17 March 2008

The final 10 races in the DEM primary season - three scenarios.

I posted this write up once, called "A look at the last 12 states or territories in the DEM primary race". WY and MS have come and gone, and the field, excluding possible re-do's in MI and FL, has been narrowed to ten.

 

I have created a table to take another way to look at those ten last delegate battlegrounds, starting with the smallest and increasing to the largest. This is a mathematical calculation of what it will take for Hillary Clinton to close the approximately 168 pledged delegate gap between herself and Barack Obama (by my calculations on 03/16, it's 1,450-1,282, including the extrapolated returns from the TX caucuses, which will probably go even more for Obama than currently projected). I am deliberately leaving the unpledged or so-called superdelegates out of this – the quote at the end of this article will explain why.

 

So, using the purely proportional system, underneath is a table of the states, from smallest to largest and then the necessary percentage to get 1, 2 3, 4 or 5 more delegates over the middle line (which means, an even 50-50 between the candidates). For the larger states, I also calculate for 2,4,6,8 and 10 delegates over the middle. And for the largest three, I calculate for 3,6,9,12 and 15 delegates over the middle. There is overlap here, for instance, I calculate Puerto Rico twice: once in the 1-2-3 and 4 column, and once in the 2-4-6 and 8 column.

 

The smaller the state, the higher the winning  percentage necessary to get even one delegate over the middle. For this reason, the smallest states are more likely to look more and more even in terms of delegate distribution,, regardless of the victor's winning margin.

 

Don't forget, the number of delegates over the middle, when doubled, equals the margin. In the case of states with an odd number of delegates, this means double the number of delegates over the middle +1. Hillary Clinton must achieve a more than +84 over the middle without Obama winning even one state in order to catch up to him.

 

In this same write up as there were 12 battlegrounds left instead of 10, I wrote:

 

"The color of the state initials indicates my take on the likely tendency of the state, based on geography, demographics and past performance of the candidates in similar states in the region. For instance, Puerto Rico has an overwhelming hispanic population, it is very likely to go for Senator Clinton in a landslide. Mississippi is one of those deep south states in the so-called "Goldwater Belt", and neighbors states which also have large black populations and have already gone for Senator Obama- and therefore this state is also likely to go for Senator Obama. This is just my opinion, but is grounded based on the outcome of the last 43 races."

 

And indeed MS went for Obama in a landslide (61%-37%) , as universally assumed would be the outcome. And WY, a western caucus, also went for Obama (61%-38%) , although in a slightly lesser landslide than neighboring ID (80%-17%) , for instance.

 

So, here's the table:

 

 

STATE

DEL

+1

+2

+3

+4

+5

Guam

4

66.7

75.01

--

--

--

SD

15

56.7

63.4

70.01

76.7

83.4

MT

16

56.3

62.6

68.8

75.1

81.4

WV

28

53.6

57.2

60.8

64.4

68.0

KY

51

52

54

55.9

57.9

59.8

OR

52

52

53.9

55.8

57.7

59.6

 

 

 

 

 

 

 

STATE

DEL

+2

+4

+6

+8

+10

KY

51

54

57.9

61.8

65.7

69.6

OR

52

53.9

57.7

61.6

65.4

69.2

PR*

55

53.7

57.3

61

64.5

68.2

IN

72

52.8

55.6

58.4

61.2

63.9

NC

115

51.8

53.5

55.3

57

58.7

PA

158

51.3

52.6

53.8

55.1

56.4

 

 

 

 

 

 

 

STATE

DEL

+3

+6

+9

+12

+15

IN

72

54.2

58.4

62.6

66.7

70.9

NC*

115

52.7

55.3

57.9

60.5

63.1

PA

158

51.9

53.8

55.7

57.6

59.9

 

 

Legend: green = Obama, light blue = Clinton, red = tossup.

* For PR: 51% brings 27.5 delegates to 28, so +1 means 29, +2 means 30, and so on / For NC: 50.5% brings 57.5 delegates to 58, so +1 means 59, +2 means 60, and so on.

