Priming political trust: Evidence from an experiment

Abstract

Political trust is one of the most researched areas in political science. Yet little is known about what causes political trust to vary. Past research has relied almost solely on survey data, and focused on exploring the correlation between political trust and various micro- and macro-level factors. This research note reports the findings from an experiment designed to examine the causal effect of one of the most commonly cited causes of political trust: political probity. Results show that political trust and trust in specific institutions change when participants read an article about political probity and complete a word-association task. The treatment we used is low cost, straightforward and may be used by other researchers to alter political trust in experiments.

Keywords:

• experiment
• political trust
• probity
• trust in government

Notes

1 One exception is Mutz and Reeves (2005), who used experiments to test the effect of televised incivility on political trust.

2 If subjects are randomly assigned to treatment and control conditions, then a simple difference in means provides an unbiased estimate of the causal effect of the treatment. For an overview on experimental research in political science, see Druckman et al. (2006). An excellent guide to the design and analysis of experiments in political science is Gerber and Green (2012). For a critical view on experiments (especially field experiments) that also shows precisely what assumptions are necessary for causal inference, see Deaton (2009).

3 Approximately one week later, students were shown preliminary research findings from the experiment as part of a classroom exercise that corresponded to lecture material on experimental methods. Debriefing can remove any lingering psychological harm done as a result of deception (Holmes 1976a; 1976b; Holmes and Bennett 1974; Smith and Richardson 1983). We invited students to individually raise concerns with us following each experimental debriefing but none did. We believe the psychological harm done as a result of deception in this experiment was zero.

4 A letter written by a putatively non-aligned political elite, in this case a former public servant, was chosen over an article written by a journalist or partisan political elite because we expect less reliance on partisan cues for information processing when a non-partisan delivers a message. The perceived veracity of the message, conditional on the form in which it is delivered (print media, television, editorial, etc.) or the characteristics of the individual delivering the message is an open question. Many Australians have little confidence in journalists (Simons et al. 2013), however, which implies that messages delivered by journalists may not be perceived as credible.