Elsevier Editorial System(tm) for Social Science & Medicine Manuscript Draft Manuscript Number: SSM-D-12-00387R1 Title: Paying for Performance and the social relations of health care provision: an anthropological perspective Article Type: Review Keywords: paying for performance in developing countries; health systems; motivation; social relations; trust; monitoring and evaluation; anthropology; ethnography Corresponding Author: Ms Priscilla Magrath, Corresponding Author's Institution: University of Arizona First Author: Priscilla Magrath Order of Authors: Priscilla Magrath; Mark Nichter Manuscript Region of Origin: USA Abstract: Over the past decade, the use of financial incentive schemes has become a popular form of intervention to boost performance in the health sector. Often termed 'paying for performance' or P4P, they involve "…the transfer of money or material goods conditional upon taking a measurable action or achieving a predetermined performance target (Eldridge and Palmer, 2009). " P4P appear to bring about rapid improvements in some measured indicators of provider performance, at least over the short term. However, evidence for the impact of these schemes on the wider health system remains limited, and even where evaluations have been positive, unintended effects have been identified. These have included: 'gaming' the system; crowding out of 'intrinsic motivation'; a drop in morale where schemes are viewed as unfair; and the undermining of social relations and teamwork through competition, envy or ill feeling. Less information is available concerning how these processes occur, and how they vary across social and cultural contexts. While recognizing the potential of P4P, the authors argue for greater care in adapting schemes to particular local contexts. We suggest that insights from social science theory coupled with the focused ethnographic methods of anthropology can contribute to the critical assessment of P4P schemes and to their adaptation to particular social environments and reward systems. We highlight the need for monitoring P4P schemes in relation to worker motivation and the quality of social relations, since these have implications both for health sector performance over the long term and for the success and sustainability of a P4P scheme. Suggestions are made for ethnographies, undertaken in collaboration with local stakeholders, to assess readiness for P4P; package rewards in ways that minimize perverse responses; identify process variables for monitoring and evaluation; and build sustainability into program design through linkage with complementary reforms. Eldridge, C., & Palmer, N. (2009). Performance-based payment: some reflections on the discourse, evidence and unanswered questions. Health policy and planning, 24(3), 160-166.

Title: Paying for Performance and the social relations of health care provision: an anthropological perspective

First Author: Priscilla Magrath, University of Arizona

Second Author: Mark Nichter, University of Arizona

Corresponding Author:

Priscilla Magrath,

University of Arizona,

PO Box 210030

Tucson, AZ 85721-0030.

Email: pmagrath@email.arizona.edu

Acknowledgements:

This paper was inspired by Mark Nichter's attendance at the Workshop:

"Expert Consultation on Organization and Service Delivery in HealthCare in Low-Income

Settings"

Sponsored by the Global Center for Health Economics & Policy Research, University of

California, Berkeley, April16-17, 2010

At the workshop the need for anthropological input into this area of health service research

was identified.

The paper received helpful comments from Professor Sallie Marston, School of Geography

and Development, University of Arizona

Cover Page

Research Highlights

1. Reviews theoretical insights and empirical evidence concerning impacts of Pay for

Performance (P4P) schemes in health

2. Highlights issues and concerns relating to social relations of health care provision in

developing countries adopting P4P

3. Suggests a social science framework for understanding why and how impacts might

vary with the socio-cultural setting

4. Suggests theoretically informed anthropological research can help critically assess

P4P in relation to particular settings

5. Argues stakeholder participation in design, monitoring and evaluation may help

reduce unintended impacts of P4P schemes

Highlights (for review)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

1

Paying for Performance and the social relations of health care provision: an anthropological perspective

Key words: paying for performance in developing countries; health systems; motivation;

social relations; trust; monitoring and evaluation; anthropology; ethnography.

Abstract

Over the past decade, the use of financial incentive schemes has become a popular form

of intervention to boost performance in the health sector. Often termed „paying for

performance‟ or P4P, they involve “…the transfer of money or material goods

conditional upon taking a measurable action or achieving a predetermined performance

target (Eldridge and Palmer, 2009). “ P4P appear to bring about rapid improvements in

some measured indicators of provider performance, at least over the short term. However,

evidence for the impact of these schemes on the wider health system remains limited, and

even where evaluations have been positive, unintended effects have been identified.

These have included: „gaming‟ the system; crowding out of „intrinsic motivation‟; a drop

in morale where schemes are viewed as unfair; and the undermining of social relations

and teamwork through competition, envy or ill feeling. Less information is available

concerning how these processes occur, and how they vary across social and cultural

contexts.

While recognizing the potential of P4P, the authors argue for greater care in adapting

schemes to particular local contexts. We suggest that insights from social science theory

coupled with the focused ethnographic methods of anthropology can contribute to the

critical assessment of P4P schemes and to their adaptation to particular social

environments and reward systems. We highlight the need for monitoring P4P schemes in

*ManuscriptClick here to view linked References

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

2

relation to worker motivation and the quality of social relations, since these have

implications both for health sector performance over the long term and for the success

and sustainability of a P4P scheme. Suggestions are made for ethnographies, undertaken

in collaboration with local stakeholders, to assess readiness for P4P; package rewards in

ways that minimize perverse responses; identify process variables for monitoring and

evaluation; and build sustainability into program design through linkage with

complementary reforms.

Eldridge, C., & Palmer, N. (2009). Performance-based payment: some reflections on the

discourse, evidence and unanswered questions. Health policy and planning, 24(3), 160-

166.

1. The policy context: why financial incentives, why now?

Financial incentives, also termed paying for performance (P4P) or performance based

financing (PBF), can be defined as: “…the transfer of money or material goods

conditional upon taking a measurable action or achieving a predetermined performance

target (Eldridge and Palmer, 2009).“

P4P have become a popular form of intervention in national health sectors over the past

decade. A number of donors have promoted P4P including the World Bank, World

Health Organization (WHO), the Global Alliance for Vaccines and Immunizations

(GAVI), the Global Fund to fight AIDS, TB and malaria (Global Fund), the UK

Department for International Development (DFID) and the US Agency for International

Development (USAID). P4P appears to have the potential to bring about rapid

improvements in measured indicators of provider performance (Kalk et al., 2010, Paul,

2009, Soeters et al., 2006), although the evidence base remains limited (Oxman and

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

3

Fretheim, 2008, Eldridge and Palmer, 2009) and doubts have been raised concerning

effectiveness in settings with weak health infrastructure and information systems

(Lagarde et al., 2007, Powell-Jackson et al., 2009).

P4P programs targeting health providers have been implemented in the US (Mehrotra et

al., 2007, Oldani, 2010), UK (McDonald and Roland, 2009), Haiti (Eichler et al., 2007),

Cambodia (Soeters and Griffiths, 2003), China (Yip et al., 2010) and a number of African

countries, including Rwanda (Soeters et al., 2006), Burundi, Zambia, Democratic

Republic of Congo and Tanzania (Toonen et al., 2009). Following the apparent success of

these programs, a number of other African countries are currently planning to introduce

P4P (Meessen et al., 2011, Soeters, 2011).

One reason for P4P‟s popularity lies in the pressure to achieve the Millennium

Development Goals (MDGs) (Oxman and Fretheim, 2009). The use of market

mechanisms to achieve these short term targets has appeared to make sense within the

broader context of market based policy reforms, often termed „neoliberal‟, which assume

markets to be the most efficient and effective mechanism for allocating resources in the

public as well as private sectors (Foucault, 2008). Indeed, P4P is often linked with other

market mechanisms such as contracting out to the private sector or NGOs (Soeters and

Griffiths, 2003, Eldrige and Palmer, 2009). Within the context of this neoliberal policy

discourse P4P has been framed as „successful‟ despite limited empirical evidence to

support the claim (Bowman, 2010). Evaluations of financial incentive schemes rarely

examine their wider impacts on health system performance over the long term, and there

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

4

is even less evidence concerning cost effectiveness (Oxman and Fretheim, 2009, Toonen

et al., 2009, Montagu and Yamey, 2011, Ireland et al., 2011).

While the MDGs encourage a focus on achievement of specific targets, the global health

community is finally waking up to the importance of strengthening national health

systems in a more integrated manner (Bradley et al., 2011, Frenk, 2010, Travis et al.,

2004, Mills et al., 2008, Mills, 2011). Global health is increasingly viewed by policy

makers as an economic and security issue encouraging an „avalanche of aid‟ (Pfeiffer and

Nichter, 2008), with development assistance to health in low and middle income

countries tripling between 1997 and 2007 (Eichler, 2009). Until recently the bulk of this

aid has been channelled via vertical, disease focused programs (Travis et al., 2004).

