Event Brief: GCRD-UNITAR Policy Dialogue
GCRD Insights: Policy Dialogue Report
Rehumanising Democracy
Public discourse and the conditions for democratic renewal
Policy insights and proposed actions from the expert dialogue convened by the Global Centre for Rehumanising Democracy in collaboration with the United Nations Institute for Training and Research, with a policy synthesis and ten proposed actions informed by the discussion.
About 30 minutes to read in full. Use the Contents button at the bottom of the screen to move between sections at any point.
Key messages for policy and practice
- Democratic renewal depends on people's capacity to disagree with dignity and on institutions that make participation consequential.
- Improvements in tone do not necessarily mean stronger agency. Credible hope requires fair consideration and opportunities to influence decisions.
- Discourse evidence must be interpreted in context, including who can speak safely and whose voices are absent. The DDI's university collaborations connect independent inquiry with civic learning.
- GCRD's policy synthesis and ten proposed actions draw on the dialogue. They are offered for consideration, with attributions identifying the contributions informing each action.
Watch the dialogue
The full recording of the policy dialogue held on 15 September 2026, with contributions from GCRD, UNITAR and the speakers whose interventions inform this report.
If the player does not appear, watch the recording on VEED.
How can we make democracy a lived experience of dignity, meaningful participation and credible hope?
Contents
- ▶Watch the dialogue
- 01Purpose and context
- 02Democratic stress as a learning problem
- 03Conceptual framing: democracy as communicative practice
- 04The instrument: the Democracy Discourse Index
- 05Country perspectives: lessons beyond the cases
- 06The moderated dialogue
- 07From participants
- 08GCRD policy synthesis for democratic renewal
- 09Proposed actions
- AAnnex A. Contributors
- BAnnex B. Note on method and on this record
- RReferences
01Purpose and context
To mark the International Day of Democracy on 15 September 2026, the Global Centre for Rehumanising Democracy (GCRD), in collaboration with the United Nations Institute for Training and Research (UNITAR), convened an expert policy dialogue on public discourse, democratic health and civic renewal. This report records the contributions and examines their implications for policy and practice.
GCRD prepared this report from the session recording. Sections 1–7 summarise the discussion, with contributions attributed to their speakers. Section 8 presents GCRD's policy synthesis; Section 9 formulates proposed actions informed by contributors' interventions. These sections do not record negotiated commitments or imply endorsement by UNITAR or individual participants. The action attributions identify the contributions informing each proposal, rather than responsibility for implementing it.
The dialogue brought together contributors from academia, multilateral organisations, civil society and government, with participants from 16 countries across four continents. Studies from Bulgaria, Pakistan and the United Kingdom using the Democracy Discourse Index (DDI) provided an entry point into a wider discussion of democratic renewal. The DDI complements assessments of institutions, rights, public services and civic participation; its university collaborations also connect research with student learning.
Opening the session, Dr Margee Ensign, Chair of the GCRD Board, set the occasion against a global crisis of democracy in which authoritarian and anti-democratic forces have gained strength. Democracy, she noted, has for decades been measured through institutions, elections, courts, political and civil rights, fair elections and freedom of the press. The approach presented at the dialogue examines instead the values that underpin democratic societies: public trust, empathy, civility and democratic agency.
Dr Ensign also placed universities at the centre of the work, quoting William Rainey Harper, the first president of the University of Chicago, speaking in 1899 at the new University of California:
The university, I contend, is the prophet of democracy. It is the university that fights the battle of democracy, its war cry being: come, let us reason together. It is the university that is the interpreter of democracy, the prophet of her past in all its vicissitudes, the prophet of her present in all its complexity, and the prophet of her future in all its possibilities.
Quoted by Dr Margee M. Ensign, Chair, GCRD Board
Universities from Pakistan, the United States, Albania, Bulgaria, North Macedonia and Greece took part in the first year of the programme. Universities in Nigeria, Rwanda, Lebanon, Ukraine, Indonesia, the United Kingdom, India and Ecuador are joining in the fall cycle with further expansion expected in the spring.
02Democratic stress as a learning problem
The keynote framing was delivered on behalf of UNITAR by Dr Jon-Hans Coetzer, in UNITAR's Division of Peace. Its argument was that democracies rarely collapse in a single moment. Drawing on Levitsky and Ziblatt's (2018) seminal work, How Democracies Die, he observed that democratic breakdown is typically gradual, a slow erosion of norms rather than a sudden rupture. The question before the international community is therefore not whether democratic stress can be prevented, but whether it can be read early enough to enable early intervention.
