Keynesian spending is stimulative; tax cuts are recessionary?

So it seems, if you take Paul Krugman’s word for it.

To wit:

“A second step would be to accelerate the flow of government spending into the economy. One of the lessons of the Asian crisis, declared the I.M.F.'s Stanley Fischer, was that “Keynesianism is alive and well” — increased government spending does help the economy. And the additional spending that will take place as a result of the attack — $45 billion approved so far, much more to come — will eventually give our own economy a significant lift.”
“…Let’s get the government spending underway, please.”
“it’s very important not to respond to the economy’s weakness with a so-called stimulus package that locks in another round of permanent tax cuts.”

“As Mr. Greenspan pointed out, one risk to the economy is that a worsening long-run budget prospect will drive up long-term interest rates; that’s the last thing we need.”
Here’s a hint, Mr. Krugman-- indirect stimulative effects aside, a 1 dollar tax cut reduces the surplus by: 1 dollar.

A 1 dollar increase in spending reduces the surplus by <surprise>, 1 dollar.
Here’s another hint, Krugman: I defy you to show me any friggin’ positive correlation between interest rates and deficits in the last 40 years. If anything, the correlation has been negative. Low deficits tends to coincide with high interest rates.

Here’s another hint, Krugman: when it comes to determining LONG TERM interest rates, expectations of federal deficits are as nothing next to inflationary expectations. And our times of greatest inflation have historically coincided with Keynesian policies.
<<But in terms of immediate action, it may be more important that the money already in the pipeline be spent quickly than that it be spent well.>>

<rolling eyes>

Ok, how about we start with YOUR money, Krugman?

In terms of the multiplier, spending has a greater impact than tax cuts (with equal budget impact). The “shift” term in the IS/LM framework has a (1-s) in the denominator (where s is the marginal propensity to save).

The identification problem. Historical anecodotes are irrelevant. Show me theory.

What use is a theory that generally tends to predict outcomes that are different than what actually happens? Theory is only useful when it explains, rather than contradicts, the empirical evidence.

Fair enough water2j. My post was too cryptic due to approaching sleep.

“The identification problem” is shorthand for “you need to control for other effects”. Just looking at nominal interest rates and the reported deficit probably won’t tell you much about the relationship between interest rates and deficits. You need to adjust for inflation (i.e real interest rates), lags, the capital/ consumption of spending and the business cycle. There is continuing controversy about this point. If Ricardian Equivalence holds, then deficits are merely future taxation, and private sector interest rates reflect this. Ricardo didn’t believe in it, and although Greenspan appeals to it, I doubt he does in anything more than a crude way. In the absence of models it is hard to test evidence in economics, since there are too many variables and we can’t experiment at the macro level. I should have said “show me viable econometric evidence” of the relationship.

In the first sentence, the two are not independent. In the second, the relationship may be spurious (as well as - prior to JMK - trivially untrue).

I think we also need to look at the pbb (the propensity of the US Congress to blow money on pork-barrel boondogles*) when we calculate this out as well. I’m not enough of an economics whiz to set up the equations, there’s probably lots of complicated stuff I don’t understand. But I suspect if pbb > s then it might be a good idea to go with the tax cuts.

Whether or not pbb > s is an excercise I shall leave for others.

  • I’ll define pork-barrel boondogles as government investments that return less benefit to the American public than if the government had simply collected interest on it.

<<<In terms of the multiplier, spending has a greater impact than tax cuts (with equal budget impact). The “shift” term in the IS/LM framework has a (1-s) in the denominator (where s is the marginal propensity to save).>>

erm, wanna try again, minus the jargon?

I think the main point to the OP is that Krugman contradicts himself. On the one hand, he says we should crank up spending, but then he turns around and says that we shouldn’t cut taxes, because the prospect for higher deficits will damage the economy more than the stimulative effects of the tax cut.

The problem is that when it comes to how big a deficit will be, it makes no difference if you spent the money or never collected it in the first place.

And his claim that all it’s more important that the money gets spent quickly than intelligently is just stupid. In the very short run it might not make much difference, but it starts to matter more and more even in the medium term. Or else we could just take 100 billion and use 50 billion to dig a big hole and the other 50 billion to fill it back in again.

I like Krugman, and this column surprises me because I can rarely find fault with his basic analysis. Where I differ with him is that he tends to ignore the fact that private industry spends money more efficiently than does the government. And in the end, what really matters is not money, but how efficient your economy is at providing things that people want.

An interesting point, Sam. So are you willing to join the radical economists who are attacking the obsession with GDP because it is precisely such a poor measure…I.e., if you go out and fuck the environment and then spend money to try to clean it up, it counts toward the GDP both times!

