto bueno to a more northern presumption of self-establishment. That was always going to take time, possibly a generation or more. The core failure of government was in not first applying this changed perspective to the legislature itself, before that legislature started passing laws on the subject. At least in part, the civil service continues to operate with an implicit assumption of visto bueno: That they will be informed of a potential problem and given time to fix it, and only then can the problem be allowed to occur. That contrasts (for example) to the British (Whitehall) civil service, which traditionally spent much of its time preparing for problems that hadn’t arisen yet, because it knew its citizens and businesses wouldn’t wait for the government to act when problems did arise. The VTC saga is riddled with examples, from lack of behavioural foresight, to split implementation of legislation, and a 4-year delay waiting for God-ot to say it’s okay to issue a licence for a car.

This doesn’t just weaken the liberalisation agenda (the only startup local passenger transport business that can afford a 4 year delay is Uber), it’s increasingly chaotic: A government unable to react to changes fast enough, in spite of those changes occurring in an environment that the government’s own legislation has accelerated. Spain is still a long way from the Anglo-American extreme that characterises the Brexit-Trump era – so much change as to create stability in a state of perpetual indeterminacy. But if the Spanish government is to manage European policy commitments effectively, it may itself need to get better at managing change. And if the pace of technological change increases exponentially (as it approaches the tech singularity), and is not countered by nature (silicon can only be engineered so far before it rebels), then the role of government will increasingly be to protect its citizens from the excesses of more change than those citizens can handle. Which transpires to be much like contemporary taxi regulation:

Technology has definitively disrupted the notion of separate VTC and taxi markets: Apps have made VTC “pre-booking” just as easy as hailing a taxi on the street. If it ever was, VTC ceased to be a luxury market segment some years ago, and in any case, such differentiation is primarily a commercial concern. All other things being equal, taxis drivers should gain marginal quality of life improvements from the use of Uber-style algorithms and communication systems, which match customers to cars more efficiently than the traditional wait at a taxi rank, improving the balance of driving to waiting time. All other things being equal is an important caveat, because those algorithms come at the cost of ownership, and a sense ownership is an important part of keeping taxi drivers sane. While software remains a licence built on intellectual property, that problem isn’t going away, for any of us. (Indeed, since the best algorithm is the one that incorporates all the others, eventually there will be just one algorithm to rule them, and perhaps just one technologist to own them, and perhaps then we can rethink ownership in less absolute terms.) In the meantime the pragmatic approach is a sense of ownership, something the online world is adept at manufacturing. That’s the easy part.

Family Ilusiones

The difficult part is, once again, change. But not so much the implementation of changes to the taxi market, as difficult as that evidently can be. If our opening analysis of the taxi market is correct, we actually regulate the taxi market to limit variability, to balance driving and waiting time, to maintain the base psychological conditions in which the job of taxi driver becomes workable. By (especially Anglo-American) convention, market intervention requires this be expressed in quantifiable utility. “Time is money” has been familiar to transport economists since Benjamin Franklin, and the need to weight (the value of) driving and waiting times differently is acknowledged in 1980s Cost-Benefit Analysis (albeit applied to the passenger, not the driver, who Systems theory economists tend to regard as a faceless cog in a mechanised unit of carriage). But variability is a surprisingly recent discovery to transport academia, barely acknowledged until the start of the 21st century, yet to filter down into policy – perhaps because it questions the Empirical certainty of time, which is at the root of Enlightenment economics and governance (not to mention physics).

The all-pervasive CaixaBank are currently advertising loans under the banner “Family Ilusiones”, or “Family Il·lusions” in Catalan. It’s the kind of marketing inglés that confuses native English speakers: Who buys an imaginary family? The confusion is the shared Latin root, “illusio”: In English “illusion” primarily means empirically false, even a deception, with a rare secondary intonation of failed optimism – none of which is implied by La Caixa. The English meaning is secondary in Spanish (and perhaps to a lesser degree, in Catalan), which emphasise hope for the good, both imaginary and realised. Language reflects culture. When we first introduced Barcelonian vanity as a social manifestation of hope, we referenced this “illusion”. That hope, which is change over time, is embodied in the way we sense the world. That to manage time, we can lean on “illusion”. And that there is lurking here a culturally native, socially intuitive, method of regulation that is built on “illusion”, not on conventional economic control of price and supply.

