USDOT Federal Highway Administration USDOT Home | FHWA Home | Feedback
Skip to main content

Feedback Forum

Value for Money for Public-Private Partnerships: A Primer

September 11, 2012

DRAFT
For review by P3 Evaluation Toolkit Beta-Testers

« Previous | Table of Contents | Next »

Chapter 7 - Quantitative Value for Money Assessment

Developing the Public Sector Comparator

The PSC is expressed in net present value (NPV) terms and is based on the actual public sector method of procuring the project in question. This means that if the public sector would procure the project as a design-build then the design-build method is the procuring option to be considered in the PSC. The PSC also includes any reasonably foreseeable efficiency which the public sector could achieve and takes full account of the risks which would be encountered by that style of procurement.

The public agency typically uses financial and statistical modeling techniques to develop the PSC and the Shadow Bid for a project. For example, a Monte Carlo simulation, a commonly used mathematical modeling technique, uses statistical sampling to provide a range of estimates for the cost of risk for the quantitative assessment. These modeling techniques may assess a range of potential outcomes for the PSC and/or Shadow Bid.

During the development of a PSC, several assumptions are made, including the assumption that the public sector can complete the project to the same quality and standards anticipated by private sector delivery. As the PSC presents a baseline cost of whole-life project delivery for the government, it can also be a useful tool that assists governments in forecasting the full costs associated with conventional procurements.

Competitive Neutrality Adjustments

In order to calculate an appropriate competitive neutrality adjustment, the agency developing the PSC must identify those ways in which the public sector and a potential P3 private sector partner would be treated differently on the basis of their differing status as public and private-sector entities. Because competitive neutrality components will usually represent cost adjustments to the PSC, they will be discounted to a NPV just like other components. 

Taxation is the most obvious differential treatment. Taxes are costs to a private partner that ultimately result in revenues to the public sector. Public sector authorities are usually not subject to the same sales, payroll or property taxes that a P3 contractor would face. These differentials would require an increase to the PSC to represent a true "apples-to-apples" comparison. It might be possible to distinguish among the various levels of government to whom taxes are paid, so that taxes paid to the Federal Government could be treated differently from state and local taxes. It is an important policy decision whether or not to draw such a distinction. It is clear that the State and local governments will benefit from state and local taxes, but the benefits from taxes paid to the Federal government are more diffuse. Adjusting for Federal taxes paid by investors in commercial debt neutralizes the advantages of tax-exempt public debt. However, the tax advantage of public debt represents a real savings to the state or local agency issuing that debt, so the agency may decide not to make adjustments for Federal taxes paid by private investors.

An adjustment may be required with respect to insurance costs paid by a private entity if the risk that is being insured against is not already accounted for under retained risks. When the government chooses to self-insure, there is a perception that the government has saved on insurance premiums. In fact, the government is taking on risks otherwise covered by insurance, and the government should account for this additional risk. An adjustment is made to the PSC by adding an amount equivalent to the premium otherwise paid by the private sector under a P3 to protect against the additional risks.

Regulatory differences in treatment are also a significant category, but the benefits can go both ways. Public sector authorities are often subject to reporting, transparency and/or process requirements that can impose significant burdens in terms of labor and cost. Conversely, private sector entities often face more rigorous performance reporting and higher levels of scrutiny than a public agency would have faced for the same project.

Developing the Shadow Bid

A Shadow Bid is described as the estimated cost to the public sector if the same project were to be delivered by the private sector as a P3. A Shadow Bid is the public sector's estimate of the bid price that it may receive if the project is structured as a P3.
A PSC and Shadow Bid can be developed and compared (see Figure 2-1) during the initial project financial assessment and feasibility study, prior to determining the procurement method and issuing the solicitation.

A financial model is a tool that is used by the public agency to create the Shadow Bid. It attempts to predict the bidder's costs, financing structure and other assumptions.  Outputs of the model include upfront public subsidies needed, and the amount of toll revenue or availability payments required throughout the P3 term. Financial modeling is discussed in FHWA's primer on Financial Structuring and Assessment for P3s.

Comparing the Public Sector Comparator to Actual Bids

After bids are received in response to an RFP, the PSC may be compared to the actual bids received to assess if VfM is still achieved prior to awarding the contract as a P3.  
Figure 7-1 shows how the PSC and an Actual Bid are compared. The present value of payments to be made by the public agency to the private entity under the Actual Bid include compensation to the private entity for life-cycle costs (including costs incurred by the private sector for the procurement process), financing costs, and costs of risks to be transferred to the private entity, which are generally included within financing costs in the form of higher rates of return on debt and equity investment. 

Figure 7-1

Figure 7-1. Comparison of PSC and Preferred Bid

A P3 may offer better value for money if the total costs calculated by the preferred actual bid are less than the costs calculated by the PSC or the Shadow Bid. For toll-based projects, if the project does not generate sufficient revenue from tolls to cover the payments needed by the concessionaire, the public agency would need to make up the shortfall in order for the concessionaire to deliver the project, assuming that the agency still considers it advantageous to move forward with the project.

Table 7-1 presents an example of results from a VfM analysis.  As shown in the table, the raw PSC was first calculated based on life-cycle costs (discussed in Chapter 4), including capital, O&M and asset replacement costs. Costs of transferred risks were independently calculated (as discussed in Chapter 5). Retained risks were not included, since they were the same for both the PSC and the Shadow Bid. Finally, a cost was added to account for competitive neutrality. When total PSC costs are compared with the costs of the preferred bid, an AU$9 million savings was estimated, indicating that value for money was achieved. Note that the savings are marginal and amount to only about one percent of the total cost, i.e., within the margin of error that might be expected in this type of analysis. This is not unusual in quantitative VfM analyses, and often qualitative factors (discussed in Chapter 8) are more important in making the final VfM determination. Together, the quantitative and qualitative assessments inform the overall VfM analysis and decision-making process.

Table 7-1: Peninsula Link Project PSC
Components of the Public Sector Comparator (PSC)NPV
(AU$mil.)
Capital Costs $680
Lifecycle Asset Replacement Costs (25 years) $43
Operating Costs (25 years) $80
Raw PSC $803
Transferred Risks (Capital and Operating) $47
Competitive Neutrality $8
PSC (excluding retained risk) $858
Preferred bid total costs $849
Savings (Value for Money estimated) $9
Adapted from Partnerships Victoria Peninsula Link Project VfM Analysis (May 2010)

 

Submit your comments

 

« Previous | Table of Contents | Next »