Earlier, I gleefully reported the good news out of North Carolina. Four consecutive polls of the state have shown Kerry within single digits of Dear Leader, and two consecutive polls have shown the state to be a toss up if Edwards is on the ticket.  Considering all of this, it now seems likely that the GOP's partisan advantage in the state has dropped from around 13 or 14 to somewhere around 7 or 8 (or maybe even less, since undecideds tend to break for the challenger). However, while a supposedly safe Bush state with fifteen electoral votes is now on the edge of being in play, a different, supposedly safe Kerry state with fifteen electoral votes might now be on the brink of being in play: New Jersey.

Here is the partisan index in New Jersey since 1976:

'76: GOP + 4.222
'80: GOP + 3.679
'84: GOP + 2.677
'88: GOP + 5.915
'92: GOP + 3.186
'96: DNC + 9.338
'00: DNC + 15.323

The first problem anyone would probably notice with New Jersey being considered a "safe Dem" state is that as recently as 1992 it slightly leaned toward the GOP. During the five election cycles between 1976 and 1992, New Jersey showed very little partisan movement, shifting only 3.5 points from one "extreme" to the other. However, in just two election cycles the state shifted toward the DNC by almost 20 points. New Jersey's shift was so sudden and so extreme, that it is difficult to imagine that it is based purely on demographic or ideological shifts. More likely, it appears to be significantly based upon events specific to the 1996 and 2000 election cycles.

Here the last four trial heats out of New Jersey (none have trend lines):

Q-poll: Kerry 46, Bush 43, Nader 5 (5/10-5/16, 1,122 RV, MoE 2.9)
 Star-Ledger / Eagleton-Rutgers: Kerry 43, Bush 37 (4/28-5/4, MoE 4)
Rasumssen Kerry 51, Bush 39 (4/20, 500 LV, MoE 4.5)
FDU Bush 48, Kerry 44, Nader 5 (4/3-10, 802 RV, MoE 3.5)

These numbers are all over the map, but on average they point to a small, 4-5 point lead for Kerry. Certainly, this is a far cry from the Clinton and Gore landslides of recent years.

What could account for this shift? Here are some thoughts:

-McGreevey is very unpopular (check the Q-poll and FDU links for numbers)
-New Jersey is doing better job-wise than most of the nation
-New Jersey has lower gas prices than most of the nation
-The massive, very recent pro-DNC shift was not an accurate reflection of the real partisan trends in the state
-Lingering post-9/11 loyalty to Bush

Combined, all of these could explain Kerry's smallish lead in the state. However, I would like to offer a different suggestion: Kerry is still comfortably ahead in New Jersey.

Post-Super Tuesday Unfavorables in New Jersey
Q-poll: Bush 40, Kerry 28 (poll linked above)
Star-Ledger: Bush 44, Kerry 27 (poll linked above)

In both polls, Kerry does not have great favorables, which explains why Bush is competitive in the state in most trial heats. Jersey girls and boys do not like Bush, but have yet to decide upon Kerry. However, Kerry's far lower unfavorables point to his massive growth potential relative to Bush in the state. With unfavorables so much lower Bush's, the vast majority of undecideds will break Kerry's way as the campaign progresses. So, while there may be some good reasons to argue that New Jersey has shifted to a lean DNC state rather than a solid DNC state, I believe the favorable / unfavorable numbers reveal that the surprising Bush strength in trail heats is an illusion. Bush is losing in New Jersey despite maxing out in the state--he can only fall further behind.

This is not the case in North Carolina. While Bush has higher favorables and equal unfavorables to Kerry in the Tar Heel state, Kerry can still play his North Carolina trump card: John Edwards. Polls have shown that a ticket with Edwards would instantly make North Carolina a toss-up in a close election, and that Edwards is more popular than Bush in the state. Bush has no such silver bullet in New Jersey. His only option is to continue to spend like a drunken sailor in the state in an attempt to crank Kerry's unfavorables. However, considering his current financial situation, it seems very unlikely that he will continue to run ads in the Garden state.

Update: The partisan index does not reflect whether a candidate won or lost a particular state. Instead, the partisan index reflects how well a candidate performed in a state relative to that candidate's share of the national popular vote. Clinton won NJ in 1992, but by less of a margin than he won the popular vote. Thus, while he won the state, the partisan index still favored the GOP. This also explains why in 1984 the partisan index for NJ was at less than 3 points, even though Reagan won the state by more than twenty.