How does the government spend its money? Not very carefully...

mes+

My account 

Com

The unlikely villain of the Horizon scandal? Tony Benn

The Post Office scandal goes back to procurement blunders perpetrated in the 1960s. We’re still making the same mistakes

Robert Colvile

Sunday January 21 2024, 12.01am, The Sunday Times

Share

There are many villains in the Horizon scandal. But one of the crucial figures has mostly escaped public censure — and even attention. I am referring, of course, to Tony Benn.

In the 1960s the Wilson government had a problem. No one wanted to buy the products Britain was making. So it decided to form giant new national champions that could conquer the export markets, by merging a host of existing companies at gunpoint. This was the process that gave us British Leyland. And, thanks to Benn, then minister for technology, it also gave us Britain’s answer to IBM — International Computers Ltd, or ICL.

Over the years, ICL had some successes. But it had more failures. In 1981, crippled by the expense of trying to keep up with IBM and others, it axed its main R&D programme and bought in better tech from Fujitsu. Eventually, the Japanese took over completely.

What does this have to do with Horizon? Well, Benn didn’t just shower ICL with grants. He gave it first refusal on government contracts. Even into the 1980s, any computer above a certain size had to come from ICL. Which was a problem, because ICL’s computers weren’t very good. “It would be fair to say,” recalled Sir Peter Gershon, who worked at the firm, “that most of the customers were not very happy.” Not least because the products those public servants were forced to use, as he euphemistically recalled, “had quite a lot of technical challenges”.

ICL wasn’t good enough to beat IBM. So it made most of its money from the British government. But that left the government with computers that weren’t good enough. In 1981, when ICL was facing bankruptcy, Margaret Thatcher wanted to let it die. But, says Gershon, “some very very brave civil servants … stood up and basically persuaded her, and Keith Joseph, that … if ICL went bust, the government would grind to a halt.” ICL was, like the banks in 2008, too big to fail. So a rescue was arranged. The tie-up with Fujitsu was critical to that process, because ICL’s customers actually trusted the Japanese tech.

ICL lost its monopoly status, but throughout its complicated corporate history it remained deeply embedded in the British state, which continued to provide the lion’s share of its revenue. In the 1980s it got the contract for the state pension database. So it was the natural choice when, in 1995, the government decided to automate the entire benefits system.

That project, called Pathway, was awarded jointly by the Department of Social Security and the Post Office. It was a disaster, and in 1999 it was abandoned with nine-figure write-offs. But it was agreed that ICL (by then fully owned by Fujitsu, which knew a captive market when it saw one) would get another massive contract: to use the Pathway technology to computerise the Post Office branch network. The project’s name? Horizon.

An anonymous member of the Horizon team, speaking to Computer Weekly in 2021, said that “everybody in the building … knew [the software] was a bag of shit. It had gone through the test labs God knows how many times, and the testers were raising bugs by the thousand.” But according to evidence submitted to the official inquiry, Fujitsu executives told the Blair government that if it didn’t sign the deal, ICL would collapse, with catastrophic implications for the services it was already running. As in 1981, it was too big to fail. And the sub-postmasters paid the price.

This is an incredibly depressing story. But it teaches all manner of important lessons. Beware of politicians who think they know more about industry than industry does. Beware of “national champions”. Beware of monopolies, both public and private. Beware of the way that bad decisions cascade through the generations, not least because the people who built the shoddy software are often the only ones who know how to maintain it.

But the ICL/Fujitsu story also reminds us of something else: how very bad the British state is at spending money. In 2020 the procurement expert Peter Smith published a book called Bad Buying. A disproportionate number of his examples involved the British state: HS2, PFI, Blair’s NHS IT programme, the lease rip-off that left a Suffolk primary school owing £500,000 for 125 laptops.

But the anecdotes were even worse. Smith had been appointed commercial director of a huge government project that involved paying a large consultancy firm more than £600,000 a week. He asked to talk to the person managing the contract. There wasn’t one. Whitehall officials “didn’t think it was worth it” to pay someone £30,000 to £40,000 a year to oversee a deal worth a thousand times that. So Smith looked at the invoices himself. He found that the 100 “full-time” staff the contractor was providing included the partners’ personal assistants. And that the firm had charged full whack for the day of its Christmas lunch. “An unfortunate error,” the consultants said.

