Defense Spending

I mentioned as an aside in a comment on my last post that defense spending has basically been flat after adjusting for inflation and population growth. Somebody (well...nobody, really) asked me to provide citations for that claim. As I noted in the comment thread, I compiled the data from the Economic Report of the President, tables B-1, B-7, B-34, B-80, and B-85, and then subtracted out Federal grants to states to avoid double-counting. There are easier ways to do this (including the one detailed in my comment), but I did it this way because I'd already done most of the work with my earlier charts.

Real per capita government spending

The first chart has real per-capita government spending (federal, state, and local), broken down into defense and non-defense spending. Note that all of the increase in spending since 1960 has come from non-defense spending. The second chart shows spending as a percentage of GDP, again broken down into defense and non-defense spending. Defense spending as a percentage of GDP declined from over 9% of GDP in 1960 to 3% in 2000, and rose to 4% in 2005.

Real per capita government spending

Some caveats: First, the charts go only to 2003, because I had state and local figures only through 2003. In 2004 and 2005, both military spending and total spending increased both in real per-capita terms and as a percentage of GDP, but not by so much that the chart would look very different if they were included, and real per-capita spending on national defense was still 10-15% below its 1968 peak of $1787 (2004 dollars). Also, I assumed that all spending on national defense was done at the Federal level or through Federal grants to states. I looked at the breakdown of state expenditures in the Statistical Abstract of the United States and found no evidence to invalidate this assumption.

I bring this up for a couple of reasons. The first is to dispel any notions anyone might have about increases in military spending being a significant factor in the long-term increase in total government spending. The second is more general: I first ran these numbers a week or two ago out of curiosity, and was very surprised to find that per-capita military spending hadn't changed much at all in such a long period of time. The things we "know" aren't necessarily true, and it never hurts to do a bit of fact-checking.

I'm working on a more detailed breakdown of non-defense spending. It's a bit of a pain, because I have to subtract out all the grants to states to avoid double-counting, but I hope to be done sometime soon.

Share this

This is the kind of analysis

This is the kind of analysis that'll please nobody.

Thanks!

I have been watching this

I have been watching this trend for a while, and find your analysis in agreement with my own. In _Parliament of Whores_ (1992), PJ O'Rourke said that defense spending accounted for about 5% of the GDP and compared it to household spending on insurance. That hasn't changed much, has it? But all of the other spending keeps going up.

For a little fun, plot the poverty rate vs. spending on poverty. Or the total amount of money spent by local, state, and federal government on education vs. assessment test results over the past few decades. I started to do that in (warning: shameless blog-whoring) this, this, and this. And I'd look at this website, too; it seems to have some tacit endorsement from a Catallarchist ancestor.

To be fair, though, I'd still say that we spend too much on defense primarily because we are spending 5% of the world's overwhelmingly largest GDP. The total amount and the amount relative to the rest of the world are as important as the amount relative to our own economy. As I recall, we spend more than the next 5 or 10 defense budgets, combined. I'd also say that the problem is not the amount we spend on social programs, but rather the lack of any outcome. Compared to other countries (Japan, Scandinavia), we don't seem to be making the same impact they do. Why? Population homogeneity is offered as one reason, but perhaps also the way we approach the problem is not the best. Not that I advocate more gov't spending, or simply overhauling gov't programs while spending stays the same, but it is a question worth pursuing - why are there programs apparently more effective?

The flipside of that, again, is how good will their programs look 50 years from now when US GDP per capita is twice what it is in Sweden?

The staggering thing is that

The staggering thing is that 35 cents out of every dollar in this country is being spent to fix "the broken window", rather than create a new window. Government spending is always a net loss. Imagine the economic growth and wealth outcomes if we spent only 20 cents of every dollar on government, for example.

I wonder if anyone would

I wonder if anyone would care to defend the thoght behind framing the numbers in per capita or %GDP? The implication seems to be that a larger or more productive domestic populace should have a larger international military force (i.e., not cops) in absolute terms. Such reasoning escapes me.

[W]e spend too much on

[W]e spend too much on defense primarily because we are spending 5% of the world’s overwhelming largest GDP. The total amount and the amount relative to the rest of the world are as important as the amount relative to our own economy....
________

I wonder if anyone would care to defend the thought behind framing the numbers in per capita or %GDP? The implication seems to be that the larger or more productive domestic populace should have a larger international military force (i.e., not cops) in absolute terms. Such reasoning escapes me.

Ha! I hadn’t considered that. Maybe we should track changes in real defense spending adjusted for growth in OTHER NATIONS’ defense spending and OTHER NATIONS’ populations. Or just nations deemed adverse to ours?

Government spending is always a net loss.

...Rrrrright. And the uncompensated labor of child rearing is a net loss, too. Yet, inexplicably, the most productive places on Earth seem perpetually mired in the dual mistakes of having governments and having kids. If only we could emulate Beruit during the civil war - with no effective government and plummeting birthrates - just imagine the consequences for our GDP!

Eric H: Actually, we haven't

Eric H:
Actually, we haven't spent 5% of GDP on defense since the early '90s. Not sure about why our social programs don't work, but it's certainly not for lack of money---we spend about as much as or more than any other country in per-capita, PPP-adjusted terms.

Other Eric:
Net loss relative to what?

