Paul Krugman discusses whether Japan has had two lost decades.

This picture suggests that the Japanese economy was indeed depressed for about 16 years, and deeply so after the slump of the late 1990s. But it may have returned to more or less potential output on the eve of the current crisis.

Just to be clear, this is not a picture of policy success; it is, in fact, a picture of enormous waste. But the condition wasn’t permanent.

I think that’s about right.  But then Tim Duy raises a good question:

Yes, the lost years surely indicate a policy failure. But arguably there is some success.  On one hand, we can see this as vindication of massive deficit spending.  But is this really success?  Because on the other hand, there is no end in sight of such deficit spending.  On Japan’s 2012 budget, via the FT:

“Even by current grim international fiscal standards, Japan’s budget for the year from April 1 makes scary reading.

For the fourth year in a row, government revenue from bond issuance is set to exceed that from all taxes. Outstanding government debt is expected to hit an extraordinary Y937tn.”

I don’t intend to go down the “Japan’s bond market is about to collapse” path.   But what I am wondering about is Krugman’s description of the condition as not permanent.

I’m no expert on Japanese public finance (although I at least do know that the net debt is much smaller than gross debt in Japan), but the FT article Duy links to certainly paints a bleak picture of the long run trends.  So I think he raises a good question in asking whether fiscal stimulus can be viewed even as a limited success (in say 2003-07) if Japan can’t maintain the stimulus forever, and can’t survive without it.  Here’s Duy again:

Where does this leave me?  Yes, fiscal policy can work – it can back fill missing demand.  But we haven’t seen enough activity in Japan to really jump start the economy to a point that fiscal policy is no longer an economic necessity.  And I am not sure we can expect such a jump start without more explicit help from the central bank.  In which case, Japan has not truly “recovered.”  Thus, we cannot yet conclude their condition is not permanent.

In short, before we can declare recovery in Japan, we need to define what a successful recovery should look like. Is it just about returning to potential output (adjusted, in this case, for demographics), or should it be recovery to the point that deficit spending is no longer necessary to support that output?

These are all good questions, and I basically agree with Duy’s post.  However I’m going to quibble with one small point, the “yes fiscal policy can work–it can fill back missing demand.”  Yes, fiscal policy CAN work in some very limited circumstances.  But I’d argue that Japan is almost a textbook example of a situation where it CANNOT work, at least under the current policy regime.

First a few stylized facts.  Japan’s NGDP has hardly changed in 20 years.  The GDP deflator has fallen about 1% per year, and RGDP has risen at about the same rate.  The 1990s were clearly a lost decade, when Japan began falling further behind the US after rapidly catching up for many decades.  Japanese unemployment never got very high, but that partly reflects all sorts of cultural and institutional differences with the US.

Then things picked up a bit around 2003-07, before slowing again with the current recession.  Some point to the fact that they’ve recently done as well as the US in per capita terms.  But of course the past ten years have been something of a lost decade for the US as well.

I’d like to step back and notice something astounding about Japan, the nearly flat NGDP over 2 decades.  Yes, there’s been little population growth, and essentially no growth in the labor force.  But it’s quite striking to see a major economy have no NGDP growth for 20 years, even in per capita terms.  Indeed I can’t think of any other country that experienced this sort of nominal stagnation.  This begs two questions.  Why was NGDP so sluggish, and why was RGDP and unemployment not even worse, given the horrible performance of NGDP?

In my view, the slow NGDP growth reflects a very flawed policy approach by the BOJ, which fiscal policy was not able to overcome.  And the non-horrible (but very mediocre) performance of the real economy reflects the fact that Japan does have some wage and price flexibility.  By no means perfect flexibility, but enough to keep unemployment at fairly low levels.

I rarely see people talk about the following odd fact.  To illustrate the Japanese malaise you could simply draw the monetarist AD curve (a hyperbola) as fixed for 20 years, and then have the SRAS curve move right fast enough for RGDP to average 1% growth and the price level to average 1% deflation.  That’s an astounding lack of movement in the AD curve.  Perhaps the “worst” AD performance in post-gold standard history. (Or the “best,” as Friedrich Hayek might say.)

