New papers on fiscal policy

by on October 28, 2010 at 1:50 pm in Economics | Permalink

Price Fishback has a new paper (ungated but earlier draft here) surveying what we have learned about the macroeconomics of the Great Depression, here is one bit:

Some thinks that World War II offers an example of a situation where fiscal stimulus worked.  But the World War II analogy is highly misleading for any discussion of a peace-time economy.  The deficits were run during an all-out war when 40 percent of GDP was spent on munitions, the military made most of the allocation decisions in the economy, over 15 percent of the workforce was in harm's way in the military, there were widespread wage and price controls, and rationing ruled the day.  In essence, the World War II deficit experience tells us more about fiscal stimulus in the Soviet Union's command economy during the Cold War than it does about the modern U.S. mixed economy.

From Ethan Ilzetzki, Enrique Mendoza, and Carlos Vegh, here is a new paper (gated) on fiscal multipliers (shorter, ungated version here, powerpoints here, slides here, ungated but slightly older version here):

We contribute to the intense debate on the real effects of fiscal stimuli by showing that the impact of government expenditure shocks depends crucially on key country characteristics, such as the level of development, exchange rate regime, openness to trade, and public indebtedness. Based on a novel quarterly dataset of government expenditure in 44 countries, we find that (i) the output effect of an increase in government consumption is larger in industrial than in developing countries, (ii) the fiscal multiplier is relatively large in economies operating under predetermined exchange rate but zero in economies operating under flexible exchange rates; (iii) fiscal multipliers in open economies are lower than in closed economies and (iv) fiscal multipliers in high-debt countries are also zero.

That's zero as in "two times zero equals zero," or "1/2 times zero equals zero too" or even "-0.5 times zero equals zero."  I don't think the multiplier is zero for the United States, but very likely it's zero in a bunch of places.

1 Andrew October 28, 2010 at 10:47 am

I'll be patiently awaiting the apology from Paul.

2 Yancey Ward October 28, 2010 at 10:52 am

Where has Bill gone?

3 Jason October 28, 2010 at 11:16 am

I'm not sure I get how a fiscal multiplier can be zero. Small, yes. Tiny, even. But zero, unless I am mistaken, means the money just completely evaporated without even the contractionary effect of just setting it on fire.

4 LoneSnark October 28, 2010 at 11:52 am

"I'm not sure I get how a fiscal multiplier can be zero."

If the government borrowed every dollar it spent, then it should be close to zero, as every dollar it spends was a dollar that was probably going to be invested anyway. To put it another way, it is conceivable that if the government had not borrowed this dollar from, say, the Chinese, the Chinese would have spent the money on one dollar worth of goods instead. $1 less spent by the Chinese and $1 more spent by the government; net effect, zero.

The only mechanism I know of for government spending to have more than a zero multiplier is borrowing from the Fed's printing press. But there is nothing stopping the Fed from buying bonds with the printing press money. As such, even there, every dollar spent by the government is one less dollar spent by the Fed; net effect, zero.

5 Tony October 28, 2010 at 12:31 pm

LoneSnark's argument breaks down in a situation like today, where banks and corporations have a lot of cash but aren't investing it at all.

6 Mike C October 28, 2010 at 2:47 pm

I can see four scenarios that would cause a low multiplier:

1) The borrowed dollar is not spent but is stagnant.

2) The borrowed dollar was spent on an import.

3) The borrowed dollar was spent but was handed off only once. Very similar to #1.

4) The borrowed dollar was transferred to pay down other debt.

7 Pragmaticon October 28, 2010 at 2:56 pm

At first I was like "haha yeah, the implications of that statement are insanely flawed, nice job sarcastically pointing that out", then I saw that the comment was written by mulp, realized that he was probably being serious, and quietly wept over the sad state of our primary education system.

I don't blame you- you're the real victim hear, friend. God bless you for trying.

8 J Thomas October 28, 2010 at 3:12 pm

"Let's see, monetary policy doesn't work as theorized."

Theorists don't understand everything.

"Fiscal policy doesn't work as theorized."

The other theorists don't understand everything either.

