5572.pdf
Media
- extracted text
-
05572
INFORMATION FOR READERS
To obtain copies of documents or any other information, please contact:
M.-C. van Holten
WHO/ICO
20 avenue Appia
1211 Geneva 27
Switzerland
Fax: +41 22 791 41 53
e-mail: vanholtenm@who.ch
Other titles in the "Macroeconomics, Health and Development" Series are:
N“1:
Macroeconomic Evolution and the Health Sector: Guinea, Country Paper - WHO/ICO/MESD.1
N° 2:
Une methodologie pour le calcul des couts des soins de santh et leur recouvrement: Document technique, Guinee WHO/ICO/MESD.2
N° 3:
Debt for Health Swaps: A source of additional finance for the health system: Technical Paper - WHO/ICO/MESD.3
N° 4:
Macroeconomic Adjustment and Health: A survey. Technical Paper - WHO/ICO/MESD.4
N° 5:
La place de I'aide extbrieure dans le secteur medical au Tchad : Etude de pays, Tchad - WHO/ICO/MESD.5
N° 6:
L'inftuence de la participation financibre des populations sur la demande de soins de santb : Une aide a la rbflexion
pour les pays les plus dbmunis, Principes directeurs - WHO/ICO/MESD.6
N° 7:
Planning and Implementing Health Insurance in Developing countries:
Principles - WHO/ICO/MESD.7
N° 8:
Macroeconomic Changes in the Health Sector in Guinea-Bissau, Country Paper - WHO/ICO/MESD.8
N° 9:
Macroeconomic Development and the Health Sector in Malawi, Country Paper - WHO/ICO/MESD.9
N° 10:
El ajuste macroeconbmico y sus repercusiones en el sector de la salud de Bolivia:
WHO/ICO/MESD.10
N° 11:
The macroeconomy and Health Sector Financing in Nepal: A medium-term perspective: Nepal, Country PaperWHO/ICO/MESD.11
N° 12:
Towards a Framework for Health Insurance Development in Hai Phong, Viet Nam:
WHO/ICO/MESD.12
N° 13:
Guide pour la conduite d'un processus de Table ronde sectorielle sur la santb:
WHO/ICO/MESD.13
N“ 14:
The public health sector in Mozambique: A post-war strategy for rehabilitation and sustained development: Country
paper - WHO/ICO/MESD.14
N°15:
La santb dans les pays de la zone franc face b la devaluation du franc CFA - WHO/ICO/MESD.15
N° 16:
Poverty and health in developing countries: Technical Paper - WHO/ICO/MESD.16
N° 17:
Gasto nacional y financiamiento del sector salud en Bolivia: Documento de pais - WHO/ICO/MESD.17
N“ 18:
Exploring the health impact of economic growth, poverty reduction and public health expenditure - Technical paper WHO/ICO/MESD.18
N“ 19:
A community health insurance scheme in the Philippines: extension of a community based integrated project.
Philippines - Technical paper - WHO/ICO/MESD.19
N° 20:
The reform of the rural cooperative medical system in the People's Republic of China, Initial design and interim
experience - Technical paper - WHO/ICO/MESD.20
Guidelines and Case Studies:
Guiding
Documento de pais -
Technical Paper -
Principes directeurs -
List of other available documents can be found on inside back cover
EXPERIENCES
OF CONTRACTING
AN OVERVIEW OF
THE LITERATURE
by
ANNE MILLS
Professor of Health Economics and Policy
London School ofHygiene and Tropical Medicine
and
JONA THAN BROOMBERG
Praxis Capital
South Africa
ADDRESSES FOR CORRESPONDENCE REGARDING THIS DOCUMENT
Dr Jonathan Broomberg
Praxis Capital
Level 5 Sunnyside Ridge
4 Sunnyside Drive
Parktown 2193
South Africa
Tel: 2711 484 22 55
Fax: 2711 484 22 23
e-mail :<jbroom@praxiscapital.co.za> at inet
and
Professor Anne Mills
London School of Hygiene and Tropical Medicine
Health Policy Unit
Keppel Street
GB-London WC1E7HT
Tel: 44 1 71 - 636 86 36/927 24 31
Fax: 44 1 71 -637 53 91/436 53 89
e-mail: a.mills@lshtm.ac.uk
05572
z
p/ LIBRARY
\
f
AND
A
OOCUM£NTAT10M
\
J
un<t
This document is not issued to the general public, and all rights are reserved by the World Health Organization
(WHO). The document may not be reviewed, abstracted, quoted, reproduced or translated, in part or in whole, without the prior
written permission of WHO. No part of this document may be stored in a retrieval system or transmitted in any form or by any
means - electronic, mechanical or other - without the prior written permission of WHO.
The views expressed in documents by named authors are solely the responsibility of authors.
Printed in 1998 by WHO
Printed in Switzerland
CONTENTS
Acknowledgements
1.
Introduction................................................................................................................................
1
2.
Terminology................................................................................................................................
3
3.
The theoretical rationale for contracting out and the dangers...............................
3
4.
Typology of the elements of contractual arrangements..........................................
4.1 Rules of the game.......................................................................................................
4.2 Contract design...........................................................................................................
4.3 Contract implementation...........................................................................................
5
6
7
11
5.
Methodological approaches for evaluating contracting.............................................
12
6.
Lessons of experience from developed countries...........................................................
6.1 Hospitals: clinical and support services...................................................................
6.2 Primary health care.......................................................................................................
12
13
15
7.
Lessons from the experience of contracting in developing countries......................
7.1 The extent of contracting................................................................................................
7.2 The rationale for contracting.......................................................................................
7.3 Contract design.............................................................................................................
7.4 Contract implementation..............................................................................................
7.5 The results of contracting...........................................................................................
7.6 Conditions conducive to successful contracting.....................................................
18
19
20
21
22
23
26
8.
Conclusions.....................................................................................................................................
28
References
31
ACKNOWLEDGEMENTS
This paper was prepared for a WHO Technical Meeting ‘Towards new partnerships for health
development: The contractual approach as a policy tool’ 4-6 February 1998, Geneva. The work
on which it is based was funded by research project and programme grants from the UK
Department for International Development to the Health Economics and Financing Programme,
London School of Hygiene and Tropical Medicine. Shorter versions of some of the material
presented here have been published in Mills (1997) and Mills (1998).
1.
INTRODUCTION
Recent health policy debates in both developed and developing countries have been strongly
influenced by a trend towards ‘marketisation’, involving the selective introduction of a range of
market mechanisms within the public health system (Hurst 1991, OECD 1992, World
Bank 1993, McPake and Ngalande Banda 1994, Mills 1995, Saltman 1995, Saltman and
Von Otter 1995a, Mills 1997). These developments are attributable to two main sets of
influences. The first is the emergence of new trends in public sector management, which identify
private sector mechanisms as a solution to many of the problems currently experienced by the
public sector in many parts of the world (Walsh 1995, Moore 1996). A second key driver of
these reforms has been the accumulating evidence of the failure of health care systems
throughout the world to meet key objectives of efficiency, equity and responsiveness to users
(Birdsall and James 1992, World Bank 1993, Mills 1995, Bennett, Russell and Mills 1996).
Further factors have been growing demands for the extension of consumer choice and influence,
and increasing tension between limited resources and increasing demands on health care systems
(Robinson and Le Grand 1995).
Marketisation’ proposals and reform programmes have varied widely in terms of their scope,
the components of the health system they address, and the specific mechanisms they rely upon.
At their most ambitious, reforms have aimed to create full ‘quasi-markets’ in the health sector,
with competition in either or both the funding and supply of health care. Less ambitious reforms
have aimed at introducing more limited market elements, such as contractual relations between
purchasers and providers.
Reforms of this kind have been debated, and in some cases implemented, in numerous
developed countries including the majority of OECD States (OECD 1992, Saltman 1995,
Saltman and Figueras 1997). Some of the better known instances include those in the United
States (Ellwood, Enthoven and Etheredge 1992, Iglehart 1993), the UK (Great Britain
Department of Health 1989, OECD 1992, Le Grand 1994), the Netherlands (van de Ven 1989),
Sweden and Finland (von Otter and Saltman 1991, Saltman and Von Otter 1992, Annel 1995,
Diderichsen 1995) and New Zealand (Borren 1993, Ashton 1995).
Very few middle or low income developing countries have to date implemented large scale
structural reform along these lines, one of the few exceptions being Colombia (Yepes and
Sanchez forthcoming). Flowever, there are several reasons to believe that such an approach to
policy reform will become increasingly relevant in the foreseeable future. The health systems
of several middle income developing countries currently feature limited market elements, such
as selective contracting (public sector purchasing of specific services from either public or
private providers). This occurs, for example, in some social insurance systems in Latin America
and Asia (Kim 1987, Briscoe 1989, De Geyndt 1990, Griffin 1990, McGreevey 1990), and in
several other countries in which public sector provision dominates, but where some services are
contracted out to private sector providers (Bennett 1991, Bennett and Mills 1993, Chandiwana
and Chiutsu 1993, Cruz and Zurita 1993, Price, Masobe and Booysen 1993, McPake and
Ngalande Banda 1994).
1
r
In the Latin American context, the Colombian rcfomis are based on the introduction of a split
between purchasers and providers with competition amongst both insurers and providers, while
Mexico is considering similar reforms (Gonzales-Scdano 1995, Mills, Hongoro and
Broomberg 1997). Chile is also considering proposals to increase competition at the primary
health care (PHC) level, by allowing private providers to participate in publicly funded provision
(Jiminez 1993, Jimenez de la Jara and Bossert 1995). Similarly, most of the countries of the
former Soviet bloc and Eastern Europe, including the Russian Federation, are experimenting to
a lesser or greater extent with reforms of this kind (Robinson and Le Grand 1995, Sheiman 1995,
Saltman and Figueras 1997).
In Africa, Zambia has begun to implement proposals which envisage extensive
decentralisation of management authority and a purchasing role for district health authorities
(Bennett, Russell and Mills 1996), while South Africa is considering the introduction of provider
competition in the delivery of publicly funded PHC services (Department of Health, Republic
of South Africa 1996), and already has extensive experience with selective contracting (see
below). In a number of other countries (e.g. Ghana, Malawi, Nepal, Rwanda, South Africa,
Swaziland, Tanzania, Zimbabwe) the public sector contracts for health service delivery, either
explicitly or implicitly, with private (usually not-for-profit) providers such as churches or other
NGOs, and meets all or some of their costs through subsidies and grants (Green 1987, Gilson et
al. 1994, Hospital Strategy Project Consortium 1996, Gilson et al 1997).
Numerous other developing countries also contract out the delivery of a range of clinical and
non-clinical support services (McPake and Ngalande Banda 1994, Mills 1995, Bennett, Russell
and Mills 1996, Mills, Hongoro and Broomberg 1997). Finally, there is emerging evidence of
a significant shift towards policies of this kind among influential donors and agencies in
international health, which is likely to add impetus to such reform initiatives (World Bank 1993,
Cassels 1995, Kutzin 1995, Mills 1995).
Despite the extensive debate and descriptive literature on these reforms, there remains very
limited information, from either developed or developing countries, on the impact of
marketisation reforms on the efficiency of providers and of the health system more generally, on
equity and other social objectives, or on the costs of the reforms. This literature review focuses
on these questions in the context of developing countries, with a specific emphasis on contracting
as the key manifestation of the general trend towards marketisation. The review identifies from
the literature:
(1)
(2)
(3)
(4)
(5)
the definitions of contracting in use;
the theoretical rationale for governments to contract out services;
issues relating to the design and implementation of contracts;
results of contracting out;
conditions conducive to successful contracting.
It focuses on the contracting of health services for the general population, referring to
experience of contracting services for insured populations only where this is of direct relevance.
2
2.
TERMINOLOGY
A variety of forms of contractual relationships have been identified (Walsh 1995), including:
•
Competitive tendering (also known as market testing), where internal staff can bid for
contracts in competition with private contractors;
•
Contracting out, where only private bids are allowed and where contracts may be agreed
without a competitive process (termed sole sourcing in the US);
•
Internal contracting, where only internal bids are allowed;
•
Performance contracts, which are explicit agreements between the government (at various
levels) and public-sector managers (World Bank 1995);
»
Internal market, where there is complete separation between the roles of purchaser and
provider within the public sector, and they are linked through trading agreements or
contracts.
These distinctions are based essentially on whether or not the contracting process is
competitive, and whether the process allows both public sector and private sector bids, or
restricts bidding to the public sector. Because of the attention attracted by the UK NHS reforms,
where contracts are largely internal to the public sector, discussions of the merits of contracts
tend to gloss over whether these are contracts between public agencies or between a public
purchaser and private provider. Yet both the case for and nature of contracts are likely to depend
fundamentally on the nature of the provider.
3.
THE THEORETICAL RATIONALE FOR CONTRACTING OUT AND THE
DANGERS
The general theoretical rationale for contracting out relates to theories of why governments
fail in their provision of services. The last decade has witnessed a dramatic re-evaluation of the
structure and functions of government in relation to the delivery of public services (Jackson and
Price 1994, Mills 1995, Walsh 1995, Moore 1996). A central tenet of the new thinking, termed
the ‘new public management’ (Hood 1991, Moore 1996), is that the traditional organisational
form of the public sector, hierarchical bureaucracy, is inherently inefficient and that the
introduction of various market mechanisms will substantially enhance the efficiency of public
service delivery.
Two main schools of thought underpin this analysis. The first, property rights theory,
contends that the main source of inefficiency in the public sector is the weakening of property
rights, so that decision-makers face few incentives to allocate resources efficiently. This is
contrasted with the incentives facing entrepreneurs or shareholders in the private sector (Mills
1997). The second critique of the public sector, argued by ‘public choice’ theorists, is that the
3
politicians and bureaucrats who control public bureaucracies cannot be assumed to be acting in
the public interest, since they are more likely to serve their own interests, or those of powerful
interest groups (Walsh 1995).
