Public-Private Partnerships: The Good, Bad and Ugly


Thank you, Dave. I appreciate that kind introduction and thank you all very much for your warm welcome.

It’s great to be here and have this opportunity to talk to you about Public-Private Partnerships. I’ve been associated with the Mackinac Center for Public Policy for more than 20 years now, with more than my fair share of contact with P3s — Public-Private Partnerships — on both the theoretical and practical level. Today I’ll try to share some of the good, the bad and the ugly that I’ve seen in several of these deals.

Stay Engaged

Receive our weekly emails!

But first…

Last week I was enjoying a family vacation in Santa Monica, Calif., Walking along the beach, I came upon an odd shaped bottle. Looking around to make sure no one was watching, I gave it rub. You can imagine my astonishment when a real genie actually popped out and offered to grant me one wish.

After thinking hard I told him, “I’ve always wanted to visit Hawaii but don’t like to fly and hate boats. Could you build me a highway from here to there? 

The Genie shook his head and asked if I was crazy. He said, “Do you have any idea how impossible it would be to sink pilings in the Pacific Ocean? Or how much concrete and asphalt you’d need for 2,000 miles of highway? Can you imagine trying to get an environmental impact release? C’mon LaFaive, get real. Try another one.

“Some genie,” I thought, a bit put out. “I’ll show him.”  “OK, then,” I said. “Can you tell me how to give a riveting talk about Public-Private Partnerships that keeps my audience on the edge of their seats right until the end?”

A long moment passed before he looked me in the eye and said: “Would you like your highway to Hawaii to have two lanes or four?”

Sorry, that was lame, but the allusion to a massive construction project does provide a valid segue to the topic of Public-Private Partnerships, because the really big, high profile deals often involve transportation projects. By one estimate, individual transportation-related P3s often exceed $500 million. Indeed, high-dollar amounts are practically a distinguishing characteristic of the breed.


That P3s are not a new concept will not surprise this audience. But you may be surprised by their number and permutations. First things first, let’s see if we can agree on a definition.

Most obviously, a public-private partnership involves an arrangement between a federal, state or local government and private-sector entities, which can include for-profit and non-profit organizations.

Second, P3s can involve assets or services, such as building a toll road or managing a water system. A new P3 arena involves “social impact” deals where investors are paid to provide services aimed at achieving some social gain, such as reducing prison recidivism rates. Some authors specifically reject the idea that an asset sale qualifies as a bona fide “partnership,” so I exclude these from my definition for purposes of this tour.

Another name for P3s is simply “shared service delivery.” This very broad definition probably explains why the National Council of Public Private Partnership can claim that American cities work with private-sector partners for 23 of the top 65 major services they provide.

The terms “privatization” and “Public-Private Partnership” are often used interchangeably, especially, in this country. Sometimes a simple outsourcing contract might be characterized as a P3.

In Europe privatization tends to be more identified with the sale of a state-owned asset — such as a bank or port. Over the past few decades governments have come to rely more on these deals to improve services, cut costs or generate revenue.

Support is often bipartisan, or multi-partisan in parliamentary systems. Political parties that compete in the electoral realm can be on the same page when it comes to P3. Some of the goals governments seek from P3s include:

  • “Using private capital to fund public infrastructure or provide public services;
  • Acquiring private-sector expertise, including managerial talents;
  • Unloading part of the risk associated with large projects onto private sector players;
  • Improving service provision with performance guarantees;
  • Saving money, and in some cases generating it.”[1]


Partnerships between government and the private sector have been around for millennia. Some scholars have described ancient Rome’s “bread and circuses” as the world’s first P3s. For others it was Athens. In the 17th century, the Canal du Midi in Toulouse, France, was built and managed by a private sector entity in partnership with King Louis XIV.  The first P3 toll bridge in America has been traced back to 1654 and has been jokingly referred to as ‘Ye Old PPP.’

Interest in new models of procuring public assets exploded in the 1970s and 1980s in response to growing public debt, especially in a high-inflation economic environment. Government debt and other fiscal strains also drive many P3 deals today.

