Forecasting the Obama Economy

David Leonhardt writes:

Mistake by Obama’s Advisers in Predicting Job Losses: In the weeks just before President Obama took office, his economic advisers made a mistake. They got a little carried away with hope. To make the case for a big stimulus package, they released their economic forecast for the next few years. Without the stimulus, they saw the unemployment rate — then 7.2 percent — rising above 8 percent in 2009 and peaking at 9 percent next year. With the stimulus, the advisers said, unemployment would probably peak at 8 percent late this year. We now know that this forecast was terribly optimistic.... [T]he difference between the situation that the Obama advisers predicted and the one that has come to pass is about 2.5 million jobs. It’s as if every worker in the city of Los Angeles received an unexpected layoff notice.

There are two possible explanations that the administration was so wrong.... The first... is that the economy has deteriorated because the stimulus package failed.... The... answer is that the economy has deteriorated in spite of the stimulus. In other words, the patient is not as sick as he would have been without the medicine he received. But he is a lot sicker than doctors realized when they prescribed it.

To me, the evidence is fairly compelling that the second answer is the right one. The stimulus package does seem to have helped. But its impact has been minor — so far — compared with the harshness of the Great Recession. Unfortunately, the administration’s rose-colored forecast has muddied this picture.... Worst of all, the economy really may need more help.... There is no ironclad way to judge the stimulus, because we can’t rerun the last six months in an alternate universe. But you can get a pretty good sense by looking at the size of the gap between where the economy is today and where the administration thought it would be: those 2.5 million jobs that would still exist if the forecast had been right.

This gap is just far too large to be explained by the stimulus. The plan that Mr. Obama signed definitely has its flaws. It spends money more slowly than is ideal and spends some of it on projects of little long-term value. But no stimulus package could have come close to preventing 2.5 million job losses over six months.... When private economists began analyzing various stimulus proposals in January, they said that none would have a major effect on the jobless rate until the end of the year. By June, the effect would be only a few tenths of a percentage point, which translates into several hundred thousand jobs.

The stimulus that passed may in fact be having an impact of roughly this scale.... “Early results,” says Mark Zandi, [former McCain advisor and] chief economist of Moody’s Economy.com, “suggest the stimulus is performing close to expectations.” Obviously, though, the economy is not performing close to expectations...

As I understand matters, last December the median private-sector forecast had the unemployment rate topping out at 9% in the second half of 2009. The incoming Obama administration simply adopted that forecast. At the time I thought that was a mistake: (I thought that was a mistake: I thought they should have made a bifurcated forecast with a "good case" 80th-percentile scenario and a "bad case" 20th-percentile scenario; they should then have stressed that in the bad case we would need a large stimulus indeed to prevent high unemployment, and that in the good case we could restrain inflation via monetary policy.) By combining standard estimates of the effects of the fiscal boost package with that forecast, Jared Bernstein, Christie Romer, Steve Braun, Alan Krueger, and company came up with the "unemployment tops out at 8%" projected path for the economy that they put to bed a couple of days after New Year's and released at 6 AM EST Saturday, January 10, 2009.

By the end of January, when the Obama fiscal boost bill started working its way through congress, the situation had deteriorated significantly: things got worse in December as Christmas sales fell significantly below even recent expectations, et cetera. And here the Obama administration made its second mistake. The executive branch of the U.S. government is geared to do budget updates in two cycles six months apart: a January cycle and a July cycle. By the end of January the work on budget forecasting for the January cycle was well-advanced using the forecast as the transition had made it in mid-December. To change the forecast in early or mid-February to reflect the way the situation had changed over year-end would have required that OMB tear up a lot of estimation work and do it over at a time when the people in the NEOB are already working twelve-hour days to cope with all the extra work associated with a change of administration. I thought that they should have (a) changed the forecast and (b) simply not done the extra budget work: that they should simply have put an asterisk on every page of the budget saying that these numbers use the outdated mid-December rather than the current forecast. (In fact, I think the administration Troika should decouple its forecasts from the budget process, update them monthly, release them, and yet still feed them into the budget process only twice a year.) But, I think largely for these bureaucratic process reasons, they did not update their forecast. So by the time the stimulus bill was passed everyone's expectations were already that the 8% unemployment peak was way overoptimistic...

How about it guys? Can we get Steve Braun and company on a monthly forecast public update cycle decoupled from the budget process? It would improve the quality of information.

---

Among other things, it would make it extremely difficult for things to happen like what happened to the Mankiw CEA over the winter of 2003-2004, when high politics appears to have reached down into the forecast, changed the table for payroll employment (and only payroll employment: the rest of the forecast is not out of line with contemporary professional forecasts), and produced an estimate for December 2004 (a) inconsistent with the rest of the forecast, and (b) high by 2.3 million in its estimate of payroll employment--all because Karl Rove and company thought it important to avoid headlines like "Bush administration forecasts 2004 payroll employment to be less than when Bush took office."

White House Media Affairs would have a much harder time pressuring the forecasters to produce a "rosy scenario" if the pressure has to be kept on month after month...

---