That I may never intuitively, natively, understand this, erks me. If the doomsayers are correct, and automation is soon to make us all unemployed, our “economy” needs to find a way of keeping us all entertained when there isn’t enough work to go round – and more importantly make that socially acceptable – lest we become Luddites. What comes after Utilitarianism is unclear, but we gain some insight from taxis, whose regulation has probably always been intended to make the job of taxi driving workable, and only as a secondary consequence, maintain a transport utility. Thus, any method of regulation that genuinely focuses on making the job more workable, and not first on the Utilitarian market, could have application in a much wider sphere of human economy. Even if, a century hence, the idea of a person driving a taxi has become as quaint and historical as having one’s carriage drawn by a horse.

Postscript (March 2018)

Between June 2017 (when this article was written) and February 2018 little changed. The regional government, the Generalitat de Catalunya, proposed tightening enforcement of VTC operations, including hefty fines and sharing of electronic booking data. Sharing by mutual agreement, because the competition commission (CNMC) had already ruled against the attempt to legislate for this and more in 2016. The proposal was overtaken by the events of the independence process, which effectively suspended regional policymaking in October 2017. The idea was taken up by the Spanish government, who proposed allowing autonomous communities to legally require each VTC journey to be registered prior to departure. However the CNMC remained unconvinced, citing the high administrative burden (which would not be imposed on other modes of transport) and lack of policy objective (beyond protectionism). They asked the government to back down or face a challenge in the Supreme Court, a challenge Uber launched themselves.

The Supreme Court did start to resolve the licence issue in the Comunidad de Madrid, albeit for just 80 licences, a tiny fraction of the total backlog. The imposition of Article 155 of the Constitution, bringing direct rule of Catalunya by the government of Spain, presumably muted the enthusiasm of the state (in the widest sense) to resolve these contentious VTC licence applications promptly, since so many claims related to Catalunya: An issue unrelated to independence should ideally await the return of Catalunya’s autonomous government. In the meantime the Área Metropolitana de Barcelona (AMB), under the direction of Barcelona mayor Ada Colau, started drafting proposals to regulate VTC operations in the urban area of Barcelona, such that existing license numbers are maintained, because promoting taxis is considered incompatible with wider mobility policy. The proposal will surely be shot down by the CNMC for interfering in a competitive market: The CNMC already considers the exclusion of VTC vehicles from infrastructure such as bus lanes to be a “limitation on VTC activity“.

With the political and legal tide not obviously moving in Uber’s favour, Uber appears to have made the decision to put into use the few VTC licenses it had access to. At the start of March 2018 it was rumoured to be about to launch in Barcelona using 150 VTC licenses – three times as many licenses as Cabify, but not enough to launch a comprehensive service (one that would match regular taxis). Uber actually launched in Barcelona on 13th March 2018 with just 120 vehicles, undercutting existing taxis fares by only 15% – not a particularly aggressive opening offer. This capacity limitation perhaps explains why Uber waited until two weeks after an event that would have guaranteed it publicity, especially in the tech’ community, the Mobile World Congress. Judging from the extremely long line of taxis outside the venue, the Congress alone could have swamped 120 vehicles.

The majority of Uber’s VTC licenses are reported to be owned by a company called Tibus, which transpires to be 60% owned by bus operator ALSA, who are part of the National Express Group. ALSA are historically notable for disrupting Spanish transport markets in the process of liberalisation, most currently on the railways, so their involvement is noteworthy – even as a mere “investment partner”.

