From today’s Macroeconomic Advisers blogpost, by Joel Prakken:
Frequently, partisan commentators — and even some economists — exclaim that the stimulus has failed because the unemployment rate now exceeds the peak shown in projections prepared before ARRA was implemented. This argument, which clearly — and perhaps intentionally — confuses the pre-stimulus baseline with the incremental effects of the stimulus, would be laughable if it was not taken so seriously in some quarters. For the record, last spring, as the financial crisis that engulfed the economy worsened unexpectedly — but before the stimulus could possibly have had any real effect on the economy — the unemployment rate already had moved above the Administration’s (and many others’) last pre-stimulus projection. So, this is simple: the baseline forecasts were optimistic, but unemployment would be even higher now without the benefit of the stimulus package.
In addition to knocking down some facile critiques, Prakken reiterates some important points:
For those who argue that we couldn’t afford ARRA, we offer these counterpoints: (a) our analysis showed that the legislation “dynamically” paid for more than a third of itself by restoring the economy to full employment sooner; (b) because it is temporary, ARRA has essentially no lasting impact on the unsustainable nature of the country’s current fiscal stance; (c) our calculations suggested that the benefit (e.g., the present discounted value of temporarily higher GDP) exceeded the budgetary cost of the legislation.
About 25 years ago, the Brookings Institution convened a series of conferences to investigate the comparative properties of the extant macro models used by the Fed, government agencies, the OECD, the think tanks and the private consulting firms (and a few academics). The results were published in Empirical macroeconomics for interdependent economies, edited by R. Bryant et al. (Brookings, 1988). Something like that is desperately needed these days, so that those people who are concerned with seriously considering policy problems can sort out why the simulation results differ. (Of course, those who are happy to merely cast aspersions on all models can sit back and toss peanuts.)
Ok, call me a trouble maker, but lately two of our esteemed Nobel Prize winners, Stiglitz and Krugman, are arguing that the stimulus program wasn’t big enough, doesn’t last long enough and “we” deserve another stimulus program.
I’ll try and summarize the reasons for their concern:
a)The stimulus program worked.
b)The stimulus program didn’t work.
So, as confusing as this may sound, if both a) and b) are true, should we do another stimulus program?
Stiglitz already pointed out that the US just prints money to pay off debt, so nothing to worry about there. Krugman publicly prefers dollar devaluation so US exporters can work our way back to prosperity, but that seems difficult to do when we have 192 countries doing the same thing with their currencies and fighting for weak global aggregate demand.
So is once enough or not?
Many of us don’t believe the economy was in the “disastrous” state maintained by the Administration and the economy has performed pretty much as anticipated. Did we need a stimulus program given this or were a few in Washington trying to create a crisis for exploitation purposes? Look how the stimulus money (of which only around half has been spent) was indeed expended: most went to shoring up public employees and their pension funds, very little has made its way onto Main Street. It is hard to maintain that stimulus has worked when so little of it has actually been used for employment creation and economic stimulus. Another factor of interest, I went into Recovery.gov to see what stimulus has occurred in my local area and find nearly all of it is simply city road improvement projects and some DOE and National Academy of the Sciences research projects that were already underway or had been budgeted for/funded this year – stimulus just reclassified the source of funding and on the margin no new jobs were created that weren’t already in process.
Classic trickle-down. Shower the public employees with huge salaries and gold-plated pensions and hope they’ll spend some of their filthy lucre in private-sector businesses.
Would you like fries with that, Madam Deputy Undersecretary for Tropical Entomological Studies?
Cedric Regula & WC Varones, your succinct jabs are right on target.
I’m not sure we enjoy casting aspersions on all models, but it is what these models deserve.
I’ve made this same comment on the previous post, but it may fit better here. My question isn’t rhetorical, and I hope I can get an answer.
So far as I can tell, these exercises in “measuring” the effectiveness of stimulus are demonstrations of what a particular theory or model implies, and not measurement.
My question: Is this a fair contention?
