National polls show that Republicans would take the brunt of the blame for a dive over the so-called fiscal cliff. A Washington Post/Pew Research Survey released this week found that a majority of the public (53 percent) would point the finger at Republicans if Congress fails to reach a deal on taxes and government spending. Just 27 percent would blame President Obama.

But are national polls really the right tools with which to understand the political consequences of this latest legislative limbo? After all, the Romney and the Obama campaign both eschewed national polling during the campaign, focusing instead on the battleground states that would deliver the Electoral College.

And, when it comes to the House, a national sample isn't particularly predictive either. There are very few Democrats or Republicans who sit in "swing districts." Just 81 members of the House - less than 20 percent - won their seats with less than 55 percent of the vote.

In other words, what the general public thinks is not necessarily indicative of what the voters in the individual districts think. In fact, most Republicans are much more worried about a challenge from their right in a primary than a challenge from a Democrat.

"National polls in a Presidential race are useless for sure, and they are useless to members of Congress trying to make decisions about their own personal politics," says Democratic pollster Jefrey Pollock. "But that doesn't make them totally worthless. Just because the congressional folks are reelected by large margins, it's still important to take temperatures of national sentiment. It's very interesting/important to note what people nationally think about who should be blamed because that drives media coverage and the narrative."

Republican strategist Jon Lerner agrees that national polling drives the national discussion, which in turn drives the national mood.

"National polls are necessarily broad, and they are therefore informative in a broad sense about the political environment and about national policy questions," says Lerner. "If national polls say the Iraq war, or Obamacare, are broadly unpopular, it says something about the direction the national political discussion will take on those issues and on others that might have less energy behind them."

Democratic pollster Geoff Garin says that you can actually use national data in "situations like now, when there is a national policy debate going on," to gauge voter sentiment in individual districts.

"If you believe, as I do, that the high ground in politics is being able to capture the high ground in the center, national polls tell you a lot about where the center is," Garin told me. "And if your sample is large enough so you can understand responses by partisanship, geography (region of country, type of area), race, other demographics, etc., you can take a pretty good guess about the driving forces and what they would look like in a particular state or CD. And it is just a lot more cost efficient to do a large sample national survey than a bunch of state surveys."

But, while national polls can be of some value in understanding the mood of the country going into the fiscal cliff debate, what they can't predict is the response of voters if we indeed fall off that cliff. How voters assess blame today may be very, very different from who they hold responsible in the aftermath of a crisis.