Following decades of pressures to reduce spending on public services under structural

adjustment, national health infrastructures are now inadequate to the task of

disseminating the benefits of these programs to intended populations (Pfeiffer and

Nichter, 2008).

With increasing recognition of the need to strengthen public health systems, and of the

important role of research in this process (Mills, 2011, Mills et al., 2008, Gilson et al.,

2011), an analysis of the role of P4P in promoting or undermining health systems is

urgently required (Ireland et al., 2011). Sections 2 and 3 discuss the advantages and

potential dangers of P4P as identified in the literature to date. This literature draws on

human resource management and organizational theory and on the limited empirical data

available. What is missing from this literature is an exploration of how or why responses

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

5

to P4P might vary from one place to another. Although the importance of context in

health policy design generally (Roberts et al., 2004, Faguet and Ali, 2009, Balabanova et

al., 2011) and for P4P in particular (Ireland et al., 2011, Eldridge and Palmer, 2009) has

been emphasized in the literature, the term „context‟ means different things to different

authors. For Roberts et al. (2004) context includes the political, economic and

administrative environment. But as Faguet and Ali (2009) point out one has to go deeper

than this in order to understand the underlying values and social processes which affect

work place motivation. Section 4 draws on social science theory to suggest a framework

for understanding contextual variation in responses to P4P impacts. Section 5 suggests

how theoretically informed anthropological research can help critically assess P4P and

adapt schemes to particular settings.

2. Potential benefits of P4P: greater in theory than in practice?

P4P is appealing because the idea is simple and appears fair. If some people perform

better, why not reward them? Theoretical support for P4P can be found in organizational

theory and in economics. In organization theory, the „principal agent‟ framework

provides a rationale for the need for incentives. Principals (such as employers) need to

offer „carrots‟ to agents (such as employees) because the interests of the agent are not

perfectly aligned with those of the principal. If health bureaucracies are viewed as a

series of principal agent relationships, the use of financial incentives makes sense (Perry

et al., 2009). Shifting from the institution to the individual, economics has contributed a

number of models of individual decision making that suggest the use of financial

incentives will have positive effects on performance. These include rational choice

theory, expectations theory and reinforcement theory (Perry et al., 2009). These models

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

6

assume that decision making is based solely on the individual‟s analysis of expected

outcomes and on the pursuance of self interest. If there is some support for P4P in these

deductive models, is there any evidence that it actually works in practice?

Much of the evidence for positive impacts comes from evaluations financed by the

agencies implementing the schemes (Ireland et al., 2011). Such evaluations must be

interpreted with caution since they focus on outputs in relation to project objectives rather

than broader impacts on people and their health services. One of the most widely cited

success stories is the Rwanda P4P scheme. Rwanda was the first developing country to

scale up P4P in the health sector to the national level from 2006, following pilots initiated

in 2002 (Rusa et al., 2009, Kalk et al., 2010, Ireland et al., 2011, Soeters and Vroeg,

2011, Basinga et al., 2011a). Donors include the World Bank, the Global Fund and

bilateral donors. Performance indicators exclusively address MDG targets for maternal

and child health, and bonuses are paid at the facility level based on achievement of

quantitative and quality targets (Basinga et al., 2011a). Early evaluations found

performance of program indicators had improved; staff motivation was strengthened;

client satisfaction and utilization of some services increased; absenteeism had fallen;

documentation improved; and management relations and team spirit were stronger (Rusa

et al., 2009, Kalk et al., 2010). These results depended on improved monitoring and

supervision, including auto evaluation by the health centers (Rusa et al., 2009) and

community oversight (Ireland et al., 2011). However, unintended negative effects on staff

morale and performance were also documented (Kalk et al., 2010, Paul, 2009).

Furthermore, complementary reforms in the Rwandan health sector, including increased

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

7

base salaries, render it difficult to link performance outcomes to the P4P scheme (Kalk et

al., 2010). A recent evaluation by Basinga et al. (2011a) isolated the P4P effect through a

randomized controlled trial. The incidence of facility visits for childbirth and child

preventative care was higher for facilities under P4P than for those with an equivalent

level of input based financing. There was no difference in the number of women

completing four prenatal visits or of children completing immunization schedules.

Eichler et al. (2007) report similar findings from Haiti, where NGOs were contracted to

provide health services under a P4P scheme. Attended deliveries and immunization

coverage improved, while prenatal and postnatal care was not significantly affected. In

both cases it appears that outcomes with higher rewards under the scheme, and those less

dependent on patient voluntarism, yielded the most robust results (Basinga et al., 2011a).

In another positive evaluation, a P4P for public health providers in China helped realign

incentives away from a profit motive that had encouraged over treatment (Yip et al.,

2010).

In addition to their impact on output measures P4P have been reported to improve

information and management systems (Eichler and Levine, 2009), encourage more

creative use of resources and strengthened accountability and transparency (Toonen et al.,

2009), clarify roles and responsibilities (Rusa et al., 2009, Paul, 2009), and recognize

volunteer or low paid workers (Kalk et al., 2010). Furthermore, P4P may convince

Ministries of Finance to increase funding to the health sector (Meessen et al., 2011) and

help retain staff, even in remote areas.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

8

3. The potential pitfalls of P4P

3.1 Measuring performance: P4P as a ‘fatal remedy’

P4P schemes aim to improve health worker performance. Given the difficulty of

measuring overall performance P4P schemes rely on indicators such as numbers of

vaccinations delivered, patients seen or deliveries assisted. But it has been suggested that

as soon as you measure something it ceases to be a good indicator of anything beyond

that which is measured (Eldridge and Palmer, 2009). For example, a correctly completed

partogram does not prove that a successful delivery has occurred (Kalk et al., 2010).

Similarly, the achievement of P4P targets does not tell us whether health worker

performance has improved in a comprehensive or sustainable manner. Form filling may

operate more as a means of controlling workers than improving their performance.

Measurement may even become a „fatal remedy‟, leading to poorer, rather than improved

performance (Power, 1997). Organization theory offers a useful framework for

understanding behavioural responses to the measurement aspects of P4P, including

various forms of „gaming‟.

Gaming refers to strategies to maximize performance in relation to rewarded behaviours.

Gaming observed in the health sector includes falsification of data (Powell-Jackson et al.,

2009, Kalk et al., 2010, Paul, 2009, Ireland et al., 2011); oversupply of targeted services

(Rusa et al., 2009, Kalk et al., 2010); retention of drugs to avoid a stock out (Kalk et al.,

2010); and neglect of health care practices that are not included in the measures, such as

prevention, care of chronic illness or care of „difficult‟ patients, including the poor or

noncompliant (Oldani, 2010, Ireland et al., 2011, McDonald and Roland, 2009). Targets

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

9

put pressure on health staff to use their social influence to recruit participants, often

against their wishes, or for some token form of remuneration. This encourages the

enrolment of unsuitable candidates (Ireland et al., 2011) and undermines the notion of

„patient choice‟. For example, women approaching or even past menopause were

included in family planning programs having enrolment targets in India in the 1980s,

while coercive practices were reported for sterilization programs in Bangladesh and

elsewhere (Hartmann, 2011). Although targets may be met in the short term the

reputation of health services and government programs more generally can be affected

over the long term.

3.2 Meddling with motivation: ‘Can do’, ‘Will do’ and the generation of

‘double binds’

Whereas P4P rests on the assumption that people are primarily motivated by material

gain (Eldridge and Palmer, 2009), health sector workers appear to be motivated by a

combination of professional ethics, public service and economic motivation (Andersen,

2009, Paul, 2009). Non-economic forms of motivation have been found to be crucial to

health sector performance (Rayner et al., 2010, Franco et al., 2002, Paul, 2009, Toonen et

al., 2009).

The human resources management literature provides a framework for understanding the

complexity of worker motivation, dividing factors into „can do‟, the ability of the health

worker to perform, and „will do‟, the desire or willingness to work towards organizational

goals (Franco et al., 2002, Mathauer and Imhoff, 2006). „Can do‟ factors cover training,

professional competence, working conditions, availability of resources, equipment and

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

10

supplies, and the time and management support to perform adequately. P4P is based on

the assumption that the lack of „will do‟ motivation is a primary cause of poor

performance (Soeters and Griffiths, 2003). But in resource poor settings where „can do‟

factors are a key constraint P4P can lead to „double binds‟ when health workers lack the

resources needed to achieve performance targets (Eldridge and Palmer, 2009, Paul,

2009). Schemes may appear to favour facilities which are already better resourced, since

these are more likely to achieve the targets (Eldridge and Palmer, 2009, Ireland et al.,

2011).

