We Now Return You To This 24,536th Episode Of “Bloggers Debating Whether The Stimulus Worked”

Megan McArdle:

Today’s earlier post has naturally aroused suspicions that I am simply a knee-jerk stimulus opponent.  This is not true.  I tepidly supported the notion of stimulus, though I also thought that the stimulus we did was not well executed, because Democrats wanted to use it to execute their policy priorities, rather than to provide maximal stimulus.  I said at the time that “shovel-ready” infrastructure projects really weren’t; that’s not how the government procurement process works, and the federal government tends to slow things down, not speed them up.  As it was under George Bush, politicians seemed more interested in using stimulus as an excuse for stuff they already wanted to do, than in actually figuring out what was most stimulative.  (My candidates:  payroll tax cut, more lavish unemployment benefits).

And what do I think now?  Well, protestations aside, the stimulus we ended up doing was huge.  Maybe it wasn’t as huge as some would have liked, but $800 billion dollars is almost 6% of GDP. (2% if you spread it over 3 years, as the stimulus was).  All told, running a deficit of 10+% of GDP ought to have some pretty powerful stimulative effect.

So far, I’ve been underwhelmed.  Maybe we were going to end up so far in a ditch that we wouldn’t even be able to see sky when we looked up, only mud.  On the other hand, maybe simply not repeating the massive, massive monetary mistakes of the 1930s was enough to keep us out of the Great Depression, and the stimulus merely tinkered around the edges.  I find the latter at least as plausible as the former.

When you combine middling and hard-to-prove results with the fact that the stimulus we got was so obviously wrapped up in the political agenda of the Democratic Party, I think that the case for stimulus has gotten weaker.  Stimulus is always going to get wrapped up in the agenda of whatever party is in power–it will concentrate too much on long-term strategic attempts to change the aggregate level of government expenditure; it will be deployed inefficiently; it will be stretched out to maximize electoral rather than economic effectiveness.

I’d much rather see people talking about how best to ease the economic transition.  One way to think of recessions is that they always represent the transition point between two states of the economy:  high inflation to low inflation, overleveraged to under, or what have you.  Those transitions involve a great deal of human suffering, and contra the “work the rot out” school of economic policy, I do not think that this suffering is necessary.

Stephen Stromberg at WaPo:

But the bigger problem is that the argument about the stimulus is basically unwinnable, for now. That’s because, when you look past the gotcha quotes, much of the Obama administration’s justification for the Recovery Act relies on a difficult-to-prove counterfactual — that it created or saved jobs.

So, unemployment could even be going up, and the president could still claim the stimulus was a success, since it could have been worse. Here, the GOP conference ignores this argument entirely, as it so often does, by pointing to negative raw data and ignoring the Democrats’ argument that you have to compare them to a far more catastrophic baseline. Obama’s critics on the left often do the same to conclude that the stimulus was much too small.

I tend to sympathize with the administration on this. It’s shocking how quickly people have forgotten the rank fear that pervaded the country in late 2008 and early 2009. As with the Troubled Assets Relief Program, approving the stimulus was important if only to prove that the government would not let the economy fall off a cliff. At the same time, policymakers had to pay attention to folks worried — without much reason in the short-term, it turned out, but it was hard to know that then — about inflation and the willingness of private capital to finance American government debt at low interest rates. How do you quantify the number of jobs saved from ministering to such psychological preoccupations? It’s hard, but it doesn’t mean they weren’t saved.

Tim Cavanaugh at Reason:

Where do Keynesians go now that even public radio is talking about the failure of one of history’s costliest Keynesian stimulus efforts?

At City Journal, Guy Sorman notes how quickly the managed-market winds have shifted. When the credit unwind started, the papers, the TV and the newsweaklies declared capitalism dead in just a little less time than it took for Kent Brockman to declare his loyalty to the Space Ants. Less than two years later, you can’t buy good press for the stimulus; the economy is frozen solid in August; the nation is rediscovering — despite the herniated efforts of local, state and federal government — the virtues of thrift; and if you search for Keynes on the interwebs, all you turn up are headlines like “How Dr. Keynes killed the patient.”

