EXECUTIVE SUMMARY

Public-private partnerships (P3s) have become an increasingly important procurement vehicle for Canadian governments seeking to build new or to upgrade infrastructure assets ranging from hospitals, bridges, and highways through to courthouses, water treatment facilities, and concert halls. Although P3s account for only 10 per cent to 20 per cent of total infrastructure spending, governments have come to rely on this type of procurement to help address the longstanding infrastructure deficit. But are P3s clearly superior to conventional procurement methods for Canadian infrastructure projects? This report provides an impartial assessment of the benefits and drawbacks of using P3s, examining arguments that they have higher financing and transaction costs, are less transparent, and lead to lower service standards than traditional procurement processes.

P3 projects tend to feature characteristics such as the integration of two or more project phases, output-based contract specifications, payment upon delivery, private financing, and private sector project stewardship. In contrast, conventional projects are characterized by separate procurements for each project phase, input-based contract specifications, monthly payments to contractors, public financing, and public sector project stewardship.

This report assesses Canadian P3 projects executed under the direction or guidance of the P3 agencies established in the early part of this decade, starting with Partnerships BC and followed by Infrastructure Ontario, the Alternative Capital Financing Office of the Alberta Treasury Board, the Agence des partenariats public-privé Québec (recently renamed Infrastructure Québec), and PPP Canada. The period under review begins when Partnerships BC began advising on the procurement process for P3 projects such as the Sierra Yoyo Desan Resource Road Upgrade Project, the agreement for which was signed in June 2004. These P3 projects are referred to as the second wave of Canadian P3s. The report focuses on the P3 projects initiated by British Columbia, Alberta, Ontario, and Quebec, because these jurisdictions have specialized infrastructure agencies (or equivalent offices within the central agencies of the respective provincial governments) and because their projects are relatively similar in structure and thereby provide a meaningful basis for evaluation. The first wave of P3s has already been reviewed extensively.

Chapter 2 presents a framework for assessing whether (and under what conditions) the procurement of public infrastructure assets using P3s can generate efficiency gains relative to conventional forms of infrastructure procurement by the public sector. The overall proposition is that P3 procurements can provide private sector contractors with strong incentives to deliver the infrastructure outcomes valued by the public sector owner. This results in efficiency gains in the form of lower financial costs, faster delivery schedules, higher-quality outcomes, or a combination thereof relative to conventional procurement methods. The efficiency gains from P3 procurements are achieved through one or more of the following mechanisms:

♦  performance-based contracts, which specify deliverables in terms of the outputs (e.g., lane availability and skid resistance in the case of roads) rather than prescribing specific materials to be used;

♦  optimal risk allocation between the public sector owner and the private sector partner, which means that many of the risks are transferred from the public sector to the private partner if the partner can manage them more cost-effectively;

♦  integrating the design, construction, operation, and maintenance phases of a project in order to minimize total life-cycle costs; and

♦  private financing, which includes primarily project-specific debt and a small tranche of equity, to ensure that the risks transferred to the private partner are borne and managed by that partner. Under conventional construction contracts, such as "stipulated sum contracts," private contractors are paid monthly based on the percentage of work completed.

The relative benefits of this procurement model take the form of:

♦  cost savings or quality enhancements in the design or construction of a new facility, as well as in its operation and maintenance (i.e., in the service provision phase); and

♦  time savings in the delivery of a public infrastructure facility fit and available for use.

It is also important to recognize that these benefits come at a cost. Specifically, compared with conventional procurements, P3s entail the following additional costs for a project of similar scope:

♦  The cost of transferring selected risks to the private partner. This is also known as the risk premium, which is used to compensate the private partner for assuming risks additional to those associated with a conventional contract. The risk premium usually represents the largest part of the additional costs involved in P3 procurements.

♦  Higher financing costs. The private financing used for P3 projects is more expensive than the public financing (i.e., government bonds) used for conventional procurements.

♦  Higher transaction costs. P3 contracts cost more to develop and monitor than conventional infrastructure contracts.

If these three categories of costs are offset by the cost savings from transferring selected risks to the private partner, the overall costs of the project will be lower under a P3 approach than under traditional project delivery.

The Canadian evidence on the cost and time performance of P3s comes from two sources: value-for-money (VfM) studies, which compare the total costs of P3 and conventional procurement methods for each P3 project; and studies comparing the performance of projects against their own time and budget targets, which are set either when a project is first announced or when the project agreement is signed. The VfM study results indicate that the second wave of Canadian P3 projects is delivering important efficiency gains for the public sector (i.e., taxpayers) relative to conventional procurement approaches. The estimated value of these gains varies from just a few million dollars per project to over $750 million in the case of the Autoroute 30 project south of the Montréal area. When these savings are expressed as a proportion of what it would have cost the public sector to procure the projects through conventional contracting methods, the savings range from 0.8 per cent through to 61.2 per cent per project.

VfM savings are necessarily prospective when estimated during the procurement period. Whether the actual savings match the expected savings by the end of the P3 project depends on the degree of cost and time certainty of P3 projects. However, the savings can also be verified on an ex post basis-after project completion-as in the case of the construction of the southeast and southwest legs of the Edmonton Ring Road. The two projects were broadly comparable, but the P3-procured project took two years less to deliver than the conventionally procured project.