Clearly, you haven’t been working for the same company as me. :wink: Bureaucracy is bureaucracy no matter how you slice it. I know the market does provide incentives for efficiency which should, in theory, help produce things more efficiently. Sometimes it works and sometimes it doesn’t. On the other side, you have various ways in which the market can fail, e.g., because of the extra money going off in terms of profits and exorbitant salaries, because of poor alignment between the interests of the buyers and sellers, etc. So, in the end, some things are handled more efficiently by the free market (like the manufacture of most consumer goods, …) and some things are handled more efficiently by the government (like building public infrastructure, providing health insurance, … I would argue.)

jshore: It’s been a while since I looked at exactly what is factored into calculations of GDP, but it wouldn’t surprise me at all to find out that some of the calculations are at least debatable. Numbers like that which have a heavy political content tend to get screwed around with over time, and often only start off as rough approximations of the thing they represent in the first place.

It’s like the ‘trade deficit’. Some things are included, others are not. Or the Dow Jones Average. None of these are important in and of themselves, and are generally used mainly because in the past they have been a good indicator of some larger, more difficult-to-measure trend. The Dow is constantly updated to reflect the fact that times change, while other numbers often are not.

I don’t know if that’s the answer you were looking for, but it’s the best I can do.

And I wasn’t trying to claim that EVERY industry is more efficient than government. I’m talking about the average here. In other words, if you gave the government a hundred trillion dollars and told them to build a country from scratch, and you had private industry make the same investment, the country that was developed through private industry would almost certainly be more efficient.

We generally recognize this in our society, which is why we reject arguments of efficiency as a reason for government intervention, as opposed to specific problems like market failures. Other countries have felt differently, and have happily nationalized many of their industries in the belief that government can do it ‘better’. Most of those countries in recent years have been just as busy privatizing all the industries they nationalized years or decades ago.

Insert post by dal_timgar bemoaning the fact that depreciation of consumer goods is not factored into GDP, complete with claims of planned obsolescence here.

I don’t think anyone is going to argue that GDP can’t be increased by ineffeciency. However, as long as efficiency is fairly stable and is not a major factor, I suppose we can all agree to live with the current system in the absence of anything better.

However, the government ought to try to get value for the money it is spending, otherwise, I don’t see government spending as being as useful as allowing the money to remain in the private sector.

Well, so I’m not too good with economic theories, but from this site: I found some interesting stuff in support of hawthorne’s analysis:

This seems to support hawthorne’s point about propensity to save. When times are bad, people tend to save, not spend, causing the multiplier to perhaps be less than 1 if you provide a tax cut. But government spending, regardless of how inefficient, is definitely spending, and the multiplier will be at least 1. And, more importantly, once the government spends the money, it is now in the hands of the public, creating the same situation as a tax cut. Thus your multiplier is almost guaranteed to be higher with government spending than through a tax cut.

Keynes doesn’t seem to say that tax cuts are recessionary. In fact, he does accept them as one of the “injections” into an economy. But he appears to think that they are not as effective as temporary government spending increases.

Krugman’s main point, which the OP missed, was that the tax cut’s costs increase over time. A spending increase can always be rescinded later, when the economy gets moving again, assuming it’s spent on something like the military response to this attack and disaster relief. The other point about spending being more stimulative than tax cuts has been covered by others already. (Before you get the wrong idea, I was for that tax cut, btw.)
Personally, I think we’ve been getting very, very lucky lately. The tax rebate money is still being distributed, I believe, even as both the government and private business are being forced to spend large sums as a result of this disaster. Combined with Greenspan’s interest rate cuts and cuts taking place around the world, I don’t see any way for the economy to avoid a rapid recovery from all this.
Or at least that’s what I’m hoping anyway.

<<In terms of the multiplier, spending has a greater impact than tax cuts (with equal budget impact). The “shift” term in the IS/LM framework has a (1-s) in the denominator (where s is the marginal propensity to save).>>

Ok, I thought a bit more about this “propensity to save.”

The thing is, I don’t buy the model.

People (and businesses) don’t save in a vacuum; they deposit that cash somewhere. And banks lend it out. In fact, the banks lend it out several times over, since they only have to keep 17 cents on hand for every dollar deposited. If banks have trouble lending out every last scrap of the remaining 83 cents on the dollar, they will compete fiercely to do so, and lower interest rates on loans to increase demand. But thanks to credit card holders and Home Equity Loans, I don’t think most banks have much trouble magnifying the effects of deposits.

If the government spends money it’s leveraged once. But if private industry saves the money, it’s leveraged several times over, because if you aren’t using your money, other people will be happy to spend it in the meantime.

The shift term has to take into account this multiplier–the rippling effect of liquidity spreading through the system.

Put another way, A dollar created as M1 and deposited as M2 exists several times over as M3.

A dollar saved in a tax cut likewise exists several times over as M3.

And a dollar spent by the treasury is eventually deposited somewhere, and likewise exists several times over as M3. The minus? If a dollar is spent by the treasury, that dollar has to be taxed. There is therefore less incentive for SOMEONE to produce something, somewhere. And THERE is your cost.

The bugaboo to economic recovery is NOT “the propensity to save,” but “the propensity to refrain from productive economic activity” in the first place.

pantom: <<A spending increase can always be rescinded later>>

2 points: 1.) So can a tax cut, by the same mechanism. Actually, tax cuts are far easier to rescind than spending increases, because once the government purchases, say, an aircraft carrier, they cannot “unspend” that money.

