Based on 124 polls from multiple sources assessing 57 competitive House seats taken in October or November (most within the past few days) and published at
Pollster.com, Democrats are leading, as of the day before the mid-term elections, in 33 races for House seats currently held by Republicans. Since there are zero (0) Democratic seats for which Republicans hold a lead, that means that if the latest polls hold the Democrats stand to pick up 33 Republican House seats, which would give them a 37 seat advantage in the House of Representatives of 236-199.
More specifically, in 21 of those 33 currently Republican seats, the Democratic candidate is leading in both the latest poll and (if applicable) the average of all poll results taken within 10 days of the latest poll. In the remaining 12 Republican seats where Democrats are leading, the lead in the latest poll is within the margin of error or the average of the polls taken within 10 days of the latest poll are within the margin of error.
In addition, there are 21 Republican held seats that are competitive (within 5 points), though the Republican was shown to be leading in either the latest poll (17 seats) or the average of the polls taken within 10 days of the latest poll. And, there are 3 currently Democratic held seats that are competitive by that definition, though the Democrat is leading in each of them.
It could be argued that the Republicans could easily win any of the 13 seats in which the Democrat lead is within the margin of error. But statistically, it is just about as likely that the Democrat could win any of the 21 competitive Republican held seats where they are not currently leading.
In the table below I listed the specific races according to the categories that I noted above. After the Congressional District I list the date of the last poll, the result of the last poll (+ means a Democratic lead), followed if applicable by the average result of all the polls taken within 10 days of the latest poll (which includes the latest poll), with the number of polls taken within those 10 days in parenthesis. Ive also noted races that involve scandals with the name of the disgraced Republican, whether or not he is running or has resigned (DeLay, Ney, Foley).
Dem lead in latest poll (and average within 10 days of that poll) above MOE (22 likely pick-ups for Dems)Dist. .. Last date . Result . Last 10 day average (# polls)AZ 8
10-29
.. +12
.. +11 (2)
CO 7
10-29
.. +16
.. +12 (3)
FL 16....10-1
.
+7
..
FoleyFL 22
. 11-2
.
+9
. +6 (3)
IL 6
. ..10-29
.. +14
.. +3 (4)
IN 8
.10-26
.. +10
.. +8 (2)
IA 1
.11-3
.
+21
.. +14 (2)
KY 3
.
11-1
.
+8
. +3 (3)
NE 3
10-29
.. +6
NY 24. .10-10
.. +11
..
NY 25. .10-26
.. +5 (2)
..
ReynoldsNY 29. .10-26
.. +11
.. +11 (2)
NC 11
..11-1
+9
. +8 (3)
OH 1.....10-26
.. +5 (2)..
OH 15. .10-10
.. +12
..
OH 18. .10-29
.. +20
..
NeyPA 6
.10-29
.. +5
+4 (3)
PA 7
.10-24
.. +7
WeldonPA 10
..10-29
.. +9
+10 (2)
SherwoodTX 22
..10-25
.. +8
DeLayWI 8
.10-26
.. +6
+3 (2)
Democrat lead within the margin of error (12 races for seats currently held by GOP)AZ 5
10-29
.. +2
CA 11
.10-26
.. +2
.
PomboCT 2
.10-30
.. +7
E (5)
CT 4
.10-29
.. +7
E (4)
CT 5
.10-29
.. +3
+5 (3)
FL 13
10-24
.. +2
+5 (2)
IL 10
...10-26
.. +2
E (2)
IN 2
..10-31
.. +3
+6(3)
KY 4
.10-29
.. +3
+2 (3)
NM 1.....11-2
.. +1
+3 (3)
NY 19
..10-26
.. +2
+3 (3)
NY 20.
11-3
.. +3
.. +5 (3)
SweeneyOther competitive races where seat currently held by Republican (21 races)AZ 1
10-26
.. -2
. -8 (2)
CO 4
..11-2
-1
. +2 (3)
FL 24
.10-21
.. -2
.
FeeneyID 1
11-1
. +4
+1 (2)
IN 9
11-3
. -2.
+3 (3)
IA 2
10-26
.. -2
.
MN 1
.10-26
. -3
.
NE 1
.11-3
-5
.
NH 2.....11-3
-1
. +1 (4)
NJ 7
.10-26
. -3
.
NY 26
.11-3
-4
. -5 (2)
NC 8
..10-26
. +4
-6 (2)
OH 2
..10-31
. +3
-1 (2)
OH 12
10-26
.. -5
.
PA 4
..10-26
.. -4
. -4 (3)
PA 8
10-30
. +5
-2 (4)
VA 2
..10-29
.. -8
. -2 (3)
VA 10
10-10
.. -5
.
WA 5
.10-26
. -5
.
WA 8
.10-28
. -6
. -2 (2)
WY
.
10-25
.. -4
. -6 (2)
Competitive races in seats currently held by Democrats (3 races)GA 12
10-26
. +3
IL 8
.10-26
. +5
+7 (4)
IN 7
10-20
. +5
+1 (2)
Qualifications to the above interpretationOf course this doesnt necessarily mean that the Democrats are going to gain at least 21 House seats tomorrow.
One possibility that would make these polls wrong would be a substantial turn of opinion favoring the Republicans. A somewhat ominous sign is that three out of five of the most recent
generic Congressional ballot polls, released yesterday, show the Democrats with a single digit lead (Pew Research 4; ABC/Washington Post 6; Gallup 7), following several consecutive weeks where all such polls showed the Democrats with a double digit lead, ranging from 11 to 23, though the two most recently released polls (CNN and FOX) show the Democrats with a 20 point and 13 point lead, respectively. The huge difference in recent polls is difficult to understand, and is way beyond the margin of statistical error. The good majority of individual House race polls are not recent enough to reflect the recent change in generic Congressional ballot polls, if indeed they represent a real trend. Nevertheless, even if they do represent a real trend, the Democrat lead is still plenty large enough that it should result in a Democratic takeover in the House.
Another possibility would be that the polls are systematically biased towards the Democrats. For example, that could prove to be the case if Democratic turnout is below what is expected or Republican turnout is greater than expected. However, given that polls have consistently shown an abnormally high level of enthusiasm for this election by Democrats, compared with Republicans, that possibility seems unlikely.
The last possibility that could prove these polls wrong would be massive election fraud. Enough said about that for now.
What use can be made of these numbers?For one thing, theyre fun to look at.
But more important, if the Republicans should keep control of the House after this election, which would be very unlikely in the absence of election fraud, these pre-election poll results should be compared with whatever exit polls are available to see if they are consistent with them. If both the exit polls and the pre-election polls indicate substantial Democratic gains that are not borne out by the official election results, and if the differences between the polls and the official results are well beyond the statistical margin of error (as was the case in 2004 with regard to differences between the exit polls and the official results in the Presidential race), then these numbers will need to be carefully analyzed for patterns in connection with what we know about the potential for fraud in the different races throughout the country.
Printer-friendly format
Email this thread to a friend
Bookmark this thread
(1000+ posts)
Send PM |
Profile |
Ignore
