Using data from 33 out of 51 states (including Washington, DC) on support among Democratic voters and leaners for Hillary Clinton, we have found that her support among the "blue" states (states in which Bush lost in the 2004 Presidential election) is slightly higher than in the "red" states (states in which Bush won), but this difference could easily be due to statistical uncertainty. 17 of the states we looked at were red and 16 were blue. Hillary's average support among democrats is 32.8% in the red states and 35.0% in the blue states. Her highest support among Democrats for red states is 49%, for her home state of Arkansas and for blue states is 48%, for New York, where she is currently residing. Her lowest support among Democrats for red states is 20%, for North Carolina, where John Edwards is in the lead and for blue states is 28%, for Connecticut.
Specifically we found that her average support among the "blue" states (averaged by state) is 35.0%, with an uncertainty (standard deviation of the mean) of 1.6%, whereas among the "red" states it is 32.8% with an uncertainty of 1.3%. Thus, the difference is 2.2% with an uncertainty of 2.1%. The standard score (z-value) of the difference is thus 1.05, which turns out to correspond to a probability of 29% that this difference is due to statistical uncertainty. The cutoff which is usually used is z=2, corresponding to a probability of 5%. Thus, we conclude that the difference is statistically insignificant.
For more information on the data, please go to Hillary Red and Blue State Analysis Data.
