Whenever he was faced with a blustering student - which was often - my old economics tutor, the late Andrew Glyn, would ask: "What's the mechanism?" This is the question we should ask about the TPA's The Single Income Tax report. It says:

There is a long history of macroeconomic studies which have found that economies where public spending is a lower share of national income tend to see stronger economic growth.

Pages 130-42 of its report (pdf) provide empirical evidence.

It's in this context that mechanisms matter. There are (at least) four important, and often overlooked, features of mechanisms.

1. They vary across times and places. The TPA has in mind the idea that lower taxes stimulate entrepreneurship and investment. That's perfectly plausible. But isn't it possible that this mechanism would be weaker than usual now, given that the dearth of monetizable investment opportunities and the lack of supply of credit? The fact that lower public spending has often raised GDP growth does not suffice to show that it will do so here and now.

2. The precise type of spending matters. If by "public spending" you mean spending on airport "security", I'm with the TPA. But if you mean, say, instead spending on early years childcare, I'm not: this can raise growth both by freeing up women to enter the labour market and by raising human capital in the longer term.

3. Institutional quality matters. Public spending which is funnelled into the pockets of bureaucrats would depress growth. But spending which has positive outcomes, such as raising educational attainment (harder than it seems (pdf)) or cutting transport congestion wouldn't.

4. There's a distinction between the long-term and short-term.In the long-term one way in which spending can affect growth is by affecting social norms.But this effect can be amibguous.On the one hand, small government might boost entrepreneurial spirits and self-reliance because youngsters believe there are few state jobs available to them. But on the other hand, the same distrust of the state which breeds small government might be an obstacle to growth-enhancing public spending; look at the state of US roads or public education.

Thinking in terms of mechanisms, then, helps us reconcile (sort of) the TPA's evidence with Lane Kenworthy's observation that the Nordic nations have combined high spending with good economic performance; they - perhaps unusually, perhaps not - have had the right sort of spending; points 2 and 3 have worked in their favour.

I'd just make some other observations.

First, it's perfectly coherent to believe that the relationship between public spending and growth varies from nation to nation. The fact that the Nordic countries have achieved good growth with high public spending does not create a case for other nations to raise public spending, if bad institutions cause that spending to go on boondoogles and bureaucrats rather than upon measures to raise outcomes.

Secondly, your attitudes to 1-4 are, I suspect, hugely conditioned by your priors. If you have faith in entrepreneurial spirits and scepticism about governmental quality, you'll side with the TPA on points 1-4. If you have faith in government but not in entrepreneurial spirits you won't. Cross-country studies don't much help.

Thirdly, this is no place for "bigthink." What really matters are the microeconomic details such as the precise types of spending and the institutions through which it is channelled.