Uber London Ltd & Ors, R (On the Application Of) v Transport for London (TfL)

The licencing and regulation of private hire vehicles and of their operators and drivers within the Metropolitan Police District and the City of London is governed by primary legislation, the Private Hire Vehicles (London) Act 1988: Section 1(1)(a) defines a private hire vehicle:

The Secretary of State for the Environment, Transport and the Regions was, but Transport for London, ("TfL") now is, the licensing authority for the grant of a licence to an operator (section 3) to a driver, (section 13) and in respect of a vehicle (section 7). No person may operate or drive a private hire vehicle unless an operator's, driver's and vehicle licence has been granted to him or them, (sections 2 and 12,) and in respect of the vehicle, (section 6).

Standard conditions are laid down by the 1988 Act including the following: the operator must be a fit and proper person (section 3(3)(a)); the driver must be over 21, a fit and proper person and have held a UK driver's licence for at least three years, (section 13(2)(a)); and the vehicle must be suitable and safe (section 7(2)(a)).

Two further specific requirements of the Act are relevant to this case. The licensing authority must be satisfied that:

TfL are, in addition, entitled to prescribe further requirements for operators, (section 3(3)(b)) and in respect of vehicles (section 7(2)(c)). They must be prescribed by regulations made by TfL (section 32(1)). The decision to make the regulations is formally and in fact made by the TfL Commissioner, Mike Brown. Three regulations have been made:

On 9th June 2016, TfL amended all three regulations. Some amendments are not controversial but three are. Regulation 9(11) was added to the 2000 regulations: "At all times during the operator's hours of business and at all times during the journey, the operator shall ensure that the passenger for whom the booking was made is able to speak to someone at the operating centre if they want to make any complaint or discuss any other matter either carrying out of the booking". Regulation 3(A) was added to the 2003 regulations:

A new paragraph 11 to schedule 1 and a substituted condition 14 of schedule 2 to the 2004 regulations were added. In its original form condition 14 of schedule 2 provided:

By a claim form issued on 15th August 2016, Uber London Limit (Uber) and three private hire vehicle drivers contracted to Uber challenged the lawfulness of the three requirements plus one other (not now, in its amended form in issue).

On 9th February 2017, TfL substituted a new regulation 9(11) in the 2000 regulations:

A substantial proportion of Private Hire Vehicle drivers do not have English as a first language. Various estimates appear in the documents which it is unnecessary to cite because the parties are now agreed that it is approximately 75 per cent. The impact of the English language requirement as originally made was estimated by the Finance Department of TfL, after consulting with Mr Robinson, a senior official in the Taxi and Private Hire Department, who reported to Ms Chapman, the general manager of the department. It estimated the number of licensed private hire vehicle drivers at about 100,000, of whom in the three years 2017-2018, to 2019-2020, 16,300 would fail to satisfy the requirement; and a further number, an unknown fraction of those who would be deterred from reapplying for a licence by either the English language requirement, or by the rigorous application of the existing topographical test by TfL itself. It had been fraudulently administered by a number of private companies to which testing had been outsourced. Their total number was estimated at about 15,000. On any view, the number of licensed private hire vehicle drivers estimated to be likely to fail or to be deterred were substantial, not less than 20,000. In addition, it was estimated that there would be a significant impact on new applicants for licences. The model suggested that just under 20,000 would fail or be deterred from applying.

On a reasonably cautious view of the impact of the English language requirement, about 40,000 persons are at risk of being prevented from obtaining a private hire vehicle driver's licence in the 3 years 2017-2018 to 2019-2020. The figures are taken from TB/100/184-90.

They have been borne out by events. Of the applicants who have taken the B(1) CEFR test administered by Trinity College London, in the 3 months, October to December 2016, 45 per cent have failed and 55 per cent have passed, TB/150/1736-7.

It is common ground that a measure which has had and will have such an impact on so many people must be strictly justified. European Union law is engaged in two ways: because the English language requirement affects the right of EEA nationals to "take up and pursue activities as self-employed persons" under Article 49 of the Treaty on the Functioning of the European Union; and, because they discriminate indirectly against non-UK nationals whose first language is not English, Article 2 of Council Directive 2000/431/EC of 29th June 2000 is engaged.

