Tax Systems: The good, the bad and the completely toot toot ding-dong loopy

Worstall @ the Weekend A loyal reader of these 'ere pages, Rich Bryant, writes in with an interesting question. There's a proposal for an entirely new taxation system out there, The Reset, and is it something that mainstream economists are ignoring because the paradigms of mainstream economics just cannot deal with it - or is it because the plan is absurd bullshit?

Narrative tension demands that the answer not be supplied for a few paragraphs. But while I would never claim that economics is entirely settled, nor really even mature as a science, there have been some thousands of pretty clever people who have chewed over certain questions within the subject for a couple of centuries or so by now. And thus we do have some answers to some things. And just as last weekend when we looked at health care systems, with a taxation system we want to look at efficacy (does it collect revenue?), efficiency (how much of everything else does it screw up?) and equity (are we nicking the rich bastards for their share?) Also as before there's a certain tension between these three meaning that we can't ever really have a taxation system that pleases everyone. But we can, and will, take a walk through the implications of the three desires.

As to the Reset, the astute mathematicians among you will see one basic problem from this little quote:

Our present taxation system is nuts! It raises under £500bn per annum from all taxes in the UK. TEAL (Total Economic Activity Levy) would raise over £1.5 trillion (over 3 times the current amount).

OK, you might need a slight acquaintance with basic economic numeracy as well to grasp it. The point being that UK GDP, the value of everything produced in the nation in a year, or equally and by definition, the incomes of everyone in the country in a year, is around and about £1.5 trillion. Meaning that this proposed new system of taxation is being praised as a method of government being able to hoover up everything from the whole lot of us, rich and poor alike. This might not be the greatest recommendation for a taxation system ever. You know, even the Soviets left the peons with a bit of pocket money.

Narrative tension now being satisfied, the plan is - of course - absurd bullshit. But not just because of this failure of basic economic numeracy: because of what the suggestion is itself. The designer of the Reset note that most economic activity flows through the banks. So, the thinking goes, just lightly tax each and every transaction through the banks and we'll raise all those hundreds of billions easily.

You might note the way in which this is just a larger version of the Robin Hood or Financial Transactions Tax, which fails for exactly the same reasons. The first and most obvious being that if the only taxation in the country is on bank transfers then everyone will stop making bank transfers. Cash becomes king, or perhaps we all just start keeping ledgers and settle up net balances once a year rather than shifting the gross amounts around in real time. People's behaviour really does respond to incentives and it would do here.

Such a tax therefore fails our first test, that of efficacy. It just doesn't raise the revenue. It also fails our efficiency test, for transaction taxes do fail that efficiency test. And finally it fails our equity one because we'd have absolutely no idea at all who was really carrying the economic burden of whatever revenue was being raised.

The two important concepts we need here are tax incidence and the deadweight costs of taxation. Incidence is, well, who really pays the tax? We don't think that the tobacconist is paying the duty on our cigarettes, nor the tobacco company that buys the stamps. Instead it's the consumer who does. Same with the barman and beer duty, same with our paycheques and the income tax that is taken out of them.

The person who hands over the money to the taxman is not, necessarily, the person carrying that economic burden. This becomes important when we discuss corporation tax: for the burden of a tax must be carried by some live human being, something that companies aren't (no, Americans, a legal personality with free speech rights is still not a natural person or a live human being). Thus the company simply does not pay the tax. The standard analysis is that the people who actually pay corporation tax are some combination of the shareholders in the company (who get lower profits) and the workers in the entire economy where the tax is levied (in the form of lower wages because lower profits leads to less investment, investment plus labour pushes up productivity, rising productivity raises wages).

As a matter of equity we do want the rich bastards to be paying more as a proportion of income. Even Adam Smith was sound here. So we've got to analyse who is paying, or carrying that burden, of a tax before we can pronounce on how equitable it is. And this is a serious problem with transaction taxes. In the words of Sir John Mirrlees (Nobel for the study of tax systems) transaction taxes “cascade through the economy” and we're really not sure where they do end up. One study of Stamp Duty on share purchases ends up stating that pensioners get lower pensions as a result, workers get lower wages. That's really not what we might expect and it's certainly not what we desire.

Then there's those deadweight costs. The very act of raising revenue through a tax means that some economic activity will not take place. That is of course a loss to us as economic activity is what makes us richer. No, this doesn't mean that we shouldn't ever tax anything: some of the things (we'll have the fight about how much of what government currently does meets this test another day) done with the revenue raised are definitely worth more than the activity destroyed by raising the revenue. But we would like to raise our revenue in a manner that destroys the least activity simply by being raised. And we've a useful rule of thumb to aid us here. Transaction taxes (Reset, the FTT etc) have the highest deadweight costs of all. Then come capital and corporation taxes, then income taxes, lower than all of those we find the consumption taxes (VAT etc) and so low that sometimes it's actually positive is repeated taxation of real property, or as it can be known, land value taxation.

So, in terms of efficiency, it's not that difficult to design a taxation system. Start with land taxation, add on any Pigou Taxes we want (carbon, baccy, booze, pollution etc, we want these precisely because they do destroy certain economic activity), as much VAT as we can get away with and if we still need more to feed the ravening maw of government, income taxes and on up the list. And the Mirrlees Review did lay it out pretty much in this manner. The hope being that we'd have enough revenue before we got to taxing corporations and/or incomes from capital. This is also the basis of the progressive consumption tax, the net effect of which is that all savings and investments are in something like a giant ISA (IRA for Yanks). You pay no tax when you put money in, nothing on what is earned and reinvested, but pay normal rates when you take any out to consume.

That taxation system does not of course pass our equity tests (despite the progressive consumption tax doing damn well here) simply because large numbers of people just wouldn't think it was equitable. And given that the definition of equity is what people think is equitable, fail it most certainly does. Which is, of course, where our tension between these three different desires for a taxation system comes in. Also why we'll never actually have a good taxation system and will always end up with, at best, one that's merely OK. Simply because large numbers of people have very different ideas about what should be that balance between equity and efficiency.

For example, Thomas Piketty is advocating tax rates of 80 per cent on incomes at the top end, and of 1-3 per cent on wealth annually, knowing very well that these entirely fail the efficiency and efficacy tests. But he thinks that equity is just so important that he doesn't care that they will make everyone poorer. Obviously he's wrong, for he is a Frenchman, but he's not the only one who thinks that way.

My own best guess about a tax system that's going to be about as OK as we can get it would be that land value tax, the Pigou stuff, something like that giant ISA (IRA) up to a heftily upper middle class level of say £1 million gross and then, on equity grounds, regular income tax rates on capital gains and investment income above that and no corporation tax. Not sure if even that last is politically possible; but that's about as far towards efficiency as I can see the system moving. And the most important part of it would be the LVT at the bottom of the system.

Agreed, others will have very different views of the right balance of our three desires and will thus end up advocating different systems. But even then it would still be true that a transaction tax system, one that aims to collect the whole of our annual output and or incomes into the tender hands of George Osborne, would not be an OK nor even a bad taxation system.

It would be a completely loopy one. ®