What makes a democracy capable of correcting itself?

 

Correctability is a central criterion of democratic legitimacy, but not its sole or ‘highest’ value. Democracy must also secure equal political status, fundamental rights, non-domination, and a minimum of effective participation. Correctability explains how these commitments are protected and renewed when institutions fail[1].

Correctability has three essential requirements. Institutional failures must first be visible, which requires accessible information, reliable knowledge, transparency, and protection for critical voices. They must then be repairable through procedures capable of reviewing, reversing, or revising decisions. Finally, no actor or coalition may acquire enough power to block correction permanently. These conditions must operate in practice, not merely on paper. If legal remedies, political influence, and public debate are accessible only to privileged groups, correction becomes selective and democracy loses its systemic legitimacy.

Democracy is consequently not inherently superior to every possible institutional arrangement. Its legitimacy is conditional on its ability to sustain a pluralistic, adaptive, and just order. Elections, parliamentary scrutiny, deliberation, and public mobilization matter because they can generate feedback and institutional learning not because democratic procedures are intrinsically sufficient[2]. When such procedures undermine rights, exclude minorities, or prevent correction, their democratic form no longer guarantees normative legitimacy.

Correctability must be deliberately designed and legally enforceable rather than left to political goodwill. This requires independent and adequately funded courts, appointment procedures protected against political capture, accessible legal assistance, effective whistleblower protection[3], constitutional review, transparency, and several overlapping channels of accountability. Institutional redundancy is a strength because the failure of one corrective mechanism needs not disable the entire system[4].

A polycentric administrative structure strengthens this redundancy. Local, national, judicial, societal, and supranational institutions should possess partially autonomous authority and be able to monitor, challenge, and correct one another. Local authorities are not merely implementers of national policy; they can reveal how legislation operates in practice, develop alternatives, initiate legal challenges, and demand reconsideration. Higher levels continue to enforce shared legal standards, but authority moves through reciprocal rather than exclusively top-down relationships.

Correctability therefore extends beyond the traditional separation of legislative, executive, and judicial power. Journalists, scientists, regulators, trade unions, civil-society organizations, citizen collectives, social movements, and independent media are integral parts of the democratic corrective system. They expose hidden failures, introduce alternative knowledge, organize opposition, and translate individual experiences into public problems.

Social counterpower forms the first corrective layer[5]. Protests, strikes, associations, public debate, and investigative journalism often detect injustice before political or legal institutions recognize it. Their effectiveness depends on freedom of expression, association, and demonstration, as well as on a pluralistic and accessible media system. Citizens must have access to understandable, verifiable information rather than information environments controlled by wealth, surveillance, or opaque algorithms.

Social correction nevertheless has serious limitations. Resources, organizational capacity, media access, and public visibility are unequally distributed. Fragmented information systems can prevent shared problem recognition, while mobilization may amplify polarization or manipulation. Social opposition is therefore indispensable but cannot by itself guarantee fair or reliable correction.

Political correctability translates social signals into legislation, policy, and executive accountability. Parliament, opposition parties, elections, inquiries, and public accountability form its principal mechanisms. Yet parliamentary majorities often support the executive they are expected to scrutinize. Coalition discipline, electoral incentives, media pressure, lobbying, and unequal representation can weaken independent control. Election cycles also favor immediate visibility over long-term correction, particularly for climate change, demographic transformation, and structural inequality.

Political correction can be strengthened through independent parliamentary inquiries, greater investigative authority, protected rights for opposition parties, guaranteed access to government information, periodic policy evaluations, revision and sunset clauses, transparent decision-making, and deliberative additions such as citizens’ assemblies[6]. Political decisions should be treated as revisable stages in an iterative process rather than irreversible endpoints.

Even strengthened representative institutions cannot fully correct majority power.

A referendum can sometimes correct representative inertia or test whether enacted legislation retains sufficient public support, but it is not inherently more democratic than representative decision-making. Its value depends on the question, legal effect, timing, information environment, participation conditions, protection of fundamental rights, and allocation of implementation responsibility. Advisory, suspensive, and binding corrective referendums perform different functions and should not be treated as interchangeable. Where used, the instrument requires accessible neutral information, transparent financing, realistic thresholds, judicially reviewable procedures, and a clear rule for governmental follow-up. It supplements rather than replaces Parliament, courts, deliberation, and ordinary accountability.

