Military Spending in the G7: A Quick Post NATO Summit Comparison

Well, the Disruptor-in-Chief, US President Donald Trump is at it again.  At today’s NATO summit he demanded that the members of NATO need to quickly increase their defence spending not only to meet the committed target of 2 percent of GDP but to also double their commitments to 4 percent of GDP.  Needless to say, all of this presented an opportunity to try and take a look at some of the numbers to see where things actually stand and to put the numbers into some kind of longer term perspective.

So, I have put together data on military spending, GDP and central government expenditure for the G-7 countries.  For Canada, I have data from Historical Statistics of Canada, the Urquhart and Statistics Canada GDP numbers and the Federal Fiscal Reference Tables.  For the other countries, the numbers for 1950 to 2000 are taken from the Correlates of War Data Project while those from 2001 to 2017 are from the World Bank database.  Note that for Germany, it is military spending for both East and West Germany from 1955 to 1989 and GDP also is combined for both countries for the 1955 to 1989 period. Why only the G-7 and not all of NATO? Well, time is money and as I do not have much of either I have opted for a truncated comparison.

Exhibit A is military spending as a share of GDP from 1950 to 2016.  In the wake of WWII, spending as a share of GDP has trended down. A couple of points.  In 1955, military spending as a share of GDP ranged from a high of 9.5 percent for the United States to a low of 2.5 percent for Japan – Canada incidentally was at 6.3 percent.  By 2017, the highest military spending to GDP share is again for the United States – at 3.1 percent of GDP and Japan is the lowest at just about 1 percent. Canada and Italy incidentally are the next lowest at 1.2 percent each. 

What is interesting is that at 3.1 percent of its GDP, if the United States wants a defence expenditure to GDP ratio of 4 percent for its partners, it will also have to commit to a significant increase in military spending on its own.   And, as we know the American President campaigned on boosting defence spending and even wants to create a Space Force.  However, American military spending has been falling as a share of GDP since 2010 when it was at 4.7 percent.  As a share of GDP US military spending fell below 4 percent in 1995 and bottomed out at 2.9 percent in 2000-2001 and it only began to rise after 9/11. The Great Recession put a dent in American military spending.

And, while military spending as a share of GDP is one way of measuring a resource commitment, those military spending dollars have to come out of a government budget and all things given, more money for the military means less for other government priorities.  Exhibit B presents a more direct comparison of the resource constraints facing countries – the share of the central government budgets devoted to military spending. 

 

In 1955, everyone spent a much larger proportion of their government budgets on defence than they do today.  Nearly 60 percent of the US federal budget was spent on defense – the highest share of the G-7 with Canada coming in second place at 38 percent.  After that, France was spending 36 percent, Germany and the UK about 32 percent each and Italy and Japan were at 19 and 20 percent respectively.  Of course, times have changed, and governments do spend more and on a greater variety of social goods as well.   The military budget share has fallen.  By 2016, the United States was still spending the largest government budgetary share on defence at 14 percent.  Canada was again second at 8 percent. Japan was third at 5.5 percent followed by the UK and France at about 5 percent each, then Germany at just over 4 percent and finally Italy at just under 4 percent.

While using the defence expenditure to GDP ratio may be a good target in terms of measuring the resources a society devotes to defence, the money does have to come from a government budget.  In the case of Canada, raising our defence expenditure to GDP ratio from 1.2 percent to 4 percent given a 2018 forecast GDP of 2.142 trillion dollars would mean defence spending rising from approximately 26 billion dollars to 86 billion dollars.  As a share of a projected federal budget of 338.5 billion dollars, it means defence growing from approximately 8 percent of the federal budget to 25 percent.  Even opting for 2 percent would result in defence spending growing to 43 billion dollars making defence nearly 13 percent of federal spending.

While not at Cold War levels - when the defence share of the budget was over 30 percent - we have not been at these kinds of defence budget shares since the 1980s and to get to it while not increasing federal spending which is already at a deficit means there would have to be a lot of other things cut from the budget.  Unless of course there are tax increases to fund new military spending but that would seem to be a non-starter.  Needless to say, this would be the dilemma facing all of the members of NATO.  Just to get to 2 percent of GDP would add substantial spending to governments around the world. Getting to 4 percent in the absence of a direct military threat seems unrealistic and represents a massive re-militarization of the world given most G-7 members have been below 4 percent of GDP since the 1980s. 

