Experimental evidence about deficit reduction strategies: bias in measuring tax and spending preferences

Abstract

Fiscal conditions and budget constraints in the United States have placed solutions to budget deficit problems at the center of the public policy debate. Preferences for deficit reduction strategies are likely to be heavily associated with particular ideologies and other demographic and economic variables. Therefore, since this study is a true randomized experiment, it provides strong evidence about the influence of question wording on deficit reduction preferences, and therefore the likelihood it is susceptible to manipulation. We find clear evidence that using the word ‘tax’ significantly and substantially influences respondents’ choices. This result is robust over two experimental trials about a year apart and whether or not control variables are included.