In response to these analyses, the ‘new public management' envisages the use of market
mechanisms to generate appropriate price and demand signals, and to weaken the influence of
politicians and professionals over public service deliver}', thus ensuring that these services are
more responsive to market signals and to customers (Jackson and Price 1994, Walsh 1995,
Moore 1996). It is also argued that private organisations can bring the advantages of functional
specialisation, as well as speed and flexibility in adjusting to changing factor prices, technology
and demand conditions (McCombs and Christianson 1987, Mills 1997). A central theme of this
thinking is thus the view of the State as responsible for enabling or ensuring service delivery,
rather than for direct delivery of sendees itself, except in certain identifiable circumstances
(Vining and Weimer 1990, Moore 1996).
Since health care constitutes a major component of public sendees in most countries, it is not
surprising that these new trends in public sector management have substantially influenced health
policy debates (Mills 1995, Bennett, Russell and Mills 1996). Analysis of the efficiency
arguments in the context of health care systems reveals a number of distinct, though linked
claims. The first is that the replacement of direct, hierarchical management structures by
contractual relationships between purchasers and providers will promote increased transparency
of prices, quantities and quality in trading, as well as managerial decentralisation, both of which
wdll enhance efficiency. The second is that these reforms will promote increased competition
among providers, and that the increased level of competition will in turn enhance supply side
efficiency.
These rationales rely largely on economic arguments. Contracting has also been justified in
political terms - witness its wide popularity with politicians in a number of countries. One
reason is that public sector management of sendees is widely seen to have failed. This is perhaps
most true in the countries of Eastern Europe and the former Soviet Union, Africa and Latin
America. Contracting, or marketisation more broadly, has been perceived as an attractive
alternative, and one that has been in use for a number of years on a small scale and so need not
involve too radical a change.
Contracting has also been justified in terms attractive to managers. It offers a means of
distancing the provision of services from the political process. It challenges the established
power of organised labour, and provides the means to use labour more flexibly (Saltman and Von
Otter 1995b). With respect to contracting out non-clinical services such as cleaning, it may
enable managers to concentrate on what they see to be the more important parts of the service.
Proponents of contracting also claim, often implicitly, that the overall benefits of contracting
out will outweigh the potentially substantial costs involved in their creation and maintenance.
These costs include transactions costs, the higher costs that may result from the loss of
monopsony purchasing power, and the social costs arising from equity problems. In the case of
transactions costs, theory suggests that there may be substantial transaction costs involved in
creating and maintaining the contracts. The extent of such costs will depend on the numbers of
4
contracts that have to be written, the extent of detail in their specification, and the intensity with
which implementation is monitored, and will therefore vary between and within systems. While
it is clear that contracting may incur high transactions costs, it is important to compare these with
the explicit and hidden costs of directly managed public systems, rather than to view them as
entirely incremental. For example, public agencies may face large costs in monitoring staff and
output quality, and there may also be significant costs involved in bureaucratic administrative
mechanisms and in the effects of political interference (Krashinsky 1986).
A second set of costs may come from the loss of monopsony power resulting from the
fragmentation of the single purchasing agency in the traditional public health sector. This will
not however be a problem where a single agency lets contracts for services.
Some critics of contracting reforms argue that they embody a potential conflict between
efficiency requirements of the market environment and equity goals, and may act to undermine
equity. Examples of this problem include the loss of comprehensiveness of local service
provision, and potential loss of consumer choice. These may arise as inefficient local services
are forced to close, as economies of scale dictate increasing concentration of service provision,
or as hospitals concentrate on more profitable services at the expense of core services (Robinson
1990).
Equity may also be threatened through the practice of provider or purchaser selection of low
risk patients where payment does not adequately compensate for the level of risk of specific
patients or types of patient (Le Grand 1991, Bartlett and Le Grand 1993). Von Otter and Saltman
(1992) also argue that the participation of private sector providers in the market may force public
hospitals to behave more like for-profit hospitals, focusing on profitable services and stripping
out non-profitable but essential services. Further, for-profit hospitals may be better equipped to
survive under market conditions and public hospitals would thus be disadvantaged, further
aggravating equity problems.
Selective contracting may have a substantial impact on the wider health care system. Firstly,
introduction of contracted providers may lead to fragmentation or lack of co-ordination within
the broader public health system. Secondly, contracting could lead to competition between
contractors and public providers for staff resources, particularly where supply is already
constrained, leading to increasing salaries and to public hospitals being drained of key personnel,
or suffering from increased staff turnover (Mills 1995). Thirdly, contracts can lock scarce
resources into a particular allocation, even when changed circumstances dictate a reallocation.
4.
TYPOLOGY OF THE ELEMENTS OF CONTRACTUAL ARRANGEMENTS
The contracting process involves defining the service to be provided and the conditions that
govern the contract, and implementing and monitoring the contract. In addition, there will be
a framework of rules and regulations that govern the contracting process and which are not
specific to each contract. The crucial issue in contract design is ensuring that there is an efficient
outcome.
5
politicians and bureaucrats who control public bureaucracies cannot be assumed to be acting in
the public interest, since they are more likely to serve their own interests, or those of powerful
interest groups (Walsh 1995).
In response to these analyses, the ‘new public management' envisages the use of market
mechanisms to generate appropriate price and demand signals, and to weaken the influence of
politicians and professionals over public service delivery, thus ensuring that these services are
more responsive to market signals and to customers (Jackson and Price 1994, Walsh 1995,
Moore 1996). It is also argued that private organisations can bring the advantages of functional
specialisation, as well as speed and flexibility' in adjusting to changing factor prices, technology
and demand conditions (McCombs and Christianson 1987, Mills 1997). A central theme of this
thinking is thus the view of the State as responsible for enabling or ensuring service delivery,
rather than for direct delivery of services itself, except in certain identifiable circumstances
(Vining and Weimer 1990, Moore 1996).
Since health care constitutes a major component of public services in most countries, it is not
surprising that these new trends in public sector management have substantially influenced health
policy debates (Mills 1995, Bennett, Russell and Mills 1996). Analysis of the efficiency
arguments in the context of health care systems reveals a number of distinct, though linked
claims. The first is that the replacement of direct, hierarchical management structures by
contractual relationships between purchasers and providers will promote increased transparency
of prices, quantities and quality in trading, as well as managerial decentralisation, both of which
will enhance efficiency. The second is that these reforms will promote increased competition
among providers, and that the increased level of competition will in turn enhance supply side
efficiency.
These rationales rely largely on economic arguments. Contracting has also been justified in
political terms - witness its wide popularity with politicians in a number of countries. One
reason is that public sector management of services is widely seen to have failed. This is perhaps
most true in the countries of Eastern Europe and the former Soviet Union, Africa and Latin
America. Contracting, or marketisation more broadly, has been perceived as an attractive
alternative, and one that has been in use for a number of years on a small scale and so need not
involve too radical a change.
Contracting has also been justified in terms attractive to managers. It offers a means of
distancing the provision of services from the political process. It challenges the established
power of organised labour, and provides the means to use labour more flexibly (Saltman and Von
Otter 1995b). With respect to contracting out non-clinical services such as cleaning, it may
enable managers to concentrate on what they see to be the more important parts of the service.
Proponents of contracting also claim, often implicitly, that the overall benefits of contracting
out will outweigh the potentially substantial costs involved in their creation and maintenance.
These costs include transactions costs, the higher costs that may result from the loss of
monopsony purchasing power, and the social costs arising from equity problems. In the case of
transactions costs, theory suggests that there may be substantial transaction costs involved in
creating and maintaining the contracts. The extent of such costs will depend on the numbers of
4
contracts that have to be written, the extent of detail in their specification, and the intensity with
which implementation is monitored, and will therefore vary between and within systems. While
it is clear that contracting may incur high transactions costs, it is important to compare these with
the explicit and hidden costs of directly managed public systems, rather than to view them as
entirely incremental. For example, public agencies may face large costs in monitoring staff and
output quality, and there may also be significant costs involved in bureaucratic administrative
mechanisms and in the effects of political interference (Krashinsky 1986).
A second set of costs may come from the loss of monopsony power resulting from the
fragmentation of the single purchasing agency in the traditional public health sector. This will
not however be a problem where a single agency lets contracts for services.
Some critics of contracting reforms argue that they embody a potential conflict between
efficiency requirements of the market environment and equity goals, and may act to undermine
equity. Examples of this problem include the loss of comprehensiveness of local service
provision, and potential loss of consumer choice. These may arise as inefficient local services
are forced to close, as economies of scale dictate increasing concentration of service provision,
or as hospitals concentrate on more profitable services at the expense of core services (Robinson
1990).
Equity may also be threatened through the practice of provider or purchaser selection of low
risk patients where payment does not adequately compensate for the level of risk of specific
patients or types of patient (Le Grand 1991, Bartlett and Le Grand 1993). Von Otter and Saltman
(1992) also argue that the participation of private sector providers in the market may force public
hospitals to behave more like for-profit hospitals, focusing on profitable services and stripping
out non-profitable but essential services. Further, for-profit hospitals may be better equipped to
survive under market conditions and public hospitals would thus be disadvantaged, further
aggravating equity problems.
Selective contracting may have a substantial impact on the wider health care system. Firstly,
introduction of contracted providers may lead to fragmentation or lack of co-ordination within
the broader public health system. Secondly, contracting could lead to competition between
contractors and public providers for staff resources, particularly where supply is already
constrained, leading to increasing salaries and to public hospitals being drained of key personnel,
or suffering from increased staff turnover (Mills 1995). Thirdly, contracts can lock scarce
resources into a particular allocation, even when changed circumstances dictate a reallocation.
4. TYPOLOGY OF THE ELEMENTS OF CONTRACTUAL ARRANGEMENTS
The contracting process involves defining the service to be provided and the conditions that
govern the contract, and implementing and monitoring the contract. In addition, there will be
a framework of rules and regulations that govern the contracting process and which are not
specific to each contract. The crucial issue in contract design is ensuring that there is an efficient
outcome.
5
r
politicians and bureaucrats who control public bureaucracies cannot be assumed to be acting in
the public interest, since they are more likely to serve their own interests, or those of powerful
interest groups (Walsh 1995).
In response to these analyses, the ‘new public management' envisages the use of market
mechanisms to generate appropriate price and demand signals, and to weaken the influence of
politicians and professionals over public service delivery, thus ensuring that these services are
more responsive to market signals and to customers (Jackson and Price 1994. Walsh 1995,
Moore 1996). It is also argued that private organisations can bring the advantages of functional
specialisation, as well as speed and flexibility in adjusting to changing factor prices, technology
and demand conditions (McCombs and Christianson 1987, Mills 1997). A central theme of this
thinking is thus the view of the State as responsible for enabling or ensuring sen ice delivery,
rather than for direct delivery of services itself, except in certain identifiable circumstances
(Vining and Weimer 1990, Moore 1996).
Since health care constitutes a major component of public services in most countries, it is not
surprising that these new trends in public sector management have substantially influenced health
policy debates (Mills 1995, Bennett, Russell and Mills 1996). Analysis of the efficiency
arguments in the context of health care systems reveals a number of distinct, though linked
claims. The first is that the replacement of direct, hierarchical management structures by
contractual relationships between purchasers and providers will promote increased transparency
of prices, quantities and quality in trading, as well as managerial decentralisation, both of which
will enhance efficiency. The second is that these reforms will promote increased competition
among providers, and that the increased level of competition will in turn enhance supply side
efficiency.
These rationales rely largely on economic arguments. Contracting has also been justified in
political terms - witness its wide popularity with politicians in a number of countries. One
reason is that public sector management of sendees is widely seen to have failed. This is perhaps
most true in the countries of Eastern Europe and the former Soviet Union, Africa and Latin
America. Contracting, or marketisation more broadly, has been perceived as an attractive
alternative, and one that has been in use for a number of years on a small scale and so need not
involve too radical a change.
Contracting has also been justified in terms attractive to managers. It offers a means of
distancing the provision of services from the political process. It challenges the established
power of organised labour, and provides the means to use labour more flexibly (Saltman and Von
Otter 1995b). With respect to contracting out non-clinical services such as cleaning, it may
enable managers to concentrate on what they see to be the more important parts of the service.
Proponents of contracting also claim, often implicitly, that the overall benefits of contracting
out will outweigh the potentially substantial costs involved in their creation and maintenance.
These costs include transactions costs, the higher costs that may result from the loss of
monopsony purchasing power, and the social costs arising from equity problems. In the case of
transactions costs, theory suggests that there may be substantial transaction costs involved in
creating and maintaining the contracts. The extent of such costs will depend on the numbers of
4
contracts that have to be written, the extent of detail in their specification, and the intensity with
which implementation is monitored, and will therefore vary between and within systems. While
it is clear that contracting may incur high transactions costs, it is important to compare these with
the explicit and hidden costs of directly managed public systems, rather than to view them as
entirely incremental. For example, public agencies may face large costs in monitoring staff and
output quality, and there may also be significant costs involved in bureaucratic administrative
mechanisms and in the effects of political interference (Krashinsky 1986).
A second set of costs may come from the loss of monopsony power resulting from the
fragmentation of the single purchasing agency in the traditional public health sector. This will
not however be a problem where a single agency lets contracts for services.
Some critics of contracting reforms argue that they embody a potential conflict between
efficiency requirements of the market environment and equity goals, and may act to undermine
equity. Examples of this problem include the loss of comprehensiveness of local service
provision, and potential loss of consumer choice. These may arise as inefficient local services
are forced to close, as economies of scale dictate increasing concentration of service provision,
or as hospitals concentrate on more profitable services at the expense of core services (Robinson
1990).
