Supreme Court confirms that Uber drivers are workers
22 February 2021
The Supreme Court has unanimously decided that drivers engaged by Uber are workers rather than independent contractors. It also decided that drivers are working when they are signed into the Uber app and ready to work.
In 2016, various drivers brought claims against Uber for the national minimum wage, holiday pay and detrimental treatment for whistleblowing. To succeed, the drivers had to be “workers” for the purposes of the relevant legislation rather than independent contractors.
While there are (confusingly) slightly different “worker” tests in different statutes, for present purposes a worker is either: (a) an employee (i.e. employed under a contract of employment); or (b) someone who works under a contract through which they undertake to perform work personally, for someone who is not by virtue of that contract their client or customer. In other words, workers agree to work personally and are not running their own business.
This definition has been considered in a series of “gig economy” cases, with Uber being perhaps the most high-profile. Uber argued that the workers were independent contractors, and after successive appeals the case finally arrived at the Supreme Court (SC).
Key facts of the case
People using Uber’s service hail private-hire vehicles via a smartphone app. Uber locates the nearest driver and informs them of the request and, once the booking is confirmed, the driver and passenger can contact one another through the app. A route is plotted by the app and at the end of the trip the fare is calculated by Uber, based on GPS data from the driver’s smartphone.
Uber’s terms (with both passengers and drivers) stated that it did not provide transportation services but acted as agent for third-party providers, i.e. the drivers. Uber contended that it was providing “lead-generation” opportunities for self-employed drivers. It did, however, impose certain requirements as to how the drivers provided the services. For instance, it would deactivate a driver’s access to the app if customer ratings fell below an acceptable level. It also told drivers they should log out of the app if they did not wish to carry passengers.
The earlier decisions
The Employment Tribunal (ET) concluded the drivers were workers. It found that Uber was in the business of providing private-hire services rather than generating leads for drivers to grow their own businesses, taking into account the significant control Uber exercised over the drivers. The ET also concluded that drivers were engaged as workers for so long as they were in the territory in which they were authorised to work, signed into the Uber app and ready and willing to accept bookings.
The Employment Appeal Tribunal (EAT) agreed that drivers were in reality incorporated into the Uber business of providing transportation services, rather than working in business on their own account. The EAT said that the drivers were clearly workers when they had accepted trips, but it was less sure the same applied in between accepting assignments. This issue is important because it is relevant to a determination of the drivers’ “working time” and their entitlement to the national minimum wage.
The Court of Appeal (CA) decided by a 2:1 majority that the drivers were workers. Although the written contractual terms said that Uber only acted as an intermediary, this did not reflect the practical reality of the relationship where Uber had significant control of the drivers. The CA majority judges also found that the drivers were workers at all times when they had the relevant app switched on. Although this issue was difficult, the key factors were the high level of trip acceptances required from drivers, and the penalty of being logged off if three consecutive requests were not accepted within a ten-second time frame.
The other CA judge (Lord Justice Underhill) disagreed with these conclusions, on the basis that the terms of the agreement made it clear that the drivers were not Uber’s workers. An agreement needed to be inconsistent with the reality in order to be a sham, which Underhill did not consider was the case here.
The Supreme Court’s decision
The SC unanimously rejected Uber’s appeal.
The SC first rejected Uber’s argument that it operated as a booking agent for drivers. This part of the SC’s decision turned on Uber’s specific contractual arrangements. A key problem for Uber was that the operating licence was held by Uber London, not by individual drivers, and there was no written agreement between Uber London and drivers. Uber argued that, despite the lack of a written contract, Uber London was acting as an agent for the drivers. The SC was not convinced that an agency model complied with the licensing regime, but decided that, in any event, the drivers had never actually authorised Uber London to act as their agent.
Given this conclusion, it was difficult for the SC to see how Uber’s business could operate without Uber entering into contracts with drivers under which they undertook to carry out the bookings that it accepted. Nonetheless, the SC suggested that the importance of the issue meant that it wouldn’t be right to decide the appeal on the basis of the arguments about agency alone, so it also went on to look at the wider arguments.
