The battle to tax multinational corporations does, in my opinion, revolve around just four variables. They are the tax base, i.e. what should be taxed. Then there's the tax rate. After that the issue is place; i.e. where should profit be taxed. Finally there's the question of time, which is when profits should be taxed.
Trust me, it's taken years to summarise all I've written about on this issue to just these four words:
It's taken longer still to realise why we have so many problems isn making a tax charge on a multinational corporation stick. Fundamentally that problem is that any state can at best, if it is in open competition with other states, have effective control of only two of these variables at any time.
So, a tax haven can fix its base (nothing) in its place (at least as far as the recording process goes) by abandoning rate and time.
The US can argue for rate (high) and place (US) only by abandoning base (to loopholes) and time (to offshore deferral).
The UK tried to do base (worldwide) and rate (28% to 2010) but lost place (when the controlled foreign company rules failed) and time (by effective deferral that those rules allowed).
You may begin to get my drift, but there is a key assumption in here and that is that states compete. What if they didn't? What if they said that if a tax burden is to be imposed on capital - as has to be the case in the pursuit of equity and balanced budgets then they should stop pretending that they should comply with the micro economic theory of the firm and should instead behave as what they are: governments who have a duty not to fail and an obligation to their own taxpayers and a resulting commitment to each other government to ensure that tax is paid wherever it is due.
Then the story changes. Now we still have four variables but we have to decide which are to be controlled cooperatively and on which variance may be allowed.
We can't solve four variable equations. It is not possible. But we can solve two variable ones. So suppose cooperation controls two variables only leaving the others open to local control. Doesn't that make sense?
If so then what variables do we need to control mutually? To me there seem to be two obvious ones: they are place and time.
Let me deal with the last first: deferral makes no tax sense to any state. It never has and never will. Profits earned now need to be taxed now. Current taxation has to be the norm. In that case the deferral of tax that comes from relocation has to be eliminated. This means that place has to be the connected, but different, variable that has also to be controlled internationally to ensure states can have sovereignty over base and rate.
This means firstly that the allocation of profit to a place has to be reasonable. Unitary taxation provides a mechanism for doing that. This deals with the problem of place. Secondly it has to be complete: country-by-country reporting deals with that by ensuring all profit is captured with the potential for re-allocation wherever it has been attributed to by a multinational corporation. This by default deals with timing: all current profit is apportioned to the place where it is really most likely to have been earned.
But this does not undermine sovereignty. Tax rate still very clearly then remains a variable under local control. And, in arguments I will be making over time, I think this is not the end of the story on base either. It is profit that is allocated. Whether to tax all, part of none of it should then be a decision for the state to make.
There is not time to expound on this last idea this morning: I am at a seminar today and this blog is a mere rehearsal for what I might (I stress, might) be saying. That issue on base is unlikely to come up in detail. It is still a research work in progress.
But if we remember there are four variables in tax and we can only control two individually we have a basis for international cooperation that is more realistic than demanding control of all four, or different ones on different occasions.
We have a long way to go in beating international companies on tax. Getting the intellectual framework right would be a useful start though.