Procedural Legitimacy and Immigration Policy
A conjoint experiment on procedural legitimacy in immigration governance, fielded in the United States, Germany, South Korea, and Singapore
Description
This registration covers a conjoint experiment on procedural legitimacy in immigration governance, fielded as part of a four-country survey study on immigration attitudes. Respondents evaluate a series of immigration policy vignettes, one at a time, and decide whether to support each, with each vignette varying decision process, implementation, policy area, direction, and immigrant composition. The registration specifies five confirmatory hypotheses: decision-process legitimacy (H1a), implementation legitimacy as a pre-registered rival (H1b), governance dominance over group-composition threat (H2), and generality of decision-process effects across policy area and direction (H3a, H3b). Approximately 2,500 quota-complete respondents are recruited per country, roughly 10,000 in total.
Study design
This is a within-subject conjoint experiment in which each respondent evaluates one immigration policy at a time. Respondents complete 10 randomized tasks, followed by an 11th task that exactly repeats the first to measure test-retest reliability. The experiment varies five features: decision process (4 levels, from a lawful process to executive overreach, external imposition, or a process that skips standard review steps), implementation (3 levels, from even-handed application to discretionary or unequal application), policy area (3 areas), policy direction (2), and immigrant composition (5 country-specific groups), producing 360 possible combinations.
Respondents are recruited through Qualtrics and international survey providers, with quotas designed to match each country's census population. This experiment is part of a larger survey containing another separately pre-registered experiment on civic fairness in immigrant evaluation, with order randomized across respondents.
Hypotheses
H1a — decision-process legitimacy. Executive overreach, external imposition, and procedural neglect will each reduce support compared with a lawful decision process.
H1b — implementation legitimacy. Unequal application will reduce support compared with even-handed application, with discretionary application falling in between.
H2 — governance matters more than group threat. Problems with either the decision process or implementation will reduce support more than concerns about group composition.
H3a and H3b — generality across area and direction. The effects of decision-process violations will remain similar across different policy areas and policy directions.
Analysis plan
H1a and H1b are tested separately in each country by estimating how each randomized policy feature affects support, accounting for repeated responses from the same person. H2 compares the effects of problems with the decision process and implementation against the effect of immigrant composition, using three pre-specified support penalties. H3a and H3b test whether decision-process effects remain similar across policy areas and directions, using a ±5-percentage-point equivalence bound. One pre-specified sensitivity analysis repeats the full analysis after excluding respondents flagged for possible automated survey completion.