Building Trust in Public Institutions Through Experience

A citizen renewing a driving licence rarely thinks about "institutional trust." She thinks about whether the queue number system makes sense, whether the clerk looks up from the screen when she asks a question, and whether the fee she just paid actually buys her the service promised. Multiply that transaction by every tax return, benefit claim, planning application and hospital referral a government handles in a year, and you have the real trust ledger — built or drained one encounter at a time, long before anyone reads a public confidence survey.

That is the argument this piece makes, and it cuts against how most governments still approach the problem: trust in public institutions is not won through communication campaigns or reputation management — it is accumulated, transaction by transaction, as a byproduct of how fairly and predictably a service actually behaves. Get the experience right and trust follows as a side effect. Try to build trust directly, through messaging alone, and you are pouring water into a bucket with a hole in it.

Why does trust in government keep falling even as digital services improve?

Because digitisation and trust are not the same project, and treating them as interchangeable is where most public-sector transformation programmes quietly fail. A government can launch a slick portal, cut processing time in half, and still watch confidence slide — because trust is not a function of speed alone. It is a function of whether the citizen believes the system is fair, consistent, and accountable when it gets something wrong. The OECD's ongoing work on drivers of trust in public institutions has repeatedly found that reliability, responsiveness and fairness of treatment predict trust far more strongly than raw service speed or digital sophistication. A faster form does not compensate for a citizen who feels dismissed, judged, or powerless to appeal a decision.

This is also why government sits in a different behavioral category from almost every private-sector experience it gets benchmarked against. A bank customer who feels wronged can switch banks. A citizen who feels wronged by a licensing authority, a court, or a tax office cannot switch governments. That captivity changes the psychology entirely: because loss aversion — the well-documented tendency, described by Daniel Kahneman and Amos Tversky in their 1979 prospect theory work published in Econometrica, for losses to be felt roughly twice as intensely as equivalent gains — applies with unusual force when exit is not an option. A citizen cannot vote with their wallet, so every bad experience compounds instead of being diluted by the option to leave. This is the structural reason public-sector trust is so much harder to rebuild than brand loyalty: the citizen is not choosing to stay engaged. They are stuck, and they know it.

What actually determines whether a citizen trusts a public service?

Four decades of research on procedural justice, most notably the work of Yale psychologist Tom Tyler, converge on a finding that should reorder every public-sector CX roadmap: people judge the fairness of an institution less by the outcome they receive and more by the process through which they received it. Tyler's research, summarised in his 1990 book Why People Obey the Law, found that citizens who felt they were treated with respect, given a genuine voice, and shown the reasoning behind a decision reported higher trust in authorities — even when the decision itself went against them.

That single finding should reshape how public bodies measure themselves. Most government CX programmes still chase satisfaction with the outcome — was your claim approved, was your permit granted. Few measure satisfaction with the process — were you told why, were you heard, could you contest it. Renascence's own view, drawn from work across public-sector engagements, is that this is the single biggest blind spot in government experience design: agencies optimise for approval rates and processing time while ignoring the four levers that procedural justice theory says actually build trust:

• Voice — did the citizen get a genuine chance to explain their situation before a decision was made, not just a form field to fill in?
• Neutrality — did the process look and feel consistent, regardless of who was serving them or which branch they visited?
• Respect — was the citizen treated as a capable adult, or as a suspect to be processed?
• Trustworthy motives — did the citizen believe the official or the system was acting in good faith, rather than covering itself?

None of these require new technology. They require redesigning the moments where a human, or a screen standing in for one, makes a decision that affects someone's life.

How does friction quietly destroy institutional trust?

Slowly, and usually by design rather than accident. Richard Thaler and Cass Sunstein's concept of sludge — friction deliberately or carelessly built into a process that makes a beneficial action harder than it needs to be — shows up constantly in government services: the benefit renewal that defaults to lapsing rather than continuing, the appeal process buried three menus deep, the requirement to physically visit an office to submit a document that could be uploaded. Every one of these frictions sends the same signal to the citizen, whether intended or not: this system is not built for you, it is built for its own convenience. Private companies absorb sludge as a cost of a slightly worse experience. Governments absorb it as a direct hit to legitimacy, because the citizen has nowhere else to attribute the failure. There is no competitor to blame. There is only "the government," a single undifferentiated entity that just failed them again.

Choice architecture — the deliberate design of how options are presented, a concept popularised by Thaler and Sunstein in their 2008 book Nudge — cuts both ways here. Set the default toward renewal, auto-enrolment, or pre-filled forms using data the state already holds, and you remove a source of friction that costs nothing to fix and buys back trust with every interaction. Leave the default toward inaction, and every lapsed benefit or missed deadline becomes proof, in the citizen's mind, that the system was designed to fail them.

Why do citizens trust some agencies within the same government far more than others?

Because trust is not actually held at the level of "government" in the citizen's mind — it is held at the level of the last agency they dealt with, and then generalised outward. A citizen who has a smooth experience renewing a passport but a miserable one appealing a tax assessment does not conclude "the passport office is good and the tax office is bad." They conclude "government is unpredictable," because they cannot see the organisational boundary between the two. This is the affect heuristic at work, the tendency described by psychologist Paul Slovic for people to let an emotional impression from one experience colour judgement of an entire, much larger category. This is precisely why fragmented, agency-by-agency CX efforts — one department investing in service design, its neighbour ignoring the discipline entirely — cap out well below their potential. The citizen doesn't experience departments. They experience "government," end to end, across a life event that cuts through half a dozen ministries at once: getting married, registering a birth, closing a deceased relative's affairs. Designing around those life events rather than around departmental org charts is one of the clearest structural fixes available, and it's a theme worth exploring further in how life-event-based design reshapes government services.

