Ryan Hill, Brigham Young University
Social and Behavioral Sciences
We analyze the presence of two types of framing bias in ballot wordings that affect election outcomes: status quo bias and agent-external loss aversion. Both of these biases are studied extensively in behavioral economics and likely have salient impacts on initiative or proposition measures in state elections. Status quo bias is the tendency of a voter to be more likely to maintain a currently enacted policy rather than passing a new law. Agent-external loss aversion is the tendency for a voter to be more likely to grant a right to a minority population rather than vote to restrict a right. Both of these biases have had important impacts on the outcome of recent ballot initiatives especially on California Proposition 8 that failed to legalize the right of marriage for same-sex couples. We test the salience of these framing biases by conducting a nationwide survey experiment that randomly assigns different ballot wordings for the same policies across different survey respondents. The survey mimics a real voting situation and the randomization of questions ensures that we can identify the specific effect of each bias on the voting outcome. We analyze the treatment effects using difference of means and ordinary least squares analysis. We also extend our analysis to examine whether moderating variables such as political knowledge or opinion have an impact on the magnitude or direction of the bias effects.