Matt Mitchell (via) writes:

The types of austerity that are most-likely to a) cut the debt and b) not kill the economy are those that are heavily weighted toward spending reductions and not tax increases. I am aware of not one study that found the opposite. In fact, we know more. The most successful reforms are those that go after the most politically sensitive items: government employment and entitlement programs.

This raises the question my old economics tutor Andrew Glyn always asked: "what's the mechanism?"

I suspect many right-leaning economists have in mind some idea that lower government spending reduces allocational inefficiencies and so boosts growth (pdf).

But there's another possibility.It starts from Silvia Ardagna's finding that, when austerity raised growth in the 70s and 80s, it did so in large part because of cuts in the government's wage bill (pdf).

There could be a simple reason for this. In the 70s and 80s, profits in many countries were squeezed by wage militancy. Cuts in public sector employment helped reduce this militancy and so alleviated the profit squeeze. And higher profits, allied to the improved business confidence created by reduced wage militancy, stimulated business investment. As Alesina and Ardagna wrote (pdf) in 1998:

Both the profit and the wage shares and the ratio of the manufacturing goods exports deflator over unit labour costs in manufacturing indicate an increase in profits during successful [fiscal] adjustments, which, instead, does not occur in other episodes.

Which brings me to my concern. Whilst this mechanism operated in many places in the 70s and 80s, it's not clear that it would operate now. Of all the things that are constraining investment, high real wages and low profit margins are surely not high among them. In this sense, the historical lessons of the 70s and 80s might not be relevant to our plight now.

I don't say all this merely to argue against austerity. I do so to remind ourselves of two other things.

One is that economics is not (just) about fancy models. It's about mechanisms. The question is: through what mechanisms can tighter fiscal policy raise growth? Mechanisms that are powerful in one time and place need not be powerful in other times and places - which is why economics has such a poor forecasting record.

Secondly, class conflict (sometimes?) matters in macroeconomics.