“Let’s Bring Back Hope” in 48 Hours
Democracy Discourse Index · United Kingdom
“Let’s Bring Back Hope”
in 48 Hours
The public reception to Prime Minister Andy Burnham’s appeal to hope.
An independent analysis, prepared without commission from, or consultation with, any political party or government body.
Mean discourse quality of the reception on the DDI 0–100 scale. Three-quarters of 8,086 public responses fall in the Risk band; one comment reached Healthy.
1The message
On 20 July 2026, the UK’s new Prime Minister, Andy Burnham, posted a clip from his inaugural address in front of 10 Downing Street across four social platforms, with the message: “Let's bring back hope”.
The clip travelled widely. On X alone it had gathered 897,000 views by 24 July 2026, with parallel posts running on Facebook, Instagram and TikTok.
Over 48 hours across the four platforms, the Democracy Discourse Index captured and coded the public's reply, comment by comment, against twenty behavioural indicators of democratically healthy speech (Figure 1).
2The methodology in brief
MEASUREMENT
The Democracy Discourse Index (DDI) exists because democracy has a measurement problem. There are robust tools for democratic structures and institutions, but none exists for public discourse, which informs democratic culture. The DDI is designed to fill this gap.
Developed by GCRD with Sofia-based Sensika Technologies and built with a founding consortium of seven universities across seven countries, the DDI is grounded in critical discourse analysis and deliberative democracy scholarship. Its premise, set out in the DDI working paper From Theory to Algorithm (Jacob, Angelov and Grigorova, 2025), is that discourse quality is a leading indicator of democratic health. When public speech becomes hostile, manipulative or detached from shared truth, deliberation collapses before institutional metrics register the damage. The full framework tracks six dimensions of discursive health. The comment-level instrument deployed here operationalises four of them through twenty behaviourally anchored indicators (Table 1), each scored 0–4, summing to dimension scores out of 20 and a composite renormalised to 0–100.
| D1 · reception mean 1.04/20 Empathy & Perspective-Taking Perspective-taking · Compassion · Vulnerability disclosure · Narrative use · Transformative potential | D2 · reception mean 5.25/20 Civility & Constructive Dialogue Respectful tone · Absence of ad hominem · Constructive intent · Acknowledgment of others · Behavioural norms |
| D3 · reception mean 3.07/20 Epistemic Trust Factual grounding · Source transparency · Internal consistency · Uncertainty acknowledgment · Epistemic humility | D4 · reception mean 3.24/20 Democratic Agency & Civic Engagement Political efficacy · Civic engagement · Collective framing · Rights & justice language · Call to action |
Interpretive bands: · Risk · < 25 · Concerning · 25–49 · Mixed · 50–74 · Healthy · ≥ 75
From 9,822 captured comments, pre-registered screening produced a coded corpus of 8,086, of which 8,082 received composite scores (Figure 2).
Coding followed a two-phase architecture. A three-model parallel pilot (n = 508) with two-of-three majority adjudication established the rubric's machine reliability; a human adjudication gate then required all twenty indicators to clear a pre-set 75% exact-agreement threshold before a validated single model (Gemini 2.5 Flash) coded the remaining 7,578. Against the human-corrected sample, the production model achieved 88.7% exact and 92.6% adjacent agreement, with a mean absolute error of 0.23 scale points (Figure 3).
3FINDINGS
3.1Finding 1 · Civil but closed: the condition of heart
Hope asks the hearer to risk being moved and to extend the willingness to be vulnerable. The reception shows that willingness largely withdrawn (Figure 4). Restraint is present but everything that opens one person to another is nearly absent (Figure 5), and no platform, stance or conversational setting in the study lifts it.
3.2Finding 2 · The latitudes of reception
Social judgement theory holds that audiences judge a message by its distance from their own anchor (Sherif and Hovland, 1961). Appeals close to the anchor fall in the latitude of acceptance and are embraced. Appeals far from it fall in the latitude of rejection, and do not merely fail: they boomerang, driving the hearer deeper into the anchor. Between the two lies the latitude of non-commitment, the only zone in which genuine persuasion occurs. The reception distributed across all three zones, and where it fell decided what the message could do (Figure 6).
