Social science literature on referendums and other mechanisms of direct democracy has pointed at their increasing use, especially as a result of the malfunctioning – on the input side – of the institutional system and the de-alignment of the party system (Qvortrup, 2014a;, 2014b). Referendums proliferate especially where parties are weaker but, increasingly, they are also supported by the forces of the left (parties and social movements) as an opportunity to promote citizens’ political engagement and participation. Referendums have been praised for their capacity to extend constitutional safeguards against the excessive power of politicians (Qvortrup, 2014c) and as the most direct expressions of people's will (Marxer and Pallinger, 2007). Referendums have been defined as ‘a public recognized institution wherein citizens decide or emit their opinion on issues – other than through legislative and executive elections – directly at the ballot box through universal and secret suffrage’ (Kriesi, 2012d: 7). They enhance citizens’ access to political decision making and voters’ capacity to determine the political agenda (Büchi, 2007) and can improve bargaining capacity (Hug and Schultz, 2007).

Direct democracy, in general, has been linked to values such as political and social equality, as well as assumptions about citizens’ capacity of learning the importance of preference formation within deliberative arenas – even though problems of scale have been highlighted when combining deliberative and direct democracy conceptions (Schiller, 2007). Scholars, as well as practitioners, have in fact warned about its potential lack of discursive qualities. Yet, referendums lead to higher levels of politicisation as they shift the initiative to the citizens; campaign dynamics are propitious for identity politics as they reduce the power of parties (Hutter, 2016). In fact, referendums have been said to create more mature, competent, responsible, and self-confident citizens (Büchi, 2007).

Empirical research has challenged the three main arguments by the opponents of the referendum, namely that citizens are unable to make considered judgements, that parliaments better defend minorities and that referendums delegitimise representative institutions (Caciagli and Uleri, 1994). However, a striking variation in referendum practices has been noted (Qvortrup, 2014a), which shapes their effects on citizens and institutions. Thus institutional designs vary in terms of communicative and deliberative exchanges, increases in collective learning processes, opening of political processes, maximum legitimacy and greater identification with the democratic polity (Gross, 2007).