This article is an excerpt from the 2020 edition of The Art of Tendering: A Global Due Diligence Guide, which is available for purchase.
In her October 2019 Canadian Centre for Policy Alternatives report entitled Shrouded in Secrecy: The Queen Elizabeth II Hospital Redevelopment and the Privatization of Nova Scotia’s Health Care Infrastructure, Larissa Atkinson questioned Nova Scotia’s decision to allocate $764 million to a public-private partnership (P3) as part of the government’s $2 billion commitment to update the Queen Elizabeth II Health Sciences Centre. More generally, Atkinson’s report also challenged the value-for-money analyses relied on by governments to justify the use of P3 project delivery models.
With respect to Nova Scotia’s recent health centre P3 decision, Atkinson’s report asserted that “there has been a great deal of secrecy and obfuscation surrounding the P3 decision, including withholding crucial information from the public” and that the “Nova Scotia government has refused to release the controversial and confidential value-for-money (VfM) report it says proves the P3 model would provide financial and service-level benefits until after the contract with the successful consortium is signed.”
Atkinson’s report also asserted that “there does not appear to be a commitment to effective oversight and accountability within government to protect this crucial public policy decision from being dictated by private interests” and that the government “failed to follow-through on recommendations by the province’s Auditor General to improve its internal oversight capacity of large development contracts” and had instead “outsourced decision making to large multinational companies” who are “members of the Canadian Council for Public Private Partnerships (CCPPP), and have a vested interest in promoting P3 projects.”
Atkinson’s report raised concerns over “a deeply entrenched institutional bias in favour of P3s and against engaging in public procurement by conventional means” and challenged the widely held assumptions that P3 projects provide better value-for-money and deliver more timely performance than conventional procurements:
…the Nova Scotia government’s justification for pursuing these two P3s rest on misleading claims about P3 advantages and public procurement flaws. Among these are claims that P3s offer more value for money and are more reliable in delivering services on time. In fact, data retrieved from provincial audits and Infrastructure Ontario show both of these justifications to be false. Moreover, claims about the value of transferring risks associated with these projects to the private sector overstate the risks of public procurement, such as the likelihood of cost overruns, and understate, or ignore altogether, the risks the government assumes in entrusting its infrastructure to private consortiums. One such risk is corporate bankruptcy, another is reduced safety and cleaning standards associated with corporate profit maximization behaviour.
This report estimates that in the case of the Bayers Lake and Halifax Infirmary projects, the cost of private sector borrowing will be 125% more than financing these projects publicly. Far from adding value to public infrastructure, P3s take money away from future governments, services, and infrastructure projects.
Noting the “poor track record of P3s” in Nova Scotia, and in P3 hospital projects more generally in Canada and the United Kingdom (UK), Atkinson’s report called on the Nova Scotia government to revisit its P3 strategy in the new hospital project and recommended that a conventional design-build project format be used so the project could be directly managed by the government using public funds. Atkinson’s report also challenged the “voodoo accounting” employed in P3 value-for-money (referenced below as VfM) assessments, claiming that the firms who conduct those assessments use misleading calculation assumptions and alleging that they are in a conflict of interest as members of the CCPPP:
And yet, while VfMs are intended to provide an unbiased and comparative assessments of value, they have been criticized as methodologically designed to privilege P3s. One such tactic involves using a high discount rate in the VfM to measure the present cost of future private sector borrowing. Discount rates allow borrowers and lenders to account for the way factors such as interest rates and inflation change the value of money over time. The higher the discount rate, the lower the perceived present value of future spending. However, provincial auditors have found that VfMs often rely on excessively high discount rates to measure the cost of P3 investment over time, making P3 options appear significantly cheaper than public procurement. As a result of these criticisms, industry practices have shifted to adopt more appropriate discount rates that are equal to the cost of government borrowing.
