Thursday, March 3, 2011

More on Elections, Mandates and Polls

I’ve commented before on the fact that the last election and recent polling results invite incompatible inferences about how voters will react to the current standoff between Wisconsin Republicans and Democrats over public employee collective bargaining rights. You could say substantially the same thing about the current fight between Republicans and Democrats about how to deal with the federal deficit. In each case, you have to ask whether your best interpretation of the mandate won in the last election or a snapshot of public opinion projected by the latest poll is more probative of how people will vote in the next election.

Republicans made sweeping gains in the House and Senate last November by promising, among other things, to get a handle on the budget deficit without raising taxes. If they’re at all serious, that has to mean making substantial cuts in unfunded Social Security and Medicare entitlements. It looks like House Republicans are determined to submit a budget this April that proposes to do just that, presumably in the expectation that they’ll be punished politically for not fulfilling their campaign promises if they don’t.

Yet a new Wall Street Journal poll suggests that, although most people want the government to do something about the deficit, there’s a lot more public opposition to Social Security and Medicare benefit cuts than the traditional Democratic approach of keeping benefits intact and raising taxes to make up for the budgetary shortfall: “Asked directly if they thought cuts to Medicare were necessary to "significantly reduce" the deficit, 18% of respondents said yes, while 54% said no; the rest were not sure or had no opinion. On Social Security, 22% said cuts would be needed, while 49% said they weren't.”

Such polling results have a lot of liberal Democrats gleefully anticipating that House Republicans are marching off a political cliff. Democrats are expecting to do well in the next election by dusting off the tactic they’ve used in elections past, opposing any cuts in Social Security and Medicare benefits even if that means raising taxes. They’re betting that polling data tells you a lot more about what will happen in the next election than what happened in the last one.

Here, for example, is Jonathan Chait:
“The specifics of the fight -- Republicans promising to cut overwhelmingly popular programs, being willing to shut down the government, and pushing a plan that private analysts predict will reduce jobs -- put them in a very tough position. Republicans are working really hard to buck each other up and ignore data about public opinion. Democrats have the upper hand here. President Obama may decide to cut a deficit deal, but both the politics and the policy say he should hand the Republicans their head first.”
Maybe Chait's right. Yet I can’t help thinking that Democratic political handicappers are kidding themselves when they put greater stock in polling data telling them things they want to hear than in the depressing results of the last election. When a pollster asks whether “Social Security cuts are necessary to reduce the deficit,” a negative answer implies that the respondent prefers some other approach to cutting the deficit presenting an unspecified tradeoff of benefits and burdens. That’s not much different than the respondent’s saying that, all other things being equal, he prefers that Social Security benefits not be cut. Since all other things never are equal, he’s telling you virtually nothing about his preferences among alternative sets of tradeoffs.

Voting for a candidate in an election is, in crucial respects, a lot more like buying a car than filling out a questionnaire. All things being equal, someone entering the car market would prefer the model that gets the best gas mileage, has the most aesthetically pleasing styling, the fanciest appointments and the highest crash-test scores. But, since none of the models on the market come out first on all of those dimensions, that tells you next to nothing about whether, when the time comes to put his money on the table, he’ll buy a Ford or a Honda. Because it revealed something about his preferences respecting alternative packages of tradeoffs, what car he bought the last time is a lot more probative about what he’ll buy this time, especially when the competing models are still offering sets of tradeoffs similar to those they offered the last time he was in the market for a car.

Why should we expect things to be much different when we’re talking about voters confronting a choice between Republican and Democratic candidates offering agendas a lot like those they were peddling in the last election?

1 comment:

Anonymous said...

I couldn't agree with you more. I believe that people in general, and especially people who are being polled, rarely consider tradeoffs until it comes time to putting their money down, i.e., buying a car or in casting a vote.

However, it may be the case that many of the voters will think they are making a tradeoff if they vote Republican next time. It's one thing to want deficit reduction and no tax increase and it's another thing to pay for deficit reduction by giving up your entitlements. Saying it (last election) and realizing you actually will have to do it (this election) may be two different things.

The next election cycle may be the most interesting in my adult life (I'm a baby-boomer).