Footnotes: I’m a partisan, but this isn’t a partisan contention and applies equally to stimulus proponents such as MA LLC and opponents such as Taylor. I’m also not suggesting that such exercises have no place, but I think they are being misrepresented as “empirical.” This also helps explain why the “evidence” fails to persuade, without all of us having to accuse each other of dishonesty or idiocy.
The question is not whether the stimulus increases GDP or employment. It will by construction.
But is it a good use of money? Well, the MA piece goes a bit of the way to address that. At least it might be a not too bad use of money, and that’s a good thing.
But there are caveats. The analysis appears senstitive to future interest rates.
Further, one noted benefit is to prevent states from reducing expenditures. In many cases, this is not desirable. New York, New Jersey and California all suffer from structural deficits, and any effort to mitigate the pressure to change is, I would think, more likely to be counter-productive than helpful in the long-run.
The peanut gallery seems to be alive and well.
Especially when the head peanut chimes in.
Steve,
I read an article that pointed out that only 7 states out of 50 did not have a structural deficit.
But you are really opening a can of worms if you question what the money got spent on. Maybe that’s what we worry about?
I’ll admit I didn’t study this in detail, but some news about the details jumped in front of my face so here are a few carrots I can toss around:
1) Infrastructure spending. In emergency stimulus terms this is translated to mean “filling potholes” since that can be accomplished with a minimal amount of upfront engineering and project planning time. But then we get multi hundred billion “highway bills” every other year, gas taxes too, and I would think that by now potholes should be on the endangered species list in this country.
2) I hear some of it was attributed to moving up the ATM. For whatever sense we can make of that. Also there was the small income tax rebate.
3) I read Brad Delong say that most people were happy that schools didn’t have to close. Not to put too fine of a point on it, but the “school stimulus” would have been paid out in summer time, and schools close every year in the summer time. But assuming they saved the stimulus until fall, you would at least expect some economist to take the amount received, divide by the schools annual operating budget, and calculate how many hours the school stimulus kept the schools open. I hear those hours have now run out in California.
So I’m sure there is much more to inspect, but $787 Billion is more than I care to make time for.
Charles N. Steele: …these exercises in “measuring” the effectiveness of stimulus are demonstrations of what a particular theory or model implies, and not measurement.
I think it’s pretty clear in this post and in his previous post that Menzie was restricting his comments to models and theories supported by mainstream economic thinking. Most of us have a sense that even if we may disagree with some of mainstream thought, those models and theories tend to be about as partisan free as you’re going to find. When Ron Zandi (McCain’s chief economic advisor) says that his model essentially agrees with Global Insight and Macroeconomic Advisor, then I think we can at least conclude that there is minimal partisan bias in the analysis. Austrians and Marxists and Institutional economists may all complain that the models are fundamentally flawed, but those are complaints coming from outside the mainstream. That doesn’t make those complaints wrong, but Menzie has to set some boundaries and one way to minimize the charge of partisanship is to refer to the kind of models widely accepted in both academic research and professional forecasting. Menzie was not arguing that mainstream economics is the only valid model, he was trying to show that claims of ARRA’s success have broad bi-partisan agreement within the community of mainstream economists.
It’s also worth noting that one of the reasons mainstream economic theory dominates the models used by professional forecasters is that unlike Austrian, Marxist or Institutional economics, mainstream economics tends to be highly quantitative and “econometrics friendly” whereas alternative models tend to be long on ideological rhetoric and short on quantitative models.
I will echo Steve Kopits remarks.
Assessing whether the stimulus temporarily raised GDP is a low bar indeed.
Better to look at the issue from a dynamic, not static, viewpoint. A temporary rise in GDP causes a problem: stimulus becomes, automatically, “contraction” since upon expiration it acts as a headwind to GDP growth. Mark Zandi of economy.com has some great charts showing that by 3q2010 the stimulus will subtract from GDP growth on a sequential basis.
So is a stimulus that doesn’t “stimulate” a stimulus? By “stimulate” I mean affect animal spirits in a way that results in sustained recovery, not a lapse back into recession or stagnation. From this perspective the jury is still very much out on the success of the stimulus. This is particularly the case given that term premiums are rising globally in response to structural deficit fears. Up to now stimulus proponent have argued that crowding out has not been in evidence. My question is, at what level of 30yr bond yields will they stop making that argument?
