The other population inside a transformation
Every transformation gets a transition office for staff. A survey. A town hall.
The people the change is actually done to get a letter. Usually after the decision.
There's decades of research on why that gap matters. Arnstein's ladder shows how "consultation" quietly stops short of any real transfer of power. Hirschman's work on exit and voice explains why people who can't switch providers, a council, a tax office, tend not to keep complaining. They just disengage. And Kahneman and Tversky's loss aversion research suggests a ratepayer isn't just weighing a new bill against a promised saving. They're losing a name on a letterhead, a sense of who's accountable, all while the gains sit in a future nobody can point to.
None of this is really about better comms. It's about whether the process felt like something done with people, not just announced.
Two questions worth sitting with for anyone running a change right now. Has anyone told people what they're losing, not just what they're being offered? And where would the current engagement actually sit on Arnstein's ladder?
The longer version covers all of the above, plus what fifty-year-old psychology experiments can teach us about small, real choices inside a transition.
Two pieces on this page have looked at agency during transformation from the inside: staff and executives, and separately the quieter, unconscious resistance that survives even when people agree the change is needed.
There is another population living through every merger and restructure that rarely gets either kind of attention. The bank customer whose account moves to a merged institution. The policyholder whose underwriter gets acquired. The ratepayer inside a public sector restructure. The customer, in the plain sense, of whatever the organisation produces.
None of what follows is specific to one sector. It shows up wherever an organisation restructures around the people it serves rather than with them. New Zealand's council amalgamations are used here as illustration, because a fixed timetable and public consultation windows make the pattern easy to see.
Staff at least get a transition office, a survey, a town hall. Customers get a letter, usually after the decision, telling them what has changed and which number to call if it has not worked. Worth asking whether that gap is designed or simply never considered. On the evidence, it is closer to the second: the organisation is consumed by the harder problem of surviving its own restructure, and the people who use the service end up governed by decisions made about them, not with them.
A number that has never measured them
John Kotter's 1996 research put the success rate of change programmes at around 30 per cent. Boston Consulting Group's more recent work puts the failure rate at roughly 75 per cent, unmoved in decades.
Every one of those figures measures organisational performance. None asks what happened to the people on the receiving end. A transformation can be scored a success by every internal measure and still make life measurably worse for the people who use what it produces, because nobody was counting that side of the ledger. Absence of a customer-side failure metric is not evidence of success. It is more likely evidence that nobody thought to check.
What participation actually means
Sherry Arnstein, writing in 1969, set out a ladder of citizen participation that still holds up. At the bottom, manipulation: participation used to manage opinion rather than shape decisions. In the middle, informing and consultation: activity that looks like involvement but stops short of any transfer of power. At the top, partnership and citizen control, where the public genuinely shares decision rights.
Arnstein's point was not that every decision belongs at the top rung; some are not the public's to make. Her point was that organisations routinely describe middle-rung activity in the language of the top rung, and the gap between the label and the reality is where trust is lost.
A council amalgamation running to a fixed Cabinet deadline, with a short consultation window, is not obviously positioned to offer more than the middle rungs. That may be an honest constraint rather than bad faith. The problem is describing it as partnership anyway. The same pattern shows up wherever restructuring runs on a commercial timetable: a bank merging customer bases runs engagement that is almost always informing dressed as consultation, since there is nothing left to negotiate on the big decision. It is not the constraint. It is calling the constraint something it is not.
Exit, voice, and what happens when neither works
Albert Hirschman, in 1970, described two ways people respond to a decline in something they depend on: they can leave, which he called exit, or complain and try to change it from within, which he called voice.
Most public services remove exit almost entirely. You do not choose a different council. Voice carries more weight than for a commercial customer who can simply switch providers, and voice, in an amalgamation, is usually a submission form with a closing date.
People who can neither leave nor be heard tend not to keep complaining indefinitely. They disengage. It looks like apathy. It is closer to what Martin Seligman's research on learned helplessness describes: people stop attempting influence once experience has taught them it changes nothing. Falling voter turnout in local government is often read as indifference. It may equally be an accurate account of what participating has felt like.
Why fair process outperforms a fair outcome
Tom Tyler's research, from 1990 through decades of empirical work, found that acceptance of a decision by an authority depends less on whether people liked the outcome than on whether they believed the process was fair. Being heard and given a coherent account of how the decision was reached matters more to legitimacy than winning.
That is directly useful for anyone running an amalgamation, because the outcome is fixed and nobody consulting on it can promise otherwise. What remains open is the quality of the process, which means treating the explaining as the job, not as something done once the job is finished.
What people are actually losing
Daniel Kahneman and Amos Tversky's work on loss aversion found that losses are felt roughly twice as strongly as equivalent gains. This is usually discussed as a staff-side phenomenon, but it applies at least as strongly to the people using the service.
