Approval Ratings: What Do the Numbers Mean? (Updated)

A recent Washington Post/ABC poll gives President Bush a 65% disapproval rating, with 52% strongly disapproving, and 28% feeling angry at him. Those are some of the highest disapproval ratings of any president in modern history (Washington Post). How survey participants labeled themselves: 32% conservative, 25% liberal, and 39% moderate.

That WaPo article states that Congress is "deeply unpopular," with a 37% approval rating. Some people fear that Congress is "unpopular" because Dems are pursuing the fired U.S. Attorney investigation (for recent background, see Washington Post).

I'm not sure that's true, given the approval ratings' partisan breakdown: 46% approve of Democrats in Congress, 34% approve of Republicans (Washington Post/ABC poll, Question #5).  This same poll that gives Congress a 37% approval rating also indicates that 55% of survey participants trust Congressional Democrats to handle the Iraq war. Talk about mixed feelings.

Wondering what Congress's ratings typically look like, I checked results of other polls (e.g., Gallup, and AP) listed at Polling Report from September 2005 - July 2007. During that time, Congress' approval rating has not even reached 50%. The highest I saw (unless my eyes played tricks) is 44%.

Might the 100-point scale have a different meaning when applied to the President than to Congress? Consider WaPo's early-post-9/11 poll, when Congress largely supported Bush and most Americans had terrorism/war on their minds:

                                                                 Approval (Dec. 2001)

                          Congress                                   59%

                          President                                   86%

I imagine it's easy for Congress' 535 members to simultaneously please and displease the same people. If Congress passes unpopular legislation -- even by a slim majority -- some people may give the whole entity negative credit.

Given that scenario, is it even possible for Congress to leave the doghouse for a significant period?  Below are some other interesting numbers:

First, as I've said before, I don't put much stock in isolated polls, though tracking changes in the same polls over time might indicate trends. Check out these examples of the new, Democrat-controlled Congress's approval ratings during its first 7 months:

                               Gallup          WaPo/ABC          AP/Ipsos

Jan. 2007                35%                43%                       32%

Feb. 2007               37%                41%                        34%

Mar. 2007              28%                --                             33%

Apr. 2007               33%                 44%                       40%

May  2007            29%                  39%                       35%

June 2007               24%                   --                             --

July 2007               27%                 37%                        24%

                                                  [Data from Polling Report.] 

Not only are each poll's numbers different but the trends aren't parallel. Below are the former Republican-controlled Congress' approval ratings during its last 7 months in control:

                            Gallup               WaPo/ABC        AP/Ipsos

June 2006                   27%                        --                        24%                                             

July 2006                    29%                       --                        27%

Aug. 2006                 27%                       36%                   29%

Sept.  2006                 29%                          --                      29%

Oct. 2006                   23%                     32/31%                27%   

Nov. 2006                26%                         36%                  26%

Dec. 2006                 21%                          37%                  27%

                                                       [Data from Polling Report.]

What do those numbers mean? Damned if I know, but I do notice some interesting things.

First, the WaPo/ABC poll's numbers were consistently higher for both the D-Congress and the R-Congress than the other two polls were.

Second, the average approval ratings for the D-Congress in all three polls (first 7 months) is higher than the averages for the R-Congress (last 7 months), assuming I didn't mistype anything into Exel:

                      Gallup       WaPo/ABC     AP/Ipsos            

D-Congress               30%          40%                 33%                  

R-Congress               26%          35%                27%

Not knowing the error margins, I don't know if the differences are significant. Even if they are, I don't know how to explain those differences. And is it even valid to compare the new Congress to the old Congress, given that the old Congress had more time to step on toes?

Because the WaPo approval ratings for Congress were consistently higher than other polls, I think of the WaPo polls as the best-case scenario.  That said, I looked at WaPo's approval ratings for Congress going back to 1974.  Here's some of the interesting stuff I found (ABC/Washington Post poll, question #4):

  * From 2001-07, Congress' high was 59% (Dec. '01, after Afghanistan bombing )

    * April 30, 2003, last time Congress hit 50% (just after Iraq invasion).

    * Jun. '03 - July '07, Congress at 31% - 44% (includes bipartisan-, R- and D-control)

 * Jan '94 - Oct 97: Gingrich Congress at 27% - 42% (avg. 35.8%, lower than today)

    * Dec. '01-Sep. '02  (heavy focus on war and terrorism): 51% - 59%: 

    * Jan.- Sep. '98 (heavy Monica Lewinski coverage): 45% - 56%

     * Jan. '05-Oct. '97:  (R-controlled Congress' first term): 27% - 42%

The long and short of it all is that I have no conclusions to offer, other than that polling numbers (especially relating to Congress) are hard to interpret. But they're interesting to look at, especially in relation to some of the medias' simplistic conclusions about poll numbers.

Comments

[Dyre42]

The amusing thing about those polls is that frequently show that most of those polled believe that their congressperson is doing a good job. Its the other 534 that suck.