A very strange, and very wrong, idea has been circulating through many otherwise respectable corners of the economics blogosphere: Monetary policymakers should ignore what goes on in the financial system. This was the standard view before the crisis. We shouldn't take it seriously today.

Contrary to popular belief, central banks don't actually print any money. Rather, monetary policy affects the incentives of financial firms to create money and credit. Inflation and unemployment are only tangentially affected -- even though those are the things central bankers are supposed to care about most.

This creates a bit of a paradox. Monetary policymakers, earnestly attempting to fulfill their legal obligations, sometimes end up encouraging bankers and "shadow" bankers to take on too much risk. For example, low yields encourage financiers to increase their leverage in order to achieve their return-on-equity targets. That in turn increases the likelihood and magnitude of devastating crises. Similarly, overreactions to changes in the consumer price index can throttle the financial system. Myopically focusing on the rate of inflation or the level of employment can therefore make it harder for central banks to achieve their objectives over the medium term.

Those who don't find this intuitively obvious should begin with this piece, which summarizes some of the most interesting research on the relationship between monetary policy and the behaviors of financial firms. The academic debate became more relevant once Jeremy Stein, a Federal Reserve governor, gave a speech on the subject in February. I summarized his arguments and their intellectual context here. Two further posts filled in some additional details.

Within the Fed, Stein's intellectual nemesis is probably Nayarana Kocherlakota, the president of the Minneapolis Federal Reserve Bank. He recently gave a speech in which he admitted that the Fed was deliberately engaging in a strategy that would lead to extremely undesirable outcomes in the medium-term for the sake of hitting its short-term inflation and unemployment objectives. Josh Brown appropriately described this as "pump and dump." I critiqued Kocherlakota's logic and noted that this was exactly what the Fed did in the 2000s.

Mike Konczal thinks the financial sector's rational responses to different monetary policy outcomes are morally unappealing. I sympathize with his point of view, although it isn't clear how he would get around this problem. You can give banks a safe way to make money, you can drive them out of business, or you can encourage them to behave in a way that will lead to crises. Fans of the Canadian banking system routinely point out that it is a cushy oligopoly that allows for easy profits without excessive risk-taking.

I certainly agree with Konczal that higher capital requirements and indexed-debt contracts would be helpful. It's naive, however, to conclude that these reforms would be sufficient to alter the basic incentives of profit-seeking financial firms.

Policymakers have to interact with the world as it is, not with the world they wish existed. If monetary policy were implemented by depositing newly printed cash directly into people's checking accounts, rather than by adjusting the level of interest rates paid by financial intermediaries, I would agree that central bankers could afford to focus exclusively on the price level or the level of nominal gross domestic product. That's not the world we live in, however.

The confusion might come from Paul Krugman's 15-year-old Slate column about the Capitol Hill babysitting co-op. Krugman's point was that heightened demand for liquid savings can depress spending and potentially make everyone worse off. Unfortunately, this insight doesn't have a lot of implications for monetary policymakers. Anyone who has read the original account of the Capitol Hill babysitting co-op knows that their problems with money-hoarding were caused by budget surpluses and solved with tax cuts. That's why I have advocated more robust fiscal stimulus even as I've been skeptical of the merits of further monetary accommodation.

Having said all this, those of us who think that monetary policy should be used to prevent excessive risk-taking don't necessarily think that central banks should start tightening up right now. Pawel Morski points us to the ongoing shrinkage of the shadow banking sector. That makes it hard to argue that excessive risk-taking is today's biggest problem. Similarly, the ratio of nonfinancial debt to national income has been flat for years, mostly because private borrowers have been repaying and defaulting on their loans about as quickly as the federal government has been issuing new bonds.

It's hard to argue that the Fed needs to tighten in these circumstances. The Fed should be ready, however, to step on the brakes if circumstances change.

(Matthew C. Klein is a contributor to the Ticker. Follow him on Twitter.)