As part of the recent ‘deliberative turn’ in political theory, it has especially been argued that deliberative democratic systems would increase the legitimacy and performance of democracy. Moreover, many have placed high expectations on realizing such deliberative systems online where discussions between citizens can be found in over-abundance. Drawing on principles of deliberative democracy theory, facilitated deliberation — i.e. discussions between citizens in which a set of rules and active facilitation is employed — is often considered likelier to result in outcomes which are beneficial from a democratic point of view than discussions lacking facilitated deliberation. This has also been argued concerning the online context. However, whether facilitated online discussion actually bring about ‘better’ outcomes has seldom been empirically tested. Thus, this paper reports on an online experiment conducted in fall 2014 in which N=110 citizens discussed the position of the Swedish language in Finland. The experiment was designed to be a rather critical test for the impact of facilitated deliberation in as much as citizens only discussed in enclaves with like-minded peers, something which has often been seen as negative from a democratic point-of-view. The experiment employed a one-factor design with a control group. Half of the discussions employed facilitated deliberation whereas the other half were free-form discussions. We were thus able to study the impact per se of facilitated deliberation on the citizens taking part in online discussions. In specific, the often called main outcomes of deliberation — opinion polarization, knowledge gains — are studied. Additionally, we examine whether facilitated deliberation also brings about more side-effects, i.e. increased efficacy, feelings of interpersonal trust, trust for political institutions and propensity for political action, than discussing online without facilitated deliberation. The findings show that facilitated deliberation alleviated opinion polarization in the enclaves, whereas groups lacking facilitated deliberation experienced polarization. The effects on knowledge were not many and not unanimously positive for facilitated deliberation. Concerning side-effects, finally, the few significant effects found — concerning efficacy and generalized trust — tended to show positive effects due to facilitated deliberation. Online facilitated deliberation, although not being the ‘silver bullet’ for democracy which has sometimes been expected, does nonetheless seem more beneficial than harmful for citizens’ discussions in enclaves.
