I am going to offer an analysis of the next 6 primary contests, using polls published at Real Clear Politics this month. I will also look at the RCP average of polls (distorted, because it includes very old polls, and there are in fact no recent polls for either DE or RI) as well as the targets for the two candidates set by Dave Wasserman before the primary contest began in Iowa.
Since we cannot really compare polls by different organizations one to another, what I am offering is in fact a rough approximation of the status for those states where polling is available — reasonable for NY, PA and MD, and one poll for CT. In New York we have the advantage of two polls from Marist less than one week apart, which does show a trend. Let me address just that fact first.
The new poll from NBC/WSJ/Marist issued yesterday does not have complete tabs up yet. There are a few areas where we can compare data points, because the story about the latest poll provides information that is NOT in the 2-page Scribed document that is online. We know the overall preference among likely voters has gone from 55-41 Clinton to 57-40 Clinton. Here there may be one very important note to consider. The makeup of the poll in its split on gender among likely voters has changed, with the earlier poll being 62-38 female while the current poll is 58-42 female. Given Clinton’s strong support among women, that she increased her overall lead among likely voters while the percentage of that pool that was female decreased is worth noting. (the figures for residents and registered voters was 52-48 female in both samples).
So a few additional notes on the Marist Poll. Clinton has gained minimally among women, from 58-36 to 60-37, and among voters 45 and up, from 66-30 to 67-29. What is significant is that she has gone from effectively splitting men with Sanders to leading that group 53-44. Meanwhile Sanders has lost same advantage among younger voters, which before he was winning 62-37, and in the latest version was only winning 55-42. That number surprised me in the latest version. I suspect that might be an outlier, and thus looking at the internals where I can compare the two polls, I am not sure that the gain in the top line for Clinton really represents any meaningful change in the status of the race.
So let’s look at the overall picture, based on polling data.
For each of the five states, I will tell you the number of polls that are included, how many total delegates, and the Wasserman targets.
I will have columns for for RCP average, highest margin, median margin, and lowest margin.
I will have rows for percent margin, delegates for Clinton, delegates for Sanders, and net difference. I will then offer some commentary.
NEW YORK 8 polls, 247 delegates Clinton 122 Sanders 125
|
|
RCP avg |
hi |
median |
low |
% margin |
|
13.8 |
18 |
15 |
10 |
Clinton dels |
|
141 |
146 |
142 |
136 |
Sanders dels |
|
106 |
101 |
105 |
111 |
net for Clinton |
|
35 |
45 |
37 |
25 |
The Sienna Poll, which is the 10 point margin, would have Sanders 14 delegates behind his original targets. That seems like a low margin to me. After having watched or listened to the final 75 minutes of last night’s debate, I remain confident this will be a double digit margin, and I would suspect that Clinton will do no worse than the median above.
PENNSYLVANIA 2 polls, 189 delegates, Clinton 89, Sanders 96
|
|
RCP avg |
hi |
median |
low |
% margin |
|
16.0 |
11 |
8.5 |
6 |
Clinton dels |
|
110 |
109 |
103 |
100 |
Sanders dels |
|
79 |
81 |
86 |
89 |
net for Clinton |
|
31 |
28 |
17 |
11 |
In theory Sanders might be able to do better than the general run of polling in the states, since a lot of industrial jobs were lost in the state. Still, it is hard to imagine him actually winning the state, and if in fact he loses by double digits in New York on Tuesday, I would expect results above the median , rather than at the low number. Even at the low number he falls 7 delegates behind his original target, and loses ground in net delegates.
MARYLAND 3 polls, 95 Delegates, Clinton 53, Sanders 42
|
|
RCP avg |
hi |
median |
low |
% margin |
|
23.3 |
22 |
22 |
15 |
Clinton dels |
|
59 |
58 |
58 |
55 |
Sanders dels |
|
36 |
37 |
37 |
40 |
net for Clinton |
|
23 |
21 |
21 |
15 |
This should be a strong state for Clinton, given the substantial part of the electorate that will be African American. Given the heated Senate primary including an African American candidate, I expect to see that portion of the electorate maximized. I would suspect the results to be closer to the high margin than to the low margin.
Connecticut 1 Poll, 55 Delegates, Clinton 27, Sanders 28
|
|
RCP avg |
hi |
median |
low |
% margin |
|
n/a |
6 |
6 |
6 |
Clinton dels |
|
|
29 |
29 |
29 |
Sanders dels |
|
|
26 |
26 |
26 |
net for Clinton |
|
|
3 |
3 |
3 |
I have to suspect that the one poll somewhat overstates how Sanders will do, given the issue of the lawsuit on Sandy Hook, and the fact it was an issue in the April 14 New York Debate. This is probably the best chance Sanders has of winning a state in the next 6 contests.
Delaware, no polls, 21 Delegates, Clinton 11 Sanders 10,
No analysis possible based on polls. Would expect this state to behave very much like Maryland
Rhode Island, no polls, 24 delegates, Clinton 11, Sanders 13
No analysis possible based on polls. Hard to predict how it might perform. Possible similar to MA, except maybe a bit better for Clinton.
To make an overall assessment of the state of the race after 4/26 and giving Sanders the benefit of the doubt on the two states without polling (using the target info from Wasserman, which I do NOT think Sanders will reach), even at worst case, using the low margins for the other four states, Clinton would net +53 delegates. Using the median numbers she would net +77. Using the high numbers she would net +96.
I will be surprised if the number she nets out of the 6 sets is less than 70. I think it possible that it may exceed 100.
But we will see what if anything impacts the numbers. How will people who watched last night’s debate perceive it? What about those who only see clips or read summaries? What impact will Sanders being out of the country have? How might the results in New York affect the other primaries?
I heard nothing in last night’s debate, nor have I read anything since, to convince me that the basic dynamics, as reflected in the polling data that is available for these forthcoming states, has changed.
We will know soon enough.