Support For Democracy Depends How People Are Asked

According to research from Georgia State University, Americans might not be as happy with U.S. democracy as we once thought, and changes in how surveys are conducted could be part of the reason.

The researchers examined data from the American National Election Studies in 2012 and 2016 to see how satisfied U.S. citizens are with American democracy and how the way they were surveyed might affect their views.

“The basic idea was that when people are by themselves, answering questions over the internet, they could be a little more truthful,” the researchers explain. “There’s some idea that liking democracy is good, so you don’t tell someone who’s knocking on your door that you aren’t satisfied with it.”

This study builds upon a theory that was initially proposed in the 1940s and has since amassed a substantial body of supporting literature. The research uncovered a significant 20% decrease in reported satisfaction with democracy when comparing data from the 2012 and 2016 surveys.

The article notes that in 2012, 65% of respondents were interviewed online, and this number rose to 71% in 2016. Prior to 2012, the interviews were predominantly conducted in person, with only a small percentage taking place over the phone.

“The optimistic take is that some of the documented decline in positive attitudes is actually a function of switching over to interviewing by the internet,” the authors explain. “I think through 2010 or so, we’re probably overestimating satisfaction levels.”

The researchers believe that the more person-to-person nature of many surveys pre-2010 might have prompted respondents to say they were happier with democracy than they really were. The decline has been precipitous since then, however, with the 2020 results even worse than those in 2016.

“Survey mode doesn’t account for all of the decline though, so there is a genuine decline in warmth for democracy, but this is a specific incident in this long-standing theory about survey method,” the authors conclude.

Facebooktwitterredditpinterestlinkedinmail