China's military spending

First, I am of the opinion that everyone should spend less on its military, and that includes China.

China's official defense spending in 2004 was $25.4bn, by including factors such as PAP, space program and more weapon imports, RAND estimated the range to be $31-38bn (2003 figure, average=$34.5bn), SIPRI $35.4bn. China's GDP was $1933bn in 2004 (old number was $1650bn) according to most recent annoucement. So China's defense spending was 1.3%-1.8% of its GDP according to various estimates.

But here I have not considered one well published (but not neccessarily accurate) source, the Pentagon, estimated China's spending to be $67.49bn, based on PPP adjusted conversion into US$. Because much of China's military spending is for importing weapon from Russia, using hard currency, everyone (including RAND, SIPRI, armscontrolwonk) have rejected the PPP proxy. Pentagon, for its own agenda of securing more fund form the Congress, is the only body that stubbornly insists on the misleading PPP. Anyway, let's see how that translates in % GDP. This is from CIA's country profile for China
  • GDP (PPP): $7.262 trillion (2004 est.)
  • Military expenditures - dollar figure: $67.49 billion (2004)
If CIA's own data is correct, China spends $67.49bn/$7262bn=0.93% of its GDP on defense. If we use the new GDP figure, $8480bn (in PPP), the ratio is only 0.79%.
  • From the same CIA site, there is a conflicting (derived) number. It said China's "military expenditures - percent of GDP: 4.3% (2004)". This is based on a wrong calculation of dividing a PPP-adjusted number by an exchange rate valued GDP (about $1.65tn). To compare 2 numbers, you need to get the units right. What CIA/Pentagon did was like saying I am taller than Yao Ming because I am 6 feet tall and Yao is only 2.29m.

My problm with CIA's site, while it is handy, is that it also lacks some important information. e.g. if a PPP adjusted value is quoted, I should at least know what ratio it used. US State Dept or DoE use exchange rate GDP, which is a source I use more often now (worldbank and EIU are better source, but they do not have detailed demography or geography data).

update: great insight from Strategic Unit



dylan said...

An extremely misleading post.
1. You use the RAND estimate and compare it to the newly revised GDP for 2004 to arrive at a percentage of GDP figure. The RAND estimate is for 2003 (see page xx of the summary of "Modernizing China's Military" by Cliff, Crane et al.). You should therefore be comparing it to the 2003 GDP NOT the 2004 GDP.
2. You claim the "Pentagon" issued the estimate on the CIA World Factbook page and proceed to make a number of unsubstatiated assertions based on this "fact". Yet the 2003, 2004, and 2005 Pentagon report have entirely different figures - so it cannot come from the Pentagon. One can only assume the figure from the CIA factbook comes from the CIA, and, I would suggest, is nothing more than taking the official PRC figure and multiplying it by the PPP exchange rate the CIA chooses to assign to China.
3. Where is your evidence to prove Pentagon estimates are PPP? There is no indication of this in any DoD report, rather they speak of unbudgeted items. It would be interesting to know how the Pentagon calculates these figures, but I see no indication they are PPP on the face of it. Maybe you know more.

Sun Bin said...

China's GDP growth from 2003 to 2004 is about 9%. So the 2003 China GDP is $(1993/1.09)bn, and 34.5/1993*1.09=1.88% GDP, making it similar to the SIPRI figure. Not misleadinhg at all in terms of conclusion.

Sun Bin said...

see, e.g., armscontrolwonk, "“Off-budget funding for foreign weapon system imports” probably accounts for no more than another US$ 2-3 billion (or 8-12 percent) of Chinese defense expenditures, based on the State Department’s World Military Expenditures and Arms Transfers (WMEAT) and the Department of Defense’s Annual Report on the Military Power of the People’s Republic of China. Much of the difference between the “official” Chinese number of $25 billion and the DOD number of $50-70 billion is based on purchasing power parity (PPP) adjustments."

Sun Bin said...

see here

"The CIA is using a twenty-five year old PPP estimate for defense goods."

And that is the only way to explain the $67.6bn, way larger than Rand or SIPRI.

dylan said...

You haven't responded directly to any of my entirely fair criticisms of your post. You appear to acknowledge you were wrong to compare 2003 defence spending estimates with 2004 GDP and yet still try to fudge the numbers. Chinese claimed GDP growth was 9.5% between 2003 and 2004 so revised 2003 GDP may be $1765 billion (or it may not, the fact is nobody knows what the real 2003 GDP was and the NBS is yet to announce its estimate). Dividing that by $35 billion gives 1.95%
You quote material from the Union of Concerned Scientists that has twenty year old source material about something that may be PPP and an assertion this is still valid. The UCS quotes only one small element of the alleged off budget expenditures of China on the military and then proceeds as if this is the only possible addition that could be made to the official budget so everything else must be PPP adjustment. Here is a simple question for you, if the CIA page uses PPP exchange rates for its estimate of China's 2004 GDP and comes up with $7262 billion (suggesting a rate of 4.4) how did it come up with $67 billion using the same PPP exchange rate for military expenditure (official expenditure in 2004 was $25bn x 4.4 = $110bn)?

dylan said...

