Felix Salmon

The Obama administration’s biggest macroeconomic mistake

By Felix Salmon
October 13, 2011

I’m late to Ezra Klein’s big article about whether the Obama administration could have avoided our current economic woes, because I was having dinner last night with the head of the Bureau of Economic Analysis, and I wanted to see what he had to say first. And I’m glad I did!

In any case, here’s Ezra, who looks at the famous chart projecting falling unemployment with the stimulus plan — something which, obviously, never happened.


To understand how the administration got it so wrong, we need to look at the data it was looking at.

The Bureau of Economic Analysis, the agency charged with measuring the size and growth of the U.S. economy, initially projected that the economy shrank at an annual rate of 3.8 percent in the last quarter of 2008. Months later, the bureau almost doubled that estimate, saying the number was 6.2 percent. Then it was revised to 6.3 percent. But it wasn’t until this year that the actual number was revealed: 8.9 percent. That makes it one of the worst quarters in American history. Bernstein and Romer knew in 2008 that the economy had sustained a tough blow; they didn’t know that it had been run over by a truck.

This is an argument I’m very sympathetic to. There’s a counter-argument, which Ezra goes into at some length, which says that even if we’d known how bad the economy was at the end of 2008, it simply wasn’t politically possible to get a bigger stimulus than the one we got. But how far off were we, really? I talked to the director of the BEA, Steve Landefeld, last night, and he made the case that we weren’t all that far off. If he’s right, the Romer and Bernstein projections wouldn’t have been all that different even if we’d known the exact figure.

One thing it’s important to remember, here, is that the numbers Ezra’s quoting are quarterly figures which are then annualized by raising them to the fourth power. So what we’re actually talking about, for the fourth quarter of 2008, was en estimate that the economy had shrunk by 0.9% that quarter, which was ultimately revised to say that the economy had in fact shrunk by 2.2%. That’s a big difference, of 1.3% of GDP in one quarter alone. So how come, if you look at the size of the recession as a whole, the revision actually seems to shrink, to just 1%?

The revised estimates show that for the period of contraction from 2007:Q4 to 2009:Q2, real GDP decreased at an average annual rate of 3.5 percent; in the previously published estimates, it had decreased at a rate of 2.8 percent. The cumulative decrease over the six quarters of contraction is now estimated as 5.1 percent, compared with 4.1 percent in the previously published estimates.

The problem here is that the “previously published estimates” were the ones which came out a few months after the Romer-Bernstein graph, showing the economy shrinking by 6.3% in the fourth quarter of 2008. Here’s the BEA’s chart; note that it simply doesn’t show the 3.8% estimate.


But what this chart does show is that the really big miss, as far as GDP statistics are concerned, was in the fourth quarter; the other quarters weren’t nearly as bad. And I just don’t believe that a single datapoint for advance GDP would have thrown off the unemployment estimates of some of the world’s smartest economists by that much. Would Romer and Bernstein have projected slightly higher unemployment numbers if they’d known the truth about GDP? Probably. But I doubt they’d have been substantially higher. And there’s no way that their “with stimulus plan” estimates would have gotten anywhere near 10%.

Ezra does a very good job of explaining why that is. Romer and Bernstein were basically treating the recession as though it were a common-or-garden cyclical downturn. Which was a big mistake, and one which was pointed out in March 2009 by Carmen Reinhart and Ken Rogoff. “The recessions that follow in the wake of big financial crises tend to last far longer than normal downturns, and to cause considerably more damage,” they wrote, adding that “so far the U.S. experience has mirrored past deep banking crises around the world to a remarkable extent”. And economies simply do not recover quickly from deep banking crises — financial crises, as a rule, cause L-shaped recessions rather than V-shaped ones.

The fiscal prescription for an L-shaped recession is very different from the fiscal prescription for a V-shaped recession. And what we got was a prescription for something which would accelerate the pace at which we recovered. It was not something which would try to fix the fundamental problem of overleverage, which both caused the crisis and which now threatens to hold back the economy for a decade or more.

Here’s Ezra:

In late 2008, when the economy was cratering, Holtz-Eakin convinced McCain that the way out of a housing crisis was to tackle housing debt directly. “What we proposed at the time was to buy up the troubled mortgages, pay them off and let people refinance at the lower rates,” he recalls. “That would have filled up the negative equity and healed bank balance sheets.”

To this day, Holtz-Eakin thinks the proposal made sense. There was one problem. “No one liked that plan,” he says. “In fact, they hated it. The politics on housing are hideous.”

