Participatory accountability and collective action: Experimental evidence from Albania

It has been argued that accountability is a public good that only citizens can provide. Governments can put institutions in place that allow citizens to hold public servants to account, but citizens must participate in those institutions if accountability is to be achieved. Thus, citizens face a social dilemma - participate in holding public servants to account at a cost in terms of time and effort or free ride, i.e. do not participate, while benefiting from the efforts of those who do. If this characterization of accountability is valid, we would expect more cooperatively inclined citizens to participate in accountability institutions, while the less cooperatively inclined do not. We test the validity of this characterization by investigating the correlation between individual behavior in a simple public goods game and their participation in local and national accountability institutions in Albania. We study a nationally representative sample of 1800 adults with children in primary school. We find significant correlations between cooperativeness and participation in school accountability institutions and national elections, both at the individual level and the district level. These correlations are robust to the introduction of many controls in the analysis and, in the case of national elections, to the use of official election turn-out statistics in place of self-reported turn-out.