February 04, 2009
Why Can't We All Just Get Along? The Great Multiplier Debate
I've been thinking about why the numbers that are typically bandied about in policy circles (at least that I'm familiar with) have so little impact on the overall general and blogosphere debate (see some examples here and here). I think it's part ideological, and part methodological. I can't do much about the first (e.g., tax cuts good, spending on goods and services bad -- unless on defense; or alternatively "let the market adjust no matter how long it takes"). But at least I can lay out why reasons why there is disagreement on the size of the multipliers.
I leave aside the "timely" issue, since others have discussed it , and I think it less relevant given the likely extended duration of this recession, and the time-pattern of stimulus, as depicted in Figure 1 reproduced from this post.
Figure 1: Estimated spending and tax revenue reductions, per fiscal year, embodied in HR 1. Shaded areas pertain to spending occurring outside of the 20 month time frame. Source: CBO, Cost Estimate of HR 1 (January 27, 2009).
The starting point in the analysis is to realize that there are three key ways in which to obtain "multipliers".
- Estimation of structural macroeconometric models, with identification a la the Cowles Commission approach.
- Calibration of microfounded models (including real business cycle models, and New Keynesian dynamic stochastic general equilibrium models).
- Estimation of vector autoregressions (VARs) and associated impulse-response functions, with identification achieved by a variety of means.
Traditional macroeconometric models. Most of the estimates I have cited   are based upon the first approach. One estimates a model with many equations, including the components of aggregate demand (C, I, G, X, M), supply side (price setting, wage setting), and potential GDP. The framework most popular in policy circles is one that might be characterized as "the neoclassical synthesis", wherein wherein prices are sticky in the short run, and perfectly flexible in the long run. The OECD Interlink model is of this nature, as is Macroeconomic Advisers' model. The latter was the standard off the shelf model referred to at CEA when I was on staff in 2000-01. The Fed's FRB US model also falls in this camp. Now even within this category, there is a wide diversity of specifications in terms of the number of equations, level of disaggregation, lag length, what variables are included in each equation, etc. For instance, in the consumption function, how many lags of disposable income, is wealth included, is wealth disaggregated into housing and non-housing wealth? Does one explain aggregate consumption, or durables, nondurables and services consumption? People who blithely argue for or against a given specification with certitude are likely to have never had to face these choices. They're hard!
Perhaps the key dividing line is between models that incorporate adaptive expectations (operationally, include lags, perhaps imposing a functional form on the lag structure) and using model consistent expectations (i.e., the equations incorporate expectations of future variables, and those expectated values are calculated in a manner consistent with the model). John Taylor was a leader in incorporating model consistent expectations in macro models (as laid out in his 1993 book.
A key reason for the academic disenchantment with these types of models included the view that the identification schemes used were untenable (e.g., why is income in the consumption function but not in the investment?). Another source is the combined impact of the inflationary 1960's and 1970's, and the Lucas Critique. On the latter point, I'd point out that unless policy changes are really massive, the Lucas Critique (a.k.a. Econometric Policy Evaluation Critique) isn't really relevant(see ).
Models with micro-foundations in general equilibrium Micro-founded models are often associated with real business cycle models. However, the association is not one-for-one. It's true the early real business cycle models worked off of utility functions and production functions. But the modern generation of dynamic stochastic general equilibrium (DSGE) models in the new Keynesian mode incorporate microfoundations as well (utility functions, production functions, investment functions, etc.) but also incorporate rigidities such as price stickiness. Purists will say everything has to be microfounded. Well, that's a matter of taste, but the fact of the matter is that it's very hard to calibrate simple real business cycle models without rigidities to match the moments of actual real world data, even after the data's been HP-filtered (I'm sure this blanket statement will get me in trouble, but I think that that's a fair assessment). So DSGEs do better at mimicking real data, especially after numerous rigidities are incorporated. In the earliest incarnation of the Fed's Sigma model, for instance, there are rule of thumb consumers (shades of Campbell-Mankiw!). For a survey of how DSGEs have been incorporated into policy analysis, see the survey by UW PhD Camilo Tovar.
