Ryan Avent on the zero multiplier hypothesis
The assumption that the fiscal multiplier is zero was once regarded as part of the standard new Keynesian model. But ever since interest rates fell to zero there has been renewed interest in fiscal stimulus. I think this is wrong, but I seem to be in the minority. Now Ryan Avent explains why the multiplier might still be zero, even at the zero interest rate bound:
A common argument at this point in the discussion is that the Fed is recusing itself from the business of macroeconomic stabilisation and fiscal policy should therefore be used to bring down unemployment. The inflation constraint prevents this, however. The higher inflation we’d like the Fed to induce is nothing more and nothing less than higher demand. The Fed might signal, for instance, that it will tolerate more inflation in the future (perhaps by indicating that it will leave rates at zero beyond the point at which the economy is expected to exit the liquidity trap). If a dollar is expected to be worth less in period t+1, then people will use their dollar in period t, turning expected future inflation into current inflation by inducing people to spend and invest in the present. Inflation rises now because the Fed has made it attractive to spend and invest now.
The problem is that the Fed is willing to accommodate this process up until it generates price increases of about 2% a year which is insufficient to bring the economy back to full employment. Recall that the liquidity trap occurs because households that previously spent and borrowed too much are now deleveraging and households that previously saved too much aren’t spending and investing enough to compensate. At 2% inflation, the economy isn’t running hot enough to encourage these previous savers to deploy their resources sufficiently to generate full employment; there is still excess desired saving.
The government might then step in to provide the needed consumption and investment. It could borrow from all those lusty savers and deploy the money it obtains. The government’s addition to ongoing spending would raise inflation, however; it would be stimulative because there would be even more money chasing a given set of real resources. You and I might think that’s a good thing, because that inflation will induce rising production which will in turn create more employment. If the Fed is serious about sticking with 2% inflation, however, then it will disagree. It will respond by tightening policy, either by raising the nominal interest rate or curtailing other stimulative actions in order to bring expected inflation back down to 2%. Government spending will crowd out other spending, not because of any market constraint but because the Fed will react to the higher inflation generated by the government by tweaking policy to make spending less attractive to private actors. If the Fed is serious about sticking with 2% inflation, then it will place a lid on recovery, and there’s nothing that fiscal stimulus can do about it.
Does the Fed actually behave this way? We can certainly debate the point. If we look at the Cleveland Fed’s calculations of market estimations of 10-year expected inflation, however, we see that the highest those expectations have risen since the beginning of 2009 is just 2.09%. Expectations last touched that figure in January of 2010, at which point the Fed took the decision to allow its balance sheet to contract as the securities it held matured””policy was allowed to tighten, if passively. The FOMC later reversed this decision and then added to the balance sheet via QE2, in response to a marked decline in the economic outlook (corresponding to a sharp reduction in inflation expectations). And 10-year inflation expectations have not been above 2% since May of 2010.
When you’re in a liquidity trap, religious adherence to an inflation-rate target is problematic. If the Fed is worried about losing credibility in the event that it misses on its target to the upside, then it really should consider changing targets, either to a higher rate or to a level, of price or some other variable (yes, like nominal GDP). And if the Fed thinks that such a change is too costly? Well, it will be a long decade. Recovery will continue at a very slow pace, and a recession generated by an unexpected shock will be a constant risk.
Addendum: Karl Smith notes that the government could still reduce unemployment by taxing current workers and hiring the unemployed. It might also feasibly do it by taxing imports and subsidising exports””the so-called “fiscal devaluation”. Both options come with some pretty substantial economic and political messiness, of course.
BTW, Karl Smith’s right that the fiscal authorities would still have some options; the easiest would be an employer-side payroll tax cut.
PS. Some commenters have insisted that the money supply offset problem isn’t really a zero multiplier, as the effect of fiscal policy by itself is still positive, but it’s being offset by contractionary monetary policy. I think this is wrong. Consider the following two cases where fiscal stimulus has no ultimate effect on NGDP. Then tell me the multiplier in each case, and explain why:
1. The fiscal authority boosts spending, but mean-spirited businessmen react to the higher interest rates and scary deficits by cutting gross investment by an equal amount. Complete crowding out.
2. The fiscal authority boosts spending, but mean-spirited central bankers react to the threat of higher inflation by tightening monetary policy so much that investment falls by as much as G increases.
