In September 2024, the Grenfell Tower Inquiry published its final report. It concluded that the building safety regime in England and Wales had been “seriously defective” for decades. Seventy-two people died in a tower block whose cladding had been certified, inspected, and approved within a system that every responsible body believed was someone else’s job to enforce. The report describes something worse than an organisational accountability failure caused by negligence or bad faith. It describes the predictable output of a regime designed, over many years, to distribute responsibility until it dissolved.
How the accountability gap in organisations gets built
The instinct after a disaster is to look for the person who failed. At Grenfell, the inquiry looked for decades. What it found was that no single person, department, or agency had failed in a way that could be isolated from the system around them. Central government had committed repeatedly to building safety. Local authorities held responsibilities. The Tenant Management Organisation held responsibilities. Contractors, designers, product manufacturers, and certification bodies each held responsibilities. Every participant operated within a narrow reading of their own duties. The narrow reading was, in every case, defensible. The aggregate was lethal.
The inquiry’s Phase 2 report traces how this happened. Over several decades, successive governments transferred key safety checks from the state to industry. Mandatory requirements were replaced with voluntary guidance. Cost-driven actors gained decision-making power over questions that had previously been held by technical regulators. The language of accountability grew more elaborate even as the enforcement behind it was progressively removed. By the time Grenfell Tower was refurbished in 2015-2016, the regime had become an elaborate landscape of guidance documents, advisory bodies, and nominal oversight, with no clear obligation on any single party to ask the question that mattered. Is this safe?
The February 2025 government response accepted 58 of the inquiry’s recommendations. It promised new licensing regimes, surveillance powers, product regulation, and social housing oversight. The promise itself is an admission. What it acknowledges, in the “careful” language of governments, is that the previous system’s accountability apparatus had been largely decorative. The arrangements existed on paper. The enforcement capacity behind them had been designed out.
This pattern is recognisable well beyond building safety. Corporate governance operates on the same principle wherever enforcement has been separated from oversight. A company can maintain a board, an audit committee, a risk committee, a compliance function, and an external auditor, and still produce no correction when something goes wrong, for exactly the same reason the pre-Grenfell regime produced none. Each body can point to its own remit. Each can show that it followed its own process. The aggregate of all that process can still leave the question that matters sitting in the space between them.
British multinational facility management and construction company, Carillion, demonstrated this in 2018. Four formal governance layers sat above the company’s operations. Board, non-executive directors, external auditors, financial regulator. Each layer held defined responsibilities. The board approved dividends from cash the company did not possess. The auditors signed off accounts the parliamentary inquiry later found unsustainable. The regulator applied standards to individual disclosures without a mandate to assess whether the overall picture was truthful. Thirty thousand supply chain businesses were left with outstanding invoices. Twenty thousand employees lost their jobs. Four oversight layers. Each working correctly within its own boundaries. No correction.
The arrangement that produced this outcome was the rational response to earlier failures. Single-layer oversight had proved insufficient, so the response was to add more layers. Each new layer addressed a genuine gap. Each also diluted the obligation to act, because each tier could assume the others had the question in hand. Adding oversight produced the appearance of greater accountability. In operational terms, it produced less. The more tiers an organisation adds, the easier it becomes for every actor within the system to discharge their formal obligations without anyone bearing the actual cost of acting. The opposite extreme, the flat or managerless organisation, fares no better. Companies that removed formal reporting lines entirely found that accountability did not disappear. It went underground, migrating to informal influence, social proximity, and invisible hierarchies that were harder to challenge precisely because no one had documented them. The correct operating question is whether the layers an organisation has are designed so that consequence lands on a named person with authority to act, or are designed so that everyone can point to someone else’s remit.
Where corporate governance without enforcement leads
The accountability gap is most visible at the scale of a public inquiry. It is most common at the scale of ordinary professional life.
A risk committee that reviews the same flagged items quarter after quarter and records them as “monitored.” An employee engagement survey whose results are presented at an all-hands meeting and filed on a shared drive without consequence. A performance management system that produces differentiated ratings on paper while every employee receives the same percentage increase. A compliance function that files regulatory returns on schedule, while the conduct those returns are supposed to govern continues unexamined between filings. Each is an accountability arrangement that generates documentation without generating consequence. The reporting exists with regularly updated dashboards. The committees meet. What is absent is the point at which someone with authority is required to act on what the reports say, and bears a cost if they do not.
The people who pay for this gap are rarely the ones who designed the system. They are the staff who raised concerns through the proper channels and spent months documenting problems, only to discover that the channels were built to receive information and the documentation was designed to make acting on it so procedurally expensive that the rational choice was to absorb the failure. They are the residents of a tower block whose safety was nominally the responsibility of multiple bodies and operationally the responsibility of none.
Accountability arrangements in organisations only work when certain conditions are met simultaneously. A named person holds the obligation and the authority to act on it, and a consequence lands on that person if they do not. Remove any one of those and what remains is the vocabulary of accountability without the operating reality. The word stays on the poster. The enforcement leaves the building.
Organisations that talk about accountability the most have usually spent the longest dismantling the arrangements that produce it. The regime that killed 72 people at Grenfell Tower had never stopped talking about safety. It had stopped enforcing it. The distance between those two things is where the cost lands, and it never lands on the people who designed the gap.