 

So, how to interpret this table: in order for to Hillary Clinton to win PA with 91 of 158 delegates to Obama's 67, or a margin of 24 delegates, she must come over 57.6% of the popular vote. Anything over 55% is considered a landslide, and in the case of a three way race, any winning margin of +10% upward is a landslide.  Now, she won Ohio big, with 54% and a 10% spread  in an extremely tight national race and polls show her up on Obama by an average of 8 points, with still lots of undecideds.

 

SCENARIO I

 

In my last write-up as there were 12 battlegrounds left instead of 10, I wrote:

 

"However, the math gets exponentially harder for Clinton to close the approximately 130-140 delegate gap should Barack Obama win big in the next two states, for if he wins there, then the delegate gap will increase in his favor and therefore require even higher percentages from Hillary in the remaining states to neutralize his margin advantage."

 

And indeed, the math has gotten much more difficult, especially since the delegate gap has grown. Here's how the last ten states come out if Clinton wins a 60-40 landslide in every single contest remaining:

 

State

Date

REG.

PD

BO-%

HC-%

Del.Mar.

BO-PD

HC-PD

PA

4/22

NE

158

40

60

+32

63

95

GM

5/03

--

4

40

60

+0

2

2

IN

5/06

MW

72

40

60

+14

29

43

NC

5/06

S

115

40

60

+23

46

69

WV

5/13

NE

28

40

60

+8

9

17

KY

5/20

MW

51

40

60

+11

20

31

OR

5/20

W

52

40

60

+10

21

31

PR

6/01

--

55

40

60

+11

22

33

MT

6/03

W

16

40

60

+4

6

10

SD

6/03

MW

15

40

60

+3

6

9

TOT

 

 

 

 

 

+116

224

340

 

In this scenario, she is still 52 pledged delegates behind Obama. So, a universal 60%-40% landslide in all states still does not get her ahead of Obama in the PD-count.

 

SCENARIO II

 

Here a similar scenario, with Hillary Clinton getting 65% (instead of 60%) in all of the remaining contests:

 

State

Date

REG.

PD

BO-%

HC-%

Del.Mar.

BO-PD

HC-PD

PA

4/22

NE

158

35

65

+48

55

103

GM

5/03

--

4

35

65

+0

2

2

IN

5/06

MW

72

35

65

+22

25

47

NC

5/06

S

115

35

65

+35

40

75

WV

5/13

NE

28

35

65

+8

10

18

KY

5/20

MW

51

35

65

+15

18

33

OR

5/20

W

52

35

65

+16

18

34

PR

6/01

--

55

35

65

+17

19

36

MT

6/03

W

16

35

65

+6

6

10

SD

6/03

MW

15

35

65

+5

5

10

TOT

 

 

 

 

 

+172

198

368

 

 

In this scenario, she pulls 4 PDs ahead of Obama, but in order to do so, she must land 65% in every single state. The chances of her getting between 65% in every single one of these states is statistically, nil. However, very large landslides in four of these states is definitely within the realm of possibility for her.

 

SCENARIO III

 

Here is the much more likely scenario of how the last 10 states will play out.

 

Six of those ten favour Obama and if current trends continue, he will win those states:

 

Guam – territory in the pacific, close to asia

OR – western pacific state, moving more left than CA, BO does better in GOP match-ups than HRC

MT, SD – upper west to upper midwest, but these are primaries instead of caucuses, however, Obama has swept the entire region till now.

NC - Obama has consistently led in polls. NC is his equivalent to HRC's PA.

IN - thought this was Clinton territory, but recent polling shows otherwise. Probably a lean Obama win.

 

Four of those ten favor Clinton, and strongly:

 

PA – the largest prize left, definitely Clinton country. She will win here. Obama's only objective will be to narrow the margin as much as he can.

KY, WV – same region, all Clinton country, great affinity for the name Clinton.

Puerto Rico – enormous hispanic population. Nuff said.

