Remember the Employment Tribunal’s Uber decision?

I wrote about it here. It defined Uber drivers as “workers” for employment law purposes, giving them certain employment law rights. As I pointed out, if those drivers were also ’employees’ (a somewhat narrower status than ‘worker’ and a question not before the Employment Tribunal), Uber would accrue a UK National Insurance Contributions bill of around £13m per month.

But it also has a further – and rather more fundamental – tax problem.

Let me explain. And for clarity I’ll reduce the argument to its essentials:

(1) before the Employment Tribunal, Uber contended that it simply acted as a booking agent for drivers. The Tribunal disagreed. It found, applying a normal contractual analysis, that Uber engaged drivers and supplied transportation services to passengers;

(2) what does this mean in terms of VAT? Well, for VAT, you start with the normal contractual analysis. But national tax authorities can also go beyond that analysis to discern the underlying “economic and commercial reality of the transactions” (as the case law puts it). I don’t, for the purposes of my argument, need to take that extra step. If the VAT analysis follows the contractual analysis then the following points apply. But the VAT test is wider than the contractual one – and even if Uber’s appeal against the Employment Tribunal decision succeeded it could still have a VAT problem;

(3) as things stand, and applying the reasoning of the Employment Tribunal decision, Uber seems to be making VATable supplies to passengers of transportation services. And those services are standard rated. In practice, this means that, of every £100 charged to an Uber customer, Uber would have a so-called ‘output’ tax liability of £16.67 (being the VAT on such sum net of VAT as, when VAT is added, gives you £100). And it would need to hand that sum over – less any ‘input’ tax – to HMRC;

(4)  output tax is the VAT you charge your customers. And input tax is the tax you are charged by your suppliers. It’s the difference – the tax on the value that you add – that you hand over to HMRC. But does Uber have any input tax? Your employees don’t charge you input tax. Uber might have some external costs on which VAT has been charged – but not many. On the assumption (see (1) and (2) above) that the VAT reality of Uber’s business is that it is engaging drivers and supplying transport services to passengers, the vast majority of its expenditure will be the money it pays to drivers. But (with perhaps a tiny number of exceptions) drivers don’t charge Uber VAT on their fares. Indeed, they are incentivised to earn less than the VAT registration threshold. If they earned more, they would have to hand over 16.67% of their profits to HMRC in VAT;

(5)  if you assume that Uber has no material input tax to set against its output tax, that would mean that, of every £100 of fares Uber has collected, it has a liability to pay VAT to HMRC of £16.67. It seems as though Uber racked up about £115m in fares last calendar year. This would mean it had a VAT liability of just under £20m for London for that year. But HMRC can go back four years or, sometimes, more. There is no suggestion in the accounts of the relevant Uber entity – Uber London Limited – that it was aware it had this risk;

(6) but Uber’s problems don’t end there. It appears that the structure Uber uses in London is replicated across the EU. VAT should operate identically across the EU. And so, if the same structure is used, the VAT treatment of Uber’s services for every other City in the EU should be identical to that in London. Different member states have different rates of VAT – and different rules for claiming unpaid VAT. But that £20m per annum for London just got an awful lot bigger;

(7) nor are Uber’s problems merely historical. They don’t end with it coughing up to HMRC any unpaid VAT which is due. Going forward, absent material change in its operating model, Uber’s revenues in the UK net of VAT will fall by 16.67%. And by equivalent percentages across the rest of the United Kingdom and EU;

(8) there is a basic difficulty in changing Uber’s operating model to avoid this consequence. The Uber commercial proposition is based on Uber’s brand value to actual and would be consumers. But there is a tension between the notion (a) that the brand name ‘Uber’ will encourage customers to book rides and (b) that the service provided by Uber is not provided to customers and is not the provision of rides. Indeed, it is this tension that gives Uber both a commercial advantage over its competitors but also a fiscal difficulty;

(9) we should watch with care what actions, if any, HMRC take. I should have absolute confidence that HMRC will properly investigate the potential NIC and VAT liabilities of the Uber structure. I don’t.

A few concluding observations.

The Uber Employment Tribunal decision is being appealed – and it could be overturned. However, I have spoken to a number of specialist Employment QCs and few expect it to succeed. Uber has declined to comment on its VAT position pending its appeal.

It is possible that Uber could change its structure to avoid a VAT liability going forward – but any resolution will have to overcome the basic tension I identified at (9) above. It is also possible that my analysis is simply wrong.

There are a number of pieces of litigation currently before the Court of Justice of the European Union – including this one – which may affect the analysis set out above.