But here’s the tell: We’re already starting to see the first of the “Today’s Keynesians are misreading Keynes” walk-back arguments.

As an idea, neo-Keynesianism is dead. (As public policy, of course, it will live forever.)

Josef Joffe at TNR:

As the joke goes, this is the reason why there are no one-handed economists; they are always saying: “on the one hand, on the other…” But, in the second year of the stimulus, it has become a lot harder to argue that Keynes is cool, and it is a lot easier to contend that tightfisted Angela Merkel, though a physicist by training, has a better knack for economic management than the Bloomsbury Sage—at least 74 years after he published his General Theory.

OK, it is silly to draw so much conclusion from so little evidence, and, whatever the gainsayers argue, Obama will throw more money at the economy as the midterms draw closer. But think of the issue in the simplest of terms: As a businessperson or consumer, would you now spend your money in the face of a rising deficit that is guaranteed to bring on either inflation or tax hikes or both? Maybe you can find the answer in the Holy Writ that is the General Theory. Confidence matters, and this is what the great man had to say in Book III: “The government program may, through its effect on ‘confidence’, increase liquidity-preference or diminish the marginal efficiency of capital, which may retard other investment…” Translation by Joffe: “The stimulus may lead to cash-hoarding and, by depressing the return on capital, diminish private investment.” Keynes may have been wiser than his disciples.

Jonathan Chait at TNR:

The truth is that you can’t prove whether the stimulus worked or not. You can try to reconstruct the effects of the stimulus (and other government interventions) as Alan Blinder and Mark Zandi did. But you’re still making assumptions on the basis of economic models — mainstream assumptions shared by most economists, but assumptions nonetheless. Actually proving the case would require going back to 2009 and re-running history with everything the same but no stimulus. Since we can’t do that, all we can do is guess.

Of course, Republicans aren’t making a good faith effort to gauge the effects of the stimulus. They’re simply looking for a pseudo-economic argument that allows them to blame Obama for everything that has happened since the economic crisis of 2008. Lindsey’s article is a rationalization for that political strategy.

Jim Manzi at TNR:

Chait is right. The reason we don’t know whether the stimulus has worked in the United States is fundamental. And I believe it has significant implications for how we should approach public policy in the recession.


I believe that recognition of our ignorance should lead us to two important, though tentative and imprecise, conclusions.

First, we should treat anybody who states definitively that the result of stimulus policy X will be economic outcome Y with extreme skepticism. And weaseling about the magnitude of the predicted impact such that all outcomes within the purported range of uncertainty still magically lead to the same policy conclusion doesn’t count; we should recognize that we don’t even know at the most basic level whether stimulus works or not.

Second, “boldness” in the face of ignorance should not be seen in heroic terms. It is a desperate move taken only when other options are exhausted, and with our eyes open to the fact that we are taking a wild risk. Actual science can allow us to act on counterintuitive predictions with confidence–who would think intuitively that it’s a smart idea to get into a heavy metal tube and then go 30,000 feet up into the air? But we don’t have this kind of knowledge about a stimulus policy. We are walking into a casino and putting $800 billion dollars down on a single bet in a game where we don’t even know the rules. In general, in the face of this kind of uncertainty, we ought to seek policy interventions that are as narrowly targeted as is consistent with addressing the problem; tested prior to implementation to whatever extent possible; hedged on multiple dimensions; and designed to be as reversible as is practicable.

What I am trying to describe here is not a policy per se, but an attitude of epistemic humility.

