Voodoo economics?
Sometimes when I post on an unfamiliar topic, I preface my comments by stating that I am pretty sure that I have overlooked something important, as the standard view can’t be as silly as it seems. And I have never had greater doubts than in this post. Mainstream economists can’t have overlooked something so basic. And fiscal multipliers are something I rarely even think about. So I hope you guys can set me straight.
Update 10/10/09: Alex and Leigh did set me straight. I thought the long run multiplier of 1.5 meant that if you increase G by $1, then in the long run Y would rise by $1.50. Instead, it looks like a long run multiplier of 1.5 means that in the long run the effect of G on real GDP is zero. Indeed, I am pretty sure that if Alex and Leigh are right, and if G had a permanent effect on the LRAS curve, then the long run multiplier would be positive or negative infinity. Thus you might want to skip the rest of the post. (This will teach me to stay out of hydralic Keynesianism.) Others are free to offer insights, but I no longer have confidence in my interpretation of the paper in question.The standard AS/AD macro model assumes that demand shocks have a temporary effect on output, but no permanent effect. Instead, in the long run output returns to the natural rate (or LRAS) which is determined by supply-side factors. I think you could say that this “natural rate” view of macro is one of the few things that all mainstream economists accept. The main difference between left- and right-wing economists is on the question of how quickly the real economy moves back to the natural rate after being impacted by a demand shock.
So I think we all agree that the long run fiscal multiplier is zero. Or at least I thought so. Recently some big names like Krugman and Mankiw have linked to this study done by three respected economists, which estimates a long run multiplier of 1.5 for an economy under fixed exchange rates. Mankiw doesn’t have anything to say about whether he accepts that number, but does seem pleased by the fact that the relatively high fixed exchange rate multiplier supports the Mundell-Fleming model, which he uses in his textbook. Krugman seems to find the 1.5 figure plausible, and even suggests that the fixed exchange rate model is appropriate for a country like the US, which has a relatively closed economy and a short term interest rate currently stuck at zero. But I’m not interested in which model best describes current conditions in the US, I’m interested in their estimate of the long run multiplier, which obviously wasn’t meant to apply only to countries in a liquidity trap.
One possibility is that Ilzetzki, Mendoza, and Vegh intend this estimate to apply to nominal output, not real output. But I have two problems with that assumption:
1. In their Voxeu.org summary they repeatedly refer to the effect on real output, not nominal output.
2. Even a 1.5 long run multiplier estimate for nominal output would be implausible, as it would require a permanent increase in the real exchange rate (assuming output was unaffected in the long run.)
So what does this mean? Should we start building pyramids in the desert? Or have I misunderstood what they mean by a 1.5 long run multiplier? I recall Robert Barro describing the Keynesian multiplier theory as “voodoo economics,” but I never paid much attention. Although I have doubts about even the short run fiscal multiplier, the fact that the SRAS has an upward slope led me to believe that short run multiplier estimates were at least defensible. But a long run multiplier? What does that mean? I look forward to being enlightened by your comments.
Tags:
7. October 2009 at 07:35
It’s important to consider that one of the assumptions is that the central bank would accommodate fiscal policy for over three years, according to the estimates, for the long-run multiplier to apply to the US. I hardly find that reasonable.
7. October 2009 at 08:01
As I understand it they didn’t impose any restrictions derived from theory on their estimates. If that’s right their results reflect any and every benefit which flows from boosting output. For example, the stock of physical and human capital is higher than it would be without the stimulus. In contrast, the textbook theory you mention typically assumes that such lasting benefits don’t exist. But even the textbooks usually mention hysterisis, or at least they did when I was in school. And I don’t think even Robert Barro would deny that US government spending on R+D has had a lasting effect on potential GDP. Likewise, I’m sure that many smaller countries have done some lasting good by spending money on health and education. If that spending is positively correlated with deficits, which it may well be, the cited estimates will be affected by that.
7. October 2009 at 08:11
Lance: …one of the assumptions is that the central bank would accommodate fiscal policy for over three years….
I don’t see that. Are you looking at a different study?
7. October 2009 at 08:34
The article is very short on details about method, processing and where the statistics came from.
7. October 2009 at 08:57
I wonder what they do with countries that switched regimes.
