Multidimensional public feasibility in policy packages: acceptability, effectiveness beliefs, and co‑production across climate, artificial intelligence, and aging

When citizens evaluate proposed policy packages, do their judgments of acceptability, perceived effectiveness, and willingness to contribute to implementation move together, or do they capture distinct dimensions of public feasibility? Using a pre-registered conjoint survey experiment in the United States (N = 1,506), we examine how citizens evaluate policy packages combining taxes, regulations, subsidies, and information measures across 3 high-difficulty domains: electric-vehicle transition, AI regulation, and population aging. We distinguish push instruments, which impose costs or constraints, from pull instruments, which provide incentives or guidance, and we separate package-level composition from instrument-level marginal evaluations. 3 findings stand out. First, packages weighted toward pull instruments are more likely to be chosen and rated acceptable. Second, this package-level pattern does not produce a clean instrument-level hierarchy: baseline-referenced marginal effects favor subsidies and information, but marginal means show that absolute acceptability differs only modestly across instruments, and reactions to taxes and regulations depend on whether costs fall on individuals or on firms and employers. Third, perceived effectiveness and willingness to co-produce diverge from acceptability and vary systematically by domain and by whether contributions involve time or money. These results highlight that public feasibility is multidimensional and context-dependent: it cannot be inferred from acceptability alone but must be understood as a joint evaluation of whether packages are perceived as acceptable, effective, and implementable. The paper contributes to policy design theory by reframing public feasibility as multidimensional and clarifying how conjoint experiments can distinguish package-level composition from instrument-level evaluations.