Chait responds to Manzi:

Manzi is a much nicer person then I am, so it’s no surprise that I will return the favor of him saying that I’m right by replying that he’s wrong. Here are a few of the problems with his case. First, and more importantly, I was arguing that the precise effect of the stimulus can’t be measured. That doesn’t mean we have no idea whether it worked. There is a general, though not unanimous, consensus within the economic field that increasing spending or reducing taxes temporarily increases economic growth. Basically, we do know the rules — increasing the deficit in order to pave some roads and cut taxes for middle-income people will increase the size of the economy; the primary debate is just how much.

Now, it’s true that the conservative movement has invested a great deal of time into throwing cold water on this basic consensus. I think this campaign should be viewed as largely political. Private forecasters unanimously believe that fiscal stimulus can temporarily boost growth. They give no credence whatsoever to the various right-wing alternative models in which fiscal stimulus does not boost growth. Moreover, in 2001, when the objective case for fiscal stimulus was much weaker, there was no real debate about the efficacy of fiscal stimulus. The fact that Republicans are fiercely contesting the merits of fiscal stimulus now, while almost nobody was doing so when the case was much weaker in 2001, suggests that the right’s skepticism is a political phenomenon.

Second, we are not really taking a “wild risk” by devoting $800 billion to mitigating the deepest economic crisis since the Great Depression. The long term fiscal cost of the stimulus is quite minimal:

The opposing risk, of allowing an economic free-fall, seems much greater. The risks of a depression are enormous. Of course, one side effect of such an outcome would be massive political gains for the opposition party. No doubt this helps explains the more sanguine attitude Republican elected officials — I am not implicating Manzi here — took toward the 2008 crisis, as opposed to their urgency in the face of the far more shallow 2001 recession.

Third, Manzi suggests that remedies be as narrowly targeted as possible, reversible, and tested prior to implementation. The stimulus was pretty narrowly targetted. It consisted of spending and tax cuts designed to increase consumption. It is completely reversible in the sense that the spending and tax cuts are temporary.

Being “tested,” as I noted, is not possible. But it’s not possible to test very much in macro-economics. We can’t run natural experiments with whole economies without the aid of time travel. Manzi’s analogies of flying in an airplane assume is that there’s some safe, default option — staying on the ground, keeping away from the casino. That isn’t a realistic way to think about economics. Doing nothing in the face of economic catastrophe is a policy choice. To the extent that it’s been “tested,” it’s been shown to produce terrible results. The better attitude than trying to imagine some safe default course of action, I’d suggest, is to hew to precepts generally accepted within the economics field.

Manzi responds to Chait:

Chait begins his reply by claiming that I “oppose any stimulus at all.” This is a position which I did not present in the post, and which I do not hold. In fact, I have consistently advocated stimulus in the face of the current crisis, and generally in venues that are not as hospitable to this idea as The New Republic.

I was arguing in my post that we should approach stimulus with appropriate humility about our knowledge, not that we should never execute such a policy. That’s why the sentence in my post that immediately follows those Chait excerpted, is: “What I am trying to describe here is not a policy per se, but an attitude of epistemic humility.”

Chait then moves on to criticize the reasoning behind my imagined position against any stimulus spending.

His first argument is that he was only saying we can’t measure stimulus precisely, but that we still know enough to act with confidence:

First, and more importantly, I was arguing that the precise effect of the stimulus can’t be measured. That doesn’t mean we have no idea whether it worked. There is a general, though not unanimous, consensus within the economic field that increasing spending or reducing taxes temporarily increases economic growth. Basically, we do know the rules — increasing the deficit in order to pave some roads and cut taxes for middle-income people will increase the size of the economy; the primary debate is just how much.

Now, it’s true that the conservative movement has invested a great deal of time into throwing cold water on this basic consensus. I think this campaign should be viewed as largely political.

Chait correctly identifies this as the most important of his arguments, so I’ll spend the most time replying to it. The problems with this criticism are that: (1) it is false, (2) it is a straw man, and (3) by far most important, it doesn’t address my point.