Tags:
10. January 2012 at 06:05
Your scenarios 1) and 2) describe a fiscal multiplier of 1, not of 0. A multiplier of 0 would imply that, if the government spends anything at all, the entire economy will grind to a total stop.
10. January 2012 at 06:36
Gene:
No. Fiscal multiplier of 1 means that a change in government spending increases income and output by an equal amount. If the goverment spends $100 billion more, then income rises by $100 billion.
With complete crowding out, the multiplier is zero. The goverment spends $100 billion more, and income remains the same. In the simplest case, the government borrows $100 billion more, and firms and households borrow $100 billion less and so investment and consumption together fall $100 billion, leaving total spending and total income unchanged.
With a multiplier of greater than zero and less than or equal to one, the increase in government output is free. (Leaving aside govermnet debt issues.) If it is greater than one, then we get the extra government output plus some extra private output as well. The increased government spending helps the private sector.
10. January 2012 at 06:37
Gene: you would need a fiscal multiplier of minus infinity, not zero, to get that result. The multiplier is the change in Y divided by the change in G.
10. January 2012 at 07:00
scott/nick/anyone:
Has anyone proposed a theory allowing the fiscal multiplier to be >0 based on the notion that individual level risk aversity is associated with borrowing is more severe than national level risk aversity?
When discussing ricardian equivalence every argument seems to be phrased in terms of expected future income – but that’s not what matters. It’s expected future utility, averaged across the distribution of expected states of the future world for each person. Even leaving aside macroeconomic effects (circular flows), the higher variance associated with individual level states of the world vs. national level states of the world (the nation is a pool of individuals), implies that EVEN IF government borrowing does not increase expected future income, it can (assuming the return on government expenditures is non-negative) have an effect on expected future utility? In other words, shifting from private debt to public debt is a form of insurance?
10. January 2012 at 07:01
Gene, I agree with Bill and Nick.
10. January 2012 at 07:18
Scott, I don’t get your first case. To get a zero multiplier because increased government spending crowds out investment, you need an increase in the interest rate. But doesn’t this whole discussion presume (rightly or wrongly) a liquidity trap, in which case, why do interest rates rise just because of additional government spending and a larger deficit? You’re not playing by the rules.
10. January 2012 at 07:30
statsguy, The multiplier is still zero under those assumptions, as long as the central bank targets inflation.
David, I think you misunderstood my point. I wasn’t arguing that crowding out will occur, I was asking how we would characterize the multiplier if it did occur.
I agree that this post presented no evidence at all that crowding out is likely to occur at zero rates.
10. January 2012 at 07:45
By the way, if the multiplier is between zero and 1, I was wrong to say the government output is “free.” There is some free government output, but there is also sacrified private output.
For example, if the mulitplier is .5, the goverment spends $100 billion and there is $100 billion more government output. The government borrows more and the private sector borrows less. However, the demand to hold money falls (and velocity rises) or the Fed creates more money, so the decrease in private borrowing is only $50 billion. Households and firms purchase $50 billion less private goods and services, firms don’t produce what they can’t sell, and so, the production of private goods and services fall by $50 billion.
So, there is $100 billion more government output and $50 billion less private output. In aggregate, output and income has increased by $50 billion.
As the multiplier approaches zero, the amount of private output “crowded out” approaches the amount of government output produced. As it approaches one, the amount of private output crowded out approaches zero.
If government output is counted as a waste, then in output terms (though not employment terms) if the multiplier is less than or equal to one, goverment spending is no help in terms of output.
Also, I actually believe, like Scott, that monetary policy can be used to increase total private expenditure as much as is desirable. And so, any increase in government output is at the expense of private consumer goods and capital goods that could have been produced if the Fed would do its job.
And finally, if fiscal policy causes inflation, the Fed has an excuse. It might not make much sense to Sumner, but it is an effective political excuse. We, the Fed, never does anything wrong. Congress spend too much, and our ability to stop the inflation caused by Congress was overwhelmed and so inflation went up. But we will get it under control soon. Prices may be 4% higher (rather than the usual 2% higher,) but soon they will return to rising from their current elevated level at no more than 2%.)
10. January 2012 at 08:51
Statsguy: I am not really up on the literature. But it is well-recognised that borrowing constraints (which are sort of similar to your risk-aversion case) would invalidate Ricardian Equivalence. If the government borrows on your behalf, it allows you to increase your desired spending (unless the central bank offsets it a la Scott).
Your risk-aversion argument sounds very plausible to me. I don’t know if it’s been made before.
