"2 Polls Say Race Is Tied Again" -- LA Times, 9/17/04
Ahh, polls. Which to believe? How can they be so different?
The best article I've come across on this topic is by Ruy Teixeira. He points out two major problems:
The two problems are: (1) samples that have an unrealistic number of Republican identifiers and hence tend to favor Bush; and (2) the widespread and highly questionable practice of using likely voters (LVs) instead of registered voters (RVs) to measure voter sentiment this far before the election.
For the first problem, he suggests:
weighting poll results by a reasonable distribution of party ID may be necessary to avoid giving the public distorted impressions of the state of the race.
Adjusting for this, it offers,
a clear picture of a tight race, with Bush likely running a small lead, but not the solid--and even large--advantage that has been conveyed to the public.
As for the second problem, most of these polls are based on so-called "Likely Voters" (LVs), not registered voters (RVs). LVs are defined by each polling group, pending on their screening methodology based on a set of questions. It's clear the composition of this group can fluctuate and vary from one polling group to the next. In addition, daily news events can greatly affect the composition of this group, offering the potential for misleading results. As quoted in the article,
[E]stimates of who may be likely voters in the weeks and months prior to Election Day in large part reflect transient political interest on the day of the poll, which might have little bearing on voter interests on the day of the election.
Whereas LVs are much more volatile, changing day to day and week to week, registered voters (RVs) are not; one could make the case to use RVs instead. In fact, when focusing strictly on those polls that use RVs:
they suggest something far different: the race is damn close and Bush's substantial lead is a myth.
Is it too much to ask for mainstream media to enunciate these very important points?