A zebra or a painted horse? Are hospital PPPs infrastructure partnerships with stripes or a separate species?

Abstract

Public Private Partnerships (PPP) have been common in infrastructure for many years and are increasingly being considered as a means to finance, build, and manage hospitals. However, the growth of hospital PPPs in the past two decades has led to confusion about what sorts of contractual arrangements between public and private partners consititute a PPP, and what key differences distinguish public private partnership for hospitals from PPPs for infrastructure. Based on experiences from around the world we indentify six key areas where hospital PPPs differ from infrastructure partnerships. We draw upon the hospital partnerships that have been documented in OECD countries and a growing number of middle-income countries to identify four distinct types of hospital PPPs: service focused partnerships in which private partners manage operations within publicly constructed facilities; facilities and finance PPPs, focused on mobilizing capital and creating new hospitals; combined PPPs, involving both facility and clinical operations; and co-located PPPs where privately operated services are developed within the grounds of a public hospital. These four types of hospital PPPs have differing goals, and therefore different contractual and functional aspects, as well as differing risks to both public and private partners. By clarifying these, we provide a base upon which hospital PPPs can be assessed against appropriate goals and benchmarks.

Full text

World Hospitals and Health Services Vol. 48 No. 2 15
Private hospitals and health care
A zebra or a painted horse? Are hospital
PPPs infrastructure partnerships with
stripes or a separate species?
ABSTRACT: Public Private Partnerships (PPP) have been common in infrastructure for many years and are increasingly being considered
as a means to finance, build, and manage hospitals. However, the growth of hospital PPPs in the past two decades has led to confusion
about what sorts of contractual arrangements between public and private partners constitute a PPP, and what key differences
distinguish public private partnership for hospitalsfrom PPPs for infrastructure.
Based on experiences from around the world we identify six key areas where hospital PPPs differ from infrastructure partnerships.We
draw upon the hospital partnerships that have been documented in OECD countries and a growing number of middle-income countries
to identify four distinct types of hospital PPPs: service focused partnerships in which private partners manage operations within
publicly constructed facilities; facilities and finance PPPs, focused on mobilizing capital and creating new hospitals; combined PPPs,
involving both facility and clinical operations; and co-located PPPs where privately operated services are developed within the grounds
of a public hospital.
These four types of hospital PPPs have differing goals, and therefore different contractual and functional aspects, as well as differing
risks to both public and private partners. By clarifying these, we provide a base upon which hospital PPPs can be assessed against
appropriate goals and benchmarks.
Public Private Partnerships (PPP) are used for differing
reasons across a range of industries. PPPs in the water
sector have successfully mobilized private operators to turn
around failing public companies and expand access to water
services. In the building sector, PPPs have transferred
responsibility of construction and estates management to private
companies, leaving government departments to focus on their
core activities (Grimsey and Lewis 2004). Increasingly
policymakers are exploring PPPs as a means to improve their
public hospitals. However, the performance goals and policy
context for hospitals differ considerably from those in which PPP
models evolved (Grimsey and Lewis 2005; Brinkerhoff and
Brinkerhoff 2011). Discussions about PPPs in the health care
sector are often hampered by confusion about what the term
means, with multiple models grouped without distinction on the
umbrella PPP term (Field and Peck 2003). Lacking a clear
vocabulary, health policymakers find it difficult to sort out what
these “imported” models offer and it is difficult to understand
which models are likely to address the performance problems for
which a specific PPP is contemplated (Ng and Loosemore 2007).
We review the PPP models most frequently applied in other
sectors, and increasingly in hospitals, and use configuration
analysisito group them into categories with analytically important
DOMINIC MONTAGU
ASSISTANT PROFESSOR OF EPIDEMIOLOGY, GLOBAL HEALTH
GROUP, UNIVERSITY OF CALIFORNIA SAN FRANCISCO, USA
APRIL HARDING
SENIOR HEALTH ECONOMIST, WORLD BANK INSTITUTE, USA
distinctions among them. We propose a typology of hospital PPPs
to permit clearer communication and more sound analysis. Clearer
specification of their characteristic mechanisms also illuminates
the problems that each PPP type has been “built” to address.
Establishing analytically meaningful categories allows researchers
to compare “like with like”. We hope that this typology will support
much needed evaluative research in this field.
Defining PPPs
PPPs are a form of contract between a government and a private
entity in which the private partner undertakes the long-term
provision of publicly beneficial services. Initial injection of capital by
the private partners is a key component of many, though not all
PPPs. What critically distinguishes a PPP from a service contract
is the duration and intended distributed benefit.
