Monday, March 23, 2009

Democrats Could take US Rep seat if they hold Democratic base and break even with Others

The point of the last two posts was that if Democrats ,who have such a large registration advantage , were to get a decisive majority of Democrats who vote in the US House race they could win by winning that large majority of their own party’s voters and by breaking even among Others (unaffiliated voters and third party voters). 25,000 to 70,000 more Democrats than Republicans have usually voted in recent elections,depending on whether it is a presidential or nonpresidential year.

In 2010, a nonpresidential year the difference will be closer to 25,000 –30,000 than 2008’s 71,000 plus advantage,but that is still sizable.
Here is what the voter difference by registration has looked like the last statewide few elections:
2002 Democrats 224,130 registered & 101,080 Dems voted (45%)
2002 Republicans 175,325 registered & 87,695 Reps voted (50%)
2002 13,385 more Democrats than Republicans voted

2004-Democrats 240,999 registered & 165,185 Dems voted (69%)
2004-Republicans 181,510 registered & 127,056 Reps voted (70%)
2004-38,129 more Democrats than Republicans voted

2006- Democrats-246,149 registered & 116,955 Dems voted (48%)
2006-Republicans-178,655 registered & 90,176 Reps voted (50%)
2006-26,779 more Democrats than Republicans voted

2008-Democrats 279,916 registered & 199,095 Dems voted (71%)
2008-Republicans 181,858 registered & 127,346 Reps voted (70%)
2008-71,749 more Democrats than Republicans voted

Since 1992 this has not taken place in the Congressional race,but statewide Democrats have done it in other races so it is possible.

For election results:
http://elections.delaware.gov/information/electionresults/election_archive.shtml
For breakdown of voters by registration the AGP (Age Group & Party) report was used which is a separate report produced by the Election Department on the same website

Sunday, March 15, 2009

In 1984 & 1992 It can't be proven Castle got Democratic votes,but it can be proven his rivals lost some.These were still his closest races to date.

In his two most competitive statewide races, it can not be proven Mike Castle got Democratic votes.It can be proven his opponents lost Democratic votes. William Quillen in 1984 and S B Woo in 1992 each had votes total below the number of Democrats who voted. Castle exceeded the number of Republicans who voted,but it can not be proven he received any Democratic votes. Castle beat Quillen 56% to 44% for Governor. Castle beat Woo 55% to 43% for US Representative .The remaining 2% went to a Libertarian candidate in 1992.


1984
Castle received 135,250 votes when elected Governor

If he received all GOP votes cast 88,888
And all Others cast 54,579
Which equals 143,467
Meaning he would need to get zero Democrats,since

143,467 is greater than 135,250

This does not prove Castle did not get any Democratic votes, but it proves that you can not mathematically prove he got any Democratic votes which might mean he got a smaller percentage of Democratic votes in 1984 than in 2004, 2006 & 2008 when it can be proven he received thousands of votes from Democrats.


Castle got 8,000 fewer votes than the combination of Republicans and Others. For him to have gotten a substantial number of Democratic votes would mean he received a diminished number of votes cast for him by Republicans and Others. Since Castle got 15,000 more votes than any other Republican statewide except Ronald Reagan (who received 152,190), this seems unlikely.
The Republican US Senate candidate, John Burris, received the fewest statewide votes for a GOP candidate in 1984. Burris received 98,101 votes in his unsuccessful campaign against Joe Biden.It appears that 98,101 of the 143,467 Republicans & Others constitute the mathematical equivalent of Republican straight ticket voting since all Republincan candidates got at least this amount of votes. This means that mathematically only 45,366 voters from this group were in play. Since Castle received 37,149 votes over 98,101, it can be assumed he received a fair portion of this group of 45,366 along with a smattering of Democrats.

The Democratic candidate who received the fewest votes in 1984 was Walter Mondale, the candidate who ran for President against Ronald Reagan. Mondale received 101,656. This is the mathematical equivalence of the Democratic straight ticket vote since all Democratic candidates got at least this amount of votes. Judge Bill Quillen, the Democratic candidate for Governor against Mike Castle, received 108,315,so he exceeded the mathematical equivalent of the straight ticket vote. This number ,however, includes some votes cast by people who are not registered Democrats. 114,610 Democrats voted and Quillen received 108,315, meaning at least 6000 Democrats either voted for someone else or decided not to vote in the Governor’s race. 12,000 voters who voted in the presidential race did not vote in the governor’s race,although this may have been to Quillen’s advantage as he lost by 26,935 votes and Mondale lost by 50,334.

