P3s IN CANADA: THEORY AND EVIDENCE Aidan Vining

1
P3s IN CANADA THEORY AND EVIDENCEAidan Vining
Anthony Boardman
2
THIS PAPER

• This paper examines 10 infrastructure P3s in
  Canada (basically the ones we could get
  reasonable information on and which are somewhat
  complete.
• P3s really began to take hold in the mid-1990s.
• What has happened to these initial P3s?
• How should we judge their effectiveness?
• What lessons can be learned for future?
• These questions are important for Canadian policy
  because emerging case study evidence in a number
  of countries, including the United Kingdom
  (Broadbent and Laughlin, 2004 Grout and Stevens,
  2003 230 Pollitt, 2005 Shaoul, 2005), Ireland
  (Reeves, 2003), the Netherlands (Klijn and
  Teisman, 2003), Denmark (Greve and Ejersbo, 2003)
  and Australia (English, 2005 Hodge, 2005)
  suggests considerable dissatisfaction with P3
  outcomes.

3
PAPER OUTLINE

• Reviews the posited major government (normative)
  rationales for P3s.
• Reformulates and broadens the normative rationale
  to adopt a broader view of costs.
• Presents a positive theory perspective on P3s
  based on public choice theory, transaction cost
  economics, emerging empirical evidence from other
  jurisdictions, and experience with
  contracting-out by government and mixed
  enterprises.
• Presents the case study evidence (10 Canadian
  infrastructure P3s).
• Assesses the case evidence in terms of the
  positive theory perspective.
• Bottom line transaction costs, including ex ante
  contracting and negotiation costs, as well as ex
  post costs (i.e. after formal contract
  agreement), such as monitoring and renegotiation
  costs, have proven to be high. These costs may be
  borne either by the private sector or
  government. There can also be significant
  external costs.
• Conclusion P3s work best when they are closest
  to old-fashioned Build-Transfer construction
  contracts.

4
SUMMARY OF FINDINGS

• Canadian governments have generally found it
  difficult to effectively reduce either their
  budgetary risk exposure through the use of P3s.
• At the same time, the for-profit private sector
  partners have had difficulty generating adequate
  rates of return, although this is a tentative
  conclusion as they have usually had incentives to
  publicly emphasize losses and to be secretive
  about profits.
• One surprisingly common outcome is the
  dissolution of the P3 more quickly than
  envisioned in the original contract, whether
  through a government buy-out, private entity
  bankruptcy or otherwise.
• An even commoner outcome is protracted conflict,
  with high contracting costs borne by one party or
  both.
• These findings throw into doubt the social
  utility of P3s as a mechanism for delivering
  public infrastructure in contexts where
  transaction costs or negative externalities can
  be expected to be high.

5
Major Canadian P3 Projects
6
GOVERNMENT RATIONALES FOR P3s

• Governments have used 3 major (articulated)
  rationales for entering into P3s (Vining,
• Boardman, and Poschmann, 2006)
• The minimization of on-budget government
  expenditures and/or formal government debt.
• The provision of infrastructure and services at
  lower cost as a result of scale and learning
  efficiencies and other sources of technical
  efficiency of large, specialized firms. Also
  superior incentives of these firms versus public
  sector bureaucracies because of X-inefficiency
  (Frantz, 1992).
• Reduce the public sectors risk exposure to
  construction costs, maintenance costs and usage
  levels. The U.K. government has been a leader in
  arguing that the various dimensions of risk
  transfer should be the primary benefit from P3s.
  
• A fourth usually unstated rationale is that
  governments believe (or at least want to believe)
  that private-sector operation means that it is
  politically easier to impose user fees, resulting
  in lower net budgetary outlays to government.
  The reasoning is that voters are more willing to
  accept that the private sector needs to raise
  revenue to cover its costs, repay its debt or
  make a profit than to accept the argument that
  the public sector needs to do so.

7
OFF-BUDGET RATIONALE FOR P3s

• There are clearly political benefits of keeping
  expenditures and/or formal debt off-the-books
  (Marlow Joulfaian, 1989 Joulfaian Marlow,
  1991) in most circumstances.
• However, the underlying economic reality is not
  altered public sector (whether government or
  users) have to pay for it (Quiggen, 2005).