People sometimes forget that it is NGDP, not RGDP, which tells us what’s going on with demand.  RGDP reflects the interaction of both supply and demand shocks.  The RGDP per capita growth is low in Japan, but not off-the-charts low.  NGDP growth is off-the-charts low.

In a way I think this actually strengthens Duy’s argument.  Japan has run up some pretty large debts, even if one uses the smaller net debt figures.  And there is nothing to show for it except a world record lack of any movement in the AD curve for 20 years.  Why is that?  Not because fiscal stimulus can’t work.  But rather because it can’t work if the central bank tightens every time it looks like the inflation rate might rise above zero.

From this perspective fiscal stimulus seems almost beside the point.  If the BOJ tightens (as they did in both 2000 and 2006) every time there is a possibility of even a tiny bit of inflation, then fiscal stimulus in Japan will probably fail.  If the BOJ did allow even modestly higher inflation it might work, but in that case it probably wouldn’t even be needed, as higher inflation would (temporarily) lower both the real interest rate and the real exchange rate, boosting growth.

On the other hand the stylized facts also suggest that AD is not Japan’s only problem.  Remember that money is neutral in the long run.  I think Japan’s deeper problem is productivity.  I’m not surprised that Japan is poorer than the US in PPP terms, almost all developed countries are.  Its per capita GDP is similar typical Western European countries like Britain and France.  What surprises me is that the Japanese need to work many more hours than the French to achieve the same per capita income.  I’d guess there are rigidities in the domestic economy, perhaps due to special interest groups/rent seeking/regulation, etc.  For instance, do they have hypermarkets like the French, or do they rely on smaller shops?  These sorts of policies can’t be fixed with more AD.

So I end up with a wishy-washy view that will probably satisfy no one:

1.  They need supply-side reforms.

2.  Faster NGDP growth would help in two ways.  First, it would make real wage flexibility slightly easy to achieve, as there is evidence (at least in the US) of a discontinuity of wage changes at zero percent.  Second, slightly higher NGDP growth would move Japan above the zero rate bound, allowing the BOJ to do a better job cushioning the Japanese economy from an external demand shock like 2008.

I really don’t see how fiscal policy can deliver either of these needed changes.  Monetary policy determines trend NGDP growth, and the Japanese Diet is responsible for regulatory changes.

Interestingly, it’s easier to see this from a distance.  Very few American economists think that bigger Japanese deficits would solve their problems.  Bernanke thought it was obvious that they needed easier money.  However if I’m not mistaken the Japanese themselves don’t see it as being so obvious that easier money would help.  And of course when the US suffered a severe aggregate demand shortfall, Bernanke no longer seemed to see monetary policy as either the cause or a “cure all.”  At least that’s what he says publicly.  Up close excessively tight money always looks like something else–usually one of its symptoms.

PS.  Japanese NGDP fell 0.3% between 1992 and 2010.  I don’t have more recent data, so the “2 decades” statement was a slight exaggeration.  The Economist estimates that 2012 RGDP will be 1.6% above 2010 levels.  Nominal growth will probably be even less.

One often reads people talking about how investors poured lots of money into stocks, bonds, houses, gold, or some other asset.  Sometimes this fact is used to explain a big run-up in asset prices.  “The money had to go somewhere, didn’t it.”  This view is wrong, but it’s wrong in interesting ways, which reveal a lot about how people think about macroeconomics:

1.  First of all, at the most basic level it’s obviously wrong to talk about money literally “going into” markets.  If you visit Wall Street as a tourist and ask to see the big pot of money that people have invested in the stock market, they’ll give you a quizzical look.  Money goes through markets.  (Especially money created by the Fed; the monetary base.)

