Five Lessons from 1937 and Otherwhen
J. Bradford DeLong
U.C. Berkeley and NBER
delong@econ.berkeley.edu
June 17, 2009
Comment on Christina Romer (2009), "The Lessons of 1937":
Let me make five points to eliminate or refute or at least to fight against or lay down a marker that there is--well, call it "confusion" about what the right state of American macroeconomy should be.
Last December’s Unemployment-Rate Forecast and Outcome to Date
Source: Romer and Bernstein (2009).
My first point is that over the past six months the economy has been a severe disappointment. Output and employment have fallen much faster than people were projecting last December. Romer and Bernstein (2009) projected at the very start of this year that unemployment in the U.S. would reach a peak of 7.9% in the summer of 2009. But unemployment now in mid-June is about 9.7%, with 10% baked in the cake and the possibility existing that it might go much higher. The signs that the cliff-dive of employment has come to an end are very few. The level of new unemployment claims is still consistent with a rapidly-collapsing labor market nationwide.
New Weekly Unemployment Claims (Red, Right Scale, Four-Week Average) and Monthly Fall in Payroll Employment (Blue, Left Scale, Thousands)
Source: Paul Krugman.
Six months ago a net federal fiscal stimulus of about $1 trillion--$400 billion each year for about 2.5 years--seemed appropriate: that seemed to balance the benefits of filling-in the hole in aggregate demand without running too great a risk of triggering worrisome inflationary fever further down the road. Now the hole in aggregate demand is greater than was thought likely last December--about twice as great--and the likelihood of heightened future inflation is less. Thus if it was appropriate to set a $1 trillion federal fiscal stimulus in motion last December given what we knew then, if we had known then what we know now it would have been appropriate to set a roughly $2.4 trillion fiscal stimulus--$800 billion for 3 years--in motion back then.
My first point is thus that the Obama administration's federal fiscal stimulus programs are on the low side of what is appropriate by a substantial margin: this is the largest economic downturn since the Great Depression and the standard tools of expansionary monetary policy are tapped out and broken right now.
My second--related--point is that the need for federal-level fiscal expansion is reinforced by what state governments are doing right now. The federal government's discretionary actions are expanding aggregate demand by about $400 billion over fiscal year 2010, but state governments are right now cutting their spending and raising their taxes in order to offset this federal fiscal expansion more or less completely. On net, the government sector will be on autopilot as far as discretionary policy moves to stimulate the economy are concerned: federal-level expansion is offset and neutralized by state-level fiscal contraction. This is not an appropriate macroeconomic policy stance: this is the largest economic downturn since the Great Depression.
My third--unrelated--point is that the policy innovations of the past year have created a potentially dangerous weakness in the Federal Reserve system. The Federal Reserve's balance sheet has more than doubled over the past year, as it has acquired an enormous and bizarre menagerie of assets. On the liability side, it has funded this acquisition by expanding the monetary base, and has increased private-sector willingness to hold this monetary base by paying interest on reserves. This has added a fourth motive--profit--to the three traditional motives for holding reserve deposits at the Fed: the transactions demand, the emergency liquidity demand, and the speculative demand.
As long as the dollar remains the safest currency in the world, as long as the dollar remains the linchpin of the global financial system, there is no problem in the Federal Reserve's funding by what is essentially overnight borrowing the expansion of its balance sheet and the purchase of private securities that will vary up or down in market price with an eye toward holding them to maturity.
However, at some future time the dollar will cease to be the linchpin of the world financial system, in which case the Federal Reserve's financing its balance sheet via overnight borrowing will leave it vulnerable to the mother of all bank runs. It would be very good to fix this now: to give the Federal Reserve now the option to borrow not in what are essentially demand but rather in time deposits--to grant the Federal Reserve the power to issue its own bonds. This diminishes the chance of a great financial crisis in 2050 or so, with no downside that I can see.
My fourth point is the obvious one that health care is the only thing tht matters for the long run budget. The other points that the Hon. Dr Christina Romer raises, are--as is almost always the case--accurate and important. America's long-run fiscal problems are caused by health care, and will not be appreciably made worse by this half-decade's federal fiscal stimulus. If restructuring the health care system can bend the curve on the rise in overall (and hence public as well as private) health care costs, then America has ample debt capacity to borrow whatever we wish in this crisis--and to borrow it at extraordinarily favorable rates as well. If the curve of rising health-care costs is not bent, then the government's long-term finances are in trouble and so is the growth of private-sector non-health living standards: health care costs that rise as fast as CBO is projecting in the baseline cause lots of long-run economic problems, of which government fiscal bankruptcy is not the worst. Health care reform to bend the long-run curve of costs is now just what it was back in 1993: the most important issue for the American political system to deal with.
Fifth, I have the sense that the Obama administration's economic policymakers have forgotten one of the most basic lessons taught by Robert Rubin during his stewardship of economic policy during the 1990s. The lesson is to think probabilistically: to project yourself forward into the possible futures, to ask in each one what would be the actions that you would then wish you hd undertaken today, and then to actually take the appropriate action today. Looking forward into the future, (a) I see a 10% chance that something happens to create renewed cliff diving--a recession that bottoms out not with an unemployment rate in the 10-12% range that we currently anticipate but an unemployment rate that blows through 12% and keeps on rising. (b) I see a 30% chance of a rapid recovery as confidence and asset prices recover, and firms take advantage of high unemployment to hire new workers in droves at wage levels that make increasing production very profitable. But (c) I see a 60% chance of the end of the current cliff-dive in employment being followed by what happened in Japan in the 1990s, in the U.S. after 1991, in the U.S. after 2001, and to some extent in the U.S. after 1933--a recovery that does not see the market exert sufficient upward pressure on employment to return the unemployment rate to normal levels in two or three years, but that instead sees a jobless or low-job recovery during which the unemployment rate continues to drift upward for years, or falls only then to rise again.
The Obama administration's policies appear to me to be the ones that would be adopted if we believed that there was a 75% chance of scenario (b) and a 25% chance of scenario (a). But I don't think those are the probabilities. And I wonder what the Hon. Dr. Christina Romer thinks the probabilities are. For she is the one who warns of how:
[t]he 1937 episode provides a cautionary tale. The urge to declare victory and get back to normal policy following an economic crisis is strong. That urge needs to be resisted until the economy is again approaching full employment. Financial crises, in particular, tend to leave scars that make financial institutions, households and firms behave differently [than in normal times]. If the government withdraws support too early, a return to economic decline or even panic could follow...
The blunt fact is that the economic recoveries that have been rapid and seen fast growth in employment are those that ended when a Federal Reserve following strongly restrictionary policies to fight inflation eased off and significantly lowered interest rates. No such lowering of interest rates is possible this time--interest rates are already as low as they can possibly go at the short end. So I can see no reason to anticipate a rapid recovery and employment when the cliff-diving stops. And I do not understand why the Obama administration is following policies that presume such a rapid recovery--a V rather than an L for the shape of the recession--is not just possible but probable.
1429 words
Posted via email from http://braddelong.posterous.com/comment-for-the-economist-on-christina-romer at Brad DeLong's Scrapbook