2slugbaits, what are you talking about? I’m not asking about mainstream vs. heterodox economics… unless John Taylor, Robert Barro, etc. are considered heterodox.
I’m not even asking about whose mainstream macro theory is better. I’m asking a question about the nature of the enterprise of “measuring” effects of stimulus. It really is a matter of ferreting out the implications of one’s model and assumptions, rather than empirical measurement, isn’t it?
BTW, the above query on yields and crowding out is not a rhetorical question. Why don’t economists publish, up front, the set of facts or outcomes that will disprove — in their own mind — their theses? Does Popper have no influence on their discipline? It is ironic, because it was Keynes that made the famous quote: “when the facts change, I change my mind…”
David Pearson: If you do your research in an academic mode — read up the literature in the textbooks, and then use Econlit or Google Scholar to identify works in peer reviewed journals — then you will find what you desire. In general, blogs, and articles aimed at a practitioner and/or generalist audience, are not going to be explicit about null vs. alternative hypotheses, maintainted hypotheses, etc., and associated testing.
David Pearson,
The thing that comes the closest that I’ve come across is Reinhart & Rogoff. Tho I don’t think they stress rising long term yields due to either credit downgrades or an increase in inflation expectations and mainly analyze the size of government debt as a drag on growth.
I think this is probably because these possible outcomes are speculative and not well grounded in print in a 1950s econ textbook. It would be too difficult to model, and probably considered wacko and unprofessional to consider what may happen if the unwashed, and washed, masses revolted against one’s theories and claim to the throne presiding over the economy.
But at any rate, it doesn’t look to be a problem for the US this year anyway. But I’ll remind economists that people live to be 80 years old on average, and that fact is not addressed anywhere in macroeconomic theory.
I wonder if the academic research contains statements such as, “the effect of stimulus will be negative if real term interest rates rise by…”
I realize this is all very squishy: term interest rates have a variety of drivers, most importantly RGDP expectations, so a rise in term rates could be evidence of success rather than crowding out.
But the point is, a debate about stimulus should be complete. It should start with a goal (was it sparking a sustained recovery, or helping cushion a recession?); and it should be measured against a cost (What happens when we remove stimulus? What is the impact of structural deficits on term interest rates?).
Countries around the world are seeing bond markets more closely scrutinize the nature of their budget deficits. Are they structural or cyclical? How much do projections depend on recovery which in turn could depend on the maintenance of stimulus? Can the economy withstand the contractionary force of the removal of stimulus? If not, will the Central Banks continue to finance fiscal deficits?
So far the answers have been reasonably favorable, but that may change, and change soon.
Charles N. Steele: It’s at least arguable whether Barro and Taylor even believe in macroeconomics as a separate discipline. I think you could make a pretty good argument that they practice a peculiar kind of micro foundations based macro without the macro.
What thoughts do you folks have on boosting demand other than through government spending? I do not like the use of tax payer funds, particularly via deficit spending, but the American public seems reluctant to return to normal spending whether it be due to fear, or due to the lack of money as a result of unemployment. How else shall we get spending to resume?
“…very little has made its way onto Main Street. It is hard to maintain that stimulus has worked when so little of it has actually been used for employment creation and economic stimulus.”
Well, half the 2008 Stimulus and a third of the 2009 stimulus go to provide unemployment benefits and other welfare aid as in Medicaid, food stamps, heating assistance, to people who lost their income from job loss.
Half the 2008 Stimulus and a third the 2009 Stimulus have gone to tax cuts/rebates.
The jobs losses only accelerated after the 2008 Stimulus so that 2008 Stimulus failed to deliver as Bush’s economists promised. The 2009 Stimulus changed the tax cut/rebate of the 2008 plan from a lump sum to one spread over two years.
The one difference from the 2008 Stimulus is the infrastructure and other programmatic spending to stimulate new activity.
Unfortunately this portion was smaller than needed with a lot of the money wasted on preventing state cut backs in their infrastructure spending when the states should hike gas taxes and other fees to pay for repairing and replacing bridges and water and sewer systems when costs were way below normal.
