One of the lesser-known stories of the stimulus is the focus placed by President Obama's team on boosting antipoverty measures. As Paul Tough wrote in his New York Times Magazine piece on Obama and poverty, one of Larry Summers, Tim Geithner and deputy NEC director Jason Furman's top priorities at the start of the administration was boosting antipoverty programs. When the Center for Budget and Policy Priorities' Robert Greenstein sent Furman a list of possible funding increases for antipoverty initiatives, Furman wrote back complaining that Greenstein's list wasn't long or ambitious enough.

If you're an Obama backer, anecdotes like that are just another indication of the administration's dedication to helping society's most helpless members. But right-wing talk show hosts have another theory. As Dave Weigel reported Friday, the operating theory of much of the conservative media is that increases in food stamps are functioning as bribes to get poor people to vote for Obama. "We have three million more off the unemployment rolls and on the disability rolls, and they all vote," Rush Limbaugh tells his listeners. " The evangelical activist Gary Bauer told Weigel, “There’s a lot of people out now around America who depend on checks from their fellow taxpayers being in the mailbox every day. They will turn out in massive numbers.”

Here's the problem: poor people actually don't vote that often. According to a CNN exit poll in 2008, those making less than $15,000 a year made up 13 percent of the population but just 6 percent of voters, while those making more than $200,000 a year made up just 3.8 percent of the population but fully 6 percent of voters:

If all income groups had voted evenly, Obama would have beaten McCain 55.2 percent to 42.7 percent, a net gain of 5.3 points relative to what actually happened. So no, poor government program beneficiaries don't "all vote" or turn out in "massive numbers".

In a way, this makes sense. Many people take time off work to vote, something that's easier to do if you're in a white collar job. Others vote in the evening, when many working-class restaurant workers, night shift janitors and so forth are on duty. What's more, voting costs gas or public transit money that many poor people just don't have.

These facts alone doesn't disprove Limbaugh and Bauer's core theory, that government programs build in a constituency that can support welfare programs. But that theory has its own problems. A majority (52 percent) of Americans say they want to cut federal welfare spending, and relatively narrowly pluralities oppose cutting housing aid and food stamps, according to a Harris Interactive poll from March of this year. A poll from Rasmussen, a right-leaning outlet, in July found that 47 percent of Americans think we're spending too much fighting poverty and only 30 percent think we're spending too little. So the evidence that anti-poverty programs are a political boon is somewhat limited. Indeed, Princeton's Martin Gillens has found that many War on Poverty programs, such as housing projects and welfare, were unpopular when Lyndon Johnson implemented them.

What's more, the trends with voting by class don't seem to match up with the generosity of government programs. Here, according to NonprofitVOTE, is how voter turnout between income groups changed from 1998 to 2010:

If the Bauer/Limbaugh theory were true, you'd expect to see an increase in low-income turnout between the 2006 and 2010 midterms due to the increases in food stamps, the Earned Income Tax Credit, and so forth. But turnout among low-income people actually falls. There's just no evidence to support the theory that Obama's antipoverty measures will lead to a mass of government-dependent people backing him in the polling booth.