On Friday I wrote about Clinton's ability to consistently win among late deciders. The last SUSA polling from Indiana suggests the pattern may be repeating in Indiana.
The SUSA Indiana poll has Clinton up 54-42. If you look at the recent polling in Indiana, what becomes apparent is most of the volatility is around Clinton's number. Clinton ranges from a low of 42 (Zogby) to a high of 54 (SUSA). Obama, in contrast, ranges only between 42 (SUSA) and 46 (PPP). This repeats the pattern in Ohio where Clinton ranged in final polling from 56 to 44, but Obama ranged in all but one poll from 42 to 44 and in Pennsylvania where Clinton ranged between 46 and 54 while Obama polled within a narrower range (in most cases between 42 and 44.
This polling suggests that the undecided in places like Ohio and Pennsylvania actually at some level lean to Clinton, something that the exit polling seems to confirm.
This suggests Clinton is likely to win by at least 10 in Indiana. In fact, I would not rule out a Clinton win of 15 or more.
As I also wrote last Friday, late deciders in southern states have broken for Obama and not Clinton. The intuition here is that in places like Indiana the undecided tend to be white and lower class (and break for Clinton) while in the South the undecided tend to be African American (and break for Obama).
Interestingly, the volatility in North Carolina polling is different from the pattern discussed. Obama ranges from a high of 53 to a low of 48, while Clinton ranges from 40 to 45. North Carolina therefore appears to present a different case from the pattern of Clinton outperforming the polling seen in Ohio, Indiana and Pennsylvania. SUSA has Obama up 50-45 in North Carolina, almost the same finding of the previous SUSA poll that had Obama up 49-44.
If the pattern of Obama outperforming polling in Southern States holds in North Carolina, Obama could easily win by 10 or more. However, as one commenter pointed out in response to my post on Friday, the tone of the race in fundamentally different now than it was in February. In the end I think North Carolina is fundamentally less predictable than Indiana.
Update - after writing this Zogby issued his final poll showing OIbama up 51-37.