The many excuses for fiscal failure
In a recent post I pointed out that many Keynesians, including Paul Krugman, argue that there really hasn’t been any significant stimulus, and then use data that implies only government output is important, taxes and transfers don’t have much effect. But when discussing 1937 they make an argument that implies higher taxes and lower transfers were a major factor in the 1937-38 slump. Stanley Greer directs me to a Mickey Kaus post that finds another inconsistency. Here’s Paul Krugman in February 2009:
Now the centrists have shaved off $86 billion in spending “” much of it among the most effective and most needed parts of the plan. In particular, aid to state governments, which are in desperate straits, is both fast “” because it prevents spending cuts rather than having to start up new projects “” and effective, because it would in fact be spent; plus state and local governments are cutting back on essentials, so the social value of this spending would be high. But in the name of mighty centrism, $40 billion of that aid has been cut out. (italics added)
And here’s Krugman last Sunday:
So what happened to the stimulus? Much of it consisted of tax cuts, not spending. Most of the rest consisted either of aid to distressed families or aid to hard-pressed state and local governments. This aid may have mitigated the slump, but it wasn’t the kind of job-creation program we could and should have had. This isn’t 20-20 hindsight: some of us warned from the beginning that tax cuts would be ineffective and that the proposed spending was woefully inadequate. (italics added.)
Attention Krugman defenders: I am not saying Krugman doesn’t have an argument. I am saying the arguments that he does use are not consistent, and give relatively uninformed readers the impression that the case for fiscal stimulus is stronger than it really is.
There is a parallel with monetary stimulus. In both cases (fiscal and monetary) it would be politically difficult to get a stimulus policy aggressive enough to have a major impact. I choose to fight for monetary stimulus because:
1. I think it’s more effective (as Tyler Cowen says, the central bank moves last)
2. It’s cheaper, doesn’t run up the deficit.
3. It’s easier to persuade the Fed than Congress. The Fed basically follows the consensus of economists; Congress doesn’t (as often.) Thus I’m actually trying to convince my fellow economists that we need more monetary stimulus.
Tags:
12. July 2011 at 10:30
Krugman says in the first quotation that transfers to state and local governments are effective since they’ll spend the money, but that part of the stimulus has been shaved. In the second quotation he says the stimulus was mostly tax cuts, plus a little aid to local governments that mitigated the slump.
I don’t really see this as THAT inconsistent all things considered. I think a better argument is that it’s pushing it to think that an extra $40bn in aid to state and local governments would have resulted in a significantly different path of growth. Is this what you mean by “give relatively uninformed readers the impression that the case for fiscal stimulus is stronger than it really is.”
12. July 2011 at 10:54
John, In the first quote he says S&L aid is the “most effective” type of stimulus. In the second quote he says it’s the ineffective type of stimulus (grouped in with tax cuts.) I guess Mickey and I see an inconsistency there. I’m surprised you don’t.
12. July 2011 at 11:00
Scott, you seem to argue that the Fed always has the power to stimulate the economy. If you are correct, then (of course) there would never be another downturn regardless of how the participants behaved.
What factors, if any, created the need for monetary policy to drift down the the zero bound and do you think that those factors need to be addressed? If any, can monetary policy and quantitative easing remedy those issues?
Also, do you see any risks with the Fed acting as the “stimulator” by engaging in QE3, QE4, … etc. ?
Thanks, Cliff
12. July 2011 at 11:15
Scott You said:
“The Fed basically follows the consensus of economists”…
William Pool has convincingly, in my view, argued against that notion (W. Poole JEP summer 1988), showing that it is the consensus of economists that “follows” the Fed. More interestingly, in the “Original QE Debate” – The Goldsborough Bill of 1932 – in an exchange between Rep Goldsborough and Fed Governor Harrison, Goldsborough argued that “an articulated, explicit policy would render monetary policy potent through its constructive influence on expectations”. Given he´s so into “communicating clearly” doesn´t Bernanke do it? So your “job” is not to convince fellow economists but the much harder one to brake something like 150 years of “atavistic” ties to wrong notions about how MP should be conducted!
12. July 2011 at 11:59
Backing up Marcus Nunes:
http://www.coordinationproblem.org/2009/06/this-sounds-about-right-macro-policy-is-more-or-less-determined-by-politics-rather-than-ideas.html
12. July 2011 at 12:15
Scott,
Again, the arguments are not inconsistent. He’s not saying that state and local spending was bad. He’s saying that it was preventing anti-stimulus.
