A "Why Fiscal Policy Won't Work" Contest

I am opening a contest. Let's see who can come up with the most plausible (or least implausible) model in which fiscal policy would not work in present circumstances.

This is not an entirely frivolous exercise. David Andolfatto (who runs another excellent Canadian economics blog) has made a number of posts recently questioning the empirical evidence in favour of fiscal policy. I said in a comment that my main reason for thinking that fiscal policy could (probably) work is that I find the models which say that fiscal policy can work more plausible (or less implausible) than any model I can think of in which fiscal policy would not work.

So, in my view, it's not a question of the absolute evidence in favour of fiscal policy working; what matters is the relative plausibility of models where it works to models where it doesn't work. Even if both classes of models had weak supporting evidence, I would lean towards whichever was less weak.

And David encourages me to run this contest.

So, who can think up the most plausible, or least implausible, model in which fiscal policy would not work?

Some ground rules:

1. "Model" can mean almost anything. Equations, graphs, a verbal explanation, whatever. As long as it is plausible.

2. "Plausible" can mean almost anything. Direct econometric evidence, fits the historical narrative, assumptions make sense, whatever you want to try to argue.

3. "Would not work" means there would not be an increase in output and employment, or that if output and employment did increase it would not be a desirable increase.

4. Closed economy. Both for simplicity, and because it's too easy to get fiscal policy to fail with flexible exchange rates.

5. Hold monetary policy constant. That means either fixed money supply or fixed nominal interest rate (your choice). Both for simplicity (because we don't want to get side-tracked into whether monetary policy can work), and because it's too easy to argue that fiscal policy would not work if the central bank is already successfully targeting (say) nominal GDP (sorry Scott, but we can't make this too easy).

6. You can bend any of these rules if you want to, but no funny stuff.

7. Contestants are not required actually to believe their models.

8. It doesn't have to be your own original model.

Here's my entry. It's the least implausible model I can think of.

Start with a basic textbook ISLM with sticky prices, with the initial equilibrium with output lower than LRAS. A bond-financed increase in government spending will mean the debt will increase. Assume that this increased debt will eventually lead to higher taxes, and a large part of those higher taxes will take the form of higher tax rates on the returns to investment. Assume that this is also what investors expect will happen. So the fiscal policy reduces the expected profitability of investment, so investment demand falls. Assume that investment demand is very elastic with respect to the profitability of investment. So investment demand falls by more than government demand increases. So when government expenditure increases, the IS curve shifts left, the AD curve shifts left, and real output and employment fall.

Note, this is not Ricardian Equivalence. Ricardian Equivalence works via the effect of future tax revenue on permanent disposable income on current consumption. I am talking about the effect of future tax rates on current investment.

That's my best shot. What's yours?

Comments

[Mattyoung]

Low demand for government goods because of customer inactivity. Government goods will be warehoused, causing an inventory glut of government goods as we trade excess savings into excess government inventory buildup.

[Adam P]

Here's mine in a sentance. Government makes poor investments that don't contribute to productivity growth. The story goes like this:

1) previous high borrowing was based on expectations of high productivity growth.
2) something caused those expectations to be reduced, perhaps just a bit at first(say an commidity price rally)
3) resulting slump caused an asset price decline that broke the banking systems balance sheet
4) clogged up banking system interrupts real investment and further reduces expected productivity growth which further reduces asset prices... continue loop

Now the government comes in, sucks the capital back out of the clogged pipes of the financial system and spends it building bridges to nowhere. This continues past the point where private investment would otherwise pick up and thus replaces it because the resources are in use and thus unavailable to the private sector.

Expected productivity growth is further marked down due to the government spending replacing private investment and so any income generated by government expenditures is just saved to help substitute for the lost income growth in future consumption...

[Niklas Blanchard]

If you hold the monetary base fixed, then you have to assume commodity backing (with long-term deflationary pressure)...and if you hold the nominal interest rate fixed, then (assuming we're talking about a recession), you have to fix it at zero -- and you end up in a classic Keynesian liquidity trap.

With asset prices dropping, cash would become a solid investment, which would increase the real savings rate indefinitely. Under such assumptions, you would almost have to have fiscal policy as monetary policy would be completely impotent to increase AD and/or inflation expectations.

