The democratic principle seems straightforward: those subject to binding decisions should have a voice in making them. Yet this elegant formula presupposes something increasingly absent in contemporary politics—a clear boundary delimiting who counts as subject to a decision. When the European Central Bank adjusts interest rates, Greek pensioners feel the effects without having voted for the decision-makers. When the United States subsidizes corn production, Mexican farmers watch their livelihoods dissolve. When China constructs coal plants, Pacific islanders contemplate abandoning ancestral lands to rising seas.

Traditional democratic theory developed within the conceptual framework of the bounded territorial state. Representation meant periodic elections within fixed constituencies, accountability meant the threat of electoral removal, and the demos was coextensive with citizenship. These assumptions served adequately when political decisions and their consequences aligned roughly with territorial boundaries. That alignment has now irretrievably fractured. Transnational economic integration, environmental interdependence, and digital connectivity ensure that consequential decisions routinely affect populations who lack any formal representational relationship to the deciding authority.

The temptation is to scale existing mechanisms upward—to imagine a world parliament or global elections. This approach fundamentally misunderstands the problem. Electoral representation presupposes precisely what transnational governance lacks: stable constituencies, shared political culture, and institutional infrastructure for meaningful deliberation. What we require instead is a reconceptualization of representation itself—one grounded not in territorial membership but in the principle of affected interests, operationalized through stakeholder mechanisms, and disciplined by accountability structures appropriate to non-electoral contexts.

The conventional justification for democratic representation rests on the all-affected principle: those significantly affected by a decision possess a prima facie claim to participate in making it. Within nation-states, we approximate this principle through territorial citizenship—the legal fiction that state boundaries adequately capture the population of affected persons. This approximation, always imperfect, has become increasingly untenable as the scope of consequential decisions expands beyond any single jurisdiction.

Consider the principle's radical implications when applied consistently. If affected interests ground representation claims, then the relevant constituency for climate policy includes future generations, residents of vulnerable nations regardless of their citizenship, and arguably non-human species whose flourishing human choices determine. The constituency for financial regulation extends to workers and consumers in distant countries integrated into global supply chains. The demos becomes functionally unbounded, defined not by prior membership but by the particular decision at hand.

This poses what we might call the boundary problem in its acute form. Democratic theory assumed the boundary question already settled—the demos was given before democratic procedures commenced. The all-affected principle makes the boundary itself a subject of democratic contestation, generating apparent infinite regress. Who decides who is sufficiently affected to participate in deciding who is sufficiently affected?

Several responses to this regress deserve consideration. One approach distinguishes between strongly and weakly affected interests, granting full representational standing only to those whose fundamental interests are at stake. Another proposes provisional boundaries subject to revision as affected populations become identifiable. A third accepts that different decisions warrant different demoi, abandoning the fantasy of a single unified democratic public.

What unites these responses is recognition that representation must become functionally differentiated rather than territorially unified. Rather than a single relationship between citizen and state, individuals participate in multiple overlapping representational arrangements corresponding to the various decisions that affect them. This multiplication of representational sites need not mean democratic fragmentation—properly designed, it can enable more precise alignment between decision-making authority and affected interests than territorial democracy ever achieved.

Takeaway

When decisions routinely cross borders, legitimate representation must follow affected interests rather than citizenship—meaning different decisions may warrant different constituencies participating in their making.

If the all-affected principle grounds transnational representation, electoral mechanisms cannot operationalize it. Global elections would require precisely what they are meant to produce: a global civic infrastructure, shared informational space, and minimal common identity enabling meaningful preference aggregation. Moreover, elections presuppose discrete choice among competing programs—a model poorly suited to the continuous, technical, and often reactive character of transnational governance. Alternative institutional designs are required.

Stakeholder models offer one promising direction. Rather than representing territorial constituencies, representatives in stakeholder systems speak for categories of affected interests—environmental groups, labor organizations, industry associations, consumer advocates, indigenous peoples. The European Union's consultative mechanisms, the International Labour Organization's tripartite structure, and various UN accreditation systems for civil society participation provide imperfect but instructive precedents.

