This paper demonstrates how the idea of cognitive dissonance in psychology offers a perspective for assessing the epistemic quality of green deliberative democracy. In political theory, it is argued that deliberative democracy, in general, produces correct or good decisions (e.g., Cohen, 1986; Gutmann & Thompson, 1996), which is sometimes endorsed by the Condorcet jury theorem (e.g., Goodin, 2003; Estlund, 2008). The theorem assumes, however, that individuals remain sufficiently competent to submit, after deliberation, verdicts which correspond to the correct/good decisions. Does this assumption hold? I argue that it depends on the types of agendas and the format of deliberation. I draw on the theory of cognitive dissonance and show that individuals can be biased and hence less competent in deciding on agendas which concern adjustments of their own existing behavior. Certain deliberative arrangements, such as Goodin's (2003) reflective democracy, are vulnerable to such effect of cognitive dissonance. This reveals the tensions between three appealing conditions of green deliberative democracy, i.e., (1) universal agendas; (2) democratic disagreement resolution; and (3) epistemic performance. I suggest several escape routes to mitigate the tensions, which can serve as a foundation for arranging green deliberative democracy
