War and the Great Recession: Some Thoughts
A while back, an American journalist wrote to me:
We ... are trying to determine how big of a war we would need to have in order to drive the US out of this recession. It is common belief that WWII was a major factor in invigorating US economy which had been decimated during the Great Depression. I was wondering if it would be possible to make a projected estimate for our current situation using that era as a model.
This question got me thinking and I put quite a lot of effort into some answers, which they did not use. So, I thought I would update them and share them here.
Basically, the question sees the problem back to front and upside down. The problem we should be thinking about today is not how to start a war that can help pull us out of recession. The real problem is that, if we don't pull ourselves and others out of recession fairly rapidly by peaceful means, we will face growing risks of war -- and that could end in a catastrophe for everyone.
So, it is a trick question. Sometimes, however, it is interesting to take trick questions at face value and work out what is wrong with them by seeing where they lead. This is what I did.
Is the situation of the U.S. economy today comparable with the Great Depression?
At the moment, the situation of the U.S. economy over the next year or two looks bad compared with the recent past, but it is still way better than it was in the 1930s. Economists often work in terms of what is called the "output gap," the proportion of potential output that is unrealized because there is not enough demand in the economy. The Congressional Budget Office currently expects the output gap over the next two years to average almost 7 percent.
There are various ways of calculating the output gap of the U.S. economy before World War II, and it varied a lot from year to year, but any reasonable estimate would be far above 7 percent. At the bottom of the depression, in 1932, the gap was probably around one third. At the end of the first recovery, in 1937, around one fifth of potential output was still not being realized, and in 1938 and 1939 the output gap widened again. It had got back to relatively normal levels by 1941.
So the good news is that, on present forecasts, the fiscal stimulus that is required to fix the U.S. economy is much less than was called for in the 1930s. What everyone should worry about, though, is that if things play out badly in the world as a whole there is plenty of scope for present forecasts to prove optimistic.
What size of war would be required to provide an equivalent fiscal stimulus?
U.S. GDP is currently around $15 trillion a year, and so an output gap of 7 percent means about $1 trillion a year of lost production. Since, in the U.S. economy, an extra dollar of public spending should give rise to about an extra $1.50 of total (public plus private) spending, a stimulus of around $700 billion a year would be needed to stimulate $1 trillion a year of extra output.
As far as I am aware from press reports and so on, the total U.S. budgetary appropriation for the global war on terror (Afghanistan, Iraq, and the protection of U.S. embassies abroad) has reached around $1 trillion in total, spread over the entire period from 9/11 to the present. I am not certain what the annual cost is currently, and I believe that not all of it is explicitly funded (i.e., the GWOT has been partly funded by the Defense Department taking resources from elsewhere.) For the sake of argument, suppose the net budgetary cost of the GWOT has recently been of the order of $200 to 250 billion a year. To provide a stimulus of $700 billion a year, therefore, the required war would have to be the equivalent of three additional global wars on terror, waged on the scale of the recent past.
How does that compare with the fiscal stimulus package that went through Congress recently? The package is $700 billion spread over two years, and much of it is tax cuts rather than public spending, which will have a lesser impact because tax cuts can be saved rather than spent privately. It is one half or one third of the stimulus that would halt the slide, so it runs the risk of being too little, too late.
One reason for the modest size of the package is that President Obama is restrained by conservative opponents of big government in Congress. I suppose someone could argue that a war might help to overcome such constraints. I think that would be a bad argument. It amounts to saying that we should whip up nationalism in order to stigmatize the people we disagree with as unpatriotic and crush them. That is not unheard of, but it does not appeal to me.
How good for the U.S. economy would it be to have another war?
History should make us very skeptical. Here are five reasons. First, it is true only in small part that World War II pulled the U.S. out of the depression. In fact, 1940 was the first year after 1919 when U.S. military spending rose above 2 percent of the national income. The fiscal stimulus from New Deal spending was also modest. The main driver of the U.S. recovery up to 1940 was private investment. If World War II had not broken out, this natural recovery process would have continued.
Second, it is true that the wartime period saw a huge further increase in the total output of the U.S. economy. In the three years from 1941 to 1944 the GDP rose by about two thirds. The main element in this was Federal outlays on national defense that brought about a vast increase in the mass production of standardized machinery and equipment for combat and transportation. Because of mobilization and wartime controls, patriotic national feeling, and mass production and the associated efficiency gains, the U.S. economy could temporarily produce far above peacetime norms -- effectively, there was a negative output gap. But the extra output did not make anyone better off; it was mainly in the form of ships, planes, and guns that achieved victory, not higher living standards.
After the war, most of the extra output disappeared and the economy fell back, not towards depression, just towards normal peacetime operation. So the wartime "production miracle" did not bring about lasting gains. The U.S. economy was much more prosperous after 1945 than before 1941, but this was not because of the war. It was because of the return to normal working combined with underlying productivity advances that had continued through the Great Depression, but were temporarily overwhelmed by the lack of demand.