If the Federal government passes a stimulus package and all it does is mitigate the fall in state and local government spending, the overall effect is 0. (It’s a lot better than no stimulus at all though, where the effect would be negative).
Krugman’s point the whole time has been that we should stop state spending from declining AND have even more federal spending that will stimulate the economy. The point of the second article you cite is just the repetition of an argument PK has used since the beginning.
It’s obviously fine not to agree, but to keep seeing inconsistancies that aren’t there is really a problem. It is as if your dislike of Krugman is forcing you to misread him.
Charlie
12. July 2011 at 12:20
It seems like SS ought to quit nitpicking at Krugman and take full aim at Kocherlakota and Hoenig, who seem to be the biggest obstacles keeping BB from pulling out the stops on monetary policy. Oh, and Senate R’s for blocking Diamond and other Fed nominations, and at O for not nominating more.
12. July 2011 at 13:15
The most important consideration that favors monetary over fiscal policy for purposes of short-run macroeconomic stabilization is that monetary policy is otherwise irrelevant, whereas fiscal policy also matters for long-run economic growth. There is only one factor to consider when setting monetary policy, namely whether there too little, too much, or just the right level of aggregate demand. When setting fiscal policy, one must trade off the adequacy of aggregate demand against the efficiency of government expenditures/revenues.
To see this tension in practice, witness a number of Krugman discussions from the early days of the Great Recession, in which he explicitly says that we need not perform any cost-benefit analysis of proposed government spending, since in a liquidity trap the social marginal cost of government spending is nearly zero. This may be correct as a point of theory, but insofar as both monetary and fiscal policy are capable of producing the same boost to aggregate demand, presumably it is better that the boost be directed towards goods and services with a higher marginal social benefit. Given that a primary concern for fiscal stimulus is getting the money out the door quickly, we should expect the efficiency of government demand qua fiscal stimulus to be lower than the efficiency of private demand qua monetary stimulus.
12. July 2011 at 13:55
DeKrugman is LYING. The only thing that matters in his “analysis” is his firm belief that:
“so the social value of this spending would be high”
—-
And THAT is why we know he is lying.
Look, remember the wisdom of Solomon? Two women each claim a baby, he says fine let’s cut it in half… the real mother screams, “no!”
When you want to find a liar, give them exactly what they ask for but nothing that they want.
As such: if fiscal will work according to DeKrugman, then fine – just tell us how much we have to cut the taxes of the top 90-99%, so that AD will go up to the levels DeKrugman wants.
There will be no more spending on his side, so we can judge HOW MUCH HE BELIEVES HIS ECONOMICS by his willingness to advocate for more AD as pure tax cuts to Tea Party crowd.
Not the top 1%, the 90-99%.
The same goes for Scott, if you want QE, remember the beauty of Cochrane’s futures – the new printed money goes directly to the guys who BET we’d come up short – they get the newly printed money.
This way, goldman doesn’t get to sell T-Bills with an extra determined buyer in the market.
And if that’s not enough and we need a helicopter drop, then print the money and give it to the Tea Party crowd – buy their agreement.
—-
The point is, thing are either so bad that right now the king is going to cut the baby in half and DeKrugman will cry out to let the other side keep the benefits OR he isn’t really a loving mother.
12. July 2011 at 14:11
Why Krugman’s hard shift away from monetary policy and toward fiscal? I think he truly believed that Bernanke was more likely to let Fiscal policy makers create inflation than to create it himself. If you assume, preventing deflation aside, that the Fed is essentially passive, then the rational thing to do is advocate the policy that maximizes the utility of that inertia. Has he really been proven all that wrong in these assumptions?
12. July 2011 at 14:39
Cliff, You are confusing real and nominal variables. I believe the Fed always has the power to boost NGDP, and so does everyone else with half a brain.
But that doesn’t mean we can’t have recessions. There are supply shocks, after all.
The interest rate fell to zero through a combination of Fed errors (money too tight) and a weak real economy (housing crash followed by weak auto sales as oil prices soared.)
With TIPS markets there is no risk from stimulus, none at all.
Marcus, I did a recent post where I quoted from that Congressional exchange. I’m not convinced on causation, but I’ll keep an open mind.
Wonks Anonymous. But surely during the Great Moderation the Fed was influenced by the ideas of the monetarists. And those ideas were developed long before they were adopted by policymakers.
Charlie, You said;
“Again, the arguments are not inconsistent. He’s not saying that state and local spending was bad.”
