Introduction

A partnership between the state and the private sector is sometimes used to finance, design, construct, operate, and maintain state infrastructure projects (such as highways, mass transportation systems, and state buildings). Such a P3 requires the state and the private sector to collaborate when making decisions about a project. By bringing external resources and specialized expertise to a project, the state is expected to achieve certain benefits from a P3 that typically are not achievable when using a more traditional, public sector procurement approach.

In recent years, California has entered into P3s with private partners for two state infrastructure projects. Specifically, to build and operate the Presidio Parkway transportation project in San Francisco (also known as Doyle Drive) and to build and maintain a new courthouse in Long Beach. Each of these agreements is for a period of about 30 years. The combined estimated cost to the state for both of these projects is about $3.4 billion. Given their significant cost and the limited experience the state has had in such partnerships, we identify in this report best practices for the state to follow when using P3s and present recommendations for maximizing the benefits to the state. In preparing this report, we met with representatives from various state departments about their experiences with P3s, as well as numerous P3 experts. We also reviewed the literature regarding best practices for implementing such partnerships.

Background

State Procures Infrastructure Projects in Various Ways

State law specifies the processes for reviewing and approving proposed state infrastructure projects. In most instances, such as for court facilities and state office buildings, the Legislature must approve and appropriate funds for a state project. (They also must receive approval from the State Public Works Board at various subsequent stages.) However, the California Transportation Commission (CTC) approves most state projects related to transportation. The review and approval process for state projects generally involves determining (1) the need for the project, (2) how the project fits into existing infrastructure systems, (3) the project's priority relative to other state infrastructure projects, and (4) how the project will be funded. This process typically consists of multiple reviews and approvals as a project's development moves from concept to environmental review and preliminary engineering, and then to design and construction.

State law specifies three general types of procurement approaches—design–bid–build, design–build, and P3—that state departments can use to deliver infrastructure projects that have been approved for construction. As summarized in Figure 1, each of these procurement approaches involves varying ways of contracting with the private sector for a project. We discuss each approach in more detail below.

Design–bid–build. State departments can use a design–bid–build approach to procure all types and sizes of infrastructure. Under this approach, the work for each stage of a project is performed separately. For example, a state department will generally first award an architectural/engineering contract to design the project based on subjective criteria of qualifications and experience of the architect/engineer. In some cases, however, state staff may design the project. After detailed project plans and drawings are completed, the department then selects a contractor to perform the construction work. Construction contracts are awarded objectively based on competitive bidding, with the contract going to the qualified bidder who submits the lowest price. Once construction is completed, the state department is responsible for operating and maintaining the facility. The state pays for the cost of design–bid–build projects either up front with state funds or over time by selling general obligation bonds or lease–revenue bonds.

Design–build. Under existing state law, certain departments can use design–build procurement. With design–build, the department typically contracts with a private general contractor to both design and build the infrastructure project. The department does not separately contract with an architect/engineer for design. Rather, the general contractor is responsible for subcontracting with other entities for design and various construction work. The state awards a design–build contract through a competitive bidding process that evaluates factors such as price, design features, construction schedule, and community or environmental outcomes. Under design–build, the state maintains responsibility for financing, operating, and maintaining the project.

Alternatively, another type of design–build involves the state transferring design and construction risks to a specialized construction manager, rather than a general contractor. This approach is commonly referred to as "construction manager at risk procurement." With construction manager at risk, the state awards a contract based on a fee. The construction manager designs the project and solicits bids from subcontractors and suppliers. The sum of these bids, along with a surcharge, determines the total price the state pays for the project.

P3s. Current state law authorizes three state departments—Caltrans, AOC, and the High–Speed Rail Authority (HSRA)—to use some form of a P3. While there can be varying degrees of P3s, the type of P3 often discussed and most recently used in California is when a single contract is entered into with a private partner (often a consortium of several companies) for the design, construction, finance, operation, and maintenance of an infrastructure facility. For the purpose of this report, we generally define a P3 as the contracting with the private sector to design–build–finance–operate–maintain an infrastructure project. (As we discuss in the nearby box, the state can also enter into partnerships with other public entities, such as counties, for the procurement of state infrastructure.)

Under a P3 approach, the state can transfer a significant amount of responsibility associated with a project to the private sector. For example, the private partner will generally make design and construction decisions and be responsible for paying the costs to resolve any construction issues in order to ensure that the project is completed on time. In addition, the partner will often be required to finance the project, which generally includes the costs of design and construction staff, materials, and construction equipment. However, in order for a private partner to be willing to finance these costs, the contract must specify a mechanism for repaying the partner. In many cases, this involves a revenue source created by the project (such as a toll or user fee on the infrastructure facility), with the private partner taking on the risk that the projected revenues will materialize at the level anticipated. Alternatively, the state can commit to making annual payments to the partner from an identified funding source, such as tax revenues. Since it can take many years for a revenue source (such as a toll on a road) to pay off the private financings, the terms of P3 contracts generally range between 25 years to 100 years.

The P3 procurement approach is typically more complex than design–bid–build and design–build. For instance, under a P3 approach, the state must first evaluate a pool of potential bidders to determine if they have the qualifications necessary to design, build, finance, operate, and maintain the infrastructure facility. Then, qualified bidders submit proposals that the state evaluates in order to select a preferred bidder. A P3 contract is often awarded to the bidder deemed to provide the best value.

Three State Departments Authorized to Use P3s for Certain Projects

As shown in Figure 2, existing state law authorizes the use of P3s for certain transportation and court construction projects. (State law also authorizes certain local governments to use P3s for local infrastructure projects.) Below, we discuss in more detail the specific P3 authority provided to state departments.