8
OFF-BUDGET RATIONALE FOR P3s (continued)

• P3s obviously can spread governmental payment
  obligations over a longer period of time.
• Per se, however, this does not reduce the present
  value of costs and is therefore a weak normative
  rationale for P3 usage.
• Caveat 1 time-shifting could be justified for
  very long-lived projects on intergenerational
  equity and efficiency grounds.
• Caveat 2 P3 as mechanism may also be somewhat
  justified by institutional barriers that thwart
  explicit government capital financing mechanisms.
  However, these barriers are best addressed
  directly.
• In practice, 3. and 4. are rarely the budget
  arguments that are made by governments.

9
COST-EFFICIENCY RATIONALE FOR P3s

• Private sector firms can have superior scale,
  scope and learning economies because they are
  more specialized and larger.
• This cost advantage is greatest vis-à-vis smaller
  provincial, regional or municipal governments.
• Cost advantage likely to be most significant for
  construction (rather than operations/maintenance).
  
• May have a cost advantage in terms of financing
  costs because of greater access to better
  expertise on pricing risk consequently ability
  to lower financing costs
• Also superior incentives (given well-designed
  contract) to reduce/minimize costs likely to be
  most significant in dynamic aspects of projects,
  such as latest production technology.
• Public sector monopoly bureau employees usually
  have weak incentives to forecast costs or use
  levels accurately.
• This rationale is buttressed by weak government
  performance on controlling costs (Flyvbjerg, Holm
  Buhl, 2002, etc.).
• However, really big caveat first-order result of
  lower costs is higher private sector profits
  rather than lower public sector costs.

10
RISK EXPOSURE RATIONALE FOR P3s

• Specialized private sector firm may be able to
  spread risk over a greater number of similar
  projects.
• Public sector, however, is usually able to spread
  risk over a greater number of dissimilar
  projects.
• Either way, not a great normative rationale for
  P3s.
• Caveat except to the degree that the private
  sector is more efficient at pricing and
  allocating the risk.

11
REFORMULATION OF COST-EFFICIENCY RATIONALE

• Problem with the cost-efficiency argument as a
  normative rationale is that its normal focus on
  production costs is too narrow.
• First, needs to include transaction costs (TC)
  the costs of negotiating, monitoring and
  re-negotiating contracts. These often do not show
  up as project costs.
• Second, needs to include external costs (and
  benefits) deadweight losses from pricing,
  project failures, etc.

12
TC PERSPECTIVE

• Transaction cost economics (TCE) provides a
  useful framework for addressing the components of
  total social costs.
• TCE emphasizes that total social costs equal
  production costs plus transaction costs
  (Williamson, 1975).
• Transaction costs (TC) include the cost of
  negotiating, monitoring and, if necessary,
  re-negotiating contracts with profit-maximizing
  firms.
• TC language is more appropriate than agency
  language because P3s have the character of a
  relationship between independent organizational
  entities. Agency, or principal-agent theory,
  language is appropriate for intra-organizational
  hierarchical contexts.

13
APPROPRIATE COST-EFFICIENCY CRITERION

• This suggests the following criterion
• Minimize the sum of production costs, transaction
  costs and net externalities (holding constant
  quality).
• This then raises the question How are P3s likely
  to perform in terms of this broader formulation?
• To address this, we need a positive theory
  perspective.

14
WHOS MAXIMIZING WHAT?

• While the language of partnership is endemic to
  P3s, the basic premise of this paper is that the
  public and private participants have conflicting
  interests (Teisman Klijn, 2002 Reeves, 2003
  Trailer et al., 2004).
• The two objective functions
• Private sector participants wish to maximize
  risk-adjusted profits.
• Public sector participants wish to minimize the
  sum of expected on-budget net public expenditures
  and political costs.

15
NUANCES OF THE OFs

• The nuances in these objective functions are
  important
• Risk-adjusted aspect of private sector
  participants They are willing to forego profits
  if they can reduce risk sufficiently.
• Indeed, private sector participants may be
  considerably more risk-averse than public sector
  participants expect, at least ex ante. This is
  especially so with use/revenue risk because (1)
  they are not familiar with it (2) They recognize
  that it is largely controlled by public sector
  actors (even absent opportunism by their partner)
  (3) It opens them up to greater opportunism. As a
  result, the private sector requires a high
  premium to accept risk.
• In order to minimize risk, sophisticated private
  sector partners are likely to (1) form
  stand-alone P3 corporations, thereby reducing
  their worst-case costs by declaring the
  stand-alone corporation bankrupt, if necessary
  and/or by (2) limit their equity participation
  through the utilization of extensive third-party
  debt financing (Roll and Verbeke, 1998).