It's useful at this point to ask how are these models calibrated? For the deep parameters (intertemporal rate of substitution, for instance), one can rely upon some estimates -- then pick the one that you like (and is in the range of estimates). Oftentime, the combination of parameter values is selected to mimic the time series properties of actual (filtered) data. So say one believes one should not appeal to ad hoc Keynesian models. It's not clear that RBCs or DSGEs get you away from the problem that one has to appeal to the data to get multipliers since the models are calibrated to mimic real world data. In other words, while the theoretical bases of the models may differ, and important insights can be gleaned from these models (e.g., distinctions between temporary and permanent changes), the differences in terms of multipliers might not be as big as one might think.
VARs Vector autoregressions are regressions of multiple variables on lags of themselves. The underlying shocks can be identified by putting them in a recursive ordering (called a Cholesky decomposition), or using restrictions based on theory (say, money has no contemporaneous impact on prices; or money has no impact on output in the long run). VARs were initially proposed as a way of getting around "incredible identifying assumptions", in the Cowles Commission approach to econometrics embodied in the old style macroeconometric models. But of course, people can disagree about which restrictions make the most economic sense. (For instance money is neutral in the long run seems natural, but not all theoretical models have that implication.)
The much cited Romer and Romer model of fiscal policy impacts is a particular sort of VAR, in which only one equation is focused on, and extra-model information is used to identify exogenous tax changes (remember, they don't analyze government spending changes). (It's an autoregression in log GDP, to the extent that the dependent variable is log first differenced GDP).
A good summary of where these types of fiscal multipliers come from was in Box 2.1 in Chaper 2 of the April 2008 World Economic Outlook.
My bottom line There are indeed a wide variety of estimates regarding the size of multipliers. Different models -- and assumptions within those model categories -- lead to different estimates. It's important to understand the underpinnings of those estimates (and this is where many of the people who cited the Romer and Romer study went wrong). Hence, one has to have an understanding of the very complicated models before taking strong stands in favor of one estime over another.
In my experience, as far as policy organizations such as central banks, government agencies and multilateral agencies go, reference is made to a number of models. Their assessments of multiplier magnitudes will then reflect some weighting of the various model predictions. That is why I will put more wieght upon assessments by organizations (that have to make decisions upon these judgments) than a single academic study, regardless of how well I respect the academics involved (and sometimes, these academics are working outside their area of research expertise...)
As an aside, here are the impacts of various fiscal experiments in response to a negative shock in a DSGE developed by the IMF:
Figure 1: Figure from Box 2.1 in Chaper 2 of the April 2008 World Economic Outlook.
"Expansion through transfers is defined as a one percentage point increase in debt-financed transfers in year one and 0.5 percentage point in year two. Expansion through labor tax cuts is defined as a reduction in the labor income tax rate by 1.5 and 0.75 percentage points in year one and year two. Expansion through government investment is defined as a combination of higher transfers and an increase in productive government investment by 0.25 and 0.125 percent of GDP in year one and year two."
Notice that the public investment shows the biggest impact, while under the base assumptions the impact of transfers and tax cuts are about the same.
Posted by Menzie Chinn at February 4, 2009 07:34 PMdigg this | reddit
Professor, here is why we hoi polloi do not trust estimates from the brain trust:
The Fed, with its high-powered economists, failed to properly estimate losses on a portfolio of loans it recently bought. Yet, you want us to take as gospel 'assessments from organizations' on something much, much more complicated?
No way, sir.
Posted by: jg at February 4, 2009 08:54 PM
Am I reading this right? A roughly 1% of GDP money financed discretionary increase in government investment will cause a roughly 0.6% of GDP increase in real GDP in year one relative to what would have happened otherwise? So the biggest multiplier is about 0.6?
Posted by: Nick Rowe at February 5, 2009 05:18 AM
The model that works for me is the simplist of all models: govt deficit=private sector surplus. The private sector requires a surplus for the economy to grow. History proves this model is correct. Short term, the govt ran a surplus from 1926-1929 and 1999-2001. The decades following both periods were an economic mess. The 20 years after the end of WWII the govt deficit was too small to keep up with economic growth. The debt to gdp ratio fell to an unacceptable level (below 40% and we ended up with the economic mess of the 1970s.
