Bellwether Points to Clinton Win in Penn.
Bellwether reflects poll prediction
In the random-digit-dial selection of likely Democratic households, Clinton led Barack Obama 52 percent-to-40 percent, with 6 percent undecided and 2 percent refusing to respond. This closely mirrors the 52 percent-to-42 percent statewide Democratic poll released by the Suffolk University Political Research Center on the morning of April 21.
Although bellwethers are not designed to depict actual margins, they have been remarkably accurate in predicting outcomes.
The Center’s analysis was made using a new election predictor module first employed successfully with the New Hampshire Democratic Primary; the GOP Florida Primary; the Democratic Primaries in California, Massachusetts and Tennessee; and the Ohio Election Day results. It is based on a number of statistical indices, including statewide polling and bellwether geographic areas.
“A cautionary word or two: Past bellwether performance is a guide but not a 100 percent guarantee of future performance,” said David Paleologos, director of the Suffolk University Political Research Center. “New bellwethers often are created every election cycle as people migrate and as development and geography-driven issues emerge. In addition, local endorsements from popular people can skew margins.
Historical precedent
The last two times a non-incumbent president was on a presidential preference ballot in Allegheny County, the bellwether area mirrored the exact order of finish statewide and was within 5 percent for the top three vote-getters in each major party.
On Jan. 29, 2008, the Political Research Center’s Florida bellwether, Hillsborough County, almost exactly matched the statewide Republican Primary results. On Jan. 8, 2008, the New Hampshire Presidential Primary bellwether towns – Kingston and Sandown – predicted a Clinton win in the Democratic Primary, while no other leading indicator in the country had Clinton winning. The bellwether analysis also was an added tool used successfully by the Center for Political Research in both the 2006 Massachusetts Gubernatorial Democratic Primary and General Election.
Methodology
Suffolk University interviewed 402 likely Democratic voters from a randomly selected list of 5,000 residents of Allegheny County. All phoning took place on Sunday, April 20, and Monday, April 21. There was an equal probability of contacting and interviewing newly registered voters, recent party-switch registrants and longtime Democratic households, provided those contacted were registered Democrats and intended to vote in the Democratic Primary on Tuesday, April 22. The execution of the ID interviews was by live telephone call. For more information, please contact David Paleologos at 781-290-9310.