Do we need banks? Frances raises this question. She says:

Lack of spending in an economy (shortage of aggregate demand) is caused by distributional scarcity of money, not by lack of loans. It is not necessary to restore lending in order to encourage spending. It is necessary to replace the money that is not being created by banks.

She's right. The famous "helicopter drop" - more realistically, a money-financed fiscal expansion - might well* do more to boost aggregate demand than measures to boost credit supply such as the Funding for Lending Scheme or "targeted" LTROs.

But this poses the question. If governments can bypass banks by simply printing money, why bother to fix the banking system at all? Why not just let it be a dysfunctional casino?

The question gains force from the likelihood that, in practice, measures to improve bank lending are in effect subsidies to bankers. I suspect that the reason bank shares did so well after the ECB's announcement last week isn't so much that investors are looking forward to a brighter economic future, but simply that LTROs are yet more hand-outs.

There is, though, an answer here. It's not that helicopter money would be inflationary; in the euro area, higher inflation is something to be desired. Nor is it much of a problem that such a policy would create a merely "artificial" boom; I've always been irritated by Austrians' tendency to dismiss some economic actions as less genuine than others in violation of Hayek's justified warnings about the boundedness of individual knowledge.

Instead, the reason to fix the banks lies in allocative efficiency. One function of banks - albeit one they have always performed indifferently - isn't merely to create money but to screen investment projects, financing good ones and rejecting bad. A simple helicopter drop isn't sufficient to ensure that enough money flows to the businesses or potential businesses that need it.

You might object that it's silly to worry about allocational efficiency when the economy's depressed: as James Tobin said, "It takes a heap of Harberger triangles to fill an Okun gap."

However, the inefficiency caused by inadequate bank lending isn't merely a static one. It matters for growth too. As Jonathan Haskel and colleagues have shown (pdf), a lot of productivity growth comes from "external restructuring" - from new establishments opening and older, inefficient ones closing. A lack of bank lending prevents this restructuring and so slows down productivity growth. I don't think it's an accident that UK productivity growth has slumped since the financial crisis.

In this sense, Frances isn't entirely correct. We do need to restore lending. How much we need to do so depends upon one's view of secular stagnation; if there's no demand for loans because there are no profitable investment opportunities, the need isn't as urgent as it would be if firms were credit-constrained**.

What would be nice, though, is if a way could be found to fix banks that wasn't simply a back-door subsidy. But it would be idealistic to hope for this.

* Subject to the caveat that some debt-constrained recipients of the money might use their windfall to pay down their debts.

** CBI surveys show that only a tiny fraction of manufacturers aren't investing because of a lack of external finance. I'm not sure how convincing this is. The CBI surveys only existing firms rather than those that don't exist but might if their would-be founders could get credit. And firms might be constrained from investing by the fear that credit might dry up in future even if it's available now.