The Auditor General Did Not Say that P3s Cost Taxpayers $8 Billion

Commentary, Taxation, Steve Lafleur

The Auditor General of Ontario’s report is a technical document that few people actually read. Bullet points from executive summaries of such reports are often used as the basis for newspaper columns and political talking points in the grown up equivalent of the game telephone. Each time those snippets are repeated, they are re-worded and become less and less representative of the original message.

To wit, some news reports following the AG’s report claimed that public-private partnerships have cost Ontarians an extra $8 billion over 9 years. In fact, the Auditor General claimed that the cost of P3s was  “almost $8 billion higher than if the public sector had been able to contract out the projects to the private sector and oversee their successful delivery.” These are two very different statements.

A careful reading of the relevant section shows both the value of having a specialized procurement agency and a robust Auditor General’s office. Each contributes to continually improving procurement practices.

Infrastructure Ontario is a specialized procurement agency created by the provincial government. It has brought together a level of procurement expertise that is rivaled by few public sector entities on earth. It, along with Partnerships BC, is viewed as a model for governments looking to get more value for money from infrastructure projects. Projects that are vetted and deemed appropriate for P3s are passed on by other departments to Infrastructure Ontario, which oversees the procurement process.

A key difference between P3s and traditional procurement is that P3s typically involve transferring risk from the government to a private partner. A well designed P3 contract shifts the burden of potential delays and cost overruns to the private contractor, giving them a strong incentive to perform on time and on budget. It also forces them to submit realistic bids, since they bear costs of failure.

The $8 billion figure was arrived at by comparing the costs of 74 projects funded through P3s to what they might have cost through public sector management if completed on time and on budget. However, part of the rationale behind P3s is that publicly managed projects often go over schedule and budget. P3s act as a form of insurance. Infrastructure Ontario estimated that the risk premium for public sector delivery of those projects was $14.6 billion higher than for P3s, meaning an estimated net savings of $6.6 billion with P3s.

The above calculations seem rather subjective. Indeed, many (including the AG) worry that the risk premiums imputed to public sector delivery are too high. However, the risk of infrastructure projects being delayed or going over budget is very real. Bent Flyvbjerg, Professor of Major Programme Management at Oxford University’s Saïd Business School, has written extensively on infrastructure cost overruns across the globe. His research found that 9 out of 10 infrastructure megaprojects encounter cost overruns – on average, 28%.

While the AG claims that Infrastructure Ontario has no empirical data to justify their risk premium, the global evidence suggests that there should be a built in risk premium. The question is how high that premium should be, something Infrastructure Ontario needs to continually assess. The answer is not zero.

The Auditor General made a number of technical recommendations, but the one that grabbed headlines was that Infrastructure Ontario explore using public sector delivery for some projects. Since governments have lower borrowing rates than private companies, and there are additional cost associated with overseeing P3s, in the abstract this could lead to savings – from the aforementioned $8 billion. While it isn’t obvious that those savings can be found, this recommendation, as well as the others in the report, point to one of the key benefits of a specialized procurement agency: the ability to experiment and incorporate best practices.

P3s enjoy broad support – from the IMF and the International Labor Organization to chambers of commerce – but that support is contingent on effectiveness. If there are better models – P3 or otherwise – those best practices should be incorporated. P3s appear to be a good model for large infrastructure projects, which have become increasingly effective in part thanks to government agencies such as Infrastructure Ontario and the Auditor General.

Rather than playing telephone or scoring political points, we should pay close attention to what the Auditor General actually wrote. While one can debate some of the finer points, the report does not contain any sort of indictment of P3s. The type of oversight the Office of the Auditor General provides is only useful if we actually pay attention to the details.