„Will do‟ factors relate to personal goals which generate a willingness to work in line

with organizational goals. These might include vocation and professional conscience, a

desire to ease suffering or help patients, a sense of responsibility, commitment to public

service ideals or the will to achieve. „Will do‟ factors are affected by personnel

management, including clear job descriptions, career advancement, professional

recognition, training, peer support, appreciation from clients, strong leadership and clear

organization goals (Mathauer and Imhoff 2006, Dieleman et al., 2006). As well as these

external factors, „will do‟ depends on internal factors including a person‟s values and

expectations, as well as intrinsic motivation, the desire to perform a task because it is

interesting and provides satisfaction, regardless of expected consequences (Deci and

Ryan, 2012). Will do motivation is thus the result of the interaction between external and

internal factors (Franco et al., 2002).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

11

Concern has been raised that P4P, by encouraging motivation for financial rewards, may

lead to the “crowding out” of other forms of motivation. For example, those motivated by

public service values may find the job less rather than more attractive, believing their

image will be spoiled by the higher monetary rewards (Georgellis et al., 2010). On the

other hand, Yip et al. (2010) suggest that shifting motivation towards financial incentives

is easier than regenerating social or moral commitment once this is lost. Crowding out of

public service ethos and intrinsic motivation has been documented for P4P in health

(Rayner et al., 2010, Ireland et al., 2011) and was a primary concern of health policy

makers interviewed by one of the authors in several West African countries.

But P4P can undermine motivation and performance in other ways. For example, P4P

may encourage mediocrity by setting limits on expectations (Bowman, 2010). P4P

schemes may generate „double binds‟ as when time taken to complete paperwork

required for P4P conflicts with time attending to patients (Paul, 2009, Kalk et al., 2010,

Ireland et al., 2011), or when receiving a reward upsets working relationships with others

(Powell-Jackson et al., 2009). P4P schemes may add to an already heavy load of

paperwork, perceived to be unnecessary or burdensome. New monitoring systems may

also interfere with informal aspects of supervision negotiated between supervisors and

subordinates, which are seen as necessary in order for the formal systems to function

(George 2004). P4P thus has the potential to generate disincentives as well as incentives,

either for those targeted or for others. A sense that rewards are being allocated unfairly

can be a strong demotivator (Mathauer and Imhoff, 2006, Powell-Jackson et al., 2009,

Toonen et al., 2009), creating jealousies between those receiving rewards and those who

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

12

do not (Nichter, 1986). Incentives for some can also create increased workloads for

others, as when incentives offered to traditional healers to refer patients overburden

public health workers not eligible to receive rewards.

3.3 Measurement and motivation: synergistic and longer term impacts on the

health sector

Crowding out and gaming are related because the rewards which potentially lead to

crowding out are linked to measured targets which potentially generate gaming

behaviour. It has been observed that crowding out and gaming can reinforce one another.

Willingness to engage in gaming can be a result of the crowding out of non-economic

forms of motivation, while increased surveillance to limit gaming can exacerbate

crowding out, as workers self esteem and self direction falls (Paul, 2009).

Shifts in motivation can have impacts beyond the individual. Crowding out and

resentment affects morale and job satisfaction at the institutional level (Rayner et al.,

2010), or in the wider society. Gaming to achieve individual rewards can generate

competition and envy among health workers, reducing information sharing, trust and

team work (Mathauer and Imhoff, 2006). These shifting behaviours are likely to

undermine continuity of care and damage the functioning of the health system, which is

highly dependent on social relations of trust (Gilson, 2003, Gilson et al., 2005) and on

efficient information systems (Frenk, 2010). Allocating rewards to groups or facilities

might address some of these issues, although it could also introduce the risk of „free

riding‟ (Eijkenaar, 2012). Evidence of group P4P schemes is limited (Bowman, 2010,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

13

Perry, 2009). Under the Rwanda P4P scheme some staff contested the distribution of

rewards among facility members even though rewards were offered at the facility level.

(Kalk et al., 2010).

„Gaming‟ and „crowding out‟ reveal how measuring and rewarding selected behaviours

can potentially affect behaviours that are not measured or rewarded. Measurement

transforms the operations of an organization, leading to a reallocation of resources, or

colonization (Power, 1997). On the other hand, bonuses re-align behaviour towards

behaviours which are measured. P4P may lead to a focus on quantity at the expense of

quality of health care, since quantitative targets are easier to implement and monitor and

therefore likely to be selected under P4P schemes (Ireland et al., 2011). The restructuring

of health systems under P4P schemes, including transformations in financing and

monitoring systems (Toonen et al., 2009), may lead to „decoupling‟ (Power, 1997), with

individuals or resources being used to create a buffer between the organization and the

supervisors who measure performance. This diverts resources away from the primary

activity of delivery of health services (Ireland et al., 2011, Eldridge and Palmer, 2009).

Expectations can also be affected over the long term. Evidence suggests that bonuses

rapidly come to be seen as part of the salary package (Mathauer and Imhoff, 2006,

Ireland et al., 2011), while performance impacts dissipate over time (Montagu and

Yamey, 2011), and may even fall below pre P4P levels if funding is stopped (Kalk et al.,

2010), although the „half life‟ of an incentive is likely to vary from program to program.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

14

A further problem is that subsequent policy interventions that do not offer financial

incentives will fail to generate enthusiasm.

Incentive schemes are also likely to have long term effects on relations between providers

and patients. Performance bonuses for providers could increase utilization and trust if

performance improves (Soeters and Griffiths, 2003), or decrease it if patients feel that

providers are oriented towards targets and bonuses rather than patient welfare (Kalk et

al., 2010). Uneven distribution of funds to mothers under Nepal‟s safe delivery program

created mistrust of public health services “thereby reducing demand for all health

services, not just delivery care” (Powell-Jackson et al., 2009, p.10).

This review of the potential pitfalls of P4P suggests that market mechanisms in the form

of P4P will not necessarily improve the performance, efficiency or cost effectiveness of

health bureaucracies, especially given the high cost of administering these programs

(Lagarde et al., 2007, Bowman, 2010, Toonen et al., 2009, Ireland et al., 2011).

Analysts of P4P tend to assume opposing positions based either on the potential benefits

or the negative consequences of P4P (Basinga et al., 2011b). This polarization is

exacerbated by limited evidence, the difficulty of isolating P4P effects from confounding

factors (Ireland et al., 2011, Macq and Chiem, 2011, Toonen et al., 2009), and the focus

on short term outcome variables. We know very little about how outcomes are achieved,

hence whether they are sustainable over the long term (Ireland et al., 2011).

Understanding the processes by which targets are reached demands a reorientation away

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

15

from quantitative measures and towards an understanding of motivation as it affects the

social relations of health care provision.

4. Using social science theory to understand responses to P4P

Understanding responses to P4P and how they vary cross-culturally demands theory

which recognizes motivation as a social as well as individual phenomenon. The

sociologist Pierre Bourdieu offers a useful starting point. Bourdieu regards human

behaviour as structured by the “habitus”, a system of dispositions which frame people‟s

attitudes, perceptions and actions. According to this framework, motivation can be seen

as a disposition which orients actors towards actions which improve performance.

Dispositions are learned behaviours, acquired through socialization and framed by past

experiences. Bourdieu argued that people from the same social group or class tend to

have similar experiences, based on shared “conditions of existence” giving rise to a

shared habitus which generates regularities in the behaviour of members of a social

group, even in the absence of conscious coordination (Bourdieu, 1977, 1986a). Bourdieu

regarded the habitus as durable and not susceptible to change, suggesting a limitation in

the applicability of his theory to contexts of rapid behavioural change (Weiss, 2008). But

others have explored the ways in which government policy can interact with the habitus,

as well as how assumptions concerning the habitus frame such policies. In an

ethnography of a participatory development project in India, Mosse (2005) describes how

„participation‟ was translated into existing routines and agendas. The policy idea

underwent different transformations within the bureaucracy of the government civil

service and within a private company contracted by the project, leading to outcomes

which were not always intended by policy makers. Similarly, Nichter (1986, 1999) has

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

16

documented how community participation in primary health care programs in the 1970s

was based on unrealistic expectations about the existing habitus of primary health care

centres, and that this was one reason for the failure of these programs to impact the poor.