Our procurement systems were designed to prevent officials from just giving contracts to their mates. But the result has been a process that too often revolved around ticking the right boxes and offering the lowest headline price, rather than actual performance.

David Halpern, founder of the government’s Behavioural Insights Team (aka the Nudge Unit), tells a Fujitsu-related horror story. The company built a system for No 10 and the Cabinet Office to work on highly classified material. But it was a living nightmare: just booting it up took half an hour. The minister in charge got so fed up he refused to use it.

Eventually the system was dumped — prompting Fujitsu to sue for breach of contract. But, even as the Cabinet Office was decommissioning the software, another branch of government was buying the same product. How on earth could that happen? Well, partly because of EU rules, government procurement policies banned departments from telling one another about their experiences with a given supplier or system, lest it should prejudice future buying decisions or disadvantage new suppliers.

The results could be absurd. In 2017 Halpern’s team asked a sample of public sector buyers to rate their suppliers on a five-star scale. But when they showed the minister the results — which did not exactly match up with the number of contracts those firms received — there was a heated debate over whether he could legally see the company names. Because if he knew who was doing a good or a bad job, it might inform decisions on future contracts. And everyone would be sued.

Which answers the question, asked by many, many people, of why Fujitsu kept getting contracts after Horizon. The government was legally prevented from taking Horizon into account.

Thanks to the efforts of various Whitehall reformers, this flaw is being addressed. The Procurement Act, passed last year, makes it easier for past performance to be considered. But Britain’s poor public procurement is ripping all of us off. We need to give it a hell of a lot more attention.

Market failure - one that you can help reduce

 And if you ask me nicely I'll buy the book for our Econ library:

BOOK REVIEW

The real cost of  the world’s rubbish

A new book reveals the dirty truth about what we chuck in the bin, says Stuart Watkins. What is to be done?

 

Every day, we throw our rubbish away. But as Rachel Salvidge notes in the Literary Review, there is no such thing as “away”. Our rubbish is moved from one spot on Earth to another, where it is either buried or transformed, whether through burning or recycling. “One person’s ‘away’ is another’s ‘here’.” 

And where does it go? In his book Wasteland, Oliver Franklin-Wallis lets us in on “the dirty truth”. He “journeys to some of humanity’s least fragrant locations to literally wade among our rubbish, discovering how it got there and exploring its  effects on the lives of the people charged with  dealing with it”.

His book is “heavy on facts, many of them interesting and sobering”, says The Economist. Twenty thousand plastic bottles are sold around the world every second. The world produced two billion tonnes of solid waste in 2016, an amount that will rise to 3.3 billion tonnes by 2050. An Indian landfill Franklin-Wallis visits is piled almost as high as the Qutub Minar, a well-known minaret in Delhi. 

He also visits towering mountains of toxic waste from mines, defunct nuclear power plants and rivers in Africa choked with microparticles from discarded clothing. Our rubbish, as the author says, is often deposited “on the margins, and on the marginalised”: shipped by wealthier countries to poorer ones, where the poor pick over it to scrape a living. 

When we think of waste management we tend to think of plastic bottles and discarded fast-fashion clothes, but fixing the “vast and seemingly intractable problem” of consumers’ rubbish would be a simple task compared with taking care of commercial and industrial waste, says Salvidge. The scale of that problem is unimaginable, not least because quantities are not systematically recorded. 

“IN THE UK 48% OF RUBBISH WAS BURNED IN 2021, A 435% RISE IN 20 YEARS. ”

In 1987, the US Environmental Protection Agency estimated that 97% of all waste is produced by industries, not households, Franklin-Wallis tells us. That figure is “outdated, vague and impossible to verify”. Nevertheless, it is clear that the volumes of waste produced are gargantuan. Canada alone,  we are told, produces 35.5 million tonnes of  household waste and 1.12 billion tonnes of industrial waste per year. Whatever the global figure is, it is  clear that “mankind is wreaking awesome damage” on the planet. 