While the shift to lower discount rates is somewhat promising, other serious issues with the methodology remain. Perhaps most pressing is that the firms hired to conduct these analyses, such as Deloitte — the company hired to conduct the VfM for the QEII Redevelopment — have a vested interest in promoting P3 projects. As a member of the Canadian Council for Public-Private Partnerships (CCPPP), Deloitte is beholden to that organization’s mandate to promote “evidence-based public policy in support of P3s” to “educate stakeholders and the community on the economic and social benefits of public-private partnerships” and to “influence the way in which public services are financed and delivered in Canada.” In short, Deloitte and other members, such as Kasian (which currently sits on board of directors), are encouraged to use data in a way that is advantageous to P3s and unfavourable to public procurement as a form of advocacy. As long as members of the CCPPP conduct these analyses, value for money reports will continue to be methodologically unreliable. [footnote citations deleted]
In her report, Atkinson also noted that health-sector P3s have fallen out of favour in the UK after a string of sub-optimal results:
In recent years, P3s have come under fire as costly, at times risky, and administratively opaque alternatives to traditional public procurement. Canada’s approach to P3s is based largely on the U.K.’s similar Private Finance Initiatives (PFI), initiated in the early 1990s. But from their inception, these public-private contracts were beset with setbacks and cost overruns. For example, the first British PFI hospitals held far fewer beds than the older hospitals they replaced including an averaged 27% reduction in acute-care beds. Efforts to correct this problem resulted in the forced closures of non-P3 hospitals and diverted resources from other necessary health services, such as mental health and primary care. In January 2018, the UK National Audit Office found financing PFI projects cost approximately 40% more than the estimated cost of financing these contracts with government money. As a result of these ongoing challenges, in October 2018 the Chancellor of the Exchequer announced that it “will abolish the use of PFI and PF2” — formally ending that country’s 26 year experiment with public-private partnerships. [footnote citations deleted]
Turning to the Canadian context, Atkinson cited a series of highly critical auditor general reports in Ontario, British Columbia, and Quebec that raised concerns over the additional costs associated with P3 projects, and over the questionable financing and risk assessments used to justify P3 arrangements:
Here in Canada, P3s have a similarly poor track record. In 2014 the Ontario Auditor General released a searing report of Ontario’s entire P3 program and methodology, arguing that the program had cost $8 billion more than if these projects had been publicly financed and operated. The report also found that each of Infrastructure Ontario’s 74 P3s were justified on the basis that they transferred large amounts of risk to the private sector, but there was little effort to support these claims in crucial value-for-money assessments (VfM). In 2014, the British Columbia Auditor General raised concerns about the high cost and levels of debt associated with P3 projects after examining 16 different P3 projects in that province. They also found that in comparing a P3 with a public sector comparator, Partnerships BC consistently used the most expensive possible method of public sector delivery, rather than the most likely one. [footnote citations deleted]
Atkinson also referred to Nova Scotia’s “disastrous” experience with P3 school projects, which resulted in the province buying back 37 of 39 schools after concerns were raised about extensive costs, negligence, poor oversight, and safety risks associated with the P3 projects:
Nova Scotia has had its own expensive and troubled relationship with P3 delivery. A 2010 Auditor General review of 39 Nova Scotian P3s revealed substantial problems with the cost, administration of contracts, delivery, and safety of its P3 schools. As recently as 2016, the current Auditor General, Michael Pickup, criticized the government for failing to “appropriately manage P3 decisions to date.” That same year the Canadian Centre for Policy Alternatives-Nova Scotia (CCPA-NS), released a scathing report outlining the extensive costs of this program as well as issues of negligence, poor oversight, and safety risks. In the end, the province yielded to these criticisms and bought back 37 of the 39 schools to end these P3 contracts at a cost of $215.9 million. The overall waste of public dollars over the 20 years of the contracts is unknown as the government has never released all the relevant information. [footnote citations deleted]
Atkinson also questioned the claim that P3 projects are delivered on time and on budget, noting that this depends on “where you draw the starting line” when measuring the time and costs of a procurement project. She asserted that while P3s projects are subject to extensive contract negotiations that allow the initial time and costs benchmarks to be adjusted before official project time and cost benchmarks are set, by comparison, the assessment of conventional procurements are skewed since they are measured against their initial non-adjusted pre-award time and cost benchmarks:
When the McNeil government announced its decision to go with P3s last October, they relied on a common assumption that P3 projects provide the greatest certainty for “work to be done on budget and as ordered.” The data, however, tells a different story. P3 projects are able to claim to be “on budget and on time” because the completion date is set after the lengthy and expensive process — usually years — it takes to reach the contract stage. During these lengthy consultation processes private companies work together with government officials to adapt project budgets to meet evolving contract demands. Because budget estimates are not released to the public until after this process is resolved and all parties are satisfied, there is a perception that P3 budgets do not fluctuate. In reality, there is often a great deal of fluctuation and sometimes substantial cost increases that take place out of the public eye during the negotiation phase of these contracts.