On the other hand, new policy agendas often involve shifts in assumptions concerning

people‟s dispositions. According to Greener (2002) social welfare policy in Britain was

based on the assumption of a collective public service disposition within the civil service,

whereas the more recent market based New Public Management policy is based on the

assumption that civil servants are „free agents‟ motivated primarily by material gain.

Greener suggests that this latest policy is likely to fail due to a misapprehension of the

nature of individual agency within the civil service habitus.

Although the habitus is shared within a social group, there is no assumption within

Bourdieu‟s theory of a “level playing field”. The arena in which dispositions are played

out is a structured “social field”, involving sets of positions which determine access to

resources or capital. Bourdieu distinguished between different forms of capital including

economic, cultural, social and symbolic capital. Cultural capital signals competence to

operate in a particular social field. It includes educational qualifications, such as medical

expertise, as well as embodied knowledge of appropriate behaviour such as the use of

correct manners and language in relation to peers and supervisors. Social capital refers to

the social networks that an individual can draw on for social, economic or professional

support. Symbolic capital refers to forms of recognition, such as promotions, certificates

or titles and to the prestige associated with one‟s social position or professional status.

These forms of capital are acquired by individuals through investment of time and labour,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

17

or in some cases through inheritance, in ways which are allowed and encouraged by the

habitus of their social group. The dynamics of the social field determine the extent to

which individuals can convert one form of capital to another (Bourdieu, 1986b). For

example, cultural capital could, under certain circumstances, be converted to economic

capital through securing a job for which the candidate is considered „suitable‟.

P4P schemes aim to bring about a change in health worker motivational disposition

through increasing their access to economic capital. Bourdieu offers a useful framework

for understanding how this process is mediated by pre-existing social formations.

According to this framework, P4P will be translated into the social field of the health

service, comprising the overlapping medical and social hierarchies in which doctors,

nurses, midwives, administrators and patients operate (Nichter, 1986). P4P acts on the

social field by introducing new positions and responsibilities and by altering access to

different forms of capital. Schemes thus bring about a “game change” with potential

knock on effects throughout the system. Although P4P is defined by the offer of a

material reward, all forms of capital and behaviours oriented towards acquiring them are

potentially affected. New cultural capital will be required to navigate systems of

indicators, targets and rewards while maintaining existing relationships with patients or

colleagues.

Social capital may be increased if performance bonuses are earned by and strengthen a

team, or if individuals share their bonus with others, thereby converting economic capital

into social capital. But the „bonus culture‟ could also reduce social capital, fragmenting

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

18

social networks if it encourages competition with peers, or is perceived as unfair. There is

some evidence that this has, in fact, occurred under some P4P schemes. According to

Powell-Jackson et al. (2009, p.8), under the Nepal safe delivery program: “There was

widespread discontent with the health provider incentive, even amongst those who

benefit directly such as midwives. It strained relations between health staff, particularly

when some felt the distribution of money was unjust or higher qualified staff were

ineligible to receive the incentive.”

P4P schemes also affect the distribution of symbolic capital. Reputations may be

enhanced through improved performance, publicity events, or simply through

participation in the scheme. Conversely symbolic capital may be lost if behaviours go

„against the grain‟, perhaps due to neglect of other duties or to „crowding out‟ of public

service motivation, leading to a perception that health staff are „just in it for the money‟.

The acquisition of symbolic capital marks a shift in power relations that may be open to

contestation (Shenkin and Coulson, 2007). Nichter (1999, 303) refers to the “tendency of

the health care bureaucracy to resist innovations which undermine pre-existing power

structures.” The distribution of bonuses to lower level staff may be perceived as a threat

to senior staff, especially where rewards normally reflect one‟s seniority rather than

current performance (Nichter, 1986). Conversely, existing power structures may be

reinforced as where supervisors are perceived as acquiring too much power in “do as I

say performance pay” (Bowman, 2010, 75). In cultures where envy is associated with

witchcraft accusations, earning a bonus could even be perceived as dangerous to the

recipient.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

19

Bourdieu‟s concept of convertible capital is useful for understanding how P4P might

impact motivation differently depending on the structure of local social fields and reward

systems. However, it may not capture all of the impacts of P4P on motivation. Bourdieu

focuses on different forms of social or material gain but does not address the issue of

intrinsic motivation, not dependent on external rewards. A number of psychologists have

studied intrinsic motivation. Under experimental conditions monetary rewards have been

found to undermine intrinsic motivation. One explanation is that “when extrinsic rewards

are introduced for doing an intrinsically interesting activity, people tend to feel controlled

by the rewards” (Deci and Ryan, 2012, p.234), leading to a loss of self-determination and

self-esteem (Paul, 2009). Concerns that P4P schemes may lead to the crowding out of

intrinsic motivation (Paul, 2009, Ireland et al., 2011) suggest the need for further

ethnographic studies focused on this issue.

Wider impacts

One advantage of regarding the health system as a complex evolving social field

embedded in wider socio-cultural systems is that it alerts the analyst to impacts of health

policies which extend beyond the formal health system. Focused ethnographies by

anthropologists have confirmed that health policies typically do have social impacts on

the wider society even where these are not intended (Hahn and Inhorn, 2009, Janes and

Corbett, 2011, Castro and Singer, 2004). Pfeiffer‟s (2004) ethnography of the impact of

health policy in Mozambique illustrates one way in which this can occur. Pfeiffer

describes how international donor policy to support NGOs in health service delivery in

Mozambique not only undermined the public health system but also increased socio-

economic inequality within communities served by the NGOs. The recruitment and

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

20

funding of a new cadre of NGO staff drained the public service of skilled human

resources while at the same time creating a new elite enclave class. These wider impacts

would be missed in conventional programme evaluations.

Although there is limited empirical evidence to date, it appears that P4P schemes can also

have impacts beyond the health system. For example, Toonen et al. (2009) argue that

centralized management of P4P has compromised decentralization and community

involvement in Rwanda.

5 Where do we go from here?

In this section we suggest how some of the insights emerging from this review can be

operationalized in qualitative research aimed at critically assessing P4P schemes.

Following Ulin et al., (2005, 52) and Nichter et al (2004, 1914) we advocate a “cyclical

formative-reformative research approach” involving a number of distinct but interactive

stages, from initial situational analysis and problem identification prior to implementation

of a scheme, through monitoring and evaluation to critical assessment of wider impacts.

Assessing readiness and keeping options open. Before introducing P4P, it is important

to inquire how health system performance is currently being framed and addressed by

policy makers, managers, public and private providers, health financiers and populations

served. How will P4P interact with existing policies and initiatives? How will it help to

achieve national as well as international goals? Discussion of a range of options will

allow people to raise concerns (Ireland et al., 2011, 696), identify who would support

P4P, who might lose out, and what would need to be in place for success over the long

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

21

term. Achieving local goals through P4P schemes may require developing broader sets of

indicators than those used in MDG focused programs in Rwanda and elsewhere (Basinga

et al., 2011a, Toonen et al., 2009).

Planning for gaming and crowding out. It has been suggested that the risk of crowding

out can be reduced by aligning performance measures with existing professional norms

and values which provide a foundation for intrinsic motivation (Eijkenaar, 2012). In

addition, it is necessary to understand how external reward systems work for people

holding different positions within the health social system. If new responsibilities and

targets or the level and distribution of rewards are viewed as clashing with existing

norms, or if health workers are not happy about the way in which indicators and targets

are set this could cause resentment (McDonald and Roland, 2009). Some targeted

providers in Rwanda felt that gaming was encouraged by the purchasers setting

inappropriate indicators and targets (Kalk et al., 2010, 186). Gaming might also be higher

if rewards are seen to emanate from a seemingly unlimited foreign source, rather than

from limited national or local government funds. On the other hand gaming might be

reduced by ensuring that information about the supervisory and reward system is

transparent to health service users as well as to those targeted by rewards (Greener,

2002). Involving health workers and the communities they serve in the selection of

indicators and targets might help reduce gaming and crowding out. (Paul, 2009, Toonen

et al., 2009, Eijkenaar, 2012).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

22

Packaging rewards: Responses to P4P will depend on how the packaging of rewards is

interpreted, as well as on the meaning of money and the connotations which it evokes in

particular contexts. P4P packaged as per diems, training fees or sitting allowances, may

evoke different reactions, depending on whether they are regarded as a one off

performance bonus, an entitlement, an aspect of the job, or a part of the salary package.

