Self-Government

A limit on power could give people back the power to govern.

I followed the question of who could shape a decision into an occasion when a governing institution had been prevented from acting. The distinction mattered. A judgment might be defensible, its benefits real, and the authority to make the next judgment still require protection. I wanted to see what such protection restored.

In September 2019, the United Kingdom's Supreme Court held that the prorogation of Parliament was unlawful. Prorogation had suspended parliamentary business. The Court's reasoning concerned the frustration of Parliament's constitutional functions without reasonable justification; its decision made the prorogation legally ineffective. Parliament resumed sitting the following day. The Speaker's statement recorded the return and the correction of the parliamentary record.¹

Here was a restraint with a consequential object. The judges had not merely offered an unfavorable description of executive conduct. Their judgment restored Parliament's ability to act, and the resumed sitting supplied evidence beyond the judgment's language. What had been protected was an institution's capacity to undertake the work through which government could be questioned and legislation considered.

The achievement did not require me to approve every decision Parliament might subsequently make. Its restored authority was valuable before the merits of those decisions were known. That was part of what distinguished governing capacity from a result I happened to welcome.

I paused over the direction in which power had moved. An office had encountered a limit, and a representative institution had recovered room to govern. To call both things a reduction of power would lose the relation between them. To call the intervention a complete restoration of popular control would lose something else. Parliament's capacity to act did not establish how equally the people governed could influence its action.

The success deserved to remain a success. It directed my attention to the relation between the office being restrained and the people whose governing authority the restraint protected.

The power a restraint protects

There was a tempting economy in describing self-government as the absence of interference with the people's wishes. It saved the trouble of asking which people, which wishes, and which authority claimed to speak for them. It also made independent judgment look suspicious before its work had been examined. A court that limited an executive, an official who refused an unlawful instruction, or a representative who rejected a demand would all become variations on obstruction.

That account could not accommodate the restored Parliament. Some restraints preserve the means through which collective decisions remain open to challenge. Their justification concerns the power they protect as well as the action they prevent. The person frustrated by a particular limit does not, by that fact alone, stand for everyone whose authority the arrangement is meant to sustain.

This did not absolve the restraining institution. Judges possess authority too. Their independence requires an account of their jurisdiction, their reasons and the limits of their own decisions. Protecting a judgment from immediate pressure is different from exempting it from all justification. The same distinction matters when people delegate technical work they cannot sensibly perform together. Expertise supplies reasons for entrusting a task; it does not decide every question about the terms of the trust.

I found the relevant comparison in the powers surrounding a decision. What had been delegated? What remained reserved? Who could question an exercise of discretion, and what could the answer change? These questions made independence more demanding than distance. An institution isolated from every challenge and an institution protected from improper pressure might both appear remote. Their relations to the people governed could be very different.

In Uganda, elected local officials received training and financial information to help them oversee administration. The intervention improved oversight where local political leadership was not fully aligned with the ruling party; comparable aggregate gains were not measurable in aligned settings. The intervention had been randomized, but political alignment had not. Other differences between the settings remained possible explanations for the differing effects.²

Its importance to my inquiry lay in the role being exercised. An elected body did not become answerable merely by possessing a formal oversight function. Its members needed usable means to examine administration and reasons to employ those means. Equally, the existence of difficult conditions did not render oversight everywhere fictitious. An arrangement contained powers whose exercise had to be followed, rather than presumed from their names.

I returned to the Brazilian audits that had helped me distinguish disclosure from correction. In the municipalities studied, public audit findings affected electoral judgment, and favorable information could help an incumbent.³ This time I followed the favorable response. An electorate's use of information need not take the form of punishment to count as an exercise of judgment.

The findings gave retention a place within the account of accountability. Information bore on a choice people were entitled and able to make, and its consequences included an electoral advantage for some incumbents. That result did not reveal every voter's reasons. It did make room for a judgment an account of exposure and disgrace would overlook. Removing an officeholder and deciding, with better grounds, to keep one are different outcomes within that capacity.

Parliamentary scrutiny of the UK's Coronavirus Act exposed another distinction between a decision and the choices available in making it. In 2020, a parliamentary committee criticized the six-month review because it did not permit rejection of individual temporary provisions. The government later reported expiry of provisions, alongside extensions and continuation of some measures through other legislation.⁴ The review's limited choices and the reported changes both mattered; the records did not isolate how much review caused expiry. I needed to distinguish a power to accept or reject a package from a power to revise its parts.