The Canadian evidence on the cost and time certainty of P3 projects is incomplete, because only 19 of the 55 second-wave P3 projects have reached substantial completion. However, these early results point to a very strong performance. Most of the 19 projects have been delivered either early or on schedule, with only two projects delivered up to two months late. (The financial penalties resulting from these delays were borne by the private sector partner or by the public sector partner in the case where delays were due to risks retained by the public sector.) Moreover, the interim results for the P3 projects that remain in the construction phase provide little reason to expect substantial cost or time overruns, based on the information regarding contract variations and claims against the public sector. Therefore, the preliminary evidence indicates that the second wave of Canadian P3 projects is providing a high degree of cost and time certainty for the period from financial close through to completion of construction.

It is important to emphasize that cost certainty in a project is not just about saving a few dollars or improving the predictability of public sector budgeting. Cost certainty is vital from a public interest perspective, because it enables public decision makers to allocate public funds to the right projects. Without cost certainty, the public sector is often compelled to allocate relatively large amounts of additional funds midway through a project, regardless of whether the additional funding would have been justified on a VfM basis. This occurred in the Vancouver Convention Centre Extension Project, the Sudbury Regional Hospital (Phase I) project, and the Montréal subway extension to Laval, all of which were conventional procurements.

We also wish to dispel a few myths about P3s in Canada. First, P3s in Canada are not about the privatization of public assets. Ownership of the new infrastructure facilities either remains with the public sector or is transferred back to the public sector at the end of the contract term. Second, long-term P3 projects (i.e., those with a maintenance phase) help ensure a satisfactory level of maintenance and upgrade work during the life of the facility. The anecdotal evidence collected in this report suggests that there is little basis for the criticism that service standards suffer under a P3 relative to a conventional maintenance contract or even relative to in-house provision.

Chapter 3 reviews the main drivers of efficiency gains in P3s relative to conventional procurements. The most important is arguably the optimal risk allocation process, which is at the heart of the P3 procurement process adopted by P3 agencies and offices across Canada. This involves identifying and valuing project risks upfront and transferring to the private partner those risks that these firms have the expertise and experience to handle. This risk transfer process also has the considerable advantage of forcing an upfront consideration (i.e., before or during procurement) of all the project requirements and associated costs. Without such upfront assessments, there is a much higher risk of cost overruns, as evidenced in several of the case studies of conventional infrastructure procurement reviewed in this report.

Performance-based contracts, which specify deliverables in terms of desired outputs rather than inputs, are another driver of efficiencies in P3 contracts. These contract provisions encourage private partners to consider the most cost-effective delivery practices. The integration of the design, construction, operation, and maintenance phases of a project is yet another potential driver of efficiencies, because it enables private firms to adopt innovations that can reduce life-cycle costs, even if they involve greater investment in the design or construction stages. However, there is little empirical evidence of the relative importance of these two efficiency drivers. As well, both these efficiency drivers can be adopted in conventional forms of contracting, provided that care is taken to specify the desired outputs and to design an appropriate contract covering a substantial part of the expected useful life of the infrastructure asset.

Private finance is the fourth efficiency driver in P3 projects. Thanks to this feature of P3s, the public sector pays the private partner only upon delivery of the facility (although some milestone payments are sometimes made before construction is complete). Consequently, the private partner has a powerful incentive to build the facility in a timely manner and in a way that meets the contractual requirements. Such payment by results forces the private partner to take on most of the financing requirements for the project, which include sizable debt obligations. Without these financing requirements, private firms would have little incentive to complete their contractual obligations should they encounter significant cost overruns that cannot be passed on to the public sector. Therefore, private financing can be considered the glue that binds together the other efficiency drivers mentioned above, particularly the optimal risk allocation process and the performance-based contract provisions.

However, these efficiency drivers do not guarantee that all P3 infrastructure projects will generate net efficiency gains. In some cases, the gains can be more than offset by a combination of the incremental cost of private financing, any additional costs due to the risks transferred to the private consortium (i.e., the risk premium), and the incremental transaction costs. This is why each infrastructure project requires a rigorous VfM assessment to ensure that a P3 procurement option delivers value relative to a conventional procurement method, as was standard practice for all second-wave P3s. Chapter 4 of the report reviews VfM tests and other aspects of P3 procurement processes, such as the guidelines used to screen infrastructure projects for their suitability as P3 procurements and the transparency of P3 procurements compared with conventional delivery of infrastructure projects.

VfM tests are designed to ensure that the right projects are selected as P3s and that the risk transfer effected in a P3 agreement is cost-effective for the public sector owner. Our review of the available VfM studies and guidance documents suggests that each of the four jurisdictions under consideration-British Columbia, Alberta, Ontario, and Quebec-has developed a rigorous VfM methodology for comparing the costs of P3s and traditional procurements. (VfM studies are not published for the Alberta P3 transactions, but the VfM methodology is available through Alberta Infrastructure and Transportation.) VfM tests are not undertaken as an afterthought: A first pass at the test is done before the start of the procurement process (i.e., before the request-for-proposal stage), and the test is finalized after the financial close. This report also suggests that there is value in updating the VfM studies ex post at key milestones, such as at completion of construction, and periodically thereafter. In contrast, conventional infrastructure procurements are normally not subject to any VfM-type tests to inform procurement strategy.

In addition, this report notes that the procurement process for the second wave of P3s is considerably more trans-parent than that for conventional infrastructure projects of equivalent scale. This is because the key procurement documentation, including a redacted form of the partnership contract, is publicly available and a fairness commissioner assesses the fairness and transparency of the process for all bidders. Neither of these features is typical of conventional public infrastructure procurements.