7. October 2009 at 09:09
Scott,
You missed the key word “cumulative”. The experiment is something like you increase G by 1% this quarter. Output this quarter goes up by 0.05%. The next quarter it will go up by 0.02%, in 8 quarters it will go up by 0.10%, and so on. Add all the effects and you get that the “cumulative” impact. The multipliers they show are the cumulative ones. The fact that they level on the long run means that there is no long run effect of fiscal policy.
Perhaps this paragraph from the paper clears things out
“The impact multiplier for high income countries is
0.24. In other words, an additional dollar in government
spending will deliver only 24 cents of additional output
in the quarter in which it is implemented. For developing
countries, the impact multiplier is close to zero.
Focusing on the impact multiplier, however, may be
misleading because fiscal stimulus packages can only be
implemented over time and there may be lags in the
economy’s response. To account for these factors,
Figure 2 shows the cumulative multipliers for both high
income and developing countries, defined as the cumulative
change in GDP divided by the cumulative change
in government consumption (as a percentage of GDP).
For example, a value of 0.5 in quarter 3 would indicate
that, after 3 quarters, the cumulative increase in output,
in dollar terms, is half the size of the cumulative
increase in government consumption.”
Alex.
7. October 2009 at 09:30
Scott,
Another comment. I wonder what Krugman would have said about the paper had he actually read it. This is what he said:
“On a happier note, this piece by Ilzetzki et al is interesting, and offers a wide range of multipliers depending on a country’s situation. The question for the United States is which estimate is most relevant.
I’d say it’s the fixed exchange rate estimate. Yes, I know, we have a floating rate. But they explain the relatively high fixed-rate number by pointing to Mundell-Fleming, which says that fiscal policy is effective under fixed rates because it doesn’t drive up interest rates (capital flows in). We’re in a similar position for a different reason: fiscal expansion doesn’t drive up rates because we’re at the zero bound.
Oh, we’re also relatively closed.
The thing is that both the fixed rate and closed multipliers are around 1.5 “” which so happens to be just about the number assumed by Christina Romer in her analysis for the Obama administration. Just saying.”
This is what the paper says:
“Given the current debate in the US regarding the effectiveness of President’s Obama fiscal stimulus package –
captured by the debate between Barro and Romer in the introductory paragraph –it is certainly relevant to inquire about the size of fiscal multipliers for the US. The left-hand side panel in Figure 6 shows the cumulative
multiplier for the US for our whole sample (from 1960 to 2007). The impact multiplier is 0.64 and the long-run cumulative multiplier is 1.19. While these estimates are certainly closer to Romer’s than to Barro’s, they mask some important structural changes over the sample period. To see this, consider the right-hand side panel in Figure 6 which breaks the sample into pre-1980 and post-1980. To avoid cluttering, we have omitted the significance bands in the right-hand side panel of Figure 6. They would show that the pre-1980 multipliers are significantly different from zero only for the first five quarters and that the post-1980 estimates are never significant. We should caution, however, that the sample for the US is about eight times smaller than
that for high income countries as a whole.
The difference between the two sub-samples is quite striking. We can see that the pre-1980 multipliers are considerably larger than the post-1980 multipliers. The post-1980 multipliers are just 0.32 on impact and 0.4 in the long-run. This is certainly a far cry from the impact multiplier (1.05) and long-run multiplier (1.55) used in the Romer report,3 although our findings for the US are consistent with Perotti (2004).”
Alex.
7. October 2009 at 10:03
Brad DeLong keeps referring the the argument about how to spend stimulus money, an argument still in the game. Given the infinite varieties of methods to spend I see no theory that works unless it compares stimulus dollars spent to economic constraints.
7. October 2009 at 10:32
Scott
The keynesian concept of multiplier refers to RGDP which was no different from NGDP since prices were (assumed) fixed. All modern versions refer to RGDP. In this case, if you assume a cumulative (long run) multiplier of 1.>0, this means that more G today will decrease G/Y in the long run. It doesn´t make sense.
7. October 2009 at 12:18
Kevin,
If exchange rates are flexible (or if nominal interest rates are affected by outflows and inflows of capital), fiscal policy will be sterilized to a degree due to increases in the real interest rate, as capital flows outwards due to increased borrowing by the government.
If exchange rates are fixed, the real interest rate will be fixed and stimulus will not be sterilized due to changes in the real interest rate.
Essentially, is there some force keeping real interest rates down? In our case, it’s not Bretton-Woods, but the Federal Reserve through accomodative monetary policy.