Caltrans. Chapter 107, Statutes of 1989 (AB 680, Baker), authorized Caltrans to enter into P3 agreements for up to four projects. Under this authorization, as well as that provided in related follow–up legislation, Caltrans built ten miles of tolled express lanes in the median of the existing State Route (SR) 91 in Orange County. In addition, the department built SR 125 in San Diego County that connects the area near the Otay Mesa border crossing with the state highway system. For each project, Caltrans used a single contract with a private partner to design, construct, finance, operate, and maintain the facility. (We discuss these two projects in more detail later in this report.)

In 2009, Caltrans' authority to enter into P3 agreements was expanded. Specifically, Chapter 2, Statutes of 2009 (SB 2X 4, Cogdill), authorizes Caltrans and regional transportation agencies (such as the Los Angeles County Metropolitan Transportation Authority) to enter into an unlimited number of P3 agreements for a broad range of highway, road, and transit projects through December 31, 2016. However, the legislation specifies that such P3 projects must achieve one or more objectives as determined by the CTC, which is responsible for programming and allocating funds for the construction of highway, rail, and transit improvements. These objectives include:

• Improve travel times or reduce vehicle hours of delay.
• Improve transportation operation or safety.
• Provide quantifiable air quality benefits.
• Meet a forecasted demand of transportation.

In addition, the above agreements are subject to a 60–day review by the Legislature and PIAC before Caltrans can sign them. The PIAC is an advisory commission created by Chapter 2 and chaired by the Secretary of the Business, Transportation, and Housing (BT&H) Agency. Specifically, PIAC is charged with assembling research, best practices, and lessons learned from transportation P3s around the world. The commission can, upon request, assist Caltrans and regional transportation agencies with P3 project selection, evaluation, procurement, and implementation. Currently, PIAC consists of about 20 volunteer members and is staffed within existing BT&H Agency resources.

In January 2011, Caltrans entered into its first P3 under Chapter 2 for the Presidio Parkway project. This particular P3 requires the private partner to complete the second phase of the design and reconstruction of the southern approach to the Golden Gate Bridge and to operate and maintain the roadway for 30 years. In exchange, the state will make payments estimated to total roughly $1.1 billion to the private partner over the life of the contract.

Judicial Council and AOC. Under current law, the judicial branch is authorized to use P3s. In addition, state law requires the Judicial Council (the policy making body for the judicial branch) and their staff in the AOC to develop performance standards to facilitate the review of P3s and requires the Department of Finance (DOF) to review projects that include a P3 component. The legislation also specifies that AOC may only proceed with a P3 if the Legislature does not object to the performance standards adopted for the project.

The 2007–08 Budget Act directed AOC to gather information regarding the possible use of a P3 for the replacement of the Long Beach courthouse. In December 2010, AOC entered into a P3 that requires a private developer to finance, design, build, operate, and maintain the Long Beach courthouse over a 35–year period in exchange for payments from the state totaling $2.3 billion. At this time, the Long Beach courthouse is the only project that the AOC has procured using a P3.

The HSRA. Chapter 796, Statutes of 1996 (SB 1420, Kopp), created the HSRA and authorized it to use P3 procurement for the development of a High–Speed train system connecting northern and southern California. However, state law does not establish a specific process for reviewing or approving P3s for HSRA. State law requires capital expenditures by HSRA to be approved by the Legislature. Based on the authority's 2012 business plan, the HSRA would not award its first P3 contract until 2023.

Benefits and Limitations of P3s

Government entities typically use P3s to achieve benefits that they may not be able to obtain under a more traditional procurement approach (such as design–bid–build). However, P3s can also introduce new limitations and costs as summarized in Figure 3 and described in detail below.

Potential P3 Benefits

Transfers Project Risks to Private Partner. The P3s can transfer risks associated with a project from a government entity to a private partner. Figure 4 summarizes the major risks that could potentially be transferred, such as those related to financing, operation, and maintenance. As indicated in the figure, the most significant risks are associated with the design and construction of a project. For example, under a P3 approach, the private developer would bear the risks and costs if the design of the project were changed to fit certain site conditions (such as soil quality or the discovery of archeological artifacts). Similarly, the private partner would be responsible for project cost overruns, which can be very expensive. The transfer of this risk could reduce or eliminate the need for additional public funds to complete a project. Moreover, the partner would bear the risk if the actual revenue collected from any tolls or user fees are less than projected, which depending on the project, can be significant. In order to ensure adequate compensation, private developers attempt to estimate the anticipated costs of resolving issues on the risks they assume and factor these costs into their bid. However, in some cases, the developer may be better equipped to manage certain risks at a lower cost than if the government retained all of the project risks.

Greater Price and Schedule Certainty. Based on a survey conducted by the Federal Highway Administration (FHWA), governments around the world reported that P3s can provide better price and schedule certainty for the design and construction of a project compared to a more traditional procurement approach (such as design–bid–build). In part, this is because P3s allow a government entity to share certain risks with a private developer who has more experience with a particular type of project and has developed strategies to mitigate potential cost increases that could result from such risks. The government can also achieve greater price certainty from P3s because, as is the case with design–build and construction manager at risk contracts, the contacts often have a maximum price. This means that the private partner must pay for any cost increases above the agreed upon price. In addition, the government typically sets up a streamlined process to review the design and construction decisions made by the partner, which can help prevent delays in the project schedule. Moreover, P3s that include financing can incentivize the partner to complete the project on time and receive the necessary funding (such as payments from the government or revenues from project user fees) to repay the private loans taken out to finance the project.

More Innovative Design and Construction Techniques. Experts in P3s generally believe that the private sector is often better able to develop innovative project designs and construction techniques than government entities. In part, this may be due to the specialized expertise that a private partner can bring to a project. Greater design and construction innovation could result in a variety of potential benefits, including lower project costs, a higher quality project, shorter construction schedules, and enhanced project features.

