218 Chapter 6Chapter 6 || Public OpinionPublic Opinionestimates made at the time the program was enacted in 2010. Forty-two percent said
spending was higher, 40 percent did not know—and only 5 percent of respondents gave
the correct answer, that spending was less than expected.^38
How can these findings be squared with our earlier statement that Americans
generally hold opinions that have some basis in reality? For one thing, many
respondents probably had not thought about these questions in detail—and it’s likely
that only very few knew anything about health care costs in the first place. When asked
for an opinion as part of a survey, respondents had no time to do research or think
things through. One possibility is that people used partisan considerations: Obama
supporters (Democrats) said they didn’t know, while Obama opponents (Republicans)
said costs were higher than expected. If so, the responses say more about partisanship
and attitudes toward President Obama than about the cost of Obamacare.
Misinformation may also result from politicians making polarizing or extreme
statements about an issue. Over the last seven years, Republicans have tried to repeal
Obamacare nearly 50 times and Republican pundits have described the program in
highly negative terms. Donald Trump made repeal of Obamacare a central plank in
his campaign platform, and attempts to repeal the program occupied Congress for
much of 2016. It’s no wonder, then, that Republican survey respondents select negative
responses when asked about Obamacare—after all, that’s all they’ve been hearing from
the political figures they listen to and trust.
Finally, many supposed facts are actually “contested truths,” meaning that even if
people move beyond considerations and take into account multiple, nuanced sources
of information, they may nevertheless arrive at different conclusions about complex
questions.^39 For example, it is a fact that the unemployment rate fell by over a third
during President Obama’s years in office and has stayed roughly stable during the
first two years of Donald Trump’s presidency.^40 Do these changes mean that Obama
was a better steward of the economy than Trump? Obama took office during the most
severe economic downturn since the Great Depression, while Trump has inherited a
stable economy. People can easily arrive at different evaluations of Trump and Obama
depending on how they account for these starting points.
Such problems do not arise in all areas of public opinion. Studies show that
respondents’ ability to express specific opinions, as well as the accuracy of their
opinions, rises if the survey questions have something to do with their everyday life.^41
Thus, average Americans would be more likely to have an accurate sense of the state
of the economy or their personal economic condition than of the military situation
in Syria. Everyday life gives us information about the economy; we learn about Syria
only if we take time to gather information. These effects are magnified insofar as the
respondent considers the economy the more salient issue of the two.How Useful Are Surveys?
By now, CNN’s disclaimer about the limits of mass surveys should not be much of a
surprise. Survey results are most likely to be accurate when they are based on a simple,
easily understood question about a topic familiar to the people being surveyed—such
as the choice between two candidates, measured close to an election—and when the
survey designers have worked to account for all the problems discussed here. Under
these conditions, with samples of 1,000 voters or more, poll results are generally within
3 or 4 percentage points of the true population values.^42 You can be even more confident
if multiple surveys addressing the same topic in different ways and at different times
produce similar findings. However, if a single survey asks about a complex, unfamiliar
topic—replacing the income tax with a national sales tax or determining whom to blame
for a policy failure—then the results may not provide much insight into public opinion.As always, keep in mind that the
poll does not, and cannot, predict
the outcome of the election in
November.—Disclaimer on CNN preelection
pollFull_07_APT_64431_ch06_198-231.indd 218 16/11/18 1:38 PM