Detention centre profits: 20% and up for the migration prison bosses

Just how much money do companies make from locking up people in the UK’s privately run immigration detention centres? Our analysis, the first to study the detention industry overall, suggests that profit rates of 20% or more are standard.

The collapse of Carillion has focused attention on the outsourcing corporations, who complain that government austerity is squeezing their once bountiful incomes. But immigration detention centres, along with prisons, remain very profitable. Of the UK’s eight long-term detention centres, seven are run by four private contractors: G4S, Serco, Mitie, and GEO Group.

Our analysis of recent accounts released by US prison profiteer GEO Group show it could be making as much as a 30% profit margin from running Scotland’s Dungavel detention centre. This comes after internal G4S documents revealed the company was making over 20% profit on its notorious Brook House deal – and over 40% on the neighbouring Tinsley House centre. (See below for full analysis of these figures.)

Why is detention so profitable?

It is certainly the case that some outsourcing contracts have been losing a lot of money. Obvious examples are the “COMPASS” contracts to run housing for asylum seekers not in detention.i G4S and Serco each have two of these deals, for different regions, and complain bitterly about them. Transport and healthcare are other areas where many have struggled – Mitie, for example, sold off all its home care business at a loss last year. Mitie’s latest annual report also notes particularly tight margins in a number of other common outsourcing areas, including cleaning and engineering maintenance. These losses will of course hit businesses’ overall results.

So why do detention contracts remain profitable? We can think of a number of reasons. One is the practice of using detainees, paid just £1 an hour, as effective slave labour. For example, GEO Group is reported to have saved over £727,000 in less than three years by paying Dungavel detainee labour below the minimum wage. Our 2014 report on detainee labour estimated the detention corporations between them could be saving £3 million a year by getting detainees to cook, clean, and maintain their own prisons.

Another is that, as there is very little scrutiny of detention contracts, contractors can cut costs further by under-staffing and stripping facilities to a minimum. As we reported in 2015, detention outsourcers are allowed to “self audit” their own performance, with minimal checking by the Home Office. Meanwhile the voices of those in detention themselves, stigmatised as “illegals” and stripped of any rights, are rarely heard.

Another reason is that these are relatively large deals with only a handful of specialist bidders (so forming an “oligopoly” who can keep prices high). There is not the same competitive pressure on margins as in, say, a general “facilities management” contract.

Also, these companies know the business very well. The very-first purpose built immigration detention centre, Harmondsworth, was run by Securicor (now part of G4S) on opening in 1970. The rash of new PFI-funded detention centres opened during the Blair government were also handed straight into private management.

Headline loss-making deals tend to be ones where outsourcing companies, seeking to keep growing their businesses in a tougher environment, push into new areas they haven’t tried before. For example, G4S and Serco came into the COMPASS deals with no experience as housing landlords. And in multi-million mega deals like COMPASS or a train line, a mistake can mean big losses indeed. Amongst the detention profiteers, Serco is particularly vulnerable as its whole £2 billion business is based on about 300 big government contracts.

In general, while many other service contracts are being squeezed in today’s austerity conditions, locking people up remains good business. So does security more generally, in a world of increasing insecurity and inequality. This is ultimately why outsourcers who focus just on security and imprisonment like G4S and GEO Group are growing and turning a healthy profit. And this is why all the outsourcers keep bidding for detention contracts, alongside promoting the private prison industry.

At a time where other government deals in sectors such as housing or transport are blowing up in corporations’ faces, locking people up is the outsourcing gift that keeps giving. Prison and immigration control industries are fuelled by insecurity, inequality, and xenophobia – and recent trends suggest the rush to lock up society’s unwanted is not going away. Or as Serco’s latest Annual Report puts it:

“we can be very confident that the world will still need prisons, will still need to manage immigration … a prison custody officer can sleep soundly in the knowledge that his or her skills will be required for years to come.”

Image: GEO Group boss George Zoley.

