The first essay in this series argued that the six forms of "psychological debt" are old wounds under new names; this one concerns what happens when the wound is given a number. Guy Champniss's method is straightforward enough: a battery of agreement statements, converted into a score from nought to one hundred, the higher the score the deeper the debt. It sounds like measurement, and in a narrow sense it is. But a score is never merely descriptive. The moment it exists, it acquires the apparatus of institutional life—an owner, a dashboard, a reporting line, a budget for its improvement—and the apparatus, once staffed, acquires an interest of its own. A cured condition is a redundant dashboard. Nobody builds one intending to switch it off after the first board presentation.
One matter of provenance belongs here, stated once and plainly. The survey behind the score was conducted for a consultancy of which the author is a director. That is disclosed in the article itself, and I note it for completeness rather than insinuation—the work stands or falls on its method, and what follows attends to the method. But a reader weighing a new diagnosis is entitled to know that the diagnosis and the practice equipped to treat it share an address.
What the Number Does
The behaviour of scores inside organisations is among the better-documented phenomena in social science. Donald Campbell observed nearly fifty years ago that the more any quantitative indicator is used for decision-making, the more it corrupts, and is corrupted by, the very processes it is meant to monitor; Marilyn Strathern later compressed the same insight into the axiom that when a measure becomes a target, it ceases to be a good measure. Now run the thought experiment. An organisation adopts the psychological-debt score. It lands, inevitably, with HR—or an enterprising executive wanting to demonstrate their capabilities with people management. A target follows—debt down from fifty-four to forty-five by year's end—because a number without a target is an orphan in any planning cycle. Employees, who are not fools, notice that their survey answers now have consequences, and the answers soften accordingly. Programmes are commissioned, attendance is recorded, the score duly improves. Whether anyone thinks more clearly, trusts a colleague more readily, or feels more securely themselves at work is a separate question, and no longer the one being asked—leat alone managed. I have written before about measurement culture as an ethical failure; this is that failure with a fresh coat of paint.
A number is easier to own than a failure, which is why failures are so often converted into numbers.
There is a subtler cost. The score aggregates six quite different accountabilities into a single figure, and a single figure gets a single owner. Yet the failures underneath—of leadership here, of management there, of governance elsewhere, as this series will show—belong to different people in different rooms. Aggregation is how they all escape. A number is easier to own than a failure, which is why failures are so often converted into numbers.
Which Way Does the Arrow Run?
Set the score's institutional career aside and consider what it claims to have found. The headline correlations are striking: those who rarely use AI report nearly twice the debt of those who use it several times a day, sixty against thirty-six; simple-task users score forty-six against thirty-five for those on complex and strategic work; the early-career cohort carries fifty-four against forty for those past their twentieth year. The reading on offer is that use pays the debt down—that adoption is, conveniently, the cure for the anxieties adoption creates.
Every one of those correlations runs at least as plausibly in the other direction. People confident in their own capability engage with demanding tools and demanding tasks; the anxious abstain. That is not a novel conjecture but the central finding of the self-efficacy literature: belief in one's competence precedes and predicts the willingness to take on challenge, not the reverse. Senior people have less to prove and more standing to survive a visible error. A cross-sectional survey cannot arbitrate between "use cures anxiety" and "the unanxious use", and the difference is not academic, because the prescriptions are opposites. If use cures, mandate it. If security permits use, then mandating the tool upon the insecure drives them into exactly the exposure they fear, without any of the conditions that make exposure safe—and the intervention deepens the condition it is scored against.
Member discussion