I never said he said it was bad. Read my post again. And the two comments are so obviously inconsistent it’s not even worth debating.
You said;
“If the Federal government passes a stimulus package and all it does is mitigate the fall in state and local government spending, the overall effect is 0.”
But that’s not what the Federal government did. And no one’s disputing that point. And that point has no bearing on my post. The issue is whether S&L aid is a weak or powerful form of stimulus, that is all.
Furthermore, Krugman now seems to think that government spending doesn’t matter, government output matters, so you are defending an argument Krugman is not making.
You said;
“Krugman’s point the whole time has been that we should stop state spending from declining AND have even more federal spending that will stimulate the economy.”
Of course that’s his point, but that has no bearing on this post. None at all. You seem to be trying to show that Krugman thought the stimulus, all things considered, was grossly inadequate. I agree. He did make that prediction and he was 100% right. But that has no bearing on this post. You are discussing how much fuel was put in the tank, I’m talking about whether it was 87 octane of 93 octane.
engineer27, In the last 2 years I’ve done 100s of posts bashing the Fed, some calling out the names you mentioned, and other names. But with Krugman there are new things to talk about each day. Not with those guys. I also have a recent post praising Krugman for being “magnificantly right” on monetary policy. Something I’d never say about the hawks, and something Krugman would almost never say about a conservative.
Ram, Those are good points.
Morgan, You said;
“The point is, thing are either so bad that right now the king is going to cut the baby in half and DeKrugman will cry out to let the other side keep the benefits OR he isn’t really a loving mother.”
Nice application of the old story.
Shane, But the Fed isn’t completely passive, that’s what I keep emphasizing, but no one seems to get. They do QE1, then they pull back, then they do QE2, then they pull back. They are very clearly steering the economy, which changes EVERYTHING. Sure I’d like them to be more aggressive, but they aren’t always passive, they move the tiller in both directions. The main problem isn’t that they don’t steer at all, but that they steer toward too low a goal. Yes, I’ve criticized their lack of steering, and that’s a problem too. But the main problem is too low an implicit NGDP target.
12. July 2011 at 14:39
I see what you’re saying now. In the full original post (from February 2009) he criticizes the tax cuts and says it was a mistake to cut aid to state and local spending from the stimulus. Perhaps I read him too favorably originally, but you’re right that in the second quote he’s not nearly as in favor of the aid as he was in February 2009 (I interpreted him as meaning that we should have done more aid in both quotes).
Regarding the second quote, he also makes the same point again here:
http://krugman.blogs.nytimes.com/2011/07/10/where-the-money-went/
12. July 2011 at 15:05
OT, the FOMC minutes for the June 21-22 meeting are now out. http://www.federalreserve.gov/monetarypolicy/fomcminutes20110622.htm
There seems to be some sharp dissent among FOMC members [more than in earlier minutes] about the medium-term policy outlook and the implications of recent price rises. Interesting stuff.
12. July 2011 at 16:43
Scott,
“The issue is whether S&L aid is a weak or powerful form of stimulus, that is all.”
He isn’t writing about state spending versus works spending. You only think he is, because you are misreading him.
Let’s parse the sentences:
“This aid may have mitigated the slump, but it wasn’t the kind of job-creation program we could and should have had.”
He’s saying state spending alleviated some pain, but we should have done more federal spending. (notice he doesn’t say state spending was ineffective)
“This isn’t 20-20 hindsight: some of us warned from the beginning that tax cuts would be ineffective and that the proposed spending was woefully inadequate.”
Tax cuts = ineffective
Spending = inadequate (notice he doesn’t say inadequate and ineffective)
It seems you are confusing PK saying tax cuts are ineffective with him saying spending was woefully inadequate.
I totally agree that you believe Krugman is making the argument that state spending is worse than job creation spending. But that’s because you are completely and totally reading it wrong. My comment was to explain to you what he was really saying. We should have had state spending AND job creation spending, which if you’ll recall is what he’s always said.
12. July 2011 at 17:58
John, Thanks for that link.
anon, Yes, and I think we know who the two loudest hawks were.
Charlie, If he’s not saying that S&L aid is not the powerful stuff in quote two, then it literally has no meaning. He clearly denigrates it as less powerful than federal job creating programs. It wasn’t the kind of program we “should have had.”
Of course he wanted both. He wants everything. But this is about which types of stimulus are “most effective” In quote one S&L aid is clearly most effective, in quote two it clearly isn’t most effective. There really isn’t any way to dispute that, it’s plain as day. Unless you believe that doing something that was not “what we should have done” was somehow still the “most effective.” That’s a fairly elastic stretch of the English language.