The packaging of rewards, including the size and frequency of payments, also affects

whether financial rewards can be hidden or converted into other forms of capital, and

their perceived fairness. Money paid to encourage behavioural changes may be regarded

as a bribe or a form of coercion in some contexts (Marteau et al., 2009, Hartmann, 2011),

or as a gift generating a moral obligation to provide a good service in others (Nichter,

1983). When “Stop Buruli” social scientists discussed offering cash payments to

traditional healers in Ghana who referred patients with Buruli ulcer to the public health

service it was warmly received. But in a neighbouring country this same payment was

interpreted as „paying for diseased bodies‟ in a cultural context where traditional healers

may be implicated in inflicting disease as well as healing.

Interpretations and responses to P4P may also vary with the professional or social sub-

group within the health system. In a qualitative study of health worker motivation in

Mali, Dieleman et al. (2006) found physicians were more strongly motivated by „feeling

responsible‟ than were nurses, whereas „increase in salary‟ was more motivating for

nurses and midwives than for physicians. Similarly, providers in the private sector may

respond differently to the package of „carrots and sticks‟ offered under a P4P scheme

than do public sector practitioners (Bennett et al., 1994). Traditional healers interviewed

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

23

by one of the authors in Cameroon said they valued respect, particularly from the medical

community, at least as much as money. This suggests that participation of local

stakeholders is needed to identify the optimal presentation and levels of rewards for each

location and social or professional group. Understanding what kinds of symbolic capital

are valued by different groups could lead to incorporation of nonmaterial rewards such as

trainings, exchange visits or the publication of collective achievements on a website,

which might be more effective at rewarding collaboration than bonuses.

Understanding the role of Teamwork in Health Systems Ethnographic research has

revealed the important role social relations and teamwork play in the performance of a

health service (Gilson, 2003, Nichter, 1986). P4P risks undermining cooperation and trust

among health workers, even where rewards are allocated to teams (Kalk et al, 2010).

More research is needed to understand how teams of health workers operate within

stable or evolving reward systems. Social network analysis to identify patterns of

collaboration in low and high performing facilities would provide a foundation for

discussion with stakeholders on this issue and for monitoring the impact of P4P schemes

on team work.

Process evaluation: There is an urgent need for process variables to complement the

outcome variables that form the basis of current monitoring and evaluation systems.

Process variables would help researchers, policy makers and health workers understand

how outcome indicators have been achieved (Mills, 2011, Mills et al., 2008). They would

involve tracking social relations throughout the system to assess the impact of P4P

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

24

schemes on cooperation, trust, and information sharing and on the quality of interactions

with patients. As Ireland et al. (2011, 696) have pointed out, “the successful referral of a

pregnant woman to a health centre or hospital for delivery is, above all, dependent on the

quality of the relationship between the woman and her health provider.” Process variables

should also measure the incidence, effects and costs of gaming and crowding out.

Currently these social impacts are addressed in an ad hoc manner during occasional

qualitative research rather than being built into design, monitoring and evaluation

(Toonen et al., 2009). Those targeted, including populations served, should be involved

in identifying and monitoring process variables in relation to targeted and non targeted

activities (Ireland et al., 2011).

Expecting expectations: short term interventions can have long term impacts on

expectations. Local historical research is needed into the types of incentive schemes that

have been tried before in health and other sectors, how past experiences are likely to

shape responses to proposed P4P schemes, and how these in turn might affect

expectations regarding subsequent interventions. Schemes in the public health sector can

have knock on effects in the private sector, and on patient trust and loyalty with respect to

other government services. The way in which P4P is presented both in policy circles and

in the media also impacts expectations. For this reason, it is useful to monitor media

representations of health sector performance and of particular interventions, since these

both reflect and frame public perceptions, which in turn influence the way problems are

addressed and policies are forged in future (Nichter, 2008).

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

25

Building sustainability: Health sector P4P schemes in developing countries remain

heavily dependent on donor support, both for financing and capacity building, bringing

into question their long term sustainability (Toonen et al., 2009). Furthermore, experience

suggests that impacts on performance can be short lived, as bonuses quickly come to be

seen as part of the normal salary package. In Rwanda Ireland et al. (2011, 696) report

“waning enthusiasm from health workers who have become accustomed to receiving

financial incentives.” Ethnographic research is needed to help assess what would be

required in a particular location not only to stimulate motivation but to sustain it over the

long term. Evidence from community health worker programs suggests that maintaining

motivation requires a mix of incentives, including symbolic recognition of achievements

and investment in career enrichment as well as material support. Incentives need to be

reviewed periodically in relation to changing workloads, competing job opportunities and

other changes in the health system (Bhattacharyya et al., 2001, Basinga et al., 2011b).

Communities can also be involved in sustaining motivation, for example through health

insurance schemes such as that implemented with the Rwanda P4P. The idea was that as

P4P boosted performance demand for health insurance would rise, generating financial

resources to sustain quality services. Given the uncertainty and unpredictability of

national and international funding, novel forms of community participation in

maintaining health sector performance are likely to be needed (Toonen et al., 2009).

6. Discussion

Anthropology can provide a bridge between social theory and practical health policy

work. Anthropology‟s strengths in this regard include the methodology of sustained

participant-observation in the research setting, the validation of local perspectives, and a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

26

tradition of drawing on a wide range of theoretical foundations (Introduction, Hahn and

Inhorn, 2009). This paper has attempted to draw together the theoretical insights and

empirical evidence relevant to an understanding of the wider impacts of P4P schemes and

how these might vary with the setting. A number of suggestions have been made as to

how these insights could be operationalized through anthropological research to assess

P4P schemes critically in relation to local socio-cultural contexts.

This approach is in line with that advocated by Mills et al., (2008), Bradley et al., (2011),

Gilson et al., (2011), Bennett et al., (2011) and others in a recent series of articles on

health systems research. Bradley et al. (2011) suggest that renewed interest in health

systems research from WHO and other funding agencies is related to the shift from an

„international health‟ to a broader „global health‟ perspective, a shift that has been

accompanied by involvement of a wider range of disciplines in health research, including

anthropology, sociology and psychology. Gilson et al (2011, 1) argue that: “As health

policies and systems are themselves social and political constructions, it is important to

acknowledge the particular value of social science perspectives in the field.” These

authors also argue for more systematic and rigorous approaches to social science research

into health systems, including the development of case study methodologies linking

processes to outcomes across different settings (Mills et al., 2008). This paper is intended

as a first step in moving this agenda forward in relation to P4P.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

27

References

Andersen, L. B. (2009). What determines the behaviour and performance of health

professionals? Public service motivation, professional norms and/or economic incentives.

International Review of Administrative Sciences, 75(1), 79-97.

Balabanova, D., McKee, M., Mills, A., Walt, G., & Haines, A. (2010). What can global

health institutions do to help strengthen health systems in low income countries? Health

Research Policy and Systems / BioMed Central, 8, 22.

Basinga, P., Gertler, P., Binagwaho, A., Soucat, A., Sturdy, J., & Vermeersch, C.

(2011a). Effect on maternal and child health services in Rwanda of payment to primary

health-care providers for performance: an impact evaluation. The Lancet 377(9775),

1421-1428.

Basinga, P., Mayaka, S., & Condo, J. (2011b). Performance-based financing: the need for

more research. Bulletin of the World Health Organization 89(2), 698-699.

Bhattacharyya, K., Winch, P., LeBan, K., & Tien, M. (2001). Community Health Worker

Incentives and Disincentives: How they affect Motivation, Retention and Sustainability.

Institutionalizing Child Survival Project (BASICS II), USAID, 1-68.

Bennett, S., Dakpallah, G., Garner, P., Gilson, L., Aramphong, S. N., Zurita, B., & Zwi,

A. (1994). Carrot and stick: state mechanisms to influence private provider behaviour.

Health Policy and Planning 9(1), 1-13.

Bennett, S., Agyepong, I. A., Sheikh, K., Hanson, K., Ssengooba, F., & Gilson, L.

(2011). Building the field of health policy and systems research: an agenda for action.

PLoS medicine, 8(8), 1-5.

Bourdieu, P. (1977). Outline of a Theory of Practice. Cambridge. Cambridge University

Press (CUP).

Bourdieu, P. (1986a). The Logic of Practice. Stanford, California: Stanford University

Press.

Bourdieu, P. (1986b). The Forms of Capital. In J. E. Richardson (Ed.), Handbook of

Theory of Research for the Sociology of Education (p. 46-58). Greenwood.