Box 1. Definition of PPP
“A more or less permanent cooperation between public and
private actors, through which the joint products or services are
developed and in which the risks, costs and profits are shared.”
KLIJN ET AL. (KLIJN ET AL. 2007), CITED IN TYNKKYNENAND AND LEHTO (2009).
15-19 Harding_Montagu:33 21/6/12 09:26 Page 15

Private hospitals and health care
16 World Hospitals and Health Services Vol. 48 No. 2
An important source of confusion about what a PPP is and what
it is intended to deliver to the health system derives from the
different perspectives of two groups of professionals. One group
has a background in infrastructure PPPs – and they seek to bring
the benefits of this model to the hospital sector. In infrastructure
sectors, PPPs are implemented mainly as a means to mobilize
private capital, transfer investment risk, and consolidate the
finance, construction, and maintenance activities into a single
contract for easy of management by government,with ensuing
gains in efficiency by the private operators. Often the advantage of
an infrastructure PPP is that financing can be recouped by direct
billing of service users, bypassing government budgets entirely.
The infrastructure model proponents view the mobilization of
private finance as the critical component of a PPP, central to the
benefits of the model, and the focal point for contractual attention
and negotiation (Brown 2007).
The second group comes to PPPs from health services and is
engaged with PPPs as a means to bring private management
expertise, flexibility, and care delivery models to hospital
operations. This group focuses on service contracting,
responsiveness, and efficiency in both the conceptualization and
assessment of PPPs. Hospital operation and management
experts view the incorporation of better management systems
from the private sector as central to improving health services, and
as the core benefit, and focus, of a hospital PPP (Dorganet al.
2010). Because of the differing perspectives, the collection of
hospital PPPs implemented around the world include examples
that each group view as having been undertaken for the wrong
reasons, and often as being inappropriately designated a PPP. As
we will explain below, we believe that the commonalities of
partnership arrangements and duration of engagement justify the
inclusion of partnerships as defined by both groups in the same,
shared, PPP designation. That said, we believe it
is critical for policy-maker, analysts, and
economists to distinguish the structural features
and objectives that drive each transaction.
Hospital PPPs vs Infrastructure PPPs?
The documentation on infrastructure PPPs is
growing, and the models, risks, benefits, contract
structures, and financing issues are understood
(Brown 2007). Since at least the 1800s,
governments have sought to encourage private
investment in areas of public benefit through
mixtures of land grants or long-term leases
(railways, toll highways), monopoly grants
(canals), and enduring purchase commitments
(water and electricity). The pricing of assets,
loans, and share of income or fee waivers have all
grown more sophisticated, but an infrastructure
PPP today is very similar to those from a century
ago. As in the days of railway PPPs, profits are
gained largely through better management, use
of zero-cost land leases, and monopoly or quasi-
monopoly control of a resource used by many
purchasers. For the government, these PPPs are
attractive because the risk and effort of investment is taken on by
a private entity. Society as a whole benefits from having the new
infrastructure or utility services that otherwise would not exist.
The mechanics and sources of gains in infrastructure PPPs
translate imperfectly to hospitals and healthcare. Acknowledging
minor variation between hospital PPP models, there are six key
issues that are common to hospital PPPs and make them different
from their infrastructure homonyms:
!Government, not individual, is primary purchaser of
outputs: Infrastructure PPPs commonly collect fees from
multiple consumers – drivers on a highway, passengers on a
railway, factories and homes receiving water or electricity. By
contrast, hospital PPPs typically receive nearly all of their
income from government in the form of scheduled lease
payments or unit service payments. This simplifies, constrains,
and adds risk to the income stream of private operators in
hospitals.
!Partnership risks are political rather than marketplace: As
a result of the government primacy in purchasing noted above,
the risks of hospital partnership success are often more due to
uncertainty about long term compliance with payment
obligations, than market demand projections. For this reasons
the borrowing costs for hospital PPPs are usually higher than
the cost of infrastructure financing.
!Measurability of output: Infrastructure PPPs deliver highly
measurable outputs, whether power, water, road access, or
otherwise. Inpatient services are immensely varied based on
the condition, co-morbidities, and patient characteristics and
largely unobservable (Preker et al. 2000).
!Varia bi lity of outputs ov er ti me : During the 20 to 30 year life
of a typical hospital PPP the population served by the facility
can be expected to change in composition, wealth, age, and
illness. This is particularly true in low- and middle-income
countries (LMICs) where both demographic and epidemiologic
transitions may be occurring simultaneously, contemporaneous
iConfiguration analysis involves comparing instances of similar institutional arrangements (or
systemic reforms) to identify recurring functional patterns to enable them to be distinguished
into analytically meaningful categories (see, for example Rothgang 2010).