That question, he argued, is the one placed at the heart of a new social contract by UNESCO's International Commission on the Futures of Education in Reimagining Our Futures Together (2021): not simply the transmission of knowledge, but the cultivation of the civic capacities that sustain democratic life, among them empathy, civility and collaborative reasoning. It is also, he noted, the challenge at the core of UNITAR's strategic framework, which anchors sustainable capacity development around four objectives: empowering individuals, strengthening institutions, building learning ecosystems and supporting the United Nations system.
Democratic stress is not primarily a constitutional event. It is a learning problem, and it demands a learning response.
Dr Jon-Hans Coetzer, UNITAR
The address then set out the four signs in turn, with their scholarly and normative anchors. Empathy, in democratic theory, means recognising the dignity of those with whom we disagree: not agreement, but mutual recognition, as articulated in Martha Nussbaum's Political Emotions (2013). Where groups are portrayed as threats or traitors and harm to them is minimised, empathy is in decline; and the inclusive societies and accountable institutions called for by Sustainable Development Goal 16 are structurally impossible without that prior recognition of shared humanity.
Democratic civility, following Gutmann and Thompson (1996), is not politeness but the condition under which disagreement remains non-violent and non-exclusionary. Protest is democratic but intimidation is not. That distinction must be protected.
On trust, the address drew on the Pact for the Future, which identifies the rebuilding of multilateral trust as a precondition for collective action, and noted that the same logic holds domestically. Healthy scepticism is democratic; totalising distrust, in which no process is regarded as legitimate and no peaceful outcome is accepted, is not. Norris and Inglehart (2019) demonstrated empirically that declining institutional confidence correlates with rising support for authoritarian alternatives.
Democratic agency is the conviction that participation can change outcomes. When it fails, particularly among young people and marginalised communities, democracy loses its most essential renewal mechanism. UNESCO's new social contract calls for an expanded vision of the right to education throughout life, which must include the civic capabilities to participate meaningfully in public life.
The response, the address concluded, must match the evidence. Dialogue programming suits a mild decline in group-based understanding; protection mechanisms and platform accountability become necessary when dehumanisation is coordinated. Monitored carefully, interpreted honestly and answered proportionately, these conditions allow both complacency and panic to be avoided, and disagreements to be kept democratic.
03Conceptual framing: democracy as communicative practice
Stefanie Pukallus, Professor of Public Communication and Civil Development at the University of Sheffield, provided the academic framing. Democracy, she argued, is too often reduced to elections, formal institutions and political systems, producing a top-down account punctuated by occasional deliberative events. Following John Dewey, she offered instead an understanding of democracy as the art of living together: cooperation, giving life to society, everything citizens do in public space. It is enacted communicatively, through discourse, and communication can build a society as readily as it can diminish or destroy one.
On this account, democracy is peaceful cooperation despite difference and deep disagreement, including with those we would not choose as company. It is never fully established and never an endpoint. It is an ongoing responsibility, offline and online, and it is expressed in very small acts as much as in large ones, verbally and non-verbally alike.
Her research asks what communicative skills this actually requires. The answer she has developed is discursive civility: three skills that provide the minimal communicative conditions for disagreement to remain non-violent, and which, importantly, can be taught and learned.
- Emotional forbearanceManaging negative emotions rather than eliminating them. A society purged of anger, contempt or fear is not achievable and would not be desirable; the task is to express and engage with those emotions constructively in the pursuit of peaceful cooperation and engagement with 'the other'.
- Genuine listening as perspective-takingGenuine listening is when listeners allow for the possibility that what others have to say might be right and that they might change their mind. It requires participants to hear the other and consider someone else's point of view and experience. This is the basis of empathy and of trust.
- Making a reasonable contributionExpressing points in language accessible to everyone present, without hiding behind tradition or dogma.
These skills apply across the whole communicative spectrum of civil society that society uses to express itself to itself and to others: factual media including news journalism, unmediated everyday situations, fictional mass media, and the performative and visual arts. Critically, discursive civility does not regulate content but the manner of its expression. The aim, she said, is not censorship. Conflict can be productive but requires management so that it leads to progress and the achievement of common goals.
04The instrument: the Democracy Discourse Index
Dr Jacob Udo-Udo Jacob, Founding Executive Director of GCRD, introduced the Democracy Discourse Index. Democracy, he noted, is shaped by institutions, elections and laws, and also by the everyday quality of public discourse. Freedom House, V-Dem, the Economist Intelligence Unit and CIVICUS assess constitutions, electoral conduct, judicial independence and the civic space, and have served the democracy community well for decades. The DDI was presented as a complement to these institutional and civic-space assessments, focusing on the relational conditions expressed in public conversation.
But democratic erosion, he noted, does not always begin in the institutions. It often begins in public conversation, and by the time an index registers backsliding the conversation that sustains democratic practice has already thinned, and the interval in which a proportionate response was available has closed.
