Blanchard roundtable: Pursue contingent policies
This discussion can be followed in its entirety here.
UNCERTAINTY is a constant in economic life, but Olivier Blanchard notes that at present it is sufficiently pervasive as to be a major exogenous source of restraint on demand. He recommends policies that reduce uncertainty, as distinct from policies that simply boost aggregate demand.
Many of his policies, however, would take effect regardless of the state of the economy. It seems to me, by contrast, that the way to deal with uncertainty is through contingent policies that are triggered when a predetermined bad state of the economy is reached. Such policies will be all the more powerful if they are widely known in advance. They would thus reassure households, investors, and businesses that tail risks are less likely to be realised, who should become more willing to spend or invest, further reducing those tail risks.
Many traditional economic policies are contingent. Deposit insurance is triggered when a bank fails, which means banks are less likely to fail in the first place. Unemployment insurance benefits are paid out when unemployment rises, which gives households less incentive to build in precautionary saving during recessions, aggravating the downturn. Even monetary policy that follows a Taylor rule is contingent—the more unemployment rises above its natural rate and the more inflation falls below target, the lower real interest rates will go. This may not be prescribed well enough to affect behaviour, but it has gotten us out of most past recessions. Of course, monetary policy in America (and perhaps, before long, in other countries) has reached the zero bound and can no longer fill this contingent role.
Can other contingent policies be created or enhanced to eliminate tail risks? Tyler Cowen suggests, sensibly, strengthening the automatic stabilisers. That is what the Obama fiscal package does by expanding unemployment insurance benefits and extending health care subsidies to the newly unemployed. (By contrast, provisions such as infrastructure spending and tax cuts are not contingent; they will be implemented regardless of the state of the economy). The automatic stabilizers could be put on steroids by, for example, cutting taxes if the unemployment rate rises to predetermined levels. Yes, this would aggravate the deficit, but given that monetary policy is less potent, fiscal policy should assume more of its pre-emptive and contingent character (Of course, fiscal policy is costlier and more difficult to reverse, and Alberto Alesina notes this is a constraint on its open-ended use.).
Arguably uncertainty is most damaging in the financial sector, as it has caused precautionary hoarding of capital and liquidity which is fueling the pullback in aggregate demand. Nowhere are contingent policies more necessary.
As lenders of last resort, central banks are contingent suppliers of liquidity. When private funding dries up, they will lend without limit to any solvent bank willing to pay a penalty rate. The Fed built on this principle with its new liquidity facilities. Term auction credit is available at a slight penalty rate to the overnight index swap rate (OIS), and commercial paper can be issued to the Fed for 100 to 300 basis points over OIS. The benefits of these programmes go well beyond the actual credit extended; their mere existence reduces funding uncertainty for banks and corporations and makes them less likely to hoard cash and cut investment. When Blanchard says the fiscal authority should sell Treasury bills and recycle the proceeds into risky assets, this is precisely what the Fed’s liquidity facilities have done. But because they are contingent, they will wind down of their own accord. The penalty ensures that once private sources of funding become easer to access, official facilities will become less attractive (as is now happening), and this limits the associated risk of inflation.
Policy makers have attempted to implement contingent financial stability policies more generally with much more mixed results. Treasury offered contingent capital to Fannie Mae and Freddie Mac (which would be supplied as their own capital eroded) but then allowed Lehman Brothers to fail and Washington Mutual’s bond holders to be wiped out. The G7 declaration in October that no systemically important institutions would be allowed to fail removed tail risks, but those benefits have been partly undone as policy makers have added uncertainty through the risk of nationalisation or dilution of common shareholders.
A better contingent financial stability policy would offer a relatively clear template to supply additional capital as banks’ capital is depleted, and insurance against further losses on assets, which is what Ricardo Caballero is driving at. Nationalisation without just compensation would be ruled out, but not “creeping nationalisation” via repeated injections so long as the state acquires its shares on the same terms as the private sector. There are of course countless obstacles to overcome for such a policy to work, and political interference will remain a permanent source of uncertainty.
The financial crisis is grave, but there are, believe it or not, worse cases of Knightian uncertainty in our recent past. In the aftermath of 9/11, no one knew if there would be another terrorist attack or how serious it would be. America was either lucky, smart, or both in that no such attack ever came. We still cannot place odds on terrorists using a weapon of mass destruction or on climate change triggering catastrophic natural disasters. By comparison, we have a better chance of dealing with Knightian uncertainty in the economic world.