Association for Public Policy Analysis & Management

Collective action plays a key role in the governance of local communities. In many rural areas in the developing world, local communities have to organize critical public services locally due to insufficient service provision by government, which gives rise to autonomous self-governing networks. Literature on this self-governing network has been focusing on what constitutes a successful collective action. An example is homogeneity in value and perception that is often considered a prerequisite for effective collective action. Collective action involves frequent social interactions that widen individuals’ exposure to information, letting the participants in the collective action go through a process of calibrating the information to have shared understanding on their action. While this process enhances the effectiveness of collective action, such activities may also reinforce information bias by forcing individuals to adhere to shared beliefs and ignore contradictory but potentially more accurate information about their environment. We observed little attention has been given to this other implication in literature and explore the process with information on climate change.

In this study, we hypothesize that participation in collective action leads to significant differences in how individuals come to perceive common experience. To explain this variation, we further propose that collective action entails social pressures that participants manage by prioritizing group beliefs; in so doing they become more susceptible to information bias than those who are not involved.

To test our hypotheses, we use a 2016 survey on 2,613 rural households in Nepal, which was conducted as part of the Nepal Central Bureau of Statistics Action on Climate Today Programme. We use “participation in local savings group” from the survey to measure the participation in collective action and “25-year precipitation change during monsoon season” to measure the individual perception under the same information. We then matched the survey data with high resolution gridded climate data (CRU TS v4.02) from the Climate Research Unit of the University of East Anglia to construct whether individual perception deviates from the historical climate records. Using multinomial logit and propensity score matching, we find that community members who participate in local savings groups come to recall past climate patterns differently than non-participants. In particular, their understanding of past weather patterns exhibits greater misalignment with instrumental records compared to the views taken by those who do not participate. Our finding points to how collective action may interfere with informational processes critical to the long-term governance in local communities. The implication is critical to many developmental states, where delays in adaptation to environmental changes can lead to serious governance failures in the long run.