Pre-determined vs. flexible is essentially asking whether short-term nominal interest rates are fixed. In our case, they are to a degree. But over three years is another matter.
7. October 2009 at 17:09
Lance, Monetary accommodation should help the short run multiplier, but the long run multiplier for RGDP should be zero with or without monetary help. There all just difference forms of AD shocks, none of which affect RGDP in the long run.
Kevin, In theory that’s possible, but in practice the reverse is true. Fiscal stimulus involves massive waste, which actually drags down growth. Most governments are already much bigger than the optimum if maximizing RGDP is the goal. There is a reason why Singapore is richer than Denmark. Both have very free markets, but Singapore has a much smaller government. R&D is only a tiny share of fiscal stimulus, far too little to affect the results 8, 10 or 12 years forward. If Krugman had made that supply side argument for the 1.5 multiplier the entire economics profession would have been on his case. Everyone assumed he was talking about standard Keynesian demand-side stimulus, not shifting the LRAS to the right. That’s why few even commented.
I have a different view. If the results show a 1.5 long run multiplier, which we know is wrong, then the whole procedure is bogus.
And my final point is that a rightward shift in the LRAS is a supply-side argument, and a far weaker one than cuts in MTRs. So how can the left call supply-side “voodoo economics.”
Kevin#2, I don’t recall it either, but some degree of monetary accommodation is implicit in the fixed rate assumption. Fiscal stimulus will tend to pull in capital. You need monetary accommodation to keep the exchange rate from rising. That’s one reason the multiplier estimates are higher.
Current, I have no interest in the article, I have a great interest that Krugman would cite the 1.5 long run estimate approvingly. (And maybe Mankiw as well.) That’s what astounds me.
Pushmedia1, I’d guess they just looked at the period when each country was under fixed rates.
Alex, I don’t think you are right. I think they are claiming a long run 1.5 cumulative multiplier. But let’s see what others say. The marginal effect wears off when the graph plateaus.
Alex#2, Yes, I agree, Krugman cherry-picked the results that made his argument look best.
Mattyoung, I agree.
Marcus, I don’t think it decreases the ratio, but it does mean higher non-government output in the long run. It literally means if you increase government output $1, non-government output will also increase (by 50 cents), even in the long run. But if government is 10 out of 100, and goes to 12, then total output will go up to 103. So 12/103 is bigger than 10/100.
Lance, I agree.
7. October 2009 at 17:17
If you change regimes because you can’t sustain your peg, then the long run multiple is zero.
8. October 2009 at 00:24
I have a different view. If the results show a 1.5 long run multiplier, which we know is wrong, then the whole procedure is bogus.
Indeed. Who are you going to believe, my theory or your lying data? Fine by me, but don’t complain when some DFH like Joe Stiglitz dismisses some fancy bit of freshwater number-crunching in just the same way.
So how can the left call supply-side “voodoo economics.”
AFAICR they got the description from that commie-loving George H. W. Bush. And didn’t David Stockman let the cat out of the bag in his memoirs or something?
…some degree of monetary accommodation is implicit in the fixed rate assumption.
I would say that (absent trade barriers) a fixed-rate assumption implies a completely passive monetary policy, not merely some degree of accomodation. My question was where Lance got his “three years” from; a different paper entirely, I presume.
Finally, I think you’re right and Alex is wrong about the meaning of the multiplier. But I wouldn’t describe the authors as “claiming” anything. They make it quite clear that the confidence intervals are wide, so (as usual in economics) you can go on believing whatever you want. They are just reporting their results like good researchers do, even when the results call pet theories into question.
8. October 2009 at 03:01
I don’t see how we can even debate this without knowing what the authors mean, which doesn’t seem clear from their paper.
The correct criticism here isn’t “here’s data supposedly showing a 1.5 long-run multiplier, how silly”. It’s “What is the methodology for this data? How are real variables derived from nominals? How are the increases in state spending disambiguated from other things that have happened at the time?”
I agree that fiscal stimulus causes a short-term boom (I’m not sure about 24 months though). I think that it has destructive effects after that because of the misallocation of capital.
8. October 2009 at 03:02
What’s DFH?
8. October 2009 at 03:55
pushmedia, The long run multiplier should always be zero (from the demand side), regardless of assumptions.
Kevin; You said;
“Indeed. Who are you going to believe, my theory or your lying data?”
