> MMT is politically useless. That said, I do think it's intellectually
> useful. First because it throws a revealing light on certain macroeconomic
> truths. And second, precisely because it begs so many political questions
> that only class analysis can really answer.
I agree, and I think Steve raised an excellent question. This debate goes all the way back to Abba Lerner's 'functional finance' argument in the 1940s, and further, and of course it follows from Keynes. The state regularly 'prints money', it's just not obvious because of the institutional division between central bank and treasury.
Policy strategy-wise, it's generally been accepted mainstream strategy since at least the 1950s that the treasury should issue securities to finance a deficit, and the central bank determines what happens monetarily on the basis of desired macroeconomic/monetary conditions. So long as the central bank targets an interest rate it loses control over the quantity of money in the system, and the proportions of government liabilities held in the form of money and securities respectively is an aggregate outcome of private decisions. The only thing that's really new with 'quantitative easing' (and it's not unprecedented in any case) is that the Fed is operating directly at the longer-term, riskier end of the market.
The basic idea of 'functional finance', i.e. that the scarce resource is labour and not money, has never been refuted, but it has been qualified by the role of private expectations - e.g. that if you start doing something that offends common sense views about government finance (even if it is more common than sense) the markets may react badly to it with unforeseen consequences. Since the 1970s rational expectations theory has made a fetish of this kind of argument, but there is surely some truth to it. Policymakers are not nearly as omniscient as you need to be to judge exactly what mix of financial assets will be needed at what time. The government deficit/debt fetish may be entirely economically bogus, but it does provide a consensual norm that organises expectations. It's not the best norm, obviously, but it seems there was more rationality to go around on this issue in the 1950s and 1960s.
I haven't followed QE 1 and 2 all that closely, but I wouldn't say it's been a failure. It's at least helped to make banks liquid enough to lend should they want to and should solvent borrowers want to borrow. You can't expect too much more than that from monetary policy - as JK Galbraith said, it's like a string - pulling works well (i.e. restraint) but pushing not so much.
I had a long debate with 'modern monetary theorist' Bill Mitchell a couple of years ago - starting here: http://scandalum.wordpress.com/2009/05/30/mitchellnomics/
Mike Beggs