Logical Errors and Empirical Errors

Yes, we\’re back with Richard Murphy again. Here is (part of) his response to my latest critique.

1. I start my analysis using company accounts;
2. I recognise that there are flaws in company reporting which limit their usefulness for this purpose, and
3. I recognise that there are differences between accounting and taxable profits in the work that I do, but only a limited number of those differences can be identified and so adjusted for in the analysis I undertake.

As a result he declares that my work cannot form the basis for any credible policy recommendations.

Of course, it\’s easy enough to refute allegations that aren\’t made. I\’m not alleging (solely) that the methods used to calculate the "tax gap" or "the missing billions" are flawed emprically, I\’m saying that they are flawed in basic logic.

Here is the logic that Murphy is using.

There is a headline rate of tax. Companies seem not to be paying that headline rate of tax. The difference between the two is the gap.

Now so far I\’m perfectly happy with this analysis. It\’s an empirical question and I\’m happy enough (with reservations) about the methods he uses to try and calculate that gap. Adding back in only the amortisation of goodwill, well, OK, I agree that it\’s a tough accounting exercise that he\’s attempting.

However, where the logic goes wrong is in his next stage. That there\’s something wrong with there being this gap, that the existence of such a gap is evidence that the will of Parilament is being frustrated, even flouted. That is, that he\’s made a logical error, not simply an empirical one.

Now let us agree that people do indeed order their affairs so as to lower their tax due. Leave aside all those legal rights about their being allowed to do that. Even then there\’s still a logical error here.

For Parliament deliberately and specifically enacts into law provisions which are there to reduce the tax paid by corporations below the headline rate.

First year allowances, Capital allowances, pension contribution allowances, intangible assets allowances, research and development credits and so on. I wouldn\’t be at all surprised to find that there are others, perhaps training tax credits, child care even….

It doesn\’t matter how many of these there are either. The existence of any of them means that Murphy\’s basic approach is logically invalid.

That a company pays less than the headline tax rate might be (I agree it doesn\’t have to be and would be terribly surprised if all of that gap were as a result of it) a result of it doing exactly and precisely the things which Parliament has specifically encouraged it to do by offering these tax breaks is all we need to see that logical error. You can\’t point simply to lower than the headline rate payments and thus conclude that companies are tax dodging spivs: nor that they are cheating others nor that they are frustrating the will of Parliament.

The calculation of that tax gap is thus logically and fatally flawed.

Moving on to his second point:

So let me compare my approach with that Mike Devereux uses in his paper on tax incidence, which arguments Tim Worstall believes absolutely correct. Mike Devereux:

1. Starts his analysis using company accounts;
2. Does not explicitly recognise that there are flaws in company reporting which limit their usefulness for this purpose, and
3. Does not seek to adjust reported company data to allow for those necessary adjustments, such as excluding non tax allowable goodwill charges and adding back unpaid deferred taxation that will undoubtedly produce an estimate of the effective tax rate closer to that actually suffered than that declared on the face of the profit and loss account, which is he figure he uses.

I\’m entirely willing to believe that there are empirical errors in the paper by Mike Deveraux: but not that the idea of tax incidence is flawed in logic.

We can start very simply: the basic idea is that whoever is handing over the cash is not necessarily the person who is bearing the economic burden of the tax itself. That\’s our logical start.

This is clear enough with income tax PAYE payments. It\’s the company that hands over the cheque but we don\’t say that it is the company bearing this cost. We say that it is the individuals whose pay stubs it has come from who do.

So far so trivial. Move along a bit to employers\’ national insurance. Again, it\’s the company handing over the dosh but is it the company that really bears the economic burden? We might say that this part of NI is simply regarded by the company as part of the cost of employing labour and that thus the actual burden is carried by labour in the form of lower wages. We might say that it really is the company paying it. There\’s been quite a lot of work on this and the answer is that it depends….on the elasticity of supply of labour and the similar elasticity of demand for it. Out in the real world the usual answer given is that most of the burden is carried by labour.

Move along now to corporation tax. The logic of the tax incidence argument is impeccable. It isn\’t the company bearing the cost of that tax, it\’s some combination of the stakeholders in it.

The literature on the incidence of taxes on corporate income dates back to Arnold C. Harberger (1962), who developed a model of a closed economy with a corporate sector and a non-corporate sector, and analysed the introduction of a tax in the corporate sector only. Harberger showed that the incidence of the tax depended on a number of factors, including the elasticities of substitution between labour and capital used in each sector, and between the goods produced in each sector. His main conclusion was that under reasonable assumptions, the tax is borne by all owners of capital, across both sectors, as it drives down post-tax return to capital. Similar results have been generated by a number of more complex CGE models with a larger number of sectors (see, for example, John B. Shoven 1976). However, these results depend crucially on, among other things, the assumption of a closed economy, which ties down the supply of capital to the economy. If capital is perfectly mobile between countries, but labour is not, then the results can be very different. David F. Bradford (1978) and Laurence J. Kotlikoff and Lawrence H. Summers (1987) show that the introduction of a tax on corporate income in a home country tends to reduce the world rate of return to capital, and tends to shift capital from the home country to the rest of the world. This shift in capital reduces the return to labour in the home country, and increases the return to labour abroad. As the home country becomes small relative to the rest of the world, the effect on the world rate of return diminishes towards zero: however, there remains an exodus of capital, and the home country labour force effectively bears the entire burden of the tax.

Quite what that distribution is though is an empirical question.

I\’m perfectly happy with the idea that this paper quoted doesn\’t give us the correct empirical answer: it\’s one of many that cluster around the same numbers and proportions which makes me think it might be correct but so what?

The logic of the argument still stands: it isn\’t the company that pays the tax, it\’s some combination of the stakeholders. And just because you want to tax capital returns that doesn\’t mean that capital returns are what bear the economic burden of that tax.

Now, ask yourself. If you were going to suggest who was seeking greater objectivity who would you choose? The person who explicitly accepts the limitations in his data, and seeks to adjust for it, or the person who just ignores the issue? And which set of data is likely to produce the more objective result? I’m biased, of course, but I know my answer.

What’s yours Tim Worstall, and why?

And what would you do better?

Before I started to try and influence public policy I\’d start with an argument that was logically supportable. Only then would I move to trying empirical proof of my contentions.

Murphy\’s calculations of the tax gap are fatally flawed in logic, because he specifically ignores the way in which Parliament enacts laws to make the rate of tax paid lower than the headline rate of tax. Thus the gap between the two rates, headline and paid, cannot be used to insist that people are flouting either the law or the will of Parliament.

That\’s a much more important error than whether the complexities of corporate accounting make our estimates of the burden of corporation tax empirically innaccurate when we know that the approach is firstly true in logic.