Designing collaborative decision-making: an experimental test

Considerable observational research has been conducted linking the features of the management or “metagovernance” of governance networks to their performance, yet few experimental studies have been conducted to test the... [ view full abstract ]

Considerable observational research has been conducted linking the features of the management or “metagovernance” of governance networks to their performance, yet few experimental studies have been conducted to test the various prescriptions that have emerged with respect to network design and management. While there is a long history of abstract games using the experimental method to test cooperation and conflict, scholars are now introducing more complex, realistic governance tasks and contexts while also maintaining controlled conditions (Persson et al. 2013; Esaissson et al., 2012). This paper presents the results of a pilot study of an experimental collaborative decision-making simulation related to local health care budgeting in Canada, involving N=100 students. The pilot study seeks to answer the following research question: what design features of collaborative decision-making facilitate more positive perceptions of deliberative quality and legitimacy of group decisions among participants? Twenty groups of five were presented with a context, a broad mandate from the Ministry, and were tasked with allocating funds towards priorities in a defined budget. There was one set of control groups and three sets of treatment groups with varying experimental conditions (e.g. appointed leader/chair or none, demand unanimity or not, final decision-maker or advisory in nature). Each group played three iterative policy games (an injection of significant new investment dollars, then a 20% budget cut, and then a threat of Minister over-ruling them), to measure the different responses across control and treatment groups. The results point to a complex interplay between the endogenously crafted design features and the exogenous contextual shifts to shape the deliberative experience and perceptions of the legitimacy of the group decision.

Note that the primary purpose of the pilot experimental study is to test and refine the experimental design prior to entering the field with real policy actors in Canada who make similar budgetary decisions in a collaborative context, which will occur in Summer 2017 as part of Government of Canada funded research grant. Constructive engagement with the IRSPM experimentation community in this panel session promises to help refine the experimental design prior to reentering the field.

This paper fits with the scope of interest of the Behavioural and Experimental Public Administration section, given its experimental design features and focus on public servant behavioural responses to various conditions on decision-making.