While Barcelona may now have Uber, and the title of this article can finally fade into history, almost all the underlying issues the case of Uber highlighted remain. Barcelona’s taxi story has not yet reached its conclusion.

Post-Postscript (January 2019)

At the start of July 2018 the Tribunal Supremo formalised the structure of the VTC sector as a state-framed Service of General Economic Interest, populated by a mercantile oligopoly of VTC licence owners that would henceforth be effectively fixed. The judgement represented a maturing of the modern VTC sector into Spain’s societal structure, where power is traditionally balanced between state and elite. And by implication, the formal Españolisation of Uber’s Spanish business – which as noted above was already heavily reliant on licenses owned by Spanish interests, with Silicon Valley providing little more than the name and technology.

However the basic tension between VTC and traditional taxi operation remained unresolved for the newly formed left-leaning minority national government of Pedro Sanchez. Sanchez’s minister of Fomento, José Luis Ábalos, opted to pass this tension on to the governments of the regional autonomous communities, thereby both redirecting the ire of the taxistas to regional politicians, and refocusing policy toward the natural scope of taxi operations – localities. He formulated a decree law that allowed VTC operations to be further regulated within regions, notably to allow a pre-contracting period to be imposed – a minimum time between booking and travel.

Decree laws are intended to maintain the spirit of the parent legislation, not to fundamentally change policy. So while such regulation makes a mockery of European policy (deregulated taxi markets), the same regulation may be judged reasonable against the original purpose of VTC legislation (limousine operation). That in turn makes it difficult for the CNMC to challenge the decree law, only to observe the failure of the Spanish state to effectively implement the underlying European directive.

The Generalitat de Catalunya’s initial policy compromise was a 15 minute pre-booking period, alongside various secondary restrictions, such as forcing vehicles to return to base between trips and banning geolocation on VTC vehicles. The compromise failed, bringing both taxistas and VTC drivers out onto the street in separate strikes. Within days AMB proposed to raise the pre-contracting period to an hour – a magnitude that would make it very difficult for Uber to compete with traditional taxi operations.

Uber and Cabify quit Barcelona on 1 February 2019 (when the Generalitat’s decree became active), and sought compensation (for the diminished value of their licences) through the courts. Such process logically necessitated a reaction to the Generalitat’s regulations (rather than AMB’s) because only the Generalitat had the financial resources to meet such legal claims (potentially in the order of billions of euros). Such challenges reflected the prior framing (by the Tribunal Supremo) of power balances between state and elite – framing which perhaps rendered operational strategy secondary and encouraged Uber to abandon drivers and customers at the first challengable administrative barrier, rather than fight for the market on-street, as had characterised Uber in earlier years.

The post-Independència politics of Catalunya added complications – Fomento, Generalitat, and AMB were already hostile to one another on a wide range of topics – but much the same issues echoed in Madrid. Uber has grown pragmatic, continuing expansion where it can in Spain, most recently in Andalucía, while being prepared to withdraw from unwelcoming cities such as Barcelona. Catalunya grows ever more regulatory, following its pre-booking fix with plans to regulate the market for taxi licenses by making licenses non-transferrable, effectively transferring long-term control from taxistas to state (although the billion euros required to buy-out existing licenses raises its own fiscal challenges for the politically weak and still heavily indebted Catalan administration).

In the mid-term Ábalos’ policy should make VTC decisions more local (albeit in the first instance regional), potentially fostering a wide variety of local regulatory regimes that may ultimately prevent any future centralised implementation of EU policy through VTCs. The hope will be that by the time the European Union starts fining Spain for non-compliance, the future will more obviously lay in automated vehicles, which eschew the underlying need to regulate taxi markets (there being no driver to maintain the sanity of). Of course that transition could yet create its own problems: VTC is logically the closest existing legal structure to that which would be required for automated vehicles. VTC license owners could yet find themselves the rulers of a far larger market for small-vehicle passenger transport without driver.