D.E.L.T.A. Merseyside Limited and another v Uber Britannia Limited

This Supreme Court case answers a narrow but commercially important question for the private hire sector outside London and Plymouth. When a licensed operator accepts a booking, must it immediately become the principal party to the hire contract for the journey?

The court said no. Part II of the Local Government (Miscellaneous Provisions) Act 1976 does not force operators to use one single contract model. The statute regulates mainly through licensing and record-keeping, not by requiring every booking to be accepted under an immediate principal hire contract.

That matters because many operators do not all work in the same way. Some contract directly with the passenger. Some act as agent for the driver. Some undertake only to try to source a vehicle, with the hire contract being made later. The court held that the Act itself does not outlaw those alternatives simply because they are not the same as Uber’s preferred model.

Uber Britannia Limited wanted the courts to confirm that the private hire regime outside London required all operators to accept bookings only by entering, as principal, into an immediate contract of hire with the passenger for the journey. The judgment called this the hire contract model.

D.E.L.T.A. Merseyside and Veezu opposed that view because their businesses used other structures. The court described an agency model, where the operator accepts the booking as agent for the driver, and an intermediary model, where the operator undertakes only to use best endeavours to find a driver and any hire contract is made later between driver and passenger.

The dispute had real commercial consequences. If Uber’s argument had succeeded, operators outside London would have needed to redesign customer terms, booking systems, risk allocation and possibly insurance arrangements. It also followed Uber’s own change of model after earlier London litigation, when the Uber group moved to the hire contract model from mid-March 2022.

The proceedings began against Sefton Metropolitan Borough Council, a licensing authority under the 1976 Act. Sefton stayed neutral. D.E.L.T.A. and Veezu stayed in the case because they had a direct commercial interest in defending the lawfulness of their existing models. Uber won in the High Court, lost in the Court of Appeal, and then lost again in the Supreme Court.

The Supreme Court had to decide whether Part II of the 1976 Act makes it unlawful for a private hire operator outside London and Plymouth to accept a booking unless the operator immediately enters, as principal, into a contract of hire with the passenger for the requested journey.

This was a question of statutory interpretation. The court looked at the 1976 Act as a whole, not by trying to mirror the separate London legislation. A key point was the statutory definition of “operate”, which means making provision in the course of business for the invitation or acceptance of bookings for a private hire vehicle.

The court also focused on section 56(1), which says that every contract for the hire of a licensed private hire vehicle is deemed to be made with the operator who accepted the booking, whether or not that operator provided the vehicle. Uber argued that the Act required principal liability from the start. D.E.L.T.A. and Veezu argued that the Act allowed different booking models and used section 56(1) to attach liability where needed.

The wider statutory context mattered too. The court described the 1976 Act as a licensing scheme with a triple lock. The vehicle must be licensed, the driver must be licensed and the operator must be licensed. The Act also creates offences for unlicensed use, driving or operation, and it allows local authorities to attach conditions to operator licences.

The Supreme Court dismissed Uber’s appeal. It held that there is nothing in the 1976 Act that expressly imposes the prohibition Uber argued for, and nothing in the Act or its purposes that justifies implying such a prohibition.

The court said that “accepting a booking” does not mean only accepting a contractual obligation as principal to perform the hire. In other words, the statutory language does not force every operator into the hire contract model.

A central part of the reasoning was section 56(1). The court treated it as a deeming provision that fixes the operator with contractual liability where a hire contract is made, regardless of how, when and between whom that contract arises. That reading makes practical sense if the Act assumed that different booking models would continue to exist.

The court gave examples. If an operator uses an agency model, section 56(1) can deem the operator liable even though it would not otherwise have personal liability at common law. If an operator uses an intermediary model and the hire contract is later made between driver and passenger, section 56(1) can still attach liability to the operator who accepted the booking. The same logic can apply where a booking is passed to another operator.

The court said Uber’s interpretation would make section 56(1) effectively unnecessary. If every lawful booking already had to be accepted by the operator as principal under an immediate hire contract, the operator would already be liable at common law. On that reading, the deeming provision would do no real work.

