Public-Private Cooperation in Infrastructure Development: A Principal-Agent Story of Contingent Liabilities, Fiscal Risks, and Other (Un)pleasant Surprises

With increasing pressure on physical infrastructure and limited resources, most governments have turned to various forms of collaboration with the private sector to help finance, build, and/or operate public assets. While the benefits of such joint efforts are potentially numerous, they also entail major risks that are hard to assess adequately. This paper discusses the economic and financial rationale, including risks, for partnerships between the public and the private sectors as a way to share the burden of developing infrastructure. It also proposes a principal-agent framework to help conceptualize the relations between the government and the private firm in public-private partnerships (PPPs).

1. See Leruth and Paul (2006) and the collection of essays in Paul (2006).

2. Two key articles directly discuss the economics of various forms of public-private collaborations, mostly using contract theory: Hart et al. (1997); and Hart (2003).

3. One should add that the private sector may also be tempted (because of agency problems discussed later) to reduce the quality of output or bribe the authorities. Moreover, positive externalities (notably on poor people) may not be correctly accounted for/specified in the PPP contract.

4. The government tends to use the social time preference rate (STPR) or some other risk-free rate to discount future cash flows when appraising projects, while private bidders for PPP projects will include a risk premium in the discount rate they apply to future project earnings.

5. For example, it is not always true that the cost of financing is cheaper if it is obtained through the private sector. Typically, government borrowing tends to remain cheaper than private borrowing, and this would argue for limiting private funding, at least in some cases.

6. T. Irwin, from the World Bank, proposes a typology of the fiscal risks borne by the State, which is discussed later in the paper. See Irwin (1998, 2003).

7. The U.K. National Audit Office was among the first to raise this issue. See U.K. National Audit Office (2003, 2004).

8. This definition is quoted from the IMF (2004).

9. To be more precise, one can distinguish three forms of contracts between the public and the private sector, ranked in decreasing order of public sector implication: (i) the public market (the debtor is the public collectivity); (ii) the PPP contract (the debtor is the private partner, but the public collectivity bears some risks); and finally (iii) the concession (the debtor is the private partner, and the major part of the returns come from the operation). See Icade Conseil et Associés (2006).

10. For more on VFM (which has become a focus of attention by the U.K. National Audit Office), when examining the PPP/PFIs (Private Financing Initiative) projects as discussed below), especially on the issue of an appropriate public sector comparator and on factors to take into account when assessing operational and financial flexibility; see U.K. National Audit Office (2006).

11. See Fitzgerald (2004) for a thorough review of the Partnerships Victoria initiative (Melbourne, Victoria).

12. Although the definition of private-public collaboration as presented in the PFI is a somewhat extreme category of PPPs, it is nevertheless similar to the “narrow definition” of PPPs informally proposed by Timothy Irwin: “privately financed projects in which the government either is the main purchaser of the output under a long-term contract or supplements user fees with subsidies or guarantees.” Note that some schemes, which are commonly associated with PPPs, such as leasings, would not fall under this definition.

13. Since 1997, the Labor government has broadened the initiative, which now also includes schemes assimilated to concessions or financial and operational leasings.

14. In 2005, 55 new PFI projects for a capital value of 87 million pounds were created under the initiative.

15. The process of contracting between the public and private sectors in the case of the London Underground is presented in U.K. National Audit Office (2000). The analysis emphasizes the “Value-for-Money test” to discriminate between the final private bidders.

16. For example, the Merloni Law in Italy (1994), the Concessions Law in Spain (2003), the Ordonnance on PPPs in France (2004). See also PriceWaterHouseCoopers (2004).

17. See the “Green Paper on Public-Private Partnerships and Community Law on Public Contracts and Concessions,” EC COM (EC COM 2004) 327, Brussels. Note that this paper has been criticized for making “PPPs too easy” and underestimating the risks borne by the public sector (see, for instance, Hall 2004).

18. Commitments have declined worldwide from US$131 billion in 1997 to below US$50 billion in 2003 while about half of the contracts signed in the mid-1980’s have been renegotiated.

19. These figures can be found in “Public Money for Private Infrastructure: Deciding when to offer guarantees, output based subsidies, and other fiscal support,” Irwin (2004).

20. Such a list is given by Polackova (1998), Table 1. See also Irwin (2004).

21. For details on similar issues, see Best (2005), and Furman and Stiglitz (1998).

22. In 2006, following multiple complications and delays in the process, the then Secretary of Defense, Donald Rumsfeld, cancelled the lease prior to its being implemented. For more details on the proposed terms, see footnote 13 of IMF “Public-Private Partnerships” (2004).

23. At a conference held in 2004, J. Colman, from the NAO, considered that PPPs are potentially beneficial (less responsibility for the State, better sharing of skills and risks), but that they involve risks to the VFM of taxpayers (more expensive financing, complexity, unequal negotiations between State and the more streetwise private sector, inflexibility, and lack of transparency).

24. In theory (Modigliani–Miller theorem), the cost of financing does not depend on how the project is financed, but only on its risk profile when markets are complete. However, markets are often incomplete.

25. This is a point often made and it argues for limiting PPPs in developing countries where the public sector indeed has cheaper access to funds than the local private sector (but not necessarily the international private sector).

26. This table is inspired from risk matrices proposed by various private companies before contracting a PPP (e.g., the Dexia group in France), and from reports by administrations, such as the French Direction Générale des Routes (see December 2005 report).

27. For more information on the difference between the two, the definitions, and some useful, related considerations, see IMF (2004), pp. 19–20.

28. The International Accounting Standards Board (IASB) identifies the factors that allow for the classification of leases.

29. See Kofman and Lawarrée (1993 and 1996).

30. Bentz et al. (2001) focus on the best allocation of ownership to raise investment incentives and conclude that PPP arrangements work best in this regard for small investment costs.

31. Bentz et al. (2001) also look at this issue but rely on a complete contract approach.

32. Benett and Iossa (2006) have explored this issue. They focus on the best ex-ante contract, but not on the transfer of risk. The other issue affected by ownership is the power of contract renegotiation. Benett and Iossa do consider this issue, but not in a Principal-Agent framework, which would provide a more natural approach, although I do not consider it in this paper.

33. In a Principal-Agent framework, the “state of nature” can be high or low. If it is high, the agent may be in a position to convince the principal that it is low so that it will be satisfied with the level of output normally associated with a low “state of nature.” This may be to the advantage of the agent if the transfer it receives, minus the effort and inputs it puts in, is comparatively higher.

34. The threat of punishment if caught cheating then reduces the agent’s incentive to misbehave.

35. The extent of controls and incentives should not be increased above the point where their marginal return (in terms of reducing cheating) equals their marginal cost (Leruth and Paul 2006).

Authors and Affiliations

1. International Monetary Fund, Offices in Europe, 66, Avenue d’Iéna, 75116, Paris, France

   Luc E. Leruth

Corresponding author

Correspondence to Luc E. Leruth.

This article is based on the keynote address I delivered at the First International Conference on Funding Transportation Infrastructure, which took place in Banff, August 2006. I would like to thank the participants, A. de Palma, T. Irwin, E. Paul, P. Pollard, G. Prunier, D. Ross, G. Schwartz, J. Seade, C.H. Weymuller, and two anonymous referees for useful comments and discussions. Excellent secretarial assistance by J. Dyer is also acknowledged. The views expressed in this paper are those of its author and should not be attributed to the International Monetary Fund, its Executive Board, or its management. The usual disclaimer applies.

Reprints and permissions