A parliamentary majority may act legally while gradually weakening rights, oversight bodies, or minority protection. Political correctability must therefore be supplemented by direct-democratic, constitutional, and supranational layers.

One possible supplementary layer is the referendum. Because direct votes can interrupt representative inertia but can also magnify agenda power, inequality, and irreversible choice, their democratic value depends entirely on design.

Referendums can provide a direct-democratic correction, but they are neither inherently democratic nor an independent source of legitimacy[7]. Every referendum is institutionally constructed: someone selects the subject, frames the question, determines the options and thresholds, regulates the campaign, and assigns legal consequences. Campaign resources, media structures, emotional framing, and digital targeting influence what voters understand and consider possible.

Different referendum types must be distinguished[8]. They may be citizen-initiated or government-initiated, advisory or binding, initiating or ratifying, suspensive or corrective. Binary voting suits a clearly defined veto over an adopted proposal, while multiple-choice, approval, or ranked voting may better represent situations containing several feasible alternatives. None of these methods reveals an objective “will of the people,” because agenda-setting and option selection remain exercises of political power.

A referendum can strengthen democracy by breaking representative inertia, exposing a loss of public support, and giving citizens a final corrective instrument. It can also reduce complex issues to misleading binaries, reproduce inequality in campaign influence, threaten minority rights, misalign decision-making with the scale of responsibility, privilege short-term preferences, ignore future generations, and harden decisions that later prove harmful.

A responsible referendum therefore requires several safeguards. There must be a real and legally feasible choice within the authority of the organizing institution. An independent legal assessment should establish compatibility with fundamental rights, higher law, and international obligations. The question and counting method must fit the decision structure and be reviewed independently[9].

All voters must receive the same accessible and comprehensible basic information about the proposal, alternatives, costs, legal effects, implementation risks, and consequences of each outcome. Campaign financing, sponsors, advertisements, targeting methods, and synthetic media should be transparent. Foreign financing and coordinated digital manipulation require specific regulation and independent oversight.

Thresholds must serve a defensible purpose without making the instrument unusable or encouraging strategic abstention. The legal effect must be known before voting begins. Advisory referendums require a binding duty to respond publicly and within a fixed period; binding referendums require advanced rules for implementation, transition, compensation, and existing legal positions.

Participation must account for people who face linguistic, economic, physical, or digital barriers. Institutions must also represent the interests of affected people who cannot vote, including children, non-citizen residents, asylum seekers, and future generations. Hearings, ombudspersons, youth forums, independent impact assessments, and representatives for future interests can partially address their absence.

Political and administrative responsibility remains intact after a referendum. A vote does not relieve elected institutions of responsibility for legality, coherence, implementation, minority protection, or unintended consequences. Every process should therefore include evaluation, remedies for procedural failure, monitoring of effects, and a route for later reconsideration.

Dutch experience illustrates the importance of connecting the question, institutional authority, legal effect, and follow-up process[10]. National advisory referendums exposed genuine public disagreement but produced conflict over whether and how governments had respected the result. Local referendums were most meaningful when municipalities possessed genuine decision-making authority and the consequences were clearly defined. They have been least successful when local voters were invited to decide matters ultimately controlled by higher levels. Combining a vote with later deliberation can be especially valuable: the referendum identifies rejection, while a citizens’ forum investigates the underlying reasons and develops alternatives.

A referendum should therefore be understood as democratic correction rather than a democratic shortcut. It is appropriate only when citizens face a real choice, the responsible authority is competent to act, fundamental rights remain protected, information and campaigns are fair, the result has a predetermined legal meaning, and implementation remains correctable. Producing a majority is not enough.

Constitutional adjudication forms the internal legal correction layer[11]. It may be exercised by a specialized constitutional court, by ordinary courts in concrete cases, or through a combination of both. The decisive question is whether legislation and government action can be effectively tested against constitutional principles and fundamental rights.

Constitutional review limits majority power, enables earlier intervention than supranational proceedings, and promotes consistent development of constitutional standards. It also has limitations: judges are not directly elected, appointments can become politicized, and courts can generally act only within their legal mandate and in cases brought before them. Constitutional review is therefore necessary but insufficient.

Judgments must also produce institutional change. Correctability fails when courts remedy individual cases but legislators and administrative bodies do not address the underlying structural cause. Judicial decisions should be systematically evaluated, translated into legislation and implementation practices, and monitored for compliance. Constitutional adjudication becomes genuine institutional learning only when its conclusions circulate through the broader system.