The DDI, an international, longitudinal instrument combining human coding and AI-supported analysis of public discourse, was designed to fill this gap. The DDI is not sentiment analysis, and it is not an automated toxicity detector; it does not ask whether people are being negative or rude. Rather, it measures the structural properties of democratic conversation across topics and issues at scale, recording whether citizens treat one another as full human beings, whether they engage constructively with disagreement, whether they trust shared institutions, and whether they believe their participation can change outcomes. Repeated monitoring across topics and platforms is intended to support early diagnosis and help identify where intervention may be warranted.
The scoring proceeds in five steps. Public discourse is sampled around monitored themes across social media platforms with a focus on citizens' comments in response to news discourse. Faculty-led student teams then code that corpus in their own country and in their own language, with every item scored independently by three trained coders. Where scores diverge materially, faculty adjudicate. Inter-coder agreement functions as a release gate, with a protocol threshold of 75 per cent and an operational floor of 70 per cent, measured using Krippendorff's alpha. The triple-coded human data then trains and validates AI classifiers, providing the ground truth that allows GCRD's ReHuman AI system to code at scale and produce national scores. Every item is classified by topic, so that a decision-maker sees not only a national composite but precisely where discourse is healthy and where it is hostile, and the same framework is applied identically across contexts. The tool is also designed to identify hate speech and coordinated disinformation.
Four dimensions of democratic discourse quality, twenty behavioural indicators
Every contribution is scored against twenty indicators, five in each dimension, on an ordinal scale from nought to four. Each dimension is defined as much by what it excludes as by what it captures.
Trust language
Can claims be brought into a shared field where they can be checked? A contribution may reject an institution outright and still score well, provided it treats evidence seriously.
Not agreement.
Empathy
Can the experience and dignity of other people matter? Recognition of those we disagree with, rather than assent to their conclusions.
Not sympathy.
Civility
Can disagreement continue without abuse or exclusion? The condition under which disagreement remains non-violent and non-exclusionary.
Not politeness.
Democratic agency
Can judgement lead to real action? Whether citizens write as though participation can change outcomes.
Not optimism.
Monitored themes include elections and governance, public institutions and accountability, economy and cost of living, conflict and security, climate and environment, migration and belonging, gender and social justice, religion and values, and media and information integrity. The Index is built in collaboration with Sensika Technologies, based in Sofia, and a founding consortium of universities. A first global report on the quality of public discourse across the monitored countries will be released at the end of 2026.
05Country perspectives: lessons beyond the cases
The three studies offered different perspectives on a shared policy question: what enables people to remain in democratic relationship and believe that participation matters? Their designs, platforms and periods differ. The value of reading them together lies in the questions they raise for democratic renewal, not in comparing national scores.
Bulgaria
Dr Lyubomir Stefanov of New Bulgarian University presented the Bulgarian case, opening with the observation that universities are not only a source of ideas but a breeding space for citizens, who are the particles that make democracy function. The Bulgarian corpus was drawn from X during a period of exceptional political turbulence: eight snap parliamentary elections in five years, with a further presidential round approaching.
The results, he said, were troubling. On empathy, the instrument illustrated a country that is deeply divided, with political polarisation transferring into social polarisation. Trust, both in institutions and in fellow citizens, had declined severely, with consequences for elections and for the capacity of institutions to hold the driving seat in national life. He noted a recent US Congressional Research Service report, titled Russian Hybrid Warfare Activities in Europe: Considerations for Congress, identifying Bulgaria and Romania as focal points for targeted and sophisticated hybrid operations, particularly within the Black Sea region. He linked that exposure to social and political vulnerabilities compounded by economic turbulence.
He nonetheless offered a qualified optimism, and noted that democracies die when good citizens do nothing. In Bulgaria, he argued, the situation is improving rather than deteriorating. A new government is in place, backed by a party that did not exist a few months earlier; the majority of citizens continue to look to political parties and institutions to resolve the situation; turnout has increased; and the law and the democratic process continue to be respected.
He closed with a methodological caution. Using the example of a global corruption perceptions index, he stressed the word perception: such instruments record how some individuals feel about some things, some of the time. Perception should not be mistaken for reality, and wishful thinking in either direction should be guarded against.
Pakistan
Professor Wajiha Raza Rizvi, GCRD Senior Fellow, presented the Pakistan case study, developed during the Iran talks, principally at Nadirshaw Edulji Dinshaw (N.E.D.) University of Engineering and Technology, with participation of students from Iqra University, Aga Khan University, University of Lahore, and Beaconhouse University.
Professor Rizvi emphasised the role of universities in examining systems and processes and in enabling successive generations to develop the knowledge and habits of thought that sustain a democratic culture. The Pakistan DDI studies drew on X/Twitter from March to May 2026 and on multiple digital platforms from June to September 2026. The earlier phase examined public responses to Pakistan's mediation in the US–Iran peace process amid wider conflict involving Iran, the United States and Israel, as analysed in Rizvi and Jacob's GCRD policy brief, Mediation and its Discontents.
