What should central banks do when politicians seem incapable of acting? Thus far, they have been willing to step into the breach, finding increasingly unconventional ways to try to influence the direction of troubled economies.
But how can we determine when central banks overstep their limits? When does boldness turn to foolhardiness?
Central banks can play an important role in a cyclical downturn. Interest-rate cuts can boost borrowing - and thus spending on investment and consumption.
Central banks can also play a role when financial markets freeze up. By offering to lend freely against collateral, they "liquify" assets and prevent banks from being forced to unload loans or securities at fire-sale prices. Anticipating such liquidity insurance, lenders can make illiquid long-term loans or hold other illiquid financial assets.
To the extent that unconventional monetary policy - including various forms of quantitative easing, as well as pronouncements about prolonging low interest rates - serves these roles, it might be justified. For example, the first round of so-called quantitative easing (QE1) by the Federal Reserve in the United States was doubly effective.
By buying mortgage-backed securities, the Fed brought down interest rates in that important market and restored it to vitality. Similarly, with its outright monetary transaction (OMT) programme, the European Central Bank has offered to buy peripheral euro-zone countries' sovereign bonds in the secondary market - provided that they sign up to agreed reforms.
The logic is that conditionality will ensure that countries are solvent, while OMTs will restore trust to a market that has broken down because investors fear that the countries concerned will exit the euro zone. Again, its effect, thus far, has been significant.
Other unconventional policies, however, have been undertaken to stimulate the economy, rather than to deal with broken markets.
The benefits have been commensurately smaller. QE2, in which the Fed bought long-term government bonds, did not have a discernible effect on long-term government interest rates. Indeed, with its recent decision to pursue QE3, the Fed is focusing once again on the mortgage-backed securities market; but, given that the market is healthier now, it is unclear how much good this will do.
The Fed expressed its intent to keep policy rates low until employment picks up strongly. The hope is that if investors consider this announcement credible, long-term interest rates will come down further, encouraging spending.
But the immediate effect on long-term bond rates has not been encouraging. As central banks venture farther into uncharted territory, advocates argue that at worst they will do no harm. In fact, no one knows. For example, sustained low interest rates hurt savers who traditionally prefer safe short-term investments. Pensioners and those near retirement, facing low income from interest, may cut back further on consumption, weakening the economy. Bolder pensioners, desperate to generate higher returns, may take undue risks - for example, investing in junk bonds - that could jeopardise their nest eggs. And such financial risk-taking may have little impact in terms of spurring corporations to assume more risk by investing.
Similarly, a potential downside to quantitative easing is that low interest rates send capital to higher growth, high interest-rate countries. In theory, as capital floods in, these countries' exchange rates will appreciate rapidly, making them look unattractive and automatically stemming the flow.
In practice though, as investors make money on their trades, they bring in yet more money, forcing further currency appreciation. All too often, the process does not end smoothly but in a crash. No wonder recipient countries resist inflows of hot capital.
We also know little about how smooth the exit from quantitative easing will be. In theory, as the economy picks up and interest rates begin to climb, central banks will simply pay higher interest rates on their reserves, so that they can finance their holdings of long-term securities and shrink them slowly. But higher interest rates also imply large capital losses for central banks' asset holdings.
Even if some of these losses are offset for the government, the losses on long-term private debt holdings are real. Moreover, the argument that losses are offset is not easy to explain to the public. Will opinion be sympathetic to the Fed when politicians blame it for losing tens of billions of dollars monthly on its asset holdings? Will bond markets fall sharply (and interest rates rise) as markets fear that the Fed will be pushed to sell its enormous holdings in short order?
A last defence offered by advocates of continuing on the path of adventurous monetary policy is that, because politicians refuse to settle their differences and act, monetary policy is "the only game in town".
By creating the impression that something beneficial is being done, unconventional monetary policy relieves pressure on politicians. So, when central bankers argue that they are the only game in town, they are ensuring that outcome.
Central bankers now enjoy the popularity of rock stars and their response to the difficult and uncertain environment during and after the financial crisis has been largely impeccable. But they must be able to admit when they are out of bullets. After all, the transformation from hero to zero can be swift.
Raghuram Rajan, a former chief economist of the IMF, is the professor of finance at the University of Chicago Booth School of Business and the author of Fault Lines: How Hidden Fractures Still Threaten the World Economy
* Project Syndicate