The government could come in with a bond offering at or near cash rates to finance debt spending, which WOULD increase inflation expectations, all else remaining equal, and devalue the currency.

So I think, under those rigid assumptions about monetary policy, you would have to have fiscal policy. Unless you were allowed a one-time devaluation of currency by resetting the commodity ratio...which would have the same effect as a bond purchasing program. I may be wrong, though.

[Trevor]

Here's my best shot: Fiscal policy induces factor market misallocations. The resulting unequal marginal products across firms/sectors/regions lowers aggregate productivity and, therefore, GDP.

To see this, note some necessary features of any government fiscal policy: (1) Disbursements are not equal across firms, sectors, or regions; and (2) Tax burdens (or present values of future tax burdens required due to current debt financing) are not equal across firms, sectors, or regions

So, given the heterogeneous nature of an economy's capital stock and labour force, this differential treatment between different parts of the economy will reallocate factors between these parts (towards recipient - politically favoured? - sectors). To the extent that this reallocation results in differences in marginal products between these sectors - and it will since the reallocation is induced by taxation/subsidies implicit in government purchases - deadweight loses will occur.

[Trevor]

Is this effect empirically important? I believe so,

Hsieh and Klenow (2008), "Missallocation and Manufacturing TFP in China and India" shows that factor market misallocation and unequal marginal products are *huge*. They conclude that efficient allocations would more than double the aggregate productivity of these nations.

Restuccia and Rogerson (2008), "Policy Distortions and Aggregate Productivity with Heterogeneous Establishments" shows in a model simulation that even small distortions result in "large" effects. For instance, with a tax of 10% applied to only 10% of firms (the more productive firms) the resulting reallocation could shrink TFP and GDP by 3-percentage-points. This 3-percentage-point effect is approximately as large as the current recession!

My own work (part of my thesis, though still very much tentative) can show for early periods in the United States that eliminating sectoral and regional (county-level) distortions in factor allocations could more than double GDP and TFP.

So, in conclusion, models that suggest fiscal policy will work normally treat economic aggregates as homogeneous blobs. If we consider reallocation between heterogeneous parts of the economy, then fiscal policy's creating an inefficient resource allocation (within the factor blobs but between productive units) may shrink output and, therefore, undo any positive effect fiscal policy might have.

[Mike Moffatt]

My entries here:

What Are The Key Ingredients of Fiscal Stimulus.

Fiscal Stimulus - Unlikely To Work in the Real World.

Shorter version - *all* fiscal stimulus models are of the 'assume a can opener' variety. They work great on a blackboard only because they ignore all real-world considerations.

The issue is largely one of timing:

1. The government has to recognize that the economy is in recession. This takes longer than one might imagine - the National Bureau of Economic Research did not recognize that the United States was in a recession until 12 months after it started.

2. The government needs to develop a stimulus package.

3. The stimulus bill needs to be made law and pass all the necessary checks and balances.

4. The projects involved in the stimulus package need to be started. There may be delays in this step, particularly if the project involves the building of physical infrastructure. Environmental assessments need to be completed, private sector contractors need to bid on the project, workers need to be hired. All of this takes time.

5. The projects, ideally, need to be completed. If they are not completed before the economy fully recovers, then we will certainly have crowding out as these employees and equipment would be of use to the private sector.

All of these items need to happen in the window of, at best, 24 months. By the time a democratic government can complete all 5 steps, the economy will already be in a recovery stage.

[Mike Moffatt]

Weird.. links didn't show up right.. they should be:

http://economics.about.com/od/fiscalpolicy/a/fiscal_stimulus.htm

and

http://economics.about.com/od/fiscalpolicy/a/fiscal_failure.htm

[Adam P]

I gleefully look forward to the day when Canadian bonds are the only soverigns in the world rated AAA.

[David]

Nick:

My contribution consists of 4 pages, and so to economize on space, I provide a link to it here:

http://www.sfu.ca/~dandolfa/fiscal%20policy%202009.pdf

David

[Patrick]

David: "The boom which began in the early 2000s was fueled by technological
innovation, as evidenced by rising productivity ..."

For your (re)consideration:

http://krugman.blogs.nytimes.com/2009/04/16/reconsidering-a-miracle/

http://www.cepr.net/index.php/press-releases/press-releases/u.s.-productivity-growth-still-trails-europe/

Not a criticism, just trying to reconcile conflicting stories from people who are probably smarter than I am.