Or perhaps he is a very, very poor communicator. In which case why is he writing for the NYT? Seriously, he’s a very skilled communicator.
12. July 2011 at 18:07
Of course I agree with you completely Dr. Sumner. But why doesn’t the fed? Why the charade about being passive technocrats?
I like the Nick Rowe post on the problem with the idea of doing nothing (http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/06/the-macroeconomics-of-doing-nothing.html). I think it explains a lot about Bernanke. He’ll stop deflation or disinflation. He may let Fiscal policy increase inflation. But he won’t create inflation himself (even in the form of NGDP targeting). Like Obama, maybe he’s just an incorrigible incrementalist. He’s made his contribution to the profession by stopping the deflationary spiral, and he doesn’t want to undo the “great moderation,” the achievement of his forerunners in the position.
12. July 2011 at 18:35
I think there is a way to read Krugman’s two statements that isn’t contradictory or insanely unlikely, and therefor is probably the best way to read them. Namely: “Aid that keeps states from cutting spending has a high multiplier, but in and of itself is hardly fiscally expansionary; just like not massively slashing government spending has a high multiplier, but is not traditionally thought of fiscally expansionary. In a world where the federal government responded to downturns by cutting spending, I would endorse the IMF giving us money to keep those cuts from happening. And that would be a huge help! But it would not be an expansionary policy. It wouldn’t be the kind of job creating program I endorse when faced with the zero lower bound and a fed that won’t listen to scott sumner.”
12. July 2011 at 19:35
Scott,
You are still reading the passage as if he is arguing we should have had more job creation instead of state spending, because job creation is more effective. Yet, you and I both know he has always wanted both. If you don’t read some sort of implied “instead of” into the sentence, it’s perfectly consistent with what he’s always said.
He’s saying “the state spending helped, but we could have and should have done more.” That’s why he’s saying the tax cuts were ineffective and the spending was inadaquate. He deliberately separated those two, plain as day. When he references his 20-20 hindsight, he’s clearly telling us that he’s just reiterating the same argument he’s made the whole time.
If my reading is correct, it is exactly the same thing he’s been saying for three years. He’s had countless blog posts about it, as well as articles. If your reading is correct, he’s made a major policy shift in three sentences that are just a small piece of the overall article. He’s made this major policy shift in the absence on any blog posts supporting his new view. How can you really believe that?
12. July 2011 at 20:05
Thanks to John Hall (above) for a link that helps prove my point.
From PK:
“Note that the aid to individuals was basically safety net, and the aid to state and local was about mitigating spending cuts rather than spending expansion. Basically, this was at best an attempt to beef up automatic stabilizers.”
Again he’s talking about, with almost the same wording, that S&L aid was stopping contraction rather than creating expansion. His point is not that it’s ineffective (like tax cuts) “but it wasn’t the kind of job-creation program we could and should have had.”
“Seriously, he’s a very skilled communicator.”
I agree, which is why it’s very annoying that there are so many that get giddy pleasure out of reading something Krugman writes, completely misinterpreting it, and then saying he just completely contradicted everything he’s ever said (then they get metaphorical high fives from their blogging buddies or commenters). Most of these people I can ignore, but you are otherwise intelligent, so I am forced to endure it.
12. July 2011 at 20:13
Now I’m paranoid about what’s obvious and what’s not.
About my post above, it’s obvious Krugman is for both mitigating spending cuts and spending expansion, right? The job-creating program we could have and should have had” is in addition to mitigating spending cuts, not instead of.
Krugma ends the blog post with the same point he was making in that part of the article “So much for “we tried Keynesian policies and they didn’t work.”
.
12. July 2011 at 20:13
The new Scott Sumner has grown bushcricket cojones.
Finally!
Not to toot my own horn, but…
“Toot! Toot!”
13. July 2011 at 03:44
Scott, you’re overdoing it I think. If Krugman said in 2009 that the multiplier on aid to state and local governments was 1.8 and in 2011 that it was 1.2, would you really be bothered by the inconsistency? You shouldn’t be. Even his “relatively uninformed readers” ought to be aware that such numbers are rough estimates.
13. July 2011 at 11:22
shane, I agree about doing nothing, I have a “doing nothing” post coming up.
Robert, Maybe, but its interesting how Krugman is the Great Communicator, and yet people keep defending him by arguing he’s a horrible communicator–he meant something really different from what he said. I respond to what he said, as does Mickey Kaus, and most importantly, HIS READERS.