The tax cuts were wasted as they never seem to work as theory claims – I can’t find one tax cut that boosted employment as promised by the economists. That portion should have been spent on infrastructure instead. But Congress and Obama tried to be bipartisan, and gave conservatives their useless tax cuts. Which I see many argue don’t work.
It is amazing to me how uncritical of Republican and conservative stimulus efforts – no conservative economists ever find them to not work, but instead find dozens of reasons why they were prevented from working. Volcker suddenly tightened just when the 1981 tax cut was passed to stimulate the economy, and suddenly provided easy money when Reagan agreed to hike taxes. The Bush and Clinton tax hikes caused slow job growth according to conservatives, while the 2001, 2002, 2003 tax cuts worked but external forces prevented the same rate of job creation that was constant for most of the 90s.
The way I see the Bush years, the tax cuts failed, and it was the government spending on infrastructure in Iraq (military bases) and in the Gulf coast after the hurricanes that got the economy going for a while. That plus the military spending on new equipment.
I’ve thoroughly enjoyed this thread, largely because it has stayed non-partisan. I don’t care for politics one bit – despite the strong link to economics.
As someone who works for a Wall St. firm I can say the economy was in very very bad shape. So bad that if things had gotten any worse this would all be academic. I don’t quite know how to quantify what I am saying but 15-20% would have been a certainty.
Having said that, I freely admit our models incorporated some bad assumptions, but they were the best we could do knowing what we did. We are now better equipped to understand the risks and are working furiously to update our models to better reflect reality and risks. Example; we were concerned with better ABS pricing/default models but didn’t have a full understanding of the risks they had created in the repo market. This myopic mode of operation has yet to be addressed unfortunatly. The only thing we have resolved is that we won’t make the same mistake again.
@JTC
I like that return to normal spending has not resumed – and I like to think other economists do secretly as well. Not because we favor unemployment and slow growth but because this structural change needed to happen for the long-run health of the US economy. The low personal savings rate of the mid-2000s was utterly unsustainable. We have seen that confidence is slow to return and rightfully so. Wall St hasn’t changed and unemployment is high. I think demand stimulus would be a huge mistake given those circumstances. As unemployment slowly trends down spending will continue to pick-up. Until then we must continue endure this painful adjustment. Down the road we can reap the benefits of having positive personal savings.
As far as Wall St is concerned. I’m of the opinion that the Street doesn’t trust itself, the Administration or the economy. Volume is disturbingly low. There is only one way to interrupt this signal – the future is still dark. Although I am free market at heart I don’t mind further regulation of banks and firms. A populist approach is what I fear.
Joel Prakken wrote:
For the record, last spring, as the financial crisis that engulfed the economy worsened unexpectedly — but before the stimulus could possibly have had any real effect on the economy — the unemployment rate already had moved above the Administration’s (and many others’) last pre-stimulus projection.
This could only have been written by a modern “economist.” Because Prakken did not forecast the “finacial crisis” it must have “worsened unexpectedly.” Let’s just ignore all those who were warning of impending doom. Let’s just ignore all the citizens who called congress begging them to defeat the TARP bill. Let’s just forget that a Senate slight of hand was used to even pass TARP and if the spirit of the constitution had been upheld TARP would have died a proper death. I called double digit unemployment right here in the Fall of 2008 shortly after TARP was signed by Bush.
And for some reason because businesses were laying people off before TARP money was spent somehow makes double digit unemployment okay and then because the Bush administration struck a body blow to the economy that somehow justifies the TARP.
Economics and logic certainly don’t mix.
RicardoZ,
I usually like taking a look at how “unexpected” works. Both for fun and profit, or at least for self preservation.
Both Fed and Treasury officials have remarked about how quickly Bear Sterns happened. It went something like this.
Bear borrowed in very short term repo and money markets to fund 32:1 leverage. Their risk management wizards somehow concluded it’s good to be very long in RE CDS. Somehow their books also showed that they had something like $17B in liquid short term assets.
