The case for public-private partnerships in infrastructure.

Murphy, Timothy J.

While still controversial, public-private partnerships (P3s) are quickly becoming an important part of infrastructure procurement for all Canadian governments. (1) A P3 project office was announced in the federal government's 2007 budget, and P3s are proceeding, or have been undertaken, in Alberta, Ontario, British Columbia, Quebec, New Brunswick, Nova Scotia and Nunavut, and in cities such as Ottawa, Calgary and Kelowna. (2) Publicprivate partnerships are playing a bigger role in capital projects across all areas of government, such as transportation, communications, power generation, energy delivery, water and wastewater, waste disposal, courthouses, hospitals, jails and even legislative assemblies.

This article attempts to distil the public arguments for and against P3s, including, where appropriate, the academic literature, and argues in favour of them, particularly for infrastructure assets and related services, as long as governments clearly understand the risks involved from the outset and throughout the life of the project and ensure an optimal and effective allocation of risk to the private sector. To do this, governments need the right expertise on their side of the table and the right levers of accountability to help enhance the legitimacy of P3s as a vehicle for delivering public and quasi-public goods and services and to monitor risk allocation throughout the project term.

The public policy rationale for P3 arrangements

"Off-book" financing--a declining factor

One primary rationale for a P3 arrangement rests on the transfer to the private sector of the financing of the delivery of the public asset. Historically, this has reflected the accounting treatment by governments of infrastructure spending in a cash accounting system. In the past, governments who reported their finances on a cash accounting basis benefited if infrastructure spending could be done "off-book" by the private sector. For example, instead of recording the full cost of a $100-million expenditure in the year it purchased a building, the government could pay the private-sector operator a $5-million annual "lease" payment and record only this lesser amount in its books. In this system, postponing the obligation or stretching the payment through a P3 arrangement permits a government to build now and pay later--an attractive proposition to cash-strapped governments.

Many Canadian governments are now changing to accrual accounting, which spreads the cost of acquiring an asset over its useful life and requires governments to consider in their annual budgets such asset-related expenditures as maintenance, replacement and other life-cycle costs. The federal government adopted accrual accounting in 2002, and the Public Sector Accounting Board has published new standards that will require local governments in Canada to use accrual accounting as of 1 January 2009. (3)

As a result, much of the attraction for government in making a deferred stream of payments to a private-sector entity instead of recording the entire purchase cost of an asset in a single fiscal year disappears as the accounting treatment of both methods of procurement merge. (4) The net result is that over time, most, if not all, of the accounting differences that might create incentives to a P3 are being eliminated. The Province of Ontario, for example, takes the position that, with its move to accrual accounting, "accounting considerations are no longer a driver of the model to be used for delivering infrastructure investments." (5) Nonetheless, P3s still offer the potential to secure better value for money and greater innovation in the delivery of public services. The advantages are summarized below.

Accelerating construction

While the accounting treatment of P3s and traditional procurement has largely merged, there can remain a difference between the timing of payments in the two procurement models such that needed public infrastructure can be built faster under a P3 for debt-restricted governments. In a traditional procurement, the government pays the general contractor during construction. In a P3 model, no payments at all are made until substantial completion of the project is reached and services are being provided. As a result, construction of P3 projects can commence in advance of a conventionally procured asset, where cash-flow timing can make a difference.

In addition, accessing private capital may permit a project to proceed where no public funds are available. For example, one of Canada's first P3s for drinking water enabled the City of Moncton to assume ownership of the facility without having to make any up-front capital investment. The facility was built in just eighteen months and was designed to meet or exceed all Canadian drinking-water quality criteria. (6) In fact, the city now advertises its water quality as a location attraction for businesses and residents.

On-time and on-budget delivery

Transferring this aspect of the design and construction risk inherent in developing infrastructure to the private sector creates a powerful incentive to effective performance. In the P3 model, cost overruns are absorbed by the private sector, and delayed completion dates can result in penalties. As a result, "[w]ith payments better aligned to the delivery of project objectives, public private partnerships also have a solid track record of completing construction on time or even ahead of schedule." (7)

The public sector, for its part, does not respond to the same incentive. Public-sector procurements are subject to what is often called an optimism bias, which is the tendency in the public sector to budget for the best possible outcome as opposed to the most likely. (8) In fact, due to the mixing of the policy delivery function with the oversight function within government, costly enhancements or changes to the project after the initial contract award can be a frequent occurrence. This is particularly true for traditional public-sector procurement of non-standard buildings, where one study found that estimates of project duration and total expected capital expenditure were off by fifteen per cent and sixty-six per cent, respectively. (9) Another review of the performance of the United Kingdom's "private finance initiatives," or PFI projects, conducted by the National Audit Office, concluded that PFI projects were on time in seventy-six per cent of the cases and on budget seventy-eight per cent of the time, compared to thirty per cent and twenty-seven per cent, respectively, for non-PFI projects. (10)

Shifting risk to the private sector

In this context, the benefits of a P3 arise through the transfer to the private sector of the design and construction risk and of the risk of operating and managing public assets. In other words, all the commercial risks, such as design risk, meeting standards of service delivery, cost overruns, market risks, etc., related to building and delivering the public good can be transferred to the private sector. An additional benefit derives from using the private sector as a hedge against the failings of government. In other words, political and budgetary pressures can lead to an under-investment in existing assets by a government where demand for increased services now can easily outweigh the need for expenditures on the maintenance of an existing asset, resulting in its deterioration over time. (11) A well-designed P3 with a concessionary term will obligate the private-sector partner to properly maintain the asset because the concessionary or annual payment to the P3 partner includes an amount for maintenance and penalties for failure to comply. As a result, governments prevent themselves from deferring maintenance by entering into a P3.

A number of academic studies of the actual performance of Canadian P3s have come to somewhat pessimistic conclusions but have done so largely on the basis of "one-off" projects where governments arguably had insufficient contract management skills to anticipate, manage and allocate risk. Subsequently, specialized agencies, such as Infrastructure Ontario--the provincial government's centralized infrastructure procurement agency--were established with significant technical, legal and financial expertise. It now also has significant experience, having signed twenty-two infrastructure deals in the last year. Based on Infrastructure Ontario's own internal valuation process, all of these projects represent value for money over a comparative traditional public-sector procurement. (12)

Cost-savings

Although this argument is far from undisputed, (13) P3 proponents argue, on the bases of both logic and experience, that the P3 model can deliver cost-savings to government. The argument from logic asserts that efficiency and cost-effectiveness are not hallmarks of public-sector service delivery, since government agencies do not gain the benefit of either ownership or competition effects. (14) These effects are argued to be what drives efficiency-maximizing behaviour in the private sector. The public sector does not gain these benefits because 1) the government does not usually seek to maximize profits; 2) with ownership residing completely with the state, there is no market for corporate control; and 3) government agencies rarely face competition. (15) The lack of incentives to control costs nullifies any benefit to the public sector of its ability to borrow at lower interest rates. Further, as governments are often monopoly suppliers, there is no built-in incentive to innovate or control costs.

Proponents argue that a review of comparative experience between the public and private sector related to the construction of new or the renovation of existing assets suggests that theory is borne out in practice. For example, an Australian study found that eight Partnerships Victoria projects achieved savings on average of nine per cent against traditional procurement. (16) A United Kingdom study found that among a sample of twenty-nine PFI projects, the average saving was close to seventeen per cent. (17)

Customer service improvements

For concessionary projects in particular, which rely on user-fees, tolls and other similar charges for revenue, the private-sector participant has a strong incentive to provide high-quality customer service. A U.K. study in 2005 found that public-sector contract managers rated PFI projects highly, with ninety-six per cent of those surveyed across 100 projects reporting that the overall performance was satisfactory to very good. Similarly, eighty-nine per cent reported that the services were provided in line with the contract or better. (18)

Enabling the public sector to focus on outcomes and core business

Where a government faces limited resources to meet public demand for services and decision-makers have limited time, a benefit can be derived from focusing on outcomes and core problems. Arguably, having governmental human resources and budgetary allocations focused on construction and maintenance of physical assets, for example, diverts scarce resources to non-core tasks. (19) If governments can focus on designing a contractual relationship with a private-sector partner that identifies and provides appropriate incentives to publicly valuable outcomes instead of wasting public resources on the methodology of delivery, P3s can be a mutually beneficial arrangement that leads to better public services.

