Just a quick diary about a recent Gallup poll showing Hillary Clinton’s net favorability among Democrats increasing 14% since the Oct. 13 debate. It gives me a chance to try out the new table functionality of DK5 and also pose a question: is favorability a valid/useful measure on its own or is it perhaps just a secondary result of the overall preference. I.e., people may just tend to view the candidate they choose favorably, and all others unfavorably.
Here’s a table with the Gallup data for Clinton:
Hillary Clinton Net Favorable Rating among Democrats/leaners
|
Sep. 28 — Oct. 12 |
Oct. 25 — Nov. 7 |
change |
favorable |
72 |
78 |
+6 |
UNFAVORABLE |
23 |
15 |
-8 |
NET FAVORABLE |
49 |
63 |
+14 |
The sample size was 764 Democrats/leaners (pretty good by recent standards) of whom about 60% were contacted by cell phone.
Did approximately 7% of Democrats and leaners really stop disliking Clinton and switch to liking her in that time period? We know that in the same period Clinton’s lead in the polls has increased, which could be the really salient effect in terms of favor or disfavor.
For Sanders:
Sanders net favorable rating among Democrats/leaners
|
sep. 25 — Oct. 12 |
Oct. 25 — Nov. 7 |
change |
favorable |
49 |
51 |
+2 |
unfavorable |
10 |
13 |
+3 |
net favorable |
39 |
38 |
-1 |
Note that as of the most recent poll, 36% of people have neither a favorable or unfavorable opinion of Sanders. This compares to 7% for Clinton.
Some crosstabs:
Net favorables among Democrats/leaners, Oct. 25 - Nov. 7
|
Clinton |
Sanders |
difference |
All democrats |
63 |
38 |
+25 |
white non-hispanic |
55 |
50 |
+5
|
BLACK non-hispanic |
89 |
21 |
+68 |
conservative/MODERATE |
55 |
19 |
+36 |
liberal |
73 |
62 |
+11
|