"But a total government takeover of the economy, high taxes, high debt, and government command and control for five years will create the conditions for decades of strong economic growth that can pay the high taxes to drastically slash the debt burden."

Remember that we made lots of bombs to blow up other people's capital goods, while our consumer economy etc deteriorated. If we had put that effort into infrastructure we'd probably have stagnated for the next 50 years.

But then I don't know everything either.

9 Right Wing-nut October 28, 2010 at 8:16 pm

Oops. Guess the filter even drops non-existent tags. That was supposed to be

</irony> tag added for Bill & mulp.

10 andy October 28, 2010 at 10:23 pm

Can the fiscal multiplier be negative given the way how the GDP is computed? (assuming the government expenses do not destroy things)

Y = C + I + G + NX

Assume the crowding out effect is 100%.

Scenario 1) The government runs a deficit and spends the additional money on transfer to the people. As 'G' does not include transfers and C/I/NX is not affected, the resulting Y' is the same as Y; multiplier is 0

Scenario 2) The government runs a deficit; spends additional money on 50% overpriced services (overpriced in the sense that if a private individual bought the same services, he would get it 50% cheaper). It's basically a partial transfer to the providers of these services. The whole amount is included in the G but C/I/NX is smaller only 50%. The fiscal multiplier is 0.5.

Is it correct? If it is, one would assume 'the more corrupt a government is, the higher the multiplier gets'….

11 Doc Merlin October 29, 2010 at 1:34 am


Well, Keynes did suggest paying people to dig ditches and others to fill bottles, then others to bury the bottles in the ditches.
Seriously… fiscal Keynesianism is absolutely insane.

12 josh October 29, 2010 at 4:09 am

Tyler, I think you are an honest man, but you are a quack in a field of quacks selling quack medicine.

13 andy October 29, 2010 at 6:14 am


I think the second quote pretty much says that he DID think that.

14 andy October 29, 2010 at 7:44 am

@J Thomas

It seeems to me that the very notion 'it would be better than nothing' is not ironic at all; he really meant it. And that is exactly the problem; it is very hard to conceive it would be better than just e.g. giving the money to the people without requiring them to dig up the bottles…

15 andy October 29, 2010 at 9:40 am

@J THomas
I don't think he had that in mind. You know, it wouldn't raise GDP 🙂 (ok, he hadn't this in mind as they didn't do much GDP computation back then) But – who knows 🙂

16 Dan H. October 29, 2010 at 1:18 pm

The latest paper from Christina and David Romer on the cost of taxation found that exogenous tax increases of 1% of GDP could cause a fall in GDP of 3%. A startlingly high number that even surprised them. However, there was a glaring exception: Tax increases used to pay down deficits showed no contractionary effects at all.

I've been pondering the implications of that for a while. What does that say about the cost of deficits to the economy? And why?

Now compare that to the second paper linked by Tyler, which found that fiscal multipliers don't exist in countries with high debt levels. That's entirely consistent with what the Romers found. I think this would be a very fertile area for more research.

One thing it does suggest is that Paul Krugman is wrong to suggest that a larger stimulus would have worked. If the stimulus becomes increasingly inefficient as debt goes up, then the marginal value of each borrowed stimulus dollar goes down just from the debt effect alone. Then consider that each additional dollar of stimulus must have a lower multiplier due to the added difficulty of spending that much money rapidly and the resultant increase in crowding out effects, and it starts to become very problematic for Dr. Krugman's point of view.

17 strawman October 29, 2010 at 2:54 pm


Which is why keynessianism is insane. It expects that too many people are too stupid.

Interesting you say this, as I've always come from the other viewpoint. The idea of a general contraction in demand seems to me not only sane, but in perfect keeping with historical human behaviour. Herd mentalities and irrational panics prevail throughout much of history.

I'm equally surprised by economic theorists who posit that 9.6% unemployment is caused by 14 million American workers deciding to take a break from work, or that structural misallocations in construction can cause employment to fall in almost every industry. I might not term those ideas "insane", but I'd certainly consider them willfully ignorant of reality.

18 J Thomas October 29, 2010 at 7:22 pm

Mulp, I will present my general theory of economic development now.