We got all these Ivy league professors running the economy with all their fancy models and data; we still end up with a mess. They need to take a step back and KISS it - Keep It Simple Stupid!
Posted by: Anonymous at February 5, 2009 06:42 AM
Good post. I would encourage you to develop these blog posts on multipliers into a researh commentary.
My only comment is to remind you that these are just models. The devil is in the assumptions and the evolutionary nature of rational expectations.
Here is what a physicist turned banker has to say about models:
Posted by: MikeR at February 5, 2009 06:44 AM
Whenever I read economists using phrases like "standard" or "off-the-shelf" models, alarm bells start ringing. It means that fundamental parts of such models are now rarely reviewed. Clearly, "sticky" prices is a key property of new Keynesian models, and I wonder what the empirical evidence for this is. The Taylor and Calvo price adjustment models typically incorporated in macro models strike me as dangerously unrealistic.
And when these models are built, they are usually impossible to solve analytically, so some approximation like "linearisation around the steady state" is applied, which has me wondering how much of the value of the economic modelling is effectively thrown away when the model is actually used.
Posted by: RebelEconomist at February 5, 2009 06:58 AM
I have to agree with jg above: There is little reason to better trust organizations--especially government organizations--than academics. First, and if for no other reason, they have a bureaucratic imperative to protect. Once they take a position (right or wrong objectively), they are highly unlikely to change it and highly likely to manipulate data to justify their actions and intentions. Second, on a broader level, they are driven by the political ideology of their leadership, starting with the President. Certainly no one saw a strong interventionist policy in the Bush years--in part because of a view that stimulus was ineffective--and we are seeing a strong interventionist policy in the Obama years, because they believe (and will create analyses to show)that stimulus will have an important positive effect.
For these reasons, I believe that the work of academics is especially important. They may have their own agendas, but the rigor of their research--subject to peer review--usually teases out the biases if they are not already well known.
Posted by: Terry at February 5, 2009 07:50 AM
Good article, thanks
Posted by: Beal at February 5, 2009 09:39 AM
Here is a Rube Goldberg contraption for why model estimates may not be reliable:
Asian economies have become enamored of the export-led growth strategy. When the dotcom bust ended a period of strong U.S. investment, growth and trade deficits, Asian countries dramatically expanded currency interventions. From 2002 to 2008, Japan and China bought more than $1.2 trillion in foreign reserves. (This doesn't count effects of implicit currency policy on the yen carry trade.) It seems to me that this response should reduce the U.S. fiscal multiplier, but I don't think it is in your models. What is the implied multiplier from the Bush fiscal deficits? That would probably be a better predictor for an Obama fiscal deficit multiplier, unless the U.S. gets serious about foreign currency mercantilism.
For exmple, I would like to see various scenarios for the multiplier based on the response of foreign official purchases of U.S. debt.
Posted by: don at February 5, 2009 09:52 AM
A question (I have some econ training, but don't know much about macro-econometrics):
You links to a paper  that claims: "Virtually no evidence exists
that empirically substantiates the Lucas critique. Empirical refutation of
the Lucas critique by using tests of super exogeneity is illustrated with
U.K. money demand. ..."
How can the Lucas critique be refuted? In order to do econometrics you have to pose a specific model and the implications of the analysis are only accurate if the model is "true" (in the Platonic sense). Thus the term "empirical refutation" seems to me to be an oxymoron. The act of refutation is based on a specific modelling framework which may itself be wrong.
I can understand the authors saying, nobody has yet found a modelling framework that is consistent with the Lucas critique (or there's something wrong with the models that are consistent with it) and this makes us doubt it's validity, but I don't understand how they can claim that it has been "empirically refuted".
Posted by: AnonCA at February 5, 2009 10:16 AM
jg: Gospel, no. Nothing should be taken for gospel. But just as survey averages are better at tracking macro aggregates, so too might one suspect an average of sensible models would do better than reliance on a single model.
Nick Rowe: No, I'm afraid you're not reading correctly. The deviations from baseline (black) are for different types of stimuli. They are for 0.25 and 0.125 ppts of GDP, so the cumulative output multiplier on cumulative impulse is pretty large (about 1.2 divided by 0.38).
Anonymous: That looks like an identity, and in fact an identity for a closed economy, not a model.