The restored parliamentary sitting remained my clearest point of reference. A judgment had reached an actor with the authority to resume collective work. The success was specific enough to resist inflation. It had not settled the fairness of every route into Parliament, supplied everyone with equal resources, or made every future exercise of authority legitimate. It had restored a place where governing action could occur.

Who could make effective use of that place was a further question.

What taking part permits

Participation was too capacious a word to answer it. A person could take part by selecting among proposals, contributing reasons, helping form an agenda or authorizing someone else to decide. Those acts gave the person different relations to what followed. Counting them together would make involvement visible while concealing what involvement permitted.

I did not want to correct that imprecision by counting only victories. Where people disagree, a procedure that takes their contributions seriously will still leave some preferences unsatisfied. If every defeated participant had thereby been excluded, shared decisions would be possible only among people who already agreed. Self-government would become a name for unanimity, and disagreement would continually disprove it.

The outcome alone was therefore insufficient. A proposal might survive scrutiny because it was preferred on grounds participants could defend. It might survive because alternatives never reached consideration. These accounts describe similar results and different powers. To distinguish them, I had to follow what the participants were permitted to introduce and what consideration their contribution received.

In Indonesian villages, an experiment replaced meetings with plebiscites for selecting project proposals. General proposal choices changed little, while participants reported greater satisfaction and knowledge; women's proposals shifted toward poorer hamlets. The findings concerned a selection procedure within an existing program. Selecting a proposal was not the same as securing its funding or completing the work.⁵

That was enough to prevent me from calling the process empty because many choices changed little. It was not enough to call the participants sovereign over the program. Their experience and the limits of their authority belonged in the same description. Satisfaction did not certify the whole arrangement, but neither was it something to discard merely because it complicated a diagnosis.

I began to pay closer attention to the moment a record of participation was used to describe authority. An attendance list establishes that people attended. A contribution establishes that something was submitted. Whether the process conferred consent, influence or a power of refusal depends on its terms and operation. An institution cannot obtain the last of those simply by retaining evidence of the first.

Authorization also has a scope. A decision entrusted to representatives does not necessarily authorize them to remove the means by which their decisions can be challenged. A mandate to pursue an objective leaves questions about which powers may lawfully be used in its pursuit. The distinction can be consequential even when the public decision has substantial political force. A referendum result has been invoked to support enlarged ministerial authority and has met a judicial limit requiring legislation.⁶

The referendum retained its political force while the claimed enlargement of ministerial authority encountered a legal limit. Legislation was still required. I followed the division of authority that the challenge preserved: the public decision mattered, and it did not by itself supply every power claimed in responding to it.

This sharpened my understanding of the counterfeit I had been looking for. Participation becomes a substitute when its record is offered in place of the capacity the process claims to provide. The defect is the substitution. A limited advisory role honestly described and seriously used is not equivalent to an advertised power that cannot be exercised. Nor does an objection have to prevail to have received genuine consideration.

I followed the capacity a process promised through to the contribution a participant could actually make. The account needed both the terms of the opportunity and the consideration a contribution received. That left another part of participation to examine: what it took for people to use the opportunity in the first place.

Whose influence becomes usable

The capacity conferred by a process was only part of its operation. People also needed the means to use it. An available meeting, a published agenda and a formal opportunity to contribute did not tell me whose arguments would reach a person able to act on them.

An experiment involving requests to US congressional offices found that identifying prospective participants as donors changed their access to senior policymakers. It did not measure purchased legislation or durable control of policy.⁷ That limit was essential. I could recognize an advantage in access without pretending to possess the rest of its history.

Unequal access matters before capture is proved. It affects the opportunity to present a claim where consequential decisions are made. But the distinction between opportunity and control also protects the inquiry from becoming circular. If agreement with a well-resourced interest proves capture, while disagreement merely proves that capture is sophisticated, no decision could ever count against the diagnosis. Such an account would be admirably secure against evidence and consequently of little use to me.

I kept the advantage and refused the completed story. To establish policy control would require evidence of what happened beyond the meeting, how decisions were affected and whether that influence endured. The familiar fact that resources differ could not perform all of this work by repetition.

Nor did the available routes run entirely through invitations from officeholders. Research on the US Tea Party movement used rainfall at its 2009 rallies to examine the consequences of participation. The analysis linked larger rallies to later mobilization and increased Republican electoral support. Weather might also affect participants' experience or media coverage, limiting what could be attributed to attendance alone.⁸ The bounded result gave organized action outside official consultation a place in the account of usable influence.

A movement's success did not make it the public. People who opposed its objectives, people who did not join it and people affected by the policies it advanced remained part of the governing question. Efficacy and inclusion required separate judgments. The discovery that collective action could matter did not establish that its gains were equally distributed or that its purposes were just.

