Saturday, 25 April 2009

Designing Taxation

Some years ago I got involved in the very practical exercise of designing a real tax system. It was in a transition country, in Europe, and it was messy.

I learned a number of lessons, some of which might be distantly relevant for the current fashion of re-designing tax systems. So, here they are.

Lesson number one. There is no possibility of an ‘engineering’ debate when it comes to taxes. Anyone who has seen the emergence of a real tax reform can attest: it is always political. And deeply so. For political scientists this is a triviality. However, for economists, far from it. For macro people are acutely aware that - apart from the politically sensitive reallocation of resources - the choice of tax system always has major consequences for the efficiency of the economy. Politicians are usually pre-occupied with the slicing of the pie, and tend to ignore the fact that the way it is sliced is a key determinant of the size of the pie. Economists tend to ignore, in turn, that it is ultimately the politicians who do the slicing.

Thus the first lesson I learned back in those days was that whatever the arguments were (Vickrey-Mirrlees's optimal taxation, Laffer-curved based, progressive or flat) - in real life they are always employed to support a political position. You need to know them all well, otherwise your opponents will slaughter you, but they are still just covers in an essentially political debate.

The upshot is that almost all tax systems have efficiency flaws. So, if you happen to find yourself on the losing side of the ‘debate’, at least you can always attack the winners with valid arguments.

Lesson number two. Even if you wanted to see the project as an ‘engineering’ problem removed from politics, there is no clear evidence about what the ‘best’ solution is. The two tails are covered: we know what is definitely not best. No taxation, no government services: clearly only for the nutcase anarchists. And total taxation leads to the disappearance of economic activity whatsoever, as proven by the total disaster that real world communism was. However, as for the refinements that represent something in the middle, the jury is out. In fact, probably they have gone home and will never come back. There are simply too many countries with too many peculiarities to control for to give a general answer.

Again, the upshot is that you will be hard pressed to find an ‘ultimate piece of evidence’ either for or against any in-between, sensible-sounding solution. So, any real tax system can always be attacked and defended using a carefully selected set of proxy cases from around the world.

However, that means the game is on. If you are a non-political, macroeconomy efficiency maximising social engineer, then you do get to design your optimal tax system. You can play lego with it. You import elements from different countries, from different historical periods, and try to see if they’d work. It can be a surprising amount of fun.

Lesson number three. The hitch is that when you present your freshly finished, super-innovative tax regime you will certainly hit a wall. Big time. First, the revenue people from the Ministry of Finance or Treasury will scream about the added uncertainty any change to the tax system necessarily brings. You may try to look at their models, and argue that they have as little real idea about the revenues as you. But still they have been doing the budget for years (probably decades), and there is no amount of economics argument that you can possibly come up with to counterbalance the fright of the politicians when they are told by the Revenues Experts that inevitable crisis, and the End Of The World is upon us. Unless a gag is put on you, the Risky Economist, your feet are immersed into concrete, and you are promptly tossed into the river.

This wall of macro-accounting conservatism is the main reason why major tax reforms are always the product of crises. Nobody takes the risk in normal times. In meltdown, you might have a chance to push your ideas through. You will still be seen as nuts, anyway.

Lesson number four. Politics is always political. This is sometimes forgotten by economic advisors, but is definitely the case.

The only time when I had a real chance to push through what I thought to be the best tax solution, it would have been a political coup. The freshly elected minority party I was harassing with my ideas needed to cheat on the freshly elected majority party via a clandestine co-operation with the freshly ousted, and much despised, former majority party. I did have an emotional stake in the previous campaign, so this was not exactly my idea of a first best solution. But it would have allowed an important piece of policy to go through.

Somehow, the party leader was convinced that such a trick on the majority party could have been fun - I suspect that he actually cared some of the time - and so he tested the idea in a press leak. He received floods of phone calls from old ladies, who did not know the first thing about the tax system, or care at that, but hated the idea that we would co-operate with the other side. The tax coup did not happen. The country lost all its previous competitive advantage, slid into government overspending, and is now on the brink of bankruptcy… (yes, it is all my fault).

Tax is always political. Even if it was not, there is no obvious ‘best’ solution. Even if there was, there is a murderous bureaucratic conservatism against anything new. And even if a crisis helped you struggle through that, the constituents the politicians play to have no idea about taxes, but very strong preferences nonetheless. Good luck.


  1. Good summary. So it is all your fault, Tamás :-). I will tell the undersecratary who implied this week it was all mine :-)

  2. Perhaps it's just me, but I do not get which tax regime you support. Care to elaborate?

  3. Well, this is a bit of an odd question, for there is no general answer, but there are two ways in which there is no general answer.

    First, every country is different, so clearly, there will not be one uniform tax system that will suit everyone. And this is true even in the longer run. There is no reason for us to think that countries will end up even after full integration into the global economy in similar ways. If nothing else, sector specialisation will probably require -- or at least suggest -- different tax structures.

    However, there is another reason as well. Existing tax systems tend not to be either fair or efficient, almost irrespective where you look in the world. As if most countries still lived in times of monarchy, when the interest of the sovereign tended to be revenue maximisation, with a not particularly strong emphasis on macroeconomic efficiency. As a result, tax systems around the world are in a kind of scream-constraint rather than an efficiency drive. As if the tax collector wanted to minimise the total political pain -- the volume of scream. In Eastern Europe, there is a particular manifestation this principle, a kind of squeeze-it-grab-it system. Tax levels are (or at least were) so high that if your company lived off the domestic market you can possibly survive unless you played around with your taxes. Then the local elite, or one of the elites, takes control of the government, siphons off the revenues, while using the tax authority to silence the opposition.

    Most tax systems of mature economies as well as emerging markets are following the same principle, too, but just the underlying societal scream-structure is different.

    So if you were welfare maximising designer of economies, your artisan tax system would probably be as much influenced by a desire to avoid state capture of the tax revenues, as by a macroeconomic efficiency drive.