MikeR: Thanks for the link. Not sure it's worth half a CE credit, but surely entertaining.
RebelEconomist: Well, I'd say there is little doubt that in certain environments prices are sticky. The question whether time-dependence or state-dependence is the proper approach is, in my mind, the relevant one.
Terry: Don't get me wrong, I am all in favor of paying attention to academics. But when one has make big decisions with billions of dollars, I want to do a little hedging.
don: If the multiplier is a transformation of the marginal propensity to consume, and suppose the simulation that gives me the implied multiplier relies upon an equation in which wealth and interest rates are incorporated either directly (in the equation) or indirectly in other equations, then I think your effects are taken into account.
AnonCA: You'll have to take up the argument with Ericsson. My take is this: adaptive expectations is not a bad assumption when the policy regime is fairly stable. Then the question is how often do we have drastic changes in the policy regime, such that the coefficient estimates (which are in these formulations convolutions of the underlying MPC and the autoregressive process governing expectations formulation) change substantially. I'm open to alternative views here -- it may be that we are in a drastically new policy regime. But then, one will have to take a stand on what is a better estimate. In other words, criticism is not -- in this situation where policies have to be undertaken -- sufficient. A constructive, specific, alternative is necessary.
Posted by: Menzie Chinn at February 5, 2009 12:14 PM
Menzie, You said my equation was an identity. First, is it a correct identity? And second, the global economy is a closed economy! People around the world get up and go to work for the same reason we do; for money (govt debt if my identity is correct).
Posted by: Anonymous at February 5, 2009 01:06 PM
Thanks. Time-dependence vs state-dependence was exactly what I had in mind. I got the impression that Taylor and Calvo are used because they give dynamics of a type that modellers are used to dealing with, rather than because they are realistic. It's been a few years since I was around people building such models. Are there any (published?) models that use state dependent price adjustment?
Posted by: RebelEconomist at February 5, 2009 01:14 PM
RebelEconomist, I share the concerns raised in your first comment. The Capital Asset Pricing model is the most prestigous theory in finance. Due to the avaiability of vast amounts of data, the theory can "almost" be tested and it fails miserably.
Economists really can't test their predictions, so they make assumptions which they are forced to ignore. Does anyone know what affect more government spending would have had in 1932? We will never know. No one behaves according to a linear model with a constant MPC = .7
Posted by: MikeR at February 5, 2009 02:02 PM
I'm not a macroeconomist but studied some of this stuff once upon a lifetime.
Can/do these models differentiate the way stimulus is implemented? Or is spending just spending and a tax cut just a tax cut?
I ask because I would think the details would matter a lot. Like a tax credit for new investment spending differs from a corporate tax rate cut which differs from an income tax break which differs from a payroll tax break. Or like government spending on infrastructure development differs from extending unemployment benefits which differs from subsidizing car purchases which differs from subsidizing the wages of all new hires by firms.
It does seem, to me anyway, that too much of the discussion is on tax cuts versus spending. It seems more relevant to talk about the details of how these things would be implemented.
Indeed, I can imaging tax-cut/spending alternatives that, in effect, would seem nearly identical.
Don't you think spending and/or tax cuts, if done creatively, might induce much larger multipliers that historical data/models would indicate? Sadly, developing evidence for such creative policies, no matter how conceptually compelling, would seem difficult, given paucity or nonexistent data and the large aggregates used in macro analysis. But please do educate me if I'm wrong and you are so compelled.
Very nice post, by the way.
Posted by: michael roberts at February 5, 2009 02:04 PM
Does it matter what things we spend on? If spending reinforces or adds to structural imbalances, does this not simply deepen the problem? Spending to keep zombie firms alive might lessen current unemployment, but "letting the market adjust" seems called for, doesn't it?
This isn't ideology. It's hard for me to see how "stimulus" does anything to correct the fundamental misallocations of capital that have been building up for years.
Posted by: Charles N. Steele at February 5, 2009 03:17 PM
"Hence, one has to have an understanding of the very complicated models before taking strong stands in favor of one estime over another."
That sure makes it challenging from a public policy perspective.