The studies indicated low expressed agency among citizens as political actors between general elections. Yet trust in the state increased during the Iran talks, alongside greater hope and confidence in institutions' capacity to represent the nation. A central distinction was between comparatively civil language and the belief that participation can change outcomes. In the material presented, improvements in tone and trust associated with diplomatic developments did not automatically extend to agency. This matters beyond Pakistan: a calmer public conversation may coexist with a weak sense of influence. Policy should therefore examine whether people can act, organise and obtain a response, rather than equating better tone with renewed democratic capacity.
Professor Rizvi emphasised that the phases were not a comparable time series and should not be combined into a single national judgement. Differences in platforms, sampled contributions and coding reliability constrain interpretation. The observed association with diplomatic developments did not establish causation.
She also underlined the civic purpose of the research. Students from English, political science, and engineering engaged with discourse about their own society and well-being. The collaboration illustrates the university's potential to connect disciplined inquiry with awareness of shared public responsibilities, alongside its contribution to an international research instrument.
United Kingdom
Dr Jacob presented a modular application of the instrument to a single event. The new UK Prime Minister, Andy Burnham, closed his inaugural address with a call for the country to believe again and to bring back hope, and shared the concluding clip across four social media platforms under that title. GCRD coded audience responses across all four platforms in the first 48 hours.
The headline finding was that the quality of discourse in response was very low. Although the message was inspiring in itself and invited people to hope, responses were predominantly negative, and a consistent pattern emerged across the platforms. Dr Jacob framed the significance in terms of the democratic function of hope: citizens must believe that change is possible, that institutions can change, and that leaders with moral imagination can emerge, because it is hope that brings people out to campaign for what they believe in. What demagogues attempt across democracies is to create a language of hopelessness, to give people a reason to despair and a scapegoat to blame. The case, he concluded, is therefore an instance of a hopeful national message meeting a public whose relational and epistemic conditions could not carry it.
06The moderated dialogue
The second half of the session placed the three cases before a practitioner, an institutional learning specialist and a scholar. The three readings converged on a single proposition: that the evidence is indispensable and that it cannot be interpreted without establishing the conditions under which the discourse it samples was produced.
Who was able to speak: the case for a fifth condition
Peter de Clercq, GCRD Non-Executive Director and former United Nations Assistant Secretary-General and Deputy Special Representative of the Secretary-General in Somalia, opened the discussion. The three cases, he observed, point to distinct failures: resignation in Bulgaria, agency lagging behind improved tone in Pakistan, and hopeful language rejected in the United Kingdom. But before any of those profiles is interpreted, a prior question must be asked: who was able to speak, and who had already been removed from the visible conversation?
This is not only a question for fragile environments. Even in established democracies such as the United States, he noted, increasingly aggressive efforts to marginalise non-compliant public voices, through what he called political bullying or cancellation, may already be narrowing the range of voices visible in public, which is also, in itself, evidence of democratic decline. In fragile and coercive settings, however, public discourse is emphatically not a neutral sample. Surveillance, detention, licensing requirements, infiltration and threatened retaliation change who becomes visible. Critics retreat to private or encrypted spaces, journalists soften their language or leave, and civil society organisations learn to write in formulations safe enough to survive official scrutiny.
He offered South Sudan as a first-hand example. Communications interception technology supplied by the Israeli company Verint has been documented there by Amnesty International, alongside an extensive network of informants. Youth groups he met during a United Nations mission could articulate their ambitions privately and convincingly, but said that expressing them publicly would place them in grave danger. One phrase stayed with him: they described themselves as firewood for the elite, to be thrown on the fire lit by their respective leaders.
Different settings produce different distortions. In Sudan, competing armed authorities can treat civic neutrality as collaboration with the enemy. In Somalia, state security bodies, subnational authorities, al-Shabaab and clan pressures create overlapping red lines. Surveillance need not even be explicit: the possibility of being watched is sufficient to produce anticipatory self-censorship, which he termed a surveillance selection effect.
A high civility score may reflect democratic restraint. It may also reflect fear. Low polarisation may indicate cohesion, or it may mean that one side has been pushed out of the public square altogether.
Peter de Clercq
He therefore proposed that a fifth contextual condition be tested alongside the four measured dimensions: voice availability. Who has access to the civic platform? What is the cost of criticism, or even of non-alignment? How much political speech has moved into exile, private networks or silence? The most important democratic voice, he concluded, may be the one that no monitoring system can observe, because surveillance has already persuaded its owner not to speak. Silence is not evidence of consent, and restrained language is not by itself evidence of democratic health. The condition should not be folded into the composite score: its purpose is to indicate how confidently visible discourse can be treated as a reflection of the wider public sphere.