Responding to the key arguments against P3s

The case against P3s centres on five main points: 1) their real costs are higher than traditional government procurement, and, as a result, they do not meet the value-for-money test; 2) design and service quality over time fails to live up to the standards of publicly delivered services; 3) reduced transparency and unclear lines of responsibility means they are less accountable to the public good; 4) they are a threat to the rights of workers (particularly unionized ones) and to jobs; and 5) they reduce the flexibility of the public sector to respond to public demands.

Higher cost, less value

On the cost side, opponents argue that P3s are more expensive because they face a "triple hurdle--the higher cost of private borrowing; the need to make a profit and associated other potential inefficiencies; and higher procurements costs." (20) Even proponents agree that as a starting position these are costs that weigh against a P3 but the overall value-for-money proposition favours P3s in the right circumstances. They argue, as noted earlier, that other factors, such as access to capital that would not otherwise be available, increased certainty of on-time and on-budget delivery, risk transfer to the private sector, contractual provision for ongoing asset maintenance, and cost-savings arising from incentives to efficiency, will more than offset the triple hurdle.

Furthermore, assessing the cost of money between a P3 and a traditional procurement makes a false comparison, since it fails to take into account which party bears the risk. In traditional government procurement, the lower borrowing rate assumes that the project is risk free, which it isn't. The risk is underwritten by the taxpayer. In a P3, however, the risks and potential costs are underwritten by the private sector, albeit compensated by an appropriate return. In other words, in a P3 procurement, the government is paying an insurance premium to protect against the risk of higher costs, rather than self-insuring at a zero premium cost but at a potentially high failure cost. (21) To make an accurate comparison, it is not the cost, but the net benefit, taking into account all factors, that is the most relevant benchmark. Here, a well-negotiated P3 model can offer significant value for money, assuming--and this is the key point--the risk transfer to the private sector is effective. (22)

In addition, there can be a real benefit to the public sector in transferring project risks and thereby off-loading the political heat of pricing for a public or quasi-public good. Generally, a government's ability to address an infrastructure gap is a function of overall revenue, and its investment can be constrained if current taxation levels do not provide adequate fiscal room. However, governments, applying what is known as the benefits model of government services, (23) which charges the cost of a service to its users, can provide new or expand existing services to users through the private sector, while also transferring the pricing risk, by using a P3 model. Governments, of course, have themselves applied the benefits model to price assets and services, such as electricity, but Ontario's experience with electricity rates is an example of the risks that governments take on matching price to cost. Transferring this risk to the private sector may also allow a better balance between supply and demand (24) and provide the further benefit to governments of insulating them from facing issues of quality of service and its pricing in the political realm, ensuring quality is a risk transferred to the private sector as a matter of contract. Again, the government's application of contract negotiation and management expertise is crucial to getting the risk transfer right.

This theoretical point seems borne out to some degree in practice. In one study on highway P3s in the United States, government-run toll roads had not raised tolls on two of the projects studied for twenty and twelve years, respectively. As a result, needed investment was not made. Transferring those roads to a Pg, where toll rates are set by the market, can capture value and fund further investment. As the study noted, the shift of control to a non-political entity that is capable of behaving over time in an economically rational manner opens up financial possibilities that depend on the financial markets recognition of that reality. (25) However, where there is a high degree of revenue uncertainty as a result of difficult or problematic traffic demand-forecasts, a higher degree of risk-sharing between the public and private sector, and effective contract management are essential to making the P3 model work in this context. (26)

Finally, the "borrowing cost" argument can be off-target where truly innovative P3s that require little or no public money but only regulatory or legislative approval can be structured. For example, the P3 structure used to rebuild Washington, D.C.'s Union Station used no incremental taxpayer money at all. The private-sector partner is recouping costs in part from rents paid by retail shops in the renovated facility. (27)

In practical terms, experience with the use of Pgs supports the argument that higher "triple hurdle" costs can be offset, including by the transfer of risk to the private sector of on-time and on-budget delivery of procured assets. The key comparative, United Kingdom data on cost and time overruns in traditional versus PFI procurements outlined in the Mott MacDonald Report, strongly supports the thesis of significant cost-savings and improved timeliness for P3 projects. (28) The value of these results, compared to some analysis of early Canadian P3 experiments, is that they are based on a reasonable number of projects in circumstances where the public sector had sufficient managerial expertise to effectively transfer risk. (29)

Ontario's experience with hospital P3s also provides anecdotal support for on-time and on-budget delivery. The Royal Ottawa Hospital, the first of Ontario's new model P3s to be constructed, recently opened six weeks early and on budget. (30) In British Columbia, the first of Partnership BC's public-private partnerships based on the PFI model, the Sierra Yoyo Desan Road, opened on time and on budget, as has the Kicking Horse Canyon Project this past year. In addition, a recent survey of construction risk in P3 projects in twenty-two countries conducted by Standard & Poor's found that ninety-one per cent of those surveyed, all of whom were very experienced in Pgs, agreed, or agreed with minor qualifications, that P3s had a better track-record of timeliness and cost-effectiveness than conventional public-sector procurement. (31)

In sum, the evidence is strongly suggestive of better timeliness and cost-effectiveness for P3s, at least through to the end of the construction phase. However, P3s are still beneficial even if the budget and timeliness objectives are not met, since the private sector bears the cost of that failure. In traditional procurements, the taxpayer bears those risks. If the risk transfer is successful, any failure by a P3 to meet cost and timeliness standards will not be paid for by the public purse. The key is ensuring that the contractual provisions governing the P3 relationship optimally allocate risks and, as a result, create incentives appropriately.

Lower-quality design and service

The value-for-money proposition also faces the second significant argument against P3s made by opponents and that is that the profit motive will drive the private sector to a lower quality of service and/or a lower quality of design. (32) Indeed, some argue that the very structure of all P3 arrangements creates incentives that make it advantageous for the private sector to reduce costs and optimize revenues, even if this negatively affects levels of service or causes the project to cost more than it would have with public ownership and normal procurement procedures. (33)

In truth, it is difficult to understand why this should necessarily be the case. In competitive industries, the quality of service can be a key driver of financial success. A P3 in a competitive environment is subject to the same market pressures to service quality. In a monopoly, the market pressures to higher service quality do not exist, although the same argument in principle can be made about governments as a monopoly service provider. Nonetheless, [C]arefully crafted service and quality standards in a concessionary P3 contract, combined with effective oversight, provide the public sector with the power to clearly define and control the levels of quality and service required of its private-sector partners. Penalty clauses and, in the extreme case, the right to terminate the contract, can be used by the public sector as a discipline on service quality. For the private sector, its profits in this context are earned through the introduction of "sound business techniques and practices, ranging from improvements in management efficiency, application of new technologies, cash flow management, personnel development and shared resources" (35) and not through service quality reductions.

Other methods can be used to obligate equal or better service standards. British Columbia has dealt with the issue of quality of service by statutorily mandating an equal or higher level of service from private-sector operators of public services. For example, the Transportation Investment Act outlines the rules for transportation P3s and provides that a concession agreement must meet or exceed the standards applicable to a comparable public highway, including design, construction, safety, maintenance and signage standards. (36)

In addition, well-drafted cancellation clauses can be used to protect the public interest. In Ontario, two large jails were built around the same time, with one of the two jails being run by the public sector and the other operated by a private-sector company under a five-year contract. At the end of the contract term, the newly elected Ontario government (which was critical of the deal while in opposition) did not renew the company's contract and returned the prison to the public sector penalty-free. (37) Of course, having decided to return the jail to the public sector, the government was obligated to hire 470 new staff to operate the facility. (38)

The broader experience of P3 prisons suggests, however, that Ontario's conclusion may not be the correct one. Early involvement of the private sector in prison facility design can lead to significant improvements in construction costs and, potentially, staffing requirements. By focusing on delivering a high-standard design and construction solution, it is often possible to reduce prison operating costs over the life of the project, thereby reducing net costs. (39) The Ontario experiment, while attempting an innovative design, did not do so with early private-sector involvement. The government decided to contract-out operation of one of the two jails as the facilities were being built. As a result, some opportunities for risk transfer were clearly foregone.

