Last month Pascal Saint-Amans, head of tax for the OECD, spoke to the Wall Street Journal, in an article subtitled The argument against taxing capital income relatively more than wages is losing its force. He said:
“For the past 30 years we’ve been saying don’t try to tax capital more because you’ll lose it, you’ll lose investment. Well this argument is dead, so it’s worth revisiting the whole story,” Pascal Saint-Amans, the OECD’s tax chief, said in an interview.”
This particular article came to our attention via the Fair Skat blog, which sketches out the implications – and they are highly significant. The usual story goes something like this:
“It’s important to cut taxes on capital because this promotes investment and growth. This is because higher taxes create incentives to shift profits overseas or into different forms, including secret forms, to escape the tax, whether legally or not. Lower taxes also put more hands in the hands of investors, which creates productive investment.”
That’s the conventional wisdom. Fair Skat goes into a bit of economese to look at the research evidence that’s being used to buttress these arguments:
“The ‘go to’ for empirical evidence on this is De Mooij & Ederveen’s 2008 “reader’s guide”. Their literature review finds that, in an average situation surveyed by the literature, the total semi-elasticity of the corporate tax base for these effects (i.e. the % change in the tax base from a 1% increase in the relevant tax, see below) is -3.1, with profit shifting (-1.2) the largest single effect.”
In short, the story goes, cut taxes, and more stuff will come into the tax net, counteracting the effect of the tax cut. And sometimes policy-makers cut taxes on the basis of these kinds of conclusions.
Now we’ve been challenging this conventional wisdom for a long time. As Fair Skat puts it:
“There are a myriad of potential arguments to question the certainty and applicability of these findings, but that is for another potential blog; suffice to note here that these figures represent the most accepted available evidence in economic literature on the behavioural effects of corporate taxation.”
For example, our Ten Reasons to Defend the Corporate Income Tax outlines a range of reasons why relentless corporate income tax-cutting is generally a bad idea. For example there’s ‘investment’ and there’s ‘investment’ – and the good stuff that countries need is embedded in their economies, with local supply chains and managers with their kids in local schools: this stuff won’t fly away at the first whiff of tax.
And there’s a more profound reason why policy-makers shouldn’t pay much attention to the kind of evidence we’ve highlighted above. From a policy-maker’s perspective, attracting investment should not be an end goal. It is at most an intermediate target. An end goal would be to benefit your broad population, for example by sustainably raising the number of productive jobs overall. Corporate tax cuts may hurt other sectors, leaving your own economy no better off, or attract mobile investment with few local roots or benefit.
In addition, there are serious question marks over the ‘go to’ evidence on elasticities. As Clausing (2016) makes clear, the bulk of the evidence rests on data which has systematic weaknesses. Most obviously, the most common data source is the corporate balance sheet database, Orbis, which is severely limited in its coverage both of developing countries and of the major corporate havens. It seems almost inevitable that research findings using data that largely excludes the two extremes will understate the true degree of elasticity.
Researchers at the Joint Committee on Taxation of the US Congress have shown a different flaw in the elasticity literature: namely, the assumption of (log)linearity. In effect, most approaches have proceeded on the basis that profit (shifting) will be equally responsive to tax rate changes in a high tax or a low tax jurisdiction. By varying this assumption, Dowd et al. (2016) reveal that the elasticities are much higher in low tax jurisdictions.
This indicates that the effects of tax cuts are potentially powerful in attracting new profit (shifting) where the rates are 5% or below; but that they will be significantly smaller for the leading (non-haven) economies. So e.g. competition may be fierce between Luxembourg and Ireland, for example, but much less so between France and Germany. (A policy aside: this finding also lends support to the UK government’s own estimate that its substantial rate cuts to below 20% would deliver an increase in the taxable base of precisely zero. The only apparent reason for ongoing cuts has been the spurious rhetoric of ‘competitiveness’ – even as public services and the tax authority itself face continuing financial pressures.)
Fair Skat makes a further point, one that we haven’t made too much noise about before.
In the last few years, there have been a number of new global initiatives – such as the OECD’s BEPS project to crack down on corporate tax cheating, or the OECD’s Common Reporting Standard to share banking information across borders, and others. To the extent that these schemes have teeth, corporations won’t be able to find the escape routes that they used to be able to. It’s increasingly clear that BEPS has failed to meet its own ambitions, but the political appetite to go further is clear (not least in the EU).
And the implication is this: those academic numbers, even if you were able to take them at face value, may no longer be relevant. Gone is the era in which corporate tax behaviour went largely unscrutinised – the era when the Guardian ran the first major frontpage splash on the subject, with a headline that read simply, ‘Revealed: how multinational companies avoid the taxman‘. 2007 seems a long time ago now, and the numbers don’t apply any more.
Policy makers are going to need new numbers. Fair Skat puts it like this, in a review of Peter Dietsch’s book on tax competition:
“Dietsch argues there is no reason to assume today’s elasticities for tomorrow – they can be modified through policies and thus we can change the factors in the optimal tax calculation. For instance, by introducing stronger international cooperation on capital tax evasion, it is possible to limit the tax evasion elasticity, and thus make tax systems more progressive by increasing the optimal levels of capital taxation, shifting the tax burden back on to mobile capital factors.”
And indeed, a new OECD report, entitled Tax Design for Inclusive Growth, underlines the point, in polite OECD-speak:
“The arguments in favour of reduced capital taxation are not as clear-cut as previously thought.”
And (with emphasis added):
“The empirical research on corporate tax incentives (e.g. S. Van Parys, 2012/3) shows that in some cases they may be successful in attracting more investment but fails to confirm that they are beneficial overall.”
As we said, the implications here could be profound. If BEPS is followed by serious progress, for example in requiring the publication of multinationals’ country-by-country reporting, then taxes on capital in the new world are much less likely to be undermined than before. One result of that will be that there is even less justification for policymakers to go for corporate tax cuts. And of course if countries are less likely to feel the pressure to ‘compete’, then the global co-operative measures are likely to be all the stronger for it.
There’s a virtuous circle in there. And it is a big one…