A release from the Eric Stewart campaign:

Scott DesJarlais has launched a television commercial that is breathtaking in its hypocrisy.
 
“Scott DesJarlais voted to cut agriculture by $60 billion, destroy Medicare and raise the Social Security retirement age to 70,” said Eric Stewart. “DesJarlais is a typical politician. He says one thing in Tennessee and votes the opposite in Washington. This ad amounts to breathtaking hypocrisy.”
 
Stewart made his statement as his campaign released a video on YouTube that takes the images from his ad and replaces the audio to highlight DesJarlais’ votes against Medicare, Social Security and farmers.
 
“DesJarlais made a series of votes on March 29, 2012 that would have done tremendous damage to the very people who sent DesJarlais to Congress,” Stewart said. “The Ryan/DesJarlais budget would have been a disaster for working families, farmers and small businesses right here in Tennessee.”
 
Stewart added, “DesJarlais is even out of step with his fellow Republicans. More than 100 Republican Members of Congress voted against amending the Ryan/DesJarlais budget amendment that would have changed the retirement age for Social Security to 70.”
 
Stewart’s web ad, entitled “Accidental Congressman” takes DesJarlais to task for his votes. Recent polling, available below, shows DesJarlais is very vulnerable.

And that polling memo? Glad you asked:

According to the results of our recent baseline survey1 this contest is remarkably fluid and without question winnable. Indeed, from the outset first term Republican incumbent Scott DesJarlais posts profoundly poor numbers nearly across the board. DesJarlais’s name ID sits at a scant 54 percent, with just 26 percent giving him warm ratings and 14 percent cool; his job approval is net negative, with just 30 percent saying excellent or good and a third (33 percent) saying fair or poor; and most importantly, just 30 percent of voters say they would vote to re- elect him while 37 percent say someone new should be given a chance, and another third (34 percent) are unsure.

In more encouraging news for the Eric Stewart campaign, DesJarlais underperforms the generic Republican in this district, and begins just barely over majority status, 37 to 53 percent. Notably, Stewart is virtually unknown to this electorate today and when we introduce Stewart and DesJarlais in a balanced message exercise, this contest immediately snaps to a statistical dead heat with Stewart surging to 47 percent and DesJarlais falling to 49 percent. By the survey’s final ballot Stewart’s lead climbs to 7 points, 47 to 40 percent.

The shifts that occur here from start to finish are profound to say the very least, especially when one considers that party self-identification in this district favors Republicans by an 18-point margin. Nonetheless, by the final ballot Stewart opens up a 32 point gender gap, leading among women in the end by 22 points while trailing among men by 10. In vote rich Rutherford County, which is key to this contest, Stewart’s begins tied with DesJarlais, yet moves the contest to a 14 point edge, 49 to 35 percent. In the new portion of the 4th CD overall Stewart takes a commanding 11 point lead by the final ballot, while the contest shifts to an absolute dead heat in the old portion of the district.

Make no mistake about the findings of this survey – DesJarlais was swept into office in a tidal wave in 2010, and without that wave fueling him he proves to be an incredibly weak incumbent. He lacks a personal connection with these voters, and more importantly, his record in Congress as well as his personal life, are solidly out of step. If Stewart has the resources to prosecute this campaign and not be out-communicated by DesJarlais, these data confirm he can win.

And for you public opinion nerds, the explanatory footnote:

These findings are based on a survey of 600 likely November 2012 general election voters in TN CD4. Calling took place from June 14-18, 2012 and interviews were conducted by professional interviewers supervised by Myers Research | Strategic Services staff. The data were stratified to reflect the projected geographical contribution to the total expected vote. The margin of error associated with these data at a 95 in 100 percent confidence level is +/- 4.0 percent. The margin of error for subgroups is greater and varies.