As political junkies everywhere await their next fix in tomorrow’s Democratic gubernatorial primary in Virginia, there are a number of variables that will bear watching when the returns are in, aside from the possible “Martin Effect” that Tom identified today, and the mutual-destruction/Iowa 2004 hypothesis offered last week by Nate.
The most obvious question involves an anomaly: in what will almost certainly be a low-turnout primary, the candidate universally considered to have a weak “ground game,” Creigh Deeds, is also the candidate surging in the polls in the last week, when pollster screening for likelihood to vote generally becomes more accurate. (A new SUSA poll of likely voters came out today showing Deeds moving from 29% to 42% in the last five days, with McAuliffe falling from 35% to 30%, and Moran from 26% to 21%). That makes four late polls in a row, all claiming to screen for likelihood to vote, placing Deeds in the lead, two by double digits. That’s a bit unexpected for a candidate whom many observers left for dead a month ago when he was running a poor third in most polls and was laying off part of his small field staff.
If Deeds hangs on to win, then we’ll have another, and somewhat heretical by recent standards, data point in the ancient debate between mobilization and persuasion strategies, and air versus ground spending.
A second and possibly related issue is the impact, if any, of Deeds’ endorsement by the Washington Post, which the Deeds campaign has tried to amplify with direct mail in Northern Virginia.
It is true that the Post endorsement immediately preceded Deeds’ rise in the polls, and it’s also true that the Deeds Surge seems to have been especially powerful in the Washington media market: the final PPP poll suggests that the rural senator’s support in NoVa more than tripled in the last two-and-a-half weeks.
But it’s also true that it’s engraved on stone tablets in most political science departments that newspaper endorsements don’t amount to a hill of beans in contemporary politics–particularly now that nobody outside the chattering classes reads op-ed pages, and newspapers appear to be in the last phase of a decades-long decline in media market share.
Thinking back to Nate’s Iowa 2004 analogy, it’s worth remembering that the startling end-game in that contest involved not only a surge for the ultimate winner, John Kerry, but another surge by near-winner John Edwards, from even further back in the pack, that seemed to be fed by an endorsement of the North Carolinian by the Des Moines Register. With all due respect for the Register‘s atavistic dominance of Iowa media, its endorsement of Hillary Clinton in 2008 didn’t, to use a familiar phrase, seem to amount to a hill of beans.
But just to toss a theory out, perhaps newspaper endorsements do matter, or at least reinforce trends, when late-deciding voters are casting about for another candidate in reaction to unsavory front-runner rasslin’: Dean and Gephardt in Iowa 2004, and McAuliffe and Moran in Virginia 2009. If that’s the case, then obviously Creigh Deeds could be benefitting from the dynamics that helped both Kerry and Edwards in Iowa in 2004, punctuated by a major media endorsement.
A third and smaller factor should draw attention if the results tomorrow night are very close: Virginia has a relatively restrictive early voting system. Voters can go to their precincts before election day and apply for and then cast absentee ballots, but only if they certify in the presence of a witness (if only a poll worker) that they qualify under specific grounds for absentee voting. So while the level of early voting in VA has risen rapidly in recent years, it hasn’t matched the massive percentages of many other states.
This is relevant because Gov. Tim Kaine’s bill to allow “no excuse” absentee voting in Virginia was killed by the legislature in February. And in a contest where the two campaigns with the “ground game” resources for an intensive early voting effort are also the campaigns that would have benefitted from “banking” votes before a third candidate’s late surge, it could be a what-might-have-been factor if tomorrow night’s results are very close.
Now we’ll just have to wait for voters to vote, and see if any of these factors ultimately matter to anyone other than serious geeks.