10. January 2012 at 08:58
More perfect blogging by Scott Sumner.
10. January 2012 at 09:04
So the PS I’m assuming is a continuation of the conversation we were having here at the end of the comments section: http://www.themoneyillusion.com/?p=12461
So I may as well continue the conversation here. First of all may as well copy and paste my last response:
“Brito, But if monetary policy is exogenous, then what it is maximizing?”
I don’t understand what you mean with this question.
Also Scott, here is a question. First, let’s assume NGDP is stationary. Let’s also assume MMT people are in charge who also happen to be terrified of inflation. Let’s say the central bank engages in aggressive asset purchasing to boost nominal incomes, but the MMT government thinks this will cause excessive inflation so they massively increase taxes and cut spending to “remove net assets from the economy”, causing NGDP to remain stationary. In this case, what is the money multiplier?
10. January 2012 at 09:09
Statsguy, Nick:
It’s still the case that if SRAS is even slightly slopped and not horizontal, the fiscal multiplier is zero if the CB is credibly targeting inflation.
An assumption often made in Keynesian literature is that SRAS is perfectly level in a recession. That assumption isn’t convincing and therefore avoiding Ricardian equivalence is not enough to make the multiplier nonzero.
10. January 2012 at 10:33
Scott, The difference is that crowding out is endogenous to the model and fiscal and monetary policy are exogenous to the model
10. January 2012 at 10:37
Bill, I agree.
Benjamin, Another perfect comment. Thanks.
Brito, What is the goal of the central bank? That’s what I was asking.
I don’t believe there is any stable money multiplier, so we agree there. But the Fed has far more resources than the central bank, and thus can always inflate if it sets its mind to it.
Jon, Exactly, your reply is better than mine.
10. January 2012 at 10:43
David, If monetary policy were endogenous we wouldn’t be facing this problem, we’d just have Congress instruct the Fed to inflate.
The question is not whether something is exogenous or endogenous in any model, it’s a question of whether they are exogenous to the policymaker (in this case the fiscal policymaker) making the decisions. Obviously Congress doesn’t control monetary policy, and when they change fiscal policy they can only estimate the reaction of the Fed, just as they can only estimate the reaction of the private investors.
I realize that’s a horrible system, but it’s the one we have. It would make much more sense for the fiscal and monetary policymakers to have the same objective. In that case we’d have no need for fiscal stimulus. Now we have a need for it, but it doesn’t work.
10. January 2012 at 11:09
Jon: agreed and understood.
10. January 2012 at 11:21
Scott, Did you mean to say “if monetary policy were exogenous?” Obviously in the real world everything depends on everything else, but discussions (and empirical estimates) of and the Keynesian multiplier are undertaken in the context of a well-defined Keynesian model and the model doesn’t have a reaction function for the monetary authority based on what the fiscal authority does. You posed a conceptual question about the difference between the multiplier calculation based on the response of investment to an increase in government spending, and a response by the monetary authority to an increase in government spending. The Keynesian model allows for one type of response it doesn’t allow for another. I can agree with your characterization of the Fed’s reaction function but that doesn’t mean that the Keynesian multiplier properly defined and properly estimated (in terms of the model) is zero. You realize that we are simply engaging in another one of those fruitless but strangely enjoyable semantic arguments that economists are endlessly getting themselves into.
10. January 2012 at 11:27
Scott: “Brito, What is the goal of the central bank? That’s what I was asking.”
In this scenario? I don’t know, why does it matter?
“I don’t believe there is any stable money multiplier, so we agree there. But the Fed has far more resources than the central bank, and thus can always inflate if it sets its mind to it.”
Did you mean to say the fed has more resources than the government? Also I look at it like this, fiscal stimulus and the central bank creating inflation when at the ZLB is the same thing, but using different methods. Fiscal stimulus aims to increase the amount of money flowing in the economy by increasing the deficit; the central bank aims to (at the ZLB) increase the amount of money in the economy by essentially printing it and buying assets. The only difference between the two is that fiscal stimulus is subject the problem of Ricardian equivalence (although this doesn’t seem to be a very good argument) as the money comes from borrowing and central bank spending runs a risk of inflation as the stock of money is increasing and people may lose faith in the ‘soundness’ of the money stock.
10. January 2012 at 11:35
RE the PS — The fiscal multiplier is almost certainly f. (Naturally, the monetary multiplier is 1/f … .)
