Exit Poll 'Reluctant Bush Responder' Hypothesis

The graph below shows the Refusal Rates (vertical) for exit interviews. The states are along the x-axis and have been sorted such that the 'bluest' states are on the left, and the reddest states toward the right (simply bush%-kerry%).

i made this graph because some have argued that perhaps one of the reasons that the exit polls were so wrong is because redvoters were less likely to agree to be exit interviewed (for one or other reason). the graph doesnt necessarily disprove this theory - but eyeballing it seems to imply that red staters weren't averse to being interviewed.



(UPDATE - this pic isnt working for some reason - lemme know if u want it)

this table from Mysterypollster also seems to indicate a relatively consistent propensity to complete interviews - independent of the patisanship of the precinct.


and joe knapp observes that rural areas had low 'error' rates, which also implies that at least some redvoters were happy to be interviewed at high rates.