 

My percentage predictions are extrapolations from the region. For instance in PA: Clinton won OH 54-44, NY 57-40 and NJ 54-44 (avg: 55-43), but Obama won neighbouring MD 60-36 and DE 53-42. Still, PA favours Clinton, but I suspect she tops out at 58%. However, I don't see a 60% win even as a remote possibility her in PA. With the exception of AR, she won no primary or caucus with over 58.45% (RI). Her regional average is 53%, her national average is 54%. I give Clinton her current ceiling of 58% +1% in order to insure that she has no handicap.  I will not write the explicit reasons for the other nine states, suffice it to say that  I use exactly the same method. For Obama, I have handicapped his primary wins in the west by 10%, since they will not be caucuses, meaning, were they caucuses, bets would be that he would do 10% better. Currently, there is no polling data for Guam, Kentucky, Puerto Rico and South Dakota,  the polling data from Oregon is 6 weeks old and the polling data from Montana is three months old – and a lot has changed since then. The one Indiana poll from one month ago shows Obama with a +15 point lead, but also with 35% undecided.

 

State

Date

REG.

PD

BO-%

HC-%

Del.Mar.

BO-PD

HC-PD

PA

4/22

NE

158

41

59

+28

65

93

GM

5/03

--

4

60

40

-0

2

2

IN

5/06

MW

72

52

48

-2

37

35

NC

5/06

S

115

56

44

-13

64

51

WV

5/13

NE

28

35

65

+8

10

18

KY

5/20

MW

51

35

65

+15

18

33

OR

5/20

W

52

56

44

-6

29

23

PR

6/01

--

55

40

60

+11

22

33

MT

6/03

W

16

56

44

-2

9

7

SD

6/03

MW

15

57

43

-3

9

6

TOT

 

 

566

 

 

+36

265

301

 

Gold shading = likely Clinton win.

Aqua-blue shading = likely Obama win.

 

So, assuming that these conditions hold, and that Obama has leaner wins in states leaning toward him as Clinton in states leaning toward her, then at the end of the day, she gains 36 delegates on him, bringing the delegate deficit down to +132 for Obama.

 

And now, to quote from the article I found in a recent Newsweek article:

 

"I've asked several prominent uncommitted superdelegates if there's any chance they would reverse the will of Democratic voters. They all say no. It would shatter young people and destroy the party.

 

Clinton's only hope lies in the popular vote—a yardstick on which she now trails Obama by about 600,000 votes. Should she end the primary season in June with a lead in popular votes, she could get a hearing from uncommitted superdelegates for all the other arguments that she would make a stronger nominee (wins the big states, etc.). If she loses both the pledged delegate count and the popular vote, no argument will cause the superdelegates to disenfranchise millions of Democratic voters. It will be over."

 

 

 

4 comments:

  1. Anonymous17/3/08 14:46

    bonn it's rascal king good anlaylsis it seems your really good with numbers and satistics and what have you,not my strong area It looks like if thing hold obama is the nominee that what I gather from this I could be wrong but it,s definatly and up hill climb for Hillary to get the nomination. I don't know how i feel about that I think barack would be a great president,but i'm starting to get the feeling he is going to drop a big stinking pile of dukakis on us.
    happy saint patricks day/ octoberfest
    Rascal King/ James

    ReplyDelete
  2. Hi rascal, yes, numbers are my relaxation. But rest assured, the Clinton and Obama teams both know these numbers. Actually, it's time for the Clinton team to quit. This, seen statistically, is the very best thing it can do for the party. And I suspect that a good 150 still unpledged superdelegates may help to make that decision by declaring soon.

    Sometimes you only need a pebble to get a whole lavine to flow...

    ReplyDelete
  3. Excellent analysis. I agree with your third scenario. With Clinton winning PA by a bigger margin than OH. I am in North Carolina and have seen the polls tighten to a high single digit lead for Obama. I hope he can get a good victory out of NC to make up for most of PA's delegate loss.
    I completely agree also that if Obama has the most pledged delegates (which he will have) and the popular vote lead (excluding MI and FL) then he will get the nomination. Of course the popular vote lead is not an official measure - the delegates are.

    ReplyDelete
  4. I don't see how Clinton goes ahead in the popular vote either. Without an enormous victory in PA, this would certainly become impossible as well.

    ReplyDelete

Constructive comments and critique are always welcome. Please keep it polite and respectful.