Who Were New Hampshire's Likely Democratic Primary Voters?
One of the pre-election pollsters' most daunting tasks is to divine a model for the "likely voter." Since most respondents in a pre-election telephone survey will report that they intend to vote, pollsters use a series of questions that help them estimate which members of their telephone samples are actually likely to show up at the polls, and which respondents are likely to demur. Like grandma's favorite recipes, survey organizations typically keep their "likely voter model" a closely held secret.
In the pre-election polls for the recent New Hampshire primary, the likely voter models for the Republican primary performed well, though the likely voter model for the Democratic primary performed poorly. The pre-election estimates for the Republican candidates were derived from models that were (presumably) based on the 2000 Republican presidential primary. For both Republican primaries, the voter turnout was remarkably similar (238,606 in 2000 and 238,909 in 2008), so the pre-election polls' likely voter model worked well, and their predictions were reasonably accurate.
The voter turnout for the Democratic primary, on the
other hand, was unprecedented in at least two ways: the number of participants
and the proportion of registered Democrats who showed up to vote. In 2004, a
(then) record of 219,787 voters turned out to vote--the previous record for the
Democratic primary was in 1992, when 167, 819 voters participated. This year, a
record shattering 287,849 voters participated in the New Hampshire Democratic
primary--including nearly two thirds (66.3%) of the state's registered Democrats
(up from 43.3% in 2004). Simply stated, the 2008