"Free Up" Public Funds for Other Purposes. In general, using a private developer's access to capital can free up government funds to advance the construction of other infrastructure in the near–term and, thus, provide the public with access to improved infrastructure sooner than planned. In addition, the developer's financing can sometimes provide more advantageous repayment terms than a government might typically obtain under a more traditional public financing approach. For example, if repayment were extended over a longer period of time than the government typically has to repay borrowed funds, it could reduce the amount that must be repaid each year. Such freed–up public funds could then be allocated for other purposes.

Quicker Access to Financing for Projects. In addition, by making a private developer responsible for financing a particular project, a government might be able to access financing in cases where it does not yet have the authority to borrow. For example, California's constitution requires voter approval prior to selling certain types of bonds to finance infrastructure projects, which could delay the state from accessing financing for certain projects. However, projects financed through a P3 would not require voter approval, potentially allowing some projects to start construction sooner.

Higher Level of Maintenance. Due to insufficient funding for maintenance, as well as how existing maintenance funding is prioritized, some governments currently do a poor job in maintaining their infrastructure. For example, due to a lack of regular maintenance, only 28 percent of California's highways are in good condition. As a result, many highways require costly major rehabilitation or replacement. Under a P3 approach, a government could require the private partner to maintain the constructed infrastructure to specified standards. Essentially, this means that P3 facilities could remain in good condition over longer periods of time, thus allowing the government to delay the cost of major rehabilitation or replacement.

Keeps Project Debts Off the Government's "Books." Another benefit of P3 financing is that the debt incurred by a private partner for a project may not be counted as government debt. In other words, P3 financing may not appear as debt on government balance sheets. According to recent studies by the FHWA and the United Nations Economic Commission for Europe, this is one of the reasons why some countries in the European Union have chosen to use P3s. While the benefit of debt not appearing on the government's balance sheet is probably more important to governments subject to strict limitations on debt, it could also improve the overall ability of some governments to borrow funds for other purposes. However, since California does not have such strict debt limitations that restrict its options for financing infrastructure projects this benefit does not currently apply to the state.

Potential P3 Limitations

Increased Financing Costs. Financing a project through a P3 is likely to be more expensive than the financing options typically used under the more traditional procurement approaches (such as obtaining state and federal loans). This is because private companies typically pay higher interest rates than government entities to borrow money. For example, a study of P3 projects in Canada found that private partners typically pay 1 percentage point higher on loans compared to the governmental cost of borrowing. In times of limited access to financial markets (such as the financial crisis of 2008), the cost difference between private and public borrowing was 2 percentage to 3 percentage points. In addition, private companies will often seek to earn a profit of roughly 10 percent to 25 percent when loaning funds to a government, which can further increase P3 financing costs.

Greater Possibility for Unforeseen Challenges. As previously discussed, in comparison to design–bid–build and design–build contracts, P3 contracts cover a much longer time period and scope of activities (such as maintenance of the infrastructure facility). Thus, there is a greater possibility for unforeseen issues to arise under a P3 approach. Such issues could include disputes regarding certain terms in the contract, as well as the private partner being acquired by another company or going out of business, effectively resulting in project schedule delays and additional costs to the government.

Limits government's Flexibility. The long–term nature of P3s can also "lock in" certain government funding priorities based on operational needs determined at the time the contract is negotiated. This can make it difficult to change funding allocations to reflect changes in government priorities. For example, a P3 contract may require litter and graffiti to be removed from a highway within three days. Renegotiating the terms of this contract to use the funds designated for prompt litter and graffiti removal to support another activity or project could be very difficult. In addition, by bundling multiple phases of a project into a single contract, P3s can make it more difficult for the government to change how a project is managed. For example, if the government wanted to make changes to how a private partner handled customer complaints and questions on a toll road, it would likely need to propose amendments to the contract, which could increase the project's overall cost.

New Risks From Complex Procurement Process. As discussed earlier, the procurement process for P3s is more complex than the procurement processes traditionally used for state infrastructure projects (such as design–bid–build and design–build). In addition to a project's design, construction price, and schedule, under P3 approach, the government entity must also evaluate proposals based on financing, operations, and maintenance. In addition, P3 procurements can also involve complex negotiations between the government and the private developers who bid on the project. As a result, P3s can require the government to perform new activities and take on certain risks that it may not be experienced at handling. For example, if the state does a poor job of drafting agreements or fails to address relevant issues in these agreements, it could experience unexpected costs or receive lower levels of service than planned.

Fewer Bidders. According to the research, infrastructure projects that are relatively expensive and complex tend to be more ideal candidates for P3s. In addition, private partners tend to be comprised of multiple companies who coordinate efforts to develop a P3 bid—each with expertise in a particular component of the project (such as design and construction, financing, or maintenance). As a result, few private developers have the financial resources and technical skills to compete for P3 projects, especially on their own. According to experts, P3 projects typically receive between one and three bids. In comparison, similarly sized projects procured under traditional public sector approaches typically receive a greater number of bids. Less competition for a project procured as a P3 could be detrimental to the government, as more competition for a project generally reduces the price of a project while increasing its quality.

Some Benefits Achieved to Date

In order to determine whether the state has achieved some of the intended benefits of P3s, we reviewed the two completed P3 projects procured by the state'sR 91 Express Lanes and SR 125 Tollway. At the time of this analysis, however, Caltrans was unable to provide us with the necessary data to evaluate whether the P3 projects completed by the state—SR 91 and SR 125—resulted in greater price and schedule certainty than if the projects were procured under a more traditional approach. As a result, we reviewed recent state projects that, while not considered P3s, transferred certain risks and responsibilities to the private sector. Specifically, we reviewed those projects procured under design–build and construction manager at risk contracts. These particular projects are summarized in Figure 5.

