As stated in §4.3, there is clear evidence from this evaluation that PPPs are more likely to be on-time than public procurement projects. Other studies have come to the same conclusion. Similarly, provided there is no change in the project definition, and assuming the Provider is carrying the completion risk, there would normally be no additional costs charged to the Promoter. However, while these findings may be important for the management and availability of public infrastructure, they are not critical for the assessment of whether, or when, to choose the PPP mechanism. Assuming that the same economic benefits will be realised, the question is which mechanism will provide the lower whole-life cost to the economy. This is, of course, the question that PSCs are designed to answer ex-ante. However, ex-post, the evaluation could not quantitatively answer the question. Two methodologies were considered: the ex-post modelling of the alternatives available ex-ante, and direct project comparisons. The modelling approach was rejected for two reasons. Firstly, because of the level of uncertainty associated both with the risks being transferred and the behaviour of the public sector. Secondly, the resources needed to carry out this work on a reasonable number of projects placed it beyond the scope of the evaluation. The direct comparison approach also had to be rejected. To make an effective comparison it would be necessary to identify two projects of similar specification, constructed and operated in the same legal, financial and fiscal framework, and subject to the same market conditions. Although the EIB has a large and diverse portfolio, it was not possible to identify suitable project pairs. Under these circumstances, EV was unable to determine ex-post if the original decision to use a PPP was more cost-effective or not.