A fifth, contextual condition: voice availability
Proposed during the dialogue by Peter de Clercq. The four dimensions measure what is said. A fifth condition would qualify what could be said at all, and by whom.
Measured in the visible corpus
Conditions the reading of all four
Voice availability
- Who has access to the civic platform, and who has already been removed from the visible conversation?
- What is the cost of criticism, or even of non-alignment?
- How much political speech has moved into exile, private networks, or silence?
Relearning, formation and domestication
Returning later in the session, Mr de Clercq addressed the language of relearning democracy, and urged caution. In a number of states, it cannot be assumed that democratic practice once existed and can therefore be recovered; a system in which public expressions of dissent and consent reliably produce political consequences has never been institutionalised. In those contexts, the accurate term is democratic formation, with civic education as one component. Relearning is accurate where civic habits and responsive institutions once functioned and have weakened.
Whichever term is used, the process must be reciprocal, and Bulgaria and Pakistan show why. A better tone does not by itself create consequence. Citizens may practise listening, coalition building and collective action, but institutions must practise restraint, disclosure, response and correction. If citizens speak and nothing changes, a low agency score may accurately describe the political system rather than a cultural defect in the public.
He drew on the Somali transition from warlord-dominated fragmentation towards an increasingly institutionalised system for selecting public leadership and debating foundational questions, including the decentralised consultation that produced the transitional constitution of 2011. This was neither conventional electoral democracy nor a linear transition; its achievement was to move a significant part of the political competition among armed actors into recognisable rules, forums and bargaining processes.
From this he drew the principle of domestication. The core democratic functions may be universal: safe voices, legitimate disagreement, influence and feedback. The institutions that perform those functions differ profoundly by context. The DDI, and any learning programme built on it, therefore requires local political translation into the languages, forums, authorities and accountability relationships that people actually use. Domestication does not mean lowering the democratic standard; it means finding locally legitimate routes through which democratic functions can operate.
From evidence to institutional practice
Alexandra Severino, Programme Coordinator at UNITAR, addressed what an institution is actually supposed to do with evidence of this kind once it has it, speaking explicitly from the standpoint of an international training body rather than a ministry or a national training institute.
What she finds most useful in the DDI is that it treats the quality of public discourse as a measurable condition of democratic health in its own right, rather than waiting for erosion to surface in the more familiar signs that people associate with democratic decline. It looks at erosion while it is happening, and potentially before other signs of wear appear.
Public discourse data of this kind, she observed, is still sometimes dismissed as soft, or as less verifiable and secondary, but the evidence presented at the dialogue shows that this view does not hold. What determines whether a finding escapes that dismissal is, in her experience, unglamorous: whether there is already someone whose job it is to own a piece of curriculum or to enact a policy, and a moment in the cycle at which they must sit down and decide whether to change it. Those review moments should be in place before the evidence arrives, as a standing posture of willingness to review and reflect, rather than being improvised in response to a finding.
A composite figure, she noted, taking a composite reading from the Pakistani study as her example, does not tell an institution what to change in a course or a master's programme. It tells it where to start looking. The usable level is the topic and the dimension: on this topic, people no longer feel their voice makes a difference; or empathy for a particular group has dropped away. Only at that level is there something around which teaching can actually be built.
She offered two concrete examples. A signal that people have stopped believing their voice changes anything on climate is at bottom a democracy problem, because a citizen who has given up on being heard has given up on what democracy promises them. A signal that people have stopped extending empathy to those fleeing conflict is equally a democracy problem, because a public that stops recognising some groups' humanity is a public whose democratic culture is already eroding, whatever its institutions still say on paper. In her own institution, the first would need to be reflected in case material for a master's programme on humanitarian action and peacebuilding within the current term, and the second in a programme on climate, energy and humanitarian action.
Climate and displacement are not separate problems that happen to sit in two separate programmes. They are wicked problems: climate pressure drives displacement, and displacement strains the institutions meant to respond to climate shocks. A finding like this should prompt a transversal conversation, not two parallel ones.
Alexandra Severino, UNITAR
The same, she noted, applies within government. A climate directorate and a migration directorate treating such a finding as two separate briefs would be making precisely the mistake she would be making if two programmes were quietly updated in isolation from each other.
Communicative conditions, resistance and the limits of perception
Professor Pukallus returned to develop the point that both preceding contributions had reached from different directions: the imperative to understand the communicative conditions within which public discourse is uttered. The more discourse is used to internally exile groups of people, the less is heard from them, until the public grants itself permission to abdicate responsibility. This is not confined to the large and familiar topics such as immigration and asylum. She offered the treatment of homeless people as civilly deficient and incompatible with civil space, expelled from it with no one hearing and no one seeing, as a loss of both audibility and visibility.