2.) Spending increases are almost never rescinded, in practice. Therefore, the cost of spending increase increases over time just as surely as a tax cut.

But wait a minute–taxes not collected in the first place don’t “cost” a damn thing!

I think that there are some issues that you’re missing here, panzerman*2

So how does this not occur if the people get the money because the government paid them for goods and services? That extra layer forces an extra round of spending, but in the end, private industry winds up with the money anyway, except that instead of simply being given the money, actual work gets done (which is what really matters anyway; if a lot of money goes around with no work behind it, it seems like we’ll wind up in a horrible inflationary cycle).

Ok, now I must admit that I am confused. Let’s look at an example with a single dollar, and a tax rate of 10 percent, plus let us assume that once money has been deposited into a bank, exactly that much gets spent one more time. Plus we will assume (correctly, I think) that a dollar does nothing for the economy until it is spent:

Tax cut method
Government gives me $1.00
Net effect to economy: $0.00

I spend $0.50, save $0.50
Net effect to economy: $0.50

You borrow that $0.50 and spend it
Net effect to economy: $1.00

Government spending method
Government pays me $1.00 to build them a dam
Net effect to economy: $1.00

Government taxes 10 cents, I now have $0.90
I spend $0.45, save $0.45
Net effect to economy: $1.45

You borrow that $0.45 and spend it
Net effect to economy: $1.90

So perhaps I’ve made some grave error in my assumptions (and if I have, I’m sure someone will correct me), but it seems to me that, even with a loss due to taxes, you come out ahead with government spending instead of a tax cut. This is because what actually matters is not the money, but the monetary value of the goods and services exchanged. So all the tax on your income does is reduce the amount you can spend later, rather than affecting the amount already spent. And thus you wind up with a greater multiplier effect on the economy.

On to more issues:

This is simply untrue. An aircraft carrier purchase is a one time deal. There is no need to “unspend” the money, as its sole purpose was to be spent. Now if it was a recurring cost (buying an aircraft carrier every year), then there might be a problem. But tax cuts are more likely (in fact, are certain) to be a recurring cost than material purchases like aircraft carriers.

The part you missed in that equation with the $1.00 is that the government only has money that it has taken out of the economy previously. For every dollar spent by the government there is a dollar not being spent by the private sector.
Also in terms of spending being easy to cut, you must remember that it costs money to staff and maintain the aircraft carrier. In order for this money to be cut the congressman whose district the carrier is based in must agree. This is why it is so hard to cut spending.

Not quite, if you look at the article I referenced earlier (, you’ll notice this:

Thus, the money being spent is not money being previously taken out of the economy (assuming true Keynesian practices, of course, which is what the point of this debate is, right?).

Although I think that even if it was money that had originally been taxed, the idea still holds, as this is now money that has to be spent, rather than being subjected to the vagaries of savings.


Agreed. I’m not arguing that spending is hard to cut, as it most certainly is.

However, the amount required for aircraft carrier upkeep is a significantly smaller amount of money than the upfront cost of the vessel. Thus, if the goal is to inject 100 million dollars into the economy, you can buy an aircraft carrier, and forever more spend 1 million per annum on upkeep, or institute a 100 million dollar tax cut, and forever more have 100 million dollars per annum that you are not collecting. (NB: numbers created out of whole cloth. Don’t take offense at any wildly inaccurate monetary values)

Deficit spending is not free money. It is borrowed money. That has economic consequences, such as higher interest rates. This might be worth it if we are facing something that needs to be accomplished today, such as a war, but it is not free.

  1. We’re in times of economic uncertainty.

  2. Corporations are reacting by having more cash on hand, rather than spending money on plant and equipment, because they don’t know that anyone’s gonna buy more stuff if they make it.

  3. Consumers are reacting by cutting back on discretionary spending, because they’re worried about layoffs and the likelihood that it’ll be harder to find new jobs if they lose the ones they’ve got.

If you cut taxes under these circumstances, what happens to the tax cut?

  1. The corporations and the consumers are cutting back on unnecessary spending, so it goes into the bank.

  2. The Fed and the banks cut rates to try to loan it out, but until people and corporations are spending again, that doesn’t mean it actually gets loaned.

This is called ‘pushing on a string’.

If the government spends money on some sort of domestic infrastructure (e.g. spends $20 billion to build new public schools), what happens?

  1. The corporations that will do the work will have money to spend on equipment they will need to do the job.

  2. Jobs will be created, resulting in less unemployment. This means:

  3. Some people will be working, rather than on welfare, in school, or whatever; and

  4. Other people with jobs will be a little less worried about losing them. As a result,

  5. Both these groups of people, along with the corporations mentioned in (6), will spend more money.

  6. Which will ripple through the economy in the usual manner.

This is why the government spending money directly is more effective, under these circumstances, than spending it indirectly in the form of a tax cut.

What causes consumer spending to freeze up at a time like this usually isn’t so much the idea that they have an income, but could use a few less dollars going to Uncle Sam - but rather the possibility of losing their job and not quickly being able to find a comparable one.

Tax cuts do nothing to alleviate this problem. Jobs do.