The Leadership Table · Radical trust: the sine qua non of freedom

When an organisation cannot generate trust from within, it ends up having trust imposed from outside in the form of control, and the first to pay that bill are those who speak most loudly in defence of freedom.

Ray Dalio built one of the world’s most influential investment firms on an idea that many executives admire and few dare to apply. His first work principle reads: “Trust in radical truth and radical transparency.” For years the conversation has fixed on the two adjectives. Read the sentence again, slowly, and the decisive word turns out to be the first.

Transparency is an instrument. Trust is what that instrument, well used, produces. An organisation can publish all its data and assess every individual in real time while nobody trusts anybody. Another can reach remarkable levels of trust with far less information, because its people have learned, time and again, that telling the truth is not punished. I would call that second condition Radical Trust: the shared certainty that the truth, particularly the uncomfortable kind, has somewhere safe to land.

The question is no longer academic. The 2026 Edelman Trust Barometer, based on some 34,000 interviews in 28 countries, describes societies turning inward: seven in ten respondents are reluctant to trust people whose values, backgrounds or sources of information differ from their own, and only 32% believe the next generation will be better off. Meanwhile, artificial intelligence makes falsehood cheaper to manufacture and truth dearer to verify. When checking costs more, trusting is worth more.

One consequence receives little attention. Trust is the invisible infrastructure of freedom.

A company or a society with high trust can function with few rules, because a word given replaces the exhaustive contract and professional judgement replaces procedure. One with low trust has to regulate everything. The alternative to trust is not distrust. It is bureaucracy. Anyone who defends economic freedom, private enterprise and a limited state therefore carries a prior and rather unglamorous obligation, which is to produce trust. Without it, the regulator need not invade anything; it simply occupies the space others left empty.

Few episodes illustrate this better than Alan Mulally’s arrival at Ford in 2006. The company was on course to lose billions, yet in the business plan reviews, where each executive marked their projects green, amber or red, everything showed green. Until the head of the US operation presented a launch held back by a safety issue, marked in red. The room fell silent, waiting for the reprimand. Mulally welcomed the visibility it gave him and asked who could help. The executive kept his job. Ford had not changed its information system. It had changed what it cost to tell the truth.

The contrast is Volkswagen. For years, the group fitted some 11 million diesel vehicles with software capable of detecting when the car was undergoing an emissions test. What is revealing is who found out: a small team of researchers in the United States who measured emissions on the road, rather than an internal control or an awkward board member. The bill exceeded 30 billion dollars, much of it in fines and legal settlements.

Trust that is not produced inside is bought later, outside, and at a considerably worse price.

The classic conservative objection deserves respect. Human nature is what it is, and controls exist because blind trust tends to end badly. Quite so, which is why radical trust has very little of the naïve about it. In December 1987, signing the Intermediate-Range Nuclear Forces Treaty with Gorbachev, Ronald Reagan quoted a Russian proverb: trust, but verify. “You repeat that at every meeting,” Gorbachev replied, to laughter in the room. Two adversaries with no reason to trust one another built trust through verification, not in spite of it. Trust that allows itself to be checked is the only kind that scales.

Its mirror image had occurred a year earlier. In April 1986, reactor No. 4 at Chernobyl exploded, and the first alarms sounded not in Moscow but at Sweden’s Forsmark plant, more than a thousand kilometres away, where staff detected radioactivity that did not come from their own facility. Only then did the first official Soviet statement appear. In a system where passing on bad news was more dangerous than the accident itself, information travelled upwards late and doctored. Twenty years on, Gorbachev himself wrote that Chernobyl was “perhaps the main cause of the Soviet Union’s collapse five years later.”

In both negative cases, the truth arrived from outside. That may be the most useful diagnostic available to any board: how does your organisation find out about its own problems? If bad news reaches you through the auditor, the regulator, the press or the customer, what is failing is trust. What we called here a few weeks ago the latency of bad news has an institutional version too, and it tends to be considerably more expensive.

There is also a dimension that management tends to forget. To tell someone the truth is to treat them as an adult, capable of bearing reality and responding to it. To withhold it, even with the best intentions, is a quiet form of condescension. That is why radical trust demands two things that rarely coexist: severity with ideas and scrupulous respect for people. Truth without charity wounds. Charity without truth infantilises.

In the family enterprise, the question takes a particular form. There, trust is presumed, because it comes with the surname, and that is precisely why truth is scarce. Nobody wants to be the brother who marks the “red” in the boardroom, or the daughter who tells the founder the launch must wait. Inherited trust is comfortable, but it has never been put to the test. The business families that endure tend to be those that manage to turn it into earned trust, built on truths told in time and well received.

How many of your organisation’s controls exist because trust is missing, and who, around your board table, would dare to mark the first “red”?