Earning and defending stakeholder trust under public scrutiny

# Earning and defending stakeholder trust under public scrutiny

Oxfam GB spent 76 years building a reputation and lost a large piece of it in a fortnight. The Times published its Haiti story in February 2018. The misconduct it described had happened in 2011, and Oxfam had investigated it internally at the time. What broke in 2018 was the distance between what the organisation knew in 2011 and what its regulator, its institutional funders and its direct debit donors knew seven years later.

Around 7,000 regular donors cancelled within weeks. The UK government suspended new funding, and Oxfam stopped bidding for it until 2021. The Charity Commission's statutory inquiry, reported in 2019, found a culture that had tolerated poor behaviour and issued an official warning. Public trust in charities across the sector, which the Commission tracks annually, fell to its lowest recorded level.

None of that was decided in the hour the story broke. The failure was older: nobody had audited the gap between the internal record and the public one while there was still time to close it quietly.

The trust balance sheet

Trust behaves like a balance you accumulate in the quiet years, when nobody is paying attention, and draw down in a few days when something goes wrong. Most organisations only check the number after the withdrawal.

A company that loses a customer's trust loses a customer. A public body that loses trust still has to serve the person, and that person still has to fund it. Nobody walks away. Add the regulatory asymmetry the rules lesson sets out, and the only real defence is a balance built in advance.

Four constituencies keep separate ledgers on you, and they want different things:

• Funders (taxpayers, ministries, grant-makers) want evidence the money bought something real. They can stop the flow within a week of a headline.
• Media and watchdogs: journalists, auditors, inspectors general, charity rating sites. Any external party whose job is to score you. They are hunting the gap between what you said and what you did.
• Regulators want defensible decisions and complete disclosure. Ofcom over the BBC, the Charity Commission over Oxfam: both punish the withholding harder than the incident.
• Beneficiaries, who are underweighted and the most expensive to lose. The International Committee of the Red Cross gets into detention centres and across front lines because parties to a conflict believe its confidentiality holds. If that belief goes, the access goes, and no volume of donor goodwill replaces it.

Transparency is a strategy, not a confession

Most organisations treat transparency as damage control: something you do after a reporter files a records request. That is reactive, and it always looks like you got caught.

The pre-commitment move

Before launch, state publicly what success looks like and how you will measure it. You lock in the standard while you still have credibility to spend. The metric then belongs to the public: hitting it earns credit, missing it reads as progress in view rather than a secret uncovered.

The BBC runs a hard version of this. Since the 2017 Charter it publishes every year the names and pay bands of on-air staff earning above £150,000. It is uncomfortable each summer, and it retired a story that reporters used to chase through freedom of information requests.

Watch the second-order effect. That same disclosure supplied the evidence for the pay disputes that followed, including Carrie Gracie's resignation as China editor in January 2018 and the apology the BBC later gave her. Transparency manufactured the crisis. It also made it survivable, because nobody could accuse the BBC of hiding numbers it had published itself. Pre-commitment does not buy you calm. It buys you the right to be believed when the argument starts.

Edge case: never pre-commit to a number you do not control. Detainees visited, tonnes delivered into a besieged area, complaints resolved by a partner agency. Those depend on actors who can close the door on you. Commit to what you will publish and when, not to an outcome someone else decides.

Publish the dashboard

Open performance dashboards are a baseline expectation now. See the U.S. General Services Administration's Performance.gov for how federal agencies present goals and results in plain language.

The lesson is the posture, not the software: we show our work continuously, so no single bad month becomes a scandal. The common failure mode is a dashboard that lags reality by a quarter. Whoever spots the divergence first gets to name it, and they will not call it a data refresh problem. Freshness beats completeness.

Communicating outcomes to people who do not trust you

Sceptical audiences discount your claims automatically, so how you phrase an outcome matters as much as the outcome.

Show outputs, sell outcomes

An output is what you did: "we held 40 job-training workshops." An outcome is what changed: "312 participants were in work within six months." Agencies report outputs because outputs are easy to count and always sound busy.

Weak: "the programme delivered 1,200 service hours across the county."

Strong: "wait times for benefits appointments fell from three weeks to four days. We added 1,200 service hours to do it."

