Institutions are weird. We recognize them in collections of connected agents that, as a collection, have agency themselves. Like the EU has agency and its member states have agency and their governments, parliaments, executives and judiciaries have agency, just like even small enterprises, cooperations and foundations have. These examples have in common that their agency is recognized in most legal systems. Generally such agency is looked at through the filter of as-if. But a less artificial approach has been adopted too, based on the observation that the formation of institutions, their dynamics and the behaviors of their constituents show interesting regularities. Scholars from wildly diverging disciplines (like legal theory, economics, anthropology and complexity theory) who adopt this approach generally accept that, in order to gain individual agency, it is a social institution’s fate to gather a constituency, gain identity, sovereignty over a domain, some shared beliefs, one or more functions, some rules (or laws), some policies (or habits) and some norms (or moralities). And that their constituents tend to divide in elites, work forces and publics and adopt communication channels (networks and languages). Somehow such are the main `natural’ characteristics of established social institutions, while loss of such characteristics indicate institutional trouble. Thus one of our assumptions is that the complex adaptive social systems that emerge with the use of public IT services are (becoming) institutions. Another one is that the fate of social institutions is partly natural and partly the subject of ongoing normative debates.