If there were two bits of consensus emerging from the conventional wisdom of this month's tax cut episode in Washington, they were roughly as follows:
- The tax cut deal struck between the President and his Republican adversaries on the Hill was exceptionally popular policy.
- The only detractors to the legislation were a handful of critics who solely represented, as one piece described it, the views of "elite liberal opinion".
Both of those points of analysis, given a deeper look at the available polling data, might not be quite as cut-and-dried as the traditional media community might insist.
Start with the former: the notion that the tax "compromise" was a universally beloved piece of legislation. This was an article of faith across the media spectrum, and seemed to be buttressed by data from ABC News (PDF file), Gallup, and Pew (PDF file).
Check out the ledes from the analysis for each of these polls:
"With decisive votes in Congress pending, Americans in a new ABC News/Washington Post poll broadly support the tax-and-benefits deal forged by President Obama and Republican leaders of Congress – the deficit be damned. Sixty-nine percent support the package overall, far outnumbering the 29 percent opposed."--ABC News
"The agreement between President Obama and congressional Republicans to extend tax cuts and unemployment benefits is getting strong bipartisan support. Overall, 60% approve of the agreement while just 22% disapprove. There are virtually no partisan differences in opinions about the agreement - 63% of Democrats approve of it, as do 62% of Republicans and 60% of independents. Among Democrats, liberals are as supportive of the agreement as are conservative and moderate Democrats."--Pew
"Americans, and particularly those who are following the tax issue closely, are substantially more likely to believe Congress should pass the new tax agreement reached by President Obama and Republican congressional leaders than to think Congress should not do so."--Gallup
The problem with the "everyone loves the deal" meme is that it makes a huge assumption about its supporters. That assumption, simply put, is that they knew all of the components of the deal when they offered their support for it.
If this seems like a harsh assessment, consider that it is reinforced, at least in part, by one of the above mentioned pollsters. When ABC polled four of the individual components of the poll, three of the four polled substantially worse than "the package" had done:
Bear in mind, that even the ABC poll took the most deal-friendly framing of the tax cuts, making no mention of the most controversial component of the tax cut (the extension of the cuts for the wealthiest Americans). When specific reference to the cuts for the wealthiest Americans were included in an early December CBS poll, just 26% favored "extending tax cuts for all", while 53% wanted to sunset them on the wealthy, and another 14% wanted them to expire altogether.
The point is that words, indeed, do matter. How did the package poll better than the sum of its parts? Check out the phrasing of the ABC question, after polling the four individual components:
Obama and the Republican leaders of Congress have reached an agreement to do all four of these things together as a package. Is this plan something you support strongly, support somewhat, oppose somewhat or oppose strongly?
Does this bias the results? It certainly could. Indeed, you could make a pretty compelling argument that this tests voter regard for compromise and bipartisanship (two things that tend to poll well), as opposed to actually polling voter sentiment on the package itself.
The other two surveys (Gallup and Pew) were framed similarly. In other words, if one was not intimately familiar with the details of the final package, all they knew when asked about the package was (a) tax cuts were included, and (b) it was a bipartisan agreement. It's pretty hard to find two things that, on a general level, are doing to poll better than that.
As Pollster's excellent Mark Blumenthal noted at the time (emphasis his): "What all of this means is that when pollsters ask questions about public policy, the answers provided are frequently a reaction to the language of each question rather than an expression of a pre-existing opinion about specific policy options."
Further undermining the notion that the deal was extraordinarily popular was the rather tepid opinions about the deal, even in the polls proclaiming its immense popularity. As our own Chris Bowers wrote two weeks ago:
Whenever survey responses are dominated by people who "somewhat" support / oppose or "somewhat" approve / disapprove of something, it is a clear sign that the universe sampled by the poll does not have strong opinions about the poll topic. And so it is the case with the tax cut deal as well.
According to the ABC / WaPo poll, only 20% of the country "strongly" supports the deal, and only 12% strongly opposes it. More than two-thirds of the country does not feel strongly about it either way. Even though news of the deal dominated political headlines for the past week, the country has not reacted strongly to it.
An ancillary point of analysis, which also made the rounds as Congress debated the deal, was that the opposition to the deal was limited to some squeaky wheels on the "professional left". As Salon's Steve Kornacki wrote in the wake of the package's passage:
As one prominent liberal after another cried "betrayal!" over President Obama's tax cut deal last week, I cautioned against reading too much into it: One of the most underappreciated stories of Obama's tenure has been the consistent disconnect between elite liberal opinion about the president and the view of rank-and-file liberals.
That point is more evident than ever today, with the release of a new Washington Post/ABC News poll -- conducted late last week and over the weekend, as the supposed liberal backlash against the deal was at its peak -- that shows Obama's approval rating with liberal Democrats sitting at 87 percent. That's essentially where it's been all year; it represents a statistically microscopic decline from where he stood with liberal Democrats in the last WaPo/ABC survey, taken about six weeks ago.
While some of Kornacki's points in the piece are well taken (I, too, strongly doubt that a primary challenge to the President would have any legs), to say that there was an absence of liberal backlash beyond the "elite" opinion makers would be, in my view, inaccurate.
There was, unquestionably, a disconnect between liberal Democrats and the rest of the electorate on the issue of a tax cut extension. While liberal Dems and other Dems were almost identical in their support for unemployment insurance extension, according to an early December Gallup poll, there was a huge gulf between the two on the tax cut issue, where liberal support (39%) for "extending tax cuts for all" was barely more than half of the support seen by other Democrats (64%). Furthermore, a pre-Christmas analysis by Gallup contradicted Kornacki, albeit after the fact. The analysis showed that while the President's overall job approval was steady, that steadiness was the balancing of an erosion of support from the left (from 88% in the Fall down to 80% today), and a bit of a spike among moderate Republicans (from 19% in the Fall up to 29% today).
We saw base discontent on this site, as well, where our own unscientific survey on the question found 75% opposition to full extension of the tax cuts. Now, the DK poll on the subject is not analogous to other surveys, because it is self-selected and limited to those reading this site. BUT, it is a decent cross-section of activist, left-of-center voters (over 12,000 of them).
And therein lies the potential peril. Kornacki's take ("What backlash?"), as well as many others similar to his, were based on the President's essentially stable job approval ratings. But approval ratings are a binary question: your only options are "approve" and "disapprove". Intensity of that opinion is not registered in a simple job approval question.
One might err on the side of approval, but be less inclined to knock on doors, donate money, and work the phones.
That is the danger for Democrats, a lesson that should have been well learned in November.
Throughout the 2010 cycle, analysts on the left (myself included) mollified themselves with polling that showed that even as Democratic approval was flagging, Republican approval was absolutely circling the drain. On Election Day, however, that didn't matter, because the participants were evenly split between 08 Obama voters and 08 McCain voters. Given that Obama had won in 2008 by nearly 10 million votes, an evenly split 2010 electorate was the recipe for a Democratic...um...shellacking.
Many other actions in the last month, to the immense credit of the President and his party in Congress, might create great progress in rebuilding base enthusiasm for 2012.
And if there was a "big picture" lesson to be learned from 2010, it is that said enthusiasm is critical to sustaining this presidency, and resurrecting the fortunes of the President's party in Congress.
Here's hoping that the new year shows that lesson has been learned.