Ezra Klein's explanation of the adverse impact of state budget cuts and tax increases on federal stimulus efforts upset the folks at George Mason's libertarian-leaning Mercatus Center. Increases in federal spending in 2009 and 2010 swamped cuts to spending at the state level, they rejoined:
At the same time that aggregate state spending was falling by 4.3% and 6.8%, federal spending was increasing by a whopping 17.9% (2009) and 5.8% (2010). This, combined with the fact that the Federal Government spends trillions while states spend hundreds of billions (in the aggregate), means that the state spending contraction comes nowhere close to offsetting the federal spending increase.
At first glance, such a comparison seems to make sense: the federal government increased spending in 2009 by about $600 billion while states cut spending by $110 billion. The federal government's increase significantly outweighed the state cuts.
But this obscures - actually, it simply ignores - the question of why we're even comparing state and federal spending. Cuts to spending at the state level are worrisome because they lead to fewer services, fewer jobs, and poorer education, all of which depress the economy. The appropriate question to ask about federal spending is then obvious: what federal spending offset such economy-depressing cuts?
It turns out that strikingly little of the "whopping" increase in spending in 2009 came from such efforts. In his breakdown of the sources of increased spending, the Center for American Progress's Michael Linden shows that 41 percent of increased spending was from President Bush's financial rescue; 18 percent was from the mandatory Social Security, Medicare, and Medicaid programs; and 7 percent was from defense. Only 18 percent of the increased spending - or $108 billion - came from the stimulus package in 2009.
Generously categorizing Linden's "other domestic spending" category as stimulative and adding in increased spending on unemployment benefits (which is definitely stimulative), Obama's efforts to counteract the economic downturn, one consequence of which was cuts to state spending, accounted for only 34 percent ($205 billion) of increased spending in 2009. The rest of the increases have little to do with stimulating the economy (Medicaid spending surely increased in part because of an increased caseload caused by the recession. But this is also a primary cause of fiscal strain for states, which share the program's cost.).
A comparison of federal spending and aggregate state spending is irrelevant. Comparing federal stimulus spending and state spending cuts is only appropriate and useful because both are responses to the economic downturn.
Update (7/1): Matt Mitchell at Mercatus quite rightly exploits my lazy sentence above that "The rest of the increases have little to do with stimulating the economy." To the extent that these spending increases are automatic stabilizers and occur as deficit spending, they do in fact have stimulative effect (Matt would disagree but I'll leave it to others to debate the wisdom of Keynesian stimulus - my perspective is obvious.).
The important difference between ARRA and these other spending increases is, far from being about mere intentionality, that the ARRA spending was designed in the context of these other increases and state spending decreases. That is why ARRA is compared, with concern, to cuts in spending at the state level. Mitchell wants to say that deficit spending does not stimulate the economy and that's fine. My point, though, is not to re-litigate the wisdom of such spending, but to suggest that the comparison is not pulled from thin air.