[David]

Patrick:

Thanks for the links. It's good to see that Krugman is taken seriously some problems with measurement. Now, if he would only take the same sober approach when evaluating the measured contributions of government purchases to GDP, he will finally have evolved into a serious macroeconomist! ;-)

The second link appears to argue that productivity growth in the U.S. lagged that of Europe. This is fine for my story.

Naturally, I think that there was more to the story than simple productivity growth. I think that the expansion of international trade had a lot to do with it too. But Nick asked us to limit our theory to a closed economy analysis.

Thanks,
DA

[Stephen Gordon]

Actually, it's the closed economy part that excludes me from this contest. My story of the recession is based on a terms of trade shock.

[pointbite]

Government steals money from bankrupt people to spend on frivolous projects (it's difficult to see how this scenario could be anything other than a disaster in the long term). The people lose their savings, the government destroys its balance sheet. The longer businesses that could never survive without government inflation (phony unsustainable demand) continue to consume resources (at the expense of willing entrepreneurs priced out of the market) the longer the slump will persist. Eventually you might get lucky and someone will invent some novel mass production techniques or the transistor, but luck is hardly an economic policy. In my model, there is a direct correlation between more government spending and more losses. The more they try to help, the more they impede the necessary re-allocation of capital to more productive people and businesses.

[Bruce Wilder]

I'd like to nominate Mark Thoma's "icy hill" metaphor, as a model of stimulus failure -- failure of the too little, too late variety.

MT: "I think the stimulus package is like driving up an icy hill. If you don't have enough momentum from the start and fail to provide enough "stimulus" to get the car over the crest of the hill, you can slide all the way back to the bottom, crashing into things along the way and ending up worse off than when you started. Maybe you can give it more gas along the way if needed without spinning out, and perhaps you can hold your position if you don't make it to the top, and then start again from the higher level, but that's not a chance I want to take when I'm sitting at the bottom wondering if I can make it to the top without wrecking my car -- the possibility of falling all the way back to the bottom and ending up worse off would make me want to start with sufficient momentum and then some. Essentially, I am arguing that there are crucial economic and psychological "tipping points" that must be reached in order for the economic recovery package to be effective (or at least, there's enough of a chance that they exist that they cannot be ignored when formulating robust policy)."

I've been getting some static, for being too easily exasperated by those, who seem to me to be missing "the obvious". "The obvious", which I think a critic ought to confront, is the present state of the economy. The state of the economy at the time a policy is executed will go a long way to determining the consequences of the policy.

I don't know if "plausibility of the model" covers having a reasonable analytical judgment about the specific state of the economy. (It's not hard to come up with a fiscal stimulus headed to disaster in some alternative universe of boom conditions.)

Mike Moffat, above, seems to have a perfectly sensible case for a "failed" stimulus. Moffat's failure is "too much, too late". Thoma's "icy hill" metaphor suggests the possibility of "too little, too late". In Moffat's scenario, the economy experiences a spontaneous recovery that pre-empts the circumstances in which a fiscal stimulus would be appropriate; in Thoma's scenario, the fiscal stimulus has some positive effects, but fails to surmount psychological or economic thresholds necessary to seeing the economy return to self-sustained growth toward, more or less, full-employment. The fiscal stimulus might not be of sufficient magnitude to relieve deflationary pressure and give conventional monetary policy renewed potence; failure to reach that threshold will leave the economy to stagnation/slow growth at high unemployment, or even risk another recession.

I like Moffat's scenario, but I don't think the present U.S. economy is likely to spontaneously recover fast enough to negate beneficial effects from the planned stimulus, or even from a much larger stimulus. That's a judgment about the state of the world and its immediate prospects.

[pointbite]

I want to expand on my model a bit more. The belief in government stimulus requires us to believe the previous level of demand is sustainable, desirable and not declining for some fundamental reason. That isn't necessarily a question of economics, you can't plot values on a chart. If people were comfortable in previous decades living with more debt-based consumption and a bias towards equities instead of savings, but the culture changes and people are no longer comfortable with that formula, the economy has changed for a very real and fundamental reason. If people want less, why should the government force feed us things we don't want? If people are determined to save and the government threatens to destroy savings (with negative interest rates or some other stupid ideas) people will simply find alternate means to save. Maybe they will hoard food or gold and cause other problems like shortages and hunger.

