Some Wacky Poll Results In New York
Posted: Sun Sep 26, 2010 2:59 pm
In all the years I've been following polling, I don't believe I've seen this many polls with such divergent results conducted over such a short period of time for a major race:
Poll Date Sample Cuomo (D) Paladino (R) Spread
RCP Average 9/16 - 9/22 -- 52.2 35.6 Cuomo +16.6
SurveyUSA 9/20 - 9/21 527 LV 49 40 Cuomo +9
Marist* 9/19 - 9/22 617 LV 52 33 Cuomo +19
Siena* 9/16 - 9/21 801 RV 57 24 Cuomo +33
Quinnipiac 9/16 - 9/20 751 LV 49 43 Cuomo +6
Rasmussen9/16 - 9/16 500 LV 54 38 Cuomo +16
So, over a space of six days, we have five polls conducted, where, depending on which one you want to believe, Cuomo is either in serious trouble or cruising to an easy landslide victory, or something in between....
The conventional wisdom had been that Cuomo would beat this unknown Buffalo businessman easily, but from what I've been reading so far he's been taking the race for granted to such an extent that his campaign has been virtually invisible. That would seem to be a huge mistake against an extremely wealthy opponent in a heavily anti-incumbent year, in a state that has been plagued by scandals and corruption to an extreme degree in state government (both by the past two governors and other leading figures in state legislature)
Those conditions are tailor made for a political outsider to knock off an incumbent office holder who is where he is primarily because of his last name...
Now, in an effort to try an determine the outlier in these poll results, let's compare these results to recent poll results in the New York Senate race between Sen. Gillibrand and Congressman Joe DioGuardi :
Poll Date Sample Gillibrand (D) DioGuardi (R) Spread
RCP Average 9/16 - 9/21 -- 49.8 39.0 Gillibrand +10.8
SurveyUSA 9/20 - 9/21 527 LV 45 44 Gillibrand +1
Siena 9/16 - 9/21 801 RV 57 31 Gillibrand +26
Quinnipiac 9/16 - 9/20 751 LV 48 42 Gillibrand +6
Rasmussen 9/16 - 9/16 500 LV 49 39 Gillibrand +10
I think this makes it pretty clear where the problem is....
It's the Siena poll.....
I've said all along that I felt strongly that Gillibrand would be very vulnerable because in all the polls taken for months, despite the fact that the GOP hadn't even picked a candidate till a couple of weeks ago, she rarely polled over 50% You can check this for yourself :
http://www.realclearpolitics.com/epolls ... -1469.html
Look at all the previous polls before the one's I posted. With one exception, the polls that do have her over 50% are all Siena polls...and even they never had her anywhere near 57%
There are two things wrong with Siena:
First, it would appear that in this most recent poll, their must have been some sort of flaw in the methodology, because even by their own standards, the numbers look skewed....
The other problem is they are doing "registered voter" versus "likely voter" polling, which is always a less accurate predictor of the sentiment of the voters who will show up at the polls, particularly so in mid-term elections, and even more so in a year when there is an historically huge gap in voter enthusiasm.
So it seems to me that if you want to get a true take on where voter sentiment in New York stands at the moment in these two races, , you throw out the Siena polls and average the rest. If you do this, you see Cuomo with a clear lead, but not a mortal lock, and Gillibrand in a close fight.
Poll Date Sample Cuomo (D) Paladino (R) Spread
RCP Average 9/16 - 9/22 -- 52.2 35.6 Cuomo +16.6
SurveyUSA 9/20 - 9/21 527 LV 49 40 Cuomo +9
Marist* 9/19 - 9/22 617 LV 52 33 Cuomo +19
Siena* 9/16 - 9/21 801 RV 57 24 Cuomo +33
Quinnipiac 9/16 - 9/20 751 LV 49 43 Cuomo +6
Rasmussen9/16 - 9/16 500 LV 54 38 Cuomo +16
So, over a space of six days, we have five polls conducted, where, depending on which one you want to believe, Cuomo is either in serious trouble or cruising to an easy landslide victory, or something in between....
The conventional wisdom had been that Cuomo would beat this unknown Buffalo businessman easily, but from what I've been reading so far he's been taking the race for granted to such an extent that his campaign has been virtually invisible. That would seem to be a huge mistake against an extremely wealthy opponent in a heavily anti-incumbent year, in a state that has been plagued by scandals and corruption to an extreme degree in state government (both by the past two governors and other leading figures in state legislature)
Those conditions are tailor made for a political outsider to knock off an incumbent office holder who is where he is primarily because of his last name...
Now, in an effort to try an determine the outlier in these poll results, let's compare these results to recent poll results in the New York Senate race between Sen. Gillibrand and Congressman Joe DioGuardi :
Poll Date Sample Gillibrand (D) DioGuardi (R) Spread
RCP Average 9/16 - 9/21 -- 49.8 39.0 Gillibrand +10.8
SurveyUSA 9/20 - 9/21 527 LV 45 44 Gillibrand +1
Siena 9/16 - 9/21 801 RV 57 31 Gillibrand +26
Quinnipiac 9/16 - 9/20 751 LV 48 42 Gillibrand +6
Rasmussen 9/16 - 9/16 500 LV 49 39 Gillibrand +10
I think this makes it pretty clear where the problem is....
It's the Siena poll.....
I've said all along that I felt strongly that Gillibrand would be very vulnerable because in all the polls taken for months, despite the fact that the GOP hadn't even picked a candidate till a couple of weeks ago, she rarely polled over 50% You can check this for yourself :
http://www.realclearpolitics.com/epolls ... -1469.html
Look at all the previous polls before the one's I posted. With one exception, the polls that do have her over 50% are all Siena polls...and even they never had her anywhere near 57%
There are two things wrong with Siena:
First, it would appear that in this most recent poll, their must have been some sort of flaw in the methodology, because even by their own standards, the numbers look skewed....
The other problem is they are doing "registered voter" versus "likely voter" polling, which is always a less accurate predictor of the sentiment of the voters who will show up at the polls, particularly so in mid-term elections, and even more so in a year when there is an historically huge gap in voter enthusiasm.
So it seems to me that if you want to get a true take on where voter sentiment in New York stands at the moment in these two races, , you throw out the Siena polls and average the rest. If you do this, you see Cuomo with a clear lead, but not a mortal lock, and Gillibrand in a close fight.