Read through the latitudes, each stance discloses its meaning. Rejection (56.4%) is the boomerang zone in action. The modal rejecting comment does not necessarily argue with the message. Rather it uses the message as an occasion to re-perform the anchor, cynicism about the political class. The study’s hostility data confirm this redirection. Two-thirds of all hostility aims at the messenger, his party or politicians as such. For this majority, a message of hope did nothing to soften the anchor. Instead, it handed the anchor a fresh stage. We also found that rejection was not simply disengagement. Comments that ignored the message altogether scored higher (15.2) than those that rejected it (10.8).
Embrace (14.9%) is the acceptance zone, and it carries its own quiet warning. Supporters are civil (D2 8.57) but epistemically thin (D3 2.95). The message is essentially cheered but not deliberated. There is loyalty but little judgment. If a campaign counts applause as persuasion it will systematically overestimate what its message achieved.
The conditional middle (19.3%) is the latitude of non-commitment, and it behaves exactly as the theory predicts the persuadable zone should. These commenters acknowledge the appeal while refusing unconditional belief. They articulate tests, reference records, invoke collective stakes: hope, yes, but show us how it will be delivered. Their scepticism is generative, not nihilistic, and they are the only segment in the reception that leads on empathy, civility, evidence and agency at once (Figure 7). Stance and discourse quality are coded from the same comment, so this locates where deliberative behaviour already concentrates rather than showing what engaging it would produce. Persuasion, where it remains possible at all, lives here, among the fifth of the audience that has not yet decided, rather than among the majority whose anchors the message cannot reach without hardening them.
"Actions, not words" is not the rejection of hope. It is the condition on which the undecided are prepared to extend it.
The strategic translation follows from the theory as much as from the data. Messaging aimed at the rejection zone boomerangs. In other words, hope more insistently shared, will deepen the very cynicism it seeks to cure. The work of a hope-based politics is therefore twofold: co-author with the conditional middle, whose stated conditions are a published agenda for credibility, and widen the latitude of non-commitment itself, which is not a messaging task but a relational and epistemic one. Sections 5 and 6 will show what that requires.
3.3Finding 3 · Four platforms, four grounds
The same message fell on four measurably different grounds, and quality and hostility run in exact parallel (Figure 8).
First, X hosted the most rejecting, most hostile, lowest-quality reception on every measure, and it is where political journalists and engaged partisans concentrate. Judging the national reception of a hopeful message by its X reply hands the verdict to the least relationally healthy environment measured (Figure 9). Second, Facebook is where the median hearer lives. With 58.6% of the corpus, its middling reception dominates every aggregate; reaching the median British commenter means, in practice, reckoning with Facebook’s architecture and habits. Third, the younger-skewing grounds were softer. The platforms with younger audiences received the message with more openness, larger conditional segments and less hostility (Figure 10). Because no author demographics were collected this remains an ecological inference, but it unsettles a common assumption: the hardest ground for hope was not the young digital public. It was indeed the most politically saturated environment. Visit our interactive dashboard for details by platform.
As shown in Figure 11, two-thirds of hostility were aimed at the messenger, Andy Burnham, his party or the political class (a legitimacy signal), while roughly 3% redirects the message toward immigrants and Muslims (Figure 11).
3.4Finding 4 · The reception in motion
TEMPORAL DYNAMICS
The reception was quite dynamic within the 48-hour window studied. Across the first evening it hardened measurably; across the long tail it gradually recovered. Both movements confirm the latitudes, and both carry strategic information no static average can.
