Direct democratic consent? On the relative effect of direct democratic victory and loss on political support in multi-ballot elections

Recent decades have shown an increase in alternative democratic decision-making procedures, which were often presumed to be a potential remedy to declining levels of political support around the globe. Most widespread, we have witnessed rising levels of direct citizen participation in referendums and other kinds of ballot measures. As a consequence, research into the support-related consequences of usage of and participation within direct democratic instruments has begun to unfold. One crucial finding of this research is, that the effect of direct democratic participation on political support strongly depends on the favorability of its outcome and therefore winning, rather than losing the democratic process in question. This mirrors research into the "losers' consent" within electoral contests, which has shown electoral losers in established democracies to remain largely stable in their level of political support, rather than becoming "sore losers" which reduce their level of allegiance and are often interpreted as a threat to the legitimacy of the decision and potentially to the political system. Research on the consequences of more innovative forms of democratic participation, like direct democratic victory and loss, on the other hand, is still divided on the question of how direct democratic losers might react to an unfavorable outcome. On one hand, theories of procedural fairness argue, that more direct involvement might increase support, while other arguments highlight the "zero-sum-game"-character of direct democracy, which might lead to greater levels of polarization between direct democratic winners and losers and therefore lower willingness of the ladder to remain allegiant in their level of political support. Previous studies also come to conflicting results, finding both sore and consenting losers. Moreover, they remain limited with regard to various aspects like the analysis of inexperienced contexts with low levels of institutionalization of direct democratic procedures and the absence of a direct comparison with electoral effects on the sides of winners and losers within the same context as well the same point in time. To overcome these limitations, this study makes use of the US Direct Democracy Panel Study 2012, which measured political attitudes and behavior in 13 US states over the course of the 2012 presidential elections, as well as numerous direct democratic procedures, which were held at the same time. It does not only make it possible to estimate the effect of direct democratic victory and loss on different measures of political support, but also allows for a direct comparison with the effect of electoral triumph and loss in 13 subnational contexts. While limited by the different levels of the political contests, national elections and subnational ballot measures, it still offers the opportunity for greater insights into the relative consequences of more direct forms of citizen participation for political support. This also includes the question if repeated losses within a direct democratic context or if multiple losses across different kinds of democratic procedures, such as elections and direct democratic decisions, lead to a more negative reaction and therefore can undermine the presumed "direct democratic consent".