Zogby polls
I notice CAI has chosen Zogby as the pollster they hire to generate poll results showing how satisfied people are with their community association. I also noticed that Zogby has been wildly inaccurate at key moments of the 2008 election cycle. After Barack Obama won the Iowa caucuses, I seem to recall that Zogby told the world that Obama had a 13 point lead in New Hampshire just a couple of days before the election. Clinton won by 3 points. Two days ago I heard on the radio that Zogby had Obama ahead by double digits in California. Clinton won easily.

If a pollster can't accurately measure public opinion a day or two before an election, on a simple "who are you going to vote for" question, how can he possibly be counted on to make a definitive finding on anything else?

Also, the issue is how "satisfied" people are--whatever that means--with an organization that most of them know little or nothing about. And those vague sentiments are wrapped up and inseparable from the way they feel about their neighborhood, their home values, and a dozen other things.

But even aside from all that, the inability of the Zogby firm to accurately measure public opinion in this election cycle is shocking. I could poll my students and ask them to predict the state by state outcomes and maybe produce a better record.

Note, though, that Zogby's (ultimately inaccurate) predictions get huge play in the press before election day, thus influencing the mindset and probably the votes of those who hear about them. For example, in the open New Hampshire primary, there were probably independent voters who decided to vote for McCain rather than Obama because Zogby was saying Obama was far ahead but McCain was in a close race. McCain and Obama were the two top candidates for independents.