Category / PFI

Responsibility and management in party conference season October 8, 2012 at 6:15 am

I’ve been writing a complicated piece of code over the weekend. It needs to read an excel spreadsheet, so I started in Visual Basic, a choice I am now regretting. The problem isn’t easy, there’s quite a bit of data, and the algorithm is naturally expressed in a way that’s quite inefficient to implement without pointers*. So… it has been a headache. I have some results now, which seem likely to be right, and I’ll spend quite a bit of tomorrow trying various pathological cases. Still, I might have messed up and if I am really unlucky someone else will spot this.


All of this gives me profound sympathy for the civil servants who messed up the analysis on the West Coast main line franchise analysis. It’s really easy to make an error in a complex spreadsheet. For me, there are two main points here. The first is best put by Not the treasury view:

Ministers had months in the run up to the franchise award in August, and two months since, to require DFT [Department for Transport] senior management to explain to them – not with pages of numbers, but with convincing analysis – why this view, now apparently vindicated, was wrong. No remotely competent Minister would accept the explanation “That’s what the model says” on an issue like this. So either they didn’t ask the right questions, or they were incapable of understanding that they were getting the wrong answers. Neither interpretation reflects well.

I guess it would be too much to expect a minister to behave honourably during his opponent’s party conference and a week before his own.

The second and more important point is that this farrago would not have happened without privatisation. Privatisation doesn’t work not just because it provides returns to shareholders when things go well and little risk when things go badly – but also because it is impossible to write a contract that adequately ties down what it means to run a railway well for fifteen years, and to evaluate bids on that contract accurately. Sure, you can have a go at writing a model which helps you to understand, if you are lucky, what the bidders are proposing. But what you can’t do is think of every contingency. If you want to model a railway, run one: the real thing is the only reliable model. So why not close the evaluating contracts department at the DFT, renationalise the railways, and put the resources you freed up to work actually running the darn things? It would be simpler than what they are trying to do at the moment.

*Yes, I know you can fake it with classes. Or with the Win API. No, that doesn’t help much.

When can private delivery of public services work? July 18, 2012 at 10:23 am

In the aftermath of the G4S humiliating shambles, this question needs to be asked. It seems to me that there are at least three preconditions that have to be met before it makes sense for the government to pay a private company for a public service:

  • They can say what they want. That is, it is possible to write a contract that define what the service is, and this is not likely to change. In most PFI contracts, this is not the case, as the deliverable changes too much: no company would agree to a sufficiently flexible contract, and no government should sign up to an inflexible one.
  • The efficiency savings gained from a private provider are bigger than the cost of having to provide a profit to shareholders. As a letter in the Guardian today reminds us ‘When private companies deliver public services, ultimately they answer to thirsty shareholders, not the taxpayer. It’s inescapable: profit maximisation is a legal obligation for such companies.’ Given the estimated £300B cost of PFI, I think we can safely say that this criteria was not met in the majority of agreements signed so far.
  • There is alignment of incentives. G4S had an incentive to hire cheap ‘security’ guards for the least time possible, and give them little training. The government had an incentive to have good security – and it faces enormous costs and embarrassment at not having it. As so often in outsourcing, if G4S had done a good job it would have reaped the rewards, but the risk of it doing a bad job mainly lies with the government. Heads I win, tails you lose is not a sound principle for public policy.

It is bizarre that there are few political voices setting out these concerns and demanding that outsourcing of public services provides clear value for money. As
Seumas Milne said yesterday after G4S, what’s needed is a political sea change.

When delegation fails May 31, 2012 at 7:21 am

When you pay someone else to do something for you, the problem of what you are paying for, exactly, sometimes comes up. This is one of the issues with PFI, for instance; it seems to be essentially impossible to write a contract that defines what it means to run a hospital or a railway properly. So either the taxpayer ends up paying far too much to the outsource provider for contractual variations, or they don’t get the service they thought they were buying. Or both.

It is interesting to read that the same issue occurs in IT outsourcing. The Guardian has a series of interviews with various (anonymous) individuals from different areas of banking. This weeks’ is from an IT sales guy. He says:

Years ago management in major banks and corporations decided that they could outsource vital IT functions to companies such as IBM, Tata, HP and Atos Origin T-Systems. The idea was that if you describe the processes you require adequately, it’s safe to delegate their execution to outsiders. But the first contract goes to IBM, two years later a contract for another part of the infrastructure is awarded to HP, then Cisco gets to manage the network … Now, who is responsible for the overall system?