Since 114,610 Democrats voted in 1992, we know that at least 6,000 did not vote for Quillen. This can be due to voting for Castle or they may have stopped voting without voting in this race.

While we can not prove mathematically that Castle received any votes from Democrats, we know Quillen got fewer than 100% of all Democratic votes. In a race with a 26,935 victory margin for Castle those 6,000 would not have changed the outcome,but they would have narrowed the margin.


For election results:
http://elections.delaware.gov/information/electionresults/election_archive.shtml

For breakdown of voters by registration the AGP (Age Group & Party) report was used which is a separate report produced by the Election Department on the same website



1992
Castle received 153,037 votes

If he received all GOP votes cast 106,372
And all Others cast 57,405
Which equals 163,777
Meaning he would need to get zero Democrats,since

163,777 is greater than 153,037

This does not prove Castle did not get any Democratic votes, but it proves that you can not mathematically prove he got any Democratic votes which might mean he got a smaller percentage of Democratic votes in 1992 than in 2004, 2006 & 2008 when it can be proven he received thousands of votes from Democrats.

Castle got 10,000 fewer votes than the combination of Republicans and Others. For him to have gotten a substantial number of Democratic votes would mean he received a diminished number of votes cast for him by Republicans and Others. Since Castle got more votes than any other Republican statewide, this seems unlikely.

The Republican Gubernatorial candidate, B Gary Scott, received the fewest statewide votes for a GOP candidate in 1992. Scott, received 90,725 votes in his unsuccessful campaign against Tom Carper.It appears that 90,725 of the 163,777 votes cast by Republicans & Others constitutes the mathematical equivalent of Republican straight ticket voting since all Republican candidates got at least this amount of votes. This means that mathematically only 73,002 voters from this group were in play. Since Castle received 62,312 votes over 90,725, it can be assumed he received a fair portion of this group of 73,002 along with a smattering of Democrats

The Democratic candidate who received the fewest votes in 1992 was SB Woo, the candidate who ran for US Representative against Castle. Woo received 117,426. This is the mathematical equivalence of the Democratic straight ticket vote, since all Democratic candidates got at least this amount of votes. Since 127,059 Democrats voted in 1992, we know that at least 10,000 did not vote for Woo. This can be due to voting for another candidate (though not necessarily Castle, since Libertarian Peggy Schmitt recieved 5,661 votes) or they may have stopped voting without voting in this race. 13,496 people voted in the presidential election, but did not vote in the US Representative race. Due to the secret ballot, it is not possible to quantify this voter dropoff by party affiliation.
While we can not prove mathematically that Castle received any votes from Democrats, we can verify that Woo got fewer than 100% of all Democratic votes. In a race with a 35,611 victory margin for Castle those 10,000 would not have changed the outcome,but they would have narrowed the margin.


For election results:
http://elections.delaware.gov/information/electionresults/election_archive.shtml

For breakdown of voters by registration the AGP (Age Group & Party) report was used which is a separate report produced by the Election Department on the same website

Tuesday, March 10, 2009

Mike Castle's recent Democratic Voting Support can be verified (up to a point) and conjectured beyond that point

You may think Mike Castle has crossover appeal, but it can be mathematically proven that he has drawn significant numbers of votes from Democratic voters in recent years because he could not get the vote totals he has gotten without getting a mathematically verifiable amount of votes from Democratic voters and a potentially higher number from Democrat voters that can be conjectured if we assume that not all Republicans and Others voted for him AND we assume that at least some of the voter dropoff occurs among Republicans and Others.

2008

Castle received 235,437 votes

If he received all GOP votes cast 127,346
And all Others cast 87,121
Which equals 214,467
He would need to get at least 20,970 Democrats,since
20,970
+214,467
235,437

Since 199,095 Democrats voted in 2008, Castle had to have gotten at least 10.5% (20,970) of all Democrat votes, even if he received all the votes cast by Republicans and Others. If we assume he got fewer than 100% of all Republican and Other votes cast, whatever amount that might be pushes up the number of Democrats who voted for Castle.

413,562 voters cast ballots in 2008 statewide in Delaware, but only 385,457 cast votes in the U.S. Representative race. 28, 105 voters entered the voting booth and left without voting for any candidate in the Congressional race. If we can assume that at least several thousand of that 28,105 voters who dropped off the ballot without voting for Congress were Republicans or Others (since the 214,467 is all Republican & Other ballots cast,not necessarily in the US Rep race), that is an additional quantity of Democrats who would have voted for Castle to make up his total vote count.