The literature identifies the 1990s as a starting point for the modern P3 era. Particularly in Britain, where John Major’s Tory government enacted a Private Finance Initiative law that represented the first systematic attempt to facilitate more public-private partnerships. Continent wide, there were some 1,400 P3 deals consummated in the European Union from 1990 to 2009 with a value estimated at $260 billion pounds.

Last week David Lick was asked by an audience member here to quantify the degree to which P3 deals happen. That is easier with Europe, but the United States — just as Dave explained — is hard to quantify.

That’s because the divisions of scope and authority between federal, state and local governments in this country limits the amount of aggregated data here, at least not in a single comprehensive database. We can say that PPPs have mostly been a tool used at the state and local level, which have to be more creative in part because they don’t have Uncle Sam’s seemingly unlimited line of credit.

An idea of the challenge of producing such a database can be seen in the fact that the state of Michigan alone has 83 counties, 1,240 Townships, 276 cities, 257 villages, 550 school districts and more than50 intermediate school districts _ all capable of engaging in P3 activity, from big cities like Detroit to little ones like Ubly in the Thumb area.

Yep – even little Ubly with some 853 residents has a P3 story to tell. Their school district leases non-teaching employees from a private company and claims to have saved 9.3 percent per contracted kitchen and custodial employee and 8 percent on bus drivers. This may cost me, but perhaps this talk should be called, “P3s: The Good, The Bad, and the Ubly.

History – Modern Era

The modern era of P3s in the U.S. may have begun with a 1979 document known as Office of Management and Budget Circular A-76. This established — I quote — “federal policy regarding the performance of commercial activities,” which included procedures for “determining whether commercial activities should be performed under contract with the private sector or not.”

At the state level, the first law passed to facilitate P3s was a 1989 California measure specific to transportation infrastructure. Today, some 35 states maintain P3 laws for transportation issues. Michigan, incidentally, is not one of them — yet.

House Bill 4925, sponsored by Rep. Marilyn Lane of Fraser was introduced in August of last year, but appears to have languished. It’s ironic that someone named lane would introduce a P3 transportation bill, particularly because the bill prohibits the use of tolls or user fees to pay for construction, unless lanes need to be added to accommodate capacity.

Public-Private Partnerships – Examples

Governments typically adopt P3s to maximize existing financial resources and talents. Theoretically they promise savings or revenue generation while minimizing risks. In some cases a government is more concerned with generating a particular qualitative outcome than saving money. In Sandy Springs, Ga., the P3 model was adopted to both save money and to provide higher quality services. It has become something of a poster child in this area

Local P3s: Sandy Springs, Ga.

Sandy Springs is part of the greater Atlanta area and has a population of 94,000. Starting almost 15 years ago, it basically outsourced all city services except police and fire to a private corporation through a P3 arrangement. Today, not counting police and fire, the city has just 10 people on its payroll. Both CNN and the Reason Foundation have profiled the town’s story — which I think you’ll enjoy:

Note: Video runs almost 8 minutes.

As you can imagine, Sandy Springs is now considered the model for an all- encompassing local public-private partnership. So far, the only real knock I’ve seen is not that the deal failed to save money or improve services, but that the experience may not be applicable to other city situations.

I’m not so sure, but it’s true that Sandy Springs did this from scratch when it first incorporated as a city in 2005, which among other things made dealing with legacy costs like pensions a small or nonexistent issue. We don’t really have an example yet of an established municipality becoming a “contract city” to this extent — except possibly the city of Pontiac, right here in Michigan, and over just the past five years.

As many of you know, Pontiac has had three emergency managers appointed by two governors to run the city under three different laws. This has reduced city employment from 495 (excluding court employees) to just a proposed 20. Most of Pontiac’s services, including law enforcement and fire, are provided under contract now, either with another unit of government (like the Oakland County Sheriff’s Office) or some private entity.

Chattahoochee Hills, Ga.

An interesting footnote to the Sandy Springs story involves a sister city called Chattahoochee Hills, Ga., about 30 miles away. In 2007 this affluent area of Fulton County with about 2,300 people also chose to incorporate as a city.