Equity may also be threatened through the practice of provider or purchaser selection of low
risk patients where payment does not adequately compensate for the level of risk of specific
patients or types of patient (Le Grand 1991, Bartlett and Le Grand 1993). Von Otter and Saltman
(1992) also argue that the participation of private sector providers in the market may force public
hospitals to behave more like for-profit hospitals, focusing on profitable services and stripping
out non-profitable but essential services. Further, for-profit hospitals may be better equipped to
survive under market conditions and public hospitals would thus be disadvantaged, further
aggravating equity problems.
Selective contracting may have a substantial impact on the wider health care system. Firstly,
introduction of contracted providers may lead to fragmentation or lack of co-ordination within
the broader public health system. Secondly, contracting could lead to competition between
contractors and public providers for staff resources, particularly where supply is already
constrained, leading to increasing salaries and to public hospitals being drained of key personnel,
or suffering from increased staff turnover (Mills 1995). Thirdly, contracts can lock scarce
resources into a particular allocation, even when changed circumstances dictate a reallocation.
4. TYPOLOGY OF THE ELEMENTS OF CONTRACTUAL ARRANGEMENTS
The contracting process involves defining the service to be provided and the conditions that
govern the contract, and implementing and monitoring the contract. In addition, there will be
a framework of rules and regulations that govern the contracting process and which are not
specific to each contract. The crucial issue in contract design is ensuring that there is an efficient
outcome.
5
The theory of contracting is part of broader economic theory c°ncerne^
relationship (MacDonald 1984, Guesneri 1989, Lazear 1989, Ledyard 1989, St glitz 1989,
Ryan 1992). The agency relationship arises out of the existence of asymmetry of nformat.on
between individuals involved in a transaction. The agency relationship is characterised by a
principal (ill informed individual) and agent (informed indivi ua ) o o w om see to
maximise their utility. Because of diversity of interest and asymmetry o in orma ion t e
principal has to devise a contract or methods of remuneration to ensure t e agent oes not c eat.
Thus the rules in the contract are crucial. The contract creates a distri ution o ris and
responsibility between client and contractor (Walsh 1995). Different distributions will have
different implications for the price of the contract. For example, if all the risk that workload may
vary is placed on the contractor, it will require a higher price than if the risk is more equally
shared. In the review below of elements of contractual arrangements, attention is focused on
what is known, or hypothesised, about their implications for efficient outcomes.
4.1.
Rules of the game
Regulatory environment
All governments constrain the operation of the private sector in various ways, as well as
having rules and regulations that govern all government contracts. For example, minimum
wage legislation may affect the wages paid by the private sector, and there may be constraints
on market entry such as controls on the establishment of private medical facilities. Firms
bidding for government contracts may be required to show evidence of their financial state
of health, and to provide a cash or bank guarantee. The regulatory environment will
influence the availability of private providers and firms, their profitability, and their
willingness to bid for government contracts.
Eligible bidders
Tendering can be selective or open, or a single contractor can be approached to negotiate a
contract. McCombs and Christianson (1987) argued that open contracting is likely to have
the advantage of encouraging new entrants, increasing competition and reducing prices.
However administrative and monitoring costs are likely to be higher because winners may
be less experienced, and there is a risk that contractors may default on their responsibilities.
Moreover, quality considerations need to be introduced at some stage - if not in the bidding
process, then in the selection of bids. However, quality is difficult to assess, especially for
health care provision, and experience may be used as a proxy. With selective tendering,
already experienced firms are pre-selected to bid. Prices are likely to be higher than in a
more open competition, but costs of monitoring the contract and of contract failure lower.
Contracting without competition may be used when there is only one possible contractor, or
if the advantages of a close relationship with one contractor are perceived to outweigh the
disadvantages of lack of competition. In the health and social services field there is some
debate over the merits of competition versus a strategy that recognises the mutual benefits
each side gains from the contract (Le Grand and Bartlett 1993). This is particularly the case
where the contractor is a not-for-profit service provider (Gilson et al 1997)
6
°st governments are likely to have established rules on the circumstances under which
ten enng nee not be competitive, and on the circumstances in which invitations to bid can
e mute to a restrictive list. Performance contracts within the public sector would not
usua y e awarded on a competitive basis, though there may be an implicit threat of
sanctions (e.g. replacement of a management board) if performance is poor.
4.2.
Contract design
Services contracted
A range of different services can be contracted out (Table 1). In addition certain functions,
as opposed to services, may be contracted including purchasing itself (Hillman and
Christianson 1984).
Table 1. Services that may be contracted
Category
Clinical services
Examples
Hospital facility
Primary care facility
Specific specialty or primary care service
Specific diagnostic procedure
Specific surgical procedure
Public health activity
Laboratory tests
Non-clinical services Pharmacy
Catering
Laundry
Cleaning
Maintenance (equipment, buildings)
Security
Personnel recruitment and employment
Functions
Management
Printing/photocopying
Building design and construction
Computing
Purchasing
Finance
Legal services
7
__________________
Not all the elements of a service may be contracted. For example, the management of a
hospital could be contracted out, but otherwise the facility and staff left in public
ownership/employment. Alternatively, the contractor may supply the ancillary workers but
not the professional staff. The greater the extent of contracting, the more complex
organisational relationships may become, with a move away from a hierarchically-organised
institution to an inter-organisational network (Walsh 1995).
Contract specification
Walsh (1995) proposed two basic approaches to specifying the work to be done in a contract.
The first states the outcome and leaves it up to the provider to determine how the work is
done. This is feasible for activities such as cleaning, but is much less easy for contracts for
clinical care. The second states the methods to be used, for example a test to be done, or
even just the workload, for example number of procedures. Because of the difficulty of
specifying outcome, Walsh suggests that contracts will generally be a mixture of method and
performance.
A method-based contract involves less risk for the contractor than a performance-based
contract. Hence a contractor will normally add a risk premium to a performance-based
contract to cover work that may be required. The client will thus pay a higher price, and
despite the specification in performance terms, may still have difficulties ascertaining quality.
The level of detail in contracts will vary greatly depending on the nature of the service and
the ease of specification. For a service like catering, it may be fairly easy to specify a diet
in terms of quantity and quality. In clinical care it may be very difficult to specify precise
services, and these may be settled by negotiation after the contract is agreed.
Indeed, recent research on contracts in the UK has highlighted that contracts are often
broadly focused, informally worded, deliberately incomplete, and reliant on informal
mechanisms for dealing with disputes (Flynn and Williams 1997, Spurgeon et al. 1997).
They conform more to what are termed ‘relational contracts’ (MacNeil 1978) where the
contract is less important than the relationship between the parties over time.
Bidding price
There are many different approaches to specifying the bid price, depending on the service to
be contracted and the degree of uncertainty on the workload. The mode of payment for
contracts is generally regarded as crucial in providing powerful incentives which influence
provider behaviour (Barnum, Kutzin and Saxenian 1995). In the case of hospital clinical
contracts in the UK NHS, in the early years of the internal market, contracts were most
commonly either block or cost and volume (Robinson and Le Grand 1995). In block
contracts, access to a defined range of services is provided in return for an annual fee. Where
the defined range of services encompasses the whole hospital, this is equivalent to a global
budget. In the context of the NFIS, block contracts had the advantages that the information
requirements were relatively small, and they perpetuated existing service arrangements
(which the government required in the first year of the internal market). Cost and volume
contracts specify the provision of a given number of treatments or cases at an agreed price.
If the actual number exceeds this, they are paid on a cost-per-case basis.
8
Cost per case contracts were initially less common in the NHS, though their frequency has
been increasing. Other pricing systems found in contracts include cost per person per time
period (i.e. capitation) and cost per unit of service. Capitation payment is particularly
common for primary care services, though is increasingly being introduced for hospital care.
Capitation has the advantage from the point of view of the client of placing a cap on total
contract cost, unlike cost per case or per unit of service which may represent an open-ended
commitment. However, since a crude capitation (unadjusted for likely need for health care)
shifts the financial risk entirely to contractors, they will require a risk premium to take on the
contract, and are likely to seek to 'cream-skim', i.e. select the cheaper cases to treat, and to
minimise care given. Monitoring systems seek to control this. An alternative approach, now
much investigated as capitation systems are extended to hospital care where risk of loss is
much greater, is to adjust the capitation payment for risk. Such adjustments are used in some
managed care arrangements in the US, for example for Medicaid patients (US GAO 1993a),
and are necessary in the reformed Dutch health care system (Van de Ven et al 1994) and in
the health sector reforms in Colombia (Gonzalez-Sedano 1995).
Capitation payment may be accompanied by other payments: in the case of primary care,
fee-for-service for particular, desired procedures, and standard allowances as contributions
to costs of such components as premises, training, and support staff.
Fee-for-service contracts are generally considered undesirable because of the incentives they
provide to increase the volume of care (Barnum, Kutzin and Saxenian 1995). For historical
reasons, and because this mode of payment is usually favoured by the medical profession and
the hospital industry, fee-for-service contracts still persist. Their tendency to produce cost
inflation may be curbed by adjusting fees downwards if an agreed volume of services is
exceeded (as in some Canadian provinces (Lomas et al. 1989) and Germany).
An alternative form of bidding is where the client sets the price per item of service in advance
and contractors put in bids to deliver this at the maximum standard they think affordable. In
addition, they may be required to offer a rent for the use of the facilities of the client (e.g.
kitchens, consulting rooms). This approach may be attractive to budget-limited public
organisations, since it guarantees that the cost of the contract will not exceed their budget.
However, it runs the risk of discouraging bidders or attracting low quality providers if the
price is set too low, and in the absence of good information on actual costs in either public
or private sectors, the price may be quite arbitrary.
Saltman (1995) distinguishes between 'hard' and 'soft' contracts. In a hard contract the price
is fixed in advance, and the relationship is more adversarial with litigation if something goes
wrong. A soft contract may be agreed particularly between units which share common
ownership; price and volume may be discussed initially but not agreed until the end of the
contract period when terms that are mutually advantageous can be agreed. Such contracts
resemble MacNeil’s relational contracts (MacNeil 1978).
Single or multiple winners
The American contracting literature contains a debate on the relative merits of having a
single winner versus multiple winners. The prospect of multiple winners is likely to increase
9
contracts maintains a number of
the attractiveness of bidding, and awarding a numbe
of bidding)> provides
firms in the market (and thus potential competit onin fu
backup if one winner defaults, helps to ensure availability
cinep other
anticipated, and makes contract termination a feasible threat for poor’ P
service providers are available (McCombs and Christianson 1987). However, if the number
of potential bidders is relatively few, the incentive to bid low prices may b
k and
economies of scale may be lost by dividing up the contract. Multip e winners are more i e y
in contracts for primary care than for hospitals, where economies o sea e may imi le
number of suppliers.
Contract price
Winning bidders may be reimbursed at the rate of their bid. However, even where there is
one winner only, and especially if there are more, bidders may not reveal their true costs in
their bid. They may bid high and if all bidders do this, over time the distribution of bids will
be inflated. Various alternative approaches can be used to counter this. For example the
'competitive rule' involves reimbursing all winners at the level of the highest winning bid,
the lowest excluded bid, or somewhere between (McCombs and Christianson 1987). Since
most winners will receive more than they bid, there is little incentive to submit bids in excess
of real costs plus a normal rate of return: it does not increase reimbursement but does
increase the risk of losing.
Contract duration
Contracts can be for one or more years. Short-term contracts, such as for one or two years,
increase competitive pressures and make it easier to change unsatisfactory contractors.
However, they have their costs. Contractors may be reluctant to invest in equipment or
training if they think they may lose the contract in the near future. Clients have higher
monitoring costs if contractors are less committed, and also bear the costs of frequent re
tendering. It has been suggested in the context of the UK that longer term contracts would
be desirable, since they would encourage collaboration and effective sharing of information
and avoid excessive transactions costs (Robinson and Le Grand 1995). Indeed, the new
government s reforms include the requirement for longer term, more cooperative contracts
(Mays forthcoming). Trust can develop in long-term relationships: trust is efficient, since
it reduces the need for detailed and expensive monitoring of performance (Walsh 1995).
Sanctions for non-performance
Walsh (1995) makes a distinction between punishment-based and co-operative contractual
approaches. The former assumes that client and contractor have different interests, and each
seeks to exploit the other. Sanctions are required to discourage the contractor from failing
to deliver. A co-operative contract assumes that the two parties have interests in common,
and that events outside the contractor's control are just as likely to affect contract
performance. Hence the best solution is for client and contractor to work together to resolve
problems. Such contracts are more likely for complex services such as medical and social
care because of the difficulties of defining and checking ™ ™
r
a ° S0C a
punishment-based approaches are more common for simpler services.
10
Price-changing rules
Longer contracts need to include provision for changing prices. Allowing the whole of any
wage or price inflation to feed its way through to the contract price provides the contractor
with no incentive to economise or adapt to new relative prices. However, excessive controls
on the contract price will encourage the contractor to cut corners and under-perform
(Lalta 1993).
4.3.
Contract implementation
Requesting bids
A variety of approaches can be used to identify potential contractors. At one extreme, a
direct approach can be made to a supplier if it is believed that there is no viable competition,
and thus no point in going through the expensive process of competitive bidding.
Alternatively, where specific expertise is sought and is in fairly limited supply, a short-list
of providers can be directly approached and invited to bid. At the other extreme, invitations
to bid can be widely advertised in the press.
Financing the contract
Normally contracts are financed from government or public agency budgets. However, when
income results from the performance of the contract, income or profit may be shared between
contractor and client, to encourage efficiency and maximum use. Such provisions are
common in leisure service contracts, for example (Walsh 1995).