Third, it is true that millions of U.S. citizens had a good war, economically speaking. Many people would previously have expected to live out their lives in poverty in the South and Mid-West. They moved to the industrialized, urbanized North and found new lives there. Many young men gained new skills and experiences by joining the military and fighting or supporting the war effort overseas. You might ask whether there were cheaper ways of achieving the same goals without having to fight a war. I don't mean that American involvement in that particular war was wrong; it was clearly in America's own national interest. But if you want to achieve a more mobile, equal society, and war is not forced upon you, there are cheaper ways.
Fourth, it needs to be said out loud that war is costly to society in terms of death and disability. I looked up what Michael Edelstein has to say in his chapter on “War and the American Economy in the Twentieth Century,” in The Cambridge Economic History of the United States, vol. 3 (published in 2000), on pages 342 and 349. He measures the budgetary cost of war as the cost of defense above normal peacetime operations, and the social cost is the capitalized value of lost earnings of the killed and injured. Everything is in constant 1982 prices. Edelstein’s estimates are: WW1, budgetary cost $378 billion and social cost $25 billion, WW2 $2,460 billion and $202 billion, Korea $206 billion and $27 billion, and Vietnam $313 billion and $46 billion.
You can see a couple of things. One is that, on this measure, the social costs were relatively small. Why? Mainly because the United States could fight all these wars at a distance against much less well equipped enemies. In World War II, U.S. battle deaths in Europe and the Pacific were around 350,000. In contrast, Red Army battle deaths on the Eastern Front were around 8.7 million.
Another thing is that, on the same figures, the ratio of social to budgetary costs rose continually from war to war. As a share of the combined total, the human costs were around 6% in WW1 rising to 12% for Vietnam. Why? I think, mainly because there was rising productivity, so human life got relatively more expensive. In their book on the costs of the Iraq and Afghanistan conflicts, Joseph E. Stiglitz and Linda J. Bilmes come up with various figures but their “realistic moderate” scenario (The Three Trillion Dollar War, published in 2007, page 112) suggests about 12% for social costs as a share of the total of budgetary plus social costs combined. (Stiglitz and Bilmes include items of veterans’ welfare costs that Edelstein I think does not, but these appear on both the budgetary and social sides of the balance.)
What does this mean? Well, if you want $700 billion a year of fiscal stimulus through going to war, you’d better factor in that, for every year at war, the U.S. economy will also lose a future income stream with a capitalized value of $100 billion, because of troops killed and injured. That does not seem like a good idea.
Fifth, a war today would bring huge costs in further disruption of the international economy. In 1941 international trade was a small fraction of its pre-1913 volume, so there was little to lose. The world today is much more interdependent than it was in the 1930s. Stiglitz has pointed up the billions of dollars lost to the U.S. economy because the war in Iraq triggered higher oil prices. You need to factor that in too.
Maybe I should finish this bit by quoting Edelstein again (on page 349):
It is absurd to think that the methods and perspectives of economic history come anywhere near to comprehending the meaning of human losses from war. We are far better served by the speeches and letters of Lincoln or the poetry of Sassoon, Brooke, Owen, Graves, and Seager.
OK, but where does this leave us?
I have two conclusions. One is that the only good reason to have a war would be to defeat an enemy. If our leaders want to make our economy work better in everyone's interests, and if they have legitimate instruments to achieve this, and if such improvements would also be an accidental by-product of a war, then that is not an argument for a war. It is an argument for adopting peaceful ways to achieve these things that carry democratic consent and do not also involve the irreversible losses and persistent collateral damage that a war would bring.
My other conclusion is that the original question ("how big of a war we would need to have in order to drive the US out of this recession?") confuses the problem for the solution. It's true that the Great Depression ended in the most terrible war the world has seen. But it did not end because of the war; the depression would have come to an end anyway. In fact, the war curtailed the natural recovery process.
But why did the war come about? World War II happened for a number of reasons, but one of them was the great powers' failure to avert the Great Depression in the first place, and rapidly to mitigate it once it came along. Many of the ingredients for violent conflict were there, but until the Great Depression they lacked a spark. Before 1929, was Germany evolving gradually towards a normal parliamentary democracy? Yes. Would Hitler have come to power without 30% unemployment in Germany in 1932? Probably not.
Eurasia today, from the Baltic to the China Sea, has many of the ingredients for violent nationalism. Scattered around that vast landmass, there is more than enough petrol and a good supply of oil-soaked rags. Meeting in London today, the G20 has the power to coordinate an effective international response to the global economic calamity that threatens us. If they fail, it is not just an economic calamity that we should fear; the world's leaders are playing with matches.