This is not a good argument. There is a massive amount of data supporting the natural rate hypothesis. It is one of the most well-established ideas in modern macro. Krugman often says he accepts it. One study hardly refutes it. If these three guys want to challenge that idea, fine, then say so. But I think they would be afraid to say so, as they’d sound silly. It is bizarre that they don’t even mention that their study refutes the NRH.
You said;
“AFAICR they got the description from that commie-loving George H. W. Bush. And didn’t David Stockman let the cat out of the bag in his memoirs or something?”
So Krugman gets his ideas from Bush and Stockman? That would be really sad. I think Krugman’s smart enough to make up his own mind. Again he can’t argue supply-side is voodoo economics, and say the long run multiplier is 1.5. At least the supply-siders are talking about changes in incentives that might boost the economy’s efficiency
You said;
“I would say that (absent trade barriers) a fixed-rate assumption implies a completely passive monetary policy, not merely some degree of accommodation.”
The standard model says otherwise. Without monetary accommodation an increase in government borrowing will cause the exchange rate to appreciate.
You said;
“Finally, I think you’re right and Alex is wrong about the meaning of the multiplier. But I wouldn’t describe the authors as “claiming” anything. They make it quite clear that the confidence intervals are wide, so (as usual in economics) you can go on believing whatever you want. They are just reporting their results like good researchers do, even when the results call pet theories into question.”
As I said before I don’t care about there study at all, all I care about is that Krugman (and maybe Mankiw) seemed to like the 1.5 long run multiplier. If Krugman had fixated on the short run multiplier and ignored the long run multiplier, I wouldn’t have even written the post. But Krugman thinks the long run multiplier of 1.5 is plausible for the US. I find that bizarre.
I really thought this post was going to be demolished. That I had made some trivial error such as the one Alex hypothesized. If I haven’t, then Keynesian economics is even more screwed up than I thought.
Current, Again, my criticism is not the authors’ study, but rather the way Krugman and Mankiw cite it. It really doesn’t matter what their method is. Maybe they did everything right. (In any case, I think VAR is a flawed technique.) But Krugman is saying “see, they showed the long run fiscal multiplier is 1.5” I find that an astounding claim by Krugman. I really don’t see why other people aren’t equally astounded.
It looks like this is an issue I will have to return to repeatedly. There must be some explanation, but I can’t imagine what it would look like.
8. October 2009 at 05:34
Scott and Kevin,
Look at figure 1 in the paper. That is the impulse response function, i.e. deviations in a variable caused by a shock from the base case in which there is no shock. You will see that in the long run there is no effect on output from an increase in G. In fact I wouldn’t be surprised if this was actually one of their identifying restrictions.
Let me see if I can give you an example. Output without shock is 100 for ever. Now G goes up by ten dollars for one quarter which causes output to go up to 105 for 3 quarters and then returns to 100. The impact multiplier is 0.5 and the long run cumulative multiplier is 1.5.
Alex.
8. October 2009 at 07:29
Kevin,
Look at the x-axis of the graph for the paper. It’s measured in quarters, and the cumulative effect plateaus around 12 quarters or so, or approximately 3 years.
8. October 2009 at 11:10
I’m saying for all practical purposes, they’re estimating a zero long-run multiplier for all countries because fixed regime countries will have to float after a while.
9. October 2009 at 04:20
Zero? The long term multiplier should be negative in terms of real GDP.
9. October 2009 at 05:51
Scott: There is a massive amount of data supporting the natural rate hypothesis.
Surely you’re not saying the NAIRU is constant? If you have a model of how the NAIRU is determined and the CEPR study contradicts that model, then we can argue about which should be discarded. The authors don’t present their work as a test of any highly-regarded model. It contradicts crude neoclassical presumptions of course but I don’t see anything there that would startle Milton Friedman, for example. They do say that Romer comes out ahead of Barro but that’s hardly earth-shaking.
So Krugman gets his ideas from Bush and Stockman? That would be really sad.
I hadn’t realised that “the left” was Paul Krugman, though as a regular reader I probably should have. Anyway, the view that Reaganomics was voodoo is widely held, not only on the left.
The standard model says otherwise. Without monetary accommodation an increase in government borrowing will cause the exchange rate to appreciate.
