Following last week’s blog on top-end income tax rates by Prof. Matthew Watson, today’s blog highlights a detailed new report (summarised here) which is launched today, on the same subject. It looks closely at the evidence behind claims about the revenue effects of cutting or hiking top income tax rates in the UK.
The report was launched (ahead of the May 7th election) by the Tax Justice Network, with the help of today’s Fools’ Gold blogger, and it shows how the official evidence on these revenue yields is subject to such huge, irreducible uncertainties that they are essentially meaningless. Perhaps more damningly these estimates are, as the report’s author John Thompson notes:
Where the parties stand on top income tax rates
The Labour government increased the top income tax rate for anyone earning above £150,000 per year from 40 to 50 percent from April 2010, the first increase since 1974. The coalition government cut the rate to 45 percent in April 2013.
Current manifesto pledges are:
- Conservative: haven’t ruled out a cut in the top rate to below 45p in the £
- Labour: will restore the top rate to 50p
- Liberal Democrats: no specific pledges on top rate.
- Green Party: a top rate of 60p in the £
- SNP: has said it supports a top rate of 50p.
- UKIP: a 40p top rate.
- Plaid Cymru: a 50p top rate
“so selective as to be unreliable and, if relied upon, worthless or worse.”
Much hot air spoken about the UK’s top income tax rates and their effect on the UK’s so-called ‘competitiveness’. In September 2011, for example, 20 well known UK economists wrote to the Financial Times to argue that the then 50p top income tax rate was:
“making [the UK] less competitive internationally, and making us less attractive as a destination for both foreign investment and talented workers”.
It is common to see such statements, despite their being founded on clear economic fallacies. But it is odd that such esteemed economists would use the c-word in this context.
What exactly, these economists might like to explain, does ‘less competitive’ mean in this context? As has already been noted on this site, a tax is not a cost to an economy, but a transfer within it. To put it crudely, a cut to the top income tax rate is a transfer from wealthy folk to roads, schools and hospitals. Whether this inequality-enhancing transfer enhances anything one might call ‘competitiveness’ is not something that serious people should state so baldly. There are winners and losers.
Sure, a tax cut may or may not attract a smattering of wealthy people to the UK – but even if it were to do this, that is never the end of the story: there is a revenue cost.
Now here’s where Thompson’s new report comes in.
Official estimates have been made of just what that revenue cost is. The benchmark everyone uses is a report by HM Revenue and Customs from 2012, which estimated that the rate cut from 50 to 45 percent in April 2013 would only cost £100 million in revenue terms.
Thompson looked at the basis for that claim, and he found two things. Neither is particularly surprising, but both are striking and important. First, the £100m numbers are hardly worth paying attention to:
“Even if one were to take the underlying HMRC numbers from its own report as correct, a more informative conclusion would have been to say that yields from having raised the top rate from 40 to 50 percent were somewhere between more than a £4 billion revenue gain and more than a £2 billion loss. But the underlying HMRC numbers are also subject to further uncertainties.”
It would have been more honest to day ‘we don’t have a good idea.’
Second, surprise surprise, the evidence marshalled by HMRC seems to have been selectively obtained. As Thompson puts it:
“It seems that the prevailing orthodoxy that higher taxes reduce growth is so ingrained that contrary evidence cannot be recognised.”
One of the most glaring examples concerns a report by the French economist Thomas Piketty (whom we contacted, and declared himself ‘surprised’ by the way HMRC had used his data.) The HMRC study said:
“A regression result in a recent study implies that higher taxes may reduce real GDP per capita levels, although the results are not conclusive.”
The footnote on that statement cites a 2011 study by Piketty, Saez, and Stantcheva. But that report actually states:
“The regressions consistently display negative coefficients across the full period, suggesting that low top tax rates are detrimental to growth. The estimates however are not fully robust to the choice of time period . . . we can conservatively conclude that low top tax rates do not have any detectable positive impact on GDP per capita.”
So Thompson sent that HMRC conclusion to Piketty, who replied via email:
“it is indeed quite surprising to learn that our paper with Saez and Stantcheva was used in this manner, given that we basically find the opposite.”
And that is not the only example of selective use of data.
So here is another way that debates have been skewed to favour those pushing the ‘tax competitiveness’ agenda. If someone is using the c-word in the context of tax, then the least they could do is frame it in terms of costs and benefits. A tax cut loses revenues, but there may be gross benefits for the economy elsewhere, against which to weigh those losses.
The official data, it seems, has been overstating the gross benefits and claiming a degree of certainty where there is none – and this has been used not just as a tool for lobbying for more tax cuts on the wealthiest, but also in the direction of being able to argue that there is something ‘competitive’ about doing so.