PPP Category Common term Definition/ Explanation
Services Operating contract
Facility/finance PFI
Combined BOT
PPIP
Co-location Co-location
Source: Authors’ analysis
A private organization is brought in to operate and
deliver publicly-funded health services within a
public facility.
A public agency contracts a private entity to
finance, design, build and operate a hospital
facility. Heath services within the facility are
provide by government.
A private organization establishes capacity
(through new construction or expansion of
existing facility) to provide health services
under sustained public or social insurance
reimbursement.
A public agency allocates a portion of a public
hospital’s land and/or premises for sustained use
by a private organization in exchange for payment
and specified benefits to the public agency.
Table 1: A Typology of Hospital PPPs
15-19 Harding_Montagu:33 21/6/12 09:26 Page 16

World Hospitals and Health Services Vol. 48 No. 2 17
Private hospitals and health care
with rapid economic development. All of these will affect the
medical service mix, or outputs, of the hospital. By contrast, in
infrastructure PPPs variation in output volume is normal, but
output mix is constant.
!Varia bi lity of technol og y a nd or ganizat io nal config ur ation
over time: The pace of change of medical service delivery is
increased every year due to changing regulation,
reimbursement systems, technology, and better information.
Across the OECD there are large changes each year in the
inpatient-outpatient mix, the duration of stay for each service,
with new technology leading to changes in diagnostic and
treatment protocols, and care shifting from doctors to nurses
to physician assistants, and sometimes back in the other
direction. The unpredictability of these shifts, together with the
high proportion of overall project cost that is due to operations,
is unique to hospitals PPPs.
!Ratio of investment capital to operating capital: Over the
life of a hospital clinical, laboratory, pharmacy and medical
services represent 65 per cent of annual operating costs and
ancillary services (food, support, IT) another 17 per cent.
Facility maintenance and amortized construction costs are less
than one fifth of total budget (EBRD2007). For infrastructure
operations, the cost of design, construction, finance and
maintenance are the large majority of total costs. This means
private involvement in design, construction, and maintenance
of hospital entities has a lower potential for efficiency gains
than in infrastructure projects. In hospitals, the majority of the
potential efficiency gains come from service provision.
The implications for governmental obligations and transaction
gains or costs are summarized in Table 1.
The defining aspect of hospital PPPs is the relationship between
public and private partners which cannot be fully planned out in
advance, and which therefore necessitates on-going active
discussion and renegotiation during the lifetime of the partnership.
For this reason, the challenges of contract management are much
greater, and the benefits to government accrue from the private
participation in finance and facility provision are often less
Infrastructure PPPs Hospital PPPs Implications
Government vs Private
purchaser of output
Business risk vs Political risk
Measurability
Variability of outputs over time
Variability of technology over
time
Ratio of onvestment to
operating capital
Source: Authors’ analysis
• Private buyers/payers
• Government does not enter
into long-term service
purchasing relationship as part
of transaction
• Borrowing costs reflect
estimated risk of demand for
infrastructure services by total
market of potential payers
• Comparators for benchmarking
cost of facility availability
services are somewhat limited
• Products stable over time
• Service delivery technology
and organizational models
change slowly
• High ratio of capital to
operating costs
• Government (or social health
insurers) buy all or most
services
• Government enters into
long-term service purchasing
relationship as part of
transaction
• Borrowing costs reflect risks
associated with single (or
multiple) government payer
agencies
• Comparators for benchmarking
cost of services often
extremely limited
• Products highly variable due to
volatility in demographics and
disease
• Service delivery technology
and organizational models
change rapidly
• Low ratio of capital to
operating costs
• Substantial risks to government
payer as a result of long-term
funding “lock in” obligation
• Substantial political risks to private
partners in hospital PPP
• Cost of finance (and therefore
capital) higher for hospital facility
investment
• Probability of that payment contract
will set excessive rates is higher for
hospitals
• Risk to private partners necessitating
either higher return contingencies, or
flexibility in contract modification
• Risk to government due to “locked
in” commitment to hospitals/
configuration that may not be
needed in the future
• Risks to government and private
partners as a result of lost flexibility
to adapt service organization; or cost
of unpredictable adjustments to
technology, systems and staffing
• Efficiency gains from private finance/
design/ construction and operation
of hospital PPPs lower than for
infrastructure PPPs
Table 2: Major differences between hospital PPPs and infrastructure PPPs
ii The US Centers for Medicare and Medicaid Services list 998 different diagnostic codes in their
most recent guide
15-19 Harding_Montagu:33 21/6/12 09:26 Page 17

Private hospitals and health care
18 World Hospitals and Health Services Vol. 48 No. 2
predictable in hospital PPPs than in infrastructure PPPs.
