Alec MacGillis tweets this from Joe Scarborough:
Not surprising, I suppose, but still a sign that more quantitative rigor would do horserace journalism some good. It’s not that quantitative data, namely polls in this instance, are perfect. It’s not that a qualitative sense gleaned from the campaign trail can’t be right. It’s just that, as a general rule, what Mark Halperin’s gut tells him after attending some Romney rallies is likely to be a less reliable than the polls for ascertaining which candidate has momentum or forecasting which candidate will win.
Taking that to heart may drain some kinds of the drama from horserace reporting, but I don’t see it as hostile to the enterprise. Halperin’s orientation shouldn’t have been, “Wow, these Romney rallies!” but, “Hey, Romney campaign, do you think your rallies are better indicators than the polls?” Journalists ask skeptical questions of campaigns all the time. Knowing the data helps them do that better.
[Update: MacGillis’ report from the NRI summit is here.]