The payment mechanism lies at the heart of the PPP contract. The primary purpose of the payment mechanism is to remunerate the PPP company sufficiently so that it will be willing to enter into the PPP contract and provide the service. Beyond that, the payment mechanism is the principal means in a PPP contract to allocate risks and provide incentives.

A useful way to think about designing the payment mechanism is to begin with an extreme or ideal form and then see where certain risks should be shifted back to the public authority or users. Ideally, the public authority might want to pay the PPP company, in arrears, a fixed price for (and only for) each unit of service that is provided and that meets the service-quality requirements. This captures the key PPP principles that payment should be made only if the service is available and that payment should not be based on the PPP company's actual costs (it is not a "cost-plus" contract); this simple ideal mechanism is one that gives strong incentives to the PPP company for good performance.

Much of the detailed design of a payment mechanism can be conceptualised as moving away from this simple ideal either to take into account more complexities or because this simple mechanism would cause the PPP company to bear too much risk. "Too much" in this context could mean that the premium that would have to be paid to the PPP company for it to be willing to take the risk would not be worth the gain that might be obtained from increased efficiency. Or it could mean that there would be too great a probability of excess profits or, alternatively, high losses accruing to the PPP company - in either case, threatening the viability of the arrangement. In this connection, one key principle in the design of the payment mechanism is that risks that are entirely beyond the control of the PPP company should generally not be allocated to the PPP company.

Some of the different ways, then, that adjustments are made to the simple (stylised) payment mechanism outlined above are the following:

□ Payments are generally indexed in some manner to compensate for cost increases due to inflation.

□ In some cases, certain well-defined costs that are beyond the control of the PPP company are handled on a pass-through basis (i.e. actual substantiated costs for the particular item are passed through into the service fee).

□ Deductions that are made to the service fee for poor performance are linked to the degree of deficiency in service quality (set out in objective rules and using verifiable measures). Generally, the amount of the deduction should be in line with the losses that would be expected to be borne by the government or users by the shortfall in service quality.

□ Demand (volume or traffic) risk - a key issue in PPP design - is often assessed as being at least partially beyond the control of the PPP company, and sometimes wholly beyond its control. A variety of mechanisms exist to shift some or all demand risk away from the PPP company. For example, the unit price could gradually increase as demand falls. Or there could be a minimum payment guarantee (where the company is paid for a certain quantity even if the actual quantity falls below that minimum).

The list above only begins to describe the various adjustments that can be made in designing the payment mechanism in practice. Designers should always be on the lookout for features that could give the PPP operator perverse incentives and features that are complicated and ambiguous in ways that might provoke disputes later on. There are many trade-offs in this exercise. Payment mechanism design is as much an art as a science.

Looking at payment mechanisms that are commonly used in similar types of projects is a good way to begin the design process. In addition, the public authority's advisors should make use of a financial model of the expected PPP
to test alternative payment mechanisms, using sensitivity and scenario analysis, and, most important, to calibrate the parameters of the chosen mechanism so that it is likely to perform well under different conditions that might arise. An important consideration is often that, although poor performance should have strong negative impact on returns to equity holders a strong "bite" - it may be counterproductive if the shortfall in cash flow too easily jeopardises debt service payments, an outcome that could lead to the bankruptcy of the project company.