16
THE CASE STUDIES

• We able to review ten Canadian P3 projects in
  depth.
• The case studies were selected because of the
  availability of information, the size and profile
  of the projects, the jurisdictional coverage that
  they present and the lessons they offer for P3
  contract theory, design and implementation.
• They are Alberta Special Waste Management System
  (Alberta), Confederation Bridge (Federal),
  Highway 407 (Ontario), Highway 104 Western
  Alignment Project (Nova Scotia), Evergreen Park
  School (New Brunswick), OConnell Drive
  Elementary School (Nova Scotia), Britannia Mine
  Water Treatment Plant (British Columbia), Moncton
  Water Treatment Facility (New Brunswick),
  Cranbrook Multiplex (British Columbia) and
  Waterloo Landfill Gas Power Plant (Ontario).

17
ALBERTA WASTE

• In 2000, BOVAR issued a notice of intent to cease
  operations due to its inability to make a profit.
  
• In 2001, the facility was returned to the
  province, and capital assets of approximately 34
  million were written off by BOVAR (BOVAR, 2000).
  
• In 2003, the Alberta government signed a 10-year
  operations contract with Earth Tech Inc., a
  division of Tyco International Ltd.
• Because there was no effective risk transfer,
  enormous contracting costs and the eventual
  demise of the P3, it is hard to classify Swan
  Hills as a success.

18
CONFEDERATION BRIDGE

• This P3 clearly delivered a functioning bridge on
  schedule.
• However, the project had high financing costs
  the bonds were sold at a 4.5 interest rate, at a
  time when similar federal issues were priced at
  4.1. Moreover, SCFI paid a sales commission of
  1.75, compared to a typical rate of 0.6 for
  federal real return bonds. These higher rates
  might be justifiable if the government had
  eliminated equivalent risk through them (in other
  words, if the federal government had acquired a
  put option against the risk of project default)
  or if the consortiums capital requirement had
  imposed upon the private partners an incentive to
  minimize project capital.
• However, these funds were guaranteed by
  government and there was no net reduction in risk
  exposure. It is difficult to escape the
  conclusion that the P3 was chosen primarily to
  achieve off-balance sheet financing. Success
  depends on the trade-off between a functioning
  bridge and the relatively un-P3-like project that
  emerged with relatively little risk transfer.

19
THE 407

• Did deliver a highway on time and on budget.
• The major initial weakness of 407 as a P3 was the
  failure of the government to effectively transfer
  financing risks the construction phase became a
  conventional develop, design and build contract.
  
• Even so, ongoing transaction costs were extremely
  high including litigation. The 407 operator, not
  surprisingly, was interested in maximizing
  profits rather than optimizing metropolitan
  Toronto traffic flows and was interested in
  increasing tolls. The end result is the province
  managing the tolls.
• On the other hand, there were also opportunistic
  elements to provincial behaviour as political
  costs escalated because of the toll increases.
• Those who focus on the lack of risk transfer,
  such as Bose (1993), regard it as a P3 failure.
  Mylvaganam and Borins (2004), more charitably,
  present a mixed assessment. Transaction costs
  could be considered very high.

20
HIGHWAY 104

• Major conflicts arose relating to the allocation
  of risk.
• Although touted as a P3, the private partner is
  really a government entity (so much so that the
  AG argues its debt is provincial debt).
• The private partner assumed the risks
  associated with cost overruns on construction due
  to factors such as design flaws or other
  interruptions (Van Adel, 1999). The risk for low
  traffic volume and resulting revenue flow
  initially also fell entirely on them.
• The government also assumed that the private
  sector partner would bear most of the project
  risk, including those associated with
  environmental permitting, archeological and
  geotechnical risks and any cost overruns.
• The government assumed that providing the
  right-of-way for the land for the project was its
  only major obligation. However, after the
  participants had reached an agreement on the
  terms of the contracts, the lenders renegotiated.
  
• Although the private sector had initially assumed
  the revenue risk for the project, this risk was
  largely offset by legislation that essentially
  required trucks to use the new toll highway by
  prohibiting trips on the old highway (Van Adel,
  1999).
• An assessment by the AG argues that the highway
  project did deliver benefits to N.S. However,
  the AG argued that debt service charges for the
  loan would have been notably lower if the
  province had directly borrowed the funds for the
  project, rather than using a P3. Political
  factors have obviously raise transaction costs in
  this case. Hard to see this a P3 success.