Location of responsibilityfor contract management
Management of the contracting process has two components: specification and approval of
the contract, and monitoring of the performance of the contractor. These may be done by the
same or different bodies. In some countries, particularly the poorer ones, the contracting
process may be highly centralised, with approval required from a central tender board. It
may even itself be contracted out (Hillman and Christianson 1984). In other countries both
processes may be handled at the local health authority or health facility level. Management
responsibility may also depend on the service contracted: national tenders may be agreed for
some services (e.g. drug purchase), and others left to local tenders (e.g. building
maintenance).
Various approaches are taken to try to enforce contracts. Experience indicates that
responsibility for monitoring performance and enforcing standards needs to be clearly
defined, and sanctions such as fines may be employed especially in manual services. In
professional services, clients can try to ensure that contractors hold similar values to their
own - for example through taking account of a contractor's reputation in selecting a winner.
In addition, the existence of a quality assurance system in the contractor may reduce the
monitoring that clients need to do. Such systems have developed rapidly in recent years, and
can be interpreted as one reflection of the commitment of the contractor to the market.
Well-informed consumers can be a valuable monitoring tool, especially where contracts are
awarded competitively. Governments can encourage consumer influence by collecting and
11
publishing information on benefits, consumer rights, provider performance, and consumer
satisfaction (Van de Ven and Van Vliet 1992).
5.
METHODOLOGICAL APPROACHES FOR EVALUATING CONTRACTING
Methodological issues concerning the evaluation of contracting have been little discussed in
the literature. From the point of view of the public enterprise considering contracting, the
relevant question is the difference between the costs of providing the service itself and the costs
of contracting out, which include the costs of the contract itself plus the transactions costs of
managing the contracting process. The question is thus how costs will change at the margin.
These costs are difficult to ascertain in practice. They imply a before and after comparison,
which is always prone to be affected by environmental changes (Appleby et al. 1993) or by
reorganisation of the pattern of service management or changes in service standards which are
introduced at the same time as the contract. With and without comparisons - for example
comparing public health centres with those run by contracted private providers - avoid the
confounding influence of environmental change over time but have a different set of problems
to do with the comparability' of the two sets of health centres. Both methodological approaches
encounter the problems of public accounting systems which rarely identify clearly the costs of
particular services let alone the transactions costs of contracting, and make it difficult to identify
costs in a comparable way (Walsh 1995). Commercial confidentiality may also make both public
and private sectors unwilling to participate in evaluation. Finally, it is important that evaluation
takes a long time-horizon, since initial bids on the introduction of contracting out may be low,
and may rise over time as firms gain a stronger market position.
The problems of assessing the success ofNHS reforms led (Bartlett and Le Grand 1993) to
propose an indirect approach to evaluation (in this case of the creation of hospital trusts). They
specified criteria for evaluation - productive efficiency, extension of consumer choice,
responsiveness to users' wants and needs, and equity (treatment related to need). They then
argued that to meet these criteria, the following conditions were necessary based on microeconomic theory: a market structure competitive on both the purchaser and provider side; both
purchasers and providers should have good information about the costs and quality of the service
being provided; transactions costs must be low; both purchasers and providers must be motivated
to respond to market signals; and there should be no opportunity for purchaser or provider to
‘cream skim’.
6.
LESSONS OF EXPERIENCE FROM DEVELOPED COUNTRIES
Most evidence on the success of contracting in relation to health services comes from the US
and UK and relates to contracting of hospital services (and most UK evidence relates also to
competitive tendering). Evidence is available on financial savings and the source of savings,
transactions costs, and quality changes. Evidence on contracting of primary care services is
scanty, but there is some recent literature from the UK, US and Scandinavia.
12
6.1.
Hospitals: clinical and support services
Savings
In the UK, large savings have been claimed as a result of competitive tendering of support
services such as cleaning and catering in the NHS. In 1986, a National Audit Office study
reported savings which amounted to about 20% of the costs before competition
(Walsh 1995). A study by (Domberger et al. 1987), of domestic costs in around
1500 hospitals, found that hospitals which had contracted out services reduced costs by 34%,
and those where the in-house provider won the contract reduced costs by 22%. Milne and
McGee (1992), on the basis of a study of domestic and catering contracting in a regional
health authority concluded that 'when looked at in terms of the costs saved, particularly for
domestic services but also for catering, it must be judged some success. This conclusion
stands, even when account is taken of the financial costs of setting up competitive tendering
and the attendant cost of redundancies'. Gleeson (1996) estimates cost savings of
approximately 1 billion pounds per annum as a result of the contracting out of NHS services.
This evidence is strengthened by proof of significant cost savings in contracting out local
authority support services, particularly refuse collection, building cleaning, and grounds
maintenance (Walsh 1995). On the basis of these and similar studies from other countries,
Walsh concludes that there are savings to be made on direct service costs, and that the
greatest savings are likely to be found in the simpler, more repetitive services where unskilled
labour can be used. However, there is some limited evidence that some of the savings from
contracting can be eroded over time. In both the US and UK, there is evidence of under
bidding, whether accidental or intentional, in the first round of competitive tendering, and
(Szymanski and Wilkins 1992) found in their study of refuse collection that the cost savings
associated with contracting fell significantly four years after the initial contract, coinciding
with contract re-negotiation. Since the structure of the market may change significantly over
time, with either increased monopoly or the entry of new firms, it is difficult to judge the
long-term impact of contracting on costs.
The evidence on costs of contracted hospital clinical services is even more complicated.
Evidence from the reformed UK NHS has been very slow to become available, partly
because of the previous government’s antipathy to evaluation. While the first ‘wave’ of
trusts did meet their financial objectives, (Bartlett and Le Grand 1993) show this was not
surprising since the market was heavily managed and they were a self-selected group of
hospitals that were significantly more efficient than their non-trust equivalents. Limited
evidence from health sector contracting in the US shows the emergence of long-term
relationships, the consequent decline of competitive bidding, and a decreasing importance
of price after the initial contract round (Propper 1993, Broomberg 1994). Competition took
place for new contracts, not the re-negotiation of old ones. More recent evidence from the
UK corroborates these observations: (Flynn and Williams, and Spurgeon 1997) comment
on the extent to which contracts between Trust hospitals and District Health Authorities
shifted over time towards relational contracts, involving long term, trust-based relationships
rather than competitively awarded ‘spot contracts’. A recent very comprehensive review of
all the evidence available - which is likely to be the last w'ord on the subject since new
reforms are being introduced - concludes that trust status had relatively weak effects on
13
provider performance (Mays forthcoming). Reasons included the relative absence of
meaningful supply-side competition, government actions which had the effect of limiting
competition, and purchasers who were not very effective at putting trusts under pressure.
Indeed, the same review concludes that the introduction of health authority purchasing was
associated with little change for good or ill.
Sources ofsavings
Sources of savings from contracting out have been ascribed to technical improvements in
efficiency, for example from changes in factor mix, reduction in pay and conditions of
workers, and increased pace of work. Managers in the UK NHS frequently argue that
improvements have been made in work organisation and productivity. However, there is
evidence of significant cuts in wages, especially in labour intensive, unskilled services such
as cleaning (Jackson 1994, Walsh 1995). Lower wages, poorer conditions of service and
reduced labour have imposed costs on the government (loss of taxes and increase in welfare
benefit payments) which have been argued to erode savings (Carnaghan and BracewellMilnes 1993). An important point is that to the extent that savings depend on lower
employment costs, their feasibility depends on the local labour market. Much of the evidence
of savings in the UK relate to a period of relatively high unemployment and government
challenge to the power of trade unions. Similar savings are unlikely to be attainable where
labour markets are tighter and/or union power stronger.
Transactions costs
Some calculations have been made of the transactions costs of introducing and maintaining
contracting. Walsh and Davis (1993) estimated the costs of preparing for competitive
tendering in UK local government at 7.7% of the annual contract value. They also found that
the client costs of managing contracts were about 30% higher than the costs of managing the
direct provision of the service. The average cost of monitoring contracts was 6.4% of the
total value of the contract, with some services up to 10%. The Audit Commission (1993)
found a range of 1.4% for education catering to 12.5% for vehicle maintenance. In the US,
the costs of managing contracts have been estimated to be higher, often up to 10% and even
20% of the value of the contract (Walsh 1995). The costs of monitoring social care contracts
in the UK are said to be high. The issue of transactions costs in the NHS have been
extremely controversial since management costs appear to have greatly increased (Robinson
and Le Grand 1995, p34), though exact figures are disputed. Estimates by (Glennester and
Le Grand 1995) suggest additional administrative expenditure of approximately 400 million
pounds per year in the NHS, with administrative staff having increased from approximately
4600 to 13,000 as a result of the reforms. Similar concerns have been expressed in relation
to the costs of organisational changes in government structures in New Zealand (Boston et
al. 1991).
Quality changes
There is mixed evidence on quality changes as a result of contracting out. In the UK NHS,
there have been a number of instances of contractors withdrawing or failing to meet quality
standards, for example in hospital cleaning. However, contracting has lead to clarity on
standard specification which was previously absent, and a level of monitoring which
previously did not occur. In addition, there is evidence that quality improves over time, as
14
both client and contractor gain experience. Thus the overall judgement seems to be that
contracting leads to improved quality of some non-clinical services (Walsh 1995).
However, it is of interest that private sector bidding for public sector hospital contracts has
been distinctly patchy, even in the UK where it might be thought that private firms should
be readily available. Milne (1987) found that 63% of 32 hospital contracts had no outside
bids, with a particular shortage of bids for catering contracts. In general, profit margins on
contracts for ancillary services do not appear to be very high (5-7% according to Milne and
McGee 1987). However, the threat of an outside bid appears to have led to improvements
even where the contract was won by the inside bid or was uncontested.
Very little evidence is available on the results of contracting for hospital clinical care, hence
questions of whether quality as perceived by users improves or worsens remain unanswered.
Cream skimming does not appear to have occurred, though in some areas there was
differential speed of access to hospital care depending on who purchased the services (GP
or local health authority purchaser) (Mays forthcoming).
6.2.
Primary health care
Contracts for primary care exist in a number of European countries, particularly the UK and
Scandinavia. The introduction of contracts has thus not been an issue, with many contractual
arrangements of long standing. The focus of attention has not been the relative efficiency of
direct provision versus contracting, but rather the desirability of different ways of paying
General Practitioners (GPs) and the primary care team, and different degrees of specification
of performance requirements.
Payment to primary care physician can be on a fee-for-service basis (e.g. Germany) but not
infrequently is capitation-based (e.g. Netherlands, UK). Often the payment system may be
mixed, using fee-for-service for services where expansion is desirable (e.g. preventive care)
and capitation for the rest. The focus of most evaluation has been on the effect of different
payment systems on the volume of care provided.
Effect on services provided and costs
In the UK until 1990, the GP contract was extremely vague: in the terms of (Culyer et
al. 1990) ‘a GP’s got to do what a GP’s got to do!’. Per item fees were paid for certain
services, such as maternity, vaccination and cervical cytology. Hughes and Yule (1992)
comment that the effect of remuneration on GP behaviour had been the subject of little
research but much speculation. The new contract introduced in 1990 sought to specify core
services which all GPs were expected to provide and to relate payment to performance. In
particular, fees were introduced or increased in order to encourage the provision of certain
services, and in some cases payment was conditional on reaching a certain target level. There
have been evaluations of the impact on the provision of, for example, cervical cytology,
childhood immunisation, and skin lesion removal. In relation to the first two, payments are
target-related.
15
In a study of the effect of fees on maternity care and cervical cytology before and after 1990,
(Hughes and Yule 1992) concluded that changes in fees prior to 1990 had little effect on the
supply of services, but that the introduction of target payments for cervical cytology had had
a major impact. Lynch (1994) sought to explain why over 25% of GP practices in the
Greater Glasgow area did not qualify for the high target payments for immunisation in the
last quarter of 1991/2. It was apparent that while financial incentives were important in
encouraging high coverage, they were not sufficient alone to assure the high update rates
desired by the government. Brown et al. (1992) found that the introduction of a financial
incentive for minor surgical procedures was followed by a large rise in skin lesion removal
by GPs: presumably as a result of the changed payment system.
The few studies of budget-holding GPs in the UK suggest that the introduction of flexibility
between drugs expenditure and expenditure on other services (previously drug costs were
paid for separately) has encouraged a lower rate of increase in drug costs (Glennester et al.
1993) because of lower volumes of prescribing and more rapid increase in generic
prescribing (Gosden and Torgerson 1997). There is also evidence of reduced hospital
referrals amongst budget holding GPs (Gosden and Torgerson 1997), although the cost
implications of this have not been measured.
A few studies from the US evaluate the impact of contracting on the costs of delivering PHC.
Begley, Dowd and McCandless (1989) evaluated nine primary care projects set up in Texas
to service low income people without health insurance. Models of service delivery included
public clinics, public hospital outreach, private-for-profit health maintenance organisation,
and private providers contracted by a local agency. Those projects following the public clinic
model (i.e. using either project staff or local providers to provide services at a clinic) had
lower average costs than those following a private provider model (providing services
through the offices of local private providers).
McCombs and Christianson (1987) reviewed the literature on competitive bidding for health
services under public programmes in the US. A number of the schemes reviewed had a
primary care content, though no contract was specifically for primary care. They emphasised
the complicated nature of what appears to be a simple process. They concluded that the
performance of bidding systems is likely to be influenced by complex interrelationships
among the market structure and characteristics of the product or service for which bids are
submitted, the incentives in the bidding system design for providers and consumers, the
political influence of provider associations, and the manner in which the bidding process is
implemented and subsequently administered. They imply that competitive bidding may not
produce cost savings because of the various influences on the contracting process.