Comment volume rose sharply from mid-afternoon on 20 July and peaked between 18:00 and 21:00 (UTC), with 1,857 comments arriving in that three-hour window alone. As the crowd arrived, the composition of the reply shifted systematically (Figure 12). The embracing share fell from 21.0 per cent before the peak to 14.5 per cent during it and 11.9 per cent after it; rejection climbed from 47.5 to 53.7 to 61.3 per cent across the same nine hours, touching 70.7 per cent in the small hours of 21 July as mean quality fell to 12.7. Facebook, the volume-dominant platform, shows the movement most cleanly: embrace ran near a quarter of its reply at 16:00 and settled into the low teens by evening, while rejection stabilised between 55 and 62 per cent from 17:00 onward.
The hardening is the anchor becoming social. A launch moment gathers a crowd, and in a crowd the anchor acquires social proof: performing the settled cynicism of Section 3.2 becomes the price of joining in, and early hopeful replies are not so much answered as outnumbered. What did not move is as telling as what did. The conditional share barely shifted across the entire peak (21.1, 19.7, 19.2 per cent): applause proved volatile and rejection proved contagious, but conditionality held its ground. The middle did not evaporate under social pressure; the cheering did.
Then the crowd departed, and the replies changed their character. Across 22 and 23 July, with volumes down to dozens of comments and replies per bin, the conditional share climbed towards and beyond a quarter, and mean quality recovered into the high teens and low twenties. The pattern is that discourse quality in this reception is inversely related to crowd size. Deliberation seemed to return when the performance ended.
There are two strategic translations here. First, launch moments select for the anchor. The first hours of any high-visibility appeal will be its hardest ground, and reading the verdict from the peak-volume window, as overnight coverage inevitably does, samples the reception at its most performative and least deliberative. Second, the tail belongs to the conditionals. The persuadable public is still in the room after the crowd has gone, and it is then, in the low-volume aftermath, that engagement with stated conditions is most likely to be heard. Timing, not only content, is a reception variable.
4Interpretation · The architecture beneath the reception
Mapped onto GCRD’s Five-Tier Trust Architecture (GCRD, 2025), the evidence shows where a message of hope enters the structure and where the damage lies. Messages enter at the tiers of leadership and legitimacy; the reception’s heaviest evidence sits in the three tiers beneath, which no message can reach (Figure 13). The DDI’s Pakistan and Albania pilots corroborate the pattern (Rizvi and Jacob, 2026; Çyfeku, Xega and Dolani, 2026): surface tone responds to events; structural capacity does not.
5Implications for policy and leadership
Six implications follow from the measured deficits (Table 2).
| 1 Do not message the rejection zone Appeals there boomerang. Co-author with the conditional fifth: publish delivery plans, named accountability, verifiable tests. | 2 Treat delivery as communication Service is the meta-message. Costed, accountable delivery converts hope from assertion into experience. |
| 3 Build participatory infrastructure Answer the agency deficit (3.24/20): assemblies, co-design, youth routes in. Hope without agency curdles into spectatorship. | 4 Undertake epistemic repair Model the missing behaviours: transparent evidence, correction logs, honest uncertainty. Verifiability over volume. |
| 5 Invest in relational formation The empathy floor (1.04/20) must be built, not found: dialogue across difference, perspective-taking, leaders’ inner work as a democratic competence. | 6 Protect targets of redirected hostility The 3% exclusionary margin is a safety signal, not grievance. Monitor, protect, counter-frame; warmth does not displace it. |
Nothing here diminishes the message: where it was engaged it lifted the conversation, and where the heart permitted, it moved people. What the reception withholds is the fantasy that naming hope produces it. Hope is not a communications product but an experience: being recognised, told the truth and served. The seed was good. The ground is the work.
6The pessimistic age
CONCLUSION
We are living through a pessimistic age, and pessimism is not simply a mood. It has become the epistemic centre of gravity of democratic societies, the settled ground on which every political message now lands.
What began as a loss of faith in particular institutions has hardened into a loss of the sense of possibility itself. Anger, resentment and hatred are the visible symptoms, but the deeper condition is subtler and more dangerous. It is the widespread conviction that the future holds nothing that has not already been tried and failed. A citizen in that condition is not persuadable in the ordinary sense, because persuasion requires believing that evidence not yet seen could still change something. This is what the reception has been measuring all along. The anchor of Section 3.2 is this settled ground given a name; the sealed coherence of the reply, where almost no one concedes they might be wrong and almost no one allows that the future is open, is pessimism’s epistemic signature.