2.  Most sensible people who talk about money pouring into markets don’t mean it literally.  But even if the phrase is used figuratively, it is still wrong.  Indeed, let’s think about what people might mean by money pouring into a market.  Perhaps they mean that people are making lots of purchases.  Money is a medium of exchange, so lots of money is being used to facilitate stock purchases.  And it is true that on days when stock prices rise by an unusual amount, transactions volume of often higher than normal.  Unfortunately, that’s even more true of stock price crashes, which are often associated by an enormous volume of trading.  The famous stock market crash of 1987 was associated with record-breaking volume, but I don’t recall people saying investors poured money into the stock market that day.

3.  I have the same problem with people who talk about money being “saved.”  At the individual level that’s true, but my decision to hold on to more dollars means someone else is holding on to less dollars.  If we (collectively) increase our real holdings of base money, that is contractionary for NGDP.  But suppose we save in some other way.  If I buy a financial asset, someone else sells me the asset.  Suppose I buy a used house.  I pull $200,000 out of the bank (dissaving) and buy a house worth $200,000 (saving).  The seller gives up a $200,000 house (dissaving), and puts the $200,000 received by selling the house in the bank (saving.)  It’s never going to net out to any aggregate increase in saving unless new capital is built.  Now do the same example with a new house.  I pull $200,000 out and buy a new house.  Same as before.  The man who builds the house gets $200,000, so he actually increases his new saving by $200,000.  Aggregate saving in the economy goes up by $200,000.  What’s different is that the new house didn’t exist before being built, hence total wealth rises by $200,000.  Following the money is completely unhelpful for thinking about national saving, all that matters is whether new capital goods have been produced.

4.  Others point to the money created by the Fed, often called the monetary base.  Prior to mid-2008, this was composed of non-interest-bearing cash plus n0n-interest-bearing reserves.  Some argue that an easy money policy in the 2000s tended to boost real estate prices.  But why would this be true?  One possibility is that the Fed put lots of reserves into the banking system, and these reserves were somehow “lent out” for new home purchases.  In fact, roughly 95% of all new base money created in the years leading up to until mid-2008 was currency that went into people’s wallets, not new bank reserves.  This is even true in the short run.  Monthly increases in the base are overwhelmingly currency, with just a little bit of extra bank reserves (until late-2008.)  So it’s hard to see how this new “money” could have been pouring into the housing markets.  Money doesn’t have to “go somewhere,” 95% of it stays in wallets (prior to 2008).

5.  Others argue that the Fed held interest rates to very low levels.  But how did they do this?  Interest rates are set by the markets, not the Fed.  They can influence interest rates by injecting money into the economy, but if money was very easy shouldn’t we have observed a very rapid growth in NGDP?  Or at least a big increase in bank reserves, if you prefer that metric?  How do we know that rates weren’t low simply because the supply of credit schedule shifted right faster than the demand for credit schedule?  Why assume monetary policy is involved at all?

6.  Prior to 2008, the amount of new money created by the Fed is typically trivial, compared to the size of the economy.  And it was usually injected by buying T-securities, which is a huge and liquid market.  Thus it’s hard for me to see how the so-called Cantillon effects were important.  On a typical day the Chinese would buy more Treasury debt than the Fed.  If the Fed had bought German and Japanese debt instead of T-securities, then the holders of those securities would switch over and buy the T-securities the Fed was no longer buying.  The net effects would be tiny.

7.  Monetary policy can have a big effect on asset prices, but the effect has little to do with what assets the Fed is buying (until 2008, after that the purchases were so massive it might have marginally shifted interest rate spreads.)  My hunch is that if the Fed had bought index stock funds during 2002-06 instead of T-securities, the impact on asset prices would have been roughly the same.  If there was an impact, it was because monetary injections can raise expected future NGDP, and higher expected future NGDP can lead to an increase in the current prices of stocks, commodities, real estate, gold, art, etc.  But this doesn’t occur because money is a medium of exchange.  Suppose an auction house was silly enough to demand payment for multi-million dollar paintings in shares of Apple stock.  That might add 1% to the normal transactions cost for a wealthy investor buying a Van Gogh, but wouldn’t dramatically affect the price of the $100 million painting.  If the price of Van Gogh’s was soaring because an easy money policy was quickly raising future expected NGDP, those paintings would soar in value even if not paid for with money.  (Whether money raises NGDP via its medium of exchange role is a tougher question, where even market monetarists are split.)