The RE CDS markets rapidly deteriorated as everyone came to the realization that housing prices can go down and defaults can go up. Bears’ big CDS bet went sour, those on the lending side of the repo deals smelled trouble, and decided maybe they shouldn’t renew their subscription to the Bear Stearns growth story. In a matter of a couple weeks the $17B disappeared. Bear then made a panicky call to the NIH Team at the NY Fed. (NIH Team stands for Not Invented Here Team, which we shouldn’t confuse with the National Institute of Health) and frantically told them that all the money disappeared, their bankruptcy paperwork is all filled out, and if they file it will be like a tactical nuke going off in Central Park, taking all their banking buddies down with them, and this will cause a worldwide economic contagion worse than Bird Flue…BEAR FLU!!!!. Unless you guys can think of any other plans for us, of course.
We know what makes the NIH Team jump into action, and that would be it.
Armed with bucket loads of money, they successfully contained Bear Flu, and found a suitor worthy of Bears’ counterparty’s expectations. Crisis solved. Whew.
Then we got hit with plagues, famines, and pestilence of Biblical proportions.
Macro economists do this an injustice when they describe it as “unexpected”.
I also like keeping an eye out for people wearing T-shirts with popular slogans on them. Keeps me in touch with the little people, I feel. Sorry to report no T-Shirts say “If it’s not Macro it’s not Real”.
Wow, ask for a peanut gallery and PRESTO, you get a peanut gallery.
So to take them in order.
Cedric’s effort to summarize was a failure, simply wrong. Whether that was intentional on his part, or due to lack of understanding, I will leave for others to decide for themselves.
A far more accurate summary of their views would be that the program tried to fill a five gallon tank with a gallon of water. That first gallon was a start, but not big enough to fill the tank. Or, to use Cedric’s terms, the stimulus worked, but it wasn’t big enough.
MT bases his critique on personal belief. Fine and good, but to be consistent, relying on to a personal relationship with the economic gods over received wisdom from the priests requires that you allow everyone else their personal revelation, as well. MT can believe whatever he wants, but cannot expect the rest of his to care what he believes. And, oh,… what mulp said about who got the stimulus money. On the assumption that “Main Street” means common folk and not some strained “what I like” version of the common folk, then public sector workers and people on jobless benefits ARE “Main Street”.
Varones either misunderstands what the word “classic” means or doesn’t understand what “trickle-down” means. Huge salaries in the public sector? Hmm. If you ignore what “huge” means in the private sector, there are some high-end administrators who earn pretty well – low triple digit pay. They are not the ones who were due to lose jobs, though, so it is simply not true – however much it may suit MT’s agenda – to claim that public sector employees at the high end were the beneficiaries of ARRA.
bryce cheers for people who say things he likes – bryce and MT can kneel and pray together.
In response to Cedric’s response to Steve – a) Federal highway bills are notorious for prefering new construction over maintenance of existing works, and b) Drive around a bit and see if the evidence supports your theory. The evidence everywhere I’ve been is that there are plenty of potholes to fill. Beyond potholes, we know that bridge maintenance is years behind schedule. Potholes cause damage one car at a time. Bridges going out isolate large chunks of geography.
Charles, we are not in a matter of model or measure. It is model vs measure. The projection vs the data. We actually do measure the actual level of output, but cannot measure the level of output that didn’t happen because we intervene. Counterfactual analysis necessarily has that quality. We still need to be careful about which models we accept to represent the counterfactual.
David P – two answers. One is that term structure is a rough guide only to crowding out. I know that lots of supporters of increased stimulus have pointed to low rates as evidence for a lack of crowding out, but that seems an effort at grabbing the most obvious tool available. If you look at corporate debt issuance in recent months, you’ll see at least one record month for investment grade issuance, and a lot of really strong months along with it. Spec grade spreads have come in sharply, which suggests strong demand for spec paper. Just stare at the curve, and you may think you see crowding out, but look at what crowding out is meant to say, and you’ll see evidence to the contrary. Plenty of demand for private debt.
The other answer is more a matter of preferences than facts. If output drops 4% below trend and the jobless rate jumps 5%, there is a bunch of hurting going on. Amelioration seems like a good idea. Amelioration now must be paid for later. If, at that later point, output is near trend and the jobless rate is down 4%, then there is a bunch less hurting going on. I think there is an argument to be made for smoothing out economic swings that goes beyond frightened politicians striving to save their jobs.
