Blair Is Right About Efficacy. His Own Government Proved Why It's So Hard.

"The challenge of democracy is not transparency, honesty or conspiracy theories about the hidden power of elites. It is efficacy. It is the ability to get big things done." — Tony Blair, Tony Blair Institute for Global Change, 26 May 2026

Tony Blair is one of the political analysts of his generation, and his essay deserves engagement. His diagnosis is correct: the failure of modern democratic government is not ideology, communication, or personality. It is efficacy, the structural inability to translate policy intent into actual outcomes.

But there is an irony at the centre of this argument. The machinery most responsible for Britain's current governance paralysis — the proliferating public inquiry — was given its modern industrial form by Blair's own Lord Chancellor, Lord Falconer of Thoroton, through the Inquiries Act 2005.

The Act That Built the Inquiry Machine

The Inquiries Act 2005 was born of frustration. The Saville Inquiry into Bloody Sunday had been running for seven years and cost £192 million. Lord Falconer's solution: a single statutory framework under which ministers could establish inquiries quickly and at proportionate cost. He told the Commons it would make inquiries “speedier, more effective at finding facts and making practical recommendations, and less costly.”

What the Act actually did was dramatically lower the barriers to commissioning an inquiry. Before 2005, a statutory inquiry required Parliamentary authority. After 2005, any minister could establish one. In place of friction, Falconer had installed a conveyor belt.

Since the Inquiries Act 2005, 23 statutory inquiries have been completed under its provisions. As of early 2025, a further 17 are ongoing or yet to begin, including the UK Covid-19 Inquiry, running at more than twice the daily cost of Grenfell. Total cost of UK public inquiries over the past decade has exceeded £1 billion. A House of Lords committee reported in 2024 that the Act required "a major overhaul."

The irony is structural. An inquiry documents governance failure, expensively, then issues recommendations that enter the same broken delivery system that produced the failure. The system is self-perpetuating: failure generates inquiry; inquiry generates recommendations; recommendations are not embedded; failure recurs.

Maternity Services: A Case Study

Nowhere is this more visible or more heartbreaking than in NHS maternity services. Over the past decade, maternity care has been the subject of more major reviews than any other area of NHS practice. Each has found essentially the same things. Each has issued recommendations. None demonstrated that its recommendations were embedded before the next inquiry was commissioned.

2015 — Morecambe Bay (Kirkup): First major modern maternity inquiry. Avoidable deaths; culture, leadership and teamworking failures. Kirkup returned seven years later and found the same themes.

2016 — Better Births (NHS England): National maternity review. Recommended transformation of culture, continuity of care and multi-disciplinary teamwork.

2022 — Ockenden (Shrewsbury & Telford): Poor care contributed to the avoidable deaths of 201 babies and 9 mothers. 15 Immediate and Essential Actions issued.

2022 — Kirkup (East Kent): Failures led to deaths of 45 babies. Kirkup noted the report "is reflective of the findings from Morecambe Bay and the Ockenden report."

2024 — Birth Trauma Inquiry: Parliamentary inquiry. Themes: culture, staffing, discrimination. Largely the same as 2015.

2022 onwards — Ockenden (Nottingham): The largest maternity review in NHS history — over 2,500 cases of alleged poor care.

2025 onwards — Amos (National Investigation): Explicitly tasked with finding out why previous recommendations were not implemented nationally. Families have criticised findings for "repeating previous reports."

No mechanism exists to verify whether recommendations from any review have been implemented before the next review begins.

An Answer Beyond Left and Right

The left believes in regulation as the instrument of equity. But in Ockenden, Kirkup, Morecambe Bay, Grenfell, Mid Staffs, and Infected Blood, in every case, the regulations existed. The protection did not. The mechanism of regulation catastrophically failed the people it was designed to protect.

The right argues that regulatory burden stifles growth. Its strongest case is evidential: show which regulations demonstrably produce their intended outcomes and which do not. That case requires data (compliance rates, outcome correlations, implementation evidence) that the current system never generates. Without it, deregulation is guesswork.

What Governing in the Age of AI Actually Means

Blair calls for "reorganising the whole of government around the harnessing of the 21st-century technological revolution." AI applied to a broken governance mechanism does not fix the mechanism, it accelerates it.

The transformation AI makes possible operates at both ends. At the upstream end: a governance system that continuously measures whether existing policies are working provides an empirical foundation for policy design. At the downstream end: policy intent translated directly into operational workflow delivers the ballast that holds the course regardless of leadership changes or inspection cycles.

The left gets regulation that actually protects people. The right gets the evidence base for deregulation. What governance needs is not another inquiry. It is needs mechanisms that close the loop before the harm, not after it.