10. January 2012 at 11:41
Which decision is easier for the Fed to make: expand it’s balance sheet until there is 3% inflation (or 5% NGDP growth or whatever) or hold the line if fiscal policy pushes inflation up to 3%?
Maybe the answer is that the fed will do neither, but at least there’s a possibility that the Fed won’t act counter-productively in the latter scenario.
10. January 2012 at 11:43
david glasner: Scott’s argument is that you can’t have a model without an (implicit) Fed reaction in the first place, because there’s no good measure of “holding the Fed’s monetary stance constant”. Does that mean holding interest rates constant? Inflation expectations? And so on.
10. January 2012 at 12:21
Jon, Scott – there would have to be an argument that the CB is not perfectly targeting inflation – it’s either flexibly imperfectly or asymmetrically targeting inflation, or is subject to political constraints. “Should” and “is” are different things, unfortunately.
Nick – “But it is well-recognised that borrowing constraints (which are sort of similar to your risk-aversion case) would invalidate Ricardian Equivalence.”
…which is why, I suppose, Krugman relies heavily on the argument that giving money to poor people will get spent. I would argue that even transferring debt from the upper middle class to the federal government increases AD because of the money/utility mismatch in ricardian equivalence. Even if people expect to pay it, the pooling of risk facilitates increased current expenditure.
10. January 2012 at 12:26
David, Sorry, I meant exogenous. But I’m actually not sure the two terms cover what I was trying to convey. There are actually three possibilities:
1. The fiscal authorities might control M.
2. M might move around randomly.
3. M might respond to fiscal policy, perhaps offsetting it.
I meant the first possibility. But I think the second is also exogenous, isn’t it?
You said:
“discussions (and empirical estimates) of and the Keynesian multiplier are undertaken in the context of a well-defined Keynesian model and the model doesn’t have a reaction function for the monetary authority based on what the fiscal authority does.”
But that makes no sense. I think they must implicitly be assuming some sort of reaction. Perhaps they assume the interest rate is held constant, or perhaps they assume the monetary base is held constant (of course these two assumptions are quite different, and will produce radically different multipliers) but they must assume SOMETHING.
I have trouble with the idea of “what the multplier is in the Keynesian model” I don’t dispute that Keynesians can write down a simple model where the MPC is .8 and the multiplier is 5. That’s not what I’m talking about. If that’s your point, then I agree 100%. I’m interested in something very different, the actual effect of a change in G on NGDP. I would think that would be of interest to policy-oriented Keynesians. Perhaps not. Maybe they are simply interested in the multplier as generated by the model, given the assumptions in the model. If so, then Barro wasted his time arguing with them, as his point wasn’t that the model was wrong in a mathematical sense, it was that the real world response to
deficits was not what they claimed.
If Congress controlled monetary policy I would agree with you. Then it would make sense to work out the independent effects of M and the deficit. But if that was the case, there would never be any justification for fiscal stimulus. So if it is going to be justified, it must do so in a model where M is endogenous. It must be justified by arguing the Fed has a dysfunctional reaction function. That treats the Fed like a corporation, not a policy unit.
10. January 2012 at 12:34
Brito, See my answer to David.
You said;
“Fiscal stimulus aims to increase the amount of money flowing in the economy”
That’s false if you mean the quantity of money. If you mean velocity it’s true, but the effects are tiny compared to money creation. It’s really hard for the fiscal authority to boost V by even 10%, and the Fed can easily offset that. The Fed can boost M by 1,000,000%; let’s see the fiscal authority offset that with less V.
Neal, ??????
Adam, The Fed doesn’t need to expand its balance sheet to hit 3% inflation, or 10% inflation. It needs to announce it wants to hit 3% inflation. But they don’t want to. Indeed when asked by DeLong, Ben Bernanke said he opposed raising inflation to 3%. The Fed’s quite content with 2% inflation.
Alex, That’s right.
Statsguy, Agreed, but even if they aren’t targeting inflation perfectly, it’s surprising difficult to get a strong multiplier.
10. January 2012 at 12:45
@ssumner:
Can’t we ask them instead of guessing? Just email Krugman, De Long and a couple of others and ask them: In your model, do you hold, the base, interest rates or something else constant? Maybe I’m prejudiced, but I think that when they say: “holding monetary policy constant” what they really mean is: “not thinking about monetary policy and its effects.” (For what it’s worth, I think that’s basically what economists almost always means when they say ceterus paribus) And to be honest, it’s kind of scary.
