There's an important fact that's lost in the debate about economic policy. It's that counter-cyclical policy is nothing like sufficient.

The back of a beermat tells us this. Let's say we wanted to get unemployment (on the LFS measure) down from its current 2.5m to one million. How many jobs would we need to create?

The answer is NOT 1.5m. This is because many jobs would be filled by those out of the labour force - the retired, home-makers, students on marginal courses and, OK, immigrants. For example, in the last year employment (pdf) has risen by 584,000 but unemployment has dropped by just 156,000. This tells us that reducing unemployment by 1.5m would require well in excess of 3m new jobs. That's an increase of over 10%.

This requires a massive expansion in GDP. Once we allow for the possibility that productivity will begin to rise again, we probably need a rise in GDP of over 15%. And we'd need a bigger rise in money GDP to achieve this, simply because a large part of any fiscal or monetary expansion would raise prices rather than real GDP: the Bank of England estimates (pdf) that its first ''200bn of QE raised real GDP by 1.5-2 per cent and inflation by up to 1.5 per cent.

Getting unemployment down to one million would therefore require a rise in money GDP of perhaps 30 per cent - over ''400bn. Nobody is proposing a fiscal or monetary expansion anything like this*.

Instead, the belief seems to be that a small policy stimulus will kick-start the economy, get the normal capitalist engine of growth purring again.

But this is questionable, for at least two reasons.

- There's a dearth of investment opportunities. There are many reasons for this: a (maybe temporary) slowdown (pdf) in technical progress; the migration of low-wage industry to the east; an inability to monetize what new technologies there are; or perhaps a wising up to the fact that innovation never really paid off anyway.

- Inequality has become a barrier to growth. This could be because it reduces the marginal propensity to consume. More likely, I suspect, it's because it has added to debt and because the same institutions that create inequality - managerialism - also give us bad decision-making and rent-seeking.

Policy-making, then, requires more than (very mild) stimulus and the hope of return to business as normal. We must think about ways of increasing trend growth.

It should be obvious to anyone not blinded by ideology that the right's ideas here - of reducing the power of the working class - no longer work. Instead, Duncan Weldon, Stewart Wood and Ha-Joon Chang are right; the left must think about supply-side socialism.

Now, you might object that I've argued that history shows that long-run growth doesn't much change. This fact, though, might not be so hostile to supply-side socialism. For one thing, it shows that "moderate" policies don't work, so perhaps we need something radical. And for another thing, it shows not just that few policies raise growth, but also that few depress them. And this suggests that perhaps radical leftist policies are a "free hit": they might not be disastrous for the economy. As for what such policies might be, I'll suggest some later.