Transcript What was wrong with the polling

Political Socialization
• Political socialization is a lifelong process
through which an individual acquires opinions
through contact with many different agents
and group identifications.
• What are the factors that shape political
attitudes?
Political Socialization factors:
•
•
•
•
•
•
•
•
Family
Gender
Religion
Education
Socio-economic class
Race and ethnicity
Geographic region
Media
Public Opinion
• What is public opinion?
• How do politicians gauge public opinion?
– Direct experience
• Letters, phone calls, e-mail, direct contact with
constituents in meetings and other public appearances
– Public opinion surveys or polls
Public Opinion Polling
• Scientific polling developed in the 1930s
• Challenges of polling
– Sample size
– Sample bias
– Questions
Public Opinion Polling
• Sample
– The portion of the population
surveyed on an issue
• Sample size
– 1,000 to 2,000 can represent 300
million Americans
– Sampling error
• “Margin of error”
• The larger the sample, the smaller the
sampling error
• Typically + or – 3 percent
George Gallup:
“Okay, next time
you go for a
blood test, tell
them to take it
all.”
Public Opinion Polling
• Random sample
– Every individual in the population group being
polled has an equal chance of being selected
– Samples can be of nation as a whole or more
narrow groups, such as “likely voters,” members of
one party or age group, residents of one state or
congressional district, etc.
– Achievable via random digit dialing today because
95% of households have telephones
Public Opinion Polling
• Sample bias
– The effect of having a sample that does not
represent all segments of the population being
polled
What happened?
• Just before the election of 1948, Gallup gave
the election to Dewey by 49.5 to 44.5% and
Roper gave Dewey 52.2 to 37.1% for Truman.
But when the election was over, Truman had
beat Dewey for re-election by over 2 million
votes—28 states and 303 electoral votes to
Dewey’s 16 states and 189 electoral votes.
• What was wrong with the polling?
What was wrong with the polling
in the 1948 election?
• The pollsters did their last poll app. 2 weeks
before the election—enough for some voters
to change their minds
• The “undecideds” didn’t vote as expected.
Often, “undecideds” go with the incumbent
(in this case, Truman)
• Sampling errors—survey taken over the
phone—household telephones, at the time,
were limited to the upper class
Public Opinion Polling
• Sample bias
– How might the following result in sample bias?
• Calling survey participants on the phone only from
Noon until 3PM.
• Asking survey participants to call in their vote.
• Polling only a retirement home to gauge the nation’s
views on health care.
• Asking bald men what kind of shampoo they prefer.
• Polling only residents listed in the City of Atlanta
telephone book.
• Going door-to-door in an affluent suburb in the middle
of the day.
Public Opinion Polling
• Questions
– Respondents should be asked about things they
know and have thought about
• “Public Affairs Act of 1975” experiment
• Study: Only 10 percent of people will say “I don’t
know.”
Types of Public Opinion Polls
• National polls
– Effort to measure national opinion within a period
of time
– Major examples: ABC News/Washington Post, CBS
News/New York Times, NBC News/Wall Street
Journal, CNN/USA Today/Gallup, Los Angeles
Times poll
Types of Public Opinion Polls
• Campaign polls
– Benchmark poll
• Conducted by candidates for public office
• Taken within state or district to gather baseline
information about attitudes toward candidate
• Instrumental in designing campaign strategy
– Tracking poll
• Conducted by campaigns and media
• Follow changes in attitudes toward candidates
• Help revise campaign strategy if necessary
Types of Public Opinion Polls
–Exit polls
• Surveys of voters at polling stations on election day
• Often used by news agencies to “call” an election for
one candidate or another
Types of Public Opinion Polls
–Exit polls
• Controversies
– 1980 – NBC announced Reagan the winner before polls in the
West had closed.
– 2000 -- Florida called for Al Gore by several networks in 2000
election, then switched to George W. Bush, then announced
the vote was too close to call.
– 2004 – Even though networks hid results of exit polls until
after close of all polls, results of poll leaked, Kerry seemed
ahead.
Types of Public Opinion Polls
–Exit polls
• Controversies
– Defenders of exit polling: Not intended to predict winner, just
understand why people voted the way they did
Types of Public Opinion Polls
• Pseudo-polls
– Self-selection polls
• E.g. viewer or listener call-in polls, Internet polls
• What are the limitations of this kind of polling?
Types of Public Opinion Polls
• Pseudo-polls
– Push polls
• Example from Florida:
– “Please tell me if you would be more or less likely to vote for
Lt. Governor Buddy MacKay if you knew that Lt. Gov Buddy
MacKay plans to implement a new early-release program for
violent offenders who have served a mere 60 percent of their
sentences if he is elected governor.”
– A false claim
• What is the goal of this kind of poll?
Public Opinion
• “I really don't worry about polls or focus
groups. I do what I think is right.” – George W.
Bush
• “Unless mass views have some place in the
shaping of policy, all talk about democracy is
nonsense” – V.O. Key
Public Opinion
• Critics of public opinion polling
– Bandwagon effect
– Politicians follow rather than lead
– Easily manipulated to achieve desired results
• Supporters of polling
– Some studies show that politicians use polls to help win
support for their policies, not to follow
– Politicians should pay attention to polls; give more direct
power to the people.
– How else can politicians gauge public opinion and then
represent the public?
Which side do you agree with more?
Public Opinion Polling
• Examine case studies on polling