Consequence Condition

In the AI Story

The condition identifies the most reliable indicator of whether a participatory mechanism is genuine or theatrical: does participation carry binding force, or is it advisory? The distinction is not hypothetical. Most AI governance consultation mechanisms are explicitly advisory — decision-makers solicit input while retaining unilateral authority. This structure produces predictable results: input that aligns with pre-existing decisions is adopted; input that would require genuine change is filed and forgotten.

The Trivandrum training described in The Orange Pill met the consequence condition within organizational scope because the engineers' discoveries directly influenced deployment decisions and team structure. The retention decision — preserving the team rather than converting productivity gains into headcount reduction — functioned as the credible commitment that made participation meaningful. Without costly commitment, participation lacks the guarantee that distinguishes genuine governance from its performance.

The condition creates specific design challenges for AI governance. Binding authority requires decision-makers willing to cede control, which requires either legislative mandate or strong strategic reasons to do so. Absent such conditions, participatory mechanisms default to advisory status and generate the consultative theater that characterizes most contemporary AI governance. The institutional innovations required to satisfy the condition therefore tend to require political change rather than merely administrative reform.

Consequence interacts with builder responsibility in an important way. The builder who creates a governance mechanism without giving it consequential authority bears responsibility for the predictable failure of that mechanism. Participatory governance that exists in form but not in substance actively damages the prospects for genuine participation by demonstrating — apparently — that it has been tried and found inadequate.

Origin

The consequence condition emerged from Fung's analysis of cases where participatory mechanisms had been implemented without authority and had failed predictably. The observation that advisory participation systematically underperformed binding participation drove the specification of consequence as a jointly necessary condition rather than merely a desirable feature.

Historical precedents for consequential participation include European works councils, German codetermination, and Scandinavian tripartite governance — each of which grants affected populations binding authority over specific decisions rather than advisory voice over all decisions. Fung's contribution was to specify the conditions under which binding authority produces better outcomes than expert-only governance.

Key Ideas

Advisory participation fails predictably. Without binding authority, participatory input is systematically overridden by the preferences of the decision-makers who retain unilateral control.

Costly commitment is the guarantee. Genuine participation requires that decision-makers cede authority in ways that are expensive, because commitment without cost is unreliable.

Participation without consequence damages future participation. Betrayed expectations inoculate institutions against future demands for genuine inclusion.

Consequence requires institutional innovation. Satisfying the condition typically demands legislative or structural change, not merely procedural adjustment.