RedKing told us there was some anti-GOP sentiment in NC after they elected douchebags in 2010. This could be the local factor.
Just to be clear on this, it would need to be
very strong.
Using the numbers from 538, Nate has NC as a
Romney +5.2 lead in state fundamentals. Obama's projected national lead is about 2.4%, ish. That means, if there were absolutely no extra local factors, Obama would be expected to lose NC by ~2.8%. Right now there appears to be an extra 2-2.5% Democratic effect on the local level, pulling Obama back to only a 0.3% on the straight average of polls.
The thing is, that's a pretty substantial local effect already, taking into account demographics and national lead. One thing worth watching will be whether that local effect increases much more than the national support levels during the convention bump, and more importantly how much of the local support sticks around as that bump fades. Nate's maths actually suggest that this figure isn't especially accurate and largely removes the entire local effect, which is worrying.
As far as general anti-Republican sentiment in NC I'd need to see a serious shift in the local numbers.
This PPP poll (democratic leaning group) suggested that both state Republicans and Democrats had roughly equal disapproval ratings (-15% each) but with the Democratic governor really getting it in the neck (-29%). That despite that poll having a 44/36/20 Dem/Rep/Ind split, and exactly equal Obama/Romney voters.
Those are a month old now (15th July), so for that to be the local effect we are seeing in Obama's numbers this month we would need to see some strong shifts in those figures.