How should a public institution actually go about rebuilding trust through experience?

Treat it as a design and measurement discipline, not a communications problem. In practice, that means a sequence:

1. Map the journey around the citizen's life event, not the agency's org chart. A birth registration, a business licence, or a bereavement touches multiple departments; if each maps its own fragment, no one owns the whole experience, and gaps between agencies become exactly where trust leaks out.
2. Measure procedural fairness, not just satisfaction or speed. Add voice, neutrality, respect and perceived good faith to the feedback instrument — through structured feedback management rather than a single post-transaction star rating that tells you nothing about why trust rose or fell.
3. Identify the moments of truth where discretion is exercised. Every process has a handful of points where a human or a system makes a judgement call — approve or deny, escalate or dismiss. These are where fairness is felt most acutely, and they deserve disproportionate design attention.
4. Redesign the defaults, not just the interface. Ask, for every step, whether inaction currently benefits the citizen or the institution. Where it benefits the institution, that is sludge, and it is corrosive to trust even when no one intended it that way.
5. Close the loop visibly. Tell citizens what changed because of their feedback or complaint. Reciprocity is a real behavioral driver: people who see that raising an issue led to a fix become more forgiving of future friction, not less.
6. Publish performance honestly, including the failures. Institutions that report processing delays or error rates candidly, alongside what they are doing about them, are trusted more over time than those that only publicise wins — because citizens quickly detect the gap between the glossy annual report and their own lived experience.

None of this is exotic. It is applied service design discipline, aimed at a sector that has historically treated CX as a private-sector import rather than a core function of legitimacy.

What does this look like when it goes wrong in practice?

Predictably, and usually for the same handful of reasons. In programmes across the public sector, the failure pattern repeats:

• KPI theatre. Agencies report improved "customer satisfaction" scores that reflect a self-selected sample of citizens who completed a transaction successfully — while the citizens who gave up, got stuck, or were denied never appear in the data at all.
• Digital-first assumed to mean digital-only. Moving a service online without maintaining an assisted channel excludes citizens with lower digital literacy or connectivity, which does not reduce friction — it relocates it onto the people least equipped to absorb it.
• Siloed ownership with no end-to-end accountability. When no single leader owns a cross-agency journey, every department can point to its own dashboard looking healthy while the citizen's overall experience remains broken.
• Consultation without consequence. Surveys and town halls that generate no visible change train citizens to stop responding — a pattern closely related to the fatigue documented in why customers stop answering surveys, and what fixes it, which applies just as directly to citizen consultation as it does to commercial feedback programmes.

Each of these failures is avoidable, and none of them require new legislation or a bigger budget — they require an institution willing to look honestly at where its own process design produces the outcome citizens least trust.

Can technology actually rebuild trust, or does it just move the problem?

It depends entirely on what the technology is asked to do. A well-known example: Estonia's X-Road data exchange layer lets citizens see, in principle, which government bodies have accessed their personal data and why — turning an opaque bureaucratic process into something inspectable. That transparency, not the underlying technology stack itself, is what earns trust. A chatbot that deflects a citizen from a genuine grievance, or an AI-scored eligibility system with no visible appeal path, does the opposite: it adds a layer of perceived opacity exactly where citizens need to see reasoning, not a black box. The lesson generalises well beyond any one country: automation should remove friction from the parts of a journey that are purely administrative, and preserve — or even increase — visibility and human recourse at the points where discretion and fairness are being judged. Confuse the two, and a digital transformation programme can measurably speed up a process while quietly eroding the trust it was meant to build. Assessing where an institution currently sits on that spectrum is a useful starting point, and a structured CX maturity assessment can help separate genuine service redesign from digitisation that has simply moved the same friction onto a screen.

What's the honest starting point for a government serious about this?

Stop measuring trust as an attitude to be surveyed once a year and start managing it as an operational output, produced or destroyed at every counter, call centre, and portal, every day. That reframing has practical consequences: it means procedural fairness metrics sit next to processing-time metrics on the same dashboard; it means a complaints team is treated as a trust-repair function rather than a cost centre to be minimised; it means the officials with the most discretion over citizens' lives get the most training in respectful, transparent decision-making, not the least.

Public institutions that take this seriously tend to organise it through dedicated public-sector experience transformation work rather than bolting CX onto an existing digital programme as an afterthought — because trust, unlike a satisfaction score, cannot be improved with a single project. It has to be built into how the institution runs.

The institution that lasts is the one that behaves consistently when no one is watching

Confidence in government has never been won by the department that shouts loudest about its mission. It is won by the tax office that explains its reasoning, the hospital that tells a patient the truth about a delay, the licensing authority that treats a rejected application as the start of a conversation rather than the end of one. Citizens forgive slow. They rarely forgive unfair, and they almost never forget arbitrary. Build the experience around that single distinction, and the trust survey numbers will follow — because for once, they will simply be describing what citizens already know to be true.