The growth of managed care in the US has had major implications for the delivery of primary
care (Kane 1995). The primary care physician is conceived of as a ‘gatekeeper’ and may be
placed financially at risk (e.g. through the use of capitation payment systems, and
withholding of a share of payments till the end of the year when the status of the referral fund
is known) in order to encourage cost containment. Schlackman (1993) reported that almost
50% of HMOs pay primary care physicians by capitation, 40% by fee-for-service, and 13%
by salary. 60% use withhold accounts, most of whom withhold 11-20%, and 25% place the
16
individual physician (as opposed to the pool of physicians) at risk. The strategy of selective
contracting promises greater patient volume to specific providers who agree to provide
services under contract (Kane 1995). The contracts usually specify utilisation controls and
discounted prices. Rules such as practice guidelines are also used to shape ‘physicians’
practice. Kane reviewed the evidence on the outcomes of managed care. It appears that
under salaried and capitated arrangements, physicians will substitute more frequent primary
care for more expensive care. There is no evidence that selective contracting without
financial risk-sharing will increase the efficiency of physician practice patterns. It is also
difficult to introduce because a particular physician will have few patients under any
particular plan. Physician profiling, plus gate-keeping and utilisation review, have been
shown to reduce total costs of care. Kane concludes that the most effective managed care
combinations are those that include limited choice of provider, an organisational culture that
favours conservative practice, and gatekeeping. However limited choice is unpopular with
users. Moreover, because those who enrol in managed care tend to be the healthier, the
purchaser (e.g. Medicaid) may end up paying more for enrollers in HMOs than it would if
they were in fee-for-service plans. In other words, Medicaid could not capture the savings
because premiums were not closely related to health status (US GAO 1993b).
Quality changes
Evidence on the quality impacts of PHC contracts is limited. Alper (1994) expressed concern
at the impact of the rapid growth of managed care on the primary physician who is expected
to act as gate-keeper and, because of the capitation payment, is put financially at risk. The
physician is at the mercy of large carriers offering non-negotiable contracts based on
capitation, and bears some liability for services other than his own. Alper’s fear was that
primary care will become even less attractive for physicians, and that there will be a steady
decrease in the quality of primary care. The introduction of contractual arrangements in
countries where physicians have been paid fee-for-service and have been accustomed to few
controls on their mode of practice is likely to raise very similar concerns. Silverstein and
Kirkman-Liff (1995) present a more positive view of the attitudes of physicians to managed
care, in a study of physician participation in Arizona’s prepaid managed care Medicaid
programme. However, they emphasise that remuneration was relatively generous and that
physician cost containment through managed care approaches must involve adequate reward
for primary care physicians.
Barnum, Kutzin and Saxenian (1995) argue that the need to compete for consumers can
counter-act the quality-reducing incentives in capitation payment. They point to the success
of the capitation-financed health plans managed by Kaiser Permanente, which have been
leading innovators in cost containment while maintaining quality. They also have very high
coverage rates for preventive procedures - another expected benefit of capitation-based
contracts. Schlackman (1993) reports successive quality-based compensation models
introduced by US Elealthcare, an operator of HMOs. Quality of service was initially
measured by performance audit and patient satisfaction surveys, and capitation amounts were
higher for higher performance. This was unpopular with physicians and proved inflexible.
It was changed to a basic age-sex related capitation rate, adjusted by a quality factor
reflecting quality, provision of comprehensive care, and cost-effective patterns of utilisation.
Kane (1995) argues that numerous analyses of the quality of care of managed plans suggest
17
that managed care medical quality is at least as good as unmanaged care, though patient
satisfaction may be lower in terms of perceived quality. Anecdotal evidence in the UK is
said to show that budget-holding GPs have reduced waiting times (Smee 1995).
There is evidence that capitation contracts for PHC which exclude hospital care encourage
a higher rate of referrals (Steams, Wolfe and Kindig 1992). In contrast, capitated contracts
which include hospital care furnish an incentive to provide care at the lowest cost level usually the primary care level. For example, budget-holding GPs in the UK are said to be
developing new services (such as testing in the practice, and chiropody) on a larger scale than
non-fundholders (Smee 1995). In addition, the experience of budget-holding GPs in the UK
appears to have been very positive with respect to the pressure they have been able to bring
to bear through contracts with hospitals to improve hospital efficiency (Glennerster el al.
1993, Smee 1995). There is also evidence that the increased linkages between GPs and
secondary care providers, specialists and hospitals, has led to some improvement in aspects
of quality of care (Gosden and Torgerson 1997).
Equity effects
Capitation contracts provide an incentive not to enroll those at higher risk of needing medical
care if this risk is not adequately allowed for in the capitation payment. Evidence is limited,
but HMOs in the US seem to have healthier enrollees than the rest of the population,
suggesting that they have selected favourable risks to a certain extent (CBO 1994). Capitated
primary care physicians may also avoid enrolling sicker patients (Hillman et al. 1989). The
potential for cream skimming by fund-holding GPs in the UK has been identified
(Glennerster et al. 1994) but seems not to have materialised, probably because of the way in
which reforms were introduced and protection for GPs against high cost patients (Mays
forthcoming).
7.
LESSONS FROM THE EXPERIENCE OF CONTRACTING IN DEVELOPING
COUNTRIES
Evidence on the extent of contracting out in developing country health systems is extremely
limited (Mills 1997) and evaluation of its advantages and disadvantages has been virtually non
existent. The review here draws largely on the results of research on contracting commissioned
by the Health Economics and Financing Programme (HEFP) of the London School of Hygiene
and Tropical Medicine over the last few years in Bombay (Bhatia and Mills 1997), Ghana,
Tanzania, and Zimbabwe (McPake and Hongoro 1995, Gilson et al. 1997), Mexico (Alvarez et
al. 1995)', PNG (Beracochea 1997), South Africa (Broomberg et al. 1997), and Thailand
(Tangcharoensathien et al. 1997), plus some evidence from the very limited other published
literature.
1 To avoid excessive repetition, these studies are not constantly referenced throughout the
following sections; however, they are the sources of information whenever information is
reported for the relevant countries.
18
7.1.
The extent of contracting
In developing countries, experience has thus far been limited to selective contracting for a
range of non-clinical and clinical services, and there is greater experience with selective
contracting for non-clinical than for clinical services, as demonstrated in Table 2. In
addition, the extent of contracting between the public sector and the commercial sector
appears to be relatively limited thus far. South Africa provides one exception to these
patterns, having had a long history of contracting out to both for-profit and not-for-profit
providers of hospital services as well as for a broad range of other clinical service contracts.
In total, clinical contracts held by the various South African government authorities
accounted for 9.4% of total hospital expenditure in 1995 (Hospital Strategy Project
Consortium 1996).
As shown in the table, contracts for provision of hospital services take a variety of forms.
In Papua New Guinea, South Africa, and Zimbabwe, there are explicit contracts, although
in the latter two countries, only single hospitals are covered in each contract. In South
Africa, on the other hand, a large number of contracts, covering both acute and long term
hospital care, are held with a single for-profit provider.
Several countries also implicitly contract with a range of private, usually not-for-profit
providers. As noted above, these are usually with hospitals run by church organisations,
although the South African government authorities have similar arrangements with a
charitable organisation which runs TB hospitals, as well as with numerous autonomous notfor-profit acute care hospitals (Hospital Strategy Project Consortium 1996). As Mills (1995)
pointed out, these contracts are ‘implicit’ in the sense that there is no obvious competition
in the award of the contracts, and governments have not historically specified contract terms
or monitored performance. Tanzania is an exception here, since the government has explicit
contracts in place with the various church organisations which provide district hospital
services (Bennett, Russell and Mills 1996 , Gilson el al. 1997).
There are a few examples of contracting for PHC from middle income countries (Swan and
Zwi 1997). In Taiwan, private physicians are contracted to provide sterilisations and IUD
insertion (Foreit 1992). In Costa Rica, the social security bureau has contracted out
responsibility for providing primary health care for a population of 50,000 to a private entity;
the contract allows the provider to gain financially if it provides services efficiently (Pezza
and Barquero Bolanos 1994).
In the HEFP-commissioned studies, contracting for non-clinical services was more common
than for clinical services, covering cleaning, catering, pharmacy, laundry, maintenance,
printing and security. In Bombay and Mexico such contracting appeared fairly common,
though with differences in which services were commonly contracted out (for example,
catering was rarely contracted out in Mexico). In Thailand and PNG, non-clinical contracting
seemed more restricted in scope and number of hospitals involved, to cleaning, security and
grounds maintenance in Thailand, and security, cleaning, catering, laundry and maintenance
in PNG. In all countries it was likely that contracting was more common in large cities,
especially the capital city.
19
Table 2. Selective contracting arrangements in developing countries
Service contracted for
Countries
Non-clinical services
Laundry
Cleaning
Security
Billing functions
Catering services
Equipment maintenance
Patient transport
Distribution of pharmaceutical supplies
Gardening services
Waste removal services
Bangladesh. India, Indonesia, Malaysia, Mexico, Pakistan, South Africa,
Thailand. Zimbabwe
Jamaica, Mexico, South Africa, Thailand
Lesotho, Malaysia. South Africa. Sri Lanka
South Africa. Zimbabwe
India, Lesotho, Malaysia, Mexico, South Africa
Mexico, Uganda, South Africa, Venezuela. Zimbabwe
South Africa
South Africa
South Africa
South Africa
Clinical services
Acute hospital care (explicit contracts)
Acute hospital care (implicit contracts
with church/NGO providers
Long term hospital care
Ambulatory care
Specific ambulatory procedures
Diagnostic services
Laboratory services
Public Health Services
Hospital management contracts
Blood product supply
Supply of nursing personnel
Papua New Guinea, Philippines,2 South Africa. Tanzania, Thailand,
Zimbabwe
Ghana, Malawi, Nepal, Rwanda. Swaziland, South Africa, Zimbabwe
South Africa (TB and chronic psychiatric are)
Namibia, South Africa
Taiwan
Malaysia (CT, Xray, radiation therapy), Thailand (CT, ESWL, MRI)3
Nigeria, South Africa
India (vector control)
Bolivia, China, South Africa
South Africa
South Africa
Modified from: Bennett, Russel and Mills 1996.
Additional sources: Mills 1995, Kutzin 1995, Hospital Strategy Project Consortium 1996.
7.2.
The rationale for contracting
In the HEFP case studies reviewed, non-clinical contracting was usually justified in terms
of lower costs, less hassle for managers, and obtaining greater flexibility in the use of labour.
2 In the Philippines and Thailand, the contracts are held by the social insurance system.
3 CT: computerised tomographic scanning; ESWL: lithotripsy; MRI: magnetic resonance
imaging.
20
I he justification for clinical contracting was either the unavailability of the service in the
facility or area, or the inability of government authorities to provide a service identified as
necessary. In several contracts reviewed in South Africa, no explicit rationale could be
identified (Hospital Strategy Project Consortium 1996).
Contract design
7.3.
The HEFP case studies revealed considerable variety in service specification, even for the
same service. For example, the following variants occurred in catering contracts in Bombay:
•
•
•
•
the hospital provided the space and facilities, for which the contractor paid rent and
utility charges;
the hospital provided the space only, and the contractor the equipment;
the contractor used his own kitchen outside the hospital;
the contractor supplied only the staff, with the hospital responsible for all else.
Similarly in South Africa, in one contract for district hospital services the contractor built,
equipped, staffed, and managed a hospital. In a second contract for similar services, medical
staff were employed by the government and in a third, medical and nursing staff.
Diagnostic service contracts showed a distinctive pattern, with a private company providing
equipment free or at a nominal rent. In return, a fee per test or case was paid; in the case of
laboratory services in Mexico City, all materials required were also purchased from the
company.
Agreements with mission health care providers were formalised as contracts only in
Tanzania, where a formal document was agreed between the diocese and the MOH.
Elsewhere, the agreements were very vague.
Contracts varied considerably in how tightly they were specified and in what terms. Bombay
catering contracts specified meals of a specific content, a PNG security contract the number
of guards, and Thai cleaning contracts the area to be cleaned. The Thai equipment contracts
specified the type of equipment and the fact that it should be new. Quality specifications
appeared to be rare, and non-existent in contracts for clinical services.
Contract duration for non-clinical services was most frequently one year, although some
contracts lasting up to three years were identified. In some cases, contracts included an option
to renew. The clinical contracts reviewed in South Africa were of much longer duration; in
two cases, contracts were for a 10 year period, renewable for a further IO.4
4 These long durations arose because the contractor had financed the constructions of the
hospitals.
21
05572
X
s''
With regard to sanctions for non performance, in Bombay and Thailand, fines were
commonly specified for non-performance; in PNG termination was the only option. Most
other contracts reviewed failed to specify sanctions.
7.4.
Contract implementation
Non-clinical contracts tended to be consistently agreed through a competitive process
involving calling for bids through newspaper adverts or hospital notice-boards. However,
changing a contractor was rare except for maintenance, printing and security in Mexico. In
PNG. contracts had even been renewed against hospital advice.
The number of bidders for catering contracts was relatively few in Bombay, probably because
the contract tender fixed the price per meal at a relatively low level. Similarly in Thailand
for cleaning, the unit price was fixed and since it was well below private sector rates of
payment, relatively few firms were interested in government contracts. In PNG, however,
many firms were interested in bidding for security contracts, probably because start-up costs
were very' low, and staff could readily be recruited at low wages. Similarly in South Africa,
competition for most non clinical contracts appeared to be intense, with numerous firms
bidding in all cases.
Clinical contracts were less commonly awarded competitively. In Mexico, it appeared to be
up to the state to decide on the provider to be contracted, and similarly in Thailand the
teaching hospital chose the private company to supply the equipment (or was approached by
them). In such circumstances, pricing the contract seemed to be fairly arbitrary. In Mexico
the price was based on the national fee schedule but considerable local leeway was allowed.