Indeed. Combine that with in-house IT that sees their prestige rise the more they spend, and you have very unhelpful incentives for the wider organization. You can’t tell the outsourcer what you want because you don’t understand it yourself, and anyway your requirements change regularly. You can’t upgrade easily because of all the interdependencies within systems. And the people you are paying to manage all of this have little incentive to reduce complexity because that would involve smaller budgets and fewer staff, and budget is power in many firms. Now of course a good CIO could fix this but honestly, can you name a CIO who really understand their firm’s systems at a deep level? As the sales guy goes on:

[CIOs] are managers, skilled in office politics, not technical experts. Most CIOs rarely stay in their post more than a few years. I worked for one of the major software companies in the world. It took my boss a year and a half of begging and pleading with the secretary to get a meeting with the CIO of a major client. CEOs are worse. They are afraid of looking stupid or ignorant, and actively avoid their IT people.

That phrase ‘actively avoid their IT people’ rings all too true to me.

Transparency and model gaming February 5, 2012 at 12:14 pm

A site with a rather tacky name suggests:

One of the most common reasons I hear for not letting a model be more transparent is that, if they did that, then people would game the model. I’d like to argue that that’s exactly what they should do, and it’s not a valid argument against transparency.

Take as an example the Value-added model for teachers. I don’t think there’s any excuse for this model to be opaque: it is widely used (all of New York City public middle and high schools for example), the scores are important to teachers, especially when they are up for tenure, and the community responds to the corresponding scores for the schools by taking their kids out or putting their kids into those schools. There’s lots at stake.

Why would you not want this to be transparent? Don’t we usually like to know how to evaluate our performance on the job? I’d like to know it if being 4 minutes late to work was a big deal, or if I need to stay late on Tuesdays in order to be perceived as working hard. In other words, given that it’s high stakes it’s only fair to let people know how they are being measured and, thus, how to “improve” with respect to that measurement.

Instead of calling it “gaming the model”, we should see it as improving our scores, which, if it’s a good model, should mean being better teachers (or whatever you’re testing).

This is an interesting point. I certainly agree that is you are going to measure people on x then telling them what x is is only fair. But I would never promise that x was my only criteria for measuring a real world job, as I don’t believe we can write the specification for many activities well enough to always know that maximizing the x-score is equivalent to doing the job well.

(PFI contracts are of course a great example of this; one of the reasons that PFI is a terrible idea is that you can’t write a contract that defines what it means to run a railway well for ten years that stands up to the harsh light of events.)

Thus I would argue that the problem in the situation outlined above isn’t lack of transparency, it is using a fixed formula to evaluate something complicated and contingent. Sure, by all means say ‘these scores are important’, but leave some room for judgement and user feedback too. Humility about how much you can measure is important too.

There is also a good reason for keeping some models secret, and that is the use of proxies. Say I want to measure something but I can’t access the real data. I know that the proxy I use isn’t completely accurate – it does not have complete predictive power – but it is better than nothing. Here for instance is the FED in testimony to Congress on a feature of credit scoring models:

Results obtained with the model estimated especially for this study suggest that the credit characteristics included in credit history scoring models do not serve as substitutes, or proxies, for race, ethnicity, or sex. The analysis does suggest, however, that certain credit characteristics serve, in part, as limited proxies for age. A result of this limited proxying is that the credit scores for older individuals are slightly lower, and those of younger individuals somewhat higher, than would be the case had these credit characteristics not partially proxied for age. Analysis shows that mitigating this effect by dropping these credit characteristics from the model would come at a cost, as these credit characteristics have strong predictive power over and above their role as age proxies.

Credit scoring models are trying to get at ability and willingness to pay, but they have to use proxies, such as disposable income and prior history, to do that. Some of those proxies inadvertently measure things that you don’t want them to too, like age, but excluding them would decrease model performance.

Here, it is better that the proxies are not precisely known so that they are harder to game. The last thing you want in a credit scoring model is folks knowing how best to lie to you, especially if some of the data is hard to check. It is much better to ask for more than you need, as in psychometrics, and use the extra data as a consistency check (or just throw it away) than to tell people how your model works. Its predictive power may well decline markedly if people know how it works.