They wanted to go the Sandy Springs route but due to the small size of the contract — worth only about $600,000 — could only strike a deal with the company that manages Sandy Springs to provide a single on-location employee to interface with the public. In other words, if someone wanted to complain about a pothole, the local guy would take the information and send it to contractor’s regional office in Sandy Springs for action.

This wasn’t viewed as fully satisfactory, and the people of Chattahoochee have been evolving away from their P3, voting to raise property taxes and create a larger core of official government employees with whom they can interact in a personal level.

This P3 had been unique as far as we know in the United States. No other P3 was remoted to an office 30 miles away.

Lucerne, Pa., and “Kids for Cash”

One local P3 that may qualify for “the ugly” example in this talk  involved a “Kids for Cash” scandal between the operator of two private juvenile detention centers and two judges in Pennsylvania’s Lucerne County. Actually, it goes beyond mere ugly into downright evil. As such, it’s more an example of rank corruption than a good faith effort that went awry, but this too is relevant to any discussion of PPPs.

In 2002 one of those two judges arranged to shut down a state-run juvenile detention center and use county money to “lease” the function out to a private firm that would build and operate the facility. According to The New York Times, the deal was done in secret and the lease was worth some $1.3 million per year plus millions more for operational expenses.

The two judges were accused of pressuring the facility’s owners for kickbacks in return for funneling misbehaving juveniles into the operation with punitive sentences. Both judges were eventually sentenced to lengthy prison terms (28 and almost 18 years, respectively), while the detention center co-owner and builder were treated less harshly. A documentary on this subject was released in February of this year and is titled “Kids for Cash.”

Michigan Corrections P3

Michigan has also experienced a checkered corrections-related P3 deal during the Engler-Granholm era. Thankfully it didn’t involve corruption, but may illustrate a more pervasive problem, which is politicians serving the interests of a special interest — the state’s prison guard union — ahead of taxpayers. Baldwin, Mich., was the location selected for a build-own-operate partnership between the state and a company called the “Wackenhut Corrections Corp.,” known today as the GEO group.

Sold as a cost-effective way of housing young law breakers, the project was begun under Gov. John Engler in 1999 and terminated by Gov. Jennifer Granholm in 2005, her third year in office. At the time, a state Auditor General report that the per-day expense of housing juveniles exceeded all but four of Michigan’s 37 other prisons got a lot of media attention. Less attention was paid to a 2002 campaign promise then-candidate Granholm made to the Service Employees International Union-affiliated Michigan Corrections Organization to shut down their private competitor.

We know this because the union bragged about it in its official newsletter, writing, “Last year, Governor Granholm’s budget eliminated funding for the Michigan Youth Correctional Facility … fulfilling a promise Candidate Granholm had made to the MCO.”

This story is not as simple as a candidate keeping a promise to shut down a private prison that some proved did not live up to its billing. There is some evidence to suggest that the Granholm administration facilitated the failure. My colleague Jack McHugh has described exactly how on our web site:

To sell it to the public, most of the dangerous prisoners that the privately-run Baldwin “punk prison” was designed for were first removed, and then dishonest figures were created showing higher “costs-per-prisoner.” These failed to point out that if you remove most prisoners but require the same number of guards and other fixed costs, the per-prisoner costs will be higher.

Incidentally, the House and Senate Fiscal Agencies repeated that info but gave no hint that there was anything wrong with it. Republicans had control of the House and Senate then, but to my knowledge they never questioned why their “non-partisan” agencies weren’t shedding more light on those flawed analyses.

To be sure, prison P3s and related private contracting have been controversial, and not just because they threaten entrenched special interests. Some have been done well and some poorly, but as the Baldwin example shows observers should be cautious about taking official explanations at face value.

Transportation P3s

Probably the highest profile P3s in this country are state-level transportation initiatives. Nationwide, more than 80 transportation-related P3s were completed between 1990 and 2010, accounting for investments that totaled more than $46 billion, according to the National Conference of State Legislatures. Here are a few that are currently underway:

  • As of February 2013 there were over $7 billion in transportation-related P3s underway in Texas according to the Reason Foundation.
  • Another $2 billion worth are under construction in Florida and Virginia.
  • In 2011 Puerto Rico sealed a deal on a 40-year $1.5 billion P3 deal to “improve, operate and maintain toll roads.”
  • In Indiana, a 2006 toll road deal generated some $3.8 billion in revenues to the state — much of it from private equity — in exchange for a 75-year concession agreement.