8.  If wages and prices were 100% flexible then money would have almost no real effects.  Because wages and prices are sticky, money does have real effects on output and real asset prices.  But not because money “pours into” particular asset markets, rather because more money raises NGDP, and this raises real output (if wages and prices are sticky), which raises real asset prices.

Money matters, but not for the reason most people believe.

Saving matters, but not for the reason most people believe.

Money and saving affect the business cycle (if at all), by affecting either M or V.

My dad (who was politically liberal, BTW) used to say that he could never be “unemployed.”  He’d go door-to-door selling pencils, then take the money and buy some more pencils wholesale.  Indeed he made good money as a 12 year old in 1933, scrounging around for scrap metal and other odd jobs.  He also understood that not everyone could do this as well as he could, which is why he was an FDR Democrat.

Most people distinguish between “real jobs,” and ways of scrounging up some money when unemployed.  I’m pretty sure Tyler Cowen agrees with that distinction.  Where we disagree is whether sticky wages force people to move from real jobs to scrounging-up-money jobs.  Tyler has a post where he discusses an unemployed man’s attempt to scrounge up some money picking up empty cans, and then says this:

It is difficult for me to see how Murdock’s predicament “” is it a typical one? “” is well-described by the theory of nominal wage rigidity. I don’t mean to bait Scott Sumner, but I will again mention the difference between “nominal aggregate demand” and “real aggregate demand.”  If society were much more prosperous and people had higher real wealth-backed demands to buy a lot more products, Murdock probably could get a traditional job of the kind he is seeking.  In this sense you can attribute Murdock’s joblessness to a shortfall in aggregate demand.  That said, it is not clear why juicing up nominal variables should do very much for him.  His wage and workplace condition demands are already quite flexible, as he is willing to settle for what he can get.  Is money illusion his problem?  It seems there is no need to trick him, using monetary policy, into a lower real wage.

A few comments:

1.  In my view there is no such thing as “real aggregate demand.”  There is nominal aggregate demand, SRAS, and real output.  I understand that others disagree.  But “real aggregate demand” only confuses issues.

2.  I’d guess Murdock would like a shiny new pickup truck, and a job in an auto factory making $25 hour.

3.  He can’t get those things because of sticky wages.  But (and this is where people are confused) not because wages are sticky in the auto industry (they aren’t that sticky) but rather because aggregate wages are sticky.  That makes aggregate wages too high for full employment.  And if he goes to a factory offering to work for 1/2 the wage as an equally good current employee, they won’t hire him.  It’s not wage rigidity of the unemployed that is the problem, it’s sticky wages attached to jobs, not attached to people.

4.  If aggregate wages were flexible, the current level of NGDP (higher than in 2007) could support many more jobs.

5.  Of course wages are not flexible, but if NGDP was raised sharply then aggregate nominal wages would not rise as fast, allowing more jobs for Murdock making pick-up trucks, and more income to buy those pick-up trucks.  This is what “Say’s Law” is really trying to “say.”  If we can produce more stuff, society will be rich enough to buy that extra stuff.  No matter how much we produce, there is always some price level where we are rich enough to buy all the stuff.

6.  I’m wrong if structural factors like extended unemployment benefits have raised the natural unemployment rate to 8.5%.  (Unlikely.)  Interestingly, I’m not wrong if the 40% jump in the minimum wage a few years back raised the natural rate of unemployment, as monetary stimulus would lower the ratio of the minimum wage to NGDP.  That’s the key variable, not “real wages.”

I believe Tyler is wrong because he assumes that what he observes is inconsistent with the sticky-wage model.  Just the opposite.  Indeed if there weren’t people like Steve Murdock picking up cans I’d completely abandon my belief in the sticky-wage model.  If all the unemployed were watching TV, and none were trying to scrounge for cash, I might become one of those conservatives who believe unemployment is “voluntary.”