So at least 10.5% (20,970) and possibly several thousand (representing several percentage points) more Democrats voted for Mike Castle.

If you don’t think this is significant, it might be worth noting that although several Democrats statewide won decisive victories by margins of over 100,000 votes none had a vote total that was more than the combined turnout of Democrats and Others (286,216). Jack Markell won the Governor’s race by over 140,000, but still received 266,861 or about 20,000 below 286,216. While I don’t doubt that some of them got Republican votes, it can’t be proven from the numbers the way Castle’s ability to get Democratic votes can be proven. Since Democrats outvoted Republicans 181,858 to 140,543, Democrats can win by holding their base and getting enough Others. A Republican would need more Democratic defections.
________________________________________________________________

2006
Castle received 143,897 votes

If he received all GOP votes cast 90,176
And all Others cast 50,922
Which equals 141,098
He would need to get at least 2,799 Democrats,since
2,799
+141,098
143,897

Since 116,955 Democrats voted in 2006, Castle had to have gotten at least 2.4% (2,799) of all Democrat votes, even if he received all the votes cast by Republicans and Others. If we assume he got fewer than 100% of all Republican and Other votes cast, whatever amount that might be pushes up the number of Democrats who voted for Castle.

258,053 voters cast ballots in 2006 statewide in Delaware, but only 251,694 cast votes in the U.S. Representative race. 6,359 voters entered the voting booth and left without voting for any candidate in the Congressional race. If we can assume that at least several thousand of that 6,359 voters who dropped off the ballot without voting for Congress were Republicans or Others (since the 141,098 is all Republican & Other ballots cast,not necessarily in the US Rep race), that is an additional quantity of Democrats who would have had to vote for Castle to make up his total vote count.

So at least 2.4% (2,799) and possibly several thousand (representing several percentage points) more Democrats voted for Mike Castle

It is mathematically impossible that his Democratic opponent, Dennis Spivack received all Democratic votes cast since 116,955 Democrats voted and his combined vote total of 97,565 & the dropoff of 6,359 equals 103,924 which is less than the number of Democratic voters who came to the polls. On the other hand, Democrat Jack Markell in the treasurer’s race with 174,388 had a vote total that exceeded the combined turnout of Democrats and Others (116,955 plus 50,922 which equals 167,877) ,meaning Markell mathematically had to get several thousand Republican voters to vote for him.
________________________________________________________________

2004
Castle received 245,978 votes

If he received all GOP votes cast 127,056
And all Others cast 79,008
Which equals 206,064
Meaning he would need to get at least 39,914 Democrats,since
39,914
+206,064
245,978

Since 165,185 Democrats voted in 2004, Castle had to have gotten at least 24.2% (39,914) of all Democrat votes, even if he received all the votes cast by Republicans and Others. If we assume he got fewer than 100% of all Republican and Other votes cast, whatever amount that might be pushes up the number of Democrats who voted for Castle.

371,249 voters cast ballots in 2004 statewide in Delaware, but only 356,045 cast votes in the U.S. Representative race. 15,204 voters entered the voting booth and left without voting for any candidate in the Congressional race. If we can assume that at least several thousand of that 15,204 voters who dropped off the ballot without voting for Congress were Republicans or Others (since the 206,064 is all Republican & Other ballots cast,not necessarily in the US Rep race), that is an additional quantity of Democrats who would have had to vote for Castle to make up his total vote count.

So at least 24.2% (39,914) and possibly several thousand (representing several percentage points) more Democrats voted for Mike Castle in 2004.


For election results:
http://elections.delaware.gov/information/electionresults/election_archive.shtml

For breakdown of voters by registration the AGP (Age Group & Party) report was used which is a separate report produced by the Election Department on the same website

Wednesday, March 4, 2009

It paid to win where the most votes were cast in the 5th County Council Special Election

An interesting pattern appears if we break down the polling places by the size of the impact they had on the overall votes cast. Lisa Diller won 11 of 14 polling places, but what appears as important is that she won in the places that turned out the greatest number of voters.