Indiana has reinvested $2.8 billion of that back into other road-related projects while distributing other revenue to all 92 counties in the state for local road work, according to the Reason Foundation. The seven counties through which the toll way runs received a larger share.

According to the Indiana Department of Transportation, spending on road construction and repair averaged $750 million per year. One-third of that went to new construction and the rest went to what the state called “preservation projects.” Since adoption of the long-term lease and creation of a highway infrastructure plan called, “Major Moves,” INDOT has been able to lift spending to more than $1 billion per year.

One tally of infrastructure projects facilitated (but not completely funded) by this money in 2012 (additional federal dollars were mixed in) included:

  • “65 roadway projects were complete or substantially under construction
  • 19 roadway projects were accelerated — when compared to the original 2006 plan
  • 375 new centerline miles complete
  • 48 new or reconstructed interchanges
  • 5,030 preservation centerline miles complete – 40 percent of the state’s inventory
  • 720 bridges were rehabilitated or replaced — 13 percent of the state’s inventory.”[2]

The agreement is a detailed one and demands the concession holder provide basic repair and maintenance meeting specified standards and worth some $4.4 billion over the life of the contract. It caps toll rate increases, details performance standards for dealing with disabled vehicles, dead animals, snow and even the timely removal of graffiti.[3]

Governing magazine wrote in 2011 that “By most accounts, the project has been a windfall for Indiana, with little downside to taxpayers.”[4] This deal arguably qualifies as one of the “good” examples of a transportation P3 in action, but it still hasn’t stopped critics from leveling some reasonable charges.

First, it is unlikely that other systems will see any comparable windfall. Related, it’s been reported that the concessionaire has been struggling to generate sufficient revenues and could default on debt payments due in 2015. If that happens taxpayers are still protected — the state need not return any money — but it might chill future deals.

Lastly, and this is a point often debated at the Mackinac Center, what value is added by politicians giving up a huge annual revenue stream for years and decades to come in return for a pile of money now? The toll road contract may turn out to be one example where this makes sense, but others like a Chicago parking deal look an awful lot like borrowing to cover current spending disguised as “privatization.”

The Chicago parking meter deal, by the way, was briefly addressed by David Lick last week, noting that the deal was not well written to protect consumers. This may have spiked privatization of Chicago’s Midway.

Taxpayers and voters should be extremely suspicious when politicians propose getting a big pile of money today in return for foregone revenues tomorrow, no matter how strong a “lockbox” they say they’ll build around that money.

London Underground

While the Indiana P3 looks like it may be an example of a “good” P3, there are several contenders for a “bad” one, with the London Underground leading the list. Some of the literature on the subject reads like case studies in what not to do.

The London Underground Public-Private Partnership was a large, lengthy and complex P3s designed to improve the city’s underground transit infrastructure over 30 years. Under the deal that began in 2004 the government would continue to operate the trains but three private agreements with two concessionaires would maintain the London “Tube” infrastructure, and replace trains and other equipment as needed. 

The P3s began in 2004 — under some heavy and perhaps legitimate criticism — and ended in 2010 due to failures of the P3 consortiums. The first involved the Metronet concessionaire, which had won two of three P3 contracts. Both failed by 2008 and the government agency responsible for operating the trains had to buy out most of Metronet’s debt from creditors. The taxpayer losses associated with these two contracts are estimated to be as high as £410 million pounds.

The other P3 — known as Tube Lines — died in 2010 after it could not persuade a   “Tube Arbiter” to meet its compensation demands and thus went bankrupt. This deal apparently cost taxpayers another £310 million pounds.

Adding to the stink, these deals allegedly cost as much £450-£500 million pounds just to assemble in the first place.

What went so horribly wrong? It was a parade of errors, some that should have been obvious from the start and others that only came to light after post mortems, all of which are neatly summed up in a great 2010 working paper titled, “Analysis of the London Underground PPP Failure,” by Trefor Williams.