There is always a dark side to democracy, she argued, and societies tend towards inequality; but there is also always resistance. That raises a question the instrument does not yet answer: what forms does public discourse take when it is resistance, surviving in communicative conditions that are not favourable? She offered an example encountered in a communist-era bunker in Tirana: letters addressed directly to the former dictator, insulting him in coded language. That is resistance discourse, and it would never appear in a public corpus. How such expression might be measured and accounted for is an open question.
She also opened the question of what agency means. Is agency real only when exercised nationally and visibly to everybody, or can it exist in a microcosm, where a thousand very small steps may constitute a large one?
Finally, she cautioned against reading public discourse as a direct report of public attitude. Public discourse is symptomatic: it points to the symptoms of a society and provides a point of investigation into what lies behind the perceptions it carries. To increase the capacity to cooperate peacefully, it is necessary to understand why public discourse looks the way it does.
07From participants
Early warning, protest and mass movements
Rashri Sharma asked whether the Index varies in the face of public movements such as protests and strikes, and how censorship and restrictions on criticism affect data collection. Dr Jacob responded with an example from Albania, where discourse on the environment had shown comparatively strong empathy and agency before a subsequent environmental protest. He presented this as an illustration of the instrument's potential to detect capacities for mobilisation. The example raises a research question rather than establishing predictive validity; peaceful protest can itself be an expression of democratic agency.
Measurement under conditions of censorship and media capture
Asmatullah Niazi, a broadcast journalist in Pakistan, commented that in an era of post-truth and hybrid regimes, where media ownership has been concentrated in the hands of the establishment, where press freedom and freedom of expression are constrained, and where state advertising revenue is tied to editorial alignment, data may be compromised, and asked how public discourse can be measured accurately.
Dr Jacob acknowledged that censorship suppresses expression and will affect scores, but noted that the instrument should also register its presence: reduced volumes of posts and comments, and depressed scores, are themselves indicators that citizens are not able to express their views as they otherwise would.
Mr de Clercq agreed that the point was fair and identical in substance to his own, citing the volume of self-censorship produced by repressive tactics. He proposed adapting the tool to include a component that takes stock of the actual civil society groups and ideas present in a society.
Ms Severino concurred, and situated the question within a wider and growing concern about the provenance of online comment, including a recent Anthropic report on the volume of comment produced by bots and comparable sources. Her argument was that the intentionality and rigour of the research, with limitations explicitly considered, is precisely what makes it useful and necessary.
Professor Pukallus added that the answer lies in understanding the media system itself: the conditions under which news journalism operates, the degree of state and institutional complicity whether forced or voluntary, and the availability of alternative sources, including journalists working in exile and civil society organisations operating in more opaque ways whose mission is nonetheless truth-seeking. She emphasised media and information literacy, not only as technical skill but as an understanding of the value of news journalism, of information and of truth, and of what that means for agency and for a society's capacity to exist on the basis of shared facts.
08GCRD policy synthesis for democratic renewal
Democratic renewal requires rights, representative institutions and accountable government, together with public relationships in which people can participate as equals. Public discourse shapes how policy is understood and whether collective action seems possible. Its quality matters to both policy legitimacy and democratic resilience. The governing concern is the dignity and worth of the human person affirmed in the United Nations Charter.
GCRD draws the following implications from the research and practitioner experience presented. The national studies do not establish a global diagnosis or a causal account of democratic decline. They inform inquiry into the conditions under which people can speak, exercise agency and obtain a response.
The relational conditions of democratic health
Assessments of constitutional safeguards, electoral integrity and civic space should also examine how people relate to institutions and to one another. The DDI contributes to this broader assessment through empathy, civility, trust and democratic agency. Its findings require interpretation alongside institutional and contextual evidence.
Participation and institutional responsiveness
Low expressed agency may reflect exclusion, repeated non-response or limited opportunities to affect decisions. Programmes focused on respectful communication alone may therefore leave the underlying difficulty untouched.
Civic learning and institutional responsiveness must develop together. People need to know how their input enters decisions, who responds, and how review and redress can be accessed. Meaningful participation requires fair consideration, reasoned decisions and the possibility of challenge. Discrimination, insecurity and inaccessible procedures can prevent participation on equal terms.
Credible hope and the integrity of public communication
Credible hope is the expectation that peaceful participation can contribute to change. Demagogic appeals exploit its absence by directing frustration towards scapegoats and dependence on a leader. Addressing the grievances and institutional failures that make such appeals persuasive requires fair treatment and accountable decisions; reassurance alone is insufficient.
