In political economics, there is a growing interest in the analysis of outcomes under different constitutional rules. The empirical analysis in this field aims at testing the causal effects of constitutional differences by comparing political outcomes of one group of observations, which is treated by a constitutional rule, and a control group not treated by this rule. Prominent constitutional settings that are explored in this way are presidential and parliamentary regimes as ways to elect the executive, and majoritarian and proportional representation systems as ways to elect legislatures (Persson and Tabellini [2000], [2003]). Another comparison involves direct and representative democracy. Whereas Frey [1994], based on works by Pommerehne [1978], [1983], emphasizes the differences between direct and representative democracy as broad concepts of decision-making in his survey, more recent analyses compare outcomes of different instruments of direct democracy, such as the existence of a fiscal referendum or of a legislative initiative (see the surveys by Matsusaka [2004], Feld and Kirchgassner [2006]). According to that evidence, fiscal referendums and popular initiatives lead to more favorable fiscal and economic policy outcomes than purely representative democracy. While this literature mainly draws on field evidence, and thus has to cope with shortcomings of these data, the effect of voting on outcomes in public goods and tax-evasion games is increasingly studied in experimental analyses (see Alm, McClelland, and Schulze [1999] for referendum voting on public-good provision, fines, or the intensity of control; Feld and Tyran [2002] and Tyran and Feld [2006] on referendum voting on fines; and GUTH et al. [2004] on referendum voting versus delegation of choice of policy measures). These experimental analyses enable one to test the effects of individual characteristics of referendums exactly and thus provide more differentiated results than field studies. The paper by Fischer and Nicklisch [2007] belongs to this class of experimental studies that test the effect of referendums on public-good provision. In contrast