This view is supported by experience elsewhere. For example, a review of nine PFI prisons in the United Kingdom conducted by the Office of the Auditor General concluded that PFI prisons tended to be more cost-efficient and better than public prisons in areas relating to decency and purposeful activities for prisoners but marginally weaker in areas such as safety and security. (40) This may be expected, as there can be a trade-off between these two sets of service criteria. (41) As the National Audit Office report made clear, the success of the PFI model rests on a combination of clear contractual service standards and effective monitoring of compliance, including, where appropriate, the use of penalties. (42) It also concluded that competition between public and P3 prisons enhanced performance in both methods.

In the result, there is no consistently compelling evidence of lower-quality design or service as a result of using the P3 model. Incentives to quality of design and service can be contractually created in the P3 context so that, at the very least, there is an equivalency between the public and private sectors. (43)

Less accountability, more secrecy

This criticism links a failure in accountability to the secrecy that can surround P3s. Lewis Auerbach, for one, argues that P3 disclosure standards must include, at the very least 1) comparisons of the cost and non-cost advantages and disadvantages of the relevant alternatives with the use of appropriate comparators, 2) the request for proposals, 3) the terms of the contract, if awarded, 4) the monitoring and audit regime if the project proceeds, and 5) ongoing access to and audit of the relevant performance and financial information of the private-sector partners. (44)

In fact, commentators on both sides of the P3 issue agree with this or a similar standard of disclosure as a key foundation for an accountable process. (45) Partnerships BC, for example, has issued policy guidance on its disclosure obligations, outlining the twin goals of the procurement process, as follows: to disclose as much as possible in the public interest without jeopardizing the ability of the government to generate the best value for taxpayers. (46)

To allow an adequate sharing of information in a form useful for citizens to hold governments to account on "best value for money" for P3 projects, three key standards have been developed: the public-sector comparator, value-for-money audits and, as indicated above, "best practice" standards for disclosure of information. The "public-sector comparator" (PSC) is simple in concept: gather a realistic and detailed assessment of all of the costs of the proposed project, including delay and budgetary risks, inflation effects, lifecycle costs, finance charges, operating costs, etc., and, based on a net present value, derive a public-sector cost of the project against which the price of a P3 model of delivering the same project can be compared. (47)

Partnerships BC, for example, has adopted the PSC model and obligates its use through the three-step procurement process outlined in its Capital Asset Management Framework. The initial PSC includes a preliminary assessment of life-cycle costs, subtracting any revenue, to provide a quantification of the true cost to the government through traditional procurement. It also includes an identification of material risks associated with the desired output of the project. The PSC is then refined through each step of the capital asset decision-making process: strategic options analysis, business case and procurement. The PSC is used to decide whether proceeding by way of a P3 is of net benefit to the government and then is again used, as refined through the process, to assess the bids. Although there are significant debates surrounding the discount rate applied to private-sector options and the "optimism bias" used to more accurately assess public-sector cost calculations, there is little doubt that the principle of a public-sector comparator is appropriate. Infrastructure Ontario also uses a similar model. (48)

In order to further enhance accountability, P3 projects ought to be subject to publicly available value-for-money assessments at three critical stages: 1) at the point of selecting an appropriate procurement methodology; 2) at the point of assessing P3 bids; and 3) at appropriate junctures during the concessionary contract. (49) Both Partnerships BC and Infrastructure Ontario have adopted this approach. For example, the Sea-to-Sky Highway Improvement project in British Columbia was initially reviewed by Partnerships BC on a value-for-money basis in December 2003 (prior to the selection process); the review was updated in December of 2005 and the reports were independently assessed by the provincial auditor general. A comparative value-for-money assessment was also conducted of the public-sector comparator and the selected bid using variables such as capital costs, operations and maintenance costs, rehabilitation costs, risk adjustment (including cost overruns, delay, etc., based on the relevant risk allocation), adjustments for tax status and payments, and the net present value of the unitary charge payments to the winning bidder. In addition, the report assesses what it calls expected user-benefits based on safety factors, service benefits and increased maintenance standards that will arise as a result of investments made under the winning bid. (50)

Infrastructure Ontario uses a similar process to enhance accountability for its P3 model, promising to deliver value-for-money reports for each project within six months of financial close. The Hopital Montfort P3 project, for example, was reviewed on a value-for-money basis, and the results were posted on the Infrastructure Ontario web site. (51) Using a similar methodology and providing similar information as does the B.C. process, the report concludes that Ontario's version of the P3 model saved the public money. When these measures are combined with "best practice" disclosure standards, such as those B.C. has put in place, most of the transparency and accountability concerns related to the project award phase can be addressed.

One accountability challenge that remains relates to the continued monitoring of the project during the concessionary period and the performance of the private-sector partner in meeting existing and evolving service standards. To satisfy transparency concerns during the operations phase, the following information should be provided: 1) public reporting of performance measures, including but not limited to penalties for poor performance; 2) the structure of the mechanisms for complaints and redress or forums for involving the public; and 3) information about any re-tendering of part of a P3 contract. (52)

These mechanisms are currently being used in a number of jurisdictions to ensure that the operator meets, and is seen as meeting, evolving service standards through appropriate contractual mechanisms. In a well-drafted P3 concessionary contract, the private sector is responsible for recording and disclosing performance failures and actively monitoring performance across all services. Significant penalties attach to the failure to carry out such monitoring or disclosure. In addition, "benchmarking" and "market-testing" provisions in P3 contracts allow the public sector (and, in some cases, the private sector) to make price adjustments to the annual unitary charge. Benchmarking involves preparing a benchmark report on the price of providing services under the concessionary contract against the price of providing comparable services in comparable facilities. Market-testing aims to re-base the price after testing them in the market through a procurement-like mechanism. A recent survey of U.K. public-sector managers who supervise P3 contracts shows them to be reluctant to use benchmarking mechanisms for fear that it will lead to pressure for higher compensation to the private-sector contractors. (53) This suggests, at least anecdotally, that the bid process has been effective in obtaining competitive pricing for these P3 services. Nonetheless, the result of these mechanisms, if shared publicly, can allay accountability and transparency concerns related to longer-term concessionary contracts.

One final accountability issue that is often raised against P3s can also be addressed. Potential bidders lobbying public officials during the bidding process can create a real and a perceived problem for the fairness of the bidding process. This is easily prevented through anti-lobbying policies that disqualify bidders who attempt to lobby public officials. Infrastructure Ontario's standard form request for proposals includes a prohibition against lobbying public officials and Infrastructure Ontario to influence the bid process. A breach of this obligation can lead to disqualification of a bidder's proposal.

Threat to workers' rights

The Canadian Union of Public Employees, a strong opponent of P3s, takes the position that they invariably result in reduced service as a result of reduced staff complements. They also allege that P3s are subject to higher employee turnover and lower wages. (54) Given the general practice in most jurisdictions that the private sector is obligated to offer employment to all displaced public-sector employees on the same terms and conditions, this criticism seems largely misplaced.