People come up with economic theories that they think apply generally. But in reality, the most important economic facts are things which are imposed on us and which nobody in particular can do much about. For most of history, across much of the world, one of the most important things was the crops. Get a drought, or too much rain during harvests, and the poorest people starved. Get crop problems two years straight and many poor people starved. Three years straight and it was famine time.

Things happen like baby booms, and aging boomers, or a whole population that insists on saving more than there are useful investments to back. It's hard to change things like that. China tried to enforce a one-child rule with forced abortions etc. That sort of dedication is rare.

So economies face challenges that they cannot eliminate but must adapt to, and then the challenges change and the adaptations must change too. Economic theories that people thought provided the answers for the ages stop working. New theories start to work, theories which would have been inappropriate before.

Horses for courses.

I say that our current economy faces a whole lot of challenges at once, and we need some way to deal with them. We need to loosen the limits on the various limiting factors.

One issue is energy. Unless we find a way to go back to cheap energy, energy constraints will be limiting. We had a chance at that in the late 1970's and we chose not to. Instead we intervened in the middle east, using diplomacy and weapons to keep oil cheap. That approach is running its course, and anyway as our balance of trade gets worse oil imports will tend to stay expensive. Things we do to keep foriegn oil cheap are likely to benefit China more than they benefit us. How can we get cheap energy? Nuclear energy looks like it will stay expensive. At the moment so does everything else.

Another constraint is our aging population. We have to find cheaper ways to take care of old people, or else encourage them to die. We did not cultivate the diplomatic relations to send large numbers of aging Americans to north africa where labor is cheap, to be cared for in a better-than-Florida climate. Perhaps we will develop versatile automated care facilities so that fewer skilled people will be required to run nursing homes etc. I dunno.

In general we need industrial automation. Human beings on assembly lines etc are expensive and error-prone. But humans have low fixed costs. Hire them when you need them, lay them off when demand is low. Automated equipment is expensive and inflexible and has a short effective life. Human beings are expensive, and Americans are seldom the cheapest. Design more flexible equipment and work out ways to lease it? Then there's the matter of jobs for Americans. The more jobs we can automate, the more we need jobs that machines can't do. How will we find such jobs? Many executives could be replaced by random number generators, but what would we do with them after they are replaced?

We need to replace our financial system with something that works. But how can we do that when they own the government?

Etc. A big collection of interlocking problems. Can we set up institutions that are better able to solve the problems? Can we set up institutions that are better able to get out of the way of solutions? How about institutions that are better able to stop nonsolutions disguised as solutions, that would make a few people rich?

19 andy October 29, 2010 at 10:21 pm

@ J Thomas

I'm equally surprised by economic theorists who posit that 9.6% unemployment is caused by 14 million American workers deciding to take a break from work, or that structural misallocations in construction can cause employment to fall in almost every industry.

The keynesians posit that 9.6% unemployment is caused by people employers stupid enough to lower nominal wages, or employees stupid enough not to lower their requirements (the second seems to me a stronger connection as there are actually quite many employers who even in the studies are not that stupid). And debt-deflation theory is Fisher's, not Keyneses.

20 J Thomas October 30, 2010 at 7:45 am


You are replying to Strawman, literally, not to me.

It's been awhile since I read Keynes and then I didn't read all of Keynes, so my understanding includes a lot of what other people said about him. And anyway, I tend to care less what he actually said than what sense I can make of it — if he said something right and something else wrong, I'd rather use the part that's right and discard the rest.

As I understand it, he said basicly that the system includes feedback loops, so it doesn't make sense to say one part of the loop is the cause and another part is not. Your furnace has a thermostat that's designed to hold the temperature in a cycle. When the thermostat gets cold it turns the furnace on. The furnace then warms the thermostat enough that it turns the thermostat off. Which one is the cause?

Employers laid off employees not because they were too stupid to offer them less money, but because they believed that demand was down. If demand is down 50% and you expect it to stay down for some time, why pay everybody to worh half-time when you can pay half the people to work full time? And you can negotiate wages down for that half, if you're hardnosed.

