Exit Poll Deviations vs. Governor's Party and Straight Party Voting Option

[Bill Bored]

I am not a huge believer in the infallibility of Mitofsky's exit polls however, when combined with other information, they can be interesting.

Here are the average Red Shifts for states in each of the following categories:
All states=50
Dem Governors=DG
Repub Governors=RG
All states with Straight Party Voting=SP
Dem Governors w/Straight Party Voting=DG/SP
Repub Governors w/Straight Party Voting=RG/SP


Categ____Av Rd Sh|Categ______Av Rd Sh
50 (n=50) -1.77% | SP-- (n=17) -2.24%
DG (n=22) -1.40% | DG/SP (n=9) -1.70%
RG (n=28) -2.06% | RG/SP (n=8) -2.84%


Clearly there were greater exit poll deviations in states with Republican Govs. but also those with Straight Party Voting. The combination of the two had the highest average Red Shift.

I have omitted DC because they don't have a Governor.

I have included MI in the DG/SP category, but it may not have had SP voting in 2004. Since MI's Red Shift was -1%, it wouldn't make much difference.

Now let's say that the exit polls were off. Is there any reason to believe that they were off more in some states than others? And why would they be off more in those with Republican Govs and/or Straight Party Voting? Wouldn't the within-precinct exit poll errors be relatively consistent from state to state, regardless of who's in charge or which voting option they have available?

I will stop now and let TIA or others more familiar with the exit polls make some comments.

Here is a list of the states in each category in case anyone wants to check the figures, governors, SP voting option, or weight them, etc:

RG: NH VT SC AK NE AL NY CT MA RI MS OH FL MN ID UT CO GA AR KY NV MD MT HI TX CA SD ND

DG: DE NC PA AZ LA IL WI NM WV IN MO NJ WA WY IA MI OK ME VA OR TN KS

SP: NH SC AL RI MS UT KY TX PA WI NM WV IN MO IA MI OK

RG/SP: NH SC AL RI MS UT KY TX

DG/SP: PA WI NM WV IN MO IA MI OK

While it is possible to exploit the Straight Party Option in touch screen/DRE votin' machines to favor one candidate over another, resulting in votes switching or defaulting to either Bush or Kerry (or perhaps undervotes), this should only have been available in states that actually allow straight party voting in the first place. Anywhere else it's just plain illegal, whether it favors a particular candidate or not, right?

The fact that SP voting, where legal, seems to have favored Bush (at least according to the exit poll deviations) suggests the possibility that it could have been used elsewhere to affect the outcome of the election.

If there were machine problems in states ostensibly without straight party voting, such as votes switching to Bush, etc. this might be a clue to some illegal machine configurations out there. I have not checked this yet, but the incident reports might provide some insights into this possibility.

[Melissa G]

1. Nice research BB! n/t

[Melissa G]

2. In Texas, The SOS who oversees elections is appointed

by the Governor. Is that true in the other states?

[bleever]

3. Wow, that's a gem Bill.

That's an elegant piece of analysis.

Class: discuss.

[TruthIsAll]

4. Here is a table of deviations with Gov. included

Repub Gov. in 13 of 17 states which deviated beyond the MOE
for Bush

	Poll	Exit	Final	Vote	Odds	Dev			Dev>	Bush
St	MoE	Poll	Vote	Dev	 1 in	Favor	Gov	SP	MoE	Flip
AK	3.18%	40.14	36.17	-3.97%	138	Bush	R		YES	
AL	3.57%	41.08	37.08	-4.00%	72	Bush	R	SP	YES	
AR	2.61%	46.93	44.74	-2.19%	20	Bush	R			
AZ	2.27%	46.60	45.03	-1.57%	11	Bush	D			
CA	2.22%	55.73	55.21	-0.53%	3	Bush	R			

CO	1.95%	49.07	47.35	-1.72%	24	Bush	R			
CT	3.27%	58.47	55.24	-3.24%	38	Bush	R		YES	
DC	1.92%	91.63	90.63	-1.00%	6	Bush				
DE	3.48%	58.44	53.82	-4.62%	216	Bush	D		YES	
FL	1.84%	49.93	47.47	-2.46%	231	Bush	R		YES	yes

GA	2.48%	43.11	41.58	-1.53%	9	Bush	R			
HI	4.38%	53.32	54.37	1.05%	3	Kerry	R			
IA	1.96%	50.67	49.54	-1.13%	8	Bush	D	SP		yes
ID	3.91%	33.33	30.71	-2.63%	11	Bush	R			
IL	2.60%	57.13	54.99	-2.14%	19	Bush	D			

IN	3.17%	40.97	39.46	-1.51%	6	Bush	D	SP		
KS	3.65%	34.60	36.97	2.37%	10	Kerry	D			
KY	3.00%	40.76	39.99	-0.76%	3	Bush	R	SP		
LA	2.38%	44.50	42.63	-1.87%	16	Bush	D			
MA	3.10%	66.46	62.70	-3.76%	114	Bush	R		YES	

MD	3.07%	57.04	56.25	-0.79%	3	Bush	R			
ME	2.20%	54.83	54.48	-0.36%	3	Bush	D			
MI	1.98%	52.55	51.73	-0.82%	5	Bush	D	SP		
MN	2.09%	54.61	51.76	-2.85%	263	Bush	R		YES	
MO	2.11%	47.48	46.33	-1.15%	7	Bush	D	SP		

MS	3.44%	43.20	39.91	-3.30%	33	Bush	R	SP		
MT	3.78%	39.28	39.51	0.22%	2	Kerry	R			
NC	2.10%	47.31	43.72	-3.60%	2508	Bush	D		YES	
ND	3.63%	33.58	36.09	2.51%	11	Kerry	R			
NE	3.37%	36.54	32.53	-4.01%	102	Bush	R		YES	