That said, flexibility in work arrangements, including innovative forms of compensation such as incentive pay, bonuses and profit-sharing, may be part of the mechanism the private sector will wish to use to enhance efficiency and provide better service. However, even in these cases, the public-sector expertise can be a valuable asset and employment opportunities will often be provided. As a result, even in the absence of a successor employer obligation, there is no compelling evidence of large job losses as a result of moving to a P3. Various U.S. studies of privatization initiatives have concluded that most displaced employees find employment relatively quickly or are transferred to the new entity. (55)

As a practical matter, Ontario P3 deals include provisions obligating the private sector to hire public-sector employees on the same terms and conditions as outlined by any existing collective agreement or employment contract. Ensuring a smooth transition to a P3 project and maintaining public support for it creates significant incentives to early and beneficial agreements with employees who are being transferred. (56)

Loss of public policy flexibility

The loss of public policy flexibility comes in three different forms: reduced expenditure choices as a result of long-term financial commitments in P3 arrangements; reduced service and policy choice options as a result of locked-in commitments; and, finally, the threat of trade repercussions as a result of private-sector involvement in previously publicly delivered services. As for the loss of expenditure flexibility, the anti-P3 thesis is that long-term contractual commitments to private-sector partners for services mean that scarce public resources are pre-committed for "decades to come." (57) On the asset side, the criticism is unwarranted. This amounts to an argument that governments should have the "freedom" to defer maintenance or not account for depreciation in order to have funds for other purposes. However, with accrual accounting, governments are not permitted to take steps like deferring maintenance in order to free up resources for other, short-term political needs. Indeed, there is an argument that the failure to build new and to properly maintain existing infrastructure has had a significantly negative impact on Canada's productivity. (58) Anything that enhances improved investment in and maintenance of infrastructure is beneficial in public policy terms.

On the operations side, there can be a trade-off between flexibility in service provision and the long-term contractual certainty that P3s using a project finance structure require. Flexibility increases the risk profile and hence the price. As a result, highly changeable environments may not be well suited to longer-term P3 arrangements. (59) Alternatively, public-sector break options can be included in concessionary agreements that would allow the public sector to terminate a P3 contract at specific points and pay predetermined levels of compensation to the private sector. In other cases, the demand risk for the service can be contractually allocated to the private sector (for example, in toll roads and other fee-for-service models) and thereby any expenditure risk is largely avoided.

The policy flexibility argument is closely related to the expenditure one. A CUPE policy paper on Ontario's infrastructure process opposes P3s in part because public service "is immensely more flexible than a long-term P3 contract. A change in public policy or the introduction of a new technology can lead to a change in service delivery when it is appropriate and without huge penalties levied by a private corporation for re-opening a contract." (60)

However, for technologically induced change, the cost-savings that can be achieved through new technologies are much more likely to be utilized in the private-sector delivery model than in a public-sector model, since the incentives for reducing costs and enhancing service, as discussed earlier, are evident in the private sector but less than compelling in the public sector. In fact, public-sector rules and employment obligations can often result in a significant disincentive to new public expenditures on technology.

As for the loss of policy flexibility, that should be at least in part the subject-matter of the initial decision by the public sector to select a P3 as its preferred model. In other words, in deciding to offer a twenty-five-year concessionary agreement, the government is concluding that service demands will justify a service period of that length. The Government of Ontario, for example, in assessing various procurement models, reviews whether the private sector should be involved when "clearly definable and measurable output specifications (i.e., service objectives) can be established, which are suitable for payment on a services-delivered basis." (61) In other words, assessing the service demand and level is part of determining whether locking in a longer-term concessionary agreement is sensible.

In any event, the building and management of infrastructure seems to be an unlikely candidate for significant policy change by a government. While policies can change, once governments invest in an asset or a service, they do not often back away and this is particularly true of core areas such as infrastructure. In fact, government itself is moving to long-term planning horizons for infrastructure assets, recognizing that it must do so as a matter of sound public policy and that, in so far as asset management is concerned, it will need to in order to adequately account for its treatment. (62)

That said, some policy flexibility will be lost as the contractual arrangements governments enter into will bind them such that compensation may be required if the government chooses to cancel a P3 contract. Of course, this is true of any contractual arrangement between the government and the private sector. In other words, governments will almost certainly be bound by the agreements they enter into, subject to their terms, unless they pay compensation. Efforts to avoid this outcome will be interpreted harshly by the courts and will result in significant controversy for the government.

The most famous example involves the former Lester B. Pearson International Airport, in Toronto, where the federal government entered into an agreement with T1T2 Ltd. Partnership (T1T2) to finance, design, build, develop and operate terminals 1 and 2. After making the arrangement an election issue, the then newly elected prime minister cancelled the contracts, even though the contractual arrangements contained no cancellation clause. The government then introduced legislation into Parliament that declared the contracts to be of no force and effect, denied the plaintiffs access to the courts, and provided for all existing legal entitlement to compensation from the Crown to be negated. The Senate objected to the legislation, the House reaffirmed it and returned it to the Senate. In advance of the legislation being passed, T1T2 asked the court to declare the government in breach of contract and for a reference for a trial on the damages, if any, owed. The motion was granted and upheld on appeal. (63) In the result, facing a judgment that it had breached a contract and a storm of protest over legislation that was argued to be expropriation without compensation, the government settled with T1T2 for approximately $60 million in 1997. (64)

What this saga illustrates is that contractual arrangements can limit government's policy flexibility but only to the extent of payment of compensation. Equally, however, it also shows that smart governments can protect their interests through careful drafting of the P3 contract (by including a cancellation clause, for example).

However, a government's programmatic flexibility can be inhibited as a result of the contractual obligations the public sector agrees to undertake within the terms of a concessionary agreement. In other words, the government can agree to limit its authority and/or to exercise it in support of the concession. Again, this would seem to be a matter of risk allocation and a judgment that the public sector would make at the beginning of the P3 process to consciously limit its rights. While this limitation may not always suit the short-term political goals of a subsequent government, decreasing the arbitrariness of government action can be a good thing.

For example, in 2003, the Government of Ontario refused to invalidate vehicle permit renewal applications for those who had not paid tolls on Highway 407, which was operated by a private consortium, the 407 ETR Concession Co. (407 ETR). The government was in the middle of a dispute with the 407 ETR over its rights to raise tolls without prior approval of the government (65) and, at the same time, had been working with the 407 ETR to improve the computerized tolling system to prevent false readings of license plates leading to mistaken requests for plate denials. An independent auditor appointed by the government audited the 407 ETR and concluded that the 407 ETR had complied with its obligations. Nonetheless, the Registrar of Motor Vehicles, who administers the process of plate denials, refused to process 407 ETR's requests. On that very same day, the government withdrew its approval of the auditors of the 407 ETR on the basis that they were not independent.

The 407 ETR challenged the Registrar of Motor Vehicles' failure to process 407 ETR's requests for plate denial and won. (66) The decision has two implications for the role of governments in a P3 context. First, the decision makes clear that some of the limitation on government action, in this case in the form of the registrar, arises from the provisions the government itself put in place through the governing legislation. In other words, the fact that the highway was a private venture was not, in and of itself, the reason that the registrar's role was circumscribed. Nonetheless, an effective enforcement mechanism to ensure tolls are paid is an obligation that the private-sector operator would require from government under the P3 contract.

This can be seen in the second lesson this case illustrates for P3s. The registrar also argued that the province's decision to remove the auditor eliminated a condition precedent to the exercise of his statutory duty. The court rejected this argument, noting that the 407 ETR's contract obligated the parties to continue to perform their respective obligations while a dispute was pending. (67) Therefore, until the dispute was resolved, the contractual provisions governed, and the government could not arbitrarily prevent 407 ETR from operating. In other words, the contractual provisions entered into can be an effective limit on the discretion of the government. That said, it is hardly an earth-shattering proposition. The key for governments is to ensure that they have the expertise and attention to detail to consider their ongoing role as they enter into P3 contracts.

Finally, the last critique of P3s in this category centres on the risk that private-sector involvement in the delivery of public services might create through trade agreements. (68) Opponents of P3s argue that the North American Free Trade Agreement (NAFTA), and Article 1110 in particular, creates the potential threat that foreign investors, unhappy with their treatment by a government in Canada, could bring an application challenging that treatment. The effect, they argue, would be to limit the range of public choices available to government and force private-sector delivery of public goods and services. (69) While there is no case law supporting any trade concerns, (70) nor, any NAFTA trade law case dealing with P3s in Canada, analysis of analogous case law and of the wording of the NAFTA and other relevant trade agreements suggests the risk is small to non-existent, and moderately effective drafting of the P3 contract can largely eliminate it.