Analysis: up to 30% profits at Dungavel

Neither the Home Office nor the outsourcing companies publish the profits made on detention or other contracts. Such information is typically impervious to Freedom of Information requests: the public right to know is overruled by companies’ rights to “commercial confidentiality”. Last September, a senior G4S executive refused to disclose detention profits even when questioned by MPs in parliament. And accounting regulations do not require the companies – which mostly run a range of different businesses – to disclose details of individual contracts.

However, there is one case where we can get a sense of the money involved: Dungavel Immigration Removal Centre (IRC) near Glasgow. Since 2011, this has been run by the Florida-based GEO Group, the Trump-donating private prison empire which runs many of the infamous ICE detention facilities in the US. (See our full profile of GEO here).

Dungavel is currently GEO’s only UK contract. The UK subsidiary that manages the contract, The GEO Group UK Ltd, files annual accounts with Companies House. Because all this company’s revenue appears to come from running Dungavel, these accounts give a unique insight into a detention profiteering contract.

GEO told us that, while the details of its contract are commercially sensitive, the profit margin is “in the single digits”. However it is not clear if they are talking about the profit rate originally agreed with the Home Office in the contract, or the profits that they actually make – which could be much higher.

The GEO Group UK Ltd’s revenue from “custody and offender management services” in 2017 was £5.2 million. The accounts tell us “cost of sales” i.e. the costs incurred when delivering the contract, such as paying staff, maintaining the centre, feeding and monitoring those detained – came to £3.6m in 2017. That leaves a profit margin of 30%: very much in line with the sums G4S is reportedly making. The Dungavel profit margin is harder to discern in prior years as GEO held other contracts, including Harmondsworth detention centre until 2014. Even so, margins for all their operations have consistently been around 20% or above since 2011.

GEO group told us this profit margin “isn’t solely related to the contract at Dungavel House, and therefore the contract is not our sole means of profitability”. However the accounts do not list any other source of revenue in 2017.ii

We asked GEO to clarify but they did not respond. Published Home Office data show the contract is worth £45.2m over eight years: so it seems likely that the vast bulk, if not all, of the company’s money and operating costs are from running Dungavel. We also asked GEO what happens if their profit in fact exceeds the “single figure” rate specified in their contract. Do they pass cost savings on to the Home Office? Again, they did not respond.

Besides “cost of sales”, GEO Group UK Ltd’s accounts also list “administrative expenses” of £0.7m in 2017. This takes the final “net” profit of the UK subsidiary as a whole down to a mere £1 million in 2017. And administrative expenses are significantly higher in previous years. The question is: how much of these are essential to running the detention centre? Or what part relate, for example, to moving money around a multi-national company, or shmoozing politicians and touting for new contracts?

GEO told us these “cover the cost of operating the contract”, including “operations, utilities, repair and maintenance, programs, rent and lease expense and insurances”. However, accounting custom is usually to include all the costs directly incurred in the running of the contract in “cost of sales”, described above. And it is not clear which of GEO’s “administrative costs” here are necessary for the running of Dungavel or for their UK head office. There are also the costs involved in bidding for new contracts, which the company’s accounts repeatedly reference, plus, prior to 2017, significant foreign exchange losses on loans they have taken from their US-based parent.

Again, we asked GEO for further clarification but did not hear back. It is impossible to say for sure without seeing their internal data. But the published accounts suggest the amounts GEO is making simply from running Dungavel are likely similar to those reported for G4S.

Image: G4S boss Ashley Almanza.

20% profits at Brook House

Internal G4S documents, which were reported on by the BBC and The Guardian last September, show similar high profit rates at that company’s Gatwick detention centres, Brook House and Tinsley House.

As the Guardian reported, the Brook House contract made a profit rate of over 20.7% in 2016, and Tinsley House made over 41.5% – although this may be distorted because the centre was closed for part of the year. Profits in earlier years were slightly lower, but still typically around 20% or more.