1. It is false.

Chait is trying to define the position that stimulus will not increase output as intellectually illegitimate. (Though, again, this is not a claim that I have made.)

It is certainly true that a large majority of professional economists accept the view that “increasing spending or reducing taxes temporarily increases economic growth”—but that is very far from claiming that disputing it is largely a political campaign. Robert Barro, Professor of Economics at Harvard, John Cochrane, Professor of Finance at the University of Chicago, and Casey Mulligan, Professor of Economics at the University of Chicago, have each separately argued that it is somewhere between plausible and likely that the multiplier for stimulus spending under relevant conditions is indistinguishable from zero (i.e., that stimulative spending will not materially increase economic output). According to surveys of professional economists reported by Greg Mankiw, about 10 percent of economists do not agree with the statement that “Fiscal policy (e.g., tax cut and/or government expenditure increase) has a significant stimulative impact on a less than fully employed economy.” Both the Wall Street Journal and the Financial Times have run opinion columns expressing the view that a multiplier of zero is a plausible to likely theory.

I have not been afraid to call out influential conservative activists when I believe they are engaging in crank refusal to accept a scientific finding. But in a genuinely scientific field which has accepted a predictive rule as valid to the point that there is a true consensus—such that the only reason for refusal to accept it is crankery or, in Chait’s terms, “politics”—you don’t usually see: several full professors at the top two departments in the subject, when speaking directly in their area of research expertise, challenge it; 10 percent of all practitioners in the field refuse to accept it; and the two leading global general circulation publications in field running op-eds questioning it.

2. It is a straw man

If the U.S. government were to borrow and spend $1 trillion with the sole result of increasing U.S. GDP in Q4 2010 by $1, it would have “temporarily increased economic growth”—but no sane person would advocate such a policy. It would not be, in either the common-sense meaning of words, or in the terms of my post, a stimulus policy that “worked.” The relevant policy question is whether stimulus spending “temporarily increases economic growth” enough to make such a policy rationally advisable . Economists are all over the place on their estimates for impact of stimulus policy across the range that is relevant to the policy decision.

A great many leading economists may accept the proposition that enough stimulus spending will probably cause at least some increase in output for a short period of time in some circumstances, yet are still uncomfortable with the kind of stimulus spending strategy that is the actual subject of current political debate. In 2009, James Buchanan (1986 Nobel Laureate in Economics), Edward Prescott (2004 Nobel Laureate in Economics), and Vernon Smith (2002 Nobel Laureate in Economics) promulgated this statement:

“Notwithstanding reports that all economists are now Keynesians and that we all support a big increase in the burden of government, we do not believe that more government spending is a way to improve economic performance.”

3. It doesn’t address my point

It is nerdy-sounding, but I believe critical to this discussion, to distinguish between measurement and knowledge. I made a very strong claim about measurement, and a very specific claim about knowledge.

I claim that we cannot usefully measure the effect of the stimulus program launched in 2009 at all. We can call this a “natural experiment” all day long, but in the absence of a control case, we cannot know what output would have been had we not executed the policy. Econometric models are not sufficient to estimate this counterfactual. Therefore, there is no achievable level of output in the United States in 2010, 2011, and so on that would enable a definitive answer to the question, “What was the effect of stimulus spending on output?” See, for example, in my original post, the response of leading economists when confronted by unemployment with stimulus that turned out to be higher than they projected unemployment would be without stimulus:

Ms. Romer famously projected in January 2009 that without government support, the unemployment rate would reach 9%, but with support the government could keep it under 8%. It’s 9.5% today.

Some Obama administration officials privately acknowledge they set job-creation expectations too high. The economy, they argue, was in fact sicker in 2009 than they and most others realized at the time. But they insist unemployment would have been worse without the stimulus.

All potentially useful predictions made about the output impact of the stimulus program are non-falsifiable. Failure of predictions can be simply justified by this sort of ad hoc explanation after the fact.