First, NAFTA tribunals have held that investors cannot seek international arbitration for mere contractual breaches. (71) It is standard that investment and trade treaties, including NAFTA, provide that a government cannot expropriate without compensation. A right of the parties to compensation upon termination is a standard clause in a P3 agreement, and any contractual dispute about its scope and effect will not give rise to a NAFTA claim. One NAFTA tribunal in fact reached this very conclusion. (72)

Secondly, international law is clear that governments do not have to compensate investors for economic injuries that are the consequence of nondiscriminatory, bona ride regulations. Obviously, there can be circumstances where what looks like a legitimate exercise of government power is in fact an expropriation of private property. However, while there were some worrisome words in one NAFTA case to the effect that expropriation under NAFTA includes covert or incidental interference with the use of property, which has the effect of depriving the owner of the use or economic benefit of property, (73) it only applied to government action where there was no contractual relationship with the aggrieved private-sector participant. In other words, if the government has a contract with a private-sector party, whatever risk there is of NAFTA limiting the scope of governmental action is significantly lessened as the private sector has had a direct opportunity to negotiate its rights and obligations; there is no covert "taking." A P3 arrangement, therefore, adds little risk and, in fact, may mitigate it in some circumstances.

However, even this small risk was limited in a decision by a different NAFTA Chapter 11 tribunal, which dealt with the argument that the refusal by the Mexican government to rebate excise taxes on exported cigarettes was expropriation. (74) This case again dealt with circumstances where there was no direct contractual relationship with the claimant. In its decision, the tribunal held that governments must be free to act in the broader public interest, such as the protection of the environment, new or modified tax regimes and the granting or withdrawal of government subsidies, and that reasonable governmental regulation of this type will not give the private sector the right to seek compensation. (75)

Furthermore, there are exemptions to a broader application of NAFTA. For example, Article 1114 of NAFTA provides that nothing in Chapter 11 shall be construed to prevent a party from taking measures to ensure that investment activity in its territory is undertaken in a manner sensitive to environmental concerns. (76)

Thirdly, while NAFFA obligates governments to generally treat foreign investors from a treaty partner to the same standard as domestic investors, the obligation only applies in like circumstances. In other words, NAFTA does not obligate all services to be delivered in the same way and, therefore, does not obligate governments to deliver the service using a P3 methodology in the future. Even those who argue that P3s put public services at risk recognize that the mere fact of private-service provision does not give rise to a potential NAFTA challenge."

Finally, NAFTA provides member states the right to take unbounded reservations to the application of its terms in certain circumstances. Canada has reserved the right to take measures regarding income security or insurance, social security, social welfare, education, public training, health and child care, to the extent they are social services established or maintained for a public purpose. (78)

Ensuring successful P3s

What the arguments for and against P3s indicate is that achieving the enhanced value for money that is at the core of the case for P3 procurements rests on three variables: the nature of the project itself; a government that exercises effective project and contract management skills; and clear and effective risk allocation.

The nature of the project

Not all projects are ideally suited to a public-private partnership. The value for money generated by a P3 rests on clear and accountable incentives and on an optimal risk allocation, (79) by which is meant the measurement and minimization of risk by the party best able to do that. (80) Not all risks are necessarily best handled by the private sector, but a P3 will more likely succeed if the commercial risks inherent in it are transferred to the private sector. The risk profile of a project is more likely suited to a P3 where 1) there can be real scope for innovation in design and service delivery, 2) there is a definable revenue stream attached to a discrete service (and hence a feedback loop from pricing to service), 3) there is a substantial potential for synergies so that the design, building, operations and maintenance can be considered together to maximize efficiencies, and 4) there is real potential for risk transfer to the private sector. (81)

Not all of these criteria need to be present for a P3 to be the preferred methodology, but the more of them that are present, the more likely the project will be successful. (82) The last criteria, however, is a key one.

Given that these criteria are by no means always present, one would expect P3s to be a significant procurement option but just one among a range. Not surprisingly, most governments view them this way. (83) For example, Ontario and British Columbia intend to use P3s for only about ten per cent of planned capital investment. (84)

Effective project and contract management skills

To maximize the likelihood that a P3 project will be successful, governments must effectively manage risk from the beginning to the end of each project, and that requires keeping three key fundamentals at the forefront of the policy and planning process: managing public support; developing, maintaining and implementing a business plan; and developing and maintaining a centralized expertise.

Support

Effectively managing a P3 requires that the government create and maintain support for it within government and in the community and the private secton For the government, that means providing an appropriate legislative framework, clear lines of responsibility within government and a fair, consistent, transparent and accountable process. (85) It also means developing support in the community and among key stakeholders, such as unions and interest groups. For the private sector, that means ensuring that the market is as close to a competitive one as possible so that the efficiencies and innovation that P3 procurement promises can be realized. Governments can assist in creating greater competition by providing a strong and predictable pipeline of P3 deals and maintaining a credible and transparent procurement process. While not all services will be delivered in a competitive context, an effective procurement process, combined with effective contract monitoring, can introduce the benefits of competition to the pricing of the good or service.

Business plan

Effective project management requires that a business plan include reliable information, reasonable public-sector comparators, output specifications, risk analysis and a procurement strategy marked by fairness and transparency. In order to assess whether a project is an appropriate one for a P3, risk transfer or allocation must be consciously assessed at each point in the government's decision-making process. (86) That hinges on understanding the true costs of government procurement (including delay costs, cost overrun risks, life-cycle costing, etc.), a detailed understanding of the need for the service and of the service quality criteria, and, above all else, appropriate measurement and allocation of risk. With clearly specified outputs, an optimal risk allocation between the parties and a sensible public-sector comparator, private-sector bids produced through a fair and competitive bid process can be judged against a benchmark to assist in selecting a cost-efficient P3. (87) Public-private partnerships are neither inherently good nor bad; the key to performance lies in effective implementation and "whether there is clear accountability for results, clear criteria in contracts and clear public objectives." (88)

Expertise

A lack of contracting expertise can be a significant problem for governments with limited P3 experience. Individual government departments or smaller sub-national entities, such as municipalities, often cannot achieve relevant economies of scale and therefore learn by doing on each project, leading to inappropriate risk transfers and opportunistic behaviour by private-sector bidders and partners. (89) Managing a P3 project and the appropriate allocation of risks, therefore, requires two clear things from governments: 1) a sophisticated understanding of the legal, technical and financial aspects of the project; and 2) a "depoliticization" of the decision-making process at relevant points (and, concomitantly, a move to a standardized process), except where key overarching policy decisions, such as the choice to build an asset or deliver a service, are required.

To do this effectively, the government needs the requisite independent expertise to make a sensible value-for-money comparison, to run a fair and effective procurement process, to ensure an effective allocation of risk, and, finally, to effectively monitor and enforce contractual compliance. Contracting challenges for P3s are significant, and, as a result, many jurisdictions have created specialized agencies to review proposals and lay out contract terms for P3s. These groups often function as "within-government consultants on P3s, and as repositories of knowledge and experience that provide governments with the skills they need to structure P3s to their maximum benefit." (90)

As an example, Infrastructure Ontario follows this model, building on principles articulated by the Ontario government in its governing policy statement, Building a Better Tomorrow: 1) protection of the public interest; 2) value for money; 3) appropriate public control/ownership; 4) clear lines of accountability, transparent and rigorous reporting and oversight and clear measurable performance standards; and 5) fair, transparent and efficient bid processes with contractual agreements based on clear, comprehensive guidelines and public disclosure. (91)

Infrastructure Ontario has also developed expertise in project development, finance, construction and law to equal that of any private-sector proponent by having the resources and approvals necessary to hire external legal, technical, project management and financial advisers. With this expertise at hand, Infrastructure Ontario serves as the government's centralized agency for procuring assets and services, under Ontario's AFP model, for all government departments. This limits opportunistic behaviour by the private sector when dealing with inexperienced public servants in each department. The relevant public-sector managers from the line departments are at the table to ensure service and output standards are clear and followed, but they do not conduct the RFP process or negotiate with the private sector; that is conducted by the staff and advisers of Infrastructure Ontario.