Governments should therefore treat communication as part of a continuing relationship of accountability, including acknowledgement of uncertainty, explanation of constraints and correction of error. Independent journalism, civil society and digital platforms also shape whether people encounter credible information and one another's experiences. The United Nations Global Principles for Information Integrity provide a relevant framework linking societal trust, public empowerment and independent media. Efforts to strengthen trust must preserve legitimate dissent; empathy must not become a demand that people moderate justified claims for justice.
Universities and democratic learning
Universities can connect independent inquiry with public engagement and civic learning. The DDI's faculty–student collaboration illustrates how research on a society's own discourse can help students examine evidence and practise reasoned disagreement. Its effects on civic capacity require evaluation rather than assumption.
Learning should also reach officials and institutions, linking listening and deliberation to actual responsibilities for consultation, service delivery and policy review. Partnerships require the authority and resources to act on evidence, alongside academic freedom and institutional independence to sustain scrutiny.
Implications for multilateral cooperation
This agenda can support existing commitments. Sustainable Development Goal 16 connects accountable institutions with inclusive decision-making, access to information and fundamental freedoms. The Pact for the Future situates cooperation within trust, solidarity and meaningful participation. Discourse evidence can inform implementation of these commitments through inclusive, nationally owned processes.
Multilateral support should sustain collaboration among public institutions, universities, independent media and civil society, adapted to local languages and accountability relationships. In conflict-affected or repressive settings, training cannot compensate for absent guarantees of safe participation. Progress should be assessed through whether excluded groups can influence decisions and whether institutions respond.
09Proposed actions
GCRD has formulated these ten proposed actions from contributors' interventions on methodological development, institutional uptake and collaborative learning. Each attribution identifies the contribution informing the action. The proposals are offered for consideration and further development; they do not constitute agreed implementation commitments.
- Action 01
Test voice availability as a contextual condition
Develop and test an assessment of who can participate in public discourse, the costs of criticism or non-alignment, and the extent to which political expression has moved into exile, private networks or silence. Report this assessment alongside the four DDI dimensions to inform their interpretation, rather than incorporate it as an additional composite score.
Arising from the contribution of Peter de Clercq.
- Action 02
Incorporate a civil society and media-system assessment
Complement discourse measurement with political and historical analysis of civil society, media ownership, journalistic independence and restrictions on expression. Include exile media and alternative sources where appropriate, so that findings reflect the conditions under which the observed discourse was produced.
Arising from the contributions of Peter de Clercq and Professor Stef Pukallus.
- Action 03
Establish institutional review cycles
Create regular points within policy, programme and curriculum cycles at which discourse findings are considered. Assign responsibility for reviewing the evidence, determining an appropriate response and recording the resulting decision, so that institutional uptake becomes an established practice.
Arising from the contribution of Alexandra Severino.
- Action 04
Interpret findings across institutional boundaries
Bring relevant programmes and policy directorates together to examine findings that span their responsibilities. Use topic-level evidence to identify shared pressures and coordinate responses, recognising that issues such as climate, displacement and institutional trust cannot be addressed adequately in isolation.
Arising from the contribution of Alexandra Severino.
- Action 05
Adapt the instrument and learning provision to local contexts
Adapt the DDI and associated learning programmes to local languages, public forums and accountability relationships while preserving their core democratic principles and methodological consistency. Make explicit whose voices are represented, whose are missing and how these differences affect interpretation.
Arising from the contribution of Peter de Clercq.
- Action 06
Investigate resistance discourse and local agency
Examine how indirect, coded and less visible forms of expression can inform an understanding of democratic resistance. Investigate agency at community and everyday levels, as well as nationally. Research must protect people whose safety depends on anonymity, restricted visibility or silence.
Arising from the contribution of Professor Stef Pukallus.
- Action 07
Publish methodological safeguards and limitations
Maintain reliability checks as a condition of releasing findings, and publish sampling coverage, coding reliability, AI-validation results and comparability statements. Keep analytically distinct corpora and phases separate unless aggregation is methodologically justified, enabling findings to be assessed and used with confidence.
Arising from the contributions of Professor Wajiha Raza Rizvi, Dr Lyubomir Stefanov and Dr Jacob Udo-Udo Jacob.
- Action 08
Expand university participation as civic engagement
Engage more universities in democratic renewal through faculty-led research, student participation and sustained dialogue with communities. Build on the DDI collaboration to make the study of public discourse a form of civic engagement: students practise critical inquiry and reasoned disagreement while contributing evidence to public understanding.
Arising from Dr Margee M. Ensign's emphasis on universities' democratic mission.
- Action 09
Consolidate the international consortium and comparative reporting
Take forward the participation of incoming university partners and prepare the first international report on public discourse across participating observatories, planned for the end of 2026. Present shared findings alongside contextual differences, using the report to support comparative learning and inform democratic renewal.
Arising from the contribution of Dr Jacob Udo-Udo Jacob.
