Jump to Any Race
National Politics|By Charlie Cook, March 12, 2014

Democrats haven't had a week this bad since 2010 and its only Wednesday. While the headlines are focused on Democrats losing the special election in Florida's 13th Congressional district, even worse news came in the form of an NBC News/Wall Street Journal national poll released last night, along with four statewide surveys conducted by a highly-regarded Democratic pollster in key Senate race states.

One can make several arguments about the FL-13 special election outcome, but the bottom line remains that while it is appropriate to classify this district as competitive, President Obama carried FL-13 in both 2008 and 2012. Despite a relatively even funding environment (campaign, independent, and Super-PAC money combined) and a Democratic candidate was at least as good as, if not probably a lot better than the Republican alternative—a lobbyist—Democrats still lost. This loss underscores how completely different the turnout dynamic is in a presidential election year vs a special election or a midterm election; both of the latter usually consisting of an older and whiter voter demographic. If Democrats are to make a plausible argument for regaining a majority, they need not only to win races like this one in FL-13, but also in other districts that are even more unfavorable to Democrats.

In the NBC News/Wall Street Journal poll released on March 12th, President Obama had his lowest job approval ratings yet, coming in at 41 percent approval, versus a 54 percent disapproval rate among respondents. His approval ratings on handling the economy and foreign policy were both also at 41 percent, with 56 and 53 percent respectively disapproving. On more personal level, 41 percent saw Obama positively, 15 percent had neutral feelings about the President, and 44 percent reported negative feelings. When asked whether they preferred a Congress controlled by Democrats or by Republicans, respondents provided data that gave the GOP a one point edge overall. While that might sound insignificant, this poll question, for whatever reason, has historically been skewed three to four points towards Democrats. Despite the rather consistent skew towards Dems, these new numbers appear pretty much comparable to the two-point GOP edge in the last NBC/WSJ poll taken just before the 2010 Republican landslide victory.

When asked whether they thought the new health care law was a good idea or a bad idea, just 35 percent said it was a good idea, 49 percent said bad idea, both very similar data points to the results of this same question in December and January. Attitudes towards President Obama and the Democrats signature policy achievement are very negative and not getting any better.

The polling in four key Senate states were conducted by Hickman Analytics, an extremely competent Democratic polling firm headed up by Harrison Hickman, who has been in the business for over 30 years. The surveys were conducted for the Consumer Energy Alliance, a group supporting construction of the Keystone XL pipeline. In Arkansas, where Obama’s favorable ratings are just 32 percent, with a 65 percent unfavorable rating, Democratic incumbent Mark Pryor led his GOP challenger, Rep. Tom Cotton, by just three points among all likely voters. A lead of just 40 to 37 percent is a dangerous place for an incumbent to be, particularly when that that lead is over a lesser known opponent. Worse for Pryor, he trailed by Cotton two points, 41 to 39 percent, among definite voters, a group that most likely represents the demographics of the midterm election turnout. Pryor still holds an 18-point effective name recognition advantage over Cotton, suggesting additional room for growth for the challenger.

In Colorado, the survey testing incumbent Democrat Mark Udall was conducted before the GOP replaced their weak challenger, Ken Buck, with a far more attractive challenger, Rep. Cory Gardner. Udall led Buck by four points among likely voters (46 to 42 percent) and three points among the tighter screened definite voters (46 to 43 percent). Obama had a 44 percent favorable rating, coupled with a 52 percent unfavorable rating; it is important to recall that he carried the state in both 2008 and 2012.

The survey taken in Louisiana showed GOP Rep. Bill Cassidy leading incumbent Democrat Mary Landrieu by four points among likely voters (46 to 42 percent) and nine points among definite voters (49 to 40 percent). Cassidy’s effective name recognition was just 42 percent, half of Landrieu’s 95 percent, suggesting growing room for Cassidy. Obama had just a 41 percent favorable, 56 percent unfavorable rating in the state.

In North Carolina, incumbent Democrat Kay Hagan led her likely GOP challenger, state House Speaker Thom Tillis by four points among likely voters (45 to 41 percent) and five points among definite voters (47 to 42 percent). Tillis’ effective name recognition was just 34 percent, well under Hagan’s 84 percent, again, showing potential for growth for Tillis. Obama’s favorable ratings were 46 percent, unfavorable 50 percent.

With each of these sets of numbers, keep in mind that for well-known incumbents, there is a tendency for WYSIWYG (What You See Is What You Get), that incumbents generally don’t grow their actual vote much above their poll numbers, undecided voters tend to break more for challengers. That’s why these numbers should be so troubling for Democrats.

The bottom line here is that at least for today, this election is not about the myriad of problems facing the Republican Party (with minority, young, female and moderate voters) but instead is about President Obama and the Affordable Care Act, both deeply unpopular. The fight for the Senate is being fought in terrain far more challenging for Democrats (read more Romney than Obama states) and with a midterm electorate that is older, whiter, and much tougher for Dems than the one that re-elected Obama in 2012. Environmentally, this election reflects a mood not dissimilar to 2010; the big difference being in the House, where Democrats have relatively few vulnerable seats to protect, so the possibility of a party shift at a magnitude similar their 63-seat loss four years ago is extremely unlikely.

For those who like to get into the weeds, links to the full NBC/WSJ poll questionnaire as well as to the questionnaires and key crosstabs for the Senate race polls in Arkansas, Colorado, Louisiana and North Carolina are provided below:

For Further Reading:

NBC News/Wall Street Journal National Poll questionnaire

Questionnaire & Tables:

Arkansas Topline & Tables

Colorado Topline & Tables

Louisiana Topline & Tables

North Carolina Topline & Tables


Hickman Analytics conducted the four polls for Consumer Energy Alliance. In Arkansas, 400 likely 2014 general election voters were surveyed via telephone or cell phone between February 17th and 20th. In Louisiana, 404 likely 2014 general election voters were surveyed via telephone or cell phone between February 17th and 24th. In Colorado, 400 likely 2014 general election voters were surveyed via telephone or cell phone between February 17th and 20th. In Colorado, 400 likely 2014 general election voters were surveyed via telephone or cell phone between February 17th and 20th. Each poll carries a 4.9% margin of error.