Though, I would slightly qualify his argument: if war is considered necessary, say, for reasons of self-defence, we are likely to be thereby inviting a planning regime that is strictly incommensurable with a free economy. The best you can do is to ponder if the "war economy" could be run more efficiently than it is, or not.
In fact, a "war economy" isn't really an economy. For it is simply an attempt to turn the economy proper into a single purpose project or organisation amenable to comprehensive planning. Not surprisingly it does perform a lot worse in terms of the criteria we would apply to measuring the effectiveness of a free economy. At best, it can give us the materially most effective way of attaining a specific objective - winning a war.
By contrast, the whole point and beauty of a free economy is to serve as a facilitator for as many goals as its participants may desire to attain - none of which goal commanding per se a higher rank of validity than any other.
See also Liberty - The All-Purpose Order.
One of the reasons that people often believe that war "improves" the economy is that they are looking at the wrong metrics. They look at unemployment and observe that it falls. They look at capacity utilization and observe that it rises. They look at GDP and see that it rises.
But these are the wrong metrics. What we care about is if people are better off: Can they buy the things they want? Are they wealthier?
These outcomes are hard to measure, so we use unemployment and GDP and capacity utilization as proxies for people's economic well-being. And in most times, these metrics are reasonably correlated with well-being. That is because in a free economy individuals and their choices guide the flow of resources, which are dedicated to improving what people consider to be their own well-being. More resources, more well-being.
But in war time, all this gets changed. Government intervenes with a very heavy hand to shift a vast amount of the resources from satisfying people's well-being to blowing other people up. [...]
What we find is that in war time, unemployment is down, but in part because young people have been drafted (a form of servitude) to fight and die. Are they better off so employed? Those who are left find themselves with jobs in factories with admittedly high capacity utilization, but building things that make no one better off (and many people worse off). GDP skyrockets as government goes deeply in debt to pay for bombs and rockets and tanks. This debt builds nothing for the future -- future generations are left with debt and no wealth to show for it, like taking out a mortgage to buy a house and then having the house burn down uninsured. This is no more economically useful than borrowing money and then burning it. In fact, burning it would have been better, economically, as each dollar we borrowed in WWII had a "multiplier" effect in that it destroyed another dollar of European or Asian civilian infrastructure.
Sure, during WWII, everyone in the US had a job, but with war-time restrictions and rationing, these employed people couldn't buy anything. Forget the metrics - in their daily lives Americans lived poorer, giving up driving and even basic staples. This was the same condition Soviet citizens found themselves facing in the 1970s -- they all had jobs, but they could not find anything to buy. Do we consider them to have been well off? [...]
The entire post.