Price and Schedule Certainty Generally Achieved. In terms of the projects reviewed, most of them were generally successful at staying on budget and schedule. When there were schedule and cost overruns, they were typically relatively small. For example, three–fourths of the projects opened to users on schedule or within one month of the planned deadline. Three–fourths of the projects were also completed on budget or with less than 5 percent in cost overruns. While there is no way of knowing what the price and schedule outcomes would have been if these projects were procured differently (such as a design–bid–build project), the projects were generally successful at meeting the goal of price and schedule certainty.

Mixed Results During Operations. The SR 91 Express Lanes and the SR 125 Tollway both experienced problems during the operational phases of their P3 contracts. Specifically, while both facilities remained open to the public, Caltrans incurred additional costs resulting from disputes with its private partners. For example, the SR 91 contract contained a "non–compete clause" that prohibited Caltrans or other public agencies from competing with the tolled lanes built by the private partner. Thus, if public agencies made any improvements to transportation facilities in the SR 91 corridor (including minor projects to improve the safety of the general–purpose lanes of the highway), the private partner believed that the state would be required to compensate for the loss of toll revenue if fewer people drove on the tolled P3 lanes due to these improvements. The issue was litigated in court, but was ultimately settled when the private partner agreed to sell the rights of the express lanes to the Orange County Transportation Authority (OCTA), a local public transportation agency. Since OCTA assumed control of SR 91, Caltrans has not had any conflicts regarding the non–compete clause.

The SR 125 project experienced legal challenges that delayed its completion. Such challenges made the partnership less profitable for the private partner. Specifically, the lawsuit alleged that Caltrans, as a partner in the agreement, was partially liable for losses claimed by some of the private companies involved in the project. The private partner ultimately declared bankruptcy, with the court awarding the rights to the remainder of the P3 to a group of lenders who had financed the project. This group subsequently sold the agreement to the San Diego Association of Governments.

Given the nature of design–build and construction manager at risk contracts, the other projects we reviewed did not transfer responsibility for operating the infrastructure facility to a private partner after it was built. However, the toll road projects involving SR 73, SR 241, SR 261, and SR 133 did involve other public entities being responsible for operating the facilities. This network of tolled public highways in Orange County (commonly referred to as "The Toll Roads") are managed by the Transportation Corridor Agencies (TCAs), which are public agencies led by locally elected officials. While the TCAs may contract with private companies to operate the toll roads, they are ultimately responsible for making key decisions and directly managing the contracts. When the TCAs encountered problems after the highways became operational, they were able to end a contract with a private operator who was not meeting expectations and later rebid the work to a different company. Using separate contracts and retaining more responsibility for making key decisions helped avoid some of the unforeseen costs (such as legal costs) that were incurred with the state's P3 projects.

P3 Best Practices

As part of our examination of the P3 approach, we reviewed international research and interviewed experts in the field. Based on our review, we identified a set of best practices that have been found to maximize the potential benefits of P3s and minimize its potential limitations. These best practices are summarized in Figure 6 and discussed in more detail below.

Establish Overall P3 Policy and Implement Transparent Processes

Experts recommend that governments adopt an overall P3 policy to (1) guide decision–makers when evaluating different procurement options and (2) inform potential private partners and the public of the process. For example, experts recommend having a transparent process so that potential partners are aware of the specific requirements that must be satisfied to bid on a project and how long the procurement process will likely take. Such transparency also helps stakeholders and the public understand how and why a government entity selected a private company to build or operate public infrastructure. For example, Virginia created the Office of Transportation public–private Partnerships to develop a consistent institutionalized process for P3 procurements, in order to help attract qualified developers and contractors.

Adopt Criteria to Determine Good Candidates for P3 Projects

International research also finds that it is good practice for governments to adopt criteria for determining whether projects would be a good fit for P3 procurement, as not all public infrastructure projects would benefit from a P3 approach. For example, as we discuss later in this report, the Legislature could establish criteria that provide a reasonable means of screening potential P3 projects. Such criteria should not be too prescriptive or cumbersome. Experts recommend that the screening criteria include the following:

• Government Benefit From Using Nonpublic Financing. The screening process should determine if there is a benefit (such as completing the project sooner) to the government from financing the project with a private partner, rather than using public funds upfront to pay for the project. Typically, relatively expensive projects—with costs ranging from the hundreds of millions to billions of dollars—are more likely to benefit from private financing, as it can take several years to save up enough funds to build a large project without financing or to get approvals for public financing.
• Technically Complex. Generally, projects that are technically complex are more likely to benefit from the innovation or specialized expertise that is typically associated with P3s. For example, a private partner with extensive experience designing and building tunnels or bridges may be able to construct a complex tunnel or bridge more quickly and/or at a lower cost. On the other hand, projects that are very simple (such as repaving a road) are not as well suited for the P3 approach because they are less likely to benefit from innovation and specialized expertise.
• Ability to Transfer Risks to Partner. Projects that are good candidates for P3s generally have significant known risks that the government can transfer to a private partner. For example, a project that is in the very early stages of development and does not have a completed environmental review may lack sufficient information to allow for an effective transfer of risk. Given these unknown risks, potential partners may be hesitant to bid on the project or may incorporate large premiums into their bid. Alternatively, a project with clearly identified risks (such as if a toll road will generate enough revenue to finance the project) would be more well–suited as a P3.
• Revenue Source to Repay Financing. As discussed above, P3s require a revenue source to repay the financing provided by the private partner. Ideally, a project would have a dedicated revenue source (such as a toll or user fee) to repay the money borrowed from the partner. The government entity, however, could commit to make payments to the partner from government funding sources, such as tax revenues.

