« Keynes's General Theory, chapter four, *The Choice of Units* | Main | Financial innovation hasn't stopped, it's just being done by the government »
The multiplier is "one"
And that's now a blog, written by Susan Woodward and Robert Hall, both impressive thinkers. Greg Mankiw offers additional comment.
Posted by Tyler Cowen on December 11, 2008 at 08:00 AM in Economics | Permalink
Comments
If that's the case, then I would think the best plan would be to boost the SBA loan guarantees. This would provide much leverage, create jobs and let the market decide where to invest. In fact, you could have the Fed (or some institution) pay interest on these loans as they now pay interest on reserves (which seems to me to be an incentive not to make loans).
Posted by: Nick Owen at Dec 11, 2008 8:27:36 AM
It is nonesense to think that the infrastructure "multiplier" is equal to
one IN GENERAL. As with all governmental expenditures, the curse or
blessing is in the details of what exactly the expenditure is for.
Roger Garrison has a great book on Austrian Capital Based Macro that
I think does a great job making this point. People who have not read
it, would do well to educate themselves to preclude opening themselves
to the shame that might be visited upon them by those of us who have.
Posted by: indiana jim at Dec 11, 2008 8:35:13 AM
From Mankiw:
"By contrast, recent research by Christina Romer and David Romer looks at tax changes and concludes that the tax multiplier is about three: A dollar of tax cuts raises GDP by about three dollars."
If Govt is 20% of gdp, then does it not follow that a tax cut pays for 60% of itself. This is even higher than Mankiw's own estimates of 50% for capital gain tax reductions and 33% for income tax reductions.
Even if it were only 50%, two bucks in my pocket is better than one in the fed's.
Posted by: Tom at Dec 11, 2008 8:40:24 AM
It seems inappropriate to calculate the multiplier of an economy not in recession and then assume that the multiplier is constant and unchanging.
Posted by: Charlie at Dec 11, 2008 9:04:21 AM
For military spending. So we have decided that fighting wars is a bad way to use money, even in a depression.
Posted by: yoyo at Dec 11, 2008 9:12:23 AM
Contrary to what Prof. Mankiw writes the reality is that we do not know how much the multiplier effect is and we cannot know in advance because economy is dynamic. The only distinction should be between productive spending and not very productive. To compare military spending and infrastructure is in this respect misleading to say the least. In an open economy a major difference exists for the multiplier if the sectors are tradeable and the spill overs are all domestic. Again you can always increase military spending and make wars abroad. That's why infrastructure is likely to have higher multiplier than military spending being also more productive...
Posted by: Massimo GIANNINI at Dec 11, 2008 9:22:50 AM
I am sorry but I can't believe we're having this discussion. Do we not believe in scarcity? If you want to make an argument about idle resources, OK, but that's not what the main argument is about.
I mentioned this in the other thread, but in section 17. C. here, Rothbard (in my opinion) blows up the whole "multiplier" argument.
My own suggestion is that the government order shipping companies and airlines to stop changing the oil in their vehicles. Then depreciation will go up, and since this is a component of GDP, it will boost national output.
Posted by: Bob Murphy at Dec 11, 2008 9:50:56 AM
Whoops, here is the Rothbard reductio ad absurdum about the multiplier (section 17. C).
Posted by: Bob Murphy at Dec 11, 2008 9:52:02 AM
To discuss "infrastructure" expenditures decided upon by government relative to "military" expenditures by government is pointless in the absence of details as to such things as: 1) the types of each expenditures ("what is spending on, not in general, but specifically?"); and 2)the contextual situation in which expenditures are to be made (e.g., "where are resources expended").
There are no such things generic "infrastructure expenditure" or "military expenditure". This means discussions at this level of generality are vacuous.
Posted by: indiana jim at Dec 11, 2008 10:01:25 AM
Susan Woodward and Robert Hall assume the money multiplier is not time varying. This was Keynes original criticism of econometrics. I find their assumption unbelievable. I see no reason to believe that the economy today is the same as it was during WWII.
Posted by: assman at Dec 11, 2008 11:23:56 AM
It may well be that for reasons given, certain
sorts of tax cuts might in fact have bigger
multiplier effects than spending increases.
However, the label on this is simply off. Going
to Mankiw's site, one finds that Ramey's finding
is of a multiplier of 1.4. Where one finds little
multiplier is in WW II, when (hack, cough) we had
rationing of consumer goods going on. In the early
50s one can see GDP rising more than military spending.
Posted by: Barkley Rosser at Dec 11, 2008 11:53:14 AM
If you assume the private sector actions are generally better at growing the economy than the government actions, then it should come as no surprise that the "multiplier" effects of tax cuts (increasing private spending / allocation decisions) is greater than the "multiplier" effects of govt spending (increasing public spending / allocation decisions).