Of course, you need a regulatory framework around this so that models which try to measure, for instance, race, are banned, but that does not require model transparency. Sometimes it really is better to keep the model as a black box.

The real expenses scandal May 27, 2009 at 5:50 am

I get the sense that if I ever met George Monbiot, I’d dislike him. He often comes over as ignorant and opinionated. He’s angry a lot of the time too. But none of this makes him wrong. And in yesterday’s Guardian he was spectacularly right about PFI.

PFI allows consortiums of banks, construction and service companies to build and run our public infrastructure. Though the government maintains that this offers better value than using public money, in reality the numbers behind all PFI projects are rigged. While the government retains much of the risk, the investors keep the profits, which often run to many times the value of the schemes.

The public liability incurred so far by the private finance initiative is £215bn…

One of the consistent features of PFI is that the projects are reverse-engineered to meet the demands of corporate investors. This, for example, is how the £30m public scheme to refurbish Coventry’s two hospitals became a £410m private scheme to knock them both down and rebuild one of them – containing fewer beds and fewer doctors and nurses. The old scheme was too cheap to attract private money.

We should all be incensed by this. Yes, MP’s expenses are scandalous and symptomatic of a wider dishonesty. But if you want a case of public money being used for private gain, PFI is the biggest, most glaring example. How Gordon Brown ever got a reputation for prudence after promulgating PFI is beyond me. We should not let our anger at his other failings to distract ourselves from the failings of this rotten, shameful system. PFI encapsulates the worst of free market dogmatism without any compensating controls. There is no price discovery and little competition. Instead we have a simple transfer of public funds to private companies. PFI brings crony capitalism to the heart of public procurement, and anyone who cares that their taxes are spent efficiently should be vehmently opposed to it.

Political Futures March 27, 2009 at 3:07 pm

I don’t always agree with Michael Meacher, but this letter in the Guardian is so good, and hits the tone which is lacking in both the government and the official opposition so well, that I am going to quote it nearly in full:

We urgently need … an alternative to the prevailing Tory-New Labour orthodoxy. I would propose three central strands. It should seek to restore a social democracy which has been ripped apart by greed and an out-of-control inequality epitomised by the banks’ bonus culture. We need a solidarity tax levied on the top 5% of incomes and on the so-called non-domiciled super-rich – who use Britain but don’t pay into it – with the proceeds hypothecated to end child and pensioner poverty.

We need to redraw the boundaries between the state and the market. The market fundamentalism of the last 30 years is well and truly busted. But ending privatisation, deregulation and PFI is not enough. We need a new perspective for the state, not – as now – passive facilitator and rescuer of last resort, but actively interventionist where the public interest requires it, and strong promoter of the key social values of accountability, equity and real equality of opportunity. A robust market has an essential role, but so does the state, not only in health and education (where private markets do not belong), but in energy (a key to national security), housing (neglect of which is the biggest repository of social misery), transport (for a fully co-ordinated system), and banking (to prevent another collapse and provide reliable housing for low-income households).

We need a state which is less an intrusive snooper and more the guardian of our civil liberties. And we need a major redistribution of power: away from a top-down state to disenfranchised citizens; away from top-down industrial relations to a fair and constructive role for trade unions: and away from a top-down politics to a much more genuinely participative system of governing.

Run rat run July 27, 2008 at 8:58 pm

Labour MPs are, of course, far more concerned for themselves than either the governance of the country or their party. Here’s just some of the utter nonsense, the self-serving anti democratic idiocy they and their leaders have been up to recently.

Now, I admit it, I cheered for Tony in 1997. I drank a lot of wine and ate pasta and stayed up until Portillo was gone. But now, please, can I have a Labour party back that is actually socialist, or at least has some vague aspirations other than leaving the rich alone, lining their own pockets, and trying to find some tiny chink of public life that has not yet been infected by Thatcherite free market dogma.

Privatised Infrastructure and the sad case of Heathrow March 26, 2008 at 3:06 pm


Heathrow, like Gatwick and Stanstead, is operated by BAA. Despite the positive impression created by the new Terminal 5, most of Heathrow is in a terrible state. It’s old. It has little or no natural light. It’s cramped. And it is stuffed full of shops and nasty food outlets to keep the waiting masses spending. Compared with many large airports – Hong Kong springs immediately to mind – it’s embarrassing. Worse, the squalor and massive misery generation of Heathrow was entirely avoidable.