Conduct a Rigorous Value for Money Analysis

Once it is determined that a particular project is a good P3 candidate, experts recommend that the government entity perform a detailed analysis that compares the project's costs using a P3 to using a more traditional procurement approach. A commonly used analysis is a "value for money" (VFM) analysis, which identifies all the costs of a project (such as the design, construction, and operation and maintenance of the facility) over the life of the project or the term of the lease with the private partner. These costs are then "discounted" over time to determine the project's cost in net present value. In other words, because the expenditures take place over several decades and the timing of the expenditures differ between a P3 approach and the more traditional procurement approach, the comparisons are adjusted to account for the fact that money available at the present time is worth more than money available in the future. Specifically, the VFM analysis should compare the cost of the different procurement approaches in net present value terms of delivering the same level of service—both in terms of the quality of the infrastructure constructed and the quality of the maintenance and operation services provided.

The VFM analyses can be complex and the underlying assumptions can significantly influence the outcomes. Thus, most experts recommend specifying parameters for the assumptions (such as for the discount rate) so that all potential projects are evaluated with similar criteria.

Adopt and Implement Project Approval Process

Experts recommend maintaining a process to approve projects for P3 procurement that allows good candidates to proceed. The approving entity (such as the Legislature or an independent board), which is typically separate from the agency sponsoring the project, should verify that (1) the project satisfies most of the established P3 criteria and (2) the VFM analysis shows that a P3 procurement is the best option. In addition, P3 experts recommend obtaining project approval prior to having potential partners bid on the project. This is because private developers may not bid on a project if they are unsure whether the approving entity might stop it from moving forward.

Establish Government Expertise in P3s

Another P3 best practice is for government entities to develop expertise regarding P3s, in order to better protect public resources when entering into large contracts with private partners. Experienced departmental staff can make it easier for the state to handle P3 workload quickly and thoroughly, as well as effectively communicate with the private sector. Governments could use private consultants to help with this workload.

Our research also found that P3 expertise can reside at multiple levels of government. For example, PPP Canada provides information and assistance to Canada's provincial and municipal governments on the use of P3s. At the provincial level, Partnerships BC in British Columbia provides specialized services, such as managing projects and facilitating communication with the private sector. In addition, experts recommend reviewing the outcomes of P3 projects at various stages to allow a government entity to determine what worked well and what problems were encountered on each project. The lessons learned can be used to inform future P3 procurements.

State's Use of P3s Falls Short of Best Practices

As we discussed earlier in this report, existing state law authorizes Caltrans, AOC, and the HSRA to use P3s to procure certain types of infrastructure. In analyzing their use of this authority, we generally found that the practices of Caltrans and AOC are not necessarily aligned with P3 best practices. (At the time of this report, HSRA had not entered into any P3 contracts.) For example, our analysis indicates that these departments did not use clear P3 processes and appear to have selected projects not well suited for a P3 procurement—meaning the Presidio Parkway project and the Long Beach courthouse project. Our findings regarding each of these projects are summarized in Figure 7 and described in detail below.

State Lacks Transparent P3 Processes

As discussed above, having clearly defined and transparent P3 processes is considered a best practice. However, our review found that the state's use of P3 procurement for the Presidio Parkway and Long Beach courthouse projects lacked transparent frameworks and clear processes. For example, when Caltrans used a P3 procurement for the Presidio Parkway, the department lacked a transparent framework for selecting the project and conducting a VFM analysis. It did not release a draft P3 program guide until December 2011, one year after signing the agreement for Presidio Parkway. While the guide addresses many procedural questions regarding the department's future use of P3s, it does not establish a consistent process for evaluating potential P3 projects through the use of a VFM analysis. We think this is a significant shortcoming of the guide because establishing VFM processes and parameters is important to ensure that projects are evaluated on a consistent basis using reasonable assumptions. The Caltrans draft P3 program guide also does not address how project evaluation, review, and procurement responsibilities will be carried out when the state partners with local transportation agencies. Specifically, the guide does not lay out how the lead agency will be determined and which entity is responsible for certain tasks, such as review and oversight. As a result, various local agencies that we talked to appear to have different understandings of what will be required of them for P3 projects.

Similarly, AOC did not use a transparent framework in selecting the Long Beach courthouse to be a P3 project. For example, AOC did not develop guidelines for selecting potential P3 projects and conducting VFM analyses. More importantly, at the time of this report, AOC had not developed transparent criteria or processes for determining potential P3 projects in the future. For example, it is unclear how AOC will identify projects that are likely to benefit from a P3 approach and evaluate potential projects through the use of VFM analyses.

Selection Criteria for Recent Projects Not Aligned to Best Practices

Our analysis indicates that the processes used to identify the two recent state projects for P3 procurement—the Presidio Parkway and the Long Beach courthouse—included few of the best practice criteria.

Presidio Parkway Selection Was Problematic. According to Caltrans staff, the Presidio Parkway project was selected as a P3 candidate primarily based on two criteria: (1) an estimated project cost of more than $100 million and (2) a completed environmental impact review. However, according to the identified best practices, these two factors alone do not constitute a robust set of screening criteria. In other words, the selection process for the project did not include such recommended criteria as the ability to transfer risk to the private sector and whether the state would benefit from using non–state financing. While the selection process for a P3 project does not need to include all of the best practice criteria, including such criteria does help ensure that the intended P3 benefits are achieved. Our analysis indicates that if Caltrans utilized such criteria in its selection process, the Presidio Parkway project would have been found to be inappropriate for P3 procurement.

