While it is often a pleasure to test one’s ideas against opponents, a fair joust requires adherence to a least one basic principle, namely that the contenders compete on a level playing field - meaning, in this case, that they will marshal their arguments on the basis of available evidence. Unfortunately, Michael Bullen’s riposte to my brief piece on the 50p marginal tax rate reads like an exercise in distortion. Even so - and for this he deserves a vote of thanks - he has prompted me to delve a little deeper into the question of marginal rates for high earners, though I think it unlikely that he will relish the result.
I begin by noting that nothing in Michael’s piece addresses the main focus of my objection to the FT letter of September 7th from twenty economics gurus, namely that there is no evidence for their suggestion that the 50p rate will inflict “lasting damage” on the UK economy. Rather than deal with this, Michael has chosen to shift the ground as follows:
...although...we cannot say much about the impact of taxation on growth, we can say a lot about its impact on government revenues.
When I last looked at an economics textbook, I seem to remember reading that economic growth has a great deal to do with government revenues.
Leaving aside that obvious point, the core of Michael’s argument appears to rest on three platforms.
The first of these concerns his view of the general trajectory of the UK economy since 1973. Having stated that exogenous factors have exercised far more influence on the economy than marginal tax rates, he then goes on to count the number of quarterly periods of growth and decline, as if counting negatives and positives over periods entirely unrelated to changes in tax rates nevertheless tells us something about their effect. One is reminded of Gordon Brown’s tiresome boast that the last government had presided over the highest number of positive quarterly results since the year dot - only for us to discover that we had been floating on an ocean of soap bubbles and are now suffering the worst recession since the 1930s.
The absurdity of counting quarterly periods of growth and decline in this way is easily illustrated. Let us suppose a five-period run of economic growth as follows: -2%, -1%, -0.5%, 5%, 8%. Those figures equate to a growth rate over the whole period of roughly 9.5% - not staggering perhaps, but showing a positive trajectory of movement from recession to something approaching a far-eastern rate of expansion. However, by simply counting the number of negative and positive periods, as Michael has done, we get three negative periods against two positive - and therefore a truly dispiriting story of failure. A favourite aphorism of an old teacher of mine was that there are “...lies, damned lies and statistics…”
Michael supports his dubious methodology not with evidence but with asides, nudges and winks:
...there are many who harbour suspicions that a more competitive tax regime contributed to the relative prosperity...
Any errors of judgement that the Chancellor makes with regard to top earners are ...likely to have non-trivial outcomes. A very modest movement abroad of some of those top earners, or others electing to spend more time on the golf course than in the board room, would likely have a significantly detrimental impact on UK PLC’s ability to repay its debts.
This last statement constitutes Michael’s’ second platform: the threat that the movers and shakers of the UK economy will desert the ship. It’s called blackmail; and no doubt the country should be quaking in its collective boots at the thought of losing those irreplaceable high-flyers who have presided so capably over the country’s deindustrialisation, the sale of its best companies, and the mother of post-war financial crises. A board-room tale of old had it that senior executives negotiated their most important business deals on the golf course; but haven’t we long suspected that their absence on the greens has had more to do with escaping the burdens of office? Contrary to Michael’s assertion, their recent performance certainly suggests that the country would gain considerably if they abandoned the board-room altogether in favour of golf or some other activity less damaging to the nation than executive decision-making.
There is a serious point here. The current 50p marginal rate begins to bite when earnings exceed £150,000. But a salary of this size is - to put it mildly - no more than loose change for the senior executives of PLCs to whom Michael refers. These heavyweights count their earnings in hundreds of thousands, or even in millions. A few years of income at these stratospheric levels leaves the recipients in the enviable position of being set up for life; their standard of living no longer bound by the financial constraints familiar to the rest of humanity. One consequence is that they may become divorced from the effects of their actions. When Fred Goodwin presided over the ruinous purchase of ABN Amro, he was effectively playing monopoly in the secure knowledge that his own financial position - even if he were fired on the spot - would guarantee him a luxurious retirement.
As a parting shot into the crumbling foundation of Michael’s second platform, I quote from Professor John Veit-Wilson’s contribution to the Routledge International Encylopaedia of Social Policy 2006 :
It is widely believed that individuals’ knowledge of their marginal tax rates affects their earning behaviour. High marginal rates are commonly believed to act as a disincentive for increased output by high earners….. Empirical evidence for such asserted behavioural effects, even amongst professionals with the best information over many years in several countries, has not yet confirmed the belief, and psychological studies of work motivation have tended to refute it.