10. January 2012 at 12:53
Alex and Scott, Thanks for the explanation. I actually interpreted Barro to be saying that the Keynesian model taken at face value did not make sense because the private economy does not respond in the way that the Keynesian model assumes it does, not because there is a perverse monetary reaction function. Given the ambiguity in the implicit reaction function, I would specify alternative reaction functions and derive the conditional multipliers rather than impose a particular reaction function and say that a specific conditional reaction function is THE Keynesian multiplier. But that’s just me.
10. January 2012 at 13:02
“That’s false if you mean the quantity of money. If you mean velocity it’s true, but the effects are tiny compared to money creation. It’s really hard for the fiscal authority to boost V by even 10%, and the Fed can easily offset that. The Fed can boost M by 1,000,000%; let’s see the fiscal authority offset that with less V.”
Yes in a sense I mean the velocity of money, the thing is I’m not thinking in terms of a strict MV = PY framework so it’s kind of hard to explain my point. I think the problem is V is heterogeneous, so it should really be more like M1V1 + M2V2 = PY, where M1 is money sitting in reserves at the banks and V1 is the rate at which this money is exchanged, and M2 is the money held by businesses and consumers and V2 is the rate at which this money is exchanged (note my use of M1 and M2 here have nothing to do with the measures of the money supply as is normally associated with those terms).
Fiscal stimulus is basically an attempt to move money from M1 with low velocity to M2 with high velocity (through borrowing from savers and transferring the money), whereas the central bank in normal times simply increases M1, but this M1 will be above its equilibrium level so some of it will bleed into M2 (as interest rates lower), or in a ZLB situation the central bank may be forced to try and increase M2 directly by buying non bank assets. Does that make sense? It’s pretty convoluted I know.
I don’t see why it should be hard to increase M2 theoretically (which is in a sense similar to increasing V, by changing the weights on V1 and V2), I mean the government could literally just write a check of $20,000 to everyone of average income or below, or they could do some crazy measure like paying off millions of peoples’ mortgages, cutting the payroll taxes right down and employing millions of workers in some huge infrastructure project. Now I understand that such a measure would be politically impossible, so I agree it is hard for the government to boost incomes politically speaking, but I don’t see any major theoretical constraints, depending on how the government does it.
10. January 2012 at 13:04
PrometheeFeu, I think Krugman and perhaps DeLong would be able to give a coherent answer, but I fear you are right about most economists, and I plan a post attacking my profession on this very point tomorrow.
David; You said;
“I would specify alternative reaction functions and derive the conditional multipliers rather than impose a particular reaction function and say that a specific conditional reaction function is THE Keynesian multiplier. But that’s just me.”
No, that’s me too! I’d have:
1. A multiplier for the world where the central bank never changed the base.
2. A multiplier for the world where the central bank never changed interest rates.
3. A multiplier for the world we actually live in, where the central bank attempts to at least roughly stablize AD.
And let people choose which multiplier they think is more useful for policy evaluation. Of course the third multiplier estimate would not be stable, and hence wouldn’t always be zero. It might be either positive or negative at various times.
10. January 2012 at 13:10
Brito, I would agree that right now they could boost velocity more than usual, especially thinking in terms of base velocity. But it doesn’t really help, as you’d still be swamped by the offsetting effects of changes in M.
You said:
“Yes in a sense I mean the velocity of money, the thing is I’m not thinking in terms of a strict MV = PY framework so it’s kind of hard to explain my point.”
I think this is precisely your problem. People often think in terms of the Keynesian framework as if it really describes the actually workings of the economy. It doesn’t. The Keynesian model is only correct to the extent to which it describes how expenditure shocks affect M and V. Thinking in Keynesian terms almost always gives poor results. Indeed that’s precisely the point of the post I put up 20 minutes ago–you may want to take a look.
10. January 2012 at 13:22
>I fear you are right about most economists, and I plan a post attacking my profession on this very point tomorrow.
That was certainly true of virtually every undergrad economics course I took. The exception was (fittingly) Monetary Economics with Henning Bohn. In all of my Macro courses we literally had the Walrasian model introduced, then never again touched MD/MS. Even with these glaring deficiencies though I got a lot out of those courses, most strongly the importance of z factor productivity, comparative advantage, competition, and a healthy dose of examining your assumptions.
