I decided to have a bit of a look at the methodology behind the Fair Tax Mark today. I’ve only spent an hour looking at it but I have to say that it really isn’t very representative of… well… anything useful.
This will undoubtedly turn into a bit of a rant and I must stress I have not read most of the company analyses. But I found that am strangely familiar with most of the errors in the methodology…
Broadly, the methodology picks scores out of five in three categories to give an overall score out of fifteen. The three categories are:
- country-by-country rating score
- tax rate score
- tax haven score
For something that describes itself as being about tax, devoting a third of the score weighting to a transparency measure seems a bit dubious. I am certain that many companies who pay a very fair rate of tax don’t provide country-by-country reporting.
Also, the tax haven score appears to be based on the existence of subsidiaries and does not appear to attempt to establish what they do and whether they actually reduce the business’s tax rate.
But it is essentially the tax rate score which I wanted to look at and that is mainly what I am discussing below.
I am very wary of Richard Murphy’s previous form on tax avoidance measures. I decided last year that his tax gap calculations were unreliable on account of the fact that he failed to use a mathematically sound average in his calculations (ie, he got the maths wrong): Tax gap corporation tax rate analysis
My first thought was that he had probably made a similar mistake with this, so I decided to have a quick check.
Sure enough, his use of weighting in calculating an average is not weighted according to magnitude of profits. Which is pretty necessary when calculating an average rate of anything against profits.
Instead he weights by according to how recent the results, with recent years being weighted more heavily. There isn’t anything wrong with this in itself, but the fatal flaw is that there is still no weighting for the magnitude of profits.
A company with lower profits will have a current tax charge that varies from the “expected rate” by more than one with higher profits.
Take the following example. A company has adjustments for tax purposes that result in an additional deduction of £1m in two consecutive years. These are due to R&D tax relief (ie not avoidance).
In year one, the profit before tax is £1.5m, and the second it is £150m.
With a tax rate of 30% (dealing with historical rates obviously), the first period shows an effective tax rate of 10%, whilst in the second period it is 29.8%.
But in both years, the tax adjustments are exactly the same. So, some sort of weighting is required to remove distortions for size of profit when comparing or combining these two rates before you provide for other weighting.
This doesn’t happen in the Fair Tax methodology. Instead the two rates are essentially averaged with equal weighting with regards to profit. Despite the fact that profits in year one are only 1% of year two.
The method averages the difference between the expected rate and the current rate and the only weighting in the final average is based on the year. The distortions for magnitude of profits are allowed to stand.
This is highlighted by the fact that the methodology has to eliminate certain data because it would dramatically distort results. See Home Retail Group plc.
The fact that results need to be eliminated is a sign of the fundamental flaw in the methodology.
They have had to pick out data in order to make the method produce something that looks reliable. So what does that tell you? It tells you the method is unreliable.
Pretending that this is some sort of safeguard is nonsense. Basically, the maths is wrong and no safeguard can correct this fundamental flaw.
Now, deferred tax is difficult to understand and I can understand the desire to not include it if it is not necessary. But it is necessary if you are proclaiming to be fair.
The decision to exclude deferred tax is highly dubious because the purpose of deferred tax is to provide for tax that isn’t liable in the accounting period in question.
Anybody deciding to eliminate this element is willingly ignoring the effect of timing differences. So if you claim capital allowances in excess of depreciation, you get a deferred tax charge, because in theory at some point in time this is going to reverse and depreciation will exceed capital allowances and there will be an increase in tax.
However, this doesn’t mean that deferred tax will even out over six years.
It isn’t restricted to capital allowances. There will be deferred tax adjustments for anything where tax legislation provides for a different treatment compared to the accounting treatment. This includes tax losses. Where losses are carried forward, reducing the tax charge, the deferred tax charge shows that an asset created in a previous year is being utilised.
So the effect of trading losses are completely ignored too: a loss in one year may result in a low current tax charge in the previous year or future years, and no account is taken of this.
But, perhaps more fundamentally, no account is taken of any tax legislation at all.
Now, anybody who talks about tax avoidance should understand this:
Tax is calculated on the accounting profits as adjusted by tax legislation.
The Tax Mark’s method returns to Richard Murphy’s misguided “expectation gap” where the effects of tax legislation are completely ignored.
Murphy takes the profit from the accounts and remarkably ignores the fact that this profit is then adjusted by whatever the tax legislation decides is necessary to remove the fact that accounting profit can be manipulated in so many ways.
So he completely ignores the effect of thousands of provisions of tax legislation. Some would be matched by a deferred tax movement, some would not. This would include permanent differences for disallowed expenses and statutory reliefs which reduce the corporation tax rate (like R&D allowances, creative sector reliefs,
So what does the Fair Tax Mark say about tax avoidance? Absolutely nothing.
It does not itself consider the effects of any tax legislation, let alone the intention of Parliament in enacting it. It simply cannot say anything about something which it deliberately ignores.
But what does it say about tax? Nothing that is of any use. It summarises some numbers from the accounts, but they’re practically meaningless on their own.
As far as I can tell, the Fair Tax Mark makes no attempt to understand the information that is already available in the accounts on tax. Normally you can identify the reasons for major differences in the tax rate by actually reading the tax reconciliation as opposed to deliberately ignoring it.
Rather than making me want more tax disclosure, it makes me feel like we should have less. It appears that tax campaigners are unable to understand the tax disclosure notes already published. I have very little confidence in them understanding even more data.
Which is, I suppose, the point of boiling all this information down to a single number. It is so that people don’t need to understand all the data, they can just understand a single, simple number.
But the problem here is that this particular number is nonsense.
The maths is wrong.
The disregard of deferred tax is unfair.
The disregard of the entire tax code is simply absurd.
Fair Tax Mark? Pass me the Trade Descriptions Act please.