The South African district hospital contracts had originally been negotiated, but one renewal
had been done on a competitive basis. All mission/govemment contracts were agreed
without competition.
In South Africa, Thailand, and Zimbabwe user fees were chargeable for some of the services
under contract. In both South Africa and Zimbabwe, user fee collection and exemptions were
handled by a government clerk: indeed, in Zimbabwe the contractor explicitly refused to take
responsibility for deciding who should pay, or ensuring payment. Interestingly in Thailand,
contractors were generally willing to accept the hospital’s decision on who could not pay,
although in a number of contracts they received no fee for such cases. A few contracts
limited the number of exempt or subsidised cases.
Responsibility for management was often not clearly specified. In Bombay, there seemed
to be a well organised system for contracting catering, with full involvement of hospital staff.
In contrast in PNG, hospitals felt powerless to monitor the contractor or obtain
improvements. In one of the South African hospitals, both contractor and government staff
were expected to manage the hospital, resulting in lack of co-ordination and even conflict.
In Thailand, the main complaint was that the contracting process was tightly controlled by
the centre. In the case of non-clinical services, this meant rigid specifications and prices
which hospitals had to adhere to although they were unrealistic. Regulations for contracting
22
clinical services had only recently been considered. In mission/government agreements,
there was minimal monitoring of mission performance.
7.5.
The results of contracting
Despite the growing experience with contracting in developing countries, there is very little
systematic data on the impact of these contracts on unit or system efficiency and equity.
The admittedly limited evidence on non-clinical contracting suggests that contracting was
capable of delivering services at lower cost. For example, data from Bombay clearly
suggested that contractors had lower costs than public providers, and that contracting the
catering service was cheaper than direct provision. The same was probably true in Thailand
and Mexico. However, it is interesting to note the evidence from Bombay that the quality
and quantity of the diet was worse in the contracted service.
In the case of clinical contracts, the study of the contract with a mine hospital in Zimbabwe
indicated that the government was able to obtain services of the same quality at a lower cost
than the equivalent costs in a public hospital. However, the contract was also noted to be
unable to control utilisation and hence total cost, and led to provincial resources being
disproportionately concentrated in one district. This led the authors to query whether the
contract could in fact be regarded as successful at all.
A recent comparison of contracted and public sector TB hospital services in South Africa
found that costs were lower, and quality of care higher, at the contractor compared to the
public hospital, although the authors identified a number of important systemic inefficiencies
resulting from the contract, including lack of co-ordination between the contractor hospital
and other public services (van Zyl et al. 1996).
A larger study of contracted out hospitals in South Africa provided important insights into
the gains achievable from contracting (Broomberg, Masobe and Mills 1997). This study
demonstrated that the contractors were highly successful in delivering services at a cost
below that of the public sector, largely through lower staffing levels and higher productivity.
Some structural aspects of quality were superior to that of directly provided services (e.g.
cleanliness and building maintenance), but others were similar, or inferior. Critically,
however, the contractor was able to capture all of its efficiency gains in profits, leading to a
situation in which the costs of direct provision were similar (and in some cases lower) than
the costs to the government of contracting out.
The series of case studies comparing the performance of‘contracted’ church hospitals with
public sector hospitals in Ghana, Tanzania and Zimbabwe did not identify any systematic
differences in performance or cost, except in the case of Zimbabwe, where the church
hospitals were noted to have lower unit costs. These studies found, though, that the church
hospitals were characterised by more highly motivated managerial staff, and enjoyed
substantially greater autonomy than the public hospitals, suggesting potential for greater
efficiency. One serious problem identified by the study was the fragmentation and lack of
23
co-ordination in district health services resulting from the dual lines of accountability when
church owned hospitals are introduced into the public health system (Gilson et al. 1997).
Analysis of the sources of efficiency or inefficiency in these various contracts highlights
some important trends. Where cost savings were shown, they were in most cases due to
some combination of lower wages, lower staffing levels, higher staff productivity, and tighter
management of supplies (as found in the UK: Domberger, Meadowcroft and
Thompson 1987). For example in Bombay, contract caterers had lower staffing levels in
relation to the number of meals prepared than directly provided catering services. The South
African contract hospitals used notably fewer staff, across most staffing categories, and
emphasised worker motivation through approaches such as performance-related pay. They
benefited from greater control over staff than that available to a public hospital manager, for
example being able to 'hire and fire1 at hospital level - facilitating control over pilferage.
Wages were for some staff cadres actually higher than those for public sector workers. The
evidence from PNG showed that where contractors paid low wages, this could bring
disadvantages of low worker morale and poor performance.
Several studies point to the importance of the payment mechanism in influencing the
performance of the contractor. A case study from Mexico, in which a state-owned company
contracted with private providers, showed that the average cost per beneficiary was 15%
higher than the costs of direct provision, a result that was attributed to the use of a fee-forservice reimbursement mechanism. In contrast, a capitation based PHC contract for workers
covered by the social health insurance system in Mexico demonstrated efficiency gains. In
South Africa, contractors were paid per inpatient day, and had longer lengths of stay than the
public hospitals studied. In Thailand, there is a capitation payment to hospitals which
contract to provide comprehensive care for insured workers: this has led to hospitals
contracting with outreach clinics to provide primary care more cost-effectively
(Tangchaorensathein and Supachutikul forthcoming).
These studies have also identified a number of problems in the design, management and
implementation of contracts which may help to explain the failure in some cases to generate
efficiency gains. Firstly, many of the contracts appear to have arisen in an ad hoc way, with
little explicit justification or evaluation of their likely costs and benefits, and they were often
very vaguely specified (Kutzin 1995, Bennett, Russell and Mills 1996, Hospital Strategy
Project Consortium 1996, Gilson et al. 1997). This suggests that contracting in a developing
country environment may have generally failed to live up to one of its objectives, that of
clarification of organisational objectives and increased transparency of resource allocation
through an explicit trading relationship between purchasers and providers (Bennett, Russell
and Mills 1996).
Many contracts also resulted in a shift of most of the contractual risk to the government, thus
putting no pressure on contractors to be efficient. This was apparent, for example, in
minimal specification of contractor performance (especially in clinical contracts) and/or of
sanctions for poor performance, through use of payment methods in which the purchaser bore
all the risk, or through long contract terms. Specific examples of these problems were
identified in South Africa (Hospital Strategy Project Consortium 1996) and in Thailand
24
(Bennett, Russell and Mills 1996), and are attributable, at least in part, to poor government
negotiating capacity. Similar problems were also noted in the contracts or agreements
between governments and church hospitals in the study by (Gilson et al 1997).
The Thai case-studies of diagnostic equipment and ESWL contracts showed some unusual
features. Some contracts appeared highly advantageous to the public sector, and all depended
on user fees as the source of finance. The motivations of private sector ESWL contractors
were explained by the reimbursement regulations of the civil service medical benefit scheme,
an important source of funds for both public and private health care providers, which would
not reimburse the cost of ESWL treatment in the private sector but would pay for the
procedure if provided under contract to a government hospital. Other contracts advantageous
to the public sector were explained by the concern of private companies to be seen as
benefactors, and to provide their machines and services in teaching hospitals where
specialists were trained and hopefully influenced in favour of particular technologies and
brand-names.
Bennett, Russell and Mills (1996) argue that there is little evidence that contracting in
developing countries has met the objective of encouraging provider competition, although
as noted above, there has been little explicit effort to encourage competition in most cases.
The South African study found very similar results, and in that case, the authors note that
directly negotiated contracts appear to more strongly favour the contractor, again suggesting
that contractors are often stronger negotiators than governments (Hospital Strategy Project
Consortium 1996). There is also the risk, noted by (McPake and Hongoro 1993) that,
without competition, governments can become dependent on powerful monopolistic
contractors.
Experience in some countries suggests that contracting has had mixed success in overcoming
bureaucratic restrictions, despite the fact that this is seen as one of its key objectives.
Evidence from Lesotho, South Africa and Thailand suggests that contracting has assisted
governments or individual hospitals to overcome public service restrictions, but other
experiences in Ghana, South Africa, Tanzania,Thailand and Zimbabwe suggest that the
public procurement process itself can become highly bureaucratic, undermining the potential
efficiency of contracted providers (Bennett, Russell and Mills 1996, Hospital Strategy Project
Consortium 1996, Gilson et al 1997, Tangcharoensathien, Nittayaramphong and
Khungsawatt 1997). Finally, there is fairly extensive evidence of very weak government
capacity to monitor the performance of both for-profit and not-for-profit contractors
(Mills 1995, Bennett, Russell and Mills 1996, Gilson et al 1997, Bennett and Mills 1998).
Since the contracts evaluated were of relatively limited scope, consequences for the health
system as a whole (Mills 1995) were few. The Thai case-study expressed concern about the
access of the poor to high technology services, given that user fees were the means of
financing the contracts. The Zimbabwe contract also gave rise to equity concerns because
it had led to provincial resources being disproportionately concentrated in one district. The
PNG non-clinical contracts protected the contracted services from government cut-backs whether this was seen as good or bad depended on the importance given to the contracted
services relative to those that were cut. In the South African contracts, there was evidence
25
of some fragmentation and lack of co-ordination between the contracted out hospitals and the
surrounding primary health care services, which remained under government control.
Taken together, these observations suggest that, despite the fairly extensive experience with
selective contracting and other marketisation reforms in both developed and developing
countries, systematic evaluation and evidence on their impact remains very limited. The data
which do exist show mixed results; there is fairly broad evidence of short-term gains in
micro-efficiency, including cost savings and quality improvements, although several studies,
particularly in developing countries, have also highlighted short term efficiency losses. It is
also generally regarded as too early to judge the long term effects of these arrangements,
although there are serious concerns as whether their high transactions and other social costs
will reduce or even eliminate overall efficiency gains from these reforms. Finally, there is
evidence that the potential efficiency gains from contracting may be undermined by the
absence of critical environmental conditions, as discussed further in the following section.
7.6.
Conditions conducive to successful contracting
It is possible to draw some tentative conclusions from the limited existing data on key
conditions that may be conducive to successful contracting in a developing country
environment.
Capacity of the contracting agency
The ability of the contracting agency to manage the contracting process is clearly crucial to
its success. In Bombay, catering contracting appeared well organised, with a special central
committee and adequate hospital involvement. In PNG, neither central nor local
arrangements were satisfactory. In South Africa, health department staff had allowed
contracts to persist that were highly favourable to contractors; the same was true of some
medical equipment contracts in Bangkok.
Weak government capacity tended to lead to late payment, of which a number of contractors
complained. This is likely to have made contractors more reluctant to bid for public sector
contracts, thus leading to fewer bidders and higher bids to compensate.
Contract design
It is clear that some services were easier to specify for contracting purposes than others. For
example cleaning and catering services, and provision of medical equipment, are relatively
straightforward to specify. In contrast, clinical care contracts were rarely specific on the
services to be provided, leaving scope for argument on the services contractors should be
providing. It could be suggested that vague specification was optimal (based on the
arguments in favour of relational contracts); however, there was little evidence of the close
and co-operative relationship of relational contracts, but rather a simple neglect of basic
requirements to be clear on what government funds were purchasing.
26
It was notable that contracts tended to be awarded for new services (e.g. cleaning a new
building), thus avoiding the problem of terminating the in-house service. While this may
have been a rational strategy for managers, it restricted the savings potentially available.
Limited budgets in poor countries often meant that poor quality services were provided.
Inadequately funded contracts, as in PNG, are likely to be a source of unhappiness. Thus
contracting out a service may require an increase in expenditure. For example in Jamaica,
contracting-out cleaning and portering services cost 25% more than the hospital's previous
costs, but resulted in services of increased quality and scope (LAC Health and Nutrition
Sustainability 1995). A judgement needs to be made on whether increasing expenditure in
order to obtain higher quality is the best use of scarce funds. A similar cost/quality trade-off
was evident from the Bombay data on catering: in this instance lower cost versus lower
quality.
The data also highlight the importance of careful design of the price of the contract. Clinical
contracts were particularly prone to manipulation by providers, and particularly difficult to
monitor. Pricing systems that contain no incentives to control over-supply are undesirable.
Fixed price tenders seemed to work relatively well in Bombay and less well in Thailand
where the discrepancy between public and private sector prices was too great. In the South
African study, initial prices appeared appropriate, but all contracts had built in price
adjustment clauses, cushioning contractors from all cost increases and weakening their
incentives for efficient production.
Sanctions also needed more attention in contracts. In PNG, termination was the only option
and since this would have left the hospital without a service, managers were reluctant to
request it. The optimal situation is probably as in Thailand where although financial
sanctions were available, they had never needed to be used because the contractor had other
incentives to keep the equipment functioning. The South African contracts provided no
details on penalties for poor performance, aside from termination of the contract.
Contract implementation
Competitive contracting was the norm for non-clinical contracts and on the whole,
competition was forthcoming for tenders, though the number of bidders might be quite small.
Evidence was insufficient to judge the relative merits of competitive versus negotiated
contracts for clinical care, though there was some evidence that contracts had been negotiated
at inflated prices.
Contracts rarely included sufficient specifications or allocation of responsibilities to allow
contract performance to be monitored. This was a particular problem for clinical contracts,
though non-clinical contract monitoring was also considered poor. Lalta (1993) reported how
careful contract specification and monitoring procedures (supported by USAID technical
assistance) in a contract for cleaning in a hospital in Jamaica led to generally satisfactory
performance, though expressed concern that such assistance might not have lasting effects.
All case-studies indicated the importance of devolving responsibility for monitoring to
hospital level.