Prime Minister Andy Burnham deserves genuine credit for identifying this as the critical democratic vulnerability of our moment and for confronting it directly. Much of our politics has responded to public despair by ignoring it or by managing it. To clearly identify the loss of hope as the problem, and to make its restoration the explicit object of political appeal, is noteworthy. It is also, for that reason, contested ground. The far right and the demagogues arrived at the same diagnosis some time ago and have built their entire communicative repertoire on it. They speak fluently in the language of cynicism, hopelessness and betrayal, and in speaking it they deepen it. Their structural advantage is that hopelessness requires no delivery. The reception data give that asymmetry a precise form: hope is the only offer in this corpus that gets audited. The conditional fifth subjects it to tests, records and published plans; despair is subject to no test at all, which is why the rejection register can hold its position indefinitely. Two politics are competing for the same ground while carrying entirely different burdens of proof.
A politics of despair is never disappointed.
What that competition produces is a culture of hyper-scapegoating, a permanent search for whoever can absorb the frustration next. Today it is the immigrant. Tomorrow it will be someone else, and the substitution will pass almost unnoticed, because the function matters more than the target. This is the trivialisation of our public life. Politics is no longer informed by material interest or economic reasoning, but by anxiety and hopelessness looking for an object. The hostility map of Section 3.3 is this culture in cross-section. Two-thirds of all hostility is frustration seeking the nearest institutional object, the messenger, the party, the political class; the exclusionary margin is anxiety that has already found a human one. Read functionally, that margin matters less as a count than as a mechanism operating in plain sight. While the target is substitutable, the function is constant. People are exhausted by the hyper-political and are reaching for something personal and cultural instead, which is why Andy Burnham’s communicative and relational approach seems appropriate. But it is also exactly the terrain on which scapegoating lurks best, waiting for the next opportunity.
The democratic cultural backlash may well have peaked in the United Kingdom, but the habits of mind it formed will outlast it, and it is those habits, not any single election result or new leader, that this baseline seeks to record. The requirement, then, is the rehumanisation of democracy and of democratic discourse: shifting the centre of gravity from anxiety and blame towards empathy, hope and trustworthy civil exchange, backed by real democratic agency rather than mere expression. That last clause is the exact deficit the agency dimension measures, and it is why participation, not persuasion, anchors the repair agenda of Section 5. But the deeper point is that none of this can be achieved by hope-based communication alone. Communication can diagnose the wound, but it cannot heal it. What is needed is moral leadership, which means a recovered culture of moral restraint and, above all, the person who remains good in cynical times. Narrative healing is long work, and it begins with inner work, because a public trained by repeated disappointment can now distinguish authentic care from its performance with considerable accuracy. Leaders who have not done that interior work will be read, correctly, as performing.
References
Çyfeku, J., Xega, E. and Dolani, V. (2026), Albania at the Threshold, DDI Albania Policy Brief, “Fan S. Noli” University of Korçë and GCRD, June 2026.
GCRD (2025), A Framework for Understanding AI-Induced Fracture and Authentic Leadership Restoration: A Five-Tier Trust Architecture.
Jacob, J. U., Angelov, G. and Grigorova, L. (2025), From Theory to Algorithm: How We are Building the World’s First Real-Time Democracy Health Monitor, Working Paper, Sofia Information Integrity Forum.
Rizvi, W. R. and Jacob, J. U. (2026), Mediation and its Discontents: Pakistan’s Role in the US–Iran Peace Process and the Quality of Democratic Discourse, GCRD Policy Brief, Democracy Discourse Index, May 2026.
Sherif, M. and Hovland, C. I. (1961), Social Judgment: Assimilation and Contrast Effects in Communication and Attitude Change. New Haven: Yale University Press.
See the interactive map and the lexical appendix.