So what is this about?  Is this a stick for President Trump to wave at Putin when they meet?  Is this playing to Trump's domestic audience? Is it a massive Keynesian stimulus plan designed to boost American defence industries?  I suppose we would have to take the need to increase military spending on faith and be content that all this represents the policy goals of a stable genius.

Comments

[Majromax]

To be fair to the 'size of government budget' comparisons, we also have to keep in mind that Canada has a smaller central government than the other nations. More of our social spending is handled at the provincial level, while your Exhibit B shows expenditures relative to just the size of the central government. It would probably be a better apples-to-apples comparison to relate defense spending to overall government spending, regardless of level.

Also, it would be interesting to drill down into just what is classified as defense spending. One credible-sounding point is that the US funds a lot of (say) university research through defense-related grants, whereas Canada would fund equivalent research programs through non-defense NSERC. Equivalently, I vaguely remember Trudeau suggesting that increasing pay/pensions for the armed forces counts as "defense spending." Neither directly buys tanks and missiles, which is the cartoon image that comes to mind with "2%GDP for defense!"

[Matthew Young]

Try understanding the detail here. This all started with the nuclear guarantee, our allies do not need nuclear weapons, the USA offers retaliatory strike.

It never worked because, in fact, the USA never planned to take a nuclear hit for our allies. Instead we put trigger troops everywhere so not opponent dare start something that ends in nuclear exchange.

If the USA pulls back on defense spending, then mutually assured destruction goes away, we won't risk it.

[Jacques René Giguère]

As Majromax noted, what is defense spending?
In order to assuage its pacifist population Japan massaged its budget in wonderful ways: pension and medical care for soldiers are in civilian budgets, military airports are shared with civilian trafic, if only one light plane a year, and so building and upkeep are civilian and of course science research as well.
So, with "1%" of GDP on defense, Japan manage to have the world's 4th largest navy.
https://en.wikipedia.org/wiki/List_of_navies_by_displacement
With no aircraft carriers, just a 19 500 tons "helicopter destroyer"
https://en.wikipedia.org/wiki/JS_Izumo

So with Army and Air Force. Sorry, "Self-defense Force"...

[genauer]

This is a welcome contribution to brng some facts and sanity to this discussion.

As Jacques already mentioned, for many players these official numbers are just part of the picture.

In my German case, we had the military draft until the year 2012. At my time 15 months, at last 8 month. Time which is missing to work and produce, and with near zero wage paid (250 per month). Out of a life time with some 35 work years this represents 1.5 - 4 % GDP, dependent how you calculate women / non-draftees

Furtheron those draftee were then a well trained reserve in times of conflict.

Only a part of the US expenses are dedicated to NATO, while Germany has no ambitions for a global military role. Most of the few remaining US troops in Germany serve non-German purposes. There is not a single infantry or panzer division left.

The purpose of european troops is to balance Russia, and some periphery. For that 1.3% GDP of 500 million EU people is enough to balance some 4% of 140 million Russians with half the GDP per capita.

What we need from the US is strategic nuclear protection. The US would not have one ICBM less, in case Germany would leave.
If I can not longer trust the US to keep its NATO promises, I need German ICBMs, but not more German panzers.

[Frank Restly]

Livio,

Kind of surprising that you missed the obvious.

"At today’s NATO summit he demanded that the members of NATO need to quickly increase their defence spending not only to meet the committed target of 2 percent of GDP but to also double their commitments to 4 percent of GDP."

Tying defense spending to the level of GDP is ultimately misguided.
It implies that defense spending should rise and fall with the economic output of a country rather than the necessity of that defense determined by risk.

[Livio Di Matteo]

Good point Frank. However, it remains that we tie all kinds of things to GDP as an indicator of the resource share whether it is foreign aid, health spending, etc...

[Frank Restly]

Livio,

"However, it remains that we tie all kinds of things to GDP as an indicator of the resource share whether it is foreign aid, health spending, etc..."

That resource that government spends from is either current or future tax revenue (not GDP per se).
Set all tax rates to 0% and it doesn't matter what GDP is, the federal government will have no income to fund it's expenditures or fulfill it's current liabilities.