Michael’s third platform rests on the Mirrlees Review, published by the Institute of Fiscal Studies and by Oxford University Press. The section relevant to income tax rates was written under Sir James Mirrlees’s chairmanship by Mike Brewer, Emmanuel Saez and Andrew Shephard.
I confess to having paled at the thought of treading on the territory of this distinguished assembly - that is until I ventured on the report itself which is entitled: “Means Testing and Tax Rates on Earnings” (available for separate download from the IFS website). I have scoured this document for evidence that would contradict the above quotation of Professor Veit-Wilson, but have found none. The entire edifice - or “ central estimate” - for the Report’s optimal top marginal tax rate rests on a formulaic calculation based partly on variations over time in the shares of national income taken by top earners and partly on a host of unverified “rational” assumptions about peoples’ propensity to work rather than to play golf.
Variations in the shares of top earners in national income over the last hundred years have been superbly charted by A.B. Atkinson of Nuffield College Oxford in a paper entitled Income Tax and Top Incomes over the Twentieth Century. Atkinson shows, amongst other things, the degree to which inequality has increased since the Thatcher reductions in the top marginal income tax rate. Here is his depressing conclusion:
From the UK income tax data, one can estimate the shares of top income recipients in total gross income for almost the whole of the twentieth century….. The estimates show a substantial, if intermittent, decline in UK top income shares up to the end of the1970s, followed by a dramatic reversal, with the share of the top group in 2000 being above its 1945 value. The rise in after tax inequality is even more marked.
Why depressing? Because, as readers of Wilkinson’s and Pickett’s The Spirit Level know, inequality is not only a scourge on society, it is also very expensive in purely economic terms. Regrettably, the IFS Report pays scant attention at the high-earner level to the possible costs of tax-related inequality on the economy and on national welfare. Nor does it address any possible relationship of marginal tax rates to economic growth - a subject I touch on below. The authors look no further than estimates - and very rough estimates at that - of the total tax paid by top earners in terms of expected government revenues from these individuals as if that is the only item worthy of estimation.
The behavioural assumptions in the Brewer, Saez, Shephard report are breathtakingly loose - some of them merely references to what economists think. Here is an example:
Economists think about the disincentive effects of the tax and benefit system using a labour supply model. A basic labour supply model assumes that, when deciding whether and how much to work, people trade off the financial reward to working (plus any intrinsic benefits from working) with the loss of leisure time (by "work" we mean "participate in the labour market", rather than doing unpaid work at home or elsewhere).
The idea that human beings have only one kind of rationality - the economic - has now been so widely and comprehensively debunked that it is surprising to see it rehearsed here.
In fairness, Brewer, Saez and Shephard accept that their analysis of the optimal Marginal Effective Tax Rate (METR) is “tentative”. Having tried to follow their argument, I heartily endorse that qualifier. Their own calculations (p.18 or pp.110-111 in the final Mirrlees Report) produce on the one hand an “optimal top rate of 50.4% - 64.5%”, and on the other hand an optimal rate of between 40.2% and 49.4% (note the decimal fractions - a prime example of what another old professor of mine called “spurious accuracy”). Somehow, these cautious estimates - surrounded in the Report by caveats - have emerged as definitive statements. They are not.
I turn now to the question with which I began this discussion, namely the effect - if any - of marginal tax rates on economic growth (or “lasting damage to the economy”). Michael agrees that information in the UK on personal income tax rates and economic growth is hard to come by. US data, however, are much more readily available and over a longer period. As luck would have it, Mike Kimel a US economist and statistician, has used that data to take a hard look at the tax-growth relationship in his own country using figures from 1901 to the present. In a multi-part analysis entitled The Effect of Individual Income Tax Rates on the Economy he concludes that there is no evidence whatsoever to support the view that low marginal tax rates increase economic growth and that, if anything, the data support the opposite view. In a further analysis of the available evidence, Kimel estimates that, for the US, the optimal top marginal rate should be about 65%, though he admits that the message is too inconvenient for legislators to take seriously given their fear of Fox News. I can find nothing wrong with Kimel’s methodology in reaching this conclusion; though perhaps others with more expertise can locate the flaws.
Meanwhile, however tongue-in-cheek my original suggestion that the Chancellor should consider raising the 50p rate to 60p, I am beginning to think it may be a thoroughly promising idea.