Featured Article


Election 2008: Run by the polls
Share this Content

Lydia Saad, our Gallup polling expert - who lives in Connecticut - was back today at a fascinating time for Presidential polling.  Gallup's latest polls reflect a change in their data, with a gap appearing between "registered" and "likely" voters.  It's a bit better news for John McCain, the closer we get to "likely" voters.  

All in all, despite their huge differences in methodology, I'm surprised that Gallup, Zogby and others are all running around the same numbers nationwide.  Today, Gallup has Obama up by 10.  We spent a lot of time talking about methodology, and I'm wondering how useful that information is to listeners.  Do you care, in detail, how polls are carried out?  

One caller suggested a much bigger problem with cell phones, call screening, caller ID and "no call" lists than Saad agreed to.  His question: Does a specific type of voter refuse to take part in polls?  If that person's not selecting him or herself "in" - then we miss gathering their data.  

Also, for those who missed our conversation about The Bradley Effect on race and politics, The New York Times ran a fascinating backgrounder on it yesterday, and Gallup ran a poll last week.  

We also heard from two callers talking about third party candidates, and how they don't tend to show up on national polls.  Gallup ran numbers on this, and it's really different - depending on whether you actually give people names to choose from, or if you just let them tell you which candidate they choose.