John Zogby, the DNC's polling shill
For much of the year, pollster John Zogby has been telling anyone who will listen that the election is John Kerry's to lose. With virtually all polls by all pollsters, including Zogby, showing a very close race for the presidency all year, the confident message that Kerry would win would be surprising coming from any pollster. But John Zogby is not any pollster. For Zogby is viewed by many in the political culture as the pollster who gets it right.
This reputation is to some extent undeserved. In 1996, Zogby predicted Bill Clinton's margin of victory within 0.1% of the actual margin obtained. That sounds great. But in 1996 there was never much drama as to the outcome of the Presidential race. Several pollsters had Clinton winning by over 10%. In the end he won by just over 8%. So was Zogby's percentage call really a big deal?
In 2000, Zogby picked up a late Gore surge, as did some others, and predicted Gore would win the popular vote by 2%. Gore won the popular vote by 0.5%. So Zogby was no closer to the final outcome than a pollster who predicted Bush would win by 1%. Both would have been 1.5% off the final outcome.
In 2002, Zogby's reputation began to suffer some. He miscalled a collection of Senate races, and was way off on a few of them, such as in Colorado. His polls on the state level, in both 2000 in the Presidential race, and the Senate races in 2002, were very volatile. Zogby had Bush down 13% in Florida less than a month before the 2000 election. Big swings occur in just a few days in Zogby state polls. This can happen when you only interview 150—200 voters in a state per night.
Zogby has also come under attack this year for keeping many details of the internals of his polls private. He admits that he weights certain segments of his sample (ethnic groups, female or males, or for party identification). This suggests that he has locked in certain distributions of voters based on prior years' data. But this years pool of voters may be different than in prior years. It would be nice to know what his distribution is. Maybe it has too many from one party. With sharply polarized voting by party member this year, this piece is crucial.
On the national level, Zogby has shown both small Bush leads and small Kerry leads this year, similar to Scott Rasmussen's tracking. But on the state level, his recent tracking polls are useless, to be kind. Six days ago, he had Kerry ahead of Bush in Michigan by 10%. A few days later, Zogby had Bush ahead in the state by 2%. That kind of movement in this short time frame is not possible. In Colorado, Zogby again seems to be unduly optimistic about the Democrats' prospects, as he was in 2002. His is the only poll that showed Kerry ahead in the state this week. Kerry is not planning to return to Colorado the final weekend, and has reduced his ad budget there. This does not sound like a state in which Kerry is leading. On Saturday, Zogby had Kerry ahead in Wisconsin by 8%, and Bush up in New Mexico by 9%. All other polls in both states were much tighter. The two states were each practically a tie in 2000. Is it likely that they would diverge by 17% in 2004?
Pollsters are allowed to be wrong. It is, after all, an inexact science,' and all pollsters tell you that their forecast must be considered in light of a particular poll's margin of error associated with the poll's sample size, usually 3 to 4%.
But what has differentiated Zogby from other pollsters, is his hawking of Senator Kerry's prospects. On Thursday, with his own national survey showing Bush up 2%, and Bush ahead in five of his ten battleground state polls and tied in a sixth, Zogby was telling everybody that Kerry was surging and would win. On Friday, Zogby showed the race tied, while two other national surveys showed Bush up by 5%, and three other tracking polls showed the race tied, or Bush up 2% and 3%. None of the other pollsters were making any predictions yet. Why was John Zogby doing so?
The answer I think, comes from the fact that Zogby is not a non—partisan actor in all this. His brother, James Zogby is an outreach coordinator for Arab voters for the Democratic Party, and heads an Arab American advocacy group. John Zogby admits to being a Democrat, like his brother. And like his brother, he is also an advocate for the Arab side of the Middle East conflict. A few months back, he conducted a poll that suggested that Americans regarded AIPAC as a group with dual loyalties and untrustworthy . This followed a CBS News smear of the organization, accusing AIPAC of assisting a Defense Department official who supposedly was passing classified information to Israel. These highly publicized charges so far seem to be all allegation, with no indictments or confirmation. The AIPAC poll was conducted for the Council on the National Interest. Search around for former members of Congress or former diplomats who are now on the Saudi/Wahhabi payroll at some think tank with the name Middle East in it, and chances are they will be part of the CNI. This is a nasty anti—Israel group, pure and simple. Zogby has offered them his imprimatur.
Mickey Kaus wrote recently that if one or the other candidate appears to be ahead by a few points right before the election, some undecided voters might select the candidate in the lead to help him win by enough to avoid a repeat of the mess that occurred in Florida in 2000. Zogby has been arguing all year that undecided voters will break against the incumbent in the end, and so Kerry has a few poll points in his reserve tank that might give him the win. But former Carter pollster Patrick Caddell, says Zogby and all others who think that undecided voters will swing to the challenger in the end are probably wrong, and any serious examination of recent history will show this has not been the pattern in the past few elections.
In any case, Zogby seems out to deliver a message — that John Kerry is still very much alive, and will likely win. If he doesn't, or if he loses in a close race, then Zogby's message that Kerry was likely to win will help feed the notion that Kerry should have won, and didn't only because the election was stolen again by GOP partisans in key states. The first step to follow an election defeat for Kerry, if this happens, is therefore to delegitimize the victory of President Bush.
I don't think Zogby's statements are all that innocent, or just pollster science. I think he is part of the Democratic Party machinery, much as the New York Times, and CBS are. James Carville and Stanley Greenberg have a poll published by the Democracy Corps, and most everyone knows it is a poll with a small Democratic party bias, much as Strategic Vision is a Republican friendly poll. To their credit, both the Democracy Corps and Strategic Vision have been close with their numbers to national averages for most of the year, though predictably, Democracy Corps makes Kerry look better, and Strategic Vision makes Bush look a little better than the averages. John Zogby wants to have himself viewed as above the fray, a neutral party.
The Washington Post revealed, and Mickey Kaus highlighted an example of the way that polling results are malleable, in the hands of John Zogby:
...the Zogby poll published in the Rapid City Journal is flawed. It showed Republican Thune leading Daschle, 48.5 percent to 45.5 percent, just within the margin of error. At first, however, the poll had shown an even larger Thune lead, which seemed so improbable that the pollsters adjusted their voter turnout estimates and arrived at the narrower gap.
Sorry, to be the one to break it, but on this one, the emperor has no clothes. Zogby is spinning for the DNC at the moment, not just reporting what his polls may show, even if they are conducted fairly. What Zogby is not, is neutral.