The standard model does indeed say this, and more, but it does not “say otherwise” than I did. I appreciate the trouble you go to in responding to comments, but it sometimes looks as if you haven’t actually read them.
Enjoy your break.
9. October 2009 at 05:57
Alex,
Unless I am mistaken about Scott’s views, he won’t be appeased by your numerical example. It implies that there’s a free lunch of sorts on offer. His theory says that that’s impossible.
9. October 2009 at 06:25
Kevin,
I´m not defending the results. I´m just trying to explain how the cumulative multiplier is computed and what it means. If we can´t agree on that then there is no point on arguing the rest.
Alex.
9. October 2009 at 07:48
Kevin,
What’s your definition of passive?
If the Fed does not respond to an increase in government borrowing by increasing the money supply (my definition of passive), all else equal, interest rates will increase in the standard model.
If the Fed does respond to increased borrowing by increasing the money supply (active/accomodative), interest rates will not increase in the standard model.
Fixed interest rates implies the second scenario.
This seems contrary to what you are claiming as passive, unless it’s merely a matter of semantics.
9. October 2009 at 08:13
“So what does this mean? Should we start building pyramids in the desert?”
This is exactly the sort of behavior Keyenes suggested. Although his example was digging ditches and then hiring people to fill them in. Yes its loony, yes they are serious.
10. October 2009 at 01:28
Hi Scott
I think Alex has it right: the key distinction is between long run Y (output/year, which as you say should revert back to its normal level in the long run) and cumulative total output. Cumulative output can be positively affected by fiscal stimulus even if long run Y is not: this free lunch is precisely the point of the Keynesian model. On this view the reduction in this year’s Y is an aberration which can partly be offset by government borrowing, without an effect on long-run annual GDP. Naturally you can query whether the Keynesian model really does work, but it isn’t logically inconsistent with constant LRAS.
In other words, “constant LRAS” is not the same as “long-run multiplier is zero”.
(Incidentally, I don’t think the choice between RGDP and NGDP matters in this particular case).
However there is another interesting factor: “constant” LRAS is not actually constant, because of long-run economic growth. Does output revert to 2.5% growth after a recession but from a lower base (the unit root hypothesis)? Or does it grow faster than 2.5% for a while to catch up with where it would have been in the absence of recession (the trend-stationary hypothesis)?
In the unit root scenario, the one-off Keynesian boost in output can increase the base for economic growth, and therefore DOES have a permanent impact on GDP/year.
This is really a different question than Scott’s original one, but it is a quite important one. Krugman and Mankiw had a bit of a debate about it in March which I commented on here:
http://www.knowingandmaking.com/2009/03/recession-and-recovery-krugman-and.html
10. October 2009 at 05:47
Alex,
The statement I disagreed with was in your first comment: “The multipliers they show are the cumulative ones. The fact that they level on the long run means that there is no long run effect of fiscal policy.” I agree with the first sentence but not the second. In the light of your subsequent comments, I take it that what you mean is that a transient fiscal shock has only a transient impact on GDP. That’s true of course. But a cumulative multiplier of 1.5 means that if G is maintained at its new higher level then so is Y. Fiscal policy can have a long-run effect on the level of GDP.
So, given his priors, Scott is right in rejecting the reassurance your first comment seemed to offer him. Unless I misunderstand him, he holds that a respectable model should have the neoclassical crowding-out property dY/dG = 0, or maybe even dY/dG < 0 when allowance is made for the wickedness of government.
10. October 2009 at 09:36
Alex, I’m still not convinced, but maybe you are right. If so, it is a weird way to define long run.
pushmedia1, But even if they estimated the multiplier was 1.5 after 20 years, that would be absurd.
Doc Merlin, I assumed zero in the best case, no efficiency costs.
Kevin, The natural rate is not stable over time, but that has zero implications for the long run multiplier. All mainstream macro models assume the natural rate is unaffected by demand shocks. I sure hope Romer doesn’t think fiscal stimulus can shift the LRAS to the right.
Of course Krugman is left wing, even his supporters would acknowledge that. Do you think he is a centrist or right-winger? I have friends who are Democrats who used to like Krugman and now think he has become a left wing ideologue.
In American right wing and conservative are almost synonymous, as are left wong and liberal (unfortunately) Almost nobody I know would raise an eyebrow if I said Krugman was left wing.