Supranational courts provide an external corrective safeguard. The European Court of Human Rights and the Court of Justice of the European Union can uphold minimum standards when national political and judicial mechanisms fail. They limit national power, provide a final remedy for individuals, harmonize legal expectations, and encourage national institutional reform.

Their authority is nevertheless bound by treaties, legal mandates, and dependence on national implementation. They generally intervene after violations occur and cannot independently prevent every gradual erosion process. Their legitimacy rests not on direct elections but on lawful mandates, judicial independence, reasoned decisions, and the protection of fundamental rights. Their function is not to govern national politics continuously, but to prevent national systems from falling below essential constitutional and human-rights standards.

A resilient democratic constitutional order consequently consists of five interacting corrective layers: social counterpower, representative political correction, conditionally used referendums, national constitutional review, and supranational legal protection. Their strength lies not in isolated operation but in continuous interaction. Social actors identify failures; political institutions translate signals into decisions; referendums may interrupt political inertia; courts protect constitutional boundaries; and supranational institutions provide a final external safeguard.

Correction is not merely negative obstruction. It also includes learning, implementation, institutional coordination, and the development of better solutions. Legal systems sometimes require an institution to issue a final decision so that rights and expectations remain settled. Finality should not be confused with infallibility or immunity: decisions must be reasoned, limited by jurisdiction, open to criticism, and capable of later legislative, constitutional, or jurisprudential development where the legal order permits.

A robust democracy is not one that avoids errors or conflict. It is one that makes failures visible, keeps remedies accessible, prevents concentrated power from blocking revision, and converts correction into lasting institutional improvement[12]. Democratic resilience ultimately depends on legally protected, mutually reinforcing, and continuously maintained structures of correctability.

Would you like to explore how a relational understanding of humanity can transform democracy, law, the economy, and public institutions? Click the link below to read the full book, Becoming Human Together, and discover a new framework for building a more humane, democratic, and correctable society within ecological limits.

https://www.researchgate.net/publication/413458836_Becoming_Human_Together_Rethinking_Democracy_Power_and_Institutions_for_a_Complex_and_Changing_World

---

[1] Guillermo O’Donnell, “Horizontal Accountability in New Democracies,” Journal of Democracy 9, no. 3 (1998): 112–126, https://doi.org/10.1353/jod.1998.0051; Bovens, “Analysing and Assessing Accountability.”

[2] OECD, Regulatory Policy Outlook 2021 (Paris: OECD Publishing, 2021), https://doi.org/10.1787/38b0fdb1-en.

[3] European Union, Directive (EU) 2019/1937 on the protection of persons who report breaches of Union law, OJ L 305, 26 November 2019, https://eur-lex.europa.eu/eli/dir/2019/1937/oj/eng.

[4] Ostrom, “Beyond Markets and States”; Charles Perrow, Normal Accidents, updated ed. (Princeton, NJ: Princeton University Press, 1999).

[5] Charles Tilly and Sidney Tarrow, Contentious Politics, 2nd ed. (New York: Oxford University Press, 2015); Young, Inclusion and Democracy.

[6] OECD, Innovative Citizen Participation and New Democratic Institutions (Paris: OECD Publishing, 2020), https://doi.org/10.1787/339306da-en.

[7] Venice Commission, Revised Code of Good Practice on Referendums, CDL-AD(2022)015, adopted 17–18 June 2022, https://www.venice.coe.int/webforms/documents/default.aspx?pdffile=CDL-AD%282022%29015-e.

[8] Maija Setälä, Referendums and Democratic Government (Basingstoke: Macmillan, 1999); Matt Qvortrup, Referendums and Ethnic Conflict (Philadelphia: University of Pennsylvania Press, 2014).

[9] Venice Commission, Revised Code of Good Practice on Referendums, especially the requirements of clarity, unity, neutrality, and balanced information.

[10] Wet raadgevend referendum, Stb. 2015, 122, repealed by Stb. 2018, 214. Identify the particular national or local referendum and official evaluation whenever drawing an empirical conclusion.

[11] Alec Stone Sweet, Governing with Judges (Oxford: Oxford University Press, 2000), https://doi.org/10.1093/0198297302.001.0001; Venice Commission, Updated Rule of Law Checklist.

[12] Longino, Science as Social Knowledge; Pettit, On the People’s Terms.