Effect is given to this Directive by section 19 of the Equality Act 2010, which provides:

Application of these principles requires TfL as a public authority to establish that the measure is proportionate in the EU law sense. It was helpfully and comprehensively explained in the joint judgments of Lord Reed and Lord Toulson on R (ota Lumsden a nd o thers) v Le gal Se rvices Board [2016] AC 697:

The heart of the test which must be applied was explained in paragraphs 66 and 67:

On the facts of this case, the application of those tests requires TfL to identify the public interest in question and the level of protection which it has determined is necessary to protect that interest and to demonstrate that the required level of protection could not be attained by less restrictive means.

Ms Chapman has borne the brunt of discharging those tasks in her first witness statement of 25th November 2016. The public interests in question are the safety and welfare and convenience of the passengers. The level of protection of those interests which TfL has determined is necessary to protect them is that private hire vehicle drivers must have a sufficient command of spoken English to be able to understand their requirements, including those arising unexpectedly, for example, in a medical emergency; to discuss a route or fare with them, and to explain safety requirements to them; and of written English, to understand regulatory communications to them, and traffic and other information supplied to them by TfL.

She gives examples in both categories: a notice of recall of a class of vehicles for safety reasons and weekly traffic information, TB2/60/1195 and 68/1246-7 and TB3/120/1622-24. As Mr de la Mare QC points out, some of her other examples would ordinarily be preceded by oral interaction between a TfL inspector and a driver so that the written notice would not come to him as a surprise. Even so, in Ms Chapman's opinion, he would need to be able to read it.

Other circumstances in which a driver might need to read English, for example, a note written by or on behalf of a vulnerable passenger by the person making the booking were also relied on, but two of the claimants who were private hire vehicle drivers with substantial experience as such, have never encountered them and they are, in reality, a make-weight. The level of protection selected stands or falls on the need to understand documents setting out regulatory requirements principally for safety reasons and documents providing information principally for passenger convenience reasons.

I am satisfied that the level of protection determined by TfL on the advice of Ms Chapman and her team was within the margin of appreciation which must be allowed to it. The real issue is whether or not TfL have shown that no other less restrictive means were and are available to it to achieve those ends.

TfL has no internal language testing capacity, as the note of 6 October 2016 prepared when the English language requirement was amended makes clear, TB3/115/1571-2. The extent of the problem was demonstrated in an appendix to a submission made earlier, dated 17 March 2016 to the Commissioner, TB/36/459-461. When topographical testing was taken in-house in 15 centres whose accreditation has been suspended, out of 200 applicants, only 72, 36 per cent, passed. Of those who failed, a significant proportion could not understand the simple questions written in English in the test.

The Commissioner on the advice of Ms Chapman and her team adopted the B(1) CEFR standard as the appropriate standard to require. Ms Chapman explained why in paragraph 61 and 62 of her first witness statement. It was because that standard described an understanding of "the main points of clear standard input on familiar matters regularly encountered at work" and in other commonly arising circumstances.

The test, as administered, requires those tested to read and answer questions and write a short essay of 100 to 130 words on the topic set, which has nothing to do with driving. The test papers shown to me concerned sleep patterns. Mr de la Mare submits that this demonstrates its lack of utility for drivers, as does the likelihood that passing it would not equip a driver fully to understand the somewhat complex regulatory documents sent to them by TfL. I accept that it does not. It tests general compliance in written and spoken English. A person who can demonstrate that level of competence can be expected to develop, in short order, an understanding of regulatory documents and traffic information sufficient to permit him to understand their gist and to be able to act appropriately on receipt of them.

There is no driver- focused English test yet available. TfL propose to introduce a test of drivers' knowledge of regulatory requirements and law including disability law in the near future, see their letter to Uber of 14 February 2017, TB3/160/1774-6. This, together with the B(1) CEFR English test will better match the regulatory requirement than current requirements alone but it will not remove the need for the driver to have a level of understanding of written English to put the knowledge thus gained or demonstrated to effect.