I could go on picking holes in this piece forever. You say "Because much of China's military spending is for importing weapon from Russia, using hard currency.." Yet your source material from the UCS makes exactly the opposite point: "Off-budget funding for foreign weapon system imports” probably accounts for no more than another US$ 2-3 billion (or 8-12 percent) of Chinese defense expenditures". So in fact much of China's defence spending is NOT in hard currency for imports.

dylan said...

You conveniently ignore the estimate the UCS comes up with as "probably too low" but better than resorting to PPP. That figure is US$40-50 billion in 2004: "I am skeptical about using a PPP estimate to convert to dollars. While the current exchange rate produces a number that is probably too low ($40-50 billion), PPP produces some perverse effects." Why not add that estimate into your mix?

dylan said...

You also ignore the estimate that this source (http://projectchina.blogspot.com/2005/10/mistrust-money-and-making-of.html) you approvingly quote came up with: US$51bn.

denk said...

“I could go on picking holes in this piece forever.”

So what if all this nitpicking on china’s spending turn out to be true, it would still be peanuts compared to US,
Seems to me a case of being able to swallow an elephant while getting choked on a gnat.

dylan said...

Good on you Denk, if you're in trouble try changing to a different subject entirely where you think you're on better ground. Tough luck, I'm not an American, don't live in America, and agree US military spending is huge. End of story, now back to the topic at hand. Why does sb privelege a set of lowball estimates while ignoring the higher estimates contained in his own sources?

Anonymous said...

"An extremely misleading post."
"I could go on picking holes in this piece forever"

These kinds of abusive languages don't fly with me! Sun Bin, as always, you are way too nice to others. If I were you, I would delete and ban this same trolling scumbag, Dylan or Bmanster who keeps on posting malicious and venomous comments. Maybe pre-approval moderation should apply for comments to prevent future abuse from the same troll.

Sun Bin said...


i noticed that. well, if DoD and CIA did not provide the details (did rummy ask for "transparency"?) in their methodlogy, i guess the only way to come up to 67.5bn is half PPP, half exchanged rate. i.e. 10x4.4+20 or something like that. ask CIA or DoD. my point is, such ratio with GDP makes no sense.

as for 2003 vs 2004 question. the calculation is very straightforward, i assumed a similar adjustment of 16.8% in service sector for GDP. it could be 15% or 18%, but the difference is very small (1.115 vs 1.118).

SIPRI's $35.4 is 2004, official # of 25.4 is 2004. you have any problem with those?

denk said...

, u remark that us defence expenditure is “huge” must be the understatement of the century, I think “obscene” would be more like it.
At one time, some analysis estimate us military budget trumps the next 10 big spenders combined.

Anyway, All these figures cause me a massive headache,
Just tell me , whats your estimate of Chinese true expense, ?
The most skeptic critic figure that it should be 2, 3 times higher.
Which means its still incomparable to US and japan, china’s main detractors, who, incidentally, are orchestrating an encirclement of china.

What happen here is -- a 1200 lb gorilla carrying a bazooka and his 1000 lb sidekick with a double barrel shotgun are prowling outside this man’s house, the man says I am going to get myself a stick in case these goons try something funny, whence the gorilla cry foul, “now no body is threatening u, why are u arming yourself TO THE TEETH, its making all your neighbors mighty NERVOUS, are u up to no good…?”

I already said it, its not inconceivable that china is cooking the book, when oh so transparent uncle scam has been doing it big time ,
But I think any neutral observer should have marveled at the brazen hypocrisy of the gorilla and his pitbulls, instead, I am surprised that u have joined the bandwagon like others, harping on perceived flaws and lack of transpancy in china’s figures. – (“look, I don’t think this man has got just a wooden stick with him, I think he probably has 3 wooden sticks with him”)
I think this a very bizarre reaction from someone who claims to be a neutral observer.

Sun Bin said...


thanks. i prefer to leave what they say here. people can judge by themselves.


the point here is to let people aware of more sources of estimates and what the real picture is. since the erroneous CIA/DoD # were highly publicized, and there is a ridiculous calculation of dividing a half PPP-adjusted # by an exchange rate GDP.

i did not see the 51bn #, and i do not understand the methodology. but it is quite strange that it could be $51bn in year 2002. i suspect it comes from the DoD source as well. do you sincerely believe it?

Sun Bin said...

there are 2 numbers from the CRS report. (it said exchanged rate is used EXCEPT for some, i guess China is the exception -- see WMEAT below)
IISS: 51bn for 2002
WMEAT: 89bn for 1999, this one definitely used the 4.4 PPP rate, since China's budget was in the teens then.

i suspect 51bn (2002) is also partial PPP rate, and it is very likely the same estimate as the 67.5bn in 2004 used by CIA.

therefore, CIA probably used IISS, and I did not miss any source.

Anonymous said...

Dylan, I think your "exteremely misleading" comment is itself exteremely misleading. Look at your own number, they support the original thesis in SB's article.

As to transparency, I'd like to hear your comment on US military "black budget".

- bobby fletcher

dghnfgj said...

welcome to the wow power leveling, cheap service site,WoW Gold buy cheap wow gold,wow gold,world of warcraft buy wow power leveling