So how should we tax these BASTARD COMPANIES, then?

Still scratching around for a better tax system

At this point we can declare victory: we've worked out how to balance equity and efficiency to some extent. And we can present this as being highly equitable. Capitalists (those shareholders, rentiers and so on) will pay the same rate of tax upon their capital income as everyone else has to on their labour income. And won't all sorts of people be happy with that?

Except, we're only going to apply that equal rate of tax to the excess profits, not to the normal ones, thus boosting efficiency. This might even work until everyone spots what is happening. I've seen a doughty tax justice warrior approving this scheme even though I think he did because he didn't understand it. But then, my opinion of this particular warrior is that he doesn't understand shoelaces yet either.

Which leaves us with one interesting problem: how do we define what the average rate of return to capital is? We could, of course, look at how much capital there is in the economy, then look at profits, then divide one into the other and there we are. But it's a sad truth about these sorts of macroeconomic numbers that they're really not very accurate.

More than that, to get them to any level of accuracy takes time: our GDP estimates for the year (where such things as profit share reside) go through several revisions and we're not usually happy with them until about six months after the year end. And we can't go around setting the tax rate on profits six months after the year in which the profits have been made.

That's quite aside from the problems of what we do with depreciation and so on (as any decent accountant will tell you, the correct answer to the question “What were my profits?” is “What do you want them to be?”).

At which point, that Mirrlees review becomes ever so slightly whiffy. Yes, I know, brave thing for a hack journo to say to the Laureate and friends. But they say that the average profit rate is about the same as the risk free rate of return. Which is the return on medium term government bonds, for the UK say 10 year gilts. Yes, this is low at present but in more normal times (ie, without quantitative easing or a recession) we'd think that 3.5 per cent to 5 per cent is the usual sort of range for this number.

So, people making profits on their capital higher than this pay the normal income tax rates on it. The problem here is that the risk free rate isn't really the same as the average return, or the “not excessive” return that we're looking for. Government bonds are free of credit risk, that's true (except when you get to crises of Greek or Argentine proportions) but they're not free of inflation risk. The two concepts are really slightly different.

However, that yield on those 10 year bonds is something that we can measure in real time (and HMRC could issue a standard number for the month, quarter or year, as it does sometimes for things like exchange rates to be used in accounts if you want to) and thus can actually be used as a determinant of what the tax rates, or rather tax allowance, should be.

We're never actually going to end up with a perfect tax system simply because of the tension between those two things, efficiency and perceived equity. But this might well be the best we can do with capital and corporate taxation. Normal rates of profit go with the efficiency argument, super-profits, the ones we sorta assume from having carved out some rent or protected position in the economy, pay normal income tax rates.

Who knows, we might even end up with a sensible tax system someday. ®