- Action 10
Advance collaborative learning on democratic renewal
Advance the collaboration between GCRD, UNITAR and Bard College on democracy and peacebuilding education, informed by the learning needs surfaced in this dialogue.
Arising from the contribution of Dr Antonio Garcia.
Annex AContributors
- Dr Margee M. Ensign
- Chair of the Board, Global Centre for Rehumanising Democracy
- Dr Jon-Hans Coetzer
- Senior Programme Specialist, Division of Peace, UNITAR
- Professor Stef Pukallus
- Professor of Public Communication and Civil Development, University of Sheffield
- Dr Lyubomir Stefanov
- Assistant Professor, Department of Politics, New Bulgarian University; DDI faculty lead, Bulgaria
- Professor Wajiha Raza Rizvi
- GCRD Senior Fellow, GCRD & DDI faculty lead, Pakistan
- Dr Jacob Udo-Udo Jacob
- Founding Executive Director, Global Centre for Rehumanising Democracy
- Peter de Clercq
- Non-Executive Director, GCRD; former United Nations Assistant Secretary-General and Deputy Special Representative of the Secretary-General, Somalia
- Alexandra Severino
- Programme Coordinator, United Nations Institute for Training and Research
- Dr Antonio Garcia
- Board Member/COO, Global Centre for Rehumanising Democracy; moderator
Annex BNote on method and on this record
- The Democracy Discourse Index scores public contributions against twenty behavioural indicators, five in each of four dimensions, on an ordinal scale from nought to four, producing a composite on a nought to one hundred scale mapped to four bands.
- Coding is performed by faculty-led student teams in their own countries and languages, with three independent coders per item and faculty adjudication. Inter-coder agreement operates as a release gate with a 75 per cent protocol threshold and a 70 per cent operational floor. That triple-coded human judgement provides the ground truth on which GCRD's ReHuman system is trained to code at scale and produce national scores.
- Quotations are taken from the session recording and have been lightly edited for readability, without alteration of substance. Country findings are summarised from the presentations, with numerical detail reduced to foreground their wider implications. Corpora, windows, platforms and coding designs differ between countries and, in the Pakistani case, between phases. GCRD is responsible for preparing the report and its editorial synthesis. Sections 8–9 develop implications and proposed actions from the discussion without implying collective endorsement or assigning implementation responsibilities.
ReferencesWorks cited
- Amnesty International (2021) ‘These Walls Have Ears’: The Chilling Effect of Surveillance in South Sudan. AFR 65/3577/2021, 2 February.
- Bowen, A.S., Garding, S.E., Mix, D.E. and Theohary, C.A. (2026) Russian Hybrid Warfare Activities in Europe: Considerations for Congress. Report R49134, 11 August. Congressional Research Service. www.congress.gov/crs-product/R49134
- Gutmann, A. and Thompson, D. (1996) Democracy and Disagreement. Harvard University Press.
- Harper, W.R. (2024) ‘The University and Democracy (1899)’, in Banout, T. and Ginsburg, T. (eds) The Chicago Canon on Free Inquiry and Expression. University of Chicago Press, pp. 21–44. Original address delivered 1899.
- International Commission on the Futures of Education (2021) Reimagining Our Futures Together: A New Social Contract for Education. UNESCO.
- Jacob, J.U.-U. (2026) Hope in a Low-Trust Democracy: UK Public Discourse in Response to New PM Andy Burnham. Global Centre for Rehumanising Democracy, 10 August. gcrd.org.uk/gcrd-news/hope-in-low-trust-democracy
- Levitsky, S. and Ziblatt, D. (2018) How Democracies Die. Crown.
- Norris, P. and Inglehart, R. (2019) Cultural Backlash: Trump, Brexit, and Authoritarian Populism. Cambridge University Press.
- Nussbaum, M.C. (2013) Political Emotions: Why Love Matters for Justice. Harvard University Press.
- Rizvi, W.R. and Jacob, J.U.-U. (2026) Mediation and its Discontents: Pakistan’s Role in the US–Iran Peace Process and the Quality of Democratic Discourse. GCRD Policy Brief, May. Global Centre for Rehumanising Democracy. gcrd.org.uk/insights/pakistan-discourse
- United Nations (1945) Charter of the United Nations. Preamble.
- United Nations (2024) United Nations Global Principles for Information Integrity. www.un.org/en/node/218792
- United Nations General Assembly (2015) Transforming Our World: The 2030 Agenda for Sustainable Development. Resolution A/RES/70/1, particularly Sustainable Development Goal 16. sdgs.un.org/goals/goal16
- United Nations General Assembly (2024) The Pact for the Future. Resolution A/RES/79/1. www.un.org/pact-for-the-future/en/pact-future-1
- United Nations Institute for Training and Research (n.d.) Strategic Framework 2026–2029.