The Obama administration, perhaps cognizant of the politics, was not nearly so bold. It focused on stimulus rather than housing debt. The idea was that if people could keep their jobs and pay their bills, they could pay their mortgages. But today, few on the Obama team will mount much of a defense of its housing policy.

Overall, I’m still unhappy with the state of macroeconomic statistics. I’m not necessarily unhappy with the BEA itself, which basically just has the job of cobbling together GDP data from a very disparate set of inputs, many of which — especially when it comes to the financial sector — are of surprisingly low quality. But I do think that we’d be much better off with a coherent, unified, and well-funded system of data-gathering, rather than outsourcing it to dozens of different public and private sources.

And I’m definitely (albeit with hindsight) unhappy with the way in which the Obama administration hasn’t even tried to fundamentally tackle the enormous amount of debt in the US economy, and the way in which that debt overhang is likely to hold back economic growth for the foreseeable future. We’re turning Japanese, here, and we’re not doing a damn thing about it.

17 comments so far | RSS Comments RSS

What “debt” are you talking about? Private debt? I hope so. Households are still deleveraging. We need more deficit spending.

Posted by petertemplar | Report as abusive

“…the numbers Ezra’s quoting are quarterly figures which are then annualized by raising them to the fourth power.”

They aren’t raised to the fourth power, just multiplied by four.

Posted by Jeff_Holmes | Report as abusive

“The Obama administration, perhaps cognizant of the politics, was not nearly so bold.”

Someone–it ain’t gonna be me, though I know which way to bet–needs to count (1) the number of times in the 2004 campaign that GWB prominently mentioned the “need” to reform Social Security and (2) the number of times in the 2008 campaign that Obama or one of his key advisors used the word “cramdown” in a positive sense.

Ezra can spew all the window dressing he likes, but the core of that stuffed cabbage isn’t the nice, appetizing mix of rice and meats he would like to pretend.

(And I note that none of you who are discussing the piece as if it were a real find, to your immortal shame, are talking about the =mix= that went into that “stimulus package” [which ended up being all sock, no ****].

If you’re capped on the amount you can get–arguable, but when your Administration thinks David Axelrod knows more than Christina Romer, inevitable–you work to get the most out of what you can get.)

BarryO apologized for his idiocy with the stimulus request so long ago that the Kleins of the world are making apologies for it while dropping the admission of ineptitude down the memory hole.

Posted by klhoughton | Report as abusive

Ah, the stimulus. After 8 years in the wilderness, the first big act of the Obama administration carried a double duty: provide stimulation to aggregate demand and incorporate as many specific desires of the coalition of interests that the President believed elected him. And then get it enacted by a Congress in which Republicans will vote no almost to the last man or woman and the Democrats had a sizeable cohort of near-Republicans in their caucuses. The 2009 stimulus, both in size and composition, was what the Democratic party wanted it to be.

Posted by Eric377 | Report as abusive

He could have gotten around the politics by being even bolder. Legalizing the money drop would have been unifying rather than divisive and would have allowed a much greater plan.

Posted by MyLord | Report as abusive

klhoughton, the mix mistake was my take too. It was clear at the time that the bulk of this money shouldn’t have went into keeping public sector employees in paychecks for another year. All that did was inhibit state and local governments from matching spending with revenue earlier. And the “stimulus” had its share of pork, too. If we were going to do infrastructure spending, that was the time to do it, but less than half of the stimulus was spent on that. In terms of what it might do for the economy, I wonder if the infrastructure boat has long since sailed.

Posted by Curmudgeon | Report as abusive

oddly enough the vast majority of the stimulus was tax cuts. there were support for new infrastructure spending, and help for states.
and remove the support for states, and you will get even less demand. so many want to say these jobs don’t matter. fine, then don’t count the jobs that come about because the teacher bought food, clothes, or any thing else. and that could and probably does includes your job.
and retirees who depend on interest income.
while Fed may have set the discount rate really low.
that won’t stop rates from being higher if there was any demand.

but there isn’t any demand

Posted by willid3 | Report as abusive

Ultimately, it’s not about the projections of how the economy was performing and whether the economy performed worse than expected. It’s about the Obama administration’s lack of response to the evidence that the stimulus was failing to achieve its objectives of stemming the increase in unemployment.

A competent leader would have adapted their original plans to fit the developing situation. If they weren’t getting the results they anticipated, they could accelerate other aspects of the plan within the scope of the stimulus, to compensate for things turning out worse than they had originally expected.

And it’s not like they were having to wait to get that feedback. While three months might have been a bit short, shouldn’t there have been some action taken by the President’s administration after six months? How about nine months? Or a year later?

Instead, the Obama administration just let the stimulus play itself out, without any useful intervention on their part, over the two years it ran.