In order to understand how I see the world, you must come to the conclusion (as I have) that many businesses currently "idle" should never have existed in the first place, they only exist because demand has been skewed by years (or decades) of government interference and stimulus of other forms (including monetary). Each time the government intervenes these businesses become more convinced of their soundness, but in reality it was always an illusion from the begining, they are addicted to government heroin. A business that consumes resources and loses money makes the whole country poorer. These businesses are like an infection, the infection wants to live, it wants to grow, but it damages the body and survives by consuming resources other beneficial cells could have used more productively. If our immune system (the market) attacks the infection or takes away its resources, we shouldn't complain that it's idle, we should let it die. The process of killing it isn't fun, we could get a fever and a headache and feel like crap (recession) but ultimately we emerge stronger without that weight holding us down.

The economic argument is basically Hayek's argument that interest rates are not an arbitrary number, but serve a coordinating function in the economy. It's a signal to producers about the relative willingess of people to consume today or in the future. When people save, the cost of capital goes down and long term projects become economic. When people spend, cost of capital increases and businesses can only get funding for short term projects intended to meet current demand. When both long term and short term projects are economic at the same time (low interest rates, stimulus) we get booms that inevitably lead to excess capacity and busts. The bust is not the problem, the boom is the problem. The government stimulus argument says we must create demand to fill the capacity. The Austrian argument is the capacity should never have been created in the first place and must be liquidated for mis-allocated resources to be freed for more productive purposes. Each time we strap on another bandaid the more infections survive the bust and the larger the imbalances ultimately become. The transition is painful, but the longer we delay the restructuring the greater the pain will be. It's just a matter of time. Again, excluding the potential of game changing miracles (like cold fusion or something).

[Mike Moffatt]

RE: Bruce's comments. I would agree that the longer/deeper the recession, the more likely fiscal policy is to work. Monetary policy operates with a much shorter lag (the Fed could change the target Fed Funds rate tomorrow if they so desired) so ceteris paribus it is a much stronger option. Though I can see the argument that for a deep/long recession monetary policy is not going to be enough - particularly when the Fed Funds rate hits near-zero.

My argument isn't so much that fiscal policy doesn't work as fiscal policy doesn't work in a democracy. Using fiscal stimulus in, say, North Korea would be more effective because you bypass Step 3 (checks and balances).

The only way I see fiscal policy being effective in a democracy for a small-to-medium sized recession is through the use of automatic stabilizers, thus bypassing steps 1 through 3. In order to avoid a lag the stabilizers should be based on leading indicators (or at the very least avoid using lagging indicators), so it should not be based on employment data. Even if you could find the right leading indicators, I am not sure how well it would work in the long term, due to the Lucas critique.

[Nick Rowe]

By the way, my own view is that fiscal policy would probably work, but monetary policy would probably work better. I can see a "belt and braces" argument for fiscal policy plus monetary policy, just in case I'm wrong about monetary policy. And in Canada (not the US), we have had a declining debt/GDP ratio, so would need to loosen up on fiscal policy sometime anyhow, and loosening up now would seem as good a time as any. But generally, I want fiscal policy set on "strong automatic stabiliser" mode.

Mattyoung: but your argument seems to imply that the government produces more goods, and then tries to sell them to households at a market price. I normally think of fiscal policy as the government buying more goods, or buying the labour and other resources to produce goods for itself. It may give them away to households (free use of new bridges), but it doesn't sell them.

Adam P: I would model your argument (oversimplifying) as saying that government expenditure is on useless goods, that add to measured GDP but not to people's welfare. Digging holes and filling them in again is the classic example ("bridges to nowhere" seems the modern equivalent). If the multiplier were greater than 1.0, there would still be a net gain in welfare and useful production though. But arguably it's less than 1.0, especially if the fiscal stimulus continues past the point it's needed, so we get 100% crowding out at the margin, and a multiplier of 0.

Niklas: I interpret your answer to be that if we can't use monetary policy, then we MUST use fiscal policy. But will fiscal policy work?