NH	2.27%	55.50	50.68	-4.81%	63547	Bush	R	SP	YES	
NJ	2.49%	56.13	53.13	-3.00%	108	Bush	D		YES	
NM	2.22%	51.34	49.42	-1.93%	23	Bush	D	SP		yes
NV	2.13%	50.66	48.67	-1.99%	30	Bush	R			yes
NY	2.47%	63.97	58.79	-5.17%	49430	Bush	R		YES	

OH	2.21%	52.06	48.75	-3.31%	602	Bush	R		YES	yes
OK	2.38%	34.73	34.44	-0.29%	2	Bush	D	SP		
OR	3.00%	51.22	51.97	0.75%	3	Kerry	D			
PA	2.22%	54.41	51.13	-3.28%	521	Bush	D	SP	YES	
RI	3.30%	64.24	60.48	-3.76%	77	Bush	R	SP	YES	

SC	2.34%	45.79	41.31	-4.48%	11104	Bush	R	SP	YES	
SD	2.45%	37.42	39.09	1.67%	11	Kerry	R			
TN	2.29%	41.15	42.78	1.63%	12	Kerry	D			
TX	2.31%	36.84	38.49	1.65%	12	Kerry	R	SP		
UT	3.18%	29.93	27.06	-2.87%	26	Bush	R	SP		

VA	2.59%	47.96	45.65	-2.31%	25	Bush	D			
VT	3.56%	65.69	60.34	-5.35%	628	Bush	R		YES	
WA	2.12%	55.07	53.60	-1.47%	12	Bush	D			
WI	2.08%	50.21	50.20	-0.02%	2	Bush	D	SP		
WV	2.35%	45.19	43.48	-1.72%	13	Bush	D	SP		
WY	3.50%	32.07	29.70	-2.37%	11	Bush	D			
	2.71%	48.84	47.00	-1.84%	9	Bush				
										
										
										
										
										
										
	Poll	Exit	Final	Vote	Odds	Dev			Dev>	Bush
TOTAL	MoE	Poll	Vote	Dev	 1 in	Favor			MoE	Flip
Avg	2.85%	48.84	47.00	-1.84%	10	0				
Med	2.59%	49.07	47.35	-1.93%	15	0				
Average										
Bush	2.81%	50.40	48.09	-2.31%	2060491	0			16	5
Kerry	3.34%	40.93	42.41	1.48%	3	0			1	0
										
East	2.78%	55.11	52.31	-2.80%	3153	22			12	3
Cent	2.80%	43.17	42.35	-0.82%	31	11			3	1
Mount	3.04%	41.54	39.68	-1.86%	15	7			0	1
Pac	3.06%	51.10	50.26	-0.83%	25	3			1	0
										
Region	> MOE	1 in	ToBush	1 in						
East	12	32Tril 22	4.2mill						
Cent	3	146	11	10						
Mount	0	1	7	28						
Pac	1	8	3	2						

[Bill Bored]

7. TIA, some of your deviations are different from mine.

I'm using the ones from your post:
<http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=140579&mesg_id=141029>
which are unweighted. Did you weight them here, or what has changed?
Thanks.

[TruthIsAll]

8. Weighted vs. Unweighted averages

Weighted average = state vote deviation / total votes in all the states.

Unweighted average = the average of the deviations without regard to state population size.

I'm not at home now so I can't say for sure why their are differences. What comes to mind is I may have used Freeman's exit poll numbers in one of them and my original 2-party numbers in the other. But even so, they should be close.

The numbers you are using may be older, before my latest update.
Or mine might me old.

I will check it out later and give you an answer.

[Bill Bored]

9. Thanks.

I can change mine if they are wrong. They are your unweighted numbers from Dec. 11.

[TruthIsAll]

10. Dec. 11 is way too old. Go with the stats I gave you.

..

[Bill Bored]

11. OK. Not much difference though:

Categ____Av Rd Sh|Categ______Av Rd Sh
50 (n=50) -1.86% | SP-- (n=17) -2.17%
DG (n=22) -1.38% | DG/SP (n=9) -1.63%
RG (n=28) -2.23% | RG/SP (n=8) -2.79%

And from the incident reports, it seems MI had SP in 2004 so these are correct.

[berniew1]

5. There was touch screen switching in 18 states; and straight party default

from Dem to something else in at least 11.
Thats a lot of votes; plus all of the documented manipulation and suppression and dirty tricks.
http://www.flcv.com/ussumall.html

People should not have allowed the straight party voting debacle to happen, and then once it happened the results should have been questioned and thrown out in many cases; elections redone perhaps.

What point is there in having an election if you aren't going to count the votes as the voters intended??

[Bill Bored]

6. Bernie, what took you so long?

Now here's the thing:

FL is NOT supposed to have straight party voting, unless this is a recent developemnt. Do you know if it's legal there?

[Bill Bored]

12. I was mistaken!

I mixed up a few states in the straight party list when I did the above calculations. The difference in the average Red Shifts is smaller than I originally stated, however, while I compared SP voting states to all 50 states, I never compared SP states to states without SP voting. So here are the revised numbers:

Average Red Shift in all 50 states: -1.86%
Average Red Shift in 17 SP voting states: -2.03%
Average Red Shift in 33 states without SP voting: -1.77%

Still a greater Red Shift in the SP voting states.

Here's a list of the SP states in order of their Red Shifts (largest shift first):
NH SC AL RI NC PA UT NM WV IN MO IA MI KY OK WI TX.

I'm working on some theories about how SP voting could have been configured in GEMS in states that don't allow it in order to affect the outcome of the election. That should prove interesting.