Late Thursday, the House of Representatives voted, 217-210, to cut the Supplemental Nutrition Assistance Program, formerly (and popularly) known as food stamps, by $39 billion over the next 10 years, a 5 percent cut relative to current law. As Brad explained Thursday, the plan would kick 3.9 million people off the food stamp rolls the coming year. After next year, it would reduce the rolls by about 2.8 million people each year.

The American Community Survey by the Census Bureau actually keeps track of how many households in each cistrict are on food stamps (thank to Andrew Reamer for pointing this out). So I thought it might be interesting to see how food stamps usage in districts represented by supporters of the cuts differs from usage in districts represented by opponents. Unsurprisingly, supporters' districts are less reliant on the program, with an average of 12.4 percent of households on SNAP, than opponents', where the average is about 15 percent. Curiously, the 15 House Republicans who opposed the cuts had districts with lower average food stamp use (~ 11.3 percent) than either districts of Republicans who supported them or districts of Democrats (all of whom opposed the cuts):

But this result is pretty fragile, and driven by Democratic domination of the poorest, most SNAP-reliant districts rather than by strong support for cuts in districts that aren't very reliant on the program. The only three congressional districts with SNAP utilization above 30 percent of households — including New York's 15th district, covering much of the Bronx, where ise is about 50 percent — are all represented by Democrats.

So it's worth asking if the share of households on food stamps has any effect on House members' votes once you take their party affiliation into account. It appears they don't. If you do a simple regression trying to explain how members voted with only two explanatory variables — the member's party, and the share of his or her district on food stamps — the latter isn't even close to statistically significant.

Now, that doesn't mean that members of Congress aren't responding to the views of their districts, since presumably economic conditions of districts — including food stamp usage — help determine which party represents them. But Democrats in districts with barely any food stamp users (such as Henry Waxman, whose district's SNAP usage rate is a paltry 1.7 percent) all voted against cut, and Republicans in districts with huge numbers of food stamp users (such as Hal Rogers, 29 percent of whose district's households are on SNAP) almost all voted for them. It's yet another indication that House members are becoming less and less motivated by parochial interests of their districts and more and more unified on party lines.