Trevor: Your argument assumes that fiscal policy will cause a reallocation of resources AWAY from some other areas of production. Now, if we were on the LRAS curve, this would naturally happen. And the question would be whether the government goods were more or less useful than the private good crowded out. But if we are OFF the LRAS curve, with deficient-demand unemployment, why would fiscal policy cause employment to fall in other sectors?

Mike Moffatt wins, I think. (Unless you argue he cheated by breaking the deliberately vague "no funny stuff" rule.) I would interpret his argument as saying that fiscal policy would only work in practice, if we had the sheer dumb luck to have a government decide to use fiscal policy at a time when it wasn't needed, and then it turned out it was needed after all. Anyway, if the "green shoots" keep growing as they do, I think he will turn out to have been right (or at least the most plausible). On the ohter hand, you could argue that the EXPECTATION of future increased government spending is what caused private spending to increase now.

That last point makes sense theoretically. Suppose you argue that you need to increase expected inflation to lower real interest rates to get you out of a liquidity trap. Assume there's an unavoidable lag in implementing fiscal policy, but the announcement of fiscal policy to be implemented in the future is fully credible. When the extra spending kicks in, the economy is at full employment, and the extra government spending only increases the price level, so appears to be useless and badly-timed. But the expectation of that increased price level causes real interest rates to fall now, and gets us out of the liquidity trap.

The irony is that fiscal policy would appear not to work, in such a model, because it always comes too late. But it would actually work fine, via the announcement effect.

(It is considerations like this that make me despair of ever finding a nice clean direct econometric test of whether fiscal policy works. And It's why I want to see whether fiscal policy works according to a model, and whether the model is plausible. It's at the root of my original reply to David, on his blog, about why I want to look at models. There's a reason we use models, and don't just do model-free econometrics. And that's the reason.)

David:

One possibly irrelevant detail first. You say that "news" is serially correlated; bad news tends to be followed by worse news. If by "News" you mean "that which causes us to revise our forecasts", then that can't be right, unless our forecasts are irrational. "News" in my sense ought to be serially uncorrelated, so that the expectation of tomorrow's news is always zero. I think your definition of "news" must be the integral over time of my definition of "news". So your newspaper report the position of the particle following brownian motion, while my newspapers report the daily change of the position of the particle. (I'm not sure it matters; I just got hung up on this point when reading your note.)

My reading of your model is that neither monetary nor fiscal policy "works", because the underlying problem is a series of relative demand and supply shocks (between sectors, between firms, between individuals). And given real frictions in reallocating resources, we get a temporary fall in measured aggregate output (the resources employed in reallocating resources are not measured as part of GDP). It's not that fiscal (or monetary) policy won't shift the AD curve, but the original problem is not a leftward shift in the AD curve (that was just a symptom of the problem), so a rightward shift in the AD curve won't be the cure. Indeed, fiscal policy, because it won't be neutral in relative terms, will just worsen the problem, by requiring the economy to cope with a whole lot more reallocations of resources.

Is that a fair interpretation? Or what about this: the AS curve is vertical (real business cycle theory). The crisis caused the AS curve (as crudely and falsely measured by GDP) to shift left (and also the AD curve to shift left more, so inflation fell). Monetary policy can shift the AD right and prevent the deflation, but will do little or nothing to help the economy adjust. Fiscal policy can also shift AD right, and prevent deflation, but will actually cause AS to shift left (because of more reallocations). Is that fair?

I have never found RBC theory plausible, though a version like yours which emphasises all those micro-frictions comes closest to plausibility in my mind, and it has to be at least PART of the truth. All the usual reasons (sticky prices, apparent non-neutrality of money in the SR, etc.,). But one day I really want to try to get evidence on barter and "private money" clubs, and to try to see how they rise and fall with the business cycle. If they became more common in recessions (I suspect they do) that would reinforce my view that recessions are essentially monetary in nature, and arise from an excess demand for the medium of exchange. (That's a point I've been boring people here with in several posts recently).

Stephen: yes, I probably shouldn't have applied the "closed economy" rule. I just wanted to keep it simple, and to rule out the obvious "the exchange rate will appreciate" answer. Also, I wanted to think in global terms. Should all/most governments do fiscal stimulus seems to be the important question. But I did say anyone could bend any rule. And you especially Stephen!

pointbite: Wow! a real mixture of useless government spending, relative demand shocks, plus something new: fiscal policy worsens government and private sector balance sheets, and makes the financial crisis worse? Dunno. There just might be something to that. But I would like to see it clarified in a way that respected the immortal truth that net debt is always and everywhere zero!