I interpret the 3:1 ratio as a measure of how much more effectively the marginal private dollar is spent to the marginal public dollar is spent.
Posted by: Jody at Dec 11, 2008 12:56:50 PM
Thanks for the reference!
NS
Posted by: notedscholar at Dec 11, 2008 1:36:36 PM
There are this analysis is fraught with problems.
1) It assumes without justification that military spending in the 40s has the same impact on the economy as infrastructure spending today.
There are two reasons why this is way off. First, there was heavy rationing in the 40s, so an increase in government spending of course led to little increase in consumption. Second, infrastructure spending lowers supply costs in most industries today and into the future, while military spending does not.
2) It assumes that the multiplier is constant over time and point in the business cycle.
They specifically look at increases in spending that were NOT an attempt to stabilize the economy, showing that they didn't significantly put idle resources to better work. (Well, duh.) They then conclude that no type of spending can increase GDP significantly even when there are resources sitting idle. Huh? These are two very different states of the economy. Of course stimulus won't be effective when the economy is at full capacity, but this doesn't mean stimulus is never effective.
Posted by: brian at Dec 11, 2008 2:22:15 PM
My confidence in government "stimulus" spending being "efficient" in any sense of the word is nil. Surely I am not the only one who watched how TARP was put together (I recall money being allocated for bows and arrows or something similarly ridiculous, to name just one example), or how it has been implemented.
Not to mention what soon to be very powerful special interest groups have tried but heretofor failed to get included in TARP.
Personally I think the US government should set up a temporary Sovereign Wealth Fund (SWF), funded as follows and managed by private sector professionals:
Issue new US Treasury debt (likely ~0.05% yield). Stick that money in SWF's that are chartered to be dedicated to be invested in fixed income securities with comensurate maturities. For total return only(!), as opposed to additional mandates for things that have alleged positive externalities, noting that these thing can and likely will be inefficient when decided upon by politicians.
Fund managers to be drawn from mutual fund land and money market fund land, depending on experience and (debt) maturity. Manager compensation to be similar to private sector.
Rinse and repeat until we run out of "mattress money", which is to say until T-bills, notes, and bonds have "normal" yields relative to corporates. Once this eventually happens, simply phase out the SWF's.
Three things will happen. Taxpayers make a juicy arbitrage profit. "Mattress money" gets soaked up. Businesses (and others, directly or indirectly via private sector arbitrage) pay lower interest rates than otherwise, which results in considerably more "green field" investments and hiring than otherwise.
All of which are likely (no duh) to be more efficient than sending it through the usual government pork grinder. Not to mention a different profit/loss sign for the government's (i.e. taxpayers) balance sheet.
Posted by: happyjuggler0 at Dec 11, 2008 3:17:25 PM
Military spending is flat until mid way through 1939.
GDP growth starts at about mid way through 1937.
So for 2 years GDP is growing without any help from a military stimulus.
There does appear to be a small corresponding slope change to GDP growth
right when military spending takes off.
But it looks like the claim of a multiplier of one is based on military
spending going from a very small amount pre 1940 to 7000 right around 1944
and the economy growing from about 3000 to 10000 during the same period.
But for two years prior GDP was trending upwards.
Surely the correct multiplier would be based on the slope difference?
Which looks like it would make the multiplier about .57. Which is still
pretty good.
But I didn't read the paper and I am definitely not an econometrician.
Posted by: someguy at Dec 11, 2008 3:38:54 PM
@ Bob Murphy,
I am finding one multiplier that is working quite well. Since I noted commenters at another MR post writing that they were checking the comments only to see what you would say, I am finding I am checking the comments to find out what they are saying about you are saying.
You are your own multiplier effect, my man.
Posted by: Robert Wenzel at Dec 11, 2008 9:49:46 PM
...I am finding I am checking the comments to find out what they are saying about you are saying.
You are your own multiplier effect, my man.
Except that I am causing output to fall.
Posted by: Bob Murphy at Dec 11, 2008 10:41:21 PM
I think the value is much less than "one". Government bureaucracy is less efficient and less capable at creating value than competitive private industry. Thus, having more money go into government-directed programs will lead to less economic growth than if the money is simply returned to the tax payers as a rebate.
Posted by: kurt9 at Dec 12, 2008 6:16:23 PM
"I think the value is less the one."
Wow, very impressive, kurt9. And do you have any data to support your
assertion, or are you just allowing us to bathe in the glory of your
self-thinking wisdom?
Posted by: Barkley Rosser at Dec 13, 2008 12:29:07 AM
