Building coalitions in a nascent subsystem: Investigating beliefs and policy preferences in Ugandan pesticide policy

Abstract Many political actors lack the power or competencies to impact policy outputs and outcomes on their own. This is why they join forces to multiply their impact. Following the advocacy coalition framework, they do so based on joint beliefs and shared policy preferences. Therefore, to understand cooperation or conflict among political actors, including the potential for policy compromises or stalemates, it seems crucial to know the allies and opponents in a political subsystem. Although many studies have investigated advocacy coalitions of like‐minded actors in various political subsystems and policy fields around the globe, not much is known about the dynamics at the origin of joint belief or common preference building. In this context, we ask: How and when do actors develop similar beliefs and joint policy preferences with others in a political subsystem? To answer this question, we investigate the early stage of policy making—a so‐called nascent subsystem—when a new issue arrives on the political agenda. We argue that it is at this stage that actors start developing joint beliefs and identifying their allies. We use expert interviews and survey data to investigate pesticide regulation, a new topic on Uganda's political agenda. We conducted a three‐step approach and selected the types of beliefs and preferences that mattered in a nascent policy setting. We then presented an extensive list of possible regulatory instruments to the stakeholders and selected those evaluated as the most relevant or conflictive. Based on this selection, we calculated belief and preference similarity and clustering to identify groups of like‐minded actors. Finally, via regression analysis, we show that joint beliefs are the result of either trust or a similar problem perception, depending on whether the actor is part of a more or less developed belief cluster.