Introduction:

According to Douglass North’s (1981) theory of the state, institutions matter for welfare. Institutional design results from the rulers’ utility maximization within given constraints, so it is not necessarily welfaremaximizing. In an open democracy, these constraints are essentially determined by potential political competitors and/or the “exit” option of important groups within the populace (Hirschman 1970). Institutional change is usually brought under way to prevent an otherwise expected loss of votes in the next elections or an exodus of factors of production.

Leading politicians in many industrial countries increasingly call for a strengthening of direct-democratic elements in the constitutions. For instance, the former secretary general of the ruling Social Democratic Party in Germany, Franz Müntefering (2000), has repeatedly stressed that peoples‘ satisfaction with the current political system, especially confidence in the political institutions and the classe politique, is steadily falling, and this observation is supported by empirical evidence [Putnam, Pharr, and Dalton (2000)]. There are empirical indications that directdemocratic procedures are not only able to boost efficiency and productivity, but also peoples’ happiness (Frey and Stutzer 2000). This observation calls for a theoretical explanation. The goal of the paper is to contribute to a comparative theory of democratic systems by developing a concept which may be helpful in answering the following narrow but important question: Are globally inefficient projects more likely to be put through successfully in a direct-democratic or in a representative system, the latter both with and without faction discipline? Under which specific conditions can institutional change be expected to be beneficial for the people governed?

The article is structured as follows: Section 2 gives a short overview of the theoretical and empirical literature on direct democracy. In Section 3 a compensation model is introduced that is then applied to a direct-democratic setting. Direct democracy serves as a benchmark case for the analysis of indirect democracy with and without faction discipline. Direct democracy is compared to indirect (or: representative) democracy with faction dicipline in section 4 and without faction discipline (or: with conscience voting) in section 5. In subsections 4.1 and 5.1, a leftist interest group initiates a project, with the government having the same political orientation. In subsections 4.2 and 5.2, a rightist interest group initiates a project, while the government is from the opposite side of the political spectrum. Section 6 summarizes the main findings.