For example, the Presidio Parkway project was too far along to transfer many of the project's risks to a private partner. This is because the Presidio Parkway's first phase of construction was already underway using a design–bid–build procurement when the second phase of the project was selected for P3 procurement. As a result, potential private partners had limited access to the construction site, which in turn made them less willing to take on many of the project's construction risks. For example, the state retained significant risks regarding the discovery of archeological artifacts and endangered species. In addition, Caltrans had already designed about half of the project's second phase prior to awarding the P3 contract. Thus, the winning bidder may be limited in its ability to find cost–savings through innovative design and construction techniques because it must adhere to certain specifications it did not design.

Long Beach Courthouse Selection Was Problematic. According to AOC staff, the Long Beach courthouse project was selected as a P3 candidate based primarily on two criteria: (1) it was one of the largest court construction projects considered at that time and (2) the Long Beach area has a competitive market for the type of property management staff needed to operate a P3. Similar to the selection of the Presidio Parkway project, the selection process for the Long Beach courthouse project did not include much of the recommended best practice criteria. For example, the selection process did not evaluate whether the project is technically complex. While the ideal level of complexity for a P3 is difficult to define in specific terms, the Long Beach courthouse project lacks unique or complex features that would likely benefit from innovative design and construction techniques. Accordingly, our analysis indicates that if AOC utilized best practice criteria in its selection process, the Long Beach courthouse project would have been found to be inappropriate for P3 procurement.

VFM Analyses Based on Assumptions That Favored P3 Procurement

As described above, VFM analyses can help decision–makers compare the cost of a project under different procurement options. Both Caltrans and AOC contracted with private consultants to perform such analyses for the Presidio Parkway and Long Beach courthouse projects. Specifically, the analyses compared the costs of constructing the project under a more traditional approach to a P3 approach. The VFM analyses found that the state would benefit financially if the Presidio Parkway and Long Beach courthouse projects were procured as P3s—meaning it would be cheaper to have a private developer build and operate the planned facility. Our review of these particular analyses, however, indicates that both VFM analyses were based on several assumptions that are subject to significant uncertainty and interpretation and tended to favor a P3 procurement. If a series of different assumptions were made, the VFM analyses would have shown that the P3 procurement on the Presidio Parkway and Long Beach courthouse projects would be more expensive in the long run than a more traditional procurement.

Assumptions in Presidio Parkway Analysis Favored P3. Some of the key assumptions made by Caltrans in the VFM analysis of the Presidio Parkway project that tended to favor P3 procurement include:

Relatively High Discount Rate. In order to calculate the net present cost of the project, Caltrans' VFM analysis discounts the cost of the project under a traditional approach and a P3 procurement by 8.5 percent per year. As discussed above, this adjustment is intended to reflect that money spent in the near term is more valuable than money spent in the future. In the past, our office has suggested that a 5 percent discount rate be used for such analyses, but acknowledges there is no one "right" discount rate. We also note that the state's long–term borrowing rate is currently less than 5 percent.

Unjustified Tax Adjustment. The VFM analysis for this project also included a $167 million adjustment in order to account for increased tax revenues (such as from corporate taxes) that the private developer would pay to the state under the P3 approach. The analysis assumed that if the project was not procured as a P3, the state would not receive these additional revenues. However, we found the adjustment included mostly revenues related to potential federal taxes, which would not directly benefit the state. Thus, the adjustment made a P3 approach look more favorable than is warranted.

Assumed Early Payment of Cost Overruns. Under a more traditional procurement approach (such as design–bid–build), Caltrans assumed the Presidio Parkway project would exceed its budget by $125 million and that such cost overruns would need to be paid for at the start of construction. However, such overages do not typically occur at the start of a project, but rather as a project progresses through construction. While some consideration of the potential for cost overages is reasonable, Caltrans' method relies on subjective judgment rather than objective evidence. Consequently, the chosen method has the effect of overstating the net present cost of the project under a traditional procurement approach, thereby favoring a P3 procurement approach for the project.

Failed to Account for Competitive Bidding Environment. The Caltrans' VFM analysis, which was prepared in February 2010, also did not take into account the competitive construction bidding environment that occurred around that time. During this period, Caltrans awarded construction contracts that were on average 30 percent below the project's original cost estimate. While it is not possible to know exactly what the bids would have been if the Presidio Parkway project had been procured using a more traditional procurement, it appears reasonable to assume that the project could have been awarded at a much lower cost than the engineer's cost estimate.

Our analysis indicates that utilizing a different set of assumptions (such as a discount rate of 5 percent and excluding the assumed tax adjustment) would result in the cost of the Presidio Parkway project being less—by as much as $140 million in net present value terms—in the long run under a traditional procurement approach than the chosen P3 approach.

Assumptions in Long Beach Courthouse Analysis Favored P3. Some of the key assumptions in the VFM analysis of the Long Beach courthouse that tended to favor P3 procurement include:

