128 PART TwO • THE POlITICS OF AMERICAn dEMOCRACy
House Effect
In public opinion polling,
an effect in which one
polling organization’s
results consistently differ
from those reported by
other poll takers.House Effects. One consequence of the use of secret in-house weighting schemes is
that the results reported by one polling firm may differ from those reported by another
systematically. Pollster A might consistently rate the chances of Republican candidates as
2 percentage points higher than does pollster B. A consistent difference in polling results
between firms is known as a house effect. (House here means the organization or firm,
as when referring to casinos.) House effects are measured by comparing a firm’s results
with the average results of all other poll takers. A house effect does not mean that a firm’s
results are in error. It could be noticing something important that its rivals have missed.How Accurate Are the Results? Despite all of the practical difficulties involved in poll
taking, the major polling organizations have usually enjoyed a good record in predicting
the outcome of presidential contests. Most major firms predicted the outcome of the 2012
presidential elections—and most of the U.S. Senate races—with considerable accuracy. Two
major poll takers, however, reported results that were embarrassingly inaccurate. Gallup
seriously overestimated voter turnout among Republicans and underestimated Democratic
turnout. Rasmussen overweighted Republican respondents. Both pollsters predicted that
Romney would win a substantially larger share of the votes than he actually did.Additional Problems with Polls
Public opinion polls are snapshots of the opinions and preferences of the people at a
specific moment in time and as expressed in response to a specific question. Given that
definition, it is fairly easy to understand situations in which the polls are wrong. For exam-
ple, opinion polls leading up to the 1980 presidential elections showed President Jimmy
Carter defeating challenger Ronald Reagan. Only a few analysts noted the large number
of “undecided” respondents a week before the election. Those voters shifted massively
to Reagan at the last minute, and Reagan won the election. The famous photo of Harry
Truman showing the front page that declared his defeat in the 1948 presidential elections
is another tribute to the weakness of polling. Again, the poll that predicted his defeat was
taken more than a week before Election Day.Poll Questions. It makes sense to expect that the results of a poll will reflect the ques-
tions that are asked. Depending on what question is asked, voters could be said either to
support a particular proposal or to oppose it. One of the problems with many polls is the
yes/no answer format. For example, suppose that a poll question asks, “Do you favor
arming the rebels in Syria?” A respondent who has a complicated view of these events, as
many people do, has no way of indicating this view because “yes” and “no” are the only
possible answers.
How a question is phrased can change the polling outcome dramatically. The Roper
polling organization once asked a double-negative question that many people found hard to
understand: “Does it seem possible or does it seem impossible to you that the Nazi extermina-
tion of the Jews never happened?” The survey results showed that 20 percent of Americans
seemed to doubt that the Holocaust ever occurred. When the Roper organization rephrased
the question more clearly, the percentage of doubters dropped to less than 1 percent.
Respondents’ answers are also influenced by the order in which questions are asked
and, in some cases, by their interactions with the interviewer. To a certain extent, people
try to please the interviewer.unscientific and Fraudulent Polls. A perennial issue is the promotion of surveys that
are unscientific or even fraudulent. All too often, a magazine or Web site asks its read-
ers to respond to a question—and then publishes the answers as if they were based onwww
Helpful Web Sites
In recent years, Nate Silver
has won a reputation as
one of the Web’s sharpest
students of polling data.
Follow his analyses and
those of his colleagues by
searching on “538”—the
name of his blog.
Copyright 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.