Kos notes that the Rasmussen tracking poll shows a six-point net swing from Bush to Kerry over the past two days: from 48-45 Bush to 47-44 Kerry. But he cautions that tracking polls tend to be volatile.
Well, maybe. But the Rasmussen numbers have been anything but volatile. Ever since the David Kay revelations in January, the Bush-Kerry matchup has been in a very narrow range, neither candidate getting above 49 or below 42. It’s been more than a month since either candidate’s number moved as much as three points a day.
In fact, tracking polls using a three-day moving average have inertia built into them, not volatility, because only a third of the sample changes. A three-point drop from one day’s three-day moving average to the next means that the latest day must be significantly lower than the reported average, and a two-point jump means the reverse. So the Tuesday night numbers must have been something like 49-42 for Kerry, and nothing that happened today is going to do Bush any good at all. So I’d expect Kerry’s lead as measured by Rasmussen to widen for the next couple of days unless something happens to reverse the trend.
Mar 24 44 47
Mar 23 47 45
Mar 22 48 45
Mar 21 47 45
Of course, that says nothing about how enduring the damage to Bush will be. (Tradesports still has him about 60% likely to win, near his lows but not substantially down since Sunday evening. The Iowa market, which had Kerry and Bush neck-and-neck as recently as March 1, now show Bush with a 5.5 percentage point lead in vote share, and that number hasn’t changed this week.)
Still, it’s a hopeful sign for those hoping for a change of leadership come January.