I think it’s useful to think of people who are unemployed as being forced out of the high productivity sticky-wage sector where they can work with lots of expensive capital to make pick-up trucks, into the flexible-wage scrounging economy where they can pick up cans, play guitar for tips, do odd house repair jobs, babysit for money, prostitution, deal drugs, etc, etc.  Because these jobs are less productive, RGDP falls.  And because they pay less than regular jobs, some people who aren’t desperate will rationally choose to watch some TV instead of working.  That would probably be my choice (internet, not TV.)

God bless Karl Smith!  If you don’t know, he’s one of the brightest and most respected econ bloggers, and he’s not an ideologue.

And he just endorsed my argument on consumption smoothing.  So I’m done.  If you have complaints, go put them in his comment section.  I shouldn’t have to carry the whole burden of Krugman attacks constructive debate for the entire right wing blogosphere.  My email box has so many comments spread out over multiple posts that I can hardly find my “important” emails from friends and co workers (Just kidding, comments are also very important to my blog.)

I’ll go back and read all the old comments today, but won’t respond to most.  If you really have something important, put it in the comment section of the “points 1-5” post.  But recently I’m just seeing people throw mud and hoping something sticks.  There are ground rules for comparative statics, and I think Paul Krugman, Wren-Lewis, Karl Smith and I basically agree on these rules.

My goal in the entire exercise was to bring Krugman and Wren-Lewis down from 100 to 50, and raise Cochrane up from 0 to 50.  I’ve completed the first project, now on to the second:

Here are some quotations from a John Cochrane paper that allegedly showed he didn’t understand Keynesian economics:

One form of “fiscal stimulus” clearly can increase aggregate demand. If the government prints up money and drops it from helicopters, this action counts as fiscal stimulus, since the money counts as a transfer payment.  In practice, our Treasury would borrow the money, and use it for tax rebates, subsidies, bailouts, or any of the many ways that our government sends people checks. Then the Federal Reserve would buy up the debt with newly created money.  The result is the same: A trillion dollars more money in private hands, just as if it had been printed and dropped by helicopters.  People naturally don’t want to sit on a trillion dollars of extra cash. They spend it, first creating demand for goods and services, and ultimately inflation.

This is perhaps the only prediction that is utterly uncontroversial among economists. It is a standard last-resort economic prescription to avoid a deflation.

Not very controversial, but this is:

Most fiscal stimulus arguments are based on fallacies, because they ignore three basic facts.

First, if money is not going to be printed, it has to come from somewhere. If the government borrows a dollar from you, that is a dollar that you do not spend, or that you do not lend to a company to spend on new investment. Every dollar of increased government spending must correspond to one less dollar of private spending.  Jobs created by stimulus spending are offset by jobs lost from the decline in private spending. We can build roads instead of factories, but fiscal stimulus can’t help us to build more of both . This form of  “crowding out” is just accounting, and doesn’t rest on any perceptions or behavioral assumptions.

Second, investment is “spending” every bit as much as is consumption. Keynesian fiscal stimulus advocates want money spent on consumption, not saved.  They evaluate past stimulus programs by whether people who got stimulus money spent it on consumption goods rather than save it.  But the economy overall does not care if you buy a car, or if you lend money to a company that buys a forklift.

Third, people must ignore the fact that the government will raise future taxes to pay back the debt. If you know your taxes will go up in the future, the right thing to do with a stimulus check is to buy government bonds so you can pay those higher taxes.  Now the net effect of fiscal stimulus is exactly zero, except to raise future tax distortions. The classic arguments for fiscal stimulus presume that the government can systematically fool people.

The central question is whether fiscal stimulus can do anything to raise the level of output.  The question is not whether the “multiplier” exceeds one – whether deficit spending raises output by more than the value of that spending. The baseline question is whether the multiplier exceeds zero.²

If you read his entire paper, here’s what he meant to say.  The first order effects, the DIRECT effects of moving money from A to B net out to zero.  You have more G, and less C+I.  You’d need some other indirect effects.  Like a change in velocity (i.e. real money demand.)