10. January 2012 at 13:40
Scott – Thanks for the response, but I don’t think that answers what I was getting at. I guess I could just rephrase. Which decision is easier for the Fed to make: announce that it wants 3% inflation (or 5% NGDP growth or whatever) or hold the line if fiscal policy pushes inflation up to 3%?
Maybe what I’m getting at is that even the nominally independent Fed is constrained by political pressures that may be reduced if the fiscal authority acts first. Is the pressure to raise rate in the face of growth/inflation the same as the pressure not to be seen as endorsing inflation?
10. January 2012 at 13:56
Scott, I’m not thinking in a simplistic Keynesian framework either, that m1v1 + m2v2 was of my own making, and is from what I observe in the real world. I think that actually it is actually thinking strictly in terms of MV=PY which is restrictive, because M and V are unbelievably broad and vague variables, so I don’t think they are of that much use other than in the most abstract sense.
10. January 2012 at 14:33
Scott, you quite possibly have covered this before but why is the employer side payroll tax cut stimulative as opposed to the employee side?
In analyzing the payroll tax it is usally held that the employee in truth pays both sides. If that’s true it’s not obvious why one would prefer the employer side tax cut to the exlusion of the demand side.
10. January 2012 at 14:36
I meant “employee side” rather than demand side, thoguh I guess it’s kind of Freudian as I do see tax cuts for employees as more demand side stimuluative.
The reason is that the nonrich put more of their money into consumption
10. January 2012 at 14:46
Mike: the employer side payroll tax is stimulative through two channels: first, the generic stimulative effect (if any) of spending increases or tax cuts, but second (and possibly more importantly) by effectively reducing real wages. In a world of downward sticky nominal wages it’s not true that “the employee pays both sides” in the short run.
10. January 2012 at 15:25
Scott when you say that the multiplier ir roughly zero is that largely due to the particular monetary regime we have in place and have had since Volcker?
There was a story in CNBC where China is looking to stimulate the economy again.
They are doing some monetary easing but they are mostly doing fiscal stimulus because they believe too much monetary stimulus would risk inlfation.
For the record I wrote a post about this which I admit I sort of related to you only because it’s about a matter you have weighed in on quite a lot.
http://diaryofarepublicanhater.blogspot.com/2012/01/guess-hsbc-and-china-havent-read-scott.html
I understand you have a lot of demand on your time but feel free to check it out if you like. You make the title…
So what would you say about this, are they wrong and not understand or is this likely to work because they ave a quite different monetary regime in China that will not likley try to clamp down on the fiscal stimulus by tighening?
11. January 2012 at 07:17
Cthorm, I’m glad you had that class.
Adam, Obviously that’s a difficult question for me to answer. What I can say is that Obama did nearly $800 billion in stimulus and it failed because the Fed offset it with tight money. Would twice as much stimulus have succeeded, or would monetary policy have been even tighter? I suspect the latter. I suspect that if the Congress did $1.6 trillion, the Fed would have done even less QE, and we’d be in roughly the same spot. But I’ve always acknowledged that we can’t be certain. I’ve also always acknowledged that there is some level of fiscal stimulus that would boost NGDP (such as WWII spending.) I’ve also acknowledged that some types of stimulus are effective, such as employer-side payroll tax cuts, so I’m not ideologically against stimulus, I simply don’t want the money to be wasted.
Brito, There is nothing broad or vague about MV. M is defined by me as the base, which can be measured precisely. And V is the ratio of NGDP/base.
Mike Sax, It’s because nominal wages are sticky in the short run, so the tax cut is not immediately passed on to workers. It lowers business costs and shifts SRAS to the right. If the Fed targets inflation, it must expand and shift AD to the right.
Mike Sax, If you read this blog over the past few years you’d know never to trust anything the press says about monetary policy. They assume high interest rates are tight money, and vice versa. Not so. So take the Chinese reports with a grain of salt.
12. January 2012 at 05:42
Scott – sorry for the late reply. I’m just rephrasing your point concisely. As I read it, you agree that absent monetary policy fiscal stimulus could well have a positive multiplier, say, f > 1. But in reality the monetary authority will react to a fiscal stimulus, and the multiplier of that reaction will be exactly 1/f.
12. January 2012 at 13:01
Neal, No, I’m saying the multiplier will be roughly zero, unless monetary policy is incompetent, in which case it might be positive or negative.
13. January 2012 at 14:09
I’m saying exactly what you’re saying, in algebra instead of words.