A ratepayer is not simply weighing a new rates bill against a promised efficiency gain. They are losing a council they recognised, a name on the letterhead, a sense of who is accountable. None of that appears in the business case, and it is lost while the gains sit in a future nobody can point to. A transition plan that only argues the net position is arguing on the wrong ledger for a significant part of its audience.
The nursing home experiment
Ellen Langer and Judith Rodin, in 1976, ran a study that has stayed relevant for fifty years. One floor of a nursing home was told they were responsible for arranging their own room and choosing a plant to look after. The other floor had the same choices made for them, framed as care. Three weeks later, the group given small, real choices was measurably happier and more active. Eighteen months later, their mortality rate was roughly half that of the comparison floor.
Nobody in that study controlled the big decision. What moved the outcome was control over something genuinely small. The big decision in a merger is not the customer's to make, and pretending otherwise would not survive contact with the timeline. What is usually available is a set of small, real choices inside the transition: how someone is contacted, what channel they use, how a form is worded. These get treated as implementation detail. They may do more for how change is experienced than the communications plan above them.
The choice you made yourself
A related finding comes from Langer's earlier work. In 1975, she found that people given a lottery ticket with a number they chose themselves valued it roughly three times higher, and were far less willing to sell it back, than people given a random-numbered ticket, despite identical odds. Nothing about the ticket changed. What changed was authorship.
The same effect shows up in service design. A person who selects their own appointment slot tends to trust that outcome more than someone handed the same slot by an efficient default, even when the two are functionally identical. The efficient default and the trusted default are not always the same thing, and optimising purely for the first will quietly lose the second.
Producing the service, not just receiving it
Not every relationship between a public service and its users is asked versus told. Elinor Ostrom's work on coproduction made the case that citizens are frequently part-producers, not simply recipients: a ratepayer who reports a fault is doing some of the producing. A transformation focused on internal efficiency can quietly remove that role. A self-service portal that used to let someone flag an issue directly, and now routes everything through a call centre, has taken agency away in the name of consistency. It rarely reads as a decision about agency. It reads as a process improvement.
Why the design itself resists change
Tony Schwartz, writing in Harvard Business Review in 2018, pointed at a reason transformations struggle that has nothing to do with project plans. Citing BCG's finding that 85 per cent of companies have run a major transformation in the past decade, he argued most effort goes into external things, policy, structure, process, while the internal shift goes unaddressed.
Robert Kegan and Lisa Lahey's research on immunity to change gives the mechanism: people who sincerely agree a new direction makes sense often carry a hidden competing commitment, never tested, that keeps old behaviour running underneath. The same mechanism sits underneath how organisations design for the people they serve. A council or bank can agree, sincerely, that consultation should be real, and still build a transition that is not, because the assumption that engagement is a cost was never surfaced.
When the organisation gets defensive about its own service
Chris Argyris, writing in 1991, studied why highly capable professionals often make the worst learners when their work is criticised. Having rarely failed, they have little practice learning from failure, and when their competence is questioned, they respond with defensive reasoning: externalising the problem, protecting themselves from blame.
This describes how organisations often handle customer complaints during a transformation. The instinct is to explain why the complaint is mistaken or a one-off, rather than treat it as data about the service. Edmondson's finding that psychologically safe teams report more errors, not because they make more but because they surface them, has an external equivalent: an organisation safe enough to hear that its transformation is not working will hear about problems while they are still small. One that is not will hear about them only once they have become a pattern nobody can explain away.
The self-service trap
Digitisation is doing much of the heavy lifting in the current wave of restructuring, and tends to be described as giving customers more control. That depends on what happens when the self-service path does not fit.
Dixon, Freeman and Toman's 2010 research found that reducing customer effort predicts loyalty far better than trying to delight them. A portal that works for the standard case and offers no real route to a person for the non-standard case has not reduced effort. It has relocated it, from the organisation to the customer, dressed up as empowerment. This is the case automation handles worst: the more reliable it is, the less practised everyone becomes at the exception, and the exception is precisely when a person is needed most.
A few questions worth putting to a transition plan
Where on Arnstein's ladder does our engagement sit, and does the material we send out describe it accurately. What is this change asking people to lose, not just what it is offering them, and have we said that plainly. What genuinely small, real choices can we hand to the people using this service, given that the large decision is not theirs to make. If someone cannot get what they need from self-service, how many steps until they reach a person, and did we design that number or did it happen by default.
Where this leaves things
Employees going through a merger often say afterwards that it happened to them rather than with them. Customers say some version of the same thing more often, and are asked about it less. They do not get a transition office. They get the outcome, then a letter.
The structural argument for any restructuring, the savings case, the synergy number, sits entirely on the organisation's side of the relationship. Whether it holds up in practice depends on people the organisation rarely designs for on purpose. Their acceptance was never going to come from liking the new structure. On decades of research into how people respond to decisions made about them, it was always going to come from whether the process felt like something done with them, not simply announced.