Bruce: Yes, I've been thinking about Mark Thoma's "Icy Hill" metaphor too. Here's my model/interpretation of his metaphor:

ISLM with standard Phillips Curve, in which inflation adjusts slowly to the output gap, and expected inflation adjusts slowly to actual inflation. Assume the natural rate is negative, so we get stuck with unemployment, at 0% nominal interest rates, inflation falling, and expected inflation falling too, so real interest rates slowly rising, and so unemployment rising still further. We are halfway up an icy hill, rolling slowly backwards, and the hill gets steeper and steeper behind us.

The fiscal stimulus has to be big enough to make the natural rate positive, and above the expected rate of deflation, and last long enough, to get actual and expected inflation high enough, that the fiscal policy can be withdrawn and still leave the natural rate above the actual rate at 0% nominal. Monetary policy now starts to work again (we are over the top of the hill, going down the other side, can withdraw fiscal policy, and can use monetary policy to brake. The longer we delay with fiscal policy the bigger fiscal policy we will need.

In other words, Mark Thomas icy hill follows exactly from a fairly standard and very mainstream macro model. (I actually did a post on this a few months back, showing that a fiscal deficit could also pay for itself in such a model.)

But I would say that in the model, fiscal policy DOES work, in the sense that it makes things better than they would otherwise have been. It's just that if it's too small, too slow to start, or too quick to stop (you missed that bit Bruce), a temporary fiscal policy cannot work to cure the problem permanently.

I am much less worried about the icy hill than I was a few months back. US expected inflation has been rising strongly, judging by the TIPS spread, and is almost back to normal. What caused the cure is another question.

Wow! 2 more comments while I was writing this one!

[Adam P]

Nick,

really my point was that fiscal spending in this case adds to employment but doesn't end the recession in that it doesn't directly result in the recession ending. If the fiscal spending maintains employment until private sector demand picks up on its own the fine. If the fiscal deficits generate expected inflation and this stimulates by lowering the real rate also fine. But in neither case was the expenditure itself ending the recession.

That was not, btw, an argument against doing it though. It depends on what you mean by fiscal stimulus working. The Thoma argument certainly sounded like he had in mind a case where for fiscal stimulus to have "worked" the stimulus directly causes an end to the recession. This I find unlikely to be possible, (except by increasing expected inflation which could be done anyway).

[pointbite]

Nick: Fiscal stimulus with printed money borrowed from the central bank destroys both personal and government balance sheets at the same time, I may have the same number of dollars but they'll buy less than they should. You can't create purchasing power by printing paper, the new paper only has value because it steals a little from my savings. It's no different than raiding my safety deposit box. And of course the government now has more debt to service. Even increasing the supply of circulated dollars by borrowing from foreigners could have a similar effect, it will cause prices to rise without necessarily a corresponding rise in income. I find the whole "average debt is zero" thing a bit frustrating because it implies a $1 billion debt on my $1 million house isn't a problem because my debt is somebody's asset. The more relevant number is debt/GDP, but a GDP minus the statistical garbage (imputations, hedonics, etc) introduced in recent years. Check out that graph if you get a chance.

These plans may have a positive multiplier, but I never argued these kinds of tricks can't "work" (increase demand) in specific cases in the short term, my argument is that increasing demand for companies that shouldn't exist is even worse than letting them go bankrupt. My argument is that each iteration of such stimulus takes us further and further from a stable economy (real demand by willing people) essentially further increasing the need for more stimulus when the current shot of heroin runs out. Every recession requires rates to drop a little more, until eventually you hit zero and it's still not enough. Non of this should surprise anyone who is old enough to remember the Soviet Union. Managed economies are built on quick sand. Nobody knows the correct level of demand for every industry, especially not government bureaucrats or economists (no offense).

[David]

Nick:

You give a fair enough representation of the theory I have in mind, although you couch in that dang "old timer" macro language that makes little sense in GE models! (And your criticism on my assuming information process is one that I will have to think about--thanks!)