At the centre of those wider arguments was the SC’s earlier judgment in Autoclenz Ltd v Belcher. In that case, the SC concluded that car valeters were workers, despite contractual documentation suggesting otherwise. According to Autoclenz, employment status cases should not be determined by applying ordinary principles of contract law. This reflected the fact that, in an employment context, the parties frequently have very unequal bargaining power.
In the Uber decision, the SC endorsed Autoclenz and spelt out a new theoretical justification for this approach. In employment status cases, individuals are claiming the protection of statutory employment rights, created by legislation. This means that the task for ETs is not to identify whether a business has agreed under the terms of its contracts to pay, for example, the national minimum wage or annual leave. Instead, their task is to determine whether individuals fall within the statutory definition of a “worker” to qualify for these rights irrespective of what had been contractually agreed. As the SC summarised, the approach must be one of “statutory interpretation, not contractual interpretation”.
The SC observed that the general purpose of employment legislation governing working hours and minimum wage etc is to protect vulnerable workers. The fact that a business is often in a position to dictate contract terms gives rise to the need for statutory protections in the first place. It therefore could not be right that a business could use its written contracts to determine who qualifies for protection.
Adopting this approach to determining whether Uber drivers were “workers”, the SC concluded that, although the drivers had substantial autonomy and independence in some respects, the factual findings of the ET justified its conclusion that the drivers were workers. In particular, Uber’s control over their remuneration was of major importance. The drivers’ ability to charge less but not more than the fare suggested by Uber meant that their notional freedom was of no possible benefit to them. Overall, drivers’ services were in fact “very tightly defined and controlled by Uber”.
In light of this conclusion, the SC also had to assess when the drivers were working. Again, it followed the CA’s decision to find that they were working whenever they were logged in. The SC placed particular weight on Uber’s practice of logging out drivers who were failing to accept bookings and keeping them temporarily logged out even if they were ready to work. This pointed to there being a penalty for drivers who failed to comply with an obligation to accept a minimum amount of work when logged in. The existence of such an obligation even when drivers were not performing a booking meant that drivers were working whenever logged in.
The SC’s decision in this case has been eagerly awaited by HR and employment practitioners since it was heard in July 2020.
Like other recent status cases, the judgment is very specific to its own facts. Other claims concerning people working in the gig economy will not necessarily be decided in the same way. However, the SC’s emphasis on the need for “statutory interpretation, not contractual interpretation” is potentially significant. It is likely to sit alongside the SC’s earlier decision in Pimlico Plumbers v Smith - where the SC ruled that a sufficiently broad, genuine and unfettered contractual right to appoint a substitute would result in the personal service requirement not being met, meaning that someone would not be a worker - as the leading guide to judicial decision-making on this topic.
As the SC has dismissed Uber’s appeal, the case will now return to the ET to decide the substantive claims, which concern holiday pay and minimum wage. The ruling that drivers were working whenever they were logged in is likely to increase the cost implications for Uber compared to if, for example, drivers had only been working while performing bookings. Uber is set to argue, however, that it has changed its model since 2016, and that the SC’s findings are limited only to those drivers who brought this claim.
The SC recognised that one of the most difficult issues in this case was whether a driver who was logged on could be said to be working even if they were also logged on and making themselves available for work with one or more competitor platforms. It concluded that this question cannot be answered in the abstract but will always be a matter of fact and degree. In this case, the ET had been provided with no evidence that drivers were in practice also making themselves available to Uber’s competitors.
Producing evidence that individuals are making themselves available to other businesses while logged on to an app – so-called “multi-apping” – may become a more important feature of future status cases. We can also expect to see some businesses make adaptations to their models - for example, by providing greater freedom to drivers to reject orders without facing sanctions, or potentially restricting the times when individuals can log in.
Finally, the government’s Good Work Plan in response to the Taylor Review was due to look at legislation to improve the clarity of the employment status tests, but no specific proposals have yet been put forward. The government’s planned Employment Bill is also awaited. It is unclear whether the SC’s decision means that the impetus for further legislative reform in this area has now diminished or whether the government might now be tempted to revisit this topic, potentially even abolishing the concept of “worker” status altogether.
Uber BV v Aslam and others - judgment available here