And here we are three years after the stimulus was launched. It’s sad that one of the big knocks against the President today is the claim that he’s basically already quit as President, stopping doing the job without bothering to resign, choosing instead to campaign for whatever without end.

The lack of any effective response to the jobs situation while the stimulus was running suggests that really happened sometime during his first year in office.

Posted by politicalcalcs | Report as abusive

What could Obama or the administration or the Fed possibly do to increase employment?

Well #1 they could employ people directly via a minimum wage jobs program. That would be, controvercial, hard to manage, and wasteful… but no more so than paying people not to work via extended unemployment benifits, or pushing people into early retirement the minute they become eligible for Social Security at 62. The cost would be about 100B/year assuming minimum wage and zero benifits.

#2 The goverment could take an even more controvercial step and shift all FICA taxes to workers. This would pull money out of the paychecks of workers instantly. It would be massively regressive as it only hits people making below the social security wage base… and it would make US workers cheaper to employ and would immediatly boost employment. It would hurt everyone with a job and help everyone without one.

Posted by y2kurtus | Report as abusive

“pushing people into early retirement the minute they become eligible for Social Security at 62″

Aren’t Social Security benefits calculated so that the value of the payout is roughly the same at 62, 67, and 70? More years but less money…

Comparing direct employment at the minimum wage to unemployment benefits is trickier. I suspect unemployment benefits for many are *above* the minimum wage? In which case paying them less would definitely save money. For the rest, it wouldn’t save money — it would simply get something done that hopefully benefits society.

Of course the unemployed aren’t all sitting on their bum in front of the TV. Many of them are working odd jobs under the table. Many of them are performing household work (esp. child care) that would otherwise be farmed out. So making them actually work for their handout would leave them substantially less well-off.

“The goverment could take an even more controvercial step and shift all FICA taxes to workers.”

Ironically, their present policy is the exact opposite. Obama’s FICA tax rebate is on the *worker* share, not the *employer* share. Moreover, there has been a big crackdown on 1090-MISC payments, forcing many situations to be converted to W2 employee status (and thus forcing the employer to pay half the FICA liability).

Posted by TFF | Report as abusive

“oddly enough the vast majority of the stimulus was tax cuts.”

…which do little or nothing to stimulate demand, partly because the benefits flow primarily to people who already have enough to meet their needs. (Those who are struggling pay very little tax.)

Posted by TFF | Report as abusive

Actually, Jeff Holmes, they do raise to the fourth power. Look at step two of how to compute an annualized rate:
http://www.bea.gov/faq/index.cfm?faq_id= 122

Posted by FelixSalmon | Report as abusive

Go ahead and multiply by four… It is a simpler computation, and the difference between that and the actual computation is far less than the uncertainty in the inputs.

Posted by TFF | Report as abusive

So, we got a series of measures (of varying effectiveness) designed to be counter-cyclical to/cure a V-shaped, business-cycle recession, when the real problem was a L-shaped, balance-sheet recession. And those measures weren’t even aimed at, nor particularly effective at resolving that, the real problem. Lovely.

We not going to get inflation going to cure the problem. The banks, if forced to mark all their assets to market, are still insolvent. The politics of debt relief stink. Feldstein’s idea, presented in FT, may be the start of a way out of this trap.

Posted by dellbell | Report as abusive

We’re turning Japanese.

Yes, I really think so.

http://www.youtube.com/watch?v=VqZ_dC1T7 pA

Posted by JohnBChilton | Report as abusive

Smart cramdown would still make a lot of sense. So would getting a HFHA director who was interested in the long term stability of Fannie and Freddie and the overall housing market. Obama’s team really did whiff on housing. Still are, and they don’t have to be. Making the banks own their piece of the housing mess via bankruptcy and cramdown goes hand in hand with the jobs push as a middle class strengthening effort.

Posted by DaveRoberts | Report as abusive

1. If to dig deeper into BEA’s publiations one can find an unofficial estimate of the uncertainty in the GDP growth rate of 1% per year or annualized 4% per quarter. Thusall revision you have mentioned are within the limits and 8.9 not worse, actually, than 6.3%. Both values inside 4%.
2. Okun’s law is very relibale for the US(http://mechonomic.blogspot.com/2011/1 0/some-corrections-to-david-altigs-job.h tml) but BEA statistics makes a big difference when used as it is – dhttp://mechonomic.blogspot.com/2011/10/ beware-of-bea.html
3. real proble is that there is no comparability of GDP estimates over time – http://mechonomic.blogspot.com/2011/10/b eware-of-bea.html

Posted by ikitov | Report as abusive

Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/