Stock companies have one aim: generating shareholder value. They have a legal obligation to work to that one end. They explicitly shouldn’t care about national pride, client unhappiness, or holding back the broader economy if it gets in the way of making more money for shareholders. Hence not only can you not blame BAA for Heathrow, you should expect Heathrow to be as it is given it is privately owned and lightly regulated. BAA are doing what they are supposed to do: making money for their owners, Ferrovial. (Or in BAA’s case, losing rather less money than they otherwise would.) The fact that this does not serve the countries’ best interests, nor passengers, nor airlines, is irrelevant.

This is why private ownership of public infrastructure is inherently problematic. And why we won’t be getting an airport anywhere near as good as Hong Kong’s any time soon. Now wish me well – I have to use Heathrow again on Friday.

Update. It seems BAA can’t even manage to run a brand new terminal properly. The BBC reports:

Cancelled flights and baggage delays have blighted the opening day of Heathrow’s new £4.3bn Terminal 5.

British Airways, which has sole use of the terminal, was forced to cancel 34 flights by 1400 GMT due to “teething problems” encountered in the morning.

And it carried on. On Sunday we heard from the Guardian:

British Airways warned last night that the disruption that has plagued Heathrow’s Terminal Five could run through this week as it emerged that 15,000 bags have yet to be returned to their owners. This figure is three times higher than BA had originally admitted to. Following three days of cancellations which saw more than 150 flights grounded and thousands of passengers disrupted, the airline, which has exclusive use of T5, said that a further 37 flights would not take off today.

Tanking PFI January 28, 2008 at 7:39 am

I have blogged before on the idiocies of PFI – how Gordon Brown has a reputation for economic prudence given he facilitated this is beyond my ken – but the latest news is even more extraordinary. Not only are we paying more for our infrastructure than we should, but now the very future of an important project, the Airbus tanker project, is under threat. According to the FT:

The turmoil in credit markets has dealt a big blow to the UK government’s defence procurement programme, putting in jeopardy plans to help fund a new fleet of Airbus tankers for the Royal Air Force with a bond issue.

The issue is that the bonds financing this project were expected to be wrapped by Ambac. This is beyond crazy: in many instances either explicitly or implicitly (because the project cannot be allowed to fail) the government backstops PFI, so there is no need for a wrap. It is only the bizarre fiction of PFI that pretends that there is any risk transfer, and hence any need for an insurer to be involved. PFI is a waste of public funds on an extraordinary scale: it is time we held Gordon accountable for it.

Mr. Bean as the Fat Controller January 6, 2008 at 9:06 am

The key question with any private sector involvement in public projects is whether the extra return shareholders rightly demand is compensated by increases in efficiency. Often the answer to that is no, which is why PFI makes so little sense – especially when you remember that there is rather little risk transfer to the private sector in many PFI projects. The latest rail debacle over the west coast main line is certainly evidence that the current mess of operating companies, National Rail, and private infrastructure contractors makes no sense — our rail policy is fatally flawed. Let’s renationalise the lot before our railway turns into a historical relic.

What works replaced by what we believe December 28, 2007 at 8:46 am

The following quote comes from a post on the Guardian website:

What I find most striking about this, as about other items in this government’s moralistic agenda, is how opposing arguments simply are not heard. It isn’t just closed mindedness, the government and its supporters are in the grip of a kind of exclusivist belief system akin to a fundamentalist religion. Arguments based on individual choice, or the notion of adults making rational decisions, simply “do not compute”; they are here to protect us, like spoilt children, from the bad world out there, and they have our interests at heart, and if you don’t agree then clearly you favour exploitation and slavery and oppression.

The two things that strike me about this quote are firstly how accurate it is, and secondly how many things it might be referring to: Europe; Iraq; PFI; Pensions; Trident; Nuclear Power; 42 day detention; the DNA database; ID cards. Pretty much any part of politics that is in any way controversial in fact. Even if you agree with a particular policy, the monological belief system which nourished it is deeply troubling. No government which scorns alternative views and abuses legislative privilege with deeply partisan, ill-thought out and costly nonsense as Brown’s does deserves to survive, let alone win re-election.

