"Deep Recession Calls for Healthy Dose of Fiscal Stimulation"

A colleague, George Evans, discusses the need for aggressive policy to combat the recession:

Deep recession calls for healthy dose of fiscal stimulation, by George Evans, Commentary, Register Guard: ...The business cycle, with expansions occasionally interrupted by recessions, is an enduring feature of market economies. Asset price bubbles and crashes also appear to be intrinsic to markets. It is important to recognize that recessions are usually ... a surprise, resulting from large shocks that could not be offset in time by policy.

Economists do, however, have a set of policies to avoid or minimize the impact of recessions and to promote recovery. The standard policy tools for dealing with recessions are two-fold: 1) easing monetary policy by reducing interest rates, and 2) allowing normal fiscal “stabilizers” to work: in recession tax collections naturally decline and unemployment benefits and welfare payments naturally increase. At the federal level this can and should be financed by temporary increases in deficits. ...

When a recession is very large, as now, the usual anti-recessionary policies are insufficient, and aggressive fiscal policies are needed. When recessions are accompanied by a financial crisis, as now, special interventions are required to prevent a complete financial meltdown and ensure that credit remains available. Regulatory reforms are also needed... Experience has shown that recessions precipitated by financial crises are usually longer and deeper than typical recessions.

The reason why aggressive fiscal policies are essential ... is that the adverse shock has been so large that monetary policy is inadequate: cutting short-term interest rates to zero is not enough to ensure recovery. Without sizable fiscal stimulus there is the possibility of a destabilizing spiral of deflation and falling output.

The combination of aggressive monetary easing and sufficiently aggressive fiscal stimulus should be enough to stabilize the economy and eventually lead to recovery. This recovery will raise tax revenues and, with appropriate long-term fiscal planning, debt levels relative to GDP will gradually return to normal levels. As the recovery begins, the monetary authorities will start to unwind current policy ... to prevent inflation from becoming a problem.

It is possible that yet more fiscal stimulus will be needed... If so, additional stimulus should avoid across-the-board temporary reductions in personal income taxes, since these have small aggregate demand effects. More effective are temporary increases in government spending ... and additional funding to states and localities... Temporary investment tax credits to firms can also be effective...

We must avoid retreating to the economics of Herbert Hoover. The argument that government should behave like families in recessions, reducing spending because times are hard, is misguided, because the proximate cause of current high levels of unemployment and low levels of GDP is a collapse of the aggregate demand... This is a Keynesian moment in history, and in this situation, temporary increases in government spending lead to higher incomes and more jobs. Households may anticipate higher future taxes to pay for the deficit spending, but the net effect on GDP and incomes is positive and substantial, and this prevents a dangerous slide into deflation. If the government spending is on infrastructure, broadly construed, then this also lays the foundation for a more productive recovery.

Is the current federal stimulus program sufficient? There is no simple answer. Policy affects the economy after a delay that is long, variable and uncertain. ... The lags and uncertainties make policy difficult.

In this situation the guiding principles are as follows. [T]he biggest risk is a 1930s type depression triggered by a negative feedback loop of deflation, high “real” interest rates after correcting for deflation, and reduced private sector spending. We appear to have avoided this outcome, but ... policy needs to be continually revisited. The Federal Reserve Open Market Committee meets every six weeks to review monetary policy. U.S. fiscal policy should also be reviewed frequently. The next several months may indicate the start of recovery or they may suggest that further action is needed. Finally, we also need a long-term plan to ensure that, after the recovery is under way, total publicly held federal debt will eventually return to its normal range of, say, 30 percent to 70 percent of GDP. ...

Although reasonable people can disagree on the appropriate role and size of government spending, we can agree that spending should be as productive as possible and that tax rates should be set to finance this spending on average. However, whatever the choice of overall level, there is a strong macroeconomic case for maintaining government spending during recessions and for temporary increases in government expenditures in deep recessions.

I should make clear that George cannot be accused of applying old fashioned theory to modern problems. The commentary above is based upon his recent research -- here's a brief description of some of his recent papers:

We examine global economic dynamics under learning in a New Keynesian model in which the interest-rate rule is subject to the zero lower bound. Under normal monetary and fiscal policy, the intended steady state is locally but not globally stable. Large pessimistic shocks to expectations can lead to deflationary spirals with falling prices and falling output. To avoid this outcome we recommend augmenting normal policies with aggressive monetary and fiscal policy that guarantee a lower bound on inflation.

And here are links to the papers themselves (the first two are technical, the third was written more for general economists and central bank policymakers):

• Expectations, Deflation Traps and Macroeconomic Policy (with Seppo Honkapohja), July 6, 2009.
• Liquidity Traps, Learning and Stagnation (with Eran Guse and Seppo Honkapohja) Abstract, European Economic Review, Vol. 52, 2008, 1438 – 1463. Electronic reprint available by email on request.
• Monetary and Fiscal Policy under Learning in the Presence of a Liquidity Trap, Monetary and Economic Studies, December 2008, 59 – 86.

Thus, this gives a fairly simple policy prescription - guarantee a lower bound on inflation - that is grounded in modern New Keynesian structures augmented with the models of learning that George, Seppo, and others have been developing. In the models with learning, there are stable and unstable regions, and large shocks can push you into the unstable region. The guarantee of a lower bound for inflation prevents inflationary expectations (which are formed through learning) from entering the region where deflationary spirals with falling prices and falling output are possible.

Here's a bit more from the conclusion to the third paper above:

The recent theoretical literature on the zero lower bound to nominal interest rates has emphasized the possibility of multiple equilibria and liquidity traps when monetary policy is conducted using a global Taylor rule. Most of this literature has focused on models with perfect foresight or fully RE. We take these issues very seriously, but our findings for these models under adaptive learning are quite different and in some ways much more alarming than suggested by the RE viewpoint. We have shown that under standard monetary and fiscal policy, the steady-state equilibrium targeted by policymakers is locally stable. In normal times, these policies will appropriately stabilize inflation, consumption, and output. However, the desired steady state is not globally stable under normal policies. A sufficiently large pessimistic shock to expectations can send the economy along an unstable deflationary spiral.

To avoid the possibility of deflation and stagnation, we recommend a combination of aggressive monetary and fiscal policy triggered whenever inflation threatens to fall below an appropriate threshold. Monetary policy should immediately reduce nominal interest rates, as required, even (almost) to the zero net interest rate floor if needed, and this should be augmented by fiscal policy, if necessary, in the form of increased government purchases. Intriguingly, using an aggregate output threshold in the same way will not always successfully reverse a deflationary spiral.

When aggressive fiscal policy is necessary, this will lead to a temporary buildup of government debt. However, government spending and debt will gradually return to their steady-state values. An earlier implementation of the recommended policies will mitigate the use of government spending, and if our recommended policy is already in place at the time of the shocks, the immediate use of aggressive monetary policy can in some (but not all) cases entirely avoid the need to use fiscal policy. Raising the inflation target π* is an alternative way of reducing the likelihood of needing to employ fiscal policy, but this may be undesirable for other reasons.