(Well, almost)

From time to time, state polls come out that appear to be nonsensical, impossible outliers bucking long term voting and demographic trends. Usually, when these polls are released (showing Kerry up twelve in Iowa, Bush up twelve in Wisconsin, or some such result) it is not long before new polling information countermands these outliers and conventional wisdom regains hold (as I expect it soon will in California). However, once in a great while, polling information in a state consistently bucks conventional wisdom to the extent that it might be time for conventional wisdom to change. North Carolina is on the brink of just such a shift.

In the 2000 Presidential election, North Carolina was not competitive:

2000 National share of the two-party vote
Gore 50.268%
Bush 49.732%

2000 North Carolina share of the two-party vote
Bush 56.464%
Gore 43.536%

This difference translate to a sizable 13.464% GOP advantage in North Carolina, the sort of advantage that would take massive national landslide for Kerry to win the Tar Heel state this year. Further, placing this partisan advantage in the context of long term voting trends, North Carolina would actually appear to be growing more GOP as time goes by:

Partisan Advantage in North Carolina:

1976. DNC + 9.002
1980. DNC + 8.412
1984. GOP + 5.716
1988. GOP + 8.572
1992. GOP + 7.828
1996. GOP + 14.635
2000. GOP + 13.464

Over the past thirty years, North Carolina seems to have twice experienced a sizable pro-GOP shift: between 1980 and 1984, and between 1992 and 1996. Although Gore did slightly better in NC than Clinton did in 1996 (at least relative to their respective national vote totals), a 1.2% shift toward the DNC does not a trend make.

Given all of this, what are we to make out of the last four polls from NC?

3/30- 4/1: Bush 51 - Kerry 44

3/01-3/13: Bush 51 - Kerry 43

2/23-2/25: Bush 53 - Kerry 42

2/23-2/25: Edwards 50 - Bush 47

Further, along with his re-elect numbers, Bush's approval ratings in North Carolina are also hovering around the 50% mark. An Elon University poll conducted April 26-29 showed him at 50% overall approval, with 39% approval for the economy and 45% approval for Iraq.

Since the Iowa caucus, every single poll out of North Carolina has shown Democrats running closer to Bush than the partisan index would indicate as normal. While Kerry has never been close enough to suggest that he could win the state in an election decided by two points or less, it is entirely possible that he could start threatening to take the state even if he is only ahead in national trial heats by four or five points. And hell, with a statewide favorable rating well over 70%, John Edwards as Vice-President would probably instantly make North Carolina a toss up.

All of this data seems to indicate that North Carolina may have shifted back to its partisan levels of 1984-1992, when it leaned GOP but was not rock solid. I honestly have no idea what the demographic, economic, political or other reasons for such a shift would be. Of course, it is possible that no such shift has taken place and we are just experiencing a rare string of outlying polls, but that conclusion is becoming less likely with each passing month.