The Death of Metronet July 19, 2007 at 9:15 pm

The collapse of Metronet and the attendant comment suggests it might be worth saying something about risks in equity and debt finance.

The classic model of a company is that it can raise money in two ways: by issueing shares, or by borrowing money. The best thing that can happen to a lender is that they get paid back: there is no upside to making a loan other than receiving your interest. Therefore lenders demand that they get paid before shareholders – if a company has funds, it must pay its creditors first, and only after that do shareholders receive a dividend. Moreover if a company cannot pay its debts, debt holders can typically seize the company and sell or liquidate it to get (some of) what they are owed.

In exchange for taking the risk that they might not receive anything, shareholders collectively own the company and so receive anything that is left after debt holders have been paid. This is sometimes called a residual interest: it is potentially high risk and high reward position, since if the company’s earnings can’t support its debt, they might get nothing; but if earnings are high, then there is a lot left for them after interest costs.

One key capitalist idea, then, is that shareholders take risk in exchange for potentially big rewards. If you don’t like the idea that you might lose everything, don’t buy equity.

Now consider a company in a long term PFI contract such as Metronet. The rewards are potentially very high, as the profits of someone like Amey indicate. But so are the risks, as we see with Metronet or the late and unlamented Railtrack. Is this a good deal for the tax payer?

There are two arguments for PFI: a specious one; and a sensible one. The specious one is that PFI gets infrastructure built without the government having to raise money. But that is just accounting trickery: the tax payer pays eventually, after all. A PFI contract is an ongoing liability just as a bond issued by the government to pay for maintaining the underground without a PFI contract would be.

A private firm doing government work will demand an extra return over and above what they think it will cost in order to pay their shareholders: they are right to do that as their shareholders are taking risk. So on average PFI contracts are more expensive than their fair cost. Are they still the best that is available? That depends on whether the PFI contractor can get the job done and make their profit for less than it would cost the government to do the job themselves. There is a natural assumption that private enterprise is more efficient than government – that may even be true – but is it so efficient it makes up for the extra profit shareholders demand?

Here’s a view from the capital of London financing, Canary Wharf, to end.

Tories are heirs to Blair May 30, 2007 at 11:56 am

From The Guardian:

The shadow chancellor, George Osborne, will say today that the Conservative party is the heir to Tony Blair’s reforms of hospitals and schools, not Gordon Brown.

(A rather better monument than Tony is likely to get from his heirs, taken in a church in Milan.)

First reaction: yes, Tony’s policies are not policies of the left. Second reaction: in saying that, does he imply that being Tony’s heir is a good thing? I just wish I believed El Gordo agreed with this statement too…

What works? You wish… December 5, 2006 at 8:22 pm

‘What matters is what works’ Tony Blair once declared. I really believe that is true: sadly Tony doesn’t, and I suspect never did. As Ann Clwyd is ejected from her position as Chairperson of the parlimentary labour party thanks to her support for Blair, it is worth highlighting five of the worse examples of Labour’s doctrine pursuit of policies despite them obviously not working:

  • Iraq. No more need be said than that one word.
  • PFI. Wow, wouldn’t it be really great to pay more for the same public services and lock the NHS, Transport for London and many other bodies into inflexible, long term contracts? Anyone who claims Gordon Brown is an economic genius should be forced to have ‘PFI’ branded on their forehead so they never forget the disasterous waste caused by this terrible idea.
  • Faith Schools. We want a more tolerant society with higher standards of education. So let’s set up schools which exclude some faiths, enhance the sense of being special of others, suck up education resources that could be spent on improving ordinary schools, and give their governance to those verging on monomania.
  • Replacing Trident. We do not need to spend £25B on nuclear weapons now, honestly we don’t. This is all about Labour appearing tough on defense: it has nothing to with what the country actually needs.
  • The Internal Market in the NHS. No one who had actually spent any time trading the financial markets would ever suggest they are efficient allocators of resources so why on earth does the government think the market will magically solve the problems of the NHS?

None of these policies was obviously unworkable at the beginning: foolish or misguided perhaps, but not obviously unworkable. They all subsequently have been proved not to work. Yet still the government sticks with them, demonstrating the dominance of ideology over efficacy in their thinking.