21
EVERGREEN SCHOOL

• The school did get built on time.
• This case study illustrates some of the
  difficulty of even calculating the costs of P3
  financing compared to direct government
  provision, especially when the private sector
  partner engages in activities that the public
  sector would not.
• On one hand, these activities may be innovative
  on the other hand, they may involve externalities
  (in this case, foregone use of school property
  outside of regular school hours).
• In the big scheme of things, this is a small
  project. Could be argued to be a reasonable
  success. But, province seems to have
  underestimated the externalities (real and
  political) of a private partner have control of
  school after hours. Transaction costs have been
  substantial.

22
OCONNELL DRIVE SCHOOL

• A school did get built.
• While it is difficult to conclusively conclude
  that there were no construction cost savings (an
  appropriate counterfactual on direct government
  procurement would necessarily be speculative), it
  is certainly unlikely that were major cost
  savings.
• Risk transfer was minimal (AG argued)
• Transaction costs appear to have been high.
• Clearly, OConnell Drive was not successful from
  a political perspective. While political costs
  of the P3 appeared low ex ante, they certainly
  turned out to be high ex post and essentially
  forced the abandonment of P3 for school
  construction.

23
BRITANNIA MINE

• Some degree of risk transfer, because the private
  partner has to meet volume and quality targets to
  get paid.
• Indeed, given the specialized technical
  requirements of the project, it is hard to
  conceive of this project being conducted in-house
  by government. Would always be contracted to
  specialized firm.
• Britannia appears to represent a reasonably
  successful P3.
• On the other hand, this is the kind of project
  that has traditionally been contracted-out in one
  form or another.

24
MONCTON WATER

• The plant got built on time.
• Again, as with Britannia, given the specialized
  technical requirements of the project and the
  proprietary knowledge of USFilter, it is hard to
  conceive of this project being conducted in-house
  by government.
• Again, as with Britannia, this is the kind of
  project that has traditionally been
  contracted-out in one form or another.
• Transaction costs appear to have been
  significant, although not as high as in many of
  the other case studies.
• Moncton Water appears to represent a reasonably
  successful P3.

25
CRANBROOK MULTIPLEX

• As a result of the contractual changes, the
  citys auditor had included the 22 million lease
  on the citys financial statements, resulting in
  a substantial decreasing in the citys borrowing
  power.
• After dealing with legal disputes, cost overruns
  and construction delays, five years after the
  project began, the P3 officially failed. Vestar
  withdrew in 2004 after paying the city 1.7
  million to take ownership of the complex
  (Cloutier 2004).
• Overall, the project proved to be much more
  costly than projected. When the project failed
  the city found itself with the highest debt level
  in the province. This project is clearly a P3
  failure.

26
WATERLOO GAS POWER

• Plant built on time.
• Relatively simple contract. Build the plant, sell
  the gas, give royalty to RMOV depending on
  volume.
• Again, this appears to a successful, relatively
  small, straightforward P3.

27
FRAMEWORK FOR FINDINGS

• Degree of asset specificity?
• Construction complexity?
• Construction cost risk transferred?
• Use (revenue) uncertainty?
• Use risk transferred?
• Level of government contract management skill?
• Externalities?
• Level of transaction costs?
• Success?
• Table on next slide summarizes this

28
TABLE 2 Key Factors in Assessing P3 Case Studies
29
CONCLUSION HIGH TRANSATION COSTS

• Given conflicting goals, transaction costs are
  high
• On public sector side changing political risk
  and opportunism raise transaction costs.
• On private sector side firms desire to maximize
  profits and avoid risk, especially use risk,
  raise transaction costs.
• Transaction costs are aggravated by
• Complexity.
• Ability to transfer risk.

30
Figure 1 Revenue Uncertainty and Effective
Revenue Risk Transfer
Revenue (Use) UncertaintyHigh
Highway 407 3.1 B
Confederation Bridge 730 M
Highway 104 113 M
Britannia Mine Water Treatment 27.2 M
Revenue Risk Transfer High
Cranbrook Rec Plex 226 M
Alberta Waste Mgmt ?
Low
Moncton Water Treatment 85 M
Evergreen Park School 14.7 M
OConnell Drive School 8 M
Waterloo Landfill Gas Power Plant7.5 M
Low
31
CONCLUSION WHERE P3s WORK.

• P3s seem to work where contracting out has always
  worked best for government
• Technically specialized projects where firm has
  expertise (Economies of scale, scope, learning).
• BOT format.
• No, or very limited, use risk transfer.