There are three big takeaways:

First, the contracts were written in such a way that each party could squabble over what constituted payable work. As Williams writes, “The government believed it had purchased an output based fixed price contract and the private consortiums behaved like they had agreed to a series of heterogeneous cost plus projects.”[5] Williams says the agreements contained “vague wording” that “led to disputes and conflicts…”

Second, it was hard for both the government agency overseeing the P3s and one of the concessionaires to simply obtain cost-related information. Disbursements were not set in stone at the outset due to the countless variables involved in infrastructure maintenance over time. Metronet apparently even had trouble getting cost data from its own subcontractors.

Third, Metronet was the child of companies to whom it had given supply contracts. So, in short, its parents were denying it access to detailed cost data it should need to justify expenditures to its government “partner.”

Perhaps most galling, while Metronet evaporated, its shareholder parents got paid, at least for a while. Williams points out that Metronet received some £3 billion for services, a big chunk of which went to one of the entity’s chief shareholders, such as Bombardier, a manufacturer of trains and planes whose losses from the Metronet bankruptcy were likely mitigated by supply contract revenues.

The list could go on but you get the picture. Obviously this was a huge black eye for the P3 concept, but to put it into context, the Tube upgrades had been a disaster even before the private partnership debacle. According to a 2011 article by P3 critic and author Yonah Freemark, in the 1990s the government had incurred £1 billion in underground infrastructure cost overruns.

Furthermore, Freemark notes, the government arbiter who had denied disbursements to the Tube Lines claimed it was due to reports that the public agency that stepped in after the P3 bankruptcy in 2011 is actually more expensive than the concessionaire had been.

While there have been troubled transportation P3s on this side of the pond too, none have failed as spectacularly. Last November The Wall Street Journal cited several struggling toll road P3s. There were bankruptcies by companies associated with them in Virginia, California and even the holding company “American Roads,” which operates the Detroit-Windsor tunnel. It pays a fee to Detroit under a lease agreement. These are more exceptions than rules as not all P3 toll ways are in trouble.

However, as transportation expert Robert Poole has pointed out, these projects have stumbled in part due to the fact operators failed to anticipate traffic declines associated with the Great Recession among other reasons. As one example, the South Bay Expressway in San Diego opened in November of 2007, just before America’s financial crisis began and at the same time gas prices hit record levels. The toll way still exists and is operational, but the operator is new and consists of its creditors.

In their own bankruptcy filings the operator of the Detroit-Windsor tunnel pointed to population decline in the Detroit area. Recall that Detroit lost 25 percent of its population from 2000 through 2009 and Michigan was the only state in the country to lose population. There were simply fewer people around to use the tunnel.

That doesn’t mean that P3s in general are failures, but it does argue that states should insulate themselves as much as possible from losses associated with bad forecasts. The Indiana toll deal does this remarkably well. In a worst case scenario the state might have to take the toll way back, but it keeps the money paid to it and could simply offer up a new concession agreement to some other firms.

The Future of P3s — Social Impact Bonds

Social Impact Bonds represent perhaps the most cutting-edge use of Public-Private Partnerships. These deals involve payment for performance or success for particular social outcomes, such as reducing recidivism rates among prisoners. The first of these were hatched in the U.K. in 2010, and the concept quickly moved here too.

A 2013 report by a British nonprofit called “Social Finance” describes the goals:

“Align the public sector funding with improved social outcomes, directly linking spending to outcomes achieved;

Increase the capital available to support prevention and early interventions in areas like recidivism reduction, workforce development and healthcare, which can reduce future spending on government remedial programs …

Enable collaboration among a broad range of social service providers;

Provide greater revenue certainty for effective service providers; and

Encourage rigorous performance management and objective outcome measurement.”

It works like this. A unit of government contracts with investors in the private or non-profit sector to achieve a measurable social outcome at their expense. If the goal is reached the investors receive a return on their investment.

