There are a couple very basic ways to drive up an opponent's negatives: 1)attack him and 2) let him attack you. The above graph depicts the exit poll response for select Democratic primaries on the question "Do you believe (the candidate) attacked unfairly." I used the following states: SC, MO, IL, NY, CA, LA, VA, WI, OH, MS, PA, IN, NC, WV, OR, KY. I attempted to get a fairly even distribution of victories for both candidates, although clearly this is subject to potential error we should test for...
Obama victory states:

Clinton victory states:

There are several conclusions we should draw from all this:

1)    The public believed Clinton ran the more negative campaign. This holds true even in those states that Clinton won, even at the very end of the campaign with Clinton's landslide victories in KY and WV driven in part by voters who held a great deal of antipathy towards Obama.

2)    The long campaign raised negativity flags for both candidates, but somewhat more for Clinton than for Obama.

3)    The public started out from a position of believing Clinton had gone negative, In New Hampshire the CNN exit poll didn't have this question, but twice as many respondents believed Clinton had run the most negative campaign than believed that of Obama. South Carolina was a particularly nasty campaign with respect to public perception.

4)    If you support a particular candidate, you will believe he/she was attacked unfairly.

Did this perception help Obama win? Hard to say, given #3 and #4 above, but we do know that it was the campaign Obama wished to run. The strategy played into an existing media narrative that Clinton was a cynical, somewhat dirty campaigner and that Obama was, in some way, shape or for a transformative candidate. One interesting thing about this strategy is that it seems particularly tuned towards independents of the anti-partisan type - not necessarily towards a primary demographic. The Democratic party base and the most likely primary voters are accustomed to dirty politics. They may not like the system but they work within it, and it's worth noting that in every state I show above in which Clinton won, over half the exit poll respondents believed she had attacked her opponent unfairly except for California (living in CA as I do, I remember her local ads being, for the most part, focused on herself as a candidate) - and yet she still won.

Clinton chose to run an increasingly aggressive campaign "against" Obama as the primary went on. This strategy ran the risk of playing up her own negatives in the media, but may have also reflected a calculated risk that she had already lost those who might resent such tactics (or an equally calculated acknowledgment that, after February, she had no chance of victory without tearing Obama down).

However, all this explanation is not intended to rehash the primary. We need to appreciate the fact that Obama's campaign is playing this image in much the same way as we transition into the General Election. John McCain's extremely personal negative ad campaigning has not yet had a detectable impact on his favorability ratings although Pew has detected some decline in his aura over the primary campaigns, and unlike Clinton, he is acting against an existing media narrative - that McCain is a maverick anti-politician.  Obama's maturity and insistence upon remaining relatively above the fray runs the risk of the public and/or the media ignoring the message. With that risk there is a great deal of potential reward. We've all noticed the occasions of the media taking its fact-checking job somewhat more seriously in recent days and taking note of McCain's message being off-image. Negative campaigning is likely to alienate a more important subset of voters than participated in the primary, an independent demographic McCain MUST improve if he is going to have any chance of winning this election. If, like Clinton, he cannot tear Obama down without harming himself an equal amount, his campaign will not work.

Some authors have suggested that those asking Obama to respond aggressively should quiet down. I don't agree. Whether to respond passively or aggressively to a specific advertisement is an interesting debate that is worth having... but I ask, given the above, that you take a look at the reasons for which you believe he is passive. There's a narrative lingering from the primary that Obama is not enough of a "fighter." Edwards adopted the label early and didn't gain much traction. Clinton chose to adopt it wholeheartedly, and it played into yet another media narrative - one that was far more complimentary from her perspective. Conventional press was very happy to depict Clinton as a no-nonsense aggressor (and they have never been willing to reinforce the same positive image in Obama).Her supporters, consequently, played (and play) down Obama as "Obambi" - a politician incapable of playing the game when it matters most. This label has never made a great deal of sense. The select times that Obama responds tend to be as forceful as any veteran political riposte. In SC, where both sides believed the negativity got out of hand, one could argue the results were significantly more damaging towards Clinton. It seems some Democrats are still trapped in this primary narrative despite the lack of evidence. To those commentators I respond: Hillary Clinton didn't play this game better than Obama... she lost it to him, and John McCain seems determined to make the same mistakes.