• Unjustified Tax Adjustment. Similar to the Presidio Parkway project, the VFM analysis for the Long Beach courthouse project included a $232 million adjustment to account for increased tax revenues that would be paid for by the private developer under the P3 approach. A major component of this adjustment reflects revenues from federal taxes. Since additional federal tax revenues would not directly benefit the state, there appears to be little to no justification for increasing the cost of using a traditional procurement approach to reflect the federal taxes that would be paid by a private developer.
• Overstated Cost Overruns. The VFM analysis assumed that using AOC's more traditional procurement approach of construction manager at risk—rather than a P3 procurement approach—would result in construction cost overruns for the Long Beach courthouse project totaling $128 million (about 30 percent of the project's estimated cost). However, given that AOC has procedures in place to prevent such cost overages and has not experienced them with recent court construction projects, this assumption has the effect of overstating the cost of the project under a construction management at risk approach.
• Leasing of Additional Space. The AOC's VFM analysis assumes that under the P3 approach, the courthouse project would include space that would initially be leased by the private developers to other entities, but could eventually be used by the court. The VFM analysis also assumes this additional space would be needed by the court in Long Beach in the future, and builds the cost of leasing this additional space into its estimates. This factor adds $260 million in costs to a traditional procurement of the Long Beach courthouse project, but only $69 million to the cost of the P3. The higher cost under a traditional approach assumes that a separate building would be leased and that the leased building would need substantial modifications. The analysis for the traditional procurement also assumes increased costs for security officers to monitor the leased building. While there is some basis for estimating a higher cost for the potential need to lease additional space under a traditional procurement approach, the AOC has not conclusively demonstrated that all of this additional space would be needed by the court in Long Beach. Moreover, AOC's other courthouse construction projects ordinarily do not include this kind of extra space.
• Project Completion. The AOC–s VFM analysis assumes that it would take 14 months longer to complete the Long Beach courthouse under construction manager at risk procurement than as a P3 project. Accordingly, the analysis uses different timelines to discount the costs of the project under each type of procurement. The way the VFM analysis adjusts for these assumed differences in timing effectively increases the cost of a traditional procurement in net present value terms. However, it is not evident that such a procurement would necessarily take 14 months longer—especially in view of the considerable flexibility state law gives AOC with respect to its construction contracting methodology.

Our analysis indicates that utilizing a different set of assumptions than those discussed above (such as excluding the assumed federal tax adjustment and leasing costs) would result in the cost of the Long Beach courthouse project being less—by as much as $160 million in net present value terms—in the long run under a traditional procurement approach than the chosen P3 approach.

State Law Lacks Thorough Project Approval Processes

Our analysis found that for both the Presidio Parkway and Long Beach courthouse projects, the state did not utilize a thorough process for selecting P3 projects. Having thorough processes in place could have prevented Caltrans and AOC from entering into a P3 agreement for each project, or at least required changes to negotiate lower prices and better ensure that the intended P3 benefits are achieved.

For P3 transportation projects, state law requires the CTC to conduct a limited review of the basic features of each project sponsored by Caltrans or a regional transportation agency. (We note that in reviewing the Presidio Parkway project, CTC extended its evaluation beyond the basic requirements to further review the project's financing.) However, state law does not require the commission or another entity to conduct an overall review of whether (1) the state would benefit from procuring a particular project as a P3 and (2) whether a particular P3 contract is structured to maximize the state's benefits. Moreover, while state law does provide a 60–day period for the appropriate legislative fiscal and policy committees and PIAC to review P3 proposals before Caltrans can sign an agreement with a private developer, state law does not require that Caltrans address any of the concerns raised in these reviews.

For court construction projects, state law authorizes the Joint Legislative Budget Committee and DOF to review a potential P3 project before AOC can fully develop the project's concept. Accordingly, the Legislature reviewed and approved the general criteria used by AOC to select the private partner for the Long Beach courthouse project. However, the Legislature did not have an opportunity to review and comment on the VFM analysis before it was finalized and the contract was signed with the private developer.

State Lacks P3 Expertise

As previously discussed, experts recommend that government entities develop expertise regarding P3s in order to better protect public resources when entering into large contracts with private developers. Our review, however, finds that such expertise within state government has not been sufficiently developed in California.

PIAC Has Limited P3 Expertise. The PIAC was established in 2009 to assemble and share research on best practices and lessons learned from transportation P3s around the world. However, based on our discussions with staff at the BT&H Agency and our review of various PIAC documents (including the minutes from the seven PIAC meetings that have taken place), we find that PIAC has done little to implement best practices for transportation P3s. The only steps that PIAC appears to have taken in this regard are to post reports containing information on P3 best practices on its website and to contract for two reports on P3s. We also note that the commission currently lacks members with in–depth expertise on issues such as state financing, state procurement, and state labor issues. Perspectives on these issues could help to ensure that the state maximizes its benefits when using P3s.

No Systematic Approach for Reviewing Lessons Learned. Our review also finds that the state does not have a systematic process for identifying and applying lessons learned from prior P3 projects. Although Caltrans is the only state agency to have entered into multiple P3 agreements, it currently lacks a formal process for reviewing past P3 projects in order to maximize benefits and avoid repeating past mistakes. We understand that AOC is currently developing a review and reporting process for the Long Beach courthouse project. Once completed, these reports may provide helpful lessons learned about AOC's use of P3 procurement.

Recommendations to Maximize State Benefits from P3s

In this report, we reviewed the state's experience with P3s and identified several instances where the best practices identified in existing P3 research have not necessarily been followed. Based on our review and findings, we have identified several opportunities for the state to further maximize its benefits when deciding to procure a state infrastructure project as a P3. Our specific recommendations are summarized in Figure 8 and discussed in detail below.

Specify P3 Project Selection Criteria

As previously mentioned, the state's processes for selecting P3 projects are inadequate and not necessarily based on selection criteria identified in the research as best practices. Accordingly, we recommend that the Legislature adopt legislation requiring that each state department with P3 authority utilize certain criteria when evaluating whether a particular project should be procured as a P3. According to the research, these selection criteria should not be highly prescriptive, but rather should provide general guidance regarding the selection of potential P3 projects. Such an approach would provide for greater consistency across departments in terms of how P3 projects are selected. The selection criteria should include being a technically complex project, as well as a project that can transfer risks to a private partner and benefit from non–state financing. In addition, the Legislature may want to specify whether P3 projects must have a revenue source, such as a user fee.

Require Analysis of a Range of Procurement Options

In order to determine which procurement approach would most effectively benefit the state, we recommend that the Legislature adopt legislation requiring a comparative VFM analysis of a range of procurement options (including design–bid–build, design–build, and P3) for all potential P3 infrastructure projects. Evaluating a range of procurement options would allow the state to better balance the potential benefits of increased private sector involvement with the potential risks unique to each project. In contrast, the benefit of evaluating only two procurement approaches—as was done by Caltrans and AOC—can be limited. This is because it does not evaluate other options (such as design–build), which in some cases may be the best option.

