Paper short abstract:

see below

Paper long abstract:

The quality of public bureaucracy is widely regarded to be an important determinant of corruption. Recent contributions stress the merits of the classic Weberian model of bureaucracy, in particular, the positive impact of merit recruitment on lowering corruption. By contrast, the role of other features of Weberian bureaucracies such as separate employment laws, formal examination systems and permanent tenure remain contested. Moreover, studies of the impact of public sector wage levels have led to contradictory results.

The recent wave of research has typically been based on cross-country analyses that primarily rely on countrywide expert assessments of corruption and bureaucratic quality. These measures are routinely criticized for their validity; their reliance on the perception of external observers; the difficulty of comparing measurements across countries and time; and the politicization of governance indicators by policy-makers and the media.

Rather than adding to the chorus of critical voices, this paper aims to develop an alternative approach that takes the perspective of individual bureaucrats. Specifically, the paper will explore different survey techniques for the study of the impact of civil service designs on corruption in developing countries.

The paper will first present findings from a survey of ministerial bureaucrats in Eastern Europe on the relation between corruption and civil service laws, merit recruitment and the politicisation of appointments. Discussing the strengths and weakness of the existing approach, the paper will then develop an alternative approach that will rely on a mix of list experiments and conjoint survey experiments in order to gauge the causal effect of diverse civil service designs on corruption.