Saying versus doing: a new donation method for measuring ideal points

Majoritarian Bargaining over Budgetary Divisions and Policy

“…More substantively, we …nd that subjects are willing to make budgetary concessions to implement preferred ideological choices and that our individual preference elicitation method largely correlates with strategic bargaining behavior. We also present suggestive evidence in Appendix Section 4.9 that our behavioral method, while correlated with stated partisan identi…cation and political ideology, may, as has been found in related work, more clearly distinguish "strong"partisans and politically active individuals from weaker partisans and more politically disengaged respondents where non-behavioral elicitation methods might not (Haas and Morton, 2018). These results thus o¤er some support for the idea that our preference elicitation is doing what we hoped by distinguishing between people who truly hold strong preferences versus those who only say they do.…”

“…In the treatment where the WTA could not be conditioned on what the other recipient was receiving, they …nd that the mean WTA is greater than the mean o¤er, leading to a 45% rejection rate. 3 Eliciting political preferences behaviorally is posited to overcome three potential shortcomings in stated preference elicitation (Haas and Morton, 2018). First, the low-stakes nature of survey questionnaires make them vulnerable to respondent satis…cing, which may account for the high predictive value of stated attitudes about which respondents feel strongly and the comparatively low predictive value of stated attitudes on issues about which respondents care less (Farc and Sagarin, 2009;Holbrook et al, 2016;Krosnick, 1988;Miller et al, 2016; Norris, Krosnick and Visser, 2016; Visser, Krosnick and 2 A few studies have looked at ultimatum games with more than two players.…”

“…Our experiment uses a modi…ed version of the behavioral donation method validated by Haas and Morton (2018). Their study compared ideal point estimates for Democrats and Republicans under two methods: a behavioral one where subjects were asked to divide a sum of money between two interest groups on a number of issues and a non-behavioral one where subjects were asked to answer questions on a Likert scale on the same set of issues.…”

“…In comparison, we extend the elicitation methodology by Haas and Morton (2018) in four key ways. First, we capture individuals' preferences over a greater space: subjects behaviorally express a preference on the ideological spectrum (left, middle, or right).…”

“…Prior to the bargaining game, subjects complete a series of incentivized tasks that allow us to identify their preferred policy (e.g., interest group) and the attitudinal strength of their preferences on a series of topics (see Haas and Morton (2018)). Our design seeks to naturally reproduce a setting that allows for a rich pattern of policy and budget compromises to occur as may happen outside the laboratory.…”