So yes, the gist of my theory is this. A contraction in AD caused by (rational) worsening private sector expectations concerning the future return to investment. Positive NPV capital projects are now deemed to have negative NPVs. The news shock is deflationary; and the demand for relatively safe assets rises (looks like a liquidity shock). The contraction in investment manifests as future declines in future AS.

Fiscal stimulus will not work, because it is typically targeted to negative NPV projects. It may work out ex post (if the government gets lucky); but this would also have been true for the private sector. One can justify fiscal stimulus (in my model) only if one holds the belief that the government is better at making NPV assessments (i.e., they are better forecasters). I find this latter assumption implausible.

This may not be the most plausible story, but dismissing it on the grounds of sticky wages/prices is not the way to go. We cannot account for decade-long depressions by assuming short-run price rigidities (and in any case, the data shows quite plainly that prices and wages adjust far more rapidly that many economists believe to be the case).

Is there a prize for second place? :-)

[Scott Sumner]

Nick, You said we get to pick either fixed M or fixed i. I believe fixed M is less expansionary (as fiscal stimulus would usually raise i.) So I picked fixed M.

My "model" is M*V=P*Y, and I assume V is fixed. QED. (Sorry, I never had IS-LM in college or grad school, in 37 economics courses--true story.)

On a serious note, the answer you gave is pretty much what I would have given, perhaps in combination with Adam's and several others. If we drop the fixed V, and make V positively related to the opportunity cost of holding money, you get the bizarre result that fiscal policy will be more expansionary if it raises interest rates. (Of course that doesn't apply if the Fed targets interest rates. So my next step is to assume the money supply is fixed forever, keeping inflation expectations low. Thus fiscal policy must raise the real interest rate. That is less likely to happen if investment demand for credit falls due to worry about high future taxes on capital.

So far I have merely considered the impact on nominal GDP. Since the test was couched in terms of real output, it could also fail if the policy made the economy less efficient. That seems to be the approach Adam took. Sorry I don't have anything original, but what can you expect when the the rules are specifically written in a way to prevent one contestant from using his favorite answer. :)

On a serious note, I think if one assumes fixed M, then fiscal policy probably has relatively little effect. Of course Keynesians usually assume fixed i, and I can't blame them.

[himaginary]

Actually, I think, most plausible argument against fiscal policy came from none other than Krugman himself a while ago.

http://web.mit.edu/krugman/www/SCURVE.htm

[pointbite]

himaginary, that may be the first time I've ever found a Krugman piece something less than infuriating. Thanks for that, and I would note it's not too different from my position, except for the belief in government's ability to find a new equilibrium with a sufficiently large stimulus... but he doesn't even really seem to believe it. I think some people might call that theory a double or nothing gamble, with a strong bias towards nothing.

[Mike Moffatt]

"Mike Moffatt wins, I think."

I demand a recount! I should never finish ahead of David Andolfatto in a Macro competition. Anyone associated with UWO Econ during the 1990s can tell you that.

"I would interpret his argument as saying that fiscal policy would only work in practice, if we had the sheer dumb luck to have a government decide to use fiscal policy at a time when it wasn't needed, and then it turned out it was needed after all. Anyway, if the "green shoots" keep growing as they do, I think he will turn out to have been right (or at least the most plausible). On the ohter hand, you could argue that the EXPECTATION of future increased government spending is what caused private spending to increase now."

I think that's a fair interpretation. I also believe that it is quite possible that "the EXPECTATION of future increased government spending is what caused private spending to increase now." as a placebo effect. Then there are three basic ways fiscal stimulus can work:

1. Through automatic stabilizers
2. During a depression where the economy experiences several years worth of negative Real GDP growth
3. Through a placebo effect

Note that these all differ from the traditional Econ 101 story of 'getting out of a recession' through discretionary fiscal policy G -> AD mechanism (including #2, which differs by degree).

[Bill Woolsey]

During the 1st quarter of 2009, the U.S. had negative net investment--first time since 1947,

As I was calculating the figures for my macro lecture, my first thought was--maybe it isn't just the falling nominal income. Maybe no one wants to investment because the "new plan" for the U.S. is going to be high taxes on future profits.

[Nick Rowe]

Bill: could be. But it could also be the recession.

Is it:
1. The MPK net of taxes is zero, or
2. The MPK is irrelevant if you can't sell the extra goods produced by the MPK.