In my judgment, TfL have demonstrated that they were and are entitled to require drivers to demonstrate that level of competence in written as well as spoken English. There is now and for the foreseeable future no practicable alternative means of achieving the protection of the legitimate public interests which TfL have identified to the level properly set by them. In reaching that conclusion, I have not found the Central Government requirement for "public facing" officials to have a sufficient command of spoken English only to deal with a comparable set of circumstances. For the reasons explained, drivers must do more than converse with passengers and understand spoken English.

Mr de la Mare also submitted that the Commissioner had not been brought face to face with the consequences for drivers who fail the B(1) CEFR test. I do not agree. He was told expressly in paragraph 13 of Ms Chapman's submission to him of 13th October 2016 that the imposition of an, in fact the, English requirement would have a significant adverse effect on those of non-UK ethnic and national origin. An assessment commissioned by TfL by Mott McDonald of January 2016 spelled out the impact on private hire vehicle drivers of the spoken element of the English language requirement in clear terms:

Mr de la Mare also submits that the strong likelihood that the measures will have a significant adverse impact on a large number of persons requires TfL to adopt less restrictive measures: TfL's proposals for testing a driver's knowledge of regulatory requirements and law should be combined with a written test devoted to those topics and implementation of any measure should be deployed until both are prescribed and put into effect. I do not agree. TfL are entitled to require private hire vehicle drivers to demonstrate compliance with the English language requirement by 30th September 2017, for the reasons which I have explained. They are not required to delay its introduction until they have equipped themselves with a language teaching and testing capacity, which they do not have, or to await the development of a purpose-designed course and test by an educational concern, if one can be found which is willing to do it.

Nor am I persuaded that TfL are in breach of their high level duty under section 149 of the Equality Act 2010 to have regard to the need to deal eliminate discrimination and advance equality at the tune. It adds nothing to the careful balancing exercise between the needs of passengers and the burdens which would be imposed on private hire vehicles, which they were required to undertake.

Finally, if contrary to my conclusions, procedural failures in the decision to impose the English language requirement should have led to the conclusion that the decision was flawed, I would be satisfied that if I were to quash it, it is at least highly likely that the same decision would lawfully be made again. In those circumstances, I would be required by section 31(2)(A)(a) of the Senior Courts Act 1981 to refuse to grant relief.

The wording of the telephone requirement is very broad. The operator is required to:

It requires a large app-based operator such as Uber to maintain, round the clock, a telephone service to deal with complaints and any other matter about the carrying out of the booking immediately. It may not be possible to quantify the cost of doing so precisely. It has been estimated at between £700,000 if done from India and £3.4 million if from Uber's call centre in Limerick, as Mr Byrne, Uber's head of public policy for the UK, explained in his witness statement of 12th August 2016. On any view, the cost will be significant. The cost to operator drivers, those with no more than two private hire vehicles, is not readily quantifiable but is likely to be significant in relation to the 700 or so who fall into this category.

Uber maintained in evidence, which is not the subject of significant challenge, that they have an efficient and cost effective system for dealing with complaints, from those which need to be deal with immediately, to those which can wait. It is summarised in paragraph 18 of the second witness statement of Mr Elvidge, Uber's general manager:

Mr Chamberlain QC has subjected this paragraph to close examination and invites me to draw the conclusion that it takes up to 6 hours for Uber to deal with truly urgent complaints. I do not accept this analysis. What Mr Elvidge demonstrates is that Uber have a system which effectively sorts real emergencies from all which are only potentially critical or urgent and deals with each appropriately; the first immediately, and all others within 6 hours as to 70 per cent and within 24 hours as to 99 per cent. This is to be commended rather than criticised.

Mr Elvidge also demonstrates in paragraph 21 of his witness statement how Uber deals with the 4 to 5,000 reports of missing items which are received each week. The system he described is at least as good as any other, and in terms of end result is likely to be better because of Uber's ability to identify and contact the re levant driver and to put a passenger who has lost an item in direct touch with him.

