I remember reading a paper that was attempting to measure different country's Keynesian Multiplier. Keynesian theory only calls for government spending when the multiplier is greater than one: i.e. a government spent dollar more than pays for itself in economic growth.

Not only did the paper find most multipliers to be below one, but America's was negative: i.e. each dollar spent by the government actually reduces the size of the economy.

That was an old paper, but I thought it was pretty amusing.