Bowman, J. (2010). The Success of Failure: The Paradox of Performance Pay. Review of

Public Personnel Administration 30(1), 70-88.

Bradley, E., Fennell, M., Pallas, S., Berman, P., Shortell, S., & Curry, L. (2011) Health

Services Research and Global Health. Health Services Research 46(6) Part II 2019-2028.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

28

Castro, A. & Singer, M. (2004). Unhealthy Health Policy: A Critical Anthropological

Examination. Walnut Creek, CA. Altamira Press.

Deci, E. L., & Ryan, R. M. (2012). The “ What ” and “ Why ” of Goal Pursuits : Human

Needs and the Self-Determination of Behavior. Psychological Inquiry, 11(4), 227-268.

Dieleman, M., Toonen, J., Touré, H., & Martineau, T. (2006). The match between

motivation and performance management of health sector workers in Mali. Human

resources for health 4(2), 1-7.

Eichler, R., Auxila, P., Antoine, U., & Desmangles, B. (2007). Performance-Based

Incentives for Health : Six Years of Results from Supply-Side Programs in Haiti. Center

for Global Development.

Eichler, R., & Levine, R. (2009). Performance Incentives for Global Health: Potentials

and Pitfalls. Centre for Global Development.

Eijkenaar, F. (2012). Pay for Performance in Health Care: An International Overview of

Initiatives. Medical care research and review : MCRR

Eldridge, C., & Palmer, N. (2009). Performance-based payment: some reflections on the

discourse, evidence and unanswered questions. Health policy and planning, 24(3), 160-

166.

Faguet, J-P., & Ali, Z. (2009). Making Reform Work: Institutions, Dispositions, and the

Improving Health of Bangladesh. World Development, 37(1), 208-218.

Foucault, M. (2008). The Birth of Biopolitics. Palgrave Macmillan.

Franco, L. M., Bennett, S., & Kanfer, R. (2002). Health sector reform and public sector

health worker motivation: a conceptual framework. Social science & medicine 54 (8),

1255-66.

Frenk, J. (2010). The global health system: strengthening national health systems as the

next step for global progress. PLoS medicine 7(1), 1-3.

George, A. (2009). “ By papers and pens, you can only do so much”: views about

accountability and human resource management from Indian government health

administrators and workers. International Journal of Health Planning and Management,

24(April), 205-224.

Georgellis, Y., Iossa, E. & Tabvuma, V. (2010). Crowding Out Intrinsic Motivation in

the Public Sector. Journal of Public Administration Research and Theory, 21(3) 473-493.

Gilson, L. (2003). Trust and the development of health care as a social institution. Social

science & medicine 56(7), 1453-68.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

29

Gilson, L., Palmer, N., & Schneider, H. (2005). Trust and health worker performance:

exploring a conceptual framework using South African evidence. Social science &

medicine, 61(7), 1418-29.

Gilson, L., Hanson, K., Sheikh, K., Agyepong, I. A., & Ssengooba, F. (2011). Building

the Field of Health Policy and Systems Research : Social Science Matters. PLoS

Medicine, 8(8).

Greener, I. (2002). Agency, social theory and social policy. Critical Social Policy, 22(4),

688-705.

Hahn, R. & Inhorn, M. Eds. (2009). Anthropology and Public Health: Bridging

Differences in Culture and Society (2nd ed.). Oxford, UK: Oxford University Press.

Hartmann, B. (2011). The Return of Population Control : Incentives , Targets and the

Backlash against Cairo. Population and Development Program, (70).

Ireland, M., Paul, E., & Dujardin, B. (2011). Can performance-based financing be used to

reform health systems in developing countries? Bulletin of the World Health

Organization 89(9), 695-8.

Janes, C., & Corbett, K. (2011). Global Health. In M. Singer and P. Erikson (Eds.), A

Companion to Medical Anthropology (pp. 135-157). Malden, MA: Wiley-Blackwell

Kalk, A., Paul, F.A., & Grabosch, E. (2010). „Paying for performance‟ in Rwanda: does it

pay off? Tropical medicine & international health 15(2), 182-90.

Lagarde, M., Haines, A., & Palmer, N. (2007). Conditional cash transfers for improving

uptake of health interventions in low and middle income countries: a systematic review.

JAMA 298(16), 1900-10.

Macq, J., & Chiem, J.C. (2011). Looking at the effects of performance-based financing

through a complex adaptive systems lens. Bulletin of the World Health Organization

89(2), 699.

Marteau, T., Ashcroft, R., & Oliver, A. (2009). Using financial incentives to achieve

healthy behaviour. BMJ 338, 983-985

Mathauer, I., and Imhoff, I. (2006). Health worker motivation in Africa: the role of non-

financial incentives and human resource management tools. Human resources for health

4(24), 1-17.

McDonald, R., & Roland, M. (2009). Pay for Performance in Primary Care in England

and California : Comparison of Unintended Consequences. Annals Of Family Medicine

7(2), 121-127.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

30

Mehrotra, A., Damberg, C., Sorbero, M., & Teleki, S. (2009). Pay for Performance in the

Hospital Setting: What is the State of the Evidence? American Journal of Medical

Quality, 24(1), 19-28.

Meessen, B., Soucaf, A., & Sekabaraga, C. (2011). Performance-based financing: just a

donor fad or a catalyst towards comprehensive health-care reform? Bulletin of the World

Health Organization 89: 153-156.

Mills, A., Gilson, L., Hanson, K., Palmer, N., & Lagarde, M. (2008). What do we mean

by rigorous health-systems research? Lancet, 372(9649), 1527-9.

Mills, A. (2011). Health policy and systems research: defining the terrain; identifying the

methods. Health policy and planning. February 2011.

Montagu, D., & Yamey, G. (2011). Pay-for-performance and the Millennium

Development Goals. Lancet 377(9775), 1383-1385.

Mosse, D. (2005). Cultivating Development: An Ethnography of Aid Policy and Practice.

London: Pluto Press.

Nichter, M. (1983). Paying for what ails you: sociocultural issues influencing the ways

and means of therapy payment in South India. Social science & medicine 17(14), 957-65.

Nichter, M. (1986). The primary health center as a social system: PHC, social status, and

the issue of team-work in South Asia. Social science & medicine 23(4), 347-55.

Nichter, M. (1999). Project Community Diagnosis: Participatory Research as a First Step

toward Community Involvement in Primary Health Care. In R. Hahn (Ed.), Anthropology

in Public Health: Bridging Differences in Culture and Society (pp. 300-324.) Oxford:

Oxford University Press.

Nichter, M., Quintero, G., Nichter, M., Mock, J., & Shakib, S. (2004). Qualitative

Research: Contributions to the Study of Drug Use, Drug Abuse, and Drug Use(r)-Related

Interventions. Substance Use & Misuse, 39(10-12), 1907-1969.

Nichter, M. (2008). Global Health: why cultural perceptions, social representations, and

biopolitics matter. Tucson: University of Arizona Press.

Oldani, M.J. (2010). Assessing the “relative value” of diabetic patients treated through an

incentivized, corporate compliance model. Anthropology & medicine 17(2), 215-28.

Oxman, A.D., & Fretheim, A. (2009). Can paying for results help to achieve the

Millennium Development Goals? Overview of the effectiveness of results-based

financing. Journal of evidence-based medicine 2(2), 70-83.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

31

Paul, F.A. (2009). A Qualitative Study on Health Worker Motivation and the Rwandan

Performance Based Finance Initiative in District Hospitals. LSE IDS Working Paper

series 44(8).

Perry, J.L., Engbers, T.A., & Jun, S.Y. (2009). Back to the Future? Performance-Related

Pay, Empirical Research, and the Perils of Persistence. Public Administration Review

69(1), 39-51.

Pfeiffer, J. (2004). International NGOs in the Mozambique Health Sector: The "Velvet

Glove" of Privatization. In A, Castro and M. Singer (Eds.), Unhealthy Health Policy: A

Critical Anthropological Examination. Walnut Creek, CA. Altamira Press.

Pfeiffer, J., & Nichter, M. (2008). What Can Critical Medical Anthropology Contribute to

Global Health? Medical Anthropology Quarterly 22(4). 410-415

Powell-Jackson, T., Morrison, J., Tiwari, S., Neupane, B.D., & Costello, A.M. (2009).

The experiences of districts in implementing a national incentive programme to promote

safe delivery in Nepal. BMC health services research 9(97), 1-17.