Typology
Hospital PPPs range from the health service focused to the
infrastructure focused. In many instances PPPs incorporate
aspects of both. From our review of documented hospital PPP
initiatives, four distinct structures, or types, emerged,
distinguished by what the public partner is “buying or selling”
from/to the private partner, and the primary objectives of the
partnership.
In the first model, “services”are the core of the partnership. In
order to improve the quality and/or efficiency of hospital services
provision, aprivate organization is brought in to operate and deliver
publicly-funded hospital services, usually within the existing
infrastructure of the government (La Forgia and Harding 2009).The
facility may be built by government explicitly in preparation for this
model of service PPP.
The second model involves a public agency contracting a
privateentity to finance, design, build and operate a hospital facility
within which a public service is run. We refer to this as the “facilities
and finance” model. It is popularly referred to as the PFI model,
coming from the name of the “Private Finance Initiative” program
which first applied the model in the UK (McKee et al. 2006,
Edwards 2005).
Under the third model, a private organization establishes
capacity to provide hospital services under sustained public or
social insurance reimbursement. We refer to this as the
“combined” model, since the public sector “buys” hospital
services combined with the underlying facilities and related
finance. There are two variants under this model, which merit
distinction. In one variant, a public agency tenders to have a
private organization build a new facility and provide services; in the
second, the private organization takes over an existing facility and
services. The former has been applied to add capacity and assure
operator “ownership” in the facility construction (Global Health
Group 2010, Sekhri et al. 2011). The latter is often more politically
controversial, but has the potential to harness the private sector to
take over existing failing facilities and turn them around, as has
been the case in Germany (Roeder 2012, Coelho and O’Farrell
2009).
Under the fourth model, public agencies do not buy services but
rather allocate public hospital real estate for a private service
provider to develop services “co-located” within the public facility.
The private organization makes payments to the public agency, as
well as providing in-kind services. Most often the objective of this
kind of partnership is to capture the value of the real assets, and
to tap the private operator’s services and expertise for the benefit
of public patients. Facilitating dual practice as an incentive to
public practitioners is sometimes a secondary goal. (Nikolic and
Maikisch 2006, Project Equity 1999).
Conclusion
The lack of a structured vocabulary to distinguish among differing
models of PPPs has led to misguided analysis, misdirected
criticism between the “service” and “infrastructure” proponents,
and inhibited needed attention to the different goals of each PPP
type, and the financial, regulatory, and contractual mechanisms
applied to advance those goals.
Building upon the typology presented here, we expect that
differing models of PPP will be suited to application in a variety of
situations depending upon the specific facility and system need;
the public and private capacity to fund the PPP; the governmental
capacity to contract and oversee; private capacity to implement,
and the legal and healthcare system infrastructure in which the
PPP is applied. To fully understand the tradeoffs between differing
models, and the context appropriate to each, a more compete
analysis of experiences within each model is needed. Only at this
point will it be possible to define success for each PPP type, and
to both build and test a model of the criteria that will make a PPP
likely to succeed or fail. The typology presented here is an
advance towards this goal. !
Dominic Montagu is an assistant professor of epidemiology and
biostatistics and lead of the Health Systems Initiative at the Global
Health Group of the University of California, San Francisco. His
work is focused on private delivery of health services in developing
countries and on market function for health services and health
commodities. He holds Masters degrees in business administration
and public health and a doctorate in public health from the
University of California Berkeley. Dr Montagu has worked
extensively in Africa and Asia, and teaches on the private sector in
developing countries, and on regulation of private hospitals and
public private partnerships at UCSF, UC Berkeley and on behalf of
the World Bank Institute.
April Harding is an economist and health systems specialist with
the World Bank Institute. She is a sought-after speaker, author, and
policy adviser on the private health sector, public private
partnerships, as well as hospital reform and governance. April has
provided policy advice and analytical support to more than 20
governments of transition and developing countries on these
topics. She recently returned to the World Bank from the Center for
Global Development, where she undertook research examining five
global health programs, child health, TB, malaria, family planning
and HIV/ AIDS, looking at how these programs interact with the
private sector in their implementation, and how this contributes to
their success or failure. Her findings are presented in her book
“Private Patients: Why Health Aid Fails to Reach So Many, and
What We Can Do about It” (forthcoming 2012, Brookings/ Center
for Global Development). She served as a contributing editor at
Health Affairs, where she helped the journal develop their global
health coverage. Prior to joining the World Bank, April was a
research fellow at the Brookings Institution. She received her
doctorate in Economics at the University of Pennsylvania.
