Seeing Bush with approval ratings in the low 30s begs the question: so what? He's not on the ballot this fall, and won't be ever again. There is, however, a historical correlation between presidential approval ratings and midterm Congressional election results (at least at the House level). This site:
http://www.ropercenter.uconn.edu/... compiles historical presidential approval ratings dating back to 1937, when Gallup first started doing the poll.
Comparing the last presidential rating prior to a midterm election to the net results of that election (
http://clerk.house.gov/...) yields some interesting results:
Year |
President |
Party |
Approval |
House Res.
|
1938 |
Roosevelt |
D |
52 |
-76.5 |
1942 |
Roosevelt |
D |
70 |
-46 |
1946 |
Truman |
D |
33 |
-54.5 |
1950 |
Truman |
D |
41 |
-28 |
1954 |
Eisenhower |
R |
61 |
-18.5 |
1958 |
Eisenhower |
R |
57 |
-48.5 |
1962 |
Kennedy |
D |
62 |
-3 |
1966 |
Johnson |
D |
44 |
-47.5 |
1970 |
Nixon |
R |
58 |
-12 |
1974* |
Ford (Nixon) |
R |
54(24) |
-48.5
|
1978 |
Carter |
D |
49 |
-15 |
1982 |
Reagan |
R |
42 |
-26.5 |
1986 |
Reagan |
R |
63 |
-5 |
1990 |
Bush I |
R |
58 |
-8 |
1994 |
Clinton |
D |
45.5 |
-54 |
1998 |
Clinton |
D |
64 |
+5 |
2002 |
Bush II |
R |
62 |
+8 |
Notes:
1. 1974 is a weird one to handicap, as Nixon resigned three months before the election. I included both Ford's approval at election time and Nixon's final rating.
2. In terms of net partisan seat change, sometimes there is a .5 because of third party candidates holding House seats; a GOP loss to an Indy is scored as D +.5, as is a Dem win from an Indy. The exception is in recent times I count Bernie Sanders as a Dem. I'm sure that some more research would reveal Indy Reps. that could be counted wth one party or other, but I wasn't motivated to do the research for this diary.
3. In the modern age, there are a glut of organizations doing approval polling, so that the 1994, 1998, and 2002 presidential approval number is an average of many polls.
Observations
Not being tech savvy, I'm not sure how to convert these numbers into a graph, but if you do, you will see a strong correlation between approval and House seat loss--especially if you focus on 1962 to the present. Polling prior to '62 wasn't all that accurate.
Even including the full set, it's really only the Roosevelt and Eisenhower presidencies that don't fit the mold: both remained highly popular but their party lost seats. That can be explained by the fact that there was a cult of personality following both men--who were decidedly larger than life figures--allowing their approval numbers to stay artificially high even when there was discontent; discontent that got taken out on their party's Congressional candidates.
Leaving those two to the side, you can see that in the remaining 13 midterms the president was either highly popular (58% approval or higher), or unpopular (under 50%). I count 1974 as 39%--the average of the final Nixon rating and Ford's.
In the 6 elections where the president was highly popular at the time of the midterm (1962, 1970, 1986, 1990, 1998, and 2002), the president's party had the following performance in the midterm House elections: -3, -12, -5, -8, +5, +8. This yields an average of -2.5 seats.
In the 7 elections where the president was unpopular at the time of the midterm (1946, 1950, 1966, 1974, 1978, 1982, and 1994), the president's party had the following performance: -54.5, -28, -47.5, -48.5, -15, -26.5, -54. This yields an average of -39.1 seats. The best performance by an unpopular president's party in House elections is the Dems' loss of 15 seats in 1978.
So, what does this mean? If Bush stays unpopular, there will be a wave at the Dems' back. Sure, it will be dampened somewhat by the district maps, but that alone is unlikely to protect the GOP majority from public dissatisfaction with their president.