Power, M. (1997). The Audit Society: Rituals of Verification. Oxford University Press

(OUP)

Rayner, J., Williams, H.M., Lawton. A., & Allinson, C.W. (2010). Public Service Ethos:

Developing a Generic Measure. Journal of Public Administration Research and Theory

21(1), 27-51.

Roberts, M., Hsiao, W., Berman, P., & Reich, M. (2004). Getting Health Reform Right: a

Guide to Improving Performance and Equity. Oxford. OUP.

Rusa, L., Ngirabega, J.D., Janssen, W., Van Bastelaere, S., Porignon, D., &

Vandenbulcke, W. (2009). Performance-based financing for better quality of services in

Rwandan health centres: 3-year experience. Tropical medicine & international health

14(7), 830-837.

Shenkin, M., & Coulson, A.B. (2007). Accountability through activism: learning from

Bourdieu. Accounting, Auditing & Accountability Journal 20(2), 297-317.

Soeters, R., & Griffiths, F. (2003). Improving government health services through

contract management: a case from Cambodia. Health policy and planning 18(1), 74-83.

Soeters, R., Habineza, C., & Peerenboom, P.B. (2006). Performance-based financing and

changing the district health system: experience from Rwanda. Bulletin of the World

Health Organization 84(11), 884-9.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

32

Soeters, R., & Vroeg, P. (2011). Why there is so much enthusiasm for performance based

financing, particularly in developing countries. Bulletin of the World Health Organization

89(9):700

Toonen, J., Canavan, A., & Vergeer, P. (2009). Learning Lessons on Implementing

Performance Based Financing, from a Multi-Country Evaluation. KIT

Travis, P., Bennett, S., Haines, A., Pang, T., Bhutta, Z., Hyder, A., Pielemeier, N.R.,

Mills, A., & Evans, T. (2004). Overcoming health-systems constraints to achieve the

Millennium Development Goals. Lancet 364(9437), 900-6.

Ulin, P., Robinson, E., & Tolley, E. (2005). Qualitative methods in public health : a

field guide for applied research San Francisco, CA: Jossey-Bass.

Weiss, G. (2008). Can an Old Dog Learn New Tricks? Habitual Horizons in James,

Bourdieu, and Merleau-Ponty. In G. Weiss (Ed.), Intertwinings (pp. 223-240). New York:

State University of New York Press.

Yip, W.C.M., Hsiao, W., Meng, Q., Chen, W., & Sun, X. (2010). Realignment of

incentives for health-care providers in China. The Lancet 375(9720), 1120-1130.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

REVIEWER COMMENTS

Reviewer #1:

The authors use the term 'neo-liberal' as if the meaning is self-explanatory and uncontested.

p.3 ‘neoliberalism’ is defined with a reference:

“The use of market mechanisms to achieve these short term targets has appeared to make sense

within the broader context of market based policy reforms, often termed ‘neoliberal’, which assume

markets to be the most efficient and effective mechanism for allocating resources in the public as

well as private sectors (Foucault, 2008).”

Page 3, Line 53 - 'compared with alternatives' - such as?

p.4 phrase deleted due to lack of space to address alternatives, which are only obliquely referred to

in the referenced literature, and which we refer to later in the paper. Alternatives could include:

increase in base salary; non-material incentives including training; changes in job descriptions,

reporting arrangements, career prospects; increase in resources such as medicines, equipment,

buildings; etc

Page 4, Line 12 - 'increasingly viewed' - by whom?

P.4 by policy makers. In the referenced paper the implication is that global health practitioners and

social scientists also need to account for the trade and security implications of health, but the point

here is that these connections have led to increased aid – this impetus comes from policy makers.

“Global health is increasingly viewed by policy makers as an economic and security issue

encouraging an ‘avalanche of aid’ (Pfeiffer and Nichter, 2008),”

Page 5, Line 7 - "increasingly placed situated' - typo?

p.5 Yes, a typo, but this whole phrase has now been deleted.

Page 7, Lines 53-56 - randomised control trial - randomised controlled trial?

Another typo - thanks

p.7 “A recent evaluation by Basinga et al. (2011a) isolated the P4P effect through a randomized

controlled trial.”

Page 8, Lines 9,10 - how are 'intrinsic motivations' undermined?

p.8 this para has been deleted in order to cut down the number of words and avoid repetition. The

phrase referred to has been removed.

Page 8, Paragraph 2 - it is not clear how the 'gap between indicators and overall performance" is an

'unintended effect'. The 'simpler measures of performance' do not reflect system performance, but

how are they an 'unintended effect'?

p.8 the text referred to has been removed. Part of the paragraph has been moved to 3.2 p.11.

*Response to Review*

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

“Concern has been raised that P4P, by encouraging motivation for financial rewards, may lead to the

“crowding out” of other forms of motivation. For example, those motivated by public service values

may find the job less rather than more attractive, believing their image will be spoiled by the higher

monetary rewards (Georgellis et al., 2010).”

I hope this gives sufficient explanation of how intrinsic motivation might be undermined by P4P.

The section sub-headed 'Can do', 'Will do' and the generation of 'double binds' presents a set of

hypotheses that could be further explored.

We have been unable to expand this section substantially due to word count limitations. However,

we have added some text as follows:

The following paragraph concerning double binds

p.11 “P4P schemes may generate ‘double binds’ affecting motivation as when time taken to

complete paperwork required for P4P conflicts with time attending to patients (Paul, 2009, Kalk et

al., 2010, Ireland et al., 2011), or when receiving a reward upsets working relationships with others

(Powell-Jackson et al., 2009).”

Has been expanded to include:

p.11 “P4P schemes may add to an already heavy load of paperwork, perceived to be unnecessary or

burdensome. New monitoring systems may also interfere with informal aspects of supervision

negotiated between supervisors and subordinates, which are seen as necessary in order for the

formal systems to function (George 2004). “

Measuring performance: P4P as a 'fatal remedy'

Page 12 Lines 47-49 - potentially "simultaneously affects."

p.13 this phrase now reads:

“‘Gaming’ and ‘crowding out’ reveal how measuring and rewarding selected behaviours can

potentially affect behaviours that are not measured or rewarded.”

Page 12 Lines 47-52 - is there a conflation here of the effects of rewards with the effects of

measurement?

Note: Perhaps – thanks for drawing our attention to this. We have tried to clarify what we see as the

relationship between the two. P4P schemes generally involve the introduction of both new rewards

and new measurements, with rewards linked to selected indicators which are then measured in new

ways – even if they were already measured previously.

p.11 Added:

“Crowding out and gaming are related because the rewards which potentially lead to crowding out

are linked to measured targets which potentially generate gaming behaviour.”

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

p.13 Edited:

“‘Gaming’ and ‘crowding out’ reveal how measuring and rewarding selected behaviours can

potentially affect behaviours that are not measured or rewarded. Measurement transforms the

operations of an organization, leading to a reallocation of resources, or colonization (Power, 1997).

On the other hand, bonuses re-align behaviour towards behaviours which are measured.”

Page 14 Line 41 - using anthropology to 'predict' performance is a bold claim.

This paragraph has been deleted.

Page 14 - the authors offer seeing health systems as a 'social system' as a way of gaining theoretical

purchase, but the illustration of how this can help is rather vague, coming down to the differing

impact of incentives on individuals fulfilling different roles.

This section has been deleted, with some of the material integrated into other sections. The

reference to Nichter (1986) has been used on p.15 in relation to Bourdieu.

p.15-16 “Similarly, Nichter (1986, 1999) has documented how community participation in primary

health care programs in the 1970s was based on unrealistic expectations about the existing habitus

of primary health care centers, and that this was one reason for the failure of these programs to

impact the poor.”

Page 16 - similarly the section on utility of 'theories of human performance' (what social theories

couldn't be caught by that phrase?) is vague and the quote from the Indonesian midwife not that

helpful in showing the utility of the 'framework'.

This section has been deleted. The reference to David Mosse’s work has been integrated into the

new Section 4. Using social science theory to understand responses to P4P.

p. 15 “In an ethnography of a participatory development project in India, Mosse (2005) describes

how ‘participation’ was translated into existing routines and agendas. The policy idea underwent

different transformations within the bureaucracy of the government civil service and within a private

company contracted by the project, leading to outcomes which were not always intended by policy

makers.”

Page 17, paragraph 2 - is the notion of a policy idea (great in analysing at the policy making level)

useful at the focus of how P4P plays out in a "particular locality"?

The paragraph referred to has been deleted.

