Theory unsupported by fact

Tweet

The familiar old nonsense about an increase in income tax leading to outward migration or deterring people from moving in made another of its regular appearances in the Western Mail yesterday.  (I can’t find it on Wales Online, but it’s available here).  This time, the warning comes from the Chartered Institute of Taxation (CIOT), and asserts that whilst the Welsh Government can make minor variations, any variation of more than 5-10% over the long term would probably start to have a significant effect on people’s choice of residence, so the Welsh Government should think very carefully before making such changes.

The basis for this assertion is unclear; it looks to me like one of those 85% of all statistics quoted by politicians which, according to the old joke, are simply made up at the time of being required.  We do know, of course, that some extremely wealthy types go to a lot of trouble to reside in jurisdictions where they can avoid paying large sums in tax, but this is more about greed and an unhealthy and anti-social aversion to paying any tax at all than a response to a fairly marginal variation between countries.  According to the figures here, Belgium’s top rate of tax is 60% and it kicks in at an annual income of around €49,600, whilst France’s top rate is a mere 55% and it doesn’t kick in until income reaches rather more than €562,000.  If the argument about high tax rates causing migration were true, then Belgium should be half empty by now.  Yet, the last time I was there, it didn’t look empty, and I have seen no empirical evidence of mass migration to France.  The CIOT itself refers to the example of Scotland and notes that there is “no clear evidence of significant effects on migration” but goes on to say that this is “probably because of a lack of awareness of devolved income tax rates”.  So, there’s nothing wrong with the theoretical model, and it’s not that higher income tax doesn’t cause emigration, it’s just that higher earners in Scotland are too thick to have worked out that they’re supposed to migrate to England to avoid paying higher rates of tax.

Classical economics tells us that, in a perfect market where all participants have perfect knowledge and where all but one of the variables are controlled for, then a change in that one variable should precipitate a change in behaviour by way of reaction.  It’s a good way of illustrating why theoretical economists should never be allowed anywhere near policy making.  My example above (of Belgium and France) is a very silly one, because there are a whole host of other reasons why people might choose to live in Belgium rather than France.  The CIOT article does itself suggest that tax differences alone won’t drive decisions – it suggests that people will also consider what they get for the taxes as well.  But reality, unlike the article, doesn’t stop there – people don’t make such decisions solely on economic grounds anyway.  The theoretical economic animal of economics responding solely to economic stimuli is a convenient fiction for the purposes of exploring theory in academia; but it doesn’t exist in practice, even if we allow for the relationship with the stimuli to be rather more complex.  

So, what can we actually say? 

1.    There is no hard, empirical evidence, anywhere that I am aware of, which supports the theoretical notion that differential tax rates drive migration between tax jurisdictions, other than for a tiny number of extremely high net worth individuals

2.    Actual human beings – as opposed to the mechanistic robots of economic theory – make decisions on a whole range of factors, not all of which are economic

3.    Anyone arguing from a basis of theory, unsupported by fact, that the Welsh Government should not vary its tax rates other than very marginally from those set by the UK Government because it may drive richer people out may, just possibly, be pursuing an agenda unrelated to what might be best for Wales.

Still, I expect that the argument will have another outing before long.