The above implies that the relevant calculation for government is not so much a comparison of public sector infrastructure costs with the cost of a PPP, but rather an analysis of the return to the public sector of a direct investment in infrastructure. For a PPP where almost all the private partner's revenues come from government, this can be done in two ways, which are equivalent in terms of their result:
1. Start from the premise that the public sector's forecast net return is the forgone cost of paying a PPP partner, less government's expected cost of building and operating the infrastructure itself. As a cash flow stream, this net return will appear, in the initial periods, as large upfront infrastructure investment cost cash outlays; followed by operations periods during which government will receive net revenues in the form of (a) forgone costs of paying a PPP partner, less (b) its own actual costs for operations, maintenance and capital rehabilitation. Government's rate of return is the internal rate of return for this net cash flow stream, that is, the rate of interest for which the present value of the cash flow stream equals zero. On the basis of financial considerations, if the government's forecast rate of return (i.e., internal rate of return) is less than the market rate, it should prefer PPP procurement; if it is greater than the market rate, it should prefer traditional procurement, and if it is equal to the market rate it should be indifferent.
2. Alternatively, start from the premise that if government expects to earn a market return from its investment, then the present value of government's forecast return (which is the present value of the forgone or avoided costs of paying a PPP partner) should equal the present value of its own forecast costs. If government's forecast revenues (i.e., the foregone cost of paying a PPP partner) discounted at the market rate of return exceeds its own forecast costs, discounted at the same rate, then traditional procurement is financially preferable, if the opposite is the case, the PPP option is financially better.
The appropriateness of using a market rate of return as a benchmark for government is more obvious if the private partner has access to its own source revenues, such as tolls. If government can earn a higher than market rate of return by investing directly in the asset, and collecting the revenues, it should make the investment, otherwise it should provide the investment opportunity to a private partner. The only caution with respect to this decision-making rule, where there are own-source revenues, is that the potential transaction may be more complicated than making an investment, providing services and collecting revenues. The PPP structure may, for example, have to be one in which government supplements the PPP partner's revenues with some form of performance payment or shadow toll. In this case, government should consider these forgone costs when making its investment decision.