Not that interested in politics myself, but I was curious about the South Carolina primary. You can see the exit polls here; seems like Mike Huckabee won evangelicals while McCain won non-conservatives. Nothing too surprising. As someone who has generally been of the opinion that evangelical prejudice against Mormomism will prevent Mitt Romney from getting the nomination, I was curious how evangelicalism tracked the primary results by county. I took these data and ran some correlations with the primary results by candidate. What I did was simply calculate the proportion of evangelicals out of the reported church members within each county (evangelicals divided by the unadjusted number). The proportion wasn’t as important as the pattern of proportions across the counties to my mind. Since over 90% of the precincts are back, I’ll report my results now below the fold (though I’ll update later, I doubt it is going to affect the bottom line).
| Correlation between candidate proportion and evangelical proportion by county | ||||
| Huck | McCain | Thomp | Romney | |
| Evanglical proportion | 0.57 | -0.54 | 0.22 | -0.40 |
| Correlation between candidate proportions by county | ||||
| Huck | McCain | Thomp | Romney | |
| Huck | – | -0.56 | 0.00 | -0.78 |
| McCain | – | – | -0.68 | 0.16 |
| Thomp | – | – | – | 0.01 |
| Romney | – | – | – | – |
I am struck by the strong negative correlation between Huckabee support and Romney support. McCain makes sense since many conservative activists dislike him for a variety of reasons, but Romney has become to some extent the “establishment candidate” (e.g., National Review endorsement). Yet it seems that the foot soldiers of the Christian Right are voting their identity as opposed to their issues. Or, perhaps more plausibly the weighting of issues for the foot soldiers was always different from their ostensible leaders.
Note: Here’s the data file.
Comments are closed.