	School	NGO	Think Tank
Activity	Education	Humanitarian	Advocacy
Staff	30 employees	40 employees	40 employees
Location	One site in France	One site in France	One site in France
Initial action research common to all three cases	Organizational Culture Measure Recording of debates and conflicts: interview of one third of the employees, directors, staff representatives, Board of Directors Initial study delivered to all personnel
Conflict management style at the beginning of the study	Mostly accommodating style	Mostly dominating style	Mostly avoiding style
Ostrom’s Design Principle: Distribution of costs is proportional to distribution of benefits	No: No possible appeal of decisions	No: working conditions negotiated on a case-by-case basis between each employee and management	No: decisions by management on a case-by-case basis and without debate with the employee
Ostrom’s Design Principle: Self-consciously change the rules in a collective choice setting	No: Management decisions imposed without debate via formal written procedures	Yes: Access to collective bargaining arenas, but not very effective in producing collective rules that are binding on everyone	Yes: Access to collective bargaining arenas, but employees do not wish to join collective debates
High risk on partnership governance	The collective project becomes the project of just one person	Each individual tries to optimize his or her interest	Fragmentation of collective identity
Action level of the researcher	Access to governance and to all teams Capacity of action interactions level (mediation) and collective rules level. Observation over several years (from 2 to 5 years)