27
The characteristics of the supply and labour markets
Contracting out requires the existence of private sector firms capable of managing contracts.
This requirement was met for hotel-type services in Bombay, Mexico, South Africa and
Thailand, but probably not in PNG. (Availability of contractors with sufficient experience
and capacity was a problem in Jamaica (Lalta 1993). These four countries also had sizeable
numbers of private health providers, making contracting of clinical services a possibility.
Nonetheless, the availability of private firms on its own was not a sufficient condition for
successful contracting since public contracts might not be considered sufficiently profitable.
This is especially likely to be the case where the private sector is growing rapidly and there
are plenty of investment opportunities within it. When growth slows, as in South Africa,
there may be more interest in government contracts.
For services dependent on manual labour such as cleaning, the ability of the private sector
to recruit cheap labour was clearly an important factor in cost savings. Only in Bangkok was
the labour market threatening to make it difficult to recruit cheap labour.
8.
CONCLUSIONS
This review has identified a number of important conclusions and unanswered questions in
relation to the implementation of selective contracting in the context of developing countries.
Perhaps the most important conclusion is that there remains relatively limited and contradictory
evidence on the impact of selective contracting on efficiency and equity at the facility and/or at
the health system level. This is true for both developed and developing countries, and highlights
the need for extensive additional research on the effects of the various reforms now being
implemented in various countries.
This review has also shown that many of the theoretical claims on the basis of which
contracting reforms are argued to improve efficiency themselves remain ambiguous. This
ambiguity is important since it leads to uncertainty as to the determinants of efficiency gains
through contracting, and hence, as to the set of conditions that are necessary for achieving
efficiency gains.
This is illustrated, firstly, by the set of issues concerning the relationship between the nature
of the contract, the contracting process, and efficiency. This review has highlighted several
aspects of the design of contracts which may impact on contractor behaviour and hence
efficiency, but the relative importance of each of these aspects, their interrelationships, and their
individual and combined impacts on efficiency merit further investigation. Regarding the
relationship between the contracting process and efficiency, the relative merits of awarding
contracts competitively or through direct negotiation remain unclear, as do other issues such as
the optimal number of bidders for a contract, and the trade-offs between securing adequate
numbers of bidders and ensuring an efficient distribution of risk between the contractor and the
purchaser.
Closely related to these issues of contract design and process are questions relating to
government capacity to act as a competent purchaser of health services. This review has
28
indicated that efficiency gains from contracting appear to be contingent on government capacity
to act as an efficient purchaser, and more specifically, to make the appropriate decisions as to
whether and when to let contracts, to design efficient contracts, and to monitor effectively
contractor compliance. Conversely, lack of this capacity may lead to inefficiency through
exploitation by contractors, through distorted resource allocation (Bennett 1991, Mills 1995), or
through uncontrolled expansion of the private sector, creating further problems of fragmentation
and inequity (Saltman 1991).
Some analysts have pointed to a generic set of skills and resources that governments require
in this context, including skills in planning, economic analysis, and contract design and
negotiation, as well as suitable information systems (Kutzin 1995, Bennett, Russell and Mills
1996, Bennett and Mills 1998), and sophisticated government regulatory capacity to carry out
such functions as licensing and accreditation (Figueras and Saltman 1997). Not surprisingly,
current evidence suggests that most developing country governments lack all or most of these
capacity requirements (Mills 1995, Bennett, Russell and Mills 1996). However, there is as yet
limited evidence on the relative importance of these various aspects of government contracting
capacity, or on their specific impacts on contractual efficiency. It is thus difficult, without further
research, to identify accurately those situations in which governments are likely to achieve
efficiency gains, or in which these reforms should be avoided until specific aspects of contractual
capacity have been strengthened.
A second critical area of ambiguity concerns the relationship between competition and
contractual efficiency. The theoretical literature suggests that at least some degree of
contestability for contracts, or preferably actual competition, is required to ensure efficiency
gains from contract-based provision. There is also some empirical evidence that where
competition or contestability is absent, efficiency may be undermined through contracts biased
towards contractors, through exploitation by for-profit contractors, or through governments
becoming dependent on a single monopolistic contractor.
The limited evidence available on these issues suggests that the conditions necessary for
competition, and even for contestability, are generally absent from most areas of most low and
middle income countries (McPake and Ngalande Banda 1994, Chemichovsky 1995, Mills 1995,
Saltman 1995). However, there remains scant information on the actual extent of competition
or contestability in most countries, an issue that is made more complex by geographical
variations in levels of competition, as well as by the fact that both local and international
providers may compete for contracts. Similarly, there are still few data on the precise
relationships between competition or contestability and contractual efficiency, so that is not
possible to predict the likely success of contracting under various competitive conditions.
This review has also suggested that the success of contracting may be contingent on a number
of features of the broader social, political and economic environment. Important factors here
appear to include a general political and social environment in which corruption is discouraged,
in which contractors share a commitment to public responsibility and contractual compliance,
and in which there is an effective legal system to ensure that sanctions for non-compliance pose
a meaningful threat. Where such conditions are absent, there is the risk that contracts may be
inappropriately awarded, and that contractors may exploit contracts, thus undermining efficiency
29
(Saltman 1995. Schieber 1995, Bennett, Russell and Mills 1996). There is also evidence that
under conditions of inadequate financial resources, contracting may not lead to efficiency gains.
For example, financially constrained governments may wish to let only short-term contracts,
which may be unattractive to potential bidders. Contract prices may also be set too low, leading
to poor quality of services (Mills 1995, Gilson et al. 1997); and contracts may lock public
resources into a specific use. limiting the flexibility which governments have to reallocate such
resources (McPake and Hongoro 1993, Beracochea 1997). Once again, however, there remains
quite limited empirical evidence on the extent to which these various conditions relating to the
broader environment, alone or combined, are necessary for the achievement of efficiency gains
from contracting, and further research will be required before it is possible to predict the likely
success or failure of such reforms under various environmental conditions.
A final area of ambiguity identified by this review concerns the impact of the relationship
between public purchasers and contracted providers on provider efficiency. More specifically,
it is not yet clear whether contracting leads to transparency in the contractual relationship, or to
decentralisation of management authority, nor is it clear to what extent these consequences of
contracting contribute to efficiency gains. The potential advantages of a greater awareness of
needs, prices, quality and quantities in resource allocation are dependent on the availability of
detailed information, and on the administrative capacity to use this information. The evidence,
cited above, on poor government administrative capacity in many developing countries suggests
that contracting may not necessarily produce the degree of Transparency of trade’ claimed by
proponents of these reforms. Similarly, while contracting will clearly encourage some degree
of managerial decentralisation relative to direct public management, the general lack of
managerial expertise in developing countries may prevent effective decentralisation amongst the
majority of providers, even where this is formally introduced. In developed countries and in
some middle income developing countries, however, contracting may produce these desired
consequences.
Together these observations suggest that there remains much uncertainty as to the overall
impact of contracting on efficiency and equity, as well as to the determinants of efficiency and
the necessary conditions for ensuring efficiency gains when such reforms are implemented.
30
REFERENCES
Alper PR (1994). Primary care in transition. Journal of the American Medical Association
272(19) 1523-7.
Alvarez F, Rodriguez C and Cruz C (1995). Contracting out of clinical and non clinical
services: the case of Mexico. Paper presented at the second meeting of the Public Private mix
network, Worthing, 4-8 Sept.
Annel A (1995). Implementing planned markets in health: The Swedish case. In Saltman RB
and Von Otter C (eds). Implementing planned markets in health care: Balancing social and
economic responsibility. Open University Press, Buckingham.
Appleby J, Smith P, Ranade W, Little V and Robinson R (1993). Monitoring managed
competition. In Robinson R and Le Grand J (ed). Evaluating the NHS reforms. Kings Fund
Institute, London.
Ashton T (1995). The purchaser-provider split in New Zealand: The story so far. Australian
Health Review 18:43-60.
Audit Commission (1993). Realising the benefits of competition: The client role for contracted
services. HMSO, London.
Barnum H, Kutzin J and Saxenian H (1995). Incentives and provider payment methods.
International Journal of Health Planning and Management 10:23-45.
Bartlett W and Le Grand J (1993). The theory of quasi-markets. In Le Grand J and Bartlett W
(eds). Quasi-markets and social policy. The Macmillan Press, London.
Bartlett W and Le Grand J (1993). The performance of trusts. In Robinson R and Le Grand J
(ed). Evaluating the NHS reforms. Kings Fund Institute, London.
Begley CE, Dowd CM and McCandless R (1989). A cost-effectiveness evaluation of primary
health care projects for the poor. Evaluation and the Health Professions 12(4)437-52.
Bennett S (1991). The mystique of markets: public and private health care in developing
countries. PHP Departmental Publication No. 4. London: Department of Public Health and
Policy, London School of Hygiene and Tropical Medicine.
Bennett S and Mills A (1993). Health insurance and the private sector. In Bennett S and
Mills A (eds). Report of the workshop on the public/private mix for health care in developing
countries. London: Health Economics and Financing Programme, Health Policy Unit,
Department of Public Health and Policy, London School of Hygiene and Tropical Medicine.
31
.
c
\
HE-I00
05572
'•
k
O'*5
-------
■
Bennett S and Mills A (1998). Government capacity to contract: health sector experience and
lessons. Public Administration and Development (forthcoming).
Bennett S, Russell S and Mills A (1996). Institutional and economic perspectives on
government capacity to assume new roles in the health sector: A review of experience. PHP
Departmental Publication No. 22. London: Department of Public Health and Policy, London
School of Hygiene and Tropical Medicine.
Beracochca E (1997). Contracting out of non-clinical sen ices: the experience of Papua New
Guinea. In Bennett S, McPake B and Mills A (eds). Private Health Providers in Developing
Countries: Serving the public interest? Zed Books, London.
Bhatia M and Mills A (1997). Contracting out dietary sendees in Bombay. In Bennett S,
McPake B, and Mills A (eds). Private providers in developing countries: Sen ing the public
interest? Zed Books, London.
Birdsall N and James E (1992). Health, government and the poor. Tire case for the private
sector. PHN Working Papers. World Bank, Washington DC.
Borren P (1993). Searching for the holy grail in the Antipodes. The market reform of the New
Zealand health care system. Discussion Paper No. 103. York: Centre for Health Economics,
University of York.
Boston J et al. (1991). Reshaping the state: New Zealand's bureaucratic revolution. Oxford
University Press, Auckland.
Briscoe J (1989). Adult health in Brazil. Adjusting to new challenges. World Bank Report No.
7807-BR. World Bank, Washington DC.
Broomberg J (1994). Health care markets for export? Lessons for developing countries from
European and American experience. Departmental Paper No. 12, London School of Hygiene and
Tropical Medicine.
Broomberg J, Masobe P and Mills A (1997). Contracting out district hospital services in South
Africa. In Bennett S, McPake B and Mills A (eds). Private Health Providers in Developing
Countries: Serving the public interest? Zed Books, London.
Brown PAJ et al. (1992). Skin lesion removal: Practice by general practitioners in Grampian
region before and after April 1990. Scottish Medical Journal 37:144-1.
Carnaghan R and Bracewell-Milnes B (1993). Testing the market: competitive tendering for
government services in Britain and abroad. Institute of Economic Affairs, London.
Cassels A (1995). Health sector reform: Key issues in less developed countries. Forum on
Health Sector Reform Discussion Paper No. 1. WHO/SHS/NHP/95.4. WHO, Geneva.
32
Cruz C and Zurita B (1993). The public private mix - the case of Mexico. In Bennett S and
Mills A (eds). Report of the workshop on the public/private mix for health care in developing
countries. Health Economics and Financing Programme, Health Policy Unit, Department of
Public Health and Policy, London School of Hygiene and Tropical Medicine.
Culyer AJ, Maynard A and Posnett JW (1990). Reforming health care. An introduction to
the economic issues. In Culyer AJ, Maynard AK and Posnett JW (eds). Competition in health
care. Reforming the NHS. The Macmillan Press, London.
De Geyndt W (1990). Managing health expenditures under National Health Insurance. The
case of Korea. World Bank Technical Paper No. 156. World Bank, Washington DC.
Department of Health, Republic of South Africa (1996). Restructuring the national health
system for universal primary health care: Report of the Committee of Inquiry into a national
health insurance system. Department of Health, Pretoria.
Diderichsen F (1995). Market reforms in health care and sustainability of the welfare state lessons from Sweden. In Berman P (ed). Health sector reform in developing countries: Making
health development sustainable. Harvard School of Public Health, Boston.
Domberger S, Meadowcroft S and Thompson D (1987). The impact of competitive tendering
on the costs of domestic hospital services. Fiscal Studies 8(4):39-54.
Elhvood P, Enthovcn AC and Ethcredge A (1992). The Jackson Hole initiatives for a twenty
first century American health care system. Health Economics 1:149-168.
Flynn R and Williams G (1997). Contracting for Health. Oxford University Press, Oxford.
Forcit KG (1992). Private sector approaches to effective family planning. Policy Research
working paper 940, Population and Fluman Resources Department.
World Bank.
Washington DC.
Gilson L, Adusci J, Arhin D, Hongoro C, Mujinja P and Sagoe K (1997). Should African
governments contract out clinical services to church providers? In Bennett S, McPake B and
Mills A (eds). Private health providers in developing countries: Serving the public interest?
Zed Books, London.
Gilson L, Dave Sen P, Mohammed S and Mujinja P (1994). The potential of health sector
non-governmental organisations: policy options. Health Policy and Planning 9(l):14-24.
Gleeson G (1996). The role of market testing in the NHS. British Journal of Hospital Medicine
55(7):449-450.
Glennester H, Matsoganis M, Owens P and Hancock S (1993). GP fundholding: Wild card
or winning hand. In Robinson R and Le Grand J Evaluating the NHS reforms. Kings Fund
Institute, London.
