> >And you do need an institution that acts as a firewall to prevent temporary liquidity issues from turning into full insolvencies via self-fulfilling prophecy

> >And you do need an institution that acts as a firewall to prevent temporary liquidity issues from turning into full insolvencies via self-fulfilling prophecy (aka bank runs). > Economists take this as gospel, but I have never seen evidence of this. I have seen evidence it is not true. I'm with Friedman, fractional reserve banking doesn't create value, just inflation and risk. Well, the point was about whether a central bank acting as lender of last resort makes sense or not. Fractional reserve banking existed before central banks, and if you have fractional reserve banking, then a lender of last resort is a must. There's plenty of evidence for that (in the form of regular bank runs before central banks, and practically zero bank runs afterwards). As to whether fractional reserve banking itself should be permitted or not, that's admittedly less clear. Though I'd say that there's a big advantage to having the kind of flexibility in the money supply that comes with letting credit shrink and grow in a decentralized fashion as demanded by the economy. I find it ironic that Friedman, who is associated with some of the worst and most ideological central planning-bashing, turns out to be an advocate of central planning when it comes to monetary policy! Seems inconsistent if you ask me... > You said there is no one who thinks the Fed is trying a game of pulling on some levers and creating more value in the economy, but then you went on to describe the Keynesian argument for how the Fed pulls some levers to magically create value (by stimulating AD). What we disagree about here is mostly whether there's magic involved. My point is that a large part of the underlying theory is sound (though not all of it), and your claim of voodoo is greatly exaggerated. The question of whether the work being done is useful is an important one. That's why I've always been against (or at least skeptical of) QE, actually, though I think that fiscal policies that often come up in the same discussions could be quite useful. For example, you could just transfer a fixed amount of money every citizen ("helicopter money") to boost the economy, and scale it down if/when inflation exceeds your target. That way, the usefulness of the additional work would be guaranteed, by virtue of being work that the people decide should be done.