This has led to the important caveats for the Left being removed, and makes it worth reiterating that the point is NOT to grant carte blanche for government of any size. In the next few years of high fiscal strain, we may get ‘smaller’ government on the plain G/Y metric*. But this might be accompanied by all sorts of ‘free’ interventions inspired by hindsight trading of the Crisis. Many of these will damage our ability to grow. So allowing sheer fiscal stress to work this ratio downwards will not mean the problem of the state damaging the economy will have gone away.
A very pertinent example (though unrelated to the crisis) is how public sector wage rates are decided in this country. Many of you may be aware of the problem when a single high value sector crowds out valuable resources, causes prices to rise, and makes development of other sectors difficult. This can be associated with a natural resources discovery (see Dutch disease). But it can also happen in well-intentioned development: an NGO lands in a poor country, hires the brightest staff, raises prices, and perpetuates underdevelopment if it is not propertly handled. The same thing can potentially happen when the public sector employs too many people in a poor part of the country, on wages that are the same as paid in Surrey. It undermines a competitive advantage of that part of the country, by raising wage costs too high.
In my approach to government-in-the-economy, I want to have it both ways, and believe I should be allowed to. I want the government only involved in economic transactions where it makes sense: public goods, improving market function (ie attacking monopoly), dangerous externalities, and so on. I also favour some redistribution, to reflect fairly conventional notions of justice .
But I am also strongly opposed to the government shrinking in a recession, for straightforward Keynesian reasons. I think these two positions are perfectly compatible, and it is a misreading of the significance of the G/Y ratio at such junctures that leads to a lot of confusion and misguided polemic. Let us keep a stern eye on governmental interference in things better done privately. But obsessing on a single metric is not the way to do it.
UPDATE: Chris Cook of the FT discussed related matters back in May this year. He makes this good point:
These reliefs, incidentally, are one reason why the tax-to-GDP ratio is a daft way to think about how much influence the state has over the economy. If you raise the tax ratio from 40 per cent to 45 per cent by cutting reliefs, you are reducing state control of the economy.
*G = government consumption. Y = GDP