Not all spending is equal
Yesterday I linked to a Paul Krugman post lampooning Republicans for discovering their inner Keynesians when defense spending is on the line. A worthy target, to be sure, but I have to nitpick one of his points:
The first thing to say is that liberals shouldn’t engage in mirror-image thinking, and imagine that spending we dislike somehow lacks the job-creating virtues of spending we like. Economics, as I say often, is not a morality play. As far as creating aggregate demand is concerned, spending is spending – public spending is as good as but also no better than private spending, spending on bombs is as good as spending on public parks.
This is mostly right. Defense spending, as Krugman later notes, “one of the main sources of evidence that fiscal expansion really does stimulate the economy.” But if you believe in the Keynesian multiplier, it must be noted that certain forms of spending have greater effects than others. I am not familiar with any research on the economic effects of spending on public parks, but most studies confirm that defense spending carries a multiplier of less than one. Dylan Matthews wrote about this recently:
Robert Barro and Charles Redlick released a paper in 2009 calculating the “economic multiplier”— that is, the GDP bang per spending buck — of defense spending, using data from 1914 to 2006. They found that during normal economic times, the multiplier is about 0.67, and is higher when the economy’s weak, rising by 0.05 for every percentage point by which unemployment exceeds 5.6 percent. Given last month’s unemployment rate of 9.1 percent, the multiplier would be about 0.845. That is, each dollar of defense spending currently grows the economy by about 85 cents.
If the supercommittee’s trigger goes off, about $500 billion will be cut from defense programs, starting in 2013. Thus, assuming that $50 billion is cut in 2013 specifically, and unemployment is still where it is today, the cuts will reduce GDP by about $42.25 billion – or 0.2 percent of our current $15 trillion GDP. If unemployment falls, the multiplier will shrink and so the hit to GDP will be lower, and if the economy grows, then the hit as a percent of GDP will be lower as well.
It’s worth noting, though, that while certainly stimulative, military spending isn’t the most stimulative thing the government spends money on. Mark Zandi of Moody’s estimates that the multiplier is 1.74 for food stamps, 1.61 for unemployment benefits, and 1.57 for infrastructure spending. The trigger’s cuts to those sorts of programs, thus, would cause much more economic damage relative to the size of the cuts than the defense cuts would.
Of course, automatic stabilizers like food stamps and unemployment insurance tend to kick in when GDP is low and unemployment is high. I’ve never heard anyone propose them as an economic cure-all. Changes in the multiplier are clearly affected by varying economic circumstances, but still, not all spending has an equal effect on the economy.