Like Dungavel, the original Brook and Tinsley House contracts signed in 2009 set official profit margins in the “single figures”. For Brook House, this is 6.8%. So G4S’ internal profit figures are well above what they are supposed to be making on the contracts.

When questioned in parliament about these figures by the Home Affairs Select Committee, G4S’ regional director Peter Neden said that they based on “incomplete information”. But he refused to disclose any more “complete” figures. According to the BBC, Neden argued that doing so would “help competitors”, and said the reported profits “did not take account of costs, including human resources and IT. He said the company’s profits were not more than 20%, but he would not confirm what level they were.”

Of course, without seeing the full G4S figures, there is no way to tell what these “human resources and IT” costs were. “Human resources” here, seems likely to refer to the company’s central management costs, as the wages of staff actually working in the centres are already included. But it seems highly unlikely that management costs and “IT” would be as high as 15% of all revenue – which is what would bring G4S’ profits down to their contractual levels.

In fact G4S’ published accounts also support the picture of extreme profits, if we put a bit of work into analysing them. G4S’ detention centre business is run through a subsidiary with the Orwellian name “G4S Care and Justice Services (UK)”. Immigration detention is only a part of this subsidiary’s business. It also runs five prisons for the Ministry of Justice, and the loss-making COMPASS contract to house asylum-seekers outside of detention. (See our full G4S Company Profile for more detail.)

G4S Care and Justice Services’ revenue was £335.41 million in 2016/17, the most recent reported year (£333.01 million in 2015). After operational costs of £290.2m, the profit rate directly from these contracts was £29.29 million, or 9% of revenue (in 2016, £30.13 million, or 9%).iii

At first sight, this seems much lower than the internal figures. However, these figures are significantly impacted by major losses from non-detention contracts. Above all, this means the big COMPASS deal to house asylum seekers outside detention. G4S won the two COMPASS contracts for the North East, Yorkshire and Humberside; and the Midlands and East of England – and has been complaining ever since that it’s losing heavily on the deal.

For example, in its 2016 accounts G4S Care and Justice adds £14.2 million to its costs to represent an “onerous contracts charge” – that is, money it expects to lose on the COMPASS deal. The year before it recorded a £20.7 million “onerous contracts charge”. It also makes other adjustments related to “commercial disputes” and old PFI contracts.

To see what the figures look like without the impact of COMPASS and other “onerous” non-detention losses, we can first re-calculate gross profit using the company’s “cost of sales excluding specific items”. This starts to more accurately reflect what G4S made from running its detention centres and prisons. On this basis, gross profits were £45.25 million in 2016, 13.5% of revenue, and £50.83 million in 2015, or 15%.

But in fact these are still under-estimates. This is because, to calculate profit rates with COMPASS stripped out, we also need to remove COMPASS’ contribution to revenue and costs. We do not know exactly what this is, but can estimate it from total contract values that the Home Office has disclosed. Combined, G4S’ two COMPASS contracts are valued at £765 million, over a total seven years (2012-19). So roughly £109 million per year, about one third of G4S “Care and Justice” total turnover.

Take this off revenue and cost of sales and the profit rate was actually 20%.iv This is in the territory of the internal documents.

As with GEO, additional costs such as “human resources and IT” referenced by Peter Neden to the MPs may well be included in “administrative expenses” section of the accounts, which would reduce this profit rate. Without seeing their full internal accounts there is no way of knowing the exact rate, and these calculations are unavoidably imprecise.v But as with GEO, the information we have available from published accounts appears to show the company is making very high returns indeed from its detention and prison business.

Mitie and Serco

The two other detention profiteers are Mitie, which runs the two Heathrow centres (Harmondsworth and Colnbrook), and Campsfield House in Oxfordshire; and Serco, which runs Yarl’s Wood. (See our full company profiles on Mitie and Serco for more information.)

Unfortunately there is not the same available information on these two companies’ detention profits as for GEO and G4S. So far, no internal documents have come to light from Mitie or Serco. And their published accounts mix detention contracts alongside other business lines.