33
Glennerster H, Matsoganis M, Owens P and Hancock S (1994).
Implementing GP
fundholding. Open University Press, Milton Keynes.
Glennester H and Lc Grand J (1995). The development of quasi-markets in welfare provision
in the United Kingdom. International Journal of Health Services 25(2):203-218.
Gonzales-Sedano E (1995). Health sector reform in Colombia. Modelling managed
competition for improved efficiency and better quality in health services. World Bank: mimeo.
Gosden T and Torgerson D (1997). The effects of fundholding on prescribing and referral
costs: A review of the evidence. Health Policy 40:103-114.
Great Britain Department of Health (1989). Working for patients. London: HMSO.
Green A (1987). The role of non-governmental organisations and the private sector in the
provision of health care in developing countries. International Journal of Health Planning and
Management 2(1 ):37-58.
Griffin GC (1990). Health sector financing in Asia. World Bank Report No. 1 DP-68. World
Bank, Washington DC.
Guesneri R (1989). Hidden actions, moral hazard and contract theory. In Eatwell J, Milgate M
and Newman P (eds). The New Palgrave. Allocation, information and markets. The Macmillan
Press, London.
Hillman DG and Christianson JB (1984). Competitive bidding as a cost-containment strategy
for indigent medical care: the implementation experience in Arizona. Journal of Health Policy,
Politics and Law 9(3):427-51.
Hillman AL, Pauly MV and Kerstein JJ (1989). How do financial incentives affect physicians
clinical decisions and the financial performance of health maintenance organisations? New
England Journal of Medicine 13, 86-92.
Hood C (1991). A public management for all seasons. Public Administration 69( 1 ):3-19.
Hospital Strategy Project Consortium (1996). Towards equity and efficiency in South Africa's
public hospital system. Final Report of the Hospital Strategy Project. Volume 4. Module 3:
Cost recovery and contracting out. Hospital Strategy Project Consortium, Johannesburg.
Hurst JW (1991). Reforming health care in seven European nations. Health Affairs Fall: 172189.
Hughes D and Yule B (1992). The effect of per-item fees on the behaviour of general
practitioners. Journal of Health Economics 11: 413-37.
34
Iglchart JK (1993). Health Policy Report. Managed Competition. New England Journal of
Medicine 328(16): 1208-1212.
Jackson PM (1994). The new public sector management: surrogate competition and contracting
out. In Jackson PM and Price CM (ed). Privatisation and regulation. Longman, London.
Jackson PM and Price CM (1994). Privatisation and regulation: A review of the issues.
Longman, New York.
Jiminez J (1993). Chile’s health sector reform. Paper prepared for conference: Health sector
reform in developing countries: Issues for the 1990s. Data for Decision Making Project,
Department of Population and International Health, Harvard School of Public Health, Boston.
Jimenez de la Jara J and Bossert TJ (1995). Chile’s health sector reform: Lessons from four
reform periods. In Berman P (ed). Health sector reform in developing countries: Making health
development sustainable. Harvard School of Public Health, Boston.
Kane NM (1995). Costs, productivity and financial outcomes of managed care. In Saltman RB
and Von Otter C (eds). Implementing planned markets in health care. Open University Press,
Buckingham.
Kim Y (1987). Health care financing in Korea. Country paper prepared for the regional seminar
on health care financing. Asian Development Bank, Manila.
Krashinsky M (1986). Transactions costs and a theory of the nonprofit organisation. In RoseAckerman S (ed). The economics of nonprofit institutions: Studies in structure and policy.
Oxford University Press, Oxford.
Kutzin J (1995). Experience with organisational and financing reform of the health sector. SHS
Paper N° 8. WHO, Geneva.
LAC Health and Nutrition Sustainability' (1995). Divestment in Jamaica’s public hospitals.
Management and Financing Update, May.
Lalta S (1993). Cost containment and competitive tendering for support services in the health
care system of Jamaica. Institute of Social and Economic Research, University of the West
Indies, Jamaica.
Lazear EP (1989). Incentive contracts. In Eatwell J, Milgate M and Newman P (eds). The New
Palgrave. Allocation, information and markets. The Macmillan Press, London.
Le Grand J (1991). Quasi-markets and social policy. Economic Journal 101:1256-1267.
Le Grand J (1994). Evaluating the NHS reforms. In Robinson R and Le Grand J (eds).
Evaluating the NHS reforms. King’s Fund Institute, London.
35
Lc Grand J and Bartlett W (1993). Quasi-markets and social policy. The Macmillan Press,
London.
Ledyard JO (1989). Incentive compatibility. In Eatwell J, Milgate M and Newman P (eds).
The New Palgrave. Allocation, information and markets. The Macmillan Press, London.
Lomas J et al. (1989). Paying physicians in Canada: Minding our Ps and Qs. Health Affairs
8(1)80-102.
Lynch ML (1994). The uptake of childhood immunisation and financial incentives to general
practitioners. Health Economics 3:117-25.
MacDonald GM (1984). New directions in the economic theory of agency. Canadian Journal
of Economics 17:415-40.
MacNeil I (1978). Contracts: Adjustment of long-term economic relations under classical,
neoclassical and relational contract law. Northwestern University Law Review 72: 854-905.
Mays N (forthcoming). The impact of the UK internal market 1991-7: A balance sheet of the
evidence. In Mills A (ed). Reforming health sectors. Kegan Paul. London.
McCombs JM and Christianson JB (1987). Applying competitive bidding to health care.
Journal of Health Policy, Politics and Law 12(4):703-22.
McGreevcy WP (1990). Social security in Latin America: Issues and options for the World
Bank. World Bank Report No. 8456-LAC. World Bank, Washington DC.
McPake B and Hongoro C (1995). Contracting out of clinical sendees in Zimbabwe. Social
Science and Medicine 41(1): 13-24.
McPake B and Ngalande Banda E. (1994). Contracting out of health services in developing
countries. Health Policy and Planning 9(I ):25-30.
Improving the efficiency of public sector health services in developing
countries: Bureaucratic versus market approaches. Departmental publication No. 17, London
School of Hygiene and Tropical Medicine.
Mills A (1995).
Mills A (1997). Contractual relationships between government and the commercial private
sector in developing countries: Are they a good idea in health? In Bennett S, McPake B and
Mills A (eds). Private health providers in developing countries: Serving the public interest?
Zed Books, London.
Mills A (1998). To contract or not to contract? Issues for low income countries. Health Policy
and Planning 13(l):32-40.
36
Mills A, Hongoro C and Broombcrg J (1997). Improving the efficiency of district hospitals:
Is contracting an option? Tropical Medicine and International Health 2(2): 116-126.
Milne RG (1987). Competitive tendering in the NHS: An economic analysis of the early
implementation of HC(83)18. Public Administration 65(2): 145-60.
Milne RG and McGee MA (1992). Compulsory competitive tendering in the NHS: A new
look at some old estimates. Fiscal Studies 13(3):96-l 11.
Moore M (1996). Public sector reform: Downsizing, restructuring and improving performance.
Revised paper for Forum on Health Sector Reform. Mimeo. WHO, Geneva.
OECD (1992). The reform of health care. A comparative analysis of seven OECD countries.
Health Policy Series No. 2. OECD, Paris.
Pezza P and Barquero Bolenos J (1994). Clinica Tibas: An experimental response to health
system challenges in Costa Rica. Journal of the Royal Society of Flealth. 25 October.
Price MR, Masobe P and Booysen M (1993). The public/private mix for health care in South
Africa. In Bennett S and Mills A (eds). Report of the workshop on the public/private mix for
health care in developing countries. London: Health Economics and Financing Programme,
Health Policy Unit, Department of Public Health and Policy, London School of Hygiene and
Tropical Medicine.
Propper C (1993). Quasi-markets, contracts and quality in health and social care: The US
experience. In Le Grand J and Bartlett W (ed). Quasi-markets and social policy. The Macmillan
Press, London.
Robinson R (1990). Competition and health care: A comparative analysis of UK plans and US
experience. Research Report No. 6. Kings Fund Institute, London.
Robinson R and Le Grand J (1995). Contracting and the purchaser-provider split. In
Saltman RB and Von Otter C (eds). Implementing planned markets in health care: Balancing
social and economic responsibility. Open University Press, Buckingham.
Ryan M (1992). The agency relationship in health care: Identifying areas for future research.
Health Economics Research Unit, University of Aberdeen, Discussion paper No. 02/92.
Saltman RB (1995). Applying planned market logic to developing countries1 health systems:
An initial exploration. WHO/SHS/NHP/95.7. WHO, Geneva.
Saltman R and Figueras J (1997). European health care reform: Analysis of current strategies.
WHO regional publications, European Series, No. 72, WHO Regional Office for Europe,
Copenhagen.
37
Saltman RB and Von Otter C (1992). Reforming Swedish health care in the 1990's: The
emerging role of‘public firms’. Health Policy 21(2):143-154.
Saltman RB and Von Otter C (1995a). In Saltman RB and Von Otter C (ed). Implementing
planned markets in health care: Balancing social and economic responsibility. Open University
Press, Buckingham.
Saltman RB and Von Otter C (1995b). Introduction. In Saltman RB and Von Otter C (ed).
Implementing planned markets in health care. Open University Press, Milton Keynes.
Schieber GC (1995). Preconditions for health reform: Experiences from OECD countries. In
Berman P (ed). Health sector reform in developing countries: Making health development
sustainable. Harvard School of Public Health, Boston.
Schlackman N (1993). Evolution of a quality-based compensation model: The third generation.
American Journal of Medical Quality 8(2): 103-110.
New methods of financing and managing health care in the Russian
Federation. In Berman P (ed). Health sector reform in developing countries: Making health
development sustainable. Harvard School of Public Health, Boston.
Sheiman I (1995).
Silverstein G and Kirkman-Liff B (1995). Physician participation in Medicaid managed care.
Social Science and Medicine 41(3):355-63.
Self-governing trusts and GP fundholders: the British experience. In
Saltman RB and Von Otter C (eds). Implementing planned markets in health care. Open
University Press, Buckingham.
Smee CH (1995).
Spurgeon P, Smith P, Straker M, Deakin N, Thomas N and Walsk K (1997). The experience
of contracting in health care. In Flynn R and Williams G (eds). Contracting for Health. Oxford
University Press, Oxford.
Stearns SC, Wolfe BL and Kindig DA (1992). Physician responses to fee for service and
capitation payment. Health Affairs 12, 248-57 (suppl).
Stiglitz JE (1989). Principal and agent. In Eatwell J, Milgate M and Newman P (eds). The
New Palgrave. Allocation, information and markets. The Macmillan Press, London.
Swan M and Zwi A (1997). Private practitioners and public health: Close the gap or increase
the distance? PHP Publications No. 24, London School of Hygiene and Tropical Medicine.
Szymanski S and Wilkins S (1992). Cheap rubbish? Competitive tendering and contracting
out in refuse collection - 1981-88. London Business School, Centre for Business Strategy.
Tancharoensathien V and Supachutikul A (forthcoming). The social security scheme in
Thailand: What lessons can be drawn? Social Science and Medicine
38
United States General Accounting Office (US GAO) (1993a). Medicaid: States turn to
managed care to improve access and control costs. US GAO, Washington DC.
United States General Accounting Office (US GAO) (1993b). Managed health care: Effect
on employers’ costs difficult to measure. GAO/HRD-94-3. Report to the chairman,
Subcommittee on health, Committee on Ways and Means, House of Representatives,
Washington DC.
Tangcharoensathicn V, Nittayaramphong S and Khungsawatt S (1997). Private sector
involvement in public hospitals: Case-studies from Bangkok. In Bennett S, McPake B and
Mills A (eds). Private providers in developing countries: Serving the public interest?.
Zed Books, London.
Van de Ven WPMM (1989). A future for competitive health care in the Netherlands. NHS
White Paper Occasional Paper No. 9. York. Centre for Health Economics, University of York.
Van de Ven WPMM and Van Vliet RCJA (1992). How can we prevent cream skimming in
a competitive insurance market? In Zweifel P, Freeh III HE (eds). Health economics worldwide
Kluwer Academic publishers, Dordrecht.
Van de Ven WPMM, Schut FT and Rutten FFH (1994). Forming and reforming the market
for third party purchasing of health care. Social Science and Medicine 39(10): 1405-12.
Van Zyl N, Fourie A, Heunis C, Pelser A, Roosjen JA, Stander D, van der Merwe P and
van Zyl H (1996). Contracting out of institutional care for Tuberculosis patients in the Free
State: An analysis. Mimeo. Centre for Health Systems Research, University of the Free State,
Bloemfontein.
Vining AR and Weimer DL (1990). Government supply and government production failures:
A framework based on contestability. Journal of Public Policy 10:1-22.
Von Otter C and Saltman RB (1991). Towards a Swedish health policy for the 1990s: Planned
markets and public firms. Social Science and Medicine 32(4):473-481.
Von Otter C and Saltman RB (1992). Planned markets and public competition. Strategic
reform in Northern European health systems. Open University Press, Buckingham.
Walsh K (1995). Public services and market mechanisms: Competition, contracting and the
new public management. The Macmillan Press, London.
Walsh K and Davis H (1993). Competition and sendee: The impact of the Local Government
Act (1988). HMSO, London.
World Bank (1993). Investing in health. World Development Report 1993. Oxford University
Press, Oxford.
39
World Bank (1995). Bureaucrats in business: The economics and politics of government
ownership. Oxford University Press, Oxford.
Yepes F and Sanchez A (1998). Reforming pluralist systems: The case of Colombia. In
Mills A (ed). Reforming health sectors. Kegan Paul, London.
40
Position: 753 (6 views)