Point two is also slightly incorrect.  It’s not so much that Keynesians “want” more consumption, rather they “assume” investment is fixed, and that tax cuts will boost AD via more consumption, or else not at all.  That’s similar to my critique of Wren-Lewis, but not well stated.

If you think I’m just making excuses for Cochrane when I say (regarding point 1) that he was implicitly holding V fixed, consider the next quote:

My first fallacy was “where does the money come from?” Well, suppose the Government could borrow money from people or banks who are pathologically sitting on cash, but are willing to take Treasury debt instead.  Suppose the government could direct that money to people who are willing to keep spending it on consumption or lend it to companies who will spend it on investment goods. Then overall demand for goods and services could increase, as overall demand for money decreases.  This is the argument for fiscal stimulus because “the banks are sitting on reserves and won’t lend them out” or “liquidity trap.”

In this analysis, fiscal stimulus is a roundabout way of avoiding monetary policy. If money demand increases dramatically but money supply does not, we get a recession and deflation. If we want to hold two months of purchases as money rather than one months’s worth, and if the government does not increase the money supply, then the price of goods and services must fall until the money we do have covers two months of expenditure. People try to get more money by spending less on goods and services, so until prices fall, we get a recession. This is a common and sensible analysis of the early stages of the great depression. Demand for money skyrocketed, but the Fed was unwilling or, under the Gold standard, unable, to increase supply.

This is of course a Nick Rowe-type argument.  But I’ve never heard anyone complain that Nick doesn’t understand Keynesian economics.  The problem isn’t too much saving; it’s too much demand for money.  And the fiscal stimulus works, if it works at all, by reducing money demand.  Cochrane made a big mistake by not making this an “except if” comment right after his point one on fiscal fallacies.

But Cochrane is less monetarist than me, and thinks we need both more money and more government debt:

In sum, there is a plausible diagnosis and a logically consistent argument under which fiscal stimulus could help:  We are experiencing a strong portfolio, precautionary, and technical demand for government debt, along with a credit crunch. People want to hold less private debt and they want to save, and they want to hold Treasuries, money, or government-guaranteed debt.  However, this demand can be satisfied in far greater quantity, much more quickly, much more reversibly, and without the danger of a fiscal collapse and inflation down the road, if the Fed and Treasury were simply to expand their operations of issuing treasury debt and money in exchange for high-quality private debt and especially new securitized debt.

Sounds like Brad DeLong.  I’d focus on money alone, not money and T-debt.  And DeLong and I are not as worried about inflation down the road.  But the basic point is right.  If there is a NGDP problem, then we should boost nominal spending through some sort of monetary policy and/or debt swap policy, don’t waste billions on government spending programs that would not have passed a simple cost/benefit test.  Again, here’s Cochrane:

My analysis is macroeconomic, and does not imply anything about the specific virtues or faults of the Obama team’s spending programs. If it’s a good idea to build roads, then build roads. (But keep in mind the many roads to nowhere, and ask why fixing Chicago’s potholes must come from Arizona’s taxes funneled through Washington DC.) If it’s a good idea for the government to subsidize green technology investment, then do it. (But keep in mind that the internet did not spring from industrial policy to improve the Post Office, the word processor did not come from a public-private consortium to rescue the typewriter industry, and that a huge carbon tax is much more likely to spur useful green ideas, and the only way to spur conservation.) The government should borrow to finance worthy projects, whose rate of return is greater than projects the private sector would undertake with the same money, spreading the taxes that pay for them over many years, after making sure its existing spending meets the same cost-benefit tradeoff.

Exactly.

