Public private partnerships
Governments need to ensure a robust process for evaluating PPPs writes UQ's Flavio M. Menezes for Policy Online.
The Queensland Government, under the leadership of Campbell Newman, is promoting Public Private Partnerships (PPPs) to finance infrastructure projects. The Government recently created Projects Queensland, a special unit tasked with stimulating PPPs in the State. Many may question the Government’s push for PPPs in the wake of several high profile PPP failures including, Sydney’s cross-city tunnel, Brisbane’s Clem 7 tunnel and the collapse of the consortium building the Ararat prison in Victoria. However, economic research has shown that PPPs can be advantageous for society and that there are key steps governments can take to ensure their success.
The University of Queensland (UQ) School of Economics in conjunction with the Association for Public Economic Theory (APET), recently hosted a workshop focusing on PPPs—the first of its kind in Australia. The workshop brought together internationally renowned scholars, including Professor of Economics at Yale University, Eduardo Engle. (The Professor has co-authored a book on PPPs that is in prepress.) The following is a brief overview of the issues discussed at the UQ workshop with regard to the advantages and disadvantages of PPPs.
The difference between PPPs and traditional public tenders
The PPPs model differs substantially from the traditional public tender model for infrastructure construction. In the traditional model, the government finances the construction phase of the infrastructure, tendering the construction to private parties. The operation and maintenance of the infrastructure also may be contracted to private parties.
In the PPP model, a government tenders a “bundle” consisting of financing, construction and operation of a piece of infrastructure (e.g., a tunnel) to private parties. The contract is usually for a fixed period. At the end of the contract, the asset reverts back to the government.
The choice between PPPs and public tendering often occurs in an environment where privatisation has been ruled out. Accordingly, I omit any discussions of privatisation in the discussion below. The overarching conclusion of the analysis below is that PPPs are superior to public tendering. However, governments often choose PPPs for the wrong reasons.
Why governments should choose PPPs: the advantages and disadvantages
i. Gains in efficiency
The first-and-foremost advantage of PPPs is the potential efficiency gains from bundling the construction and operations/maintenance. When bundling occurs (and as long as tender process is well designed), the winning firm should minimise the total of construction and maintenance/operating costs. For example, a firm that is only concerned with construction might design and construct a hospital that minimises construction costs but the design could entail very high operation and maintenance costs. The extent of these efficiency gains varies from project to project. However, it is worth noting that these gains would be negated if the traditional public tendering of construction and operation could replicate the benefits of the bundling. (Though it is unclear whether this can be done at all.) Bundling also avoids a government’s temptation to save on maintenance in later years. In the case of roads, for example, governments appear to prefer to build new roads rather than maintain existing ones—even though the cost of maintaining a road is substantially lower than that of replacing it.
ii. Public versus Private financing
It is often argued that public tenders are preferable to PPPs because it is cheaper for government to finance the construction phase. This argument is based on the ability of a government to borrow at the bond rate whereas the private sector must borrow at a substantially higher rate. However, this is argument is flawed.
Under the public tendering model, the government collects user fees/tolls. If this revenue could be hypothecated to pay the government debt associated with the project, then the argument above would be correct. However, typically these fees instead go into the general revenue pool used to finance all government expenditure. This implies that the true cost to society from government financing maybe be substantially higher than the bond rate.
Another advantage of private financing is the potential to avoid the construction of politically motivated white elephants. Private parties will find it difficult to obtain financing for a project that is not commercially sound.
Finally, governments may favour PPPs over public tendering arguing incorrectly that the former alleviate governments’ budget constraints. Given the current public perception that government debt is bad, politicians might be attracted to PPPs if they do not appear in the government’s balance sheet. For example, in the UK up until April 2009, only 23% of the capital cost for 599 Private Finance Initiative (PFI) projects was on balance sheet.
This argument is clearly wrong when PPPs involve direct government transfers, such as minimum income guarantees or other types of payments. It is also wrong to the extent that the PPP project is financed by user fees—a revenue stream which the government gives for the duration of the PPP contract.
iii. Shifting demand risk
Governments can be attracted to PPPs because they perceive that this model shifts the demand risk from the government to the private parties.
Campbell Newman, when Lord Mayor of Brisbane, was quoted as saying that the PPPs model worked for the delivery of Clem 7 because ratepayers were protected from the financial collapse of the company that built and operated the tunnel. (Although he did apologise to small investors who lost their money.)
This argument for choosing PPPs is erroneous for several reasons. Firstly, the private parties bearing demand risk do so in exchange for a risk premium. To the extent that they cannot influence demand, the government may be the best party to hold the risk. Secondly, the upshot of the financial difficulties with the Clem 7 is that it will be very difficult to find investors that are willing to finance similar ventures in the future. Thirdly, it is likely that these events have resulted in changes in attitudes by private parties who might refrain from bidding in future tenders for any PPPs or submit less aggressive (i.e., higher) bids. This is all bad news for users who will have to pay more to use infrastructure.
The argument about risk shift is also not factually correct. International experience suggests that PPP contracts are often renegotiated when the builder/operator gets into financial trouble due to lower than anticipated demand. The renegotiation typically involves additional government transfers or sometimes nationalisation of assets. This again suggests that ultimately governments ended up bearing at least some of the demand risk.
Improving the design of PPP tenders
There are ways in which PPP tenders can be modified to allocate risk appropriately—making them more attractive to both private parties and the government. A central contribution of Engle and co-authors to the PPP academic literature over the past decade is to suggest a tender process that allocates risks appropriately and ensures that the benefits of PPPs over public tender are realised. Although the focus of their research is mostly on roads, its applicability is clearly wider.
Their main idea is to run a least-present value of revenue tender. As part of the tender, the government announces a discount rate that is used to calculate the present value of revenue and a toll price. The discount rate is typically higher than the risk-free rate (or government bond rate) to account for risk. The winner of the tender is the firm that has submitted the lowest present value of revenue.
The innovation of this process is that the duration of the concession is variable. The contract only expires when the winner of the tender recovers the amount of revenue bid. If, for example, the government decides to reduce the toll, or if demand is lower than anticipated, the contract is extended accordingly. This way, the risk of demand (or other changes outside the control of the concessionary) is born by the government while competition for the contract ensures that society gets value for money from the PPP. In this type of tender, the firm with the lowest construction and operation costs is expected to win rather than, for example, the firm that is more optimistic about future demand but that may have higher costs. Under many circumstances this type of tender will work well: it has been tried successfully in Chile.
In the past decade, we have learned a lot about what works and what does not in PPPs. To avoid previous mistakes with PPPs, governments need to ensure that there is a robust process for evaluating PPPs.
Closer attention by policy makers to the better design of PPP tenders and contracts, as suggested by both economic theory and international practice, should also become an integral part of the Queensland Government’s renewed interested in PPPs.