Variations, Causes, And Consequences of Bureaucrat-Led Public Engagement In A Hybrid Regime: A Case of Thailand

Abstract

Research on direct public participation has largely focused on stable democracies with only recent extensions to some stable authoritarian contexts, while hybrid or oscillating regimes remain understudied. This dissertation addresses this gap using Thailand as the empirical setting. It asks: (1) what do direct participation or public engagement practices look like in a developing democracy with volatile politics, and (2) what drives variation in those practices? I conceptualize public engagement variation along three dimensions (recruitment inclusiveness, information flow or communication mode, and perceived impact) and operationalize them as indices at the policy-task level rather than at the level of individual public engagement activity, to better reflect how public managers make decisions. The study uses administrative data, original interview data, and original survey data conducted on K3-level Thai public managers and their equivalents covering 230 policy tasks, nested in 125 managers, across 8 organizations. At the policy-task level, multilevel linear models with random intercepts for individuals and organizations show that higher Public Service Motivation (PSM) and greater political autonomy are associated with higher scores across all three indices. Other variables, such as technocratic orientation and legal requirements, have different relationships with different dimensions. For example, having no legal requirement but having norms to engage the public has a significant relationship only with the perceived impact of the engagement activities of a policy task, not with inclusiveness or communication. These findings suggest that it might be useful to model public engagement as multidimensional dependent variables, since disaggregating the dimensions can reveal more specific ways that independent variables influence variations of public engagement. The study also compares the policy-task level results with results from analyses at the engagement-activity level. These results diverge in theoretically informative ways. No predictor is consistently significant across all dimensions within one activity. Attitude toward democracy shows a significant negative relationship with committee meeting’s information flow while it does not appear as a significant driver at the policy-task level. Technocratic orientation has positive relationship with committee’s inclusiveness but not task-level inclusiveness. These patterns are consistent with managers making policy-task-level design choices rather than thinking about each engagement venue or activity in isolation. Overall, modeling participation with policy-task-level indices aligns more closely with theoretical expectations, but these indices still have limitations and should be further refined.