(In preparation for the Swiss Journal of Political Science, special issue on Publicity in Democratic Decision-Making, 2006)

The concept and phenomenon of transparency has for some time been subject to an increasingly broad and lively debate among social scientists, both in political science, law, economics, public administration, psychology and philosophy. This debate concerns normative issues of whether transparency is good thing (instrumentally and/or intrinsically) which should be promoted by institutional reforms (to which the answer usually is yes, but where there are also some objections) and empirical questions on what the effect on transparency will be on different variables and under different circumstances (for example the effectiveness of juridical and political processes and economic markets, the quality of decisions taken, the democratic status of political processes, etc).1 It is hardly surprising that the concept of transparency is used in many ways and with different meanings in this discussion. This is not a big problem as long as the state of the field to a large extent still is about brainstorming, trying out different ideas and deriving hypotheses. However as scholars are moving towards empirical testing some conceptual convergence would be preferable. In particular, as I will argue in this article, transparency should be distinguished from tw