Employers thought demand was down so they quite reasonably laid off employees. Laid-off employees tried to economise and not buy as much, so demand went down more.

I think that was one of the feedback loops that happened during the Great Depression.

But it isn't the same now. Employers wait awhile to lay off employees because it's expensive to train new hires if business picks up. Also some of them don't pay into Unemployment Insurance so they have a big expense if they do fire somebody. They bet and they lost. When an employee is fired he gets unemployment insurance which helps some, and he can spend his savings, and he can run up his credit cards. Of course he doesn't want to do that, but he can hope to find work soon and it's easier to hope than to economize. So when employment drops, demand does not drop as fast as it used to.

Also, some well-to-do americans have some tendency to buy american products. When they feel poor enough to start shopping at WalMart etc, their money has more of a tendency to buy imports than it used to. Maybe. To the extent that's true, as americans get poorer the trade deficit gets worse.

And it doesn't work for the government to give money to the poor. Keynes suggested that the government should raise taxes and reduce spending in good times, and lower taxes and increase spending in bad times, so that the taxes tended to even out. The government could pay for poor people to survive the bad times, and collect some taxes from them in good times, and again it would even out.

But we already reduced taxes in good times and now we want to reduce them more in bad times. And the government can't hope to give poor people enough to live on and pay for it from the taxes on the people who do the work of importing and distributing the cheap foreign goods. Every time around that cycle we lose more money to the foreigners who actually make stuff.

The obvious solution is to become a third-world nation. Let our poor people have a lifestyle of utter poverty, let the much-shrunken middle-class have a lifestyle equivalent to the majority of chinese workers, and compete with China and India on an equal footing.

We can do that. It doesn't take any special effort, we can get there easily, by following any economic policy that doesn't manage to avoid that result. It's the ground state.

I don't like it. Lots of people don't like it. It would not be acceptable as a political platform. I doubt anybody has a good alternative.

If we let the free market work its magic, it will mostly destroy the US middle class. There is no economic reason for them to have so much money. They are not competitive and there is no obvious way for them to become competitive.

We once had an idea that we could live off of Intellectual Property. But few US citizens own patents.

We were going to transform the economy with the Internet. Of course, the Internet allows commodities to be sold with less work done, although the temptation to pay for quick delivery increases aviation fuel imports.

People hoped they could sell their houses to make money, and that worked for awhile. But a community can't survive off selling their houses to each other.

I see two possibilities. One is to have a trade war, and live without many imports. Then we learn to make everything we need for ourselves, and we have whatever standard of living we can arrange that way. Whatever we work out for ourselves, it won't be anybody else's fault. We'd be doing the best we could with what we had.

Another is to create technology etc that would let a chinese and indian etc middle class live like our middle class. If we can make it a lot cheaper to live like we do, then they can do it and we can do it too.

Or we can sink.

21 andy October 30, 2010 at 8:08 am

Employers laid off employees not because they were too stupid to offer them less money, but because they believed that demand was down. If demand is down 50% and you expect it to stay down for some time, why pay everybody to worh half-time when you can pay half the people to work full time? And you can negotiate wages down for that half, if you're hardnosed.

Employers thought demand was down so they quite reasonably laid off employees. Laid-off employees tried to economise and not buy as much, so demand went down more.

While I think there is something to the debt-deflation theory, I find your explanation pretty useless. The firm's profit is given by revenue minus costs. Costs fail, especially if you are able to lower the wages. Costs of land/materials fall too. No problem there. And Keynes probably knew, that's why he introduced the 'nominal wage stickiness theory'. If wages are sticky, than they cannot fall alas you have unepmloyment.

While theoretically this explanation could work (and some studies indeed do confirm some wage stickeness), it basically concludes, that everybody is too stupid to cut wages… And somehow the explanation that the employers are unable to lower total worker compesnastion by 2-3% a year seems a little bit odd..

Every time around that cycle we lose more money to the foreigners who actually make stuff.

Before uttering this you should have looked at the industrial exports of the USA. A few years ago it was all time high, it's a little bit lower now because of the recession. Surprise, isn't it…

Comments on this entry are closed.

Previous post:

Next post: