Winner of the New Statesman SPERI Prize in Political Economy 2016

Thursday 18 January 2018

What Carillion tells us about public sector outsourcing

Jeremy Warner, an editor at the Telegraph, once said there are either big state people or small state people. I felt the same way following the reaction to the collapse of Carillion: there are either private good, public bad people or public good, private bad people. Of course, reality is somewhere in between.

Corillion went bust because of cost overruns or delays in three large construction projects. The nature of such projects involve that kind of risk, but clearly the company - despite its size - was not resilient enough to withstand those failures. It did not go bust because of privatisation of public services, unless you think the government should build its own hospitals or roads. If anything, it shows that those contracting out public contracts were getting a good deal.

There will always be public projects contracted out to the private sector. Much of the increase in public investment planned by Labour if it wins the next election will be undertaken by private firms. Getting the contracting relationship right is difficult and fraught with dangers.

The government clearly has questions to answer about why it continued to award contracts after the profit warning, and we need some informed analysis to determine whether the government, as they claim, had fully protected all but one (!) of these projects and will not lose any money as a result of the collapse. As David Allen Green suggests, this smacks of ministerial failure. (The link also shows public procurement can have its funny side.) Also why was the position of the “crown representative” who was meant to be overseeing scrutiny of, among others, Carillion left vacant? The government should also ask whether companies should be allowed to pay large dividends when their own pension fund is underfunded. And why Carillion's auditors, KPMG, gave it a clean bill of health when its balance sheet was already showing signs of stress
To see what lessons the collapse of Corillion does have for the debate over whether the public sector should privatise certain of its activities or do them in house, we need to go through some of the pros and cons

There is one main benefit of contracting out public services, which is that it can save money. To mention ‘the market’ here is not very helpful, because with one buyer and only a few sellers for something (the contract) agreed once every few years, this is hardly a normal market. [1] It is instead about the incentives faced by managers and workers, both in achieving efficiency and fostering innovation. Managers have a clearer incentive system in a private sector firm to maximise profits, and that incentive is provided by the need to bid low to win the contract and nevertheless make a profit. As Carillion shows, margins on most public sector outsourcing are not large. In that sense Carillion confirms that part of this mechanism is working. A single public sector entity cannot replicate this advantage, unless it too is in competition with private sector firms. In short, competition improves incentives.

One important qualification to this argument involves information. The temptation of a bidding system based on the lowest price is to cut quality. So the public sector has to have a clear means of not just specifying quality in the contract, but of ensuring the contract is being fulfilled once it is awarded. Sometimes politics can get in the way of that happening. For activities where quality is difficult to observe, contracting out is not a good idea.

Another qualification involves the attitude of public sector workers before privatisation. If they, for whatever reason, internalise the need for efficiency and innovation, because for example they can see how both improve the outcome for customers, then contracting out to the private sector will achieve little. The NHS could be a case in point.

A further problem with privatisation is finance. When people argue that public money should not be wasted paying the shareholders or creditors of private firms, they are both right and wrong. They are wrong in the sense that without contracting out the same amount of money has to be raised by the public sector, and so it “wastes” money by having to pay interest on government debt. But they are right in that the rate of interest on government debt is much less than the rate of interest a private firm has to pay on any debt, or in the form of dividends to shareholders. The reason for this is that investors do not like risk: people who lend to the UK government know they will always get their money back, while as the shareholders and creditors to Carillion have just found this is not true for private sector firms.

This is why PFI projects undertaken just so that the borrowing is done by the private rather than the public sector are costly from an economic point of view. It is why it makes sense to exclude public investment from any fiscal rule: fiscal rules that restrict public investment are an open invitation to politicians to undertake PFI type financing. In my view the best constraint on public investment is the expected social return, assessed with the help of an independent body. It is often said that PFI type projects ‘avoids risk to the taxpayer’. Again this is the wrong way round. It is far easier and cheaper for the public sector to take risks than the private sector, so PFI projects are paying far too high a price to avoid risk to the public sector. 