We also recommend that the Legislature specify in statute that such VFM analyses:

• Exclude Federal Tax Adjustments. Increased federal tax revenues do not directly benefit the state and should not be included in a VFM analysis.
• Apply Costs to Expected Year of Expenditure. Project costs should be accounted for in the year they are likely to be incurred, in order to effectively estimate the project's likely total cost in the long run.
• Use Current Construction Cost Estimates. Construction cost estimates should be based on the current bidding environment in the state.
• Include a Sensitivity Analysis. A sensitivity analysis can help to indicate how the results of the VFM analysis might change with a different set of assumptions. Specifically, this analysis should evaluate project costs and revenues with a range of reasonable discount rates to show how differing assumptions can influence the outcome of the VFM analysis. If a project will generate revenue, such as from tolls or fares, a reasonable range of revenues should also be evaluated in the sensitivity analysis.

Modify Structure and Responsibilities of PIAC

In order to help ensure that PIAC effectively assembles and shares research, best practices, and lessons learned from transportation P3s around the world, we recommend the Legislature adopt legislation to:

• Expand PIAC's Authority. In order to provide a consistent review and approval process for the use of P3 procurement, we recommend expanding the PIAC's role to require the commission to approve all state P3 projects, as discussed in detail later in this report. We also recommend expanding the scope of PIAC to all types of infrastructure projects, rather than only those related to transportation. Having the commission involved in all types of P3 will further the state's P3 expertise. To reflect this broader scope, we also recommend making PIAC an independent commission, rather than part of the BT&H Agency.
• Direct PIAC to Evaluate Other Departments for P3 Authority. We have found that certain types of projects may benefit the state if procured using a P3. It is possible that state departments other than Caltrans, AOC, and HSRA will have projects meeting these P3 criteria. Accordingly, we recommend that the Legislature direct PIAC to review the types of projects planned by other state departments and recommend to the Legislature whether P3 authority should be granted to additional state departments.
• Broaden PIAC's Expertise. In order to ensure that PIAC has the expertise necessary to advise state departments on all types of P3s, we believe it would be beneficial for the commissioners to have a broad mix of expertise related to P3, as well as state finance and procurements. Specifically, we recommend that the Legislature appoint some of the commissioners and that, in addition to P3 experts, the commission include the Director of the Department of General Services (or a representative), and the State Treasurer (or a representative). The Legislature could also consider reducing the number of commissioners on PIAC to a more manageable size.
• Require PIAC to Develop and Implement Best Practices. We recommend requiring PIAC to (1) develop a set of best practices for P3 projects in California, (2) provide state departments specific steps for implementing those best practices, and (3) provide technical assistance to state agencies planning to pursue a P3. Consistent with the research, it would benefit the state to have such expert advice provided from the initial project screening stage through the procurement and administration of a P3 contract. We also recommend that the Legislature require periodic reports from PIAC in its efforts in developing and implementing P3 best practices.

Improve Consistency of state's P3 Approval Process

We recommend that the Legislature adopt legislation to make the process for reviewing and approving P3 projects consistent and thorough across those state departments authorized to pursue such projects. Specifically, we recommend requiring the use of the review and approval process summarized in Figure 9 and discussed in detail below.

Require PIAC to Approve P3 Concept and VFM Analysis. Our above recommendations to modify the structure and responsibilities of PIAC would make it well–suited to review and approve a department's proposed use of P3 procurement. As shown in Figure 9, we recommend that the Legislature require departments to provide a VFM analysis and other relevant project information (such as draft procurement documents) on all proposed P3 projects to PIAC. Under our recommended process, if PIAC identifies concerns with a P3 proposal, the commission would require the department sponsoring the project to perform additional analyses and resubmit the proposal for subsequent review. If a project does not satisfy the above P3 criteria, we recommend that PIAC have the authority to reject the use of a P3 approach, and direct the department to use another procurement method. Thus, we recommend that the Legislature adopt legislation directing PIAC to implement a process to evaluate (1) whether a P3 project proposal is consistent with the scope and cost approved in the state's current capital outlay processes (meaning either by the Legislature or CTC) and (2) whether using a P3 approach would be the best procurement option.

Conclusion

Based on our review of existing research, we believe that P3 procurement—if done correctly—has merit and may be the best procurement option for some of the state's infrastructure projects. In certain instances, sharing risks with a private partner and using a diverse financing package (including private loans) may even be the only way to build those projects that are both very complex and expensive. For such projects, the use of P3 procurement can make the price and schedule more certain by transferring various project risks to a private partner. In addition, access to specialized expertise and private financing could have the effect of accelerating projects and providing other benefits to the state. However, the state does give up considerable control over the management and long–term funding priorities of a project that is constructed under a P3 approach. This limitation and others must be considered carefully when considering a decades–long partnership.

We also find that implementing certain P3 best practices identified in the research can better ensure that the intended benefits of P3s to the state are achieved. Thus, in order to maximize the state's benefits from P3s, we recommend that the Legislature take a series of steps to ensure that such best practices are followed in developing and implementing future P3 projects. For example, we recommend specifying P3 project selection criteria and improving the state's approval process to utilize an entity with expertise in P3s. More importantly, our proposals to develop P3 expertise and better evaluate potential P3 projects would provide for a better understanding of the actual benefits and limitations of P3 projects. Finally, as the state gains experience with P3s, the Legislature may want to consider whether the existing P3 authorization provided to Caltrans and AOC should be expanded to other departments.