Restructures are on the rise – but organisations ignore the repair work that makes them function

Organisations budget for the redesign but not for the work of making the new design liveable. Alice Stephenson argues that this uncosted repair labour is both essential to whether a restructure succeeds and likely to fall disproportionately to women — and that the current fashion for flattening is removing the people who perform it.

British organisations appear to be reorganising at a rate not seen in years. But the evidence suggests they are not managing it well. The IC Index 2026 , published in May by the Institute of Internal Communication with Ipsos Karian and Box, draws on a survey of 5,000 workers in organisations with at least 500 employees. It finds reported restructuring at its highest level in four years. More than half have experienced a restructure in the past twelve months, up 12 percentage points on 2024. Fewer than half say the reasons were clearly explained. Only 42 per cent think their employer is any good at helping them adapt. Trust in senior leaders has fallen nine points in a year, to 50 per cent.^[1]

By restructure I mean a wide range of events: redundancy programmes , mergers of teams or whole organisations, the removal of management layers and the redrawing of who reports to whom. More than a third of IC Index respondents had seen redundancies in the past year, also up 12 points. The pressures vary by sector. Private employers are absorbing higher National Insurance contributions and a rising wage floor. In 2025 the government announced plans to abolish NHS England, while integrated care boards were required to make substantial reductions in running costs . The Office for Students’ May 2026 analysis forecast that 42.7 per cent of English higher education institutions would report a deficit for 2025–26. And across sectors, employers are removing layers of middle management, a trend I return to below.^[2][3][4]

The obvious reading is that this is a communication failure, and that the answer is better communication. I argue that the problem sits further upstream, in what organisations believe a restructure consists of.

A restructure is not finished when the chart changes

A new structure goes live on a date. It becomes workable considerably later, if at all. In between sits a body of work that rarely appears in any plan: reconnecting teams that have been separated, rebuilding the informal routes through which the previous structure actually functioned, absorbing the anxiety of colleagues who cannot tell whether their role still exists and managing client relationships through a handover nobody outside can see.

Call this repair. It is what converts a diagram into a functioning organisation. It is performed by identifiable people, in hours nobody allocated, alongside their actual objectives. And it is invisible by design, because its purpose is that the disruption should not reach the client or the board.

Because it is invisible, it is not planned. Because it is not planned, it is not resourced. And because it is not resourced, whether a restructure succeeds depends on how much discretionary effort particular individuals are willing to supply – which is a strange basis on which to run a change programme costing millions.

The distribution is not random

Here the argument becomes uncomfortable, because there is good reason to think this work is not evenly shared.

Linda Babcock, Maria Recalde, Lise Vesterlund and Laurie Weingart examined tasks that need doing but do not advance the person who does them, such as writing a report or serving on a committee. They found that women are more likely to volunteer for such work, more likely to be asked and more likely to accept when asked. Restructures generate these tasks in volume. Babcock and colleagues’ findings give reason to expect that the repair following a restructure may also fall disproportionately to women.^[5]

When a client’s long-standing contact is moved or made redundant, someone has to explain the change, make the introductions and sit in on the first few meetings. When a role is deleted, its duties rarely disappear with it; someone works out which still need doing and absorbs them. Teams that have been split or merged need someone to rewrite the handover notes, rebuild the shared files and answer, for months, the question of who now handles what. After each consultation meeting, someone sits with the colleague who has just come out of it. Much of this work may appear in neither the restructuring plan nor anyone’s objectives.

The consequences compound. Repair is not written down, so it is not counted. It is not counted, so it does not appear in performance review or promotion decisions. And because it presents as a disposition rather than a body of work, the people doing it are described as being good with people rather than credited with carrying a workstream. The organisation extracts the labour and declines to recognise it as labour.

There is a second pattern layered on top. Michelle Ryan and Alexander Haslam’s work on the glass cliff found women disproportionately appointed to leadership positions at moments of organisational difficulty. Restructuring is precisely such a moment. Those promoted into it inherit the maximum requirement to hold other people steady at the point when their own authority is newest and least established.^[6]

Fatigue is the wrong word

Gartner has found that the average employee experienced ten planned organisational changes in 2022, against two in 2016, and that willingness to support such change fell from 74 per cent to 43 per cent over the same period.^[7]

This is generally described as change fatigue, a diagnosis with a convenient implication: that the remedy is pacing, sequencing and periods of rest. Fatigue is what happens to people who are tired, and rest addresses it.

What the data may instead be capturing is something rest does not touch. If a group repeatedly performs the work that makes an organisation survivable, receives no recognition for it and then watches that work dismantled by the next announcement, the resulting withdrawal may reflect more than exhaustion: a reasonable response to a bargain that has stopped making sense.

Flattening removes the people who do it

This matters commercially, and immediately. The current enthusiasm for removing management layers targets a population whose value is substantially relational. On an organisational chart, someone whose contribution is holding relationships together looks like duplication, because the chart has no field in which that contribution appears.

Removing these managers risks leaving fewer people with the relationships and organisational knowledge needed to make the next restructure work.

What follows

The design implication is not a wellbeing initiative. It is to treat repair as work.

That means naming, in the restructuring plan, who is responsible for reconnecting the functions being separated and allocating time to it in writing rather than assuming it will occur. It means establishing which roles carry relational load before removing a layer, which requires asking teams rather than reading charts. It means not interpreting silence as consent, given that the IC Index found fewer than half of employees feel able to express an opinion without fear of consequence. And it means ensuring the work appears in the promotion criteria, since otherwise the people carrying the change programme will continue to be the people passed over for having carried it.

None of this is soft. It is the difference between a structure that exists on paper and one that functions. The uncomfortable question for anyone planning the next reorganisation is who performed the repair after the last one, whether the organisation could name them, and what has been done to ensure they are still there.

This article gives the views of the author, not the position of LSE Business Review or the London School of Economics. You are agreeing with our comment policy when you leave a comment.