Ricardo, contrary to the just-so story that Cedric offers, “expected” can reasonably be read as “an outcome that doesn’t change prices very much.” Yes, some folks thought things would be worse (here, I take a little bow). Others thought things would be better. Credit spreads went blooey at about the time Prakken has in mind, which is to say that there was a balance of expectations for an outcome far better than the one we got. Things were worse than what was priced in. Despite Cedric’s suggestion (to the extent I understand his story at all), credit spreads did not go back where they were for a very long time, despite buckets of money applied as a balm to the situation.
kharris,
Re: potholes
We’re in good shape here, but the city keeps a male one and a female one at the public zoo so they can mate ’em and give us more potholes.
Menzie:
Looks like I have support on less than 1.0 multiplier as well as your preference for comparison of stimulus to baseline. It is negative according to Robert Barro. See “Stimulus Evidence One Year On”. Pretty much supports everything I have claimed and does it in your preferred manner- comparison to baseline. It proves empirically what is practically known- govt spending can never be more than 1.0. It is not in the biz of making money so therefore spends inefficiently. Just go through the airport and you can experience the proof- TSA. Massive govt spending at less than 1.0. Do you disagree with his analysis?
tim kemper: Do you ever read the links I provide? See this post from December of last year, and I’ve commented on this several times since. This is old news. You should definitely read the paper. Ever wonder why the samples used never start after 1953. So I think it’s the right way of thinking about counterfactuals — I just wonder (a lot) about the robustness of the model results. Then I think back to his great error on reading employment data. I still laugh about that each time I think about it.
menzie: so let me ask you does it matter how the money is spent? If I hire private contractors to do the work or I hire union labor at much higher cost? Would that have an effect? You can’t say that if I spend 15% on higher labor for the same task I do not get the same effect. How can you then judge your multipliers the same? Spending inefficiently is still inefficient spending. you cannot get more than 1.0 mulyiplier on govt spending.
2slugsbaits: (1) how can macro be a separate discipline from micro? If micro results show one thing, shouldn’t macro find the same thing as well? They are — properly done — both part of the same enterprise: understanding the world through economic analysis.
(2) You’ve outed me on my partisanship…microeconomist suspicious that macro isn’t pulling its own weight.
Tim Kemper: If you have an NBER subscription (or an extra $5 to spare) you might want to read Woodford’s paper on the government multiplier. He basically dissects Barro’s argument. To be sure, Woodford’s paper is not about empirical estimates of the multiplier, it’s about how Barro gets the math wrong when the implied NAIRU clearing nominal Fed rate is negative. Barro’s argument about a multiplier of less than 1.00 is fine if you’re talking about an economy that is not in a liquidity trap, but it completely collapses when the nominal Fed rate is zero and there is still slack in the economy. And of course there have been several other papers over the last 6 months that have found similar results both in theory and empirically.
http://www.nber.org/papers/w15714
Of course, 2 slugbaits, we’re not in a liquidity trap, because the short-term nominal rate isn’t the only tool of monetary policy. Woodford may genuinely believe that the short-term rate is the only tool the Fed has, or he may just be using it as a modeling device, but by looking out the window you can see that there are lots of other tools at hand.
Illustration: if you think we’re in a liquidity trap, you should be perfectly willing for the Fed to sell off its trillion dollars of MBS as long as they keep the Fed Funds Rate at zero—since the liquidity-trappers must believe that the Fed’s only substantial tool is the Fed Funds Rate.
As for me, I think the short term Fed Funds Rate is only one tool of monetary policy…and of course, a crucial tool of monetary policy is the expected future path of future monetary actions, both conventional and unconventional.
Indeed, Sumnerians like myself would be happy for the Fed to loudly announce a nominal GDP target path while also creating a nominal GDP futures market, thereby spurring the Fed to engage in aggressive operations so that it meets its NGDP target.
On the academic research side, I’m a fan of monetary policy models that include actual money—such models are particularly helpful and enlightening when facing the zero nominal bound. Even better, models with money fit the data…