Page 18 - the offer of 'rewards system' as a useful theoretical tool is again vague, and amounts really

to the rather obvious idea that both material and non-material rewards are important.

The term ‘rewards system’ is used in the context of Bourdieu’s theoretical framework rather than as

a stand alone or original concept.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

p. 19 “Bourdieu’s concept of convertible capital is useful for understanding how P4P might impact

motivation differently depending on the structure of local social fields and reward systems.”

Pages 19 - 20 The discussion here briefly covers ideas from Bourdieu to Foucault and while there is

no denying that these authors and others provide many useful concepts, taken altogether as the

authors have presented them these do not add up to a coherent 'theoretical lens'

We have tried to provide a more coherent theoretical lens by focusing on Bourdieu as a useful

starting point, then pointing out the strengths and limitations of his theory in the context of P4P and

how anthropological studies have highlighted both its strengths and limitations. Reference to

Foucault has been deleted from this section.

p. 15 “Understanding responses to P4P and how they vary cross-culturally demands theory which

recognizes motivation as a social as well as individual phenomenon. The sociologist Pierre Bourdieu

offers a useful starting point.”

p.15 “Bourdieu regarded the habitus as durable and not susceptible to change, suggesting a

limitation in the applicability of his theory to contexts of rapid behavioural change (Weiss, 2008). But

others have explored the ways in which government policy can interact with the habitus, as well as

how assumptions concerning the habitus frame such policies.”

p. 19 “Bourdieu’s concept of convertible capital is useful for understanding how P4P might impact

motivation differently depending on the structure of local social fields and reward systems.

However, it may not capture all of the impacts of P4P on motivation. Bourdieu focuses on different

forms of social or material gain but does not address the issue of intrinsic motivation, not dependent

on external rewards.”

Discussion section - this section presents a number of methodological steps to take, essentially

exhorting researchers to ethnographically investigate local cultures and develop an understanding of

the local contexts to allow P4P to successfully adapted. Again there are number of possibly useful

ideas here but they don't appear to be organised into a integrated whole - more like a grab-bag of

goodies than a coherent approach. This may be more to do with the presentation than the content.

Maybe the authors are trying to do too much and in doing so risk simply enumerating potentially

useful concepts.

We have tried to clarify what we are trying to do, and relate our approach to a systematic formative

research approach. We have reorganized the sub-sections to reflect a series of stages from

situational analysis through monitoring to critical assessment of wider impacts. We highlight key

issues in need of research appreciated by two of the other reviewers --but we have endeavored in

this version to sharpen our focus and point to productive points of departure for anthropologists

wishing to contribute to the study of p4p.

p. 20 5 Where do we go from here?

“In this section we suggest how some of the insights emerging from this review can be operationalized in qualitative research aimed at critically assessing P4P schemes. Following Nichter et al (2004, 1914) we advocate a “cyclical formative-reformative research approach” involving a number of distinct but interactive stages, from initial situational analysis and problem identification

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

prior to implementation of a scheme, through monitoring and evaluation to critical assessment of wider impacts. “

Addtionally given the paper's aim and the nature of the intervetion they are looking at, there is not

much on the importance of accounting for the wider cultural political economic context in which

P4Ps are introduced.

We have added a section on Wider Impacts including anthropological contributions to

understanding them (p.19).

I would encourage the authors to review their paper and redraft it to communicate a theoretically

focused, coherent and consistent approach. I also think that the goal of 'prediction' might be given

some further thought.

As mentioned above we now offer a single theoretical framework drawing on Bourdieu. We have

removed reference to prediction.

Reviewer #2:

The overall goal of this contribution is to analyze the various dimensions of pay-for-performance

schemes in global context, offering both a critical literature review that identifies important gaps,

questions and concerns, and also providing theoretical and conceptual guidance for further

research. The authors argue that anthropological contributions to this research are needed.

This is an excellent paper. It provides a detailed, insightful, and critical analysis of a piece of health

policy that is wholly understudied. The literature review is up-to-date, though the authors might

want to include a brand new systematic review (available online) by Frank Eijkenaar, "Pay for

Performance: An International Review of Initiatives," Medical Care Research and Review, Feb. 2012,

DOI: 10.1177/1077558711432891. Eijkenaar's review generally offers support to the authors'

argument, and may also provide a few additional insights in/re: to the idea of "context" and

heterogeneity.

References to Eijkenaar’s review have been included as follows:

p.12 “Allocating rewards to groups or facilities might address some of these issues, although it could also introduce the risk of ‘free riding’ (Eijkenaar, 2012). “

p.21 “It has been suggested that the risk of crowding out can be reduced by aligning performance measures with existing professional norms and values which provide a foundation for intrinsic motivation (Eijenkaar, 2012). “

And also on p. 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

“Involving health workers and the communities they serve in the selection of indicators and targets might help reduce gaming and crowding out. (Paul, 2009, Toonen et al., 2009, Eijkenaar, 2012).”

The writing is crisp and straightforward, and will be fully accessible to a broad readership.

I particularly appreciate the efforts to focus anthropological/social theory on an important

dimension of health systems policy. Anthropologists, in particular, have tended to avoid the messy

world of health systems research, perhaps in part because of its complexity, but also in part because

of our ethnographic orientation to people in community. The authors are thus providing an

important service to the discipline: outlining a problem, analyzing its dimensions, and suggesting a

research agenda that is aligned with an anthropological methodological and theoretical orientation.

I would thus imagine that this paper would provide important guidance for those seeking to develop

research projects in this area.

Finally, I particularly appreciate the discussion of "crowding out" and the tendency for these kinds of

incentive systems to efface intrinsic motivation. I've heard this as well from health practitioners in a

number of settings. This is something that definitely requires further, in-depth research.

I have a few comments, quibbles and minor editorial points.

First, the authors argue for the applicability of anthropology. While, as an anthropologist, I agree

with this argument, it seems clear from the paper that the contribution is less disciplinary, than it is

theoretical, and the theory that the authors draw upon would fall more generally into the field of

general social theory; it is not an exclusive, innovative contribution of anthropology. Bourdieu is,

after all, a sociologist. I would recommend a few minor edits to indicate, perhaps, that value is not

necessarily to be drawn from a single discipline, but instead, that "insights from social theory, and

especially the focused ethnographic methods of anthropology" are the way to go in unpacking P4P.

The abstract has been edited as follows:

p.1 “We suggest that insights from social science theory coupled with the focused ethnographic

methods of anthropology can contribute to the critical assessment of P4P schemes and to their

adaptation to particular social environments and reward systems.”

The title for Section 4 now reads:

p.15 4. Using social science theory to understand responses to P4P

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Secondly, again a minor point, the authors may want to look a bit more closely at Bourdieu - his use

of the term "cultural capital" is somewhat broader than its usage in the paper. In the paper, cultural

capital is reduced primarily to social knowledge; i.e., from page 19: "learning how to operate in a

health facility involves acquiring 'cultural capital,' including not only medical knowledge, but also

knowledge of the medical hierarchy and how to relate to peers and supervisors." Certainly

knowledge is an important piece of cultural capital, but so are more symbolic forms (language,

manners, dress) that convey status less overtly. I don't necessarily disagree with the application of

the concept here, but a slightly more nuanced discussion is needed, I think. See in particular the

essay on the "Forms of Capital" [in J.E. Richardson, ed., Handbook of Theory of Research for the

Sociology of Education, 1986, Greenwood], where Bourdieu provides a fairly concise, but in-depth

discussion of the concept.

The section on Bourdieu has been expanded and edited. The paragraph relevant to Forms of Capital

now includes the suggested reference and reads as follows:

p. 16-17 “Bourdieu distinguished between different forms of capital including economic, cultural,

social and symbolic capital. Cultural capital signals competence to operate in a particular social field.

It includes educational qualifications, such as medical expertise, as well as embodied knowledge of

appropriate behaviour such as the use of correct manners and language in relation to peers and

supervisors. Social capital refers to the social networks that an individual can draw on for social,

economic or professional support. Symbolic capital refers to forms of recognition, such as

promotions, certificates or titles and to the prestige associated with one’s social position or

professional status. These forms of capital are acquired by individuals through investment of time

and labor, or in some cases through inheritance, in ways which are allowed and encouraged by the

habitus of their social group. The dynamics of the social field determine the extent to which

individuals can convert one form of capital to another (Bourdieu, 1986b).”

Thirdly, I would like to emphasize the authors' reference to the recent papers by Ann Mills, who has

argued for more systematic research on health systems, including development of appropriate case

study methodologies that can be used to link process to outcome