Cornelius:
The main reason I made the %GDP chart was to show that non-defense spending is up significantly as a percentage of GDP, and that a relatively modest increase in total government spending masks a significantly greater increase in domestic spending. I was really more interested in trends in the burden of defense spending than in whether or not we're spending the optimal amount.

But to answer your question, because of the income effect, it makes sense for military spending to increase in absolute terms (although perhaps not as quickly as GDP) as the economy grows. When a nation is poor, the marginal utility of money is high, and any extra military spending hurts. But as it becomes wealthier, it doesn't hurt as much, and it might make sense to spend a bit more, just in case. It's the same reason rich people hire guards while the rest of us just buy burglar alarms or put bars on our windows.

Of course, if it leads to other countries building up their militaries in response, it's just wasted.

nobody.really: Yet,

nobody.really:

Yet, inexplicably, the most productive places on Earth seem perpetually mired in the dual mistakes of having governments and having kids.

Nice job of putting words in my mouth. I'm not an anarchist in the sense you are implying. Read some Bastiat.

Brandon:

Net loss relative to what?

Relative to the money being spent by market forces. See Bastiat, or Hazlitt.

A dollar spent by the government is a loss economically. First, the person who originally had the dollar cannot spend it. Second, the government cannot spend it as efficiently. These two things add up to a net loss.

Some government services are necessary, although they could, theoretically, be provided by competitive "governments", rather than government monopolies.

For example, enforceable legal systems and contract law and national defense are things that a government of some sort is needed to provide. The net loss of those dollars is far outweighed by the net gain of a stable legal and economic environment and defense against barbarians.

On the other hand, taking 750 million dollars from the economy to relocate a railroad line on the Gulf Coast is a net loss that is not made up for in other areas. As is, for that matter, using government tax dollars to rebuild New Orleans.

Actually, we haven’t spent

Actually, we haven’t spent 5% of GDP on defense since the early ’90s.

Not sure what your beef is, Brandon. PJ said "about 5%" in 1991, and I said "it hasn't changed much". It's hard to read the exact percentage from your chart, but I'd say it's at *about* 3% and climbing, which in my book is not a big change from 5%.

Regarding social spending, that was my point (and I meant "their" programs :wall: ). We spend as much but don't seem to get the same outcomes.

Why, for example, have the Dutch gotten so much taller? It appears to be related to nutrition, but are there anti-poverty programs really that much more successful?

only a 1% increase since

only a 1% increase since 9/11? A rough estimate would then be 12T * .01 = 120B USD per year? Surely the wars in the middle east have cost a bit more than that? Is this counting the 80B emergency bills, of which i think there have been 3, and i think they are finishing up making the 4th? What about airport and port security, coast guard stuff, and other types of defense spending such as the creation of DHS or CIA NSA stuff? I have a feeling the government isn't being very honest about their numbers here?

Jesus ... there, their,

Jesus ... there, their, they're ... I cannot get them straight!

So their!

Eric - we don't have to

Eric - we don't have to imagine it. See http://www.freetheworld.com/papers/Gwartney_Holcombe_Lawson.pdf, particularly Figure 2, which shows average gdp growth rate as a function of %gdp spent on government.

It can be approximated by the line growth = 7.14 - 0.10 * (government spending in %).

Patri, the thing that always

Patri, the thing that always astounds me is that the US manages the economic growth rate it does with the level of government spending we have.

Your graphic goes well with

Your graphic goes well with this one:
http://www.neolibertarian.net/articles/cc_20060215.aspx

Brandon, Could you explain

Brandon,

Could you explain in some more detail which expenditures are being counted as "defense" expenditures? For example, are V.A. benefit pay-outs being counted as "defense" or "non-defense"? What about "Homeland Security" grants to state and local governments? Are 100% of debt interest payments being counted under "non-defense"?

Good point. This spreadsheet

Good point. This spreadsheet has the answers. To paraphrase: VA is non-defense ($70B in 2005), homeland security grants are non-defense but insignificant ($3B), and interest is non-defense ($184B net), but it's hard to say whether any/how much should be counted as defense spending.

I doubt that factoring in VA spending would make any perceptible difference. Depending on how you classify interest expense, you might get something interesting that way---maybe shifting defense spending up by 1% of GDP or a few hundred real dollars per capita---but the trends are going to look pretty much the same any way you slice it.

I'm probably not going to chart this out (I will do one that includes net interest as a separate category), but if you're interested in pursuing this question further, I can send you my spreadsheets in Excel or CSV format.

According to the CIA

According to the CIA Factbook 2006, the United States military spending represents about $1750 per capita - more than Israel's $1450 which used to be the highest in the world [*].

The US now spends more than 50% of the military budget of the planet, while its population represents less than 5% of the world population [*]. To have a better idea of what these numbers represent, imagine a schoolyard with 22 kids. There are rich kids and there are poor kids, but they all have weapons of some kind - sticks, knives, even guns. Now imagine one kid in the middle who, every year, spends on his weapons as much money as all the other 21 kids in the yard combined. That's the US.

[*] http://www.cia.gov/cia/publications/factbook/rankorder/2067rank.html
[*] http://en.wikipedia.org/wiki/List_of_countries_by_military_expenditures

...and the US (or that

...and the US (or that middle kid on the school yard) is pretty much why there is a "stable" world economy (or a relatively "safe" school yard).

It would be interesting, albeit probably impossible, to determine what monetary benefit the other nations of the world reap from the dollars that we spend on defense? Maybe we could bill them.