Infrastructure Ontario also uses a highly standardized procurement process, with a request for qualifications, a request for proposals, including an opportunity to review and comment on proposed contractual arrangements, a fairness monitor, detailed bid criteria, transparent rules, and ongoing efforts at consultation and information exchange with the private sector.

As the deal flow increases, both the private and public sectors are becoming familiar with the process and project documentation and risk allocation is becoming standardized such that transaction costs are being reduced. The net result is a more efficient and effective procurement process that is more likely to achieve value for money for the public purse.

Reduced uncertainty and effective risk allocation

By enhancing project certainty and clearly and effectively allocating risks, the suitability of a P3 will be enhanced and its pricing optimized. When governments can effectively measure the risk and, once measured, make considered judgments as to how to allocate risk to the party best able to manage it, (92) value for money can then be achieved.

The Ontario Ministry of Public Infrastructure Renewal, for example, has published an assessment guide for public-sector managers in managing and allocating risks. The guide advises that decisions "related to which risks an entity will retain and which it will transfer will dictate to some degree which financing and procurement model an entity may use to develop its infrastructure initiative." (93) For the most part, P3 projects will start from the assumption that most commercial risks are to be allocated to the private party and most regulatory risks to the public party, with the sharing of additional risks. (94) In this way, the public interest should be well protected.

Conclusion

What this article has tried to illustrate is that generating value for money based on an effective risk transfer for a P3 requires the right project and significant government expertise. It also requires a clear-eyed view as to what risks should be borne by whom. Getting that equation right is fundamental to ensuring that a P3 is the right vehicle for delivering a public service. The analysis leads to four principles that should guide a public official in this determination:

1. Ensure that the services to be provided respond to a clear public need and can be clearly identified and measured.

2. Ensure that the public sector has the expertise to assess and manage risk.

3. Ensure that the partnership can deliver the high-quality, efficient and responsive services through an optimal allocation of risk.

4. Ensure that there are clear lines of accountability and redress.

When these are met, the only reason not to pursue a P3 is failure of political will.

Notes

(1) There is a significant, often politically motivated, debate about what properly constitutes a P3. While privatization opponents often conflate the two, P3s and privatization are not the same. For this article, P3s are characterized by two key components: 1) a risk allocation to the private sector that permits efficiencies in life-cycle costing through the overlap of design and operation; and 2) private financial participation.

(2) Different jurisdictions have used different names. Ontario uses "alternative financing and procurement" (AFP) and England uses "private finance initiative" (PFI). For ease of reference, we will use the term P3.

(3) Canada, Public Sector Accounting Group of the Canadian Institute for Chartered Accountants, Guideline to Accounting for and Reporting Tangible Capital Assets (Ottawa: Public Sector Accounting Group, 2007), available at http://www.psab-ccsp.ca/download.cfm?ci_id= 37536&la_id=1&reid=0.

(4) D. Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada (Toronto: TD Economics, 2006), p. 15.

(5) Ontario, Ministry of Public Infrastructure Renewal, Building a Better Tomorrow: An Infrastructure Planning, Financing and Procurement Framework for Ontario's Public Sector (Toronto: Queen's Printer, 2004), p. 27.

(6) Donald G. Pierce, Ira J. Berg and Cristina Alaimo, The Latest in Delivery Methods in Canada: Design Build, Public-Private Partnerships and More ... (Toronto: Goodmans, 2003), p. 11, available at http://www.goodmans.ca/docs%5CThe_Latest_Delivery_Methods in Canada_Don_Pierce.pdf.

(7) W.D. Eggers and T. Startup, Closing the Infrastructure Gap: The Role of Public-Private Partnerships (New York: Deloitte Research, 2006), p. 7.

(8) United Kingdom, Treasury Taskforce, How to Construct a Public Comparator. Technical Note No. 5, quoted in G. Allen, "The Private Finance Initiative (PFI)." Research Paper, Economic Policy and Statistics Section, House of Commons Library, December 2001, p. 27.

(9) Mott MacDonald Report, Review of Large Public Procurement in the UK (July 2002), p. 14, cited in Peter Fitzgerald, Review of Partnerships Victoria Provided Infrastructure. Final Report to the Treasurer (Melbourne: Growth Solutions Group, 2004), available at http://www.gsg.com.au/ pdf/PPPReview.pdf.

(10) United Kingdom, National Audit Office, PFI: Construction Performance (London: NAO, 2003); Mott Macdonald Report, Ibid.

(11) R. Andres, "The Infrastructure Deficit: Current Challenges." Paper presented at the Second Annual Summit on the Future of Canada's Infrastructure, Toronto, 26 September 2006, p. 2.

(12) See, for example, Vining Aidan, Anthony Boardman and Finn Poschmann, "Public-private partnerships in the U.S. and Canada: There are no 'free lunches,'" Journal of Comparative Policy Analysis 7, no. 3 (September 2005), pp. 199-220; Steven Globerman and Aidan Vining, "A framework for evaluating the government contracting-out decision with an application to information technology," Public Administration Review 56, no. 6 (November/December 1996), pp. 40-6; Joan Price Boase, "Beyond government? The appeal of public-private partnerships," CANADIAN PUBLIC ADMINISTRATION 43, no. 1 (Spring 2000), pp. 75-92; C. Mylvaganam and S. Borins, If You Build It: Business, Government and Ontario's Electronic Highway (Toronto: University Centre for Public Management, 2004); and other recent academic studies focusing on a series of projects, such as the Confederation Bridge and Highway 407, that largely pre-date efforts by Canadian governments to significantly improve their contract management and risk assessment and allocation skills. These new agencies have attempted to learn the lessons of prior P3s and draft contract provisions that manage and allocate transaction costs to deal, inter alia, with projects where failure rates may be higher--that is, ones with high asset s pecificity, high complexity and/or uncertainty and low competitiveness.

(13) See, for example, Vining, Boardman, Poschmann, "Public-private partnerships in the U.S. and Canada," Journal of Comparative Policy Analysis.

(14) See Roy Hrab, "Privatization: Experience and prospects." Panel on the Role of Government in Ontario, University of Toronto, June 2003. Research Paper # 22, p. 10, available at www.law-lib.utoronto/investing/about.htm.

(15) Ibid.

(16) Fitzgerald, Review of Partnerships Victoria Provided Infrastructure, p. 17.

(17) See Allen, "The Private Finance Initiative" [Research paper], p. 31, quoting Arthur Andersen and Enterprise LSE, Value for Money Drivers in the Private Finance Initiative. Report commissioned by The Treasury Taskforce (London: The Treasury Taskforce Limited, 2000).

(18) United Kingdom, Partnerships UK, Report on Operational PFI Projects (London: Partnerships UK, 2006), pp. 12-14.

(19) Eggers and Startup, Closing the Infrastructure Gap, p. 9.

(20) Lewis Auerbach, Issues Raised by Public Private Partnerships in Ontario's Hospital Sector ([Toronto]: Canadian Union of Public Employees, 2002), p. 19, available at http://www. healthcoalition.ca/cupe3p.pdf.

(21) H. Kitchen, A State of Disrepair: How to Fix the Financing of Municipal Infrastructure in Canada (Toronto: C.D. Howe Institute, 2006), p. 11. In fact, J-E. de Bettignies and T. Ross argue, after a comprehensive review, that it is not at all clear that governments can borrow at a lower cost than the private sector. See de Jean-Etienne Bettignies and Ross Thomas, "The economics of public-private partnerships," Canadian Public Policy 30, no. 2 (June 2004), pp. 135-54.

(22) See Globerman and Vining, "A framework for evaluating the government contracting-out decision with an application to information technology," Public Administration Review. Globerman and Vining rightly argue as part of a "positive" theory of P3s that an accurate comparison must include all of the transaction costs including the incremental cost of negotiating with and monitoring the performance of the private partner.