To summarize,  I’m certainly not defending Cochrane in the sense of saying he’s right, rather that he’s wrong in much the same way as Krugman and Wren-Lewis were wrong.  They all clearly understand their models, but sometimes made statements that are overly simplistic, and not accurate as stated.  Cochrane also seemed unaware of the fact that although new Keynesianism had abandoned fiscal stimulus in recent decades, some new Keynesians did favor it during a liquidity trap.

But inconsistencies happen all the time.  Krugman often says that open market purchases don’t work at the zero bound, and often says he supports programs like QE2.

One of my very first posts compared macroeconomic debate to a Tower of Babel.  One reason we think our opponents sound so stupid, is we use completely different frameworks (money vs. expenditure) and a completely different language.  And thus end up talking past each other.  But a sympathetic reading of Cochrane shows he knows his stuff, but expressed himself in such a way as to drive Keynesians up the wall.  He’d do better trying to persuade them using their own framework.

And of course this is just one more reason why we should all try to be more like Tyler Cowen.

Paul Krugman is impressed with Michael Woodford’s recent paper on fiscal stimulus, and I am too.  But probably for different reasons.  Keynesians like the paper because it shows that fiscal stimulus in the form of government spending can work under a surprisingly wide range of circumstances, even with “Ricardian Equivalence” (where tax cuts just get saved.)  On theoretical grounds I agree, although I have practical reservations, as does Woodford himself.  But I’d like to focus on some interesting points of overlap with my blog.

I often say “there is no such things as the multiplier.”  Not in the sense that it is zero or one, but rather that there is no stable relationship between government spending and aggregate demand.  Much of this is based on an argument that many people seemed to find quite novel, my claim that there was no such thing as “holding monetary policy constant.”  There is no “neutral” monetary policy, just lots of different things that could be targeted; the base, M2, nominal rates, exchange rates, inflation, NGDP, etc.  Here’s Woodford:

If prices or wages are sticky, monetary policy affects real activity, and so the consequences of an increase in government purchases depend on the monetary policy response. One might suppose that the question of interest should be the effects of government purchases “leaving monetary policy unchanged”; but one must take care to specify just what is assumed to be unchanged. It is not the same thing to assume that the path of the money supply is unchanged as to assume that the path of interest rates is unchanged, or that the central bank’s inflation target is unchanged, or that the central bank continues to adhere to a “Taylor rule,” to list only a few of the possibilities.

So any statement along the lines of “the multiplier” is 1.6 is pretty much nonsensical.

Another provocative maxim that I repeat over and over again is:

All multiplier estimates are nothing more than forecasts of central bank incompetence.

I think it fair to say that’s not something you learned in your intermediate macro class.  (Although I’m not claiming to have invented that idea, or any others that appear in this blog.)  In any case, let’s see what Woodford has to say about multiplier estimates:

Yet it is important to remember that in New Keynesian models, both the size of the output gap and the size of the multiplier will depend on monetary policy; and while there might well be significant opportunities for fiscal stabilization policy under the assumption that prices, wages or information are sticky and that monetary policy is inept, the most obvious solution in such a case is to increase the accuracy of monetary stabilization policy. Indeed, given that effective monetary stabilization policy should prevent there from being large variations in the ratio of u’ to v’ (by stabilizing the output gap), it is not obvious that the novel considerations [for fiscal stimulus] mentioned in the previous paragraph should be of great quantitative significance when monetary policy is used optimally.  (italics in original!)

‘Inept’ seems pretty close to “incompetent,” doesn’t it?  I’ve also argued that monetary [edit: I meant fiscal] policy should focus on efficiency, not stabilization.  Here’s Woodford:

Thus one may conclude that, regardless of the path of government purchases, an optimal monetary policy achieves the allocation of resources predicted by the neoclassical model. But then the condition for optimality of the level of government purchases is again simply (5.2), which is to say, the principle of efficient composition of expenditure.