Another problem related to risk is the interrelationship between what the private company contracts to do and what actually happens when government forecasts go wrong, as they always will. This may have happened with the East Coast line “bailout” (but if it was, we should be told), and it did happen with privatising the probation service. Public sector contracting out forces each side to commit to guesses about the future, whereas if everything remains in-house there can be much more flexibility. There is also the cost of having to train more civil servants in the art of writing good contracts.

One further problem that Carillion reminds us of is that privatisation runs the risk of a degree of interruption if the company goes bankrupt. Disruption is nothing new. If privatisation is to have any benefits, the contract from the public sector has to come up for renewal every few years, and if the private sector provider is changed that will involve some dislocation of service.

One final point, which is contingent on what I hope will be a temporary state of affairs. Nowadays the management overheads for private sector firms are likely to be far higher than in the public sector, for reasons that have little to do with management quality. Ben Chu sets out how much management was being paid at Carillion compared to equivalent public sector managers. And what on earth were shareholders doing allowing the directors to relax clawback conditions on management’s pay if things went wrong, which even the Institute of Directors described as “highly inappropriate” and “lacking effective governance”. In truth the public sector is much better at stopping managers using their monopoly power to be paid over the odds than the private sector appears to be.

So the economist’s answer on public sector outsourcing is, it depends: on all the factors outlined above and probably more I have momentarily forgotten. (Like economies of scale and expertise: no one would ever suggest the public sector makes its own paperclips.) Where the balance will be is bound to be case dependent. But it would be incredibly surprising if at least some of the outsourcing undertaken by this government was not ideological rather than evidence based. This suggests that Labour, if it wins the next election, should undertake a thorough independent review when it has all the facts at its disposal. That at least might ease fears that we will lurch from one ideological position to its opposite.

[1] This is an interesting example of a longstanding debate with myself. If you want to claim that much of this kind of outsourcing represents neoliberal ideology at work (which it probably does), and also that neoliberalism is all about the market, then your definition of a market has to be pretty wide. But of course a large firm like Carillion, as Ronald Coase said, involves the large scale supersession of the price mechanism. By which he meant that firms are an alternative to markets, and large firms suppress what would be market activity if it was replaced by lots of smaller firms buying and selling to each other. This is a contradiction at the heart of neoliberalism as market worship: firms are alternatives to markets. 


  1. This post lists a lot of things that have gone wrong with Carillion and highlights decisions that are questionable. One advantage of outsourcing to private companies is the fact that we have clear sight of many aspects of governance; auditors and accountants reports, minutes of board meetings, details of how much board members were being paid and their bonus schemes, hence we can ask these questions. This illustrates the primary purpose of the Conservative Party; not as many claim to act on behalf of tycoons and corrupt business leaders, but to put the profit motive in a clear political and moral framework so that corruption and poor performance is brought to light and can be addressed without favour whilst retaining the innovation and cost benefits private business activity produces.

    Bringing in public spending into public institutions can create clear conflicts of interests; take the NHS where the government is both purchaser and provider, and has to play both sides of that line. One specific advantage of outsourcing is that it makes clear in cases such as this where the Government's responsibilities lie, which is with the people.

    1. If you think the conservative party's motive really is to "bring to light and address corruption and poor performance", I have a bridge to sell you.

    2. Your post may express the aspirations of outsourcing, but the reality is very different and corruption clearly plays a part, you should have worked where I did and listen to the comments about cars and holidays that suddenly become apparent.

      I in fact placed an FOI on our local CCG about the costs of the outsourcing body for the CCG, in that particular year the costs of that Unit alone came to 9% of the total overhead costs of the CCG. In round numbers our CCG costs £50 million for a smallish City, that is before even one person has been treated locally. If you think that is value for money or even efficient I would beg to differ.