For example, the Abbotsford Hospital and Cancer Care Centre in BC started with a budget of $211 million, but by the time the contract was signed the budget had soared to $335 million.42 Similarly, a 2008 Auditor General report shows estimates for the Brampton Civic hospital, Ontario’s first P3 experiment, ballooned between initial estimates of $381 million in October 2001 (for a 716 bed hospital) to $525 million in November 2004 (for a smaller 608-bed hospital), but there was little accounting for this increase.
By contrast, public procurement estimates are often announced early in the negotiation process as officials look to “sell” their plans and gain approval from decision makers. This has led to the perception that there are large cost overruns when, in reality, the original budget was premature. [footnote citations deleted]
Atkinson also asserted that the benchmarking is further slanted against conventional procurements due to an optimism bias that sees conventional project teams engage in “strategic misrepresentation” and set unrealistic initial time and cost estimates to help obtain project approvals:
Politicians and project promoters have an incentive to make their proposals attractive to voters by strategically proposing low estimates, knowing that once a project is approved the budget can be adapted. This tendency is encouraged by the perception that there are few consequences for politicians and bureaucrats when project costs escalate.
The Colchester Regional Hospital in Truro Nova Scotia is a perfect example of such “strategic misrepresentation” influencing the public’s perception of cost overruns. The original budget was announced as $104 million — approximately $80 million less than the final cost of $184 million (not including the cost of furniture and equipment). Yet, as then Auditor General Jacques Lapointe observed, the original budget was unrealistic. It did not take into account basic costs like inflation, space contingencies, and physicians’ offices (which together totalled $22 million). These costs were cut from the preliminary budget before Cabinet approval was requested, presumably to make the budget appear more attractive to decision-makers. But, as Lapointe stressed in his audit, “[t]hese decisions ultimately made the initial approved budget a meaningless number for planning purposes.” [footnote citations deleted]
Furthermore, as Atkinson noted, a recent Ontario study found that even though benchmarking was biased towards P3s, conventional procurement projects still had higher timely completion rates than P3 projects:
Further, according to a 2015 Infrastructure Ontario report on 45 Alternative Financing and Procurement (AFP) (Ontario’s term for P3) projects and seven traditional direct delivery projects (with project costs between $10 million and $50 million), the public procurement comparators (PSC) were 71% on budget and 86% on time (PSC) whereas 98% of the AFP projects were on budget and 73% were on time. If we account for the long negotiation processes that AFP/P3 projects have to adapt their budgets, and weigh this against the added incentive public decision makers have to under-estimate budgets in order to gain approval for their projects, the “on budget” discrepancy between these two options is less meaningful. Given, however, the emphasis that P3 advocates place on corporate and financial incentives that facilitate faster delivery of P3 projects as compared to the public sector (for example, P3 payments are often withheld until the project is complete or construction milestones reached), it is striking that this report found PSCs to be significantly more punctual than AFPs. [footnote citations deleted]
While Atkinson’s report acknowledged that a capacity gap continues to exist in the execution of conventional procurement projects, she warned that the P3-bias serves as a self-fulfilling prophecy since “public sector expertise is being eroded by the government’s reliance on private sector delivery”. Furthermore, she maintained that using P3s does not solve the capacity gap since P3s require a higher degree of oversight capacity and that even those jurisdictions that have special purpose bodies that manage the P3 portfolio remain subject to state capture due to a lack of independent oversight.