Ms Shannon and Mr Smithers, TfL's investigations' manager, justify the current telephone requirement on two essential grounds: many passengers will be reassured by their ability to speak to a human being about their complaint and emergencies can be dealt with more quickly than would be the case without the intervention of a live telephone operator. A small number of real life examples are given. I accept their reasoning, but only up to a point. I can readily understand that some passengers, believing themselves to be confronted by an emergency, will find it reassuring to speak to a human telephone operator.

TfL were and are entitled to identify the public interest affected as public safety and convenience, in the case of an emergency principally the former, and to conclude that a high level of protection should be given to passengers who believe that they are faced with an emergency. I also accept that in that instance there is no alternative measure which could combine reassurance and speed of response to some passengers. I do not understand Mr de la Mare to resist that conclusion. If my understanding of his position is wrong, I would nevertheless have reached it for myself.

Accordingly, despite the fact that Uber's current system may well provide an effective and timely response to genuine emergencies, I accept that it cannot provide the level of reassurance which some passengers may, for good reason, require. This aspect of the measure is therefore lawful. It does not follow that the regulation is lawful in its full width.

Uber's systems do provide effective means of dealing with non-emergency issues. Their clientele is, by self-selection, willing to make use of the app-based facility offered by Uber to book private hire vehicles. Most of them can, accordingly, be expected to be content to use the same facility, at least initially, to resolve non-emergency issues. There is no evidence to suggest that they will not be. No survey had been conducted of the users of app-based booking facilities, only of those who use private hire vehicles and taxis generally. Mr Chamberlain's extrapolation from Uber's assumption that 25 per cent of passengers using its facilities will telephone to complain or discuss the carrying out of their booking is unsound. It was simply an assumption made for the purpose of costing TfL's proposal and cannot legitimately be used for any other purpose.

Given that the evidence shows that no useful purpose would be served by requiring Uber to replace or duplicate its current efficient system for dealing with non-urgent issues raised by the passengers, TfL has not shown that less restrictive measures could not be adopted than those contained in the telephone requirement to meet the level of protection which they require for the only public interest at stake, passenger convenience. It is accepted by Mr Chamberlain that Article 49 is engaged even though the restriction of Uber's freedom of establishment may not be great. It follows, therefore, that I must quash regulation 9(11) and leave it to the Commissioner to make a fresh regulation if he considers it desirable to provide for, in colloquial terms, a "hot- line" for emergencies broadly defined to include situations in which immediate action is required to remedy a breach of the law. The Commissioner will also need to reconsider the nature and extent of any telephone requirement on operator drivers who may find it easier to acquire, or club together to provide, a hot- line than they can a telephone service to deal with all passenger complaints and comments about the carrying out of their booking.

I can deal with this shortly because of a proper concession made by Mr Chamberlain on instructions from TfL. They now "accept that in taking the decision to impose the motor insurance requirement TfL did not take into account the extent to which claims against the Motor Insurers' Bureau would in fact provide passengers with equivalent protection", in other words, to that provided by the insurance requirement.

He was right to make this concession. Ms Chapman's reasoning was based on the need to ensure that, for the protection of passengers, no gap existed between the cover which an insurer provided and the circumstances in which a private hire vehicle was being driven when they were injured in consequence of the careless driving of the licensed driver. A paradigm case was that of the driver who had domestic and pleasure cover only but was carrying passengers for reward. She appears to have thought that in those circumstances, the passenger would experience at least difficulty in recovering full compensation, or at worst might recover nothing.

If so, this was a misconception. As the Court of Appeal explained in BristolAlliance Limited Partnership v Williams [2013] QB 286 the combined effect of part V1 of the Road Traffic Act 1988, article 75 of the memorandum and articles of association of the Motor Insurers' Bureau and the agreement which has been made between it and the Secretary of State for Transport, on and since 31st December 1945, is that the insurer under a policy in respect of the vehicle in which an injured passenger is carried, whether for reward or not, must meet a judgment obtained by the passenger against the driver of the vehicle even if the use at the time was other than that permitted under the policy. There, is in reality, no gap which, in the interests of passengers, is required to be filled by the insurance requirement. As I am invited to by Mr Chamberlain I therefore quash this requirement.