The UK-Spain tax treaty explained: how it works and what it does not cover.

Most confusion about treaties comes from a single misconception: people read them as if they impose tax. They do the opposite. A double taxation convention is a restrictive instrument. It allocates taxing rights between two states and limits what each may do, but it never creates a liability that domestic law did not already create.

The previous instrument dated from 1975 and had aged badly. It taxed interest at 12% and royalties at 10%, used a permanent establishment definition written before modern service businesses existed, and lacked the exchange of information provisions that had become standard. The replacement, signed on 14 March 2013 and in force from 12 June 2014, modernised all three.

Two of those rows deserve emphasis. Zero withholding on interest and royalties is not a minor improvement; it removes source taxation entirely from two of the three classic passive flows. And the 10% direct holding threshold for nil dividend withholding is lower than many treaties require, which makes ordinary group structures work without contrivance.

Everything else in a treaty depends on residence, because the distributive rules are written as "a resident of a Contracting State". A person resident in neither state gets nothing from the treaty at all.

Residence is first determined by each country's own law. The UK applies its statutory residence test; Spain applies the 183-day count, the centre of economic interests, and the family presumption. Where both conclude that the person is resident, the treaty applies a sequence and stops at the first test that produces an answer.

Article 5 defines when a business presence in the other state becomes taxable there. It is the gateway to Article 7, and it is where most corporate disputes in this corridor originate.

The practical consequence of the MLI changes is that a great deal of pre-2018 guidance on this article is now wrong in the same direction: the definition is wider than it was, and structures built on the old "they do not sign contracts" defence are exposed.

These three articles follow a common architecture worth understanding once, because it recurs in every treaty. Each says: the income may be taxed in the residence state; it may also be taxed at source, but the source tax is capped; and the cap applies only if the recipient is the beneficial owner.

Article 13 allocates taxing rights over gains, and its structure is a hierarchy: some categories are taxable at source, and everything not listed falls to the residence state.

The land-rich provision matters disproportionately in this corridor because of the volume of UK-held Spanish property. Holding a Spanish villa through a company and selling the shares rather than the asset does not remove the gain from Spanish taxing rights where the company's value derives principally from that property. Structures designed before such clauses became standard should be reviewed rather than assumed to still work.

Employment income is taxable where the employment is exercised — physically performed — with a narrow exception preserving residence-state taxation for short assignments. The exception has three cumulative conditions, and failing any one of them defeats it.

Directors' fees follow a separate article and a different logic: they may be taxed in the state where the company is resident, regardless of where the director performed the duties. A UK resident on the board of a Spanish company is therefore treated differently from a UK resident employed by one, and the two are frequently conflated.

No other treaty question affects as many people in the UK-Spain relationship, and the rule is cleaner than the folklore around it suggests.

One further consequence follows from the government service rule. Income taxable only in the UK is still relevant in Spain for rate purposes in some circumstances, and it does not remove a Spanish resident's obligation to file. Exclusive taxing rights are not the same as invisibility.

Where both states retain a right to tax, the residence state gives relief. This convention uses the credit method rather than exemption: the residence state taxes the income but allows a credit for tax properly paid in the other state, capped at the residence-state tax on that income.

Both states are parties to the OECD multilateral instrument, which modifies covered bilateral treaties without renegotiating them. The operative text is therefore the 2013 convention as modified, and reading the original alone will produce incomplete answers.

The principal purpose test deserves particular attention because it changes the burden. It is not enough that a structure satisfies the letter of an article; where obtaining the benefit was one of the principal purposes, the taxpayer must show that granting it accords with the object and purpose of the provision. Documentation of commercial rationale, made at the time, becomes the substance of the defence.

This is the most important structural change in the relationship and it is easy to miss, because the headline outcomes barely moved.

Alongside this sits the non-tax consequence, which for most businesses is larger: a UK company is now a third-country supplier in the EU, with the contracting, VAT, data protection and procurement friction that entails. That is not something a tax treaty can fix, and it is the main reason UK groups establish Spanish subsidiaries.

A convention on income and capital does exactly that, and the gaps around it are where UK-Spain planning most often fails. None of the following is addressed by this instrument.