A sound P3 policy would also go a long way to dispel the notion that partnerships should only be entered into where the cost of financing is cheaper. Public postmortems of P3 contracts often put a spotlight on the higher cost of financing under a private consortium compared to traditional public models, suggesting that the firms are profiting at taxpayer's expense. One of the most common arguments is that the government can always borrow at a lower rate than the private sector because government bonds are backed by tax revenues and so are deemed to be virtually risk-free. This argument is then buttressed with concern that the financial benefits that accrue to the private enterprise will be more generous relative to a publicly funded model or relative to the benefits that the public derives from the delivery of the service itself. But herein lies a disconnection between government and private sector objectives. Aside from having access to the deep pockets of the private sector, the greater benefit to the government comes from being able to divest from project-risk while at the same time leveraging technological and innovative advances from the private sector.
GOVERNMENT NEEDS TO STICK TO THEIR GUNS The 198-kilometer Fredericton-Moncton highway in New Brunswick was one of the largest P3 arrangements ever entered into by the province, not to mention its first experience with a tolled road. When the project was first conceived and implemented, the three main government objectives supporting the P3 model were: 1. to avoid an increase in provincial debt 2. to achieve a reasonable cost of funds 3. to achieve an optimal degree of risk transfer to the private sector In accordance with these objectives, the government agreed on a DBO P3 structure with the revenue from user-paid tolls deemed the best means to fund the project. However, the tolls were only in place for a brief 14 months - swiftly abolished in March 2000 under new government when the tolled road became a political hotbed during provincial elections. Of course, nothing is free and the government still had a contractual obligation to fulfill with the private developer. It certainly could not transfer all the design/build/maintenance costs and risks without compensation. But since its revenue source through user-tolls was eliminated, the government had to pick up the tab through shadow toll payments - i.e. the government now makes the toll payment to the private company based on vehicle counts rather than the user. So in the end, the government abandoned a sound financial model that avoided an increase in government debt, and in the process did not achieve optimal transfer of risk to the private sector since it ended up with all the revenue risk. Now every New Brunswick taxpayer foots the bill on the highway's use, rather than just the users. This "out-of-sight, out-of-mind" strategy, also means "out-of-pocket". To be perfectly clear, our beef is not with the notion of shadow tolls. Shadow tolling is a highly effective means for governments to construct and finance roads, and it is actively used throughout UK and Europe. But, it tends to be used for extensions or upgrades to existing road systems, rather than alternative new road systems. In other words, it is publicly unpalatable, and may not necessarily make financial sense, for a government to instate tolls on roads that were previously considered "free" to the public or to instate tolls on roads where the user does not have a choice of a reasonable "free" alternative route. Moreover, the implementation of shadow tolls in the Moncton highway case does not detract from the fact that the private consortium did (and continues to) meet its obligation to the government. • The road was completed in short measure, with the construction timesaving estimated in the range of 10-15 years over traditional procurement. • On the initial cost side, the province estimated savings to be $170 million. • On the maintenance side, the province estimated savings over 20 years to be $13.7 million (net present value). • The time saving to the user compared to the existing route was estimated to be 30-35 minutes. And, during the first four years of construction, the overall safety record for the workers was three times better than the industry average. The problem with the Fredericton-Moncton P3 model is that shadow tolling was implemented under political pressure, and not because it delivered optimal value to the public. By switching to a shadow toll structure, the province relinquished an estimated $321 million in concession fees that would have accrued from the project over a 30-year period - mainly consisting of toll revenues in excess of what is required to fund toll-based debt repayments and interest.50 And, more importantly, the province compromised its own P3 objectives, which were to avoid an increase in provincial debt and to achieve optimal risk transfer. Let's face it, road tolls are a sticky issue for Canadians, making it all the more reason that governments are upfront in their objectives, deal structure and costs. For the public to support the model, it needs to be properly informed. In this particular case, the problem did not lie in the structure or proposal of the P3, but rather in the government's inability to effectively follow through with its objectives. |
On the other side of the equation, there is no mistaking that every private company that bids on a government project does so with the intent of earning an acceptable rate of return on their investment. Although this matter can often be the subject of public debate, as a rule of thumb, companies should not earn substantially more from a P3 contract than from comparable work. But, in risk allocation, nothing is free. In bidding for a project, the private party estimates the project risks and their potential impacts on project revenues, and in effect sets premiums to insulate itself from the financial results of materialized risks. In essence, the risk premium is a form of self-insurance. Of course, the bottom line is that private parties will accept almost any risk provided that the premium paid is sufficiently large. The question for government is whether the risk premium is good value for money or whether it is more cost-effective for governments to assume the risk itself, taking into account the likelihood of a particular risk occurring and how the government may be able to mitigate the impacts.41 This returns us to the importance of a thorough project risk-costing evaluation under the direction of a PSC model. The onus of responsibility falls to the government to ensure that proper guidelines and expertise are in place to aid in the decision.
The other claim regarding the public sector's ability to borrow at a lower cost is true, but only if the basis of comparison is the absolute difference between government and corporate bond yields. However, this is not an accurate reflection of the true debt-financing costs of a project, since the yield on government bonds will not factor in the specific risks associated with that project. In this respect, the government does not borrow and spend money absent of any risk, because a publicly financed project means that the taxpayer has, in fact, underwritten all the associated risks of that project, and, where it materializes, bears the costs.42 Simply put, if the project fails or ends up costing substantially more than initial estimates, taxpayers are on the hook. To cover any additional expenses, the government would have to raise taxes, divert funds from other areas of spending or borrow the additional funds, which is an added burden to taxpayers down the road. Therefore, it is inappropriate to compare a "risk-free" government bond yield with the cost of private financing. Instead, the guiding principle for a public-private venture should incorporate risk-transfer under a broader principle of "value for money". We must keep in mind that P3s are not solely about the injection of private financing into public infrastructures; this is but one element of the infrastructure equation.