Stand and Deliver: Review of Michael Barber’s ‘How to run a government’

Jen Gold

How to Run a Government, so that citizens benefit and taxpayers don’t go crazy

Michael Barber (Penguin, March 2015)

The skills needed to win elections are very different from those required to govern effectively. In his new book Sir Michael Barber makes a compelling case for governments to pay far more attention to the science of delivery – the “set of processes that enables governments to deliver ambitious goals by learning effectively as they go, and refining as necessary.”

Barber uses the problems that recently beset US healthcare reforms as a cautionary tale. Few would deny that Barack Obama is one of the most skilled political campaigners of his generation. But the technical problems that plagued the roll out of the website were hugely damaging – and entirely avoidable. As Barber persuasively argues, from the moment the Affordable Care Act passed into law in 2010, the White House and Department of Health and Human Services should have tracked progress, identified problems, and stress-tested solutions. Instead, the full scale of the website’s problems came to light only in the autumn of 2013. This flagship reform had simply been left to flounder under a lack of proactive management.

Over the course of its eight chapters, How to Run a Government offers a blueprint for avoiding such blunders and successfully delivering campaign commitments. In particular, Barber highlights the importance of leaders establishing clear priorities and ensuring the government machine has the capacity to deliver, aided by central delivery units; emphasises the need to build resilience into government by establishing monitoring routines and developing strategies for rapid problem solving; and considers the potential for linking performance to budget decisions.

To this end, Barber lays out 57 “rules”. Above all else, he sees delivery units as the engines keeping administrations relentlessly focused on effective implementation. These small units, operating at the centre of either departments or entire governments, gather and scrutinize performance data, regularly brief leaders, and intervene when necessary.

Barber’s recommendations are based on decades of direct experience working for and advising governments, both here and overseas. Between 2001 and 2005, He set up and ran the Prime Minister’s Delivery Unit (PDMU). This model is now a global phenomenon, with similar units on six continents.

With the general election and subsequent spending review very much on the horizon, this book offers a number of timely lessons for a future government. For one, it recommends establishing a central delivery function at the outset. It took the Blair administration until its second term – after four years of slow progress on key reforms – to establish this capacity in the form of the PMDU. The coalition government was also slow on the uptake; the PMDU was abolished in 2010 but just a year later, frustrated by the lack of a means to monitor and drive key priorities, it reappeared under the guise of the Implementation Unit.

Equally, Barber notes that leadership continuity really does matter. Reshuffles and constant ministerial churn can seriously undermine delivery. Barber offers a memorable anecdote in this regard:

“Kim Howells MP held six different roles during Blair’s ten-year premiership. I saw a good deal of him when, at Transport, we were enjoying collaborating on the brave effort to ensure that the trains ran on time. We were just getting some traction when, for no reason I ever understood, he was gone again, this time to the Foreign Office to deal with the Middle East.”

Finally, Barber points to a range of trends – from open data and transparency to continued outsourcing, digital services, and privacy concerns – that will shape the trajectory of delivery reforms over the coming decade. He highlights the multiple innovations of other jurisdictions from which the UK can learn: Maryland’s commitment to transparency in performance data, Khyber Pakhtunkhwa’s use of virtual ‘dashboards’ to monitor its school system in real-time, and Malaysia’s use of cross-sector innovation labs to formulate delivery action plans – to name but a few.

Ultimately, Barber isn’t advocating a static approach. The science of delivery, he tells us, “is not a complete science and never will be.” But whatever the result in May, the need to restore fiscal balance means the next government can ill afford to ignore Barber’s advice.

Jen Gold is a Researcher at the Institute for Government and has authored a number of reports on policy implementation and delivery units, including Data-driven Delivery and International Delivery. 

This article originally appeared in the Spring 2015 edition of the Fabian Review


  1. Anu Bose

    In most developing countries, the problems are political and ethical. Sir Michael’s prescription for a ‘new’ way of looking at effective delivery of government servces ignores these harsh realities. He may say that his formula is not ‘static’. It behooves the reviewer to point out that it is.

  2. f.rahim

    Other countries may have other reasons for successfully running a program or project and likewise their failures, but in Pakistan, the single most important reason is lack of honesty or absolute dishonesty. PTI is pursuing its objectives honestly and sincerely and hopefully it will show tangible results.

  3. Mian Mujahid Ullah

    i agree that pti in kpk are governing the way they should be copied by any advanced government to govern even better.this is system which compel to deliver for anyone in the government where ever who ever no matter but to deliver or to leave the place empty for those who are ready to that the money and time of the tax payer is utilized best in the best interest of the people.

  4. Yasir Khan (KPK,Pakistan)

    There is not a rocket science required to run a good government in developing countries like Pakistan you just take care of the following departments effectively and efficiently i,e School,Hospital, Policing, Justice ,Municipality and off course corruption.

  5. faizullah khan Tareen

    development of any society requires good team,clear goals,commitment,and timely delivery -we must listen to the cries of the people whom we are going to manage

  • (will not be published)

Please read our community standards