I also emphasize that fiscal and monetary policies aren’t simply two different tools, each of which can be used for stabilization (as in the intro textbooks.)  Rather monetary policy is far more efficient.  Here’s Woodford:

It is not simply a matter of there being two instruments which can each, in principle, address the problem of an insufficient level of aggregate nominal expenditure, given the existing level of prices or wages, so that it does not matter which instrument is used for the job. Rather, to the extent that the problem can be solved using monetary policy, it is costless to do so, since monetary policy has no other aims to fulfill; whereas, while government spending can also be used to ameliorate the problem, this has a cost, since it requires the diversion of real resources to alternative uses. Whenever government purchases are used for aggregate demand management, there is a tension between this goal and the choice of government purchases so as to maintain an optimal composition of expenditure. Since there is no equally important conflict in the case of the use of monetary policy for aggregate demand management, monetary policy should be used to the extent possible; and this should largely allow decisions about government purchases to be made from the standpoint of the optimal composition of expenditure.  (italics in original.)

Comments:

1.  In other words, Joe Stiglitz is wrong.

2.  What does the phrase “aggregate nominal expenditure” remind you of?

3.  Ben Bernanke keeps saying the Fed has plenty of unused ammunition, but is taking a wait and see attitude before deciding whether additional stimulus is appropriate.  How would that information affect Woodford’s view of the relative desirability of additional fiscal and monetary policy?

Woodford does take the zero bound problem more serious than I do, and that produces the multiplier estimates that Krugman likes.  But there is also this issue:

In a neoclassical model (where real marginal cost can never differ from 1 in equilibrium), the increased tax distortion will lower equilibrium output, and may even negate the increase in equilibrium output that would occur with lump-sum taxation for the reason explained in section 1.

However, I can’t buy Woodford’s argument here:

In the case of an increase in government purchases while monetary policy is constrained by the zero lower bound, the multiplier would actually be increased if we assume that the increased government purchases are financed by a balanced-budget increase in the tax rate on wage income. The reason is that the increase in the tax wedge makes the policy even more inflationary, for the reason just explained. But an increase in expected inflation during the period while the nominal interest rate is constrained to equal zero will mean that real interest rates fall even more than in the analysis in section 4, resulting in an even greater increase in output.

I think what matters in NGDP growth, which isn’t boosted by adverse supply shocks.  I’d also point to the unhappy experience in Britain after the VAT went up recently (recall that a VAT is a consumption tax, just like a wage tax.)  I focus more on NGDP and sticky wages, whereas Woodford focuses on real interest rates and sticky prices.

The conclusion starts off by noting that fiscal stimulus may be appropriate in a major downturn like the Great Depression, but then ends by raising lots of caveats, such as this paragraph:

Nonetheless, under less extreme circumstances, the case for using variations in government purchases for stabilization purposes is much weaker. Even when the zero lower bound is a binding constraint, if the disturbance that causes it to bind is not expected to be too persistent, then even though the multiplier for increased purchases while the constraint still binds will be at least slightly greater than one, it need not be much greater than one; and the optimal increase in government purchases is probably only a small fraction of what would be required to “fill the output gap.” When monetary policy is not constrained by the zero lower bound, there is a good case for leaving output-gap stabilization largely to monetary policy, and basing decisions about government purchases primarily, if not entirely, on the principle of efficient composition of aggregate expenditure.

And finally, he even throws some support to those awful Republicans who blame President Obama for the job loses right after he took power, before the stimulus was spent.  Indeed he suggests they aren’t being awful enough.  Obama should have been able to influence economic growth almost immediately after being elected in November 2008, merely by announcing his fiscal stimulus policy:

It follows that when Eggertsson obtains a multiplier of 2.3, 1.0 of this is due to the increase in government purchases during the current quarter, while the other 1.3 is the effect of higher anticipated government purchases in the future.

Hence even if there were no increase in government purchases in the current quarter at all, an expectation of higher government purchases in all future quarters prior to date T would increase output immediately by an amount that is 1.3 times as large as the promised future increase in the level of government purchases.

I’d love to see Romney use that quotation in the debates this fall, but I don’t expect it.

PS.  Nick Rowe has a post on this paper, and Matt Rognlie has some comments.  I’ll be watching what they come up with, because it’s all way over my head.