Adam Ozimek at Modeled Behavior on Manzi:

Specifically, he cites the fact that the University of Chicago’s Barro, Fama, and Mulligan are stimulus skeptics, and according a survey from Mankiw, so are 10% of all economists. But I don’t think 10% of economists and a handful of high-profile experts disagreeing is sufficient to say there is not a strong consensus.

For economics 90% agreement is a pretty high level of agreement, and I would be surprised to find a consensus much stronger on that on most issues. From a survey of economists by Whaples we can see that ”only” 87.5% of economists agree that the U.S. should remove all remaining tariffs and trade barriers, 90.1% believe that employers should not be restricted from outsourcing jobs, 85% agree that subsidies to agriculture should be removed, and the same percent say it about sports subsidies as well. From another survey of economists, 87.5% agree that the U.S. trade deficit is not primarily due to other nations’ nontariff trade barriers, 83.5% agree or agree with provisos that tax policy can affect the long-run rate of capital formation, 93% agree that pollution taxes or tradeable permits are more efficient than emissions standards, 92.9% agree or agree with provisos that flexible exchange rates are effective, and 92.6% agree that tariffs or import quotes reduce the general welfare of society.

Despite the disagreement by 7% to 17% of economists on these issues I would argue that are all accurately characterized as representing as a strong consensus. Whaples calls the agreement in those examples a “consensus” and “an overwhelming majority”, and Fuller and Geide-Stevenson, the authors of the other paper, explicitly refer to those examples as representing a “strong consensus”.

Yet I’m certain that on each of these issues you could find experts at the top 10 economics departments that agree with the minority position. Stiglitz alone will probably disagree with more than half of them, and you won’t have to look hard to find a half a dozen other Ivy League dissenters.

My point is not to disagree with Manzi that a strong consensus means it is okay to call anyone who disagrees with the consensus a “crank” or “politically motivated”, but just to point out that the bar he’s set for a “true consensus” pretty much means that there’s is no “true consensus” on important issues in economics. Then again, he may very well agree with that point.

Karl Smith at Modeled Behavior:

Jim Manzi and I are kindred spirits on the issue of epistemic humility. Out in the real world – as opposed to the whiteboard – only one thing is for sure and that is that this is all going to end very, very badly. The long run is not your friend.

In the mean time, little is for certain. In particular, we are not clear on exactly what the consequences of economic stimulus were. However, it is important to clarify what some economists may have meant by stimulus skepticism.

I wrote frequently in the weeks leading up to the passage of the stimulus that I was a stimulus skeptic. I signed a letter offered by John Boehner for economists who believe that IIRC, stimulus is not the best way to revive our economy.

At no time, however, did I believe that stimulus would have no effect on output. That is, unlike what I believe to be Mulligan and Fama’s stance, I did not believe that labor markets always clear.

What I did think is that I wanted to take what, even then, seemed like enormous monetary risks. It has been rumored that Tim Geithner suggested securing the assets of all major banks in the United States. While I didn’t formally support that, it did not seem absurd to me. Nor, did massive nationalization of the banking system and certainly not aggressive purchases of government securities.

While taking on all of these risks I didn’t see the need to add the confusion and inevitable political food fight of stimulus on top of it. Moreover, if one was going to do a stimulus it seemed much more sensible to simply slash the payroll tax. The bang for the buck would have been less but you can cram a whole lot more bucks through the payroll tax system than you can through the appropriations mechanism. Additionally, the public choice issues in cutting the payroll tax were much more manageable.

The point of all of this is that I listed myself as a stimulus skeptic but I wasn’t at all skeptical about the stimulating powers of government spending. I was skeptical as to whether that was the ideal course of action. That skepticism was as much rooted in my understanding of American political dynamics and my own tolerance for risk as in any scientific claims about macroeconomics.


Leave a comment

Filed under Economics, The Crisis

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s