Charlie, You said;
“You are still reading the passage as if he is arguing we should have had more job creation instead of state spending, because job creation is more effective. Yet, you and I both know he has always wanted both. If you don’t read some sort of implied “instead of” into the sentence, it’s perfectly consistent with what he’s always said.”
No I’m not, he wants both strongly effective and weakly effective stimulus, I agree. I just wish he make up his mind which is which.
You said;
“How can you really believe that?”
How can I believe that he frames the argument in the way that makes his side look best? Maybe because he’s been doing that for 10 years, as I’ve attested to in other posts. Remember the post where he claimed Mexico wasn’t doing well, despite the election of a man who favored neoliberal reform? How true, of course he didn’t mention that the man implemented no neoliberal reforms, as Congress blocked them. So is it true? And what about the great fiscal contraction of 1937? What’s that based on? Taxes and transfers?
BTW, to an economist there is no difference between reducing contraction and increasing expansion. The multiplier is just math. He’s discussing which fuels have higher octanes.
Kevin, But he says there’s been no fiscal stimulus at all, but then in 1937-38 there was a big fiscal contraction. You have no problem with that? That’s twisting facts to make your model look good. And BTW, there is no rational for even counting S&L in the multiplier, as the Fed’s don’t control it. It’d be like counting investment. It’s endogenous. Which means the FEDERAL fiscal multiplier was far smaller than promised even if the Federal Reserve doesn’t neutralize the effects. And in 2009 the Fed did neutralize the effect.
13. July 2011 at 11:23
Morgan, Are you going to make me look up bushcricket?
14. July 2011 at 04:18
There is no inconsistency here- you are misunderstanding Krugman-
Krugman has argued that aid to states is one of the most effective forms of fiscal policy. However, his argument in the second post is that this only works to prevent fiscal policy from being contractionary, it can’t make it expansionary on its own. Thus it is not inconsistent to believe that state aid “mitigate the slump” but was still vastly insufficient to count as a test of fiscal policy.
14. July 2011 at 05:48
Inconsistent or not, there’s an obvious conclusion: not all demand is created equal.
14. July 2011 at 08:43
Andrew, He said it wasn’t the KIND of program needed, not it wasn’t the amount needed.
Greg, Maybe, but that’s not the issue here.
14. July 2011 at 18:06
Yay, Andrew has reading comprehension!
“Andrew, He said it wasn’t the KIND of program needed, not it wasn’t the amount needed.”
Again, that reading only makes sense if he is talking about replacing S&L with state spending. If you agree that the “kind of program we needed” included both S&L spending or Federal job programs, then he’s not offering any new information about the relative multipliers of the two options.
That is also consistent with the very next sentence, which says taxes were ineffective and spending was inadequate. If your reading was correct there’d be no reason to separate the two.
Finally, both his article and blog post are couched in larger points about where did the stimulus go. His argument in both places is tax cuts aren’t very stimulating and spending increases just mitigated spending cuts already happening at the state and local level, “the aid to state and local was about mitigating spending cuts rather than spending expansion.” There is no opinion being offered about the relative multipliers of state aid vs. federal spending. The federal spending on state aid only served to cancel out the decline in spending by the states.
There’s nothing new hear. Is this how English professor’s argue? Both thinking the other one has no reading comprehension?
15. July 2011 at 10:00
Charlie, To a macroeconomist there is no difference between boosting spending, and mitigating cuts. It’s the same thing. We look at ceteris parible cases. X cause Y to be higher than without X.
I’ll say it again. The first post suggested S&L aid is the high octane stuff, unlike taxes.
The second suggested it was the low octane stuff, not the “kind” of spending we really needed.
17. July 2011 at 03:26
Scott,
One could easily say that the kind of stimulus program we need is a much larger one, capable of substantially increasing total gov spending rather primarily mitigating cuts. That is not necessarily implying that state aid isn’t effective and shouldn’t still be a part of the program, but only that it will only take you so far and needs to be combined with other spending measures.
I think that now you are just being picky about semantics, which I understand are important in academia. However, Krugman probably makes 5-10 blog posts a day and parsing through them looking for a single word that might have been better chosen is probably not fair… plus taken in context I think that Krugman’s intended meaning is nevertheless clear!
17. July 2011 at 16:54
Andrew. There’s plenty at stake here. Krugman keeps changing the goalposts. At times taxes count, at other times they don’t. Mickey Kaus had the same reaction, as I’m sure most others would as well.