Atkinson’s report also challenged the risk transfer rationale used to justify the additional costs associated with P3 projects. As Atkinson observed, while project delivery risks are theoretically transferred to the P3 consortia, the benefits of that risk transfer are exaggerated by inflating the risk assumptions allocated to conventionally run projects:
Given the known costs of P3 infrastructure, the most persuasive justification for choosing more expensive P3s is that the risk of undertaking large infrastructure development is theoretically transferred to the private sector. The value of risk transfer is in placing the onus for project delays or other project cost overruns on the private provider. As Premier McNeil noted: “We’ve seen huge overruns. We believe this is the best value for the taxpayers of Nova Scotia.” This justification means that taxpayers pay a premium to transfer risk to the private sector equivalent to an insurance premium. Yet it is often unclear what risk is being transferred.
Heather Whiteside, for example, lists common risks attributed to P3 projects as “site (tenure, access suitability, design) and construction (delays, access, cost-overruns), operation and maintenance (cost overruns), and financial risks (interests rates, inflation).” Many of these presumed risks are, notably, associated with the design and build phase of construction (i.e. a traditional procurement model).
The Ontario Auditor General’s 2014 Report notes, moreover, that risks are often overestimated and subjective (not based on verifiable data or independently verified) and also that there is a bias against public sector delivery “resulting in significant differences in the assumptions used to value risks between the public sector delivering projects and the [Alternative Financing and Procurement, or P3] approach.” That audit shows that Infrastructure Canada assumed that the risks of public procurement were approximately five times higher than the AFP delivery and that this assumption gave AFPs an overall positive VfM. And yet, the report notes that no empirical data supports this assumption and, even more problematically, the calculation for risk transfer involved significant errors including double counting and false attribution of risk transfer. When these more egregious errors were corrected for, 17 of the projects originally discounted show a positive VfM for public procurement at an estimated savings of $350 million.
The following table from the Auditor General’s report shows the extent to which such inflation of private risk has been used to tip VfM reports in favour of P3s.
In reality, however, the risks incurred by P3s are rarely transferred to the private sector. As Toby Sanger puts it, all “P3s in Canada are structured as Special Purpose Vehicles (SPVs). This means the larger companies behind P3 projects can walk away at any time, risking only the equity they have put into the project, which is typically 10–15% of the initial cost. Meanwhile the amount of “risk” that is assumed transferred to the firm averages about 50% of this base project cost.”77 In other words, companies build a premium into their contracts in the face of contract disputes with future governments. The ultimate responsibility for delivering a project or service therefore rests with the government or another public entity. [footnote citations deleted]
Finally, as the report warned, recent events have proven how governments will always remains the ultimate underwriter when large P3 projects fail:
Even though private corporations assume risk, they can still seek bankruptcy protection and avoid contractual commitments. When this happens, governments must scramble to maintain public services and taxpayers are stuck with the higher costs of private sector operation. For example, Carillion, a major global player in the promotion of Public-Private Partnerships (P3s) went into liquidation in the UK, triggering the UK’s decision to end its P3 program. Carillion’s Canadian operation employs 6,000 Canadian workers and is “heavily involved in 10 P3s in Canada with total capital assets of $3.25 billion, eight in Ontario and one each in Saskatchewan and the Northwest Territories.” As one critic put it, “[T]he fundamental issue raised by Carillion’s failure is that vital public services cannot be outsourced to private contractors, without Government underwriting the risks of collapse… So-called risk transfer, where financial risk associated with infrastructure construction and operation was transferred to the private sector, to justify the higher interest on PFI (PPP) loans (double the interest rate on Government loans) is, and always was, a massive con.” [footnote citations deleted]
As this report illustrates, recent studies have raised significant concerns about the broadly accepted benefits of P3 projects. While P3 advocates will continue to point to success stories in support of using that model, and P3 critics will continue to raise fundamental questions about the benefits of that model, project teams should avoid rushing to judgement or relying on untested assumptions when developing their procurement strategies and selecting their procurement formats.