      We also had meetings with members of the local ambulance service who wanted us to act on their behalf as whistleblowers, these wonderfully clever legal experts that draw up contracts know nothing about the front line services they outsource and so guess why things go drastically wrong and just to mention one case, where a 90 year old gentleman was left for 24 hrs before being collected for his appointment. And to list all the other situations such as vehicles in a state of disrepair putting the publics safety at risk.

      Theory is a wonderful thing but what we see as observers on the ground would make any sensible person get very angry, I even took this information to my local Tory MP, stating everything the Ambulance crews told us, and his response was, "well they are only looking after their jobs". You are today witnessing the chaos we have been highlighting since 2014. And no this is not winter chaos this is all year round, but don't wait for the BBC to tell you truth as to what is happening.

  2. "For activities where quality is difficult to observe*, contracting out is not a good idea."

    * or easy to obscure

    If you think about it, this covers a huge range of public service sectors from health, education, prisons, transport.

    Its not a case of big state vs small state, its a case of making intelligent, informed decisions on a case-by-case basis. The evidence suggests that the number of services sectors that are more efficiently run by the private sector is significantly smaller than the current number of privatised services.

  3. there's another point to make on risk and financing costs, though.

    the comparison could* look like this:

    the govt borrows at 1% but if costs overrun it has to cough up more money to finish the project

    the PFI SPV borrows 8% but if costs overrun the project is finished and its lenders are not repaid

    so the expected cost to the public sector could be the same, even if the financing cost looks higher under PFI

    * whether this is actually how the contracts are written is the big question, and I can believe it isn't

  4. If I remember my history correctly, there was no mass support for Attlee's nationalisations nor for Thatcher's privatisations.

    There is a real opportunity for an incoming Labour government to get the balance right on this issue.

  5. "Another problem related to risk is the interrelationship between what the private company contracts to do and what actually happens when government forecasts go wrong, as they always will. This may have happened with the East Coast line “bailout” (but if it was, we should be told)"

    I don't think it was a "bailout". Everybody seems to agree that VTEC overbid (hence they made losses in the first two years, which they've rightly swallowed) but the premium payments to government rise so much in the last few years, that they don't make sense unless they were promised infrastructure upgrades to have been done. Plus of course it suited the government to get the biggest figure up front to release to the press.

    The infrastructure isn't being done by the government, so VTEC could get out of the contract. Labour's response to this has really irritated me. What they should be saying is that the government cost us this money by postponing investment. A textbook example of why investment is good. Instead they've bullshitted about "a £2bn bailout".

  6. From press reports (e.g. most recently,Polly Toynbee's Guardian article at some of the cost-saving of out-sourcing comes from cutting the wages in lower paid jobs (e.g. cleaners) as compared to those in the public sector for the same jobs. I have not seen any attempt to quantify this. Similarly, often the actual level of service provided may deteriorate (e.g. see the same Toybee article, as well as the Probation Service referred to the blog).

    It is important to note that such wage-cutting or service deterioration is not "efficiency" as understood by economists. There is no saving in the resources used to achieve a given level of outcome, nor any indication that the previous level of service was too high. Instead it is simply a transfer of income from the lower paid to shareholders or higher paid directors and executives.


  7. > …it makes sense to exclude public investment from any fiscal rule: fiscal rules that restrict public investment are an open invitation to politicians to undertake PFI type financing…

    I was surprised to learn the following in yesterday’s National Audit Office report on PFI (emphasis added):

    “Most PFI debt is scored as off-balance sheet under the [EU] European system of accounts (ESA), which determines government debt levels. However, under the International Financial Reporting Standards (IFRS), used to produce departmental financial accounts and the Whole of Government Accounts, most PFI debt is on-balance sheet.” [footnote 14, page 11]