Posted by: Anonymous at February 5, 2009 05:23 PM
Anonymous: C+S+T ≡ GDP ≡ C + I + G + X - M; Rearranging leads to (S-I) + (T-G) ≡ X-M ≡ Trade Balance. Set the trade balance always equal to zero, then (S-I) ≡ (T-G).
RebelEconomist: I'm not an expert on this particular literature, but see Chapter 2 in this NBER volume and associated comments, especially by Cogley.
michael roberts: Some models would make the distinctions you do highlight, but the more formal general equilibrium models would probably not be able to handle these fine distinctions.
Charles N. Steele: Sure it matters what we spend on, in terms of how stimulative of aggregate demand, in terms of how much it changes potential output in the future, and so forth. But I don't think anybody is saying "don't let the market adjust". Rather, it's trying to mitigate overshooting. Markets sometimes cease functioning, including financial markets. By the way, you've just propounded the "liquidationist view" popular in the Hoover era.
Anonymous (5:23pm): Yes, I agree -- making good policy is hard.
Posted by: Menzie Chinn at February 5, 2009 05:45 PM
What I can't understand is how economists can expect any form of stimulation, be it tax breaks or handouts, to give such high multipliers? I have at least 3 reasons to be suspicious
1. The marginal productivity of debt has been declining for a while now. Data that measures the annual increase of total debt divided by the annual increase in total National Income for the USA suggests it has declined from 0.53 in 1957 to 0.21 in 2007. $5.50 of new debt was required to generate $1 of incrimental GDP in 2007 - so getting more than $1 of incrimental GDP for $1 of any other form of incrimental money makes me really wonder why this form of stimulation can be more than 5 times as effective as current debt (http://mwhodges.home.att.net/nat-debt/debt-nat-a.htm)
2. The US economy has hollowed out substantially and is now dominated by out of town big box stores and multinationals who import what used to be manufactured domestically - this means that money quickly escapes any local community relative to the days when local businesses were dominant - hence money's multiplier effect is reduced and the whole economy loses resilience
3. Consumer spending is largely dependant of the wealth effect. Consumers have collectively lost over $12T in paper wealth on the houses and shares since 2006. Their savings rates have been around zero and the equity they hold in the homes has declined from around 85% to 45%. A chart from the White House that tracks real wealth against consumer spending suggests that consumer spending is set to fall much further (http://economix.blogs.nytimes.com/2009/01/16/the-white-house-view-of-the-economy/)
I find it difficult to believe that extra dollars in consumer's hands will do much other that reduce debt in these circumstances - especially given that household debt has doubled from 50% of GDP to 100% of GDP since 1980
Posted by: Alan McCrindle at February 5, 2009 06:43 PM
I don't think that debt generates income, as Alan McCrindle says. Debt can move income forward from the time the debt is paid off to the present (assuming the debt is paid). It is the production of goods and services that enables the debt to be paid off that generates income. When we have debt that is not paid off, the system collapses, as we are witnessing.
On the other hand, his second point is excellent and I second his views.
Consumer spending does seem to be influenced by the wealth effect, in part, but how much of the wealth effect is forward anticipation of a bright future which is produced by wealth gained today? Also works backwards on the way down. Less consumption today because of anticipated bad tomorrows.
I certainly agree with his main conclusion. Hold the stimulus for a while. I would say hold the stimulus until we recognize that bankruptcy is the best way to deal with the bad debts that the private sector cannot pay. We flee from reality. Elimination of the belief that the stimilus will redirect the U.S. economy apparently is required before reality will guide public decisions.
Posted by: ReformerRay at February 5, 2009 07:55 PM
Why can't we get along? Because of garbage like this:
Without swift passage of his stimulus bill, Obama said, "an economy that is already in crisis will be faced with catastrophe."
As for the Senate taking time and debating the issues, POTUS says: "The time for talk is over. The time for action is now."
Ah yes; the best and the brightest resorting to fear mongering.
While in my state we have great stewards of the citizenry like Senate Democratic Whip Dick Durbin of Illinois. He's warned against cuts: "Every time we lop off $100 billion, we're lopping off jobs," he said.
Hey Dick Durbin: prove that for every 100 billion cut jobs are lost.
Good 'ole fashioned fear mongering a la Al 'Climate Change' Gore.
