A Welsh tax haven?

Tax devolution for Wales creates new policy opportunities but in assessing them Wales’ special circumstances must be taken into account. Although many people migrate in and out of Wales every year the flows at present are broadly balanced; net migration is very low, unlike the gross flows. The Welsh economy is very highly integrated with England, more so than the Scottish economy. As the Holtham Commission report pointed out over 1.4 million people in Wales (48% of the total) live within 25 miles of the border with England, and 2.7 million people (90% of the total) live within 50 miles of the border. Only 5% of the combined population of Scotland and England lives within 50 miles of the border between those countries. Any tax difference between the countries is therefore more likely to trigger net migration between England and Wales than between England and Scotland.

An important question is how much migration would be induced by what sort of divergence between Welsh and English tax rates. This matters because movement of taxpayers affects tax revenues; in the limit, migration could ensure that revenue was boosted by a tax cut and reduced by a tax rise. It is tempting to conclude that likely differentials will be small enough not to figure in calculations as to which side of the border to live. But while this is likely to be true for the average migrant, there could be quite a few among the many moving in both directions for whom small differences might tip the balance.

To gain some idea of the size of possible tax-induced mobility we looked at migration between local authority areas in England and Wales. We controlled for the obvious major influences on such movements – relative population size, distance between local authority pairs, house price differentials and so on – in order to isolate the effect of council tax rates that can differ substantially between local authority pairs. The objective is to deduce from these actual tax divergences how what the effects would be of a tax differential that has not yet happened – such as different Welsh income tax or sales tax rates. We use a simple model of the Welsh economy to transform estimated council tax effects into hypothetical income tax effects. Finally, we take account of some limited spillover effects of tax changes on to the economy as a whole, in particular the impact of other sources of tax revenue.

Not surprisingly perhaps, we find that changes in the Basic Rate of income tax (20%) have little effect on migration and tax yields. There is not enough money at stake for most families paying only the Basic Rate to warrant relocating. Where we do identify significant impacts is for changes in the Additional and Higher income tax rates. The higher income groups subject to these taxes can more easily move around the country and incur greater tax bills than Basic Rate paying households. Rather few people are involved but some of these could gain or lose substantially from the emergence of tax rate differentials.

The Additional Rate only affects the richest one percent of taxpayers. We find that for this Rate (currently 45% on a taxable income over £150K), Thanks to induced migration any tax cut will always raise tax receipts and any rise will always reduce tax revenue. Over a longer period the migration and revenue effects become stronger. A reduction in the Additional Rate from 45% to 40% increases Welsh tax revenue by an annual rate of £55 million after ten years.

A Higher Rate differential between Wales and the rest of the UK has the greatest impact on Welsh tax revenue because it affects both the middle income and the high income groups; in fact the greater part of a typical high income worker’s earnings is taxed at the Higher Rate.A large cut in the Higher Rate (currently 40% on a taxable income between 43K-150K) will slightly lower tax receipts in the short run (1-3 years), but in the long run even a large cut will still raise Welsh tax receipts. The encouragement to some households to immigrate, increasing Welsh taxable income is greater than the revenue lost from the reduction in the Rate. A rise in the Higher Rate to 42.5% from 40% reduces Welsh tax revenue ten years later by about £240 million a year.

We simulate some wider impacts of these tax changes. A cut reduces government spending, and thereby affects household well-being, but is perhaps partly or more than counterbalanced by greater spending on private goods and services. In addition, the composition of the labour force changes through tax-induced migration. The net effect turns out to be that tax cuts always increase output per capita but they do not always increase total tax yield.

Considering the income tax devolution permitted by the 2014 Act, tax-induced migration has only a muted effect on devolved government revenues because such a small proportion of revenue is decentralised. The additional effects above on the tax revenue generated by the devolved economy are greater, but devolved government revenues are largely insulated from them (unlike central government). Welsh output or Gross Value Added is most sensitive to the changes in the Higher Rate of tax.

Extensive tax devolution, in contrast to the limited devolution proposed for Wales in the UK 2014 Act could trigger substantial spillover impacts of migration on tax revenue. This prospect may, and perhaps, should deter substantial decentralisation of taxation.