(23) See David G. Duff, "Benefit Taxes and User Fees in Theory and Practice." Research Paper 45, submitted to the Panel on the Role of Government in Ontario, 2003, available at http:// www.law-lib.utoronto.ca/investing/reports/rp45.pdf. The underlying principle is straightforward: those who benefit from infrastructure and the services it provides should pay for it. The ability to set correct prices depends on the asset and whether specific beneficiaries are easily and discretely identifiable.

(24) Elizabeth Woodman, "The market for financing of infrastructure projects through public-private partnerships: Canadian developments," Financial System Review (June 2006), p. 36, available a t http://www.bankofcanada.ca/en/fsr/2006/reports.pdf.

(25) R.W. Poole Jr. "A response to critics of toll road leasing," Public Works Financing 205 (May 2006), p. 3.

(26) B. Flyvbjerg, M.K. Skamris Holm and S.L. Buhl, "How (in)accurate are demand forecasts in public works projects: The case of transportation," Journal of the American Planning Association 71, no. 2 (Spring 2005), pp. 131-46; and see also Aidan Vining and Anthony Boardman, "Public-Private Partnerships in Canada Theory and Evidence." UBC P3 Project Working Paper 2006-04.

(27) United States, National Council on Public-Private Partnerships, Creating Effective Public-Private Partnerships for Buildings and Infrastructure in Today's Economic Environment [White Paper] (Washington, D.C.: NCPPP, 2005), p. 6, available at http://ncppp.org/resources/papers/ HDRP3whitepaper.pdf.

(28) United Kingdom, National Audit Office, PFI: Construction Performance (London: NAO, 2003); and the Mott MacDonald Report, Review of Large Public Procurement in the UK, cited in Fitzgerald, Review of Partnerships Victoria Provided Infrastructure.

(29) Vining, Boardman and Poschmann, "Public-private partnerships in the U.S. and Canada," Journal of Comparative Policy Analysis; and Globerman and Vining, "A framework for evaluating the government contracting-out decision with an application to information technology," Public Administration Review, make the case that a key factor in reducing transaction costs in P3s rests on the contract management skills of governments. As governments build specialized procurement vehicles for P3 arrangements, such as Partnerships BC and Infrastructure Ontario, these skills are being quickly developed.

(30) n.a., "Ceremony opens new $132M Royal Ottawa hospital," [CBC on-line news story] CBC.ca (Toronto: Canadian Broadcasting Corporation, 2007), available at http://www.cbc.ca/canada/ ottawa/story/2006/10/27/royal-ottawa.html.

(31) R. Bain and J.W. Plantagie, The Anatomy of Construction Risk: Lessons from a Millennium of PPP Experience (New York: Standard & Poor's, 2007). Some of the public debate on the cost-effectiveness of a P3 fails to distinguish between a project that is "on budget" and a project where bid price exceeds the initially budgeted price (which some Canada Line critics have done). In the latter case, all that has happened is that the private-sector price has exceeded the public entity's budget allocation for the project. In that case, the key calculus remains value for money. If the bid price exceeds the public-sector comparator, the project should not proceed as a P3.

(32) See, for example, n.a., "Private-Public-Partnerships (P3s) and the Transformation of Government," Ontario Federation of Labour Policy Papers (Toronto: OFL, 2005), available at http:// ofl.ca/uploads/library/policy_papers/P3s.pdf; Auerbach, Issues Raised by Public Private Partnerships in Ontario's Hospital Sector, p. 29; n.a., A CUPE Backgrounder on Urban Infrastructure (Ottawa: Canadian Union of Public Employees Research Branch, 2004), p. 19; and n.a., Re-Building Strong Communities with Public Infrastructure: A Submission to the Ontario Minister of Public Infrastructure Renewal in Response to the Discussion Paper on Infrastructure Financing and Procurement--"Building a Better Tomorrow: Investing in Ontario's Infrastructure to Deliver Real, Positive Change" (Toronto: Canadian Union of Public Employees--Ontario Division, 2004), p. 7.

(33) Auerbach, Issues Raised by Public Private Partnerships in Ontario's Hospital Sector, p. 29.

(34) de Bettignies and Ross, "The economics of public-private partnerships," Canadian Public Policy, pp. 144-45

(35) Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada, p. 16.

(36) Transportation Investment Act, S.B.C. 2004, c. 44.

(37) Ontario, Ministry of Community Safety and Correctional Services, "Central North Correctional Centre Transferring to Public Sector Operation: Private Jail Operation Contract not Renewed" [Press Release, 27 April 2006] (Toronto: Queen's Printer, 2006), available at http:// ogov.newswire.ca/ontario/GPOE/2006/04/27/c1439.html?1match=&lang=_e.html.

(38) Ontario, Ministry of Finance, Budget 2007 (Toronto: Queen's Printer, 2007), p. 113.

(39) C. Smith, "Prisons," in N. Avery, ed., Public-Private Partnerships (St. Paul, Minn.: Globe Publishing, 2006), pp. 89-90.

(40) United Kingdom, National Audit Office, The Operational Performance of PFI Prisons: Report by the Comptroller and Auditor General HC 700 Session 2002-2003: 18 June 2003 (London: NAO, 2003), available at http://www.nao.org.uk/publications/nao_reports/02-03/0203700.pdf.

(41) Smith, "Prisons," in Avery, Public-Private Partnerships, p. 90.

(42) Sir John Bourn [National Audit Office[, "'The Operational Performance of PFI Prisons." Report by the Comptroller and Auditor General, HC 700 2002-2003. National Audit Office Press Notice, 18 June 2003 (London: NAO, 2003), available at http://www.nao.org.uk/pn/02-03/ 0203700.htm. A well-drafted P3 contract will encourage the kind of behaviour the supervisory public authority prefers--either greater security or more purposeful activities for prisoners.

(43) See, for example, R. Harding, Private Prisons and Public Accountability (Buckingham, U.K.: Open University Press, 1997).

(44) Auerbach, Issues Raised by Public Private Partnerships in Ontario's Hospital Sector, pp. 16-17.

(45) Tim Gosling, "Openness Survey Paper." London: Institute for Public Policy Research, 2004, available at http://www.ippr.org.uk/press/files/opennesssurveyfinal.pdf, cited in D. Cartlidge, Public Private Partnerships in Construction (London: Taylor & Francis, 2006), p. 84. Joan Price Boase too is rightly critical of the lack of transparency in some early Canadian P3 deals. See Boase, "Beyond government," CANADIAN PUBLIC ADMINISTRATION.

(46) Partnerships BC, Procurement Related Disclosure for Public Private Partnerships (Vancouver: Partnerships BC, 2007), p. 2, available at http://www.partnershipsbc.ca/pdf/ pbc-disclosure_guidance-25-apr-07.pdf.

(47) Cartlidge, Public Private Partnerships in Construction, pp. 139-40.

(48) See Ontario, Infrastructure Ontario, Assessing Value for Money (Toronto: Queen's Printer, 2007); and Ontario, Infrastructure Ontario, Value for Money Assessment: Hopital Montfort Expansion and Redevelopment Project (Toronto: Queen's Printer, 2006), available at http:// www.infrastructureontario.ca/en/projects/health/montfort/files/Montfort%20 Value%20for%20Money%20-%20Final.pdf.

(49) Arthur Andersen and Enterprise LSE, Value for Money Drivers in the Private Finance Initiative.

(50) Partnerships BC, Project Report: Achieving Value for Money Sea-to-Sky Highway Improvement Project (Vancouver: Partnerships BC, 2005. The Partnerships BC web site is at http:// www.partnershipsbc.ca.

(51) The Infrastructure Ontario web site is at http://www.infrastructureontario.ca/en/index.asp.

(52) Cartlidge, Public Private Partnerships in Construction, p. 87.

(53) Partnerships UK, Report on Operational PFI Projects (London: Partnerships UK, 2006), p. 85.

(54) n.a., CUPE Backgrounder on Urban Infrastructure, pp. 19-20.

(55) Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada, p. 16.