A GLOBAL TREATY ON PLASTIC POLLUTION

It all makes for pretty uncomfortable reading, says Chris Stokel-Walker in New Scientist. And there’s  not much in the way of a happy ending. Solutions are hard to find. Attempts are being made to improve waste and recycling laws, and 175 nations have agreed to develop a global treaty on plastic pollution, but  such efforts move at a snail’s pace while the rubbish keeps piling up. People wash their yogurt containers and tear the plastic windows off cardboard sandwich boxes so they can be recycled, yet much of it ends up in landfill anyway. 

And the modern economy upon which we rely is itself “built on trash”. A third of what we throw  away is less than 12 months old. And the people  and places that take it off our hands have often built whole microeconomies on the back of it. 

There may be no easy answers, but Franklin-Wallis does provide a useful “public service” by fact-checking dubious claims, says Rose George in The Spectator. Only a tenth of the donations of old clothes that charity shops receive are sold through their outlets, for example; the rest, much of which is “unusable, stained tat”, is sent overseas, whether it’s wanted or not (countries that have tried to stop the flow have faced US “bullying” in the form of trade sanctions). 

We’re told to buy T-shirts made of recycled plastic, but clothing made from blends of that and organic fabrics are typically less recyclable that those made from a single fibre. The most common label put on our packaging indicates that it is not recyclable, and in the UK 48% of our rubbish was burned in 2021, an increase of 435% in 20 years. No one has figured out what to do with plastic. Companies continually make promises that they fail to keep.

So what is to be done? “It’s complicated” is a refrain throughout the book, as Salvidge notes. But the ultimate and most powerful action, according to Franklin-Wallis, is “laughably simple: buy less stuff”. Clamping down on corporate “greenwashing” would help too. 

Nevertheless, there’s also a case for “tighter, smarter regulation”, says the Financial Times. “Individual litterbugs” have for too long been blamed for a “packaging waste problem caused by companies that have successfully dodged full responsibility after decades of lavishly funded lobbying effort”. Franklin-Wallis has at least succeeded in outlining the size of the challenge, says The Economist, and his book “should prompt serious discussion in boardrooms and parliaments”.

However, Wasteland is worth reading for more than such practical consideration, or for the mountains of grim facts. “Waste is monstrous to look at because it is a mirror,” says Franklin-Wallis. That’s what makes his book “so captivating”, says Salvidge. “It is an unflinching account of the best and worst of us, related through the things we choose to discard.”

Wasteland: The Dirty Truth About What We Throw Away, Where It Goes, and Why It Matters
Oliver Franklin-Wallis
Simon & Schuster UK, £20

Government spending and waste/inefficiency

You know how, in the early days of your Economics course, Fiscal Policy is taught as "when the government spends it creates jobs and produces significant gains"? While I am not advocating you drop that view and become cynical of government (that is for you to develop - or not - as you mature), I do advise you to be able to evaluate the value of some government spending. FREER is a new part of the IEA, and it has published a report into inefficiency in government departments and spending projects. There are rumours that not only is Crossrail late and over-budget, it also has issues with the platforms being too low for the trains; this is something the French encountered after spending nearly £4.5bn on trains that now cannot be used. Here is a taster (from the report) of commons committees questioning some of the "culprits":

The Charge Sheet

“The Public Accounts Committee continues to have serious concerns about the cost and delivery of the Emergency Services Network (ESN) […including] the delivery timetable, potential overrun costs, and the absence of detailed contingency plans.” Public Account Committee preamble, www.parliament.uk

“It is disappointing to us to see a programme that at first seemed so promising unravel so quickly and fall victim to the same project management issues that we see so frequently across Whitehall.” Public Accounts Committee report: ‘Crossrail: Progress Review’, 3 April 2019, p.4

“Just so I am clear […the Disclosure and Barring Service] took a service […] ‘modernised’ it and productivity dropped off a cliff by a half. It has now got a little better, and we are about to go through a total reprocurement to use the same system, which you are concerned is architecturally flawed, in order to get us back, potentially to close to where it was before it was touched. Is that the summary?”

 “Yes, because that is the sensible way to go about it.” 
Exchange between Public Accounts Committee and the Head of the Disclosure and Barring Service reviewing their flawed ‘Modernisation’ Programme which cost £220 million more than expected, 11 March 2019

“If the project is too complex to provide a name as to where the failure is, why is the project not too complicated to provide bonuses to those at the top when they haven’t delivered?” 