And you are still wrong in your criticism about fixed exchange rates. To maintain a fixed rate the central bank does need to adopt an accommodative stance when there is a fiscal deficit. You denied that view. If you care to explain why fiscal deficits would not force up interest rates without monetary accommodation, you are free to do so, but instead you simply make a sarcastic remark that I don’t understand you. That speaks volumes. Given you writing style, I’m not surprised you like Krugman. He did exactly the same with me, and later even his supporters conceded I was right and he was wrong.
Lance, Of course you are right. The fact that Kevin just made a sarcastic remark, and didn’t care to defend his views, speaks volumes.
Doc Merlin, And Krugman mentioned Keynes’ example of filling bottles with cash, burying them, and paying people to dig them up.
Leigh, If you are right then my post is clearly wrong. But what a bizarre way to define “long run”. The authors should have said the “cumulative effect” is zero, not the long run effect. I suppose if both you and Alex are convinced than I probably got it wrong.
How would they interpret a decision to permanently increase G by one unit?
I am aware of the unit root problem, which I discussed in other posts. I think the problem is caused by the mixture of nominal shocks and real shocks.
Kevin, You said;
“So, given his priors, Scott is right in rejecting the reassurance your first comment seemed to offer him. Unless I misunderstand him, he holds that a respectable model should have the neoclassical crowding-out property dY/dG = 0, or maybe even dY/dG < 0 when allowance is made for the wickedness of government." No that is not what I said. Instead of saying you didn't read what I wrote and leaving it at that, I will actually tell you what I said. I said a change in G should not lead to a long run increase in Y. I never denied that a short run effect was possible. And I would say the same thing for either a temporary or a permanent shock to G. As far as I know almost every mainstream macro textbook has that approach. So it's not just me saying that. I don't deny that you can come up with theoretical explanations for a positive long run effect, i do deny that they are remotely plausible in real world situations. Everyone, I'll add an update to reflect Alex and Leigh's explanation of the cumulative effect. I wish I had known that at the time. I would have focused on the fixed exchange rate issue, which implies it is monetary policy that is doing the heavy lifting in the fiscal multiplier. So that makes the 1.5 estimate either an artifact of fixed rates, or the liquidity trap. Which is what I argued in a paper in the late 1990s; i.e. that Keynesian economics was either a liquidity trap model or a gold standard model. Oh well.
10. October 2009 at 13:31
“Doc Merlin, I assumed zero in the best case, no efficiency costs.”
Ah, that makes more sense.
“Which is what I argued in a paper in the late 1990s; i.e. that Keynesian economics was either a liquidity trap model or a gold standard model. Oh well.”
Also, a model where there are high costs to changing prices (maybe regulatory fiat that makes it hard to change prices, or something similar), but you probably mentioned that elsewhere.
11. October 2009 at 08:04
Scott,
the multiplier is a concept that relates to closed economies with significant unemployment rates (Keynes’s depression economics). The logic is old in doctrinal history: John Law, certainly not the first, argued that in case of unemployment, higher nominal spending increases real output. Real wages are assumed to be constant, since prices and nominal wages fall together with NGDP. Further note that in the old IS/LM framework from which undergrad textbooks derive the AD curve an increase in G and any other autonomous component is accomodated by an increase in money in circulation. That is, the multiplier rests on the argument that in a closed economy with large unused resources an increase in NGDP increased economic activity and real output. The idea that the multiplier is greater one is due to the fact that the addition to money transmits by successive spending, each time increasing activity, though with diminishing return due to a positive propensity to save. But hey, this is all old Keynesian stuff. It has no meaning for the Ramsey saver we use in modern macro.
14. October 2009 at 15:33
Doc Merlin, yes, but many non-Keynesian also assume sticky wages and prices. Indeed I do.
amv, Thanks for that summary. The following is not directed at you, but rather at Keynesian economics:
1. If prices and wages fell with NGDP, then a fall in NGDP would not reduce output. So I find that assumption odd. My problem with Keynesian economics is that they really don’t have a theory of NGDP. They just sort of take it as a given. And I think that makes them miss how the current situation reflects the long run consequences of monetary actions taken earlier.
2. I don’t see why Keynesians can assume that fiscal stimulus will increase NGDP, and I discussed why in a post a week earlier.
3. The best argument for a fiscal multiplier is that fiscal policy has shorter policy lag than monetary policy. But the reverse seems to be true. Most of the fiscal stimulus is still unspent.