These differences changed what I sought in an account of participation. I wanted to know how an opportunity became usable, whose means made that possible and whose contribution still lacked a route into the decision. That inquiry preserved the importance of organization without making successful organization a title to rule everyone else.

It also brought representation into view as a continuing obligation. A representative is not merely an amplifier for whoever most effectively reaches the office. The claim to represent concerns a constituency that exceeds the people present at a meeting. How its conflicting interests should be judged cannot be settled by adding up the intensity of the requests received.

Here the grounds of self-government met the constraints I had carried from justice. People living under collective rules have reasons to demand a role in shaping them. Other people's standing and protection remain reasons to limit what any particular demand can require. The fact that a contribution is popular, organized or successful cannot settle both questions at once.

I had followed participation beyond its visible occasion into the powers and means that made it consequential. Yet even an effective contribution might end when that occasion ended. What place did the contributors possess when the next question arose?

A place in the next decision

I found a useful distinction in the recurring citizen dialogue of Ostbelgien, Belgium's German-speaking Community. Its records for 2019–24 described more than an isolated assembly. A continuing citizen council and successive assemblies gave citizen participation a recurring place in agenda and process work. The council participated in evaluating the arrangements and in the process leading to their revision.⁹

The importance was not that citizens had acquired a general power to enact their preferred policies. The role was advisory, and the distinction between recommendations and final decisions remained. What interested me was that citizen participation had an institutional place beyond one requested answer. Questions about the process could return through a continuing role rather than depend entirely on another discretionary invitation.

This was a different achievement from the restored Parliament. There, an existing representative institution had recovered its capacity to act. Here, the records showed a recurring citizen role within an arrangement for public deliberation. I had no reason to turn the two into stages of one history. Their comparison helped me separate what was restored from what was repeatedly available.

The recurrence did not make every aspect of the dialogue successful. Academic observers of its fifth assembly recorded constructive participation alongside difficulties with the breadth of the topic, the framing supplied by experts, and the specificity and feasibility of recommendations. Responses from political actors also referred to existing measures and resource constraints. The observation was not an independent causal evaluation of all policy implementation.¹⁰

Those limits bore on what the role permitted. A broad invitation could give people room to speak while leaving the eventual recommendation difficult to act on. Agreement from officials could express serious consideration without demonstrating that the assembly had changed policy. The selected participants, moreover, could not be assumed to speak for every affected resident simply because a selection process sought diversity.

I wanted to preserve these difficulties within the achievement. Advisory standing need not be worthless, and recurrence need not be permanent to be real during the period observed. The account became more exact when it stated the role's reach rather than praising or dismissing the whole arrangement at once.

The distinction also prevented me from treating a favorable policy outcome as the highest possible evidence of self-government. A government could grant a request while reserving to itself the decision about whether to hear such requests again. Another arrangement could reject a particular recommendation while preserving a meaningful public role in forming later questions. Their merits would still need examination, but the favorable answer alone would not tell me which public possessed the more continuing capacity.

What remained available mattered. I followed the recurring role through its remit, the people able to use it and the decisions open to their contributions. Those features gave the public's continuing place its substance. Repetition established that a role had returned; its influence and inclusion still required their own assessment. The distinction preserved the achievement without treating several cycles as proof of complete self-government.

I could now distinguish three achievements without making the last a test designed to discredit the first two. A correction changed something that required correction. Restored authority enabled a governing institution to act. Recurring public standing gave people a place in later decisions. An arrangement might accomplish one without accomplishing them all. Each deserved an account of its own powers and limits.

This gave self-government a temporal dimension more demanding than the survival of an institution. Keeping the same offices occupied did not by itself establish that those governed could continue to shape their work. Nor did replacing officeholders necessarily alter the public's role. The promise concerned the relation people retained to collective authority through those changes.

There were good reasons for some decisions to last. People make commitments to one another, organize their lives around protections and undertake work whose value depends on continuity. A demand that everything remain instantly reversible could place those very conditions at the mercy of whoever prevailed next. Continuing public authority therefore required judgment about what to preserve as well as what to reopen.

I had begun with a restraint that restored the capacity to govern. I ended with a question about the reach of that capacity. Even people who can revisit their decisions act on a world that others will inherit. Those others will have judgments of their own, and the authority exercised now helps determine what will remain available to their judgment.

The present need not refrain from choosing in order to respect them. But its entitlement to choose could not mean that its choices exhausted the legitimate purposes of everyone who came after it. What should a people able to govern itself leave others able to govern?