The court also rejected Uber’s reliance on later deregulation provisions about sub-contracting under section 55A. Those provisions were treated as permissive and deregulatory, not as evidence that the original 1976 Act had always contained an implied ban on other booking models.

This decision gives operators outside London and Plymouth more confidence that the 1976 Act itself does not dictate one mandatory contract model. But that does not mean every model is low risk. Flexibility in legal structure creates practical compliance work.

The first lesson is to separate licensing duties from contract design. The statute requires the right licences and records. It does not automatically answer who promises the journey, who bears customer liability before a hire contract exists, or how your terms should deal with failed fulfilment, complaints or passed bookings.

The second lesson is to avoid mismatch. If your app tells passengers that your company provides the journey, but your terms say the driver contracts directly, that inconsistency can create disputes. The same applies if your call handlers describe the service one way, your driver agreement says another, and your records suggest something else happened in practice.

The third lesson is to be careful with passed bookings. The judgment discusses statutory sub-contracting under section 55A, but the core holding is not that all passed-booking arrangements are automatically safe. The real point is that the 1976 Act does not require every accepted booking to involve an immediate principal hire contract. Businesses still need to check how their own arrangements work and whether local conditions add extra rules.

The fourth lesson is not to overread this case into London. The court deliberately treated the 1976 Act on its own terms and did not express a view on whether the London decision was right or wrong. If you operate in London as well as elsewhere, you should not assume the same answer applies across both regimes.

The fifth lesson is that public protection arguments do not always win if the statute does not support them. The court accepted that Uber’s model might better protect passengers in some situations, such as where no vehicle ever arrives. But it still refused to read that model into the Act by implication.

After this case, the most useful exercise for an operator is an end-to-end review of how bookings are invited, accepted, recorded, allocated and fulfilled. The aim is to make sure your paperwork and your real-world process say the same thing.

Start with the customer journey. Look at your website, app screens, booking confirmation messages and telephone scripts. Do they say your business is providing the journey, arranging a driver, or using best endeavours to source a vehicle? Those statements should line up with your terms and conditions.

Then review your driver and partner operator documents. If your business says it acts only as agent or intermediary, your agreements should support that position. If you pass bookings to another operator, your records and contracts should show how that happens and who is responsible for what.

Finally, check your record-keeping. The judgment highlights that section 56 contains express record obligations for operators, and that local authorities can impose conditions on operator licences. Good records are not just an admin issue. They help show what model you actually used for a booking and whether your process complied with local requirements.

It is also worth checking your complaints process. The judgment notes that licence conditions may include operational requirements such as maintaining a complaints process. If a passenger complains that no car arrived, your internal documents should make it clear who investigates, who responds and what remedy is available under your chosen model.

This case is specifically about private hire vehicle regulation outside London and Plymouth under Part II of the Local Government (Miscellaneous Provisions) Act 1976. The Supreme Court was not deciding the London regime, and it did not try to align the two statutes by analogy.

The judgment is also not a general approval of every possible booking structure. It decides a point of statutory interpretation: whether the 1976 Act itself forces one immediate principal contract model whenever a booking is accepted. The answer was no.

That means businesses still need to ask separate questions about local licence conditions, insurance, consumer-facing promises, complaints handling and the practical consequences if no vehicle arrives. Those issues may still create risk even where the basic statutory argument is now settled.

The judgment was given by the Supreme Court on 29 July 2025. The appeal was heard on 2 July 2025. It was an appeal from the Court of Appeal’s decision in July 2024, and the Supreme Court agreed in substance with the Court of Appeal by dismissing Uber’s appeal.

For most operators, the practical position after this case is straightforward. The 1976 Act does not itself force one contract model for accepting bookings. But your business still needs to comply with the licensing framework and any conditions attached to your operator licence.

If your business operates across more than one area, this is a good point to review whether your documents and workflows are consistent across those areas, and whether any local conditions require a different operational approach even though the core statutory point has now been settled by the Supreme Court.