Saturday, June 07, 2008

Arizona 3: New Poll Reveals Little

By Nathan L. Gonzales

Polling memos are only as valuable as the numbers in them. A May 27 Bennett, Petts & Normington memo in Arizona’s 3rd district claims that Rep. John Shadegg (R) is “extremely vulnerable” but provides few and flimsy numbers from the actual survey to support the claim.

The May 18-20 survey, conducted for Shadegg’s Democratic challenger, attorney Bob Lord, showed the Congressman with a 31 percent re-elect number and 75 percent name identification.

But those numbers are only useful with context. The memo doesn’t give the other half of the re-elect question (the percentage of people who say they will vote for a “new person”). And 75 percent name ID is not terrible for a Member of Congress in a major metropolitan area, according to a second Democratic pollster not involved in the race.

It’s the numbers missing from the polling memo that tell the real story. The initial head-to-head ballot between Shadegg and Lord is nowhere to be found. Want Shadegg’s job approval number? Won’t find it here. The favorable/unfavorable numbers for the candidates (including Lord’s name ID) are not included either.

What does the memo include? Favorable and unfavorable ratings for Arizona Gov. Janet Napolitano (D), who isn’t on the ballot this year, and President Bush, who isn’t running either, but at least its standard to include his numbers when Democrats are making their case. And it refers to results of state legislative races that aren’t in the district.

The survey (and memo) also showed a presidential matchup, with Sen. John McCain (R-Ariz.) leading Sen. Barack Obama (D-Ill.) 48 percent to 43 percent in a district President Bush won with 58 percent in 2004. If McCain is underperforming Bush by 10 points in Arizona, Republicans should brace themselves for an electoral massacre that will make 2006 look like a birthday party.

Shadegg did have some issues with decisiveness after retiring and then un-retiring earlier this election cycle. But the missing numbers in the Democratic poll must have been pretty good for the Congressman, or else they would have been included in the polling memo.


This item first appeared on RollCall.com on June 5, 2008. 2008 © Roll Call Inc. All rights reserved. Reprinted with permission.