And when it doesn't work what then?
Posted by: Babinich at February 6, 2009 02:46 AM
Sorry Menzie - I am "Anonymous (5:23pm)". Keyboard pilot error.
Policy is especially hard in this case because it appears you have to be a technical expert to judge the relative effectiveness of policies, *and* the policy prescriptions change radically based on the estimates. By contrast, most customers don't need to be experts in semiconductor physics to choose between different makers of processors for their PCs.
Posted by: Robert Bell at February 6, 2009 04:17 AM
Babinich: In the land of hyperbole, how do those examples you provide rank against "...smoking gun to be a mushroom cloud"? That being said, I'm not sure I disagree with the quote from POTUS. In fact, I don't disagree. So am I engaged in scaremongering if I say something I believe in?
Posted by: Menzie Chinn at February 6, 2009 07:34 AM
Chinn says the time for action is now. I say no time is the right time to take on more debt that you have no way to pay back...
Debt that cannot be paid back got us in this mess. More of the same will not get us out.
When the money generated by assuming a debt is used to buy something that will increase productivity, that is a good debt. Increased productivity will generate the wealth to pay back the debt.
The current stimilus package is not crafted to pick up the economy and generate wealth after the current economy has collapsed. Instead, it is intended to prevent the collapse. That is why the current bill should be voted down.
Can't we just keep our powder dry until the right time?
Posted by: ReformerRay at February 6, 2009 08:15 AM
Menzie, thanks for responding.
I am still puzzled about the ability of the model to predict under my scenario, though. If the muliplier depends on the response of currency interventions (as I hypothesized), you would need to argue that the interventions (or at least reponses of the trade balance) have been so regular in the past that their effects have been incorporated in the past multiplier responses. But I think there has been a sea change in the magnitude of the interventions since about 2002. In any event, wouldn't the model need to incorporate a host of international variables, such as incomes abroad, in order to capture the effects of responses in the external balance?
Posted by: don at February 6, 2009 09:40 AM
I am inclined to agree with Ray. I have been convinced that stimulus can work, but, when you cut through all the confusing economic models and jargon to the common sense, it seems to me that stimulus works because the state commands resources, either by borrowing from foreigners or exploiting its own population's money illusion (ie sticky prices), which it uses to make people do work that they would not otherwise do - like a fish to a performing seal. The problem is that, unless the work that is done regenerates the resources, or some other reason to work arises, the stimulus only boosts activity as long as it lasts, and quite possibly makes the situation worse. What I would want to ask about the stimulus plan is (a) how it can generate resources to pay for itself and (b) what is supposed to happen when it finishes.
Posted by: RebelEconomist at February 6, 2009 10:18 AM
Macroeconomists are supposed to be taken seriously when they're still discussing MULTIPLIERS? The process works by selling a bond (whether to a US or foreign citizen) so that a deficit may be run and this somehow does not reduce private spending by an equal amount. Government now engages in spending, conducted by government employees, all at zero cost. Finally, resources that would have been directed to their highest valued uses by optimizing individuals in the private sector are re-directed to second, third or nth-best uses in these various government-run stimulus programs. Tell me, if there IS a multiplier, why it isn't NEGATIVE?
Posted by: Huh? at February 6, 2009 11:24 AM
Why aren't all those multipliers negative? Because they are not, as the historical evidence shows. Even Barro's downbeat estimate for military spending in WW II when we were at overfull employment with rationing, still gave a 0.8 mulitplier, definitel positive, if in that one case, not greater than one, as most estimates for most other spending multipliers are (although not all tax cut multipliers in the short run).
To parse your argument point by point:
1) No, it does not reduce private spending today by the same amount because it is borrowed from the future. Yes, there may be some reduction in the future, but if the economy goes into a long recession, there will be less in the future anyway.
2) Of course there is a cost to the spending. The direct "cost" equals the spending, but so what?
3) As for "highest and best use," clearly what the money is spent on is important. US private citizens have mostly been consuming short term stuff and not saving to carry out investments. If this spending is on productive infrastructure, it may not be nearly as wasteful as you imply.