With any non-zero tax rate, the non-discounted present value of all tax revenue (present and future) can be presumed to be infinite and so it
makes little sense (even economically) to make spending decisions based upon current (and projected future) GDP / tax revenue growth rates.

INSTEAD what must be ascertained is the public's demand for government liabilities (bonds, equity, other) versus liquidity (currency / deposits).

[Frank Restly]

Someone agrees:

http://thehill.com/policy/finance/396965-what-trumps-nato-defense-plan-would-mean-for-the-us

“You should set the strategy first, based on security needs, and then let the budget flow from that,” said Todd Harrison, director of defense budget analysis at the Center for Strategic International Studies (CSIS), a think tank in Washington, D.C.

“I think it’s a ridiculous metric to use. Why do you want to tie your level of defense spending to something that’s totally unrelated?” he added.

[Jacques René Giguère]

genauer: "I need more german ICBMs".
Exactly what France, from the 4th Republic on(De Gaulle had nothing to do with the inception of nukes) realized. In fact Joliot-Curie was already working on A-bomb plans in the '30's.
If nobody will mess with North Korea, it will not be because of its huge stock of rolling junk...

[genauer]

@Jacques: Germany might follow France, and buy ICBMs from her

For a long time we said, we dont need our own German nukes. We have the 3 western nuclear powers on our side, we are close Allies, the largest one has trip wire forces in Germany. German was supporting the non-proliferation treaty.

By now, the UK is leaving, and the US just pays lip service and wants to extract protection money.

Getting our own nukes, costs some serious money, will motivate others to do the same, but why should Germans be second class citizens?

@Frank
The total effort to balance a potential adversary, namely Russia, should be calibrated to that, and just looking at GDP is misleading. The Russians still have some 4000 ICBMs, a large territory, a working nuclear power plant technology, a working Space station. Considering that they are only 140 million people on large swathes of unhospitable land, their GDP, their achievments are actually quite impressive.

But I think with getting our existing forces into a somewhat better shape is sufficient. We have in Europe now strategic depth, and will not be overrolled in some surprise attack. And we dont have, unlike the US, no agressive intentions.

For looking at some just burden sharing, in between Allies of roughly the same output per capita, I see fractions of GDP as a fair criteria. After considering that US, UK, France cover with that substantial other, national interests

[robots2005 AI32080]

Number of sounding rockets and number of VTOL fighter jets is more useful a metric. Nazi Germany would be the peak for the former and the Harrier days of the UK for the latter. At present, such analysis leaves Winnipeg 1st at present with Moscow soon to follow. Who is buying a 3d printable rocket with acoustic homing? 100x as many fighters go down for each robot hit if present missiles. I feel lonely without Chretein and B.Clinton in power.

[robots2005 AI32080]

I note Russia has the 2nd best education on Earth next to MB. There are ways of analyzing economics from a macro perspective, but the micro problems need to be solved: namely, software graphing itself and a person deduces it can be turned off by humans.
If you discount insurance premiums for risks the industry can't presently insure, you shift the premium into the past at some point: it turns into investment to shape the structure of the economy. Macro would model this as more investment required for a Cold War mindset, less for a Wpg or Russian mindset, and an infinite amount for money in power as a mindset. The inertia of being mad at a chosen life of no free time, of considering the ethics of technology, and of considering the ethics of society, are all brain imagible and all equatable into macro.
So for a NATO investment rate, you can assume neuro-imaging command, you can assume diamond is only used for good, you can assume a bad Haig-like Presidential leader or worse Chinese tyranny will fall somehow, you can assume humans will still be post-IR progessive after robo-attacks; but these assumptions have to be stated. Then they can be quantified using macro. And the worldview of Truman and King can be fixed. The economic mistake made before them was the division of labour, ignoring it wouldn't put simple people into power. My babysitter who made me pour milk a certain way was not fit to guide an economy with the ends being quality-of-life.
Europe needs frigates and private boats fitted for VTOL fighters. Then factories that get turned on can be precision bombed. Metal shops are harder for robots to replicate with: this is a macro measure, doubling time and doubling base. 2005 saw hands, a 1980 Manhattan might've made software able to equate a human with a nuke able to turn it off.