What we do know is that both companies see detention as amongst their most profitable operations, and continue to actively bid for new detention contracts. We have no reason to believe that the detention centres they run aren’t just as profitable as Dungavel or Brook House.

If you have any further information on these companies or their detention contracts please get in touch. You can contact us securely through our contact page.

Conclusion: detention is good business

Following the Carillion collapse, a chorus of outsourcing corporations have complained about how times are hard and profits meagre in the age of austerity. But there is a world of difference amongst outsourcing contracts. In some sectors, margins are undoubtedly tighter than in the boom days of Labour’s public-private giveaway. Elsewhere, though, the party continues.

It is important here not to take the companies’ complaints at face value. For example, in 2015 the Financial Times cited unnamed “analysts” estimating sharp decline in detention centre profit margins “from 12 to 13 per cent 10 years ago to between 5 and 7 per cent now.” This was as Mitie explained how the terms of its new contract for the Heathrow centres pushed it to reduce staff and extend lock-up hours. In fact, after its first year of running the centres, Mitie Care & Custody’s profits were up six-fold. From the figures we’ve looked at above, if there has been some margin tightening this must mean that previous contracts were bounteous indeed.

Annex: Detention contracts, size and value

Please note these are necessarily rough estimates. Access to Home Office figures is sporadic and incomplete, to say the least, relying on occasional leaks or vague answers to Freedom of Information Act (FOI) requests.

Heathrow: Harmondsworth and Colnbrook

contracted to Mitie, September 2014-22

number of beds: 1,065

total value at award: £240m

value per year: £30 million – roughly £28,000 per bed

Campsfield

contracted to Mitie, May 2011-19

number of beds: 282

total value at award: £42 million

value per year: £5.25 million – roughly £19,000 per bed

Gatwick: Brook House

contracted to G4S, May 2009-18; now extended to 2020

current number of beds: 558 (after recent expansion)

total value at award: £90.4 million

value per year: £10m – or roughly £18,000 per bed

Gatwick: Tinsley House

contracted to G4S, May 2009-18; now extended to 2020

current number of beds: 178

total value at award: £43.6 million

value per year: £4.8 million – or roughly £27,000 per bed

Yarl’s Wood

contracted to Serco, 2015-23

number of beds: 349 (average occupancy)

total value (calculated at award): £69.9 million

value per year: £8.8 million – or roughly £25,000 per bed

Dungavel

contracted to GEO, 2011-19

current number of beds: 249

total value: £45.2 million

value per year: £5.65 million – or roughly £23,000 per bed

Morton Hall

Run by Her Majesty’s Prison Service (HMPS).

Notes

i COMPASS stands for “Commercial and Operational Managers Procuring Asylum Support Services”. The contracts were awarded in 2012, and are due to end in 2019. See our G4S company Profile for more detail.

ii GEO’s only other UK business is the 50/50 joint venture GEOAmey, which runs prisoner transport for the Ministry of Justice in England and Wales. But this income is treated separately, and does not feature on the GEO Group UK accounts.

iii Both years are knocked down by “administrative expenses” of £24.19 million (£21.51 million). Final pre-tax profits then become £10.25 million, or 3% (£12.07 million, or 3.6%, in 2015). After tax, Care and Justice booked £7.93 million, or 2.4% (£9.16 million, or 2.8% in 2015).

iv To calculate this we also subtracted the estimated COMPASS revenue of £109 million from the overall revenue of £335.4 million, to give an adjusted non-COMPASS revenue of £226.4 million. And we also subtracted it from the cost of sales (excluding non-specific items) of £290.2 million, to give adjusted cost of sales of £181.2 million. This leaves a £45.2 million gross profit.

v For example, we cannot be sure that G4S has receive the full value of the contracts in annual payments – it might be, e.g., that payments were reduced due to penalties for poor performance, although this has not been made public. This would make the actual profit rates lower than our estimates. However, they would still be very considerable. And no records of any such penalties have been published, to our knowledge.