The first such PPP in Britain raised nearly $8 million to lower recidivism rates amongst male prisoners. This was a pilot project and involved some 17 different investors, including nonprofit organizations. The goal was to reduce recidivism rates to below 7.5 percent within six years. The payoff would be up to 13 percent of the dollars invested. The final results won’t be known for a while, but in the worst case taxpayers won’t lose money on the deal, save perhaps for some oversight costs.

Stateside, similar programs are underway. In Massachusetts Gov. Deval Patrick launched a $27 million program called “Juvenile Justice Pay for Success” to lower juvenile recidivism. A non-profit group called “Roca, Inc.” will provide coaching and job training over seven year to 900 youngsters who have served their terms or are on probation. The main goal is to reduce future jail stints by 40 percent. Once again, investors will get paid for performance – if they succeed... (Returns on investment to at least one investor are capped at $1 million.)

As we have seen, P3s can be done well or poorly. A growing portfolio of case studies suggests some guidelines and best practices. Examples include:

Determine if it would be more cost effective to provide an asset or service in-house or though contracting. Privatization and P3s should not be undertaken for the sake of doing it.

  • Conduct a cost-benefit analysis to determine if the P3 provides the best overall value for the money. Sometimes the primary goal isn’t to save money but rather to improve a service. Both cost and quality should be examined.
  • To bundle or not to bundle. Hoped-for economies of scale that derive from bundling services into all-encompassing contracts can be a mirage if a project is too complex. The London Underground P3s failed in part due to the hyper-complexity and length of the agreements.

Transparency is a Need-to-Have, not a Nice-to-Have. Starting with the public and private partners. It seems obvious, but both sides need to understand who is responsible for meeting performance benchmarks and when. A transparent process facilitates greater trust and understanding.

  • Appoint a Champion. There must be an office or person with whom the P3 buck stops. Specifically, there needs to be a person — a voice, if you will — who can take responsibility and credit for the successes or failures, and field questions from lawmakers, the press and public. They must be quick to communicate to all stakeholders.
  • A P3 manager must ensure against conflicts of interest. Once again, the London Tube provides our negative role model. The private-sector player was simply a subsidiary of corporate parents who were able to lock in lucrative contracts. Neither side could get vital cost information to determine a proper payment schedule. 
  • Clearly define performance metrics and benchmarks.

In too many contracts exactly what is expected from all parties is simply not explicit enough.

Some years ago Michigan contracted for highway maintenance without benchmarking the results against what the government paid to provide the same services in-house. The result? No one knew whether or not the experiment was a success or not because there was no data to measure it against.

  • Detail proper incentives.

Governments that adopt P3s, especially large, expensive ones such as those in transportation, should adopt performance incentives and penalties while guarding against opportunists. Governments risk demands from contract winners to reopen negotiations for more favorable terms because contractors know government has no wish to pursue an expensive retendering process. In the academic literature this is often called, “weak governance.”[6]

  • Monitor. Monitor, Monitor.

The Mackinac Center has said this too many times to remember. The government simply cannot sign off on a privatization or P3 deal and then just walk away. There must be a contract monitor to ensure that the all parties to the transaction are keeping their end of the bargain.

P3s are a tool, and like any tool they can be used well or badly. Moreover, doing it well requires intense, focused, eyes-wide-open management by the public sector partner, before and during the project. These are not “fire-and-forget” missiles designed to relieve government managers of responsibility. Indeed, just like there are “no bad dogs, only bad masters,” the ultimate responsibility for the success or failure of a P3 falls squarely on government officials and their political masters.







[1] Leonard Gilroy, “Emerging Paradigm of Public-Private Partnerships,” 2013 APEEP Annual Conference, Scottsdale Arizona.

[2] Indiana Department of Transportation Web Site:

[3] Reason Foundation, “Leasing the Indiana Toll road: Reviewing the First Six Years Under Private Operation.” 4.

[5] Ibid. Williams. 9.

[6] “Public-Private Partnerships and Investment in Infrastructure,” OECD Economics Department, Working Papers No. 803. P. 11 and “Introduction to the IJIO Symposium on Public/Private Partnerships,” International Journal of Industrial Organization 26, (2008) P. iii-iv. The latter refers to the “high rate of renegotiation” in PPPs, as detailed in “Renegotiation of concession in Latin America.”