Impact is not the same as turnout percentage in this discussion. We are talking about numbers of voters. The 13-22nd which voted at George Wilson Community Center had a 18.35% turnout which is a high percentage in a special election with an overall 5.53% turnout,but since this district only has 109 registered voters the 20 votes cast there had little impact. By comparison, McVey Elementary School was the polling place for 3 election districts comprised of 4,176 registered voters It had only a 5.6% turnout, but those 234 voters made it the third highest polling places in terms of numbers of voters.
It is better to have a small piece of a very big pie than a big piece of a very small pie.


There were fifteen polling places and this pattern can be seen when we break them into turnout groups .

High Impact Polling Places
First Presbyterian Church –Total Turnout 241-Diller 128-Moulthrop 113
(2nd - 23rd & 5th –23rd) (7.95% turnout-241 out of 3,032 voters)
Downes Elementary School-Total Turnout 309-Diller 201-Moulthrop 108
(3rd-23rd & 4th-23rd) (11.01% turnout-309 of 2,806 voters)
West Park Elementary School-Total Turnout 100-Diller 79-Moulthrop 21
(7th-23rd)(6.86 % turnout-100 out of 1,458 voters)
McVey Elementary School-Total Turnout 234-Diller 129-Moulthrop 105
(2nd-25th,6th-25th, & 7th-25th) (5.60% turnout-234 of 4,176 voters)
Brookside Elementary School-Total Turnout 112-Diller 68-Moulthrop 44
(3rd-25th & 4th-25th) (3.66% turnout-112 of 3,058 voters)

In these five polling places, representing ten election districts, Diller beat Moulthrop 605 to 391 or 214 votes. She won overall by 263, so these areas provided much of her victory margin.
These ten election districts have 14,668 of the district’s 25,081 registered voters (as of 2-1-2009, but likely to fluctuate over time) . This is 58.5% of the county council district’s registered voters. The 996 votes casts at these 5 polling places represented 73.6% of the overall turnout district wide.Diller won 60.7% in these areas that comprised almost three quarters of votes cast.


Medium Impact Polling Places
Main Towers, Main Street-Total Turnout 34-Diller 26-Moulthrop 8
(1st -23rd) (4.35% turnout-34 of 781 voters)
Aetna Fire Hall- Total Turnout 35 –Diller 23-Moulthrop 12
(6th-23rd)(2.69% turnout-35 out of 1,302 voters)
Newark Library- Total Turnout- 55- Diller 40-Moulthrop 15
(1st-25th)(4.80% turnout 55 out of 1,145 voters)
Glasgow High School-Total Turnout 81-Diller 44-Moulthrop 37
(5th- 25th &9th-25th)(3.77% turnout-81 of 2,150 voters)
Newark High School-Total Turnout 56-Diller 34-Moulthrop 22
(8th- 25th)(6.33% turnout-56 out of 884 voters)

Diller won this group 167 to 94 or with 64% of the vote.


By contrast, in the five polling places with the lowest turnout, Moulthrop won 62.5% of the vote.
Low Impact Polling Places
Maclary Elementary School-Total Turnout 19-Diller 7-Moulthrop 12
(13th-21st) (4.03% turnout-- 19 out of 472 voters)
Shue Middle School- Total Turnout 24-Diller 14-Moulthrop 10
(14th-21st) (8.73% turnout—24 out of 275 voters)
George Wilson Community Ctr-Total Turnout 20-Diller 5-Moulthrop 15
(13th-22nd)( 18.35% turnout-20 out of 109 voters)
Medill School-Total Turnout 14-Diller 1-Moulthrop 13
(3rd-24th) (1.08% turnout 14 out of 1,301 voters)
Jennie E Smith Elem School Total Turnout 19-Diller 9-Moulthrop 10
(6th-24th) (1.24% turnout 19 out of 1528 voters)
These five polling places combined for 96 votes or 7.1% of total turnout.
These five election districts make up 3712 of the 25,081 registered voters or 14.8% of the district’s potential voters.

High Impact polling places account for 73.6% of votes cast-Diller wins 605-391

Medium Impact polling places account for 19.3% of votes cast- Diller wins 167-94

Low Impact polling places account for 7.1% of votes cast-Moulthrop wins 60-36

When asked why he robbed banks Willie Sutton is quoted as having said “Because that’s where the money is”. On February 19, 2009 those five High Impact polling places in the 23rd and 25th Representative Districts were the places where the votes were. Winning there appears to be why she won.


Election results and turnout data is from the New Castle County Department of Elections:

http://electionsncc.delaware.gov/special/2009_5th_ncc/5th_ncc.shtml