“I don’t think you are comparing apples with apples in that regard.” 
Exchange between Public Accounts Committee and the former Chair of Crossrail reviewing their £2.8 billion cost overrun, 15 May 2019

“Do you think, as the head of the civil service, that we have the wrong people in the wrong jobs? Do we have a lot of very senior civil servants who are interested in policy and not at all interested in project management?” 

The Executive Summary (to read more click this link):

The UK public sector currently struggles to provide IT and project management. From NHS Records to the Emergency Services Network, and from the Disclosure and Barring Service to e-borders, the Government has been responsible for a litany of project management, delivery, and implementation failures stretching back over many years. On their own, these failures occasionally make the news — buried after the latest political intrigue and guaranteed to elicit no more than an eye roll from most.
Yet each one of these failures matters: they represent a forgone opportunity for change, a promise to the public not being realised, or spending not achieving what it was intended to do.

Indeed, government institutions seem rather flat-footed when it comes to project management. Repetitive failure matters more when, for much of the last decade, the Government has been making the case for spending restraint. When your pounds need to stretch even further than before, the ones you actually spend should be done so wisely. And, given that the political narrative has recently (regrettably) swung towards raw spending totals and inputs, the focus on how we spend, or what we achieve, has been lost at the very top. Prime Minister’s Question Time has recently felt like an arms race, characterised by evermore inflated amounts to spend, and little focus on how we spend it.

And that is why the travails of public-sector project delivery are so frustrating. In the last couple of years, the select committee I sit on (the one charged with overseeing that value-for-money test), Public Accounts, has dealt with nine separate instances of project failures, costing the taxpayer over £7.5 billion pounds and collectively causing 34 years of project delays. That’s £7.5 billion of money that has effectively been written off. Or, in crude terms, money that could have funded a penny off corporation tax for a year, the building of 21 new hospitals, the funding of 17,000 new police officers for the next ten years, or the ability to plug the socialcare gap for the remainder of this parliament.

 Of course, project failure is nothing new. The public are relatively inured to poor project delivery precisely because they have seen decades of it before. And, whilst those at the most senior level of politics have failed to focus on value for money, there has been a valuable continuing effort further down to build on the greater commercial focus first applied by Francis Maude in the 2010–2015 parliament. Credit should go to those people for some successes. This paper neither seeks perfection nor full success in project delivery. Yet, we need a renewed approach to the reduction of project failure in the public sector. We need to decrease inefficiency, to reform public-sector delivery mechanisms properly, and to shift the focus back on to how we spend money, rather than just how much.

“[…] Well, my background is not in programme and project management. I would not say every permanent secretary has that […]”
Exchange between Public Accounts Committee and the Head of the Civil Service, 1 April 2019

“Neither the Army nor Capita tested the fundamental changes to the recruitment approach prior to its introduction.” National Audit Office Report: ‘Investigation into the British Army Recruiting Partnering Project’, 14 December 2018

“We also found no clear or shared understanding of what constitutes value for money in nuclear decommissioning.” National Audit Office Report: ‘The Nuclear Decommissioning Authority: progress with reducing risk at Sellafield’, 20 June 2018

Short video explaining where policy has NOT worked:

You may or may not agree, but you do need to know:

Another interesting news day for fiscal policy/supply-side topic

On Radio 4 this am a commentator interviewed to talk about the Chancellor's fiscal rules proposals, while broadly agreeing the new rules are good, made the comment that "with the crumbling state of our infrastructure, and with the government's ability to borrow at 0% in real terms over 50 years, [can it?] I wish they would borrow to invest in the future."

Some time later there was a news item describing how the Public Accounts Committee [what's that?] chaired by Margaret Hodge had questioned the value for money of the whole HS2 project, pointing out that the estimated cost of the the first stage had already risen from £17bn to £21bn over the last few months.

In addition I note a headline in the Daily Telegraph:

Whitehall's quickfire £1bn spree to spend surplus in foreign aid budget

                       

"MPs say the fact that taxpayer funds were spent so quickly raised serious questions about whether value for money was achieved..."

Anyone who wants to achieve a B or higher should be all over this sort of material; it is the second most effective way of showing appropriate knowledge to the examiner. What is the first?