WAPOR 74th Annual Converence presentation on the Impact of Re-wording Questions about Democracy and Its Alternatives.
Abstract
How well do people understand survey questions about democracy and its alternatives? For many years, both the World Values Survey (WVS) and the European Values Survey (EVS) have asked questions about people’s preferences for governing their respective countries. The validity of these questions matters a great deal, as changing preferences over time have begun to set off alarm bells among political scientists. Since the election of Donald Trump in 2016, many scholars have expressed concern about the erosion of support for democracy in the US and Europe. As scholars debate the significance of these trends in the WVS and EVS data, there are good reasons to explore whether the questions themselves mean what we think they mean – and whether and how the responses can be influenced by small changes in question wording. This paper reports the results of an experiment conducted in the US in which the five key questions driving this debate are asked in a slightly altered form using a national sample of 2,500 adults. The results indicate that when we ask about support for “democracy” rather than “a democratic political system,” support increases significantly. In addition, multivariate regression results show that being a member of the out-party (Republicans, in the case of the US) is a significant (and negative) predictor of support for “a democratic political system.” However, the same is not true in the model of support for “democracy.” The findings may be cause for validity concerns in countries where the word used for “democratic” in the WVS survey question is also part of the name of one or more of a country’s major political parties, particularly in countries experiencing high levels of polarization.