Name the limits before they do

If your result has caveats, say them plainly: this covers six months and one region, we will know more after a full year. Admitting uncertainty makes the rest of your claims more believable, because you sound like someone reporting reality.

Use money as the unit of trust

Public audiences translate everything into what it cost them, so do the translation first. The BBC prices the licence fee, around £170 a year, at roughly 46 pence a day, and that framing has held up for decades.

It has a limit worth knowing. The per-day maths works when the payer has a choice. Non-payment of the licence fee is a criminal matter in the UK, generating tens of thousands of prosecutions a year, the majority of them against women. Against that fact, "less than a coffee" sounds glib to the people most likely to be prosecuted. Use the human anchor only where the arithmetic survives a hostile recalculation and the payment is genuinely voluntary.

What the balance buys when something breaks

The acute decisions belong to the crisis lesson. What concerns you here is what you have banked before the phone rings, because that stock decides whether a bad week is survivable.

Two cases, opposite directions.

In 1995 Martin Bashir obtained his Panorama interview with Diana, Princess of Wales using faked bank statements. The BBC's own 1996 internal review cleared him. Lord Dyson's report in 2021 found the deceit real and the internal inquiry woefully ineffective, and the BBC apologised, returned the award the programme had won and undertook never to broadcast the interview again. The reporter's deception was one man's misconduct. The 1996 review is what converted it into an institutional trust failure that was still being paid for 26 years on, in the middle of a charter debate about BBC governance.

In January 2022 the ICRC discovered that attackers had taken data on more than 515,000 highly vulnerable people from its Restoring Family Links programme. It disclosed publicly within days, said plainly that the intruders had been inside for roughly ten weeks before detection, and appealed directly to the attackers not to use the data. That admission was worse-sounding than any statement its lawyers would have preferred. It also meant no journalist got to reveal it, and the 70-day detection gap was reported in the ICRC's own words.

One failure mode is universal: never blame demand. "Too many people wanted our service" is not a defence in the public sector. Meeting that demand is the job.

Building a trust message that survives scrutiny

The four questions every stakeholder is silently asking

1. What did you promise? Restate your pre-commitment.

2. What actually happened? Outcome first, with honest caveats.

3. What did it cost? In human terms, honestly calculated.

4. What happens next? Concrete, dated commitments.

Answer all four and a watchdog has little room for a gotcha, because you already told the whole story.

Match the message to the audience without changing the facts

The facts stay identical. The emphasis shifts. Taxpayers get the outcome and the per-person cost. Your board gets how the decision was made and what risk you managed. Watchdogs get the raw data and the methodology, offered before they ask.

Telling different stories to different audiences is where organisations get destroyed. If the board hears one number and the public hears another, the gap is the headline.

Where you cannot release the raw data, say why in advance. Protection organisations routinely hold back case-level detail because publishing it would identify people in danger. Publish the aggregate, publish the method, and publish the withholding rule itself. A rule stated in year one is policy. The same rule produced in response to a records request is a cover-up.

A quick self-audit before you publish

Ask your team: if a hostile reporter had everything we know and everything we say publicly, could they write a cover-up story? If yes, you are sitting on something, and you should publish it yourself, on your timing. Run that check on a schedule, not only before launches. Oxfam's exposure in 2018 sat in a file for seven years, and no launch cycle would have surfaced it.

Rebuilding after you have been the villain

Some organisations carry earned distrust. You cannot message out of it quickly.

Oxfam published a ten-point safeguarding plan within weeks of the 2018 story, commissioned an independent review of its culture, and accepted the regulator's findings rather than contesting them. UK government funding did not resume until March 2021. Three years of small, specific, kept promises, audited by people who did not work for Oxfam.

That is the pace. Reliability compounds, it is boring, and nothing else works.

Key takeaways

• Audit the gap between what you know and what you have said, on a schedule. Oxfam's 2018 collapse was written in 2011 and nobody read it back.
• Pre-commit to metrics and to publication dates you actually control. Never pre-commit to an outcome another actor can veto.
• Lead with outcomes, back with outputs. Nobody counts workshops; they count who got a job.
• Disclose the ugly detail in your own words and your own timing. The ICRC named its own ten-week detection gap. The BBC's 1996 internal review cost far more than Bashir's deceit did.
• One set of facts, framed for each audience. The gap between what your board hears and what the public hears is where watchdogs find their story.