We test our hypotheses using two-way ANOVA testing and linear probability model (LPM) regression.
Data are pooled across Scenarios. Standard errors are clustered at the individual respondent level.

H1 is tested using two-way ANOVA (T1 main effect) and LPM (with binary T1 indicator).

H2 is tested using two-way ANOVA (T1 and T2 interaction effects) and LPM (with binary T1, T2, and interaction indicators).

H3 is tested using two-way ANOVA (T1 and “political ideology” interaction effects) and LPM (with binary T1, “political ideology,” and interaction indicators). 

Political ideology is captured by the following item (responses on an 10-point scale, anchored at 1 = “Left wing” and 10 = “Right wing”): “There is often talk about an economic left-right scale in politics. “Left wing” denotes that government must ensure that everyone is taken care of. “Right wing” denotes that the individual must have more responsibility for oneself. Where would you place yourself on this scale, where 1 is the most left-wing and 10 is the most right-wing?” We the purpose of our analysis, we use a binary measure (1-5 and 6-10).

H4 is tested using LPM (with binary T1, T2, “experience as service recipient” and interaction indicators).  

Experience as service recipient is captured by a binary indicator (“No” and “Yes”) based responses to the following item: “Have you as a citizen been in contact with the public sector within one of the following policy areas?” For each policy area, “Yes” represents the citizen having experience with the given policy area used in the corresponding Scenario. 

H5 is tested using LPM (with binary T1, T2, “experience as public employee,” and interaction indicators). 

Experience as public employee is captured by a binary indicator (“No” and “Yes”) based responses to the following item: “Have you been employed in the public sector during your working life?” 

Additional heterogeneity analyses are conducted using LPM (with appropriate binary indicators).