15-19 Harding_Montagu:33 21/6/12 09:26 Page 18

World Hospitals and Health Services Vol. 48 No. 2 19
Private hospitals and health care
References
Grimsey D., Lewis, M.K., 2004. The Governance of Contractual Relationships in Public-Private
Partnerships.
The Journal of Corporate Citizenship
15:91.
Grimsey D., Lewis M.K., 2005. Are Public Private Partnerships Value for Money?
Accounting
Forum
29:345-378.
Brinkerhoff D.W., Brinkerhoff J.M., 2011. Public–Private Partnerships: Perspectives on
Purposes, Publicness, and Good Governance.
Public Administration and Development
31:2-
14.
Field J.E., Pec k E., 2003. Public–Private Partnership s in Healt hcare: the Manager s’
Perspective.
Health & social care in the community
11:494-501.
Ng A., Loosemore M., 2007. Risk Allocation in the Private Provision of Public Infrastructure.
International Journal of Project Management
25:66-76.
Rothgang H. [Fairness in Germany's long-term care insurance: the relationship between social
insurance and private mandatory insurance].
Gesundheitswesen
72:154-160.
Klijn E.H., Edelenbos J., Hughes M., 2007. Public-Private Partnership: A Two Headed Reform. A
Comparison of PPP in England and the Netherlands.
New Public Management in Europe
Adaptation and Alternatives
71-89.
Tyn kkyn en L.K ., L eht o J ., 2 009 . An Ana lys is of Oph thal mol ogy Se rvi ces in Fin lan d-Ha s the
Time Come fo r a Public -Private Pa rtnership?
Health Research Policy and Systems
7:24.
Brown K., 2007. Are Public–Private Transactions the Future of Infrastructure Finance?
Public
Works Management & Policy
12:320.
Dorgan S.L.D., Bloom N., Homkes R., Sadun R., Van Reenen J., 2010.
Management in
Healthcare: Why Good Practice Really Matters
. Mc Kinse y and Co and L SE (C EP).
Preker A.S, Harding A., Travis P., 2000. “Make or Buy” Decisions in the Production of Health
Care Goods and Services: New Insights from Institutional Economics and Organizational
Theory.
Bull World Health Organ
78:779-790.
EBRD; 2007. Public Private Collaboration in Health Care Provision Montenegro: EBRD.
La Forgia G.M., Harding A., 2009. Public-Private Partnerships and Public Hospital Performance
in São Paulo, Brazil.
Health Affairs
28:1114-1126.
McKee M., Edwards N., Atun R., 2006. Public-Private Partnerships for Hospitals.
Bulletin of the
World Health Organization
84:890-896.
Edwards C.A., 2005. The Private Finance Initiative (PFI) and Value for Money? A case Study of
the Norfolk and Norwich University Hospital (NNUH). Norwich.
Global Health Group, 2010.
Public-Private Investment Partnerships for Health:An Atlas of
Innovation
. Sa n Fra ncisco : The Globa l Hea lth G roup.
Sekhri N., Feachem R., Ni A., 2011. Public-Private Integrated Partnerships Demonstrate The
Potential To Impro ve Health Care Access, Quality, And Efficiency.
Health Affairs
30:1498-
1507.
Roeder F.C., 2012. The Private Sector within a Public Health Care System: The German
Example.
Health Care
.
Coelho C., O'Farrell, C., 2009. Breaking New Ground: Lesotho Hospital Public-Private
Partnership – A Model for Inte grated Heal th Service s Delivery. IFC SMA RTLESSONS.
Nikolic I.A., Maikisch H., 2006. Public-Private Partnerships and Collaboration in the Health
Sector, in: The World Bank [Hrsg.]:HNP Discussion Paper,
http://siteresources.worldbank.org/INTECAREGTOPHEANUT/Resources/HNPDiscussionSeries
PPPPaper.pdf, Zugriff am: 29.02.2009.
Project Equity, 1999. Case Study: Successful Public/Private Partnerships: The Uitenhage
Experience.
15-19 Harding_Montagu:33 21/6/12 16:07 Page 19