    “The Office for Budget Responsibility’s (OBR’s) July 2017 fiscal risks report cited the use of off-balance sheet vehicles like PFI as an example of a ‘fiscal illusion’. Most PFI debt finance raised to construct the asset is transparently reported to Parliament, where the debt is considered to be on-balance, via departmental financial statements and the Whole of Government Accounts (WGA). The debt is recorded as a financial liability but as noted by the OBR ‘most public and political attention, and the government’s fiscal rules, still concentrate on the National Accounts measures of PSND (Public Sector Net Debt) and PSNB (Public Sector Net Borrowing)’, which does not reflect fully PFI liabilities (see paragraph 1.14). PFI can be attractive to government as recorded levels of debt will be lower over the short to medium term (five years ahead) even if it costs significantly more over the full term of a 25–30 year contract.” [para 1.16, page 12]

    PFI and PF2, NAO, 18 January 2018

  8. I have worked in the private sector for most of my working life, and guess why I am a socialist and not a capitalist?

    This article is so full of holes and presumptions that it beggars belief, for over forty years now we have been spoon fed that the public sector is inept and the private sector knows best. Well here we are forty years later witnessing not only the melt down of the financial sector but a Tory government propping up a corrupt organisation that was clearly insuring they paid themselves money that a company could ill afford to pay.

    As an example of how modern corporations operate I will pass on some anecdotes of my experience of the one I worked in- This multi national company set up a direct sales company on the same site that I worked on, it sold a multitude of products some of which we produced on site, this direct sales operation was set up in contravention of the EU anti competitive regulations. After a number of years haggling the company was forced by the EU to shut down the whole operation, which cost about £10 million. So this international company decided it would recoup the cost of this failed exercise by sacking half the workforce locally. So a company earning profits between £1-£2 billion a year decided that it would downsize a profitable local company to recoup profits lost by its own stupidity. Forgetting that when you reduce the size of a company you reduce the output, meaning less profits into the future, later 12 to 18 months or so the same company reduced the workforce by half again, that was when I accepted early retirement and left that company.

    There are myriad of tales I could go on about, that actually happened and were not stories written third hand about inefficient management, useless contractors that leave the job when money runs out, corrupt managers going on holidays, contractors working unsafely, the list is endless, and I could go on. But then these are only anecdotal real events.

    The private sector doesn't deliver, is not cost effective, and there is an awful lot of corruption. No transparency, no accountability, and you only have to look at the rail networks to see failure at its best.

    Ever since Nicholas Ridley introduced his so called 1977 conservative research document, they have progressively dismantled the state, "selling off the State Silver", in the words of none other than Harold Macmillan.

    If you want to check the authenticity of the above document, please look in the Margaret Thatcher foundation under this reference: THCR 2-6-1-37 (177)JPG

    In addition to my comment above, the proof of the damage done to the economy by the privatisation of our public services, we only need to hear the comments and reference to the "Golden age between 1945 and 1970", More proof can be obtained by this graph from 1950s to the present day:

    None of this is rocket science, but you do need common sense to see it.

  9. Doesn't supercession of the price mechanism imply prices in general are arbitrary, more psychological than set automatically by supply and demand? Isn't markup a sign of money demand, and the private financial sector supplies money as credit manufactured out of promises to each other?

    Carillion should have raised money by lending long and borrowing short. They just needed enough bluster to get loan extensions. They could borrow at US Dollar rates and swap into Euros or Yen and back again, and make a riskless profit since covered interest parity persists negative in currency swap markets. In fact if Carillion really is just after money they should abandon physical construction and go into finance, because r > g. Then who builds roads and hospitals? Volunteers on a basic income, with material support from government. Let me maintain forest roads I want to use, with access to public equipment ...

  10. I think few people who use the term neoliberalism believe that it represents an honest call on market forces to perform social and economic good. The ideology continues to evolve, but its focus has never been on market efficiency. First it focused on removing the power of organised labour and the public in general to make claims on employers and government (largely accomplished by the Reagan/Thatcher governments), then on taking public assets into private ownership (still ongoing) and finally on turning the state into a corporate quartermaster, creating phony 'markets' often where only a handful of corporations need apply. Carillion's main skill was in winning contracts, not in providing services. Virgin's recent litigation after being refused an NHS contract demonstrates the brazenness of corporate expectations of public sector compliance. The market is an increasingly transparent fig leaf.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.