Posted by: Barkley Rosser at February 6, 2009 11:42 AM
I should have mentioned in my original post that a multiplier of 0.6 would make the same qualitative point as one estimated to be negative: For one dollar of deficit spending, we get 60 cents worth of stuff in return. Perhaps we are acting like the firm that is losing money on every unit but hopes to "make it up on volume?"
Posted by: Huh? at February 6, 2009 03:32 PM
"I'm not sure I disagree with the quote from POTUS. In fact, I don't disagree. So am I engaged in scaremongering if I say something I believe in?"
Good for you... Time will prove one of us wrong.
Posted by: Babinich at February 6, 2009 06:45 PM
"This recession might linger for years. Our economy will lose 5 million more jobs. Unemployment will approach double digits. Our nation will sink deeper into a crisis that, at some point, we may not be able to reverse"
Enough already... It looks like POTUS has a little Al Gore in him.
Will Wilkinson hits a home run:
Posted by: Babinich at February 6, 2009 07:53 PM
By the way, thanks for the NBER reference, Menzie.
Posted by: RebelEconomist at February 7, 2009 01:33 AM
Barkley Rosser, others,
There is no reason to think that there will be any cost, now or in the future, financial or real, to a stimulus plan.
If the stimulus causes unemployed or underemployed people to do work that would never have been done anyway then everybody, us and all our grandchildren, will be better off as a result.
Given the lowering price of cars and the rotting stock of cars out there, I'd say there are a lot of people who'd snap up a car at the bargain price now if only they had a job and/or some more job security. Newly reemployed people are likely to spend much of their new money on hiring recently unemployed people as it is those areas of the economy that'll have seen recent price falls and whose products will seem particularly tempting.
So the stimulus can easily pay for itself.
Posted by: Aaron McDaid at February 7, 2009 02:28 PM
"Set the trade balance always equal to zero, then (S-I) ≡ (T-G)."
Is that correct or did you mean "Set the current account always equal to zero"?
Posted by: Anonymous at February 7, 2009 04:35 PM
Peter Schaeffer: Since I have suppressed for simplicity of notation net foreign income and transfers to zero, the trade balance and current account are the same.
Posted by: Menzie Chinn at February 7, 2009 07:40 PM
Thank you for the excellent discussion and for your exceptional patience with the blogsphere. I'm amazed, but always elucidated, by your balanced responses to what are frequently simply rants.
Posted by: rf at February 7, 2009 10:18 PM
I concur, thanks for your discussion.
Your point that what we spend on matters for potential output in the future is the essence of my question.
In defense of the "liquidationist" view, if there had not been serious (mis)allocations of capital to low productivity activities, there'd be no economic debacle. The structural adjustment to correct this is painful. The appropriate policy response isn't boosting of AD, but rather speeding the adjustment. Stimulus hampers this if it keeps afloat non-performing firms and finances low-value projects. This indeed is "not letting the market adjust."
This is a very different issue from whether government spending on certain things might be warranted (e.g. safety nets for workers in failing industries, extraordinary procedures for recapitalizing banks or restructuring home loans, or rebuilding decrepit infrastructure).
I.e. are we spending to boost aggregate demand, or productivity?
Posted by: Charles N. Steele at February 8, 2009 10:29 AM
Hi Menzie, Sorry to be slow getting back to you. You asked for empirical numbers to bolster my arguments. I found Greg Mankiw's blog offered quite a few empirical studies consistent with my observations. I discuss this, and give the various citations in my latest blog post here:
BTW, it is not at all clear to me that Zandi's multipliers are at all empirical. It is hard to tell from his testimony, but i get the impression they are based on a model, and it seems Mankiw is right that there is a flaw in textbook stimulus models based on empirical results.
Posted by: Rod Richardson at February 8, 2009 10:35 PM
Rod Richardson: Sorry, but the post you sent me cites me a 1.4 multiplier for government spending, not too far off of the estimates I cited. In addition, as has been pointed out in various pieces of commentary (including by Christina Romer's colleague, Brad Delong) the usual reading from critics of the stimulus bill is incorrect in terms of the impact of tax cuts relative to spending. Apparently, many readers did not read past the abstract. So, I regret your rejoinder is for me utterly unconvincing.
Posted by: Menzie Chinn at February 9, 2009 12:08 PM