(56) Canadian Council on Public-Private Partnerships, Transitioning Staff from Public to Private Organizations in Public-Private Partnerships: A Guide to the Successful Handling of Collective Bargaining Issues (Toronto: CCPPP, 2001).

(57) Auerbach, Issues Raised by Public Private Partnerships in Ontario's Hospital Sector, p. 42.

(58) Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada, p.16.

(59) Avery, Public-Private Partnerships, p. 72.

(60) n.a. [CUPE], Re-Building Strong Communities with Public Infrastructure, p. 13.

(61) Ontario, Ministry of Public Infrastructure Renewal, Building a Better Tomorrow Framework: An Infrastructure Planning, Financing and Procurement for Ontario's Public Sector (Toronto: Queen's Printer, 2004), p. 22, available at http://www.pir.gov.on.ca/ English/infrastructure/framework.htm.

(62) ibid., pp. 16-17.

(63) T1T2 Ltd. Partnership et al. v. The Queen in Right of Canada (1995), 23 O.R. (3d) 66 (Ont. Gen. Div.); T1T2 Ltd. Partnership et al. v. The Queen in Right of Canada (1995), 23 O.R. (3d) 81 (Ont. Gen. Div.); T1T2 Ltd. Partnership et al. v. The Queen in Right of Canada (1995), 24 O.R. (3d) 546 (Ont. C.A.).

(64) Canada, Department of Justice, "Settlement Paves Way for Progress at Pearson International Airport." News release, 16 April 1997 (Ottawa: Public Works and Government Services Canada, 1997), available at http://www.justice.gc.ca/en/news/nr/1997/pearsn.html.

(65) See 407 ETR Concession Co. v. Ontario (Minister of Transportation) (2004) O.J. 373 (Ont. Div. Ct.); and Ontario (Minister of Transportation) v. 407 ETR Concession Co. (2005) CarswellOnNt. 36 (Ont. S.C.J.).

(66) 407ETR Concession Company Limited v. Registrar of Motor Vehicles (2005), 82 O.R. (3d) 703 (Ont. Div. Ct.).

(67) Ibid., pp. 716-17.

(68) This section is based in part on a legal opinion filed by the Canadian Council for Private-Public Partnerships at the Walkerton Inquiry, Submission to the Walkerton Inquiry, Part 2, Legal Opinion (Toronto: CCPPP, 2001), available at http://www.pppcouncil.ca/pdf/ fretrade.pdf; and also Canadian Council on Public-Private Partnerships, Public-Private Partnerships and Trade Agreements (Toronto: CCPPP, 2003).

(69) Steven Shrybman [Sack Goldblatt Mitchell], Public-Private Partnerships: Assessing the Risks Associated with International Investment, and Services Treaties (Ottawa: CUPE, 2002), p. 2, available at http://cupe.ca/updir/P3s & Trade Agreements .doc.

(70) Canadian Council for Public-Private Partnerships, Submission to the Walkerton Inquiry, Part 2, Legal Opinion.

(71) Robert Azinian and others v. United Mexican States, ICSID Case No. (AF)/97/2, (NAFFA Tribunal, 1999), cited in Canadian Council for Public-Private Partnerships, Submission to the Walkerton Inquiry, Part 2, Legal Opinion, p. 11.

(72) Ibid.

(73) United Mexican States v. Metalclad Corp. (2001), 89 B.C.L.R. (3d) 359 (B.C.S.C.), para. 99.

(74) Marvin Roy Feldman Karpa v. Mexico, ICSID Case No. ARB (AF)/99/1, (NAFFA Tribunal, 2003) at para. 103, available at http://www.state.gov/documents/ organization/16639.pdf.

(75) Ibid., at para. 103.

(76) North American Free Trade Agreement between the Government of Canada, the Government of Mexico and the Government of the United States of America, 17 December 1992, Can. T.S. 1994 No. 2, 32 I.L.M. 289 (entered into force January 1, 1994 ("NAFTA"), Article 1114.

(77) Shrybman, Public-Private Partnerships, p. 27: "At least two commentators ... argue that the participation of private service providers should not prove fatal to this reservation, so long as public funding is preserved."

(78) Annex II to NAFTA.

(79) Allen, "The Private Finance Initiative (PFI)." [paper], p. 30.

(80) de Bettignies and Ross, "The economics of public private partnerships," Canadian Public Policy, p. 139.

(81) It is important to distinguish risk from uncertainty. Risk is used here to denote known or knowable risks that, as a result, can be priced. Uncertainty, in the sense that the risk cannot be known or calculated at the time of the bid, presents real challenges to planning for both traditional procurement and the P3 model. It is the informational vacuum, not complexity per se, that can create project failure particularly in the area of revenue-dependent projects. See, for example, Vining, Boardman and Poschmann, "Public-private partnerships in the U.S. and Canada," Journal of Comparative Policy Analysis, p. 212.

(82) Cartlidge, Public Private Partnerships in Construction, pp. 51-2; and Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada, p. 23.

(83) Eggers and Startup, Closing the Infrastructure Gap, p. 25.

(84) See Woodman, "The market for financing of infrastructure projects through public-private partnerships: Canadian developments," Financial System Review, p. 36; and Ontario, Ministry of Public Infrastructure Renewal, ReNew Ontario 2005-2010: Strategic Highlights (Toronto: Queen's Printer, 2005), p. 3.

(85) United States, National Council on Public-Private Partnerships, Creating Effective Public-Private Partnerships for Buildings and Infrastructure in Today's Economic Environment [White Paper], p. 21; United States, General Accounting Office, Report to the Honorable Stephen Horn, Committee on Government Reform, House of Representatives--Public-Private Partnerships: Key Elements of Federal Building and Facility Partnerships (GAO/GGD-99-23) (Washington, D.C.: GAO, 1999), p. 5, available at http://www.gao.gov/archive/1999/gg99023.pdf; G. Banfai, T.A. Barry, R. Beaumont and L. Roth, "Construction Risk in Public-Private Partnerships in Canada." Proceedings, 9th Annual Conference, Canadian College of Construction Lawyers, 2006, pp. 1-2; Residential and Civil Construction Alliance of Ontario, The Infrastructure Funding Deficit: Time to Act (Toronto: RCCAO, 2006), pp. 12-13, available at http://www.rccao.com/research/files/Rccao_InfraFundDeficit-Jun06.pdf.

(86) Partnerships BC, An Introduction to Risk Management in a Public Private Partnership (Vancouver: Partnerships BC, 2006), p. 4, available at http://www.partnershipsbc.ca/ pdf/riskmanagement-ppp-28-jul-06.pdf.

(87) United States, Government Accounting Office, Public-Private Partnerships, p. 5.

(88) United States, National Council on Public-Private Partnerships, Creating Effective Public-Private Partnerships for Buildings and Infrastructure in Today's Economic Environment [White Paper], p. 8.

(89) Vining, Boardman and Poschmann, "Public-private partnerships in the U.S. and Canada," Journal of Comparative Policy Analysis, p. 25. See also Burleton, Creating the Winning Conditions for Public-Private Partnerships (P3s) in Canada, p. 18.

(90) de Bettignies and Ross, "The economics of public private partnerships," Canadian Public Policy, p. 145.

(91) Ontario, Ministry of Public Infrastructure Renewal, Building a Better Tomorrow Framework, p. 9.

(92) Allen, "The Private Finance Initiative (PFI)." [paper], pp. 29-30.

(93) Ontario, Ministry of Public Infrastructure Renewal, Building a Better Tomorrow Framework, p. 31.

(94) Canadian Council on Public Private Partnerships, Position Paper: Building a Better Tomorrow through Public-Private Partnerships. Submission to the Ontario Government's Discussion Paper on Infrastructure Financing and Procurement (Toronto: CCPPP, 2004), p. 5

The author was chief of staff to Prime Minister Paul Martin and now practices law at McMillan Binch Mendelsohn in the P3 area. He is also an adjunct faculty member of the Faculty of Law, University of Toronto, where he teaches the law and policy of public-private partnerships.