Several design principles distinguish robust stakeholder representation from mere consultation. First, standing must be defined by affected interests rather than organizational capacity—otherwise wealthy Northern NGOs will dominate at the expense of actually affected populations. Second, proportionality should weight participation according to the magnitude and character of stakes involved. Those facing existential risks merit stronger voice than those experiencing marginal inconvenience. Third, deliberative quality requires institutional structures that enable genuine reason-exchange rather than mere interest-aggregation.

The objection inevitably arises: stakeholder representatives lack electoral authorization. They speak for affected interests without having been chosen by affected persons. This objection has force but proves too much—it would invalidate most actually existing transnational governance, including treaty-making, international adjudication, and regulatory harmonization. The relevant question is not whether stakeholder representation achieves electoral legitimacy but whether it achieves adequate legitimacy given realistic alternatives.

Here we must distinguish authorization from representation. Electoral authorization is one mechanism for ensuring representatives genuinely speak for those they claim to represent, but it is not the only mechanism. Representatives may achieve representativeness through demonstrated expertise, track record of advocacy, transparent selection procedures, and ongoing accountability to affected populations. What matters is that representative claims track affected interests reliably—elections are one means to this end, not the only means.

Takeaway

Stakeholder models can give voice to affected populations without requiring global elections by designing institutions where representatives speak for categories of interests rather than territorial constituencies, validated through expertise and accountability rather than ballots.

Electoral democracy solves the accountability problem elegantly: representatives who fail their constituents face removal. This sanctioning mechanism disciplines representatives prospectively—knowing they face future judgment, they attend to constituent interests now. Transnational governance lacks this elegant solution. If stakeholder representatives are not elected, how can affected populations hold them accountable?

The question reveals that accountability comprises multiple distinct elements that elections happen to bundle together. We can disaggregate these elements and ask how each might be achieved through non-electoral means. Transparency requires that affected populations can observe representative behavior—achievable through mandatory disclosure, open proceedings, and independent monitoring. Justification requires that representatives explain their choices in terms affected populations can assess—achievable through institutionalized reason-giving requirements.

Contestation requires mechanisms for challenging representative claims and decisions—achievable through administrative review, ombudsman institutions, and civil society watchdogs. Sanctioning requires consequences for representatives who fail affected populations—achievable through reputational damage, withdrawal of recognition, and legal liability. None of these requires elections; together they constitute a meaningful accountability regime.

Consider the emerging accountability practices surrounding transnational private governance—certification schemes, standard-setting bodies, and corporate self-regulation. These arrangements face persistent legitimacy deficits, yet they have developed sophisticated accountability mechanisms: multi-stakeholder boards, independent auditing, public reporting requirements, and complaint procedures. The task is not to invent accountability ex nihilo but to strengthen and democratize existing practices.

Crucial to non-electoral accountability is what we might call recursive representation. Representatives in transnational forums must themselves be accountable to organizations rooted in affected communities, who are in turn accountable to their members, creating chains of representation that ultimately ground in affected persons. These chains will be longer and more mediated than electoral representation, but mediation is not inherently illegitimate—all large-scale democracy involves extensive mediation. What matters is that the chain remains traceable and each link functions adequately.

Takeaway

Accountability need not depend on elections alone—transparency, mandatory justification, contestation mechanisms, and sanctioning through reputation or recognition can together discipline transnational representatives when designed carefully.

The conceptual apparatus of democratic theory developed for a world of bounded territorial states. That world has not disappeared—states remain consequential actors—but it has been overlaid by transnational relationships that existing categories inadequately capture. The principle of affected interests, stakeholder representation, and non-electoral accountability offer resources for theorizing democratic governance adequate to these transformed conditions.

This is not a utopian program. Elements of what I have described already exist in embryonic form across international organizations, transnational regulatory networks, and global civil society. The challenge is less invention than normative clarification and institutional strengthening—articulating the democratic principles these practices imperfectly embody and redesigning them to embody those principles more fully.

What we cannot do is wait for a global demos to spontaneously emerge before addressing global governance democratically. The decisions are being made now, affecting populations now, creating urgency that theoretical tidiness cannot satisfy. Representation must evolve alongside the transformation of political life itself—not as a finished blueprint but as ongoing democratic experimentation, disciplined by the enduring principle that those affected by power deserve voice in its exercise.