The Tax Justice Network has just told us, once again, that countries lose gazillions in tax revenue to tax dodging. However, their claim is not in fact true. Specifically, this is not true:

We focus now on the long-term estimates of the revenue (in per cent of GDP) lost by each country i in period t as a consequence of profit shifting through tax havens in 2013, i.e. 𝐿𝐿𝑖𝑡.

They are not measuring long term revenue anything. For they are not measuring the effects of the corporation tax rates they look at.

Imagine, for example, that a country had a corporate profit tax of 100%. There would be little to no corporate activity leading to profit to be taxed. The country would be dirt poor in fact. Thus there would be little revenue.

Equally, a tax rate of zero would encourage activity like billy oh. And revenue would be gained from other taxes instead, like the consumption of the workers now gaining higher wages, even perhaps their income tax.

This is, of course, the Laffer Curve argument and just like that argument it is something which is true at some level of taxation. Lower rates will increase revenue, higher rates reduce them.

We're not insisting that current rates are either above or below that revenue maximising level, not right here and right now we're not. But we do insist that failing to even consider the point means that no estimation of long term revenue losses has in fact been made.

We'd also point out that going over that revenue raising rate is easier than many think. The combination of three things made it probable that the UK corporation tax system in the UK in the 1970s managed this. High inflation, a corporate tax system which did not rebase costs on that inflation and relatively high tax rates meant that corporate capital wasn't even earning enough to cover depreciation in some of those years.

It really is necessary to consider the wider effects, not just do sums as the TJN has done. But then, you know. This is the same researcher who insisted that Zambia was losing squillions in revenues by comparing the price of tens of thousands of tonnes of copper in that country with the price of 10 kg samples of copper in Switzerland. Experimental design might not be the strongest suit here.