Starbucks is back in the news for having kept their promise to pay corporation tax. I’ve written quite a lot on this and I don’t think I pieced it together. So here goes…
The first thing is that the discussions about the UK accounting profit are a red herring. Yes, the management accounts under US GAAP showed a profit. Yes, the UK accounts showed a loss. An interesting inconsistency, but it does not tell you anything about tax avoidance and it can be completely misleading.
For example, reports in the media were saying that Starbucks were deliberately maximising UK losses in order to avoid tax. If this translated directly into taxable profits/losses it would mean that other group companies would be getting taxed on profits the UK business hadn’t actually made. This would increase the group’s tax liability. You might want to reduce profits arising in the UK to avoid tax, but you certainly would not want to create losses.
To get a real view of what is going on, you need to look at the taxable profits instead. Taxable profits can be simply described as: ‘the accounting profits as determined by GAAP as adjusted for all provisions required or authorised by law’.
Some variation in the business’s tax liability can arise through choice of accounting principles, but only if there is no tax legislation to prevent it.
As it happens, the three key pieces of evidence used to convict Starbucks in the court of public opinion are all subject to lengthy anti-avoidance provisions which allow the accounting entries to be overridden completely.
Now, the original Reuters piece actually acknowledges this, in roundabout fashion. Despite its accusation of tax avoidance in its opening paragraph, the article never once tries to determine whether tax adjustments are made as intended by law. In fact, it doesn’t even try to establish what adjustments required or authorised by law have been made.
The usual excuse for this sort of behaviour is that computations are not publicly available. But you can infer some of what it is on them by reading the tax reconciliation notes. And the Reuters article did not refer to them or give any indication of having looked at them.
They might have reconsidered their story if they had.
By looking at the adjustments related to creation and utilisation of trading losses, you can estimate pretty reliably what the profit or loss for the trade is. Here is a graph of the taxable profits against accounting profit. Far from telling HMRC they were unprofitable, Starbucks UK were declaring taxable profits in 6 out of the 9 years they paid royalties.
Also, there is a relatively large prior year adjustment in 2008 in the tax reconciliation. My guess is that this is an adjustment for transfer pricing relating to the payment of royalties following an enquiry by HMRC.
As we know, Starbucks UK pays royalties to a group company in the Netherlands. However, what is not stated in most of the media stories is that this was not always the case. Royalties only commenced in 2003.
So what prompted these royalties to commence?
What also changed in 2003 was that the UK operation no longer ran its own roasting plant. Instead, Starbucks had built a roasting plant in the Netherlands which began to service subsidiaries in other countries, like the UK.
So, there was a significant change in business structure in 2003. Certain economic activities (bean roasting) ceased in the UK. Certain economic activities (bean roasting) commenced in the Netherlands on behalf of the UK subsidiary.
I might be relying on Occam’s Razor to conclude that there is causality, yes. However, anybody reading Starbucks UK 2003 accounts should notice these two facts and probably consider their relationship. The directors report mentions both the closure of the roasting plant and that royalties have commenced.
Now, nobody in their right mind would think that this economic activity, this service to the UK company, would be provided for free. In fact, the Dutch tax officials might be rightly miffed if the Dutch subsidiary simply roasted coffee beans for fun.
So, transfer pricing kicks in. Starbucks are expected to decide on “an arms-length price” for these services to form the basis of the accounting entries knowing that all interested tax authorities will look at the methodology with a view to tearing it apart.
‘Fortunately’, Starbucks had a good idea what the arms-length price should be. They already had arms-length licensees who paid a consistent price for royalties. At the time of the Public Accounts Committee hearing, Troy Alstead said they had about 20 arms-length licensees who were all paying this rate.
Furthermore, Starbucks would have already considered a suitable basis for transfer pricing adjustments on the payments for coffee beans. This was done on the basis of what other arms-length licensees paid.
And if this method was ok for determining the correct transfer pricing for their coffee beans, why not their royalties?
Now, I can start to see why the full rate of royalties probably would not be appropriate for quantifying the economic activity provided by the Dutch roasting plant. For licensees, this appears to be a bundled payment for rights relating to intellectual property beyond their roasting techniques, such as use of brand. The UK subsidiary appears never to have paid for this before. It would have presumably been recharged to the US so why should it be included now? (Perhaps it should have but the US rate is higher so it would have been avoiding US tax, not UK tax)
This could explain why Starbucks agreed a lower “royalty” rate for tax purposes as a result of an HMRC enquiry. As I said above, I think that this is shown in the 2008 adjustment, which coincides with a noticeable increase in the permanent adjustments, which is what you would expect to see for the transfer pricing adjustment as agreed with HMRC.
This is not an avoidance scheme or clever accounting. This is just how companies arbitrate with HMRC to determine a mutually acceptable arms-length rate. Let’s not forget that Starbucks could have taken the point to legal proceedings. Their initial method may have been upheld. It may not have been. They simply agreed an adjustment with HMRC and it appears to have not taken an excessive amount of time.
So, what have we actually got? A company that pays the arms-length rate for its coffee beans which subsequently agreed an arms-length rate with HMRC for the genuine economic activity attached to its roasting plant? I know that some individuals don’t like the arms-length principle, but to focus on a single business that appears to be complying with its intention looks like bullying tactics.
By all means, use Starbucks as a case study about the application of the arms-length principle. Make it clear that this is how the current system works and the system needs to be changed. But do not single the company out, make public accusations of immoral behaviour and call for a mob to descend on their employees. That’s just irresponsible.
As for the third piece of evidence, Troy Alstead confirmed at the PAC hearing that the interest was paid to a US company where it was subject to a higher rate of corporation tax anyway. This seems odd for an “evil tax avoider” who could easily have made the loan arise from the Netherlands or Switzerland.
Regardless of your opinion of the other two pieces of evidence, Starbucks certainly did not take every opportunity to reduce their group’s tax liability.
All of this brings me back to their payment of “tax”. I see it as a PR solution to a PR problem and nothing more. Having said that, it really didn’t work very well and I don’t think that other businesses will repeat this act. The public saw this as an admission of guilt rather than a gesture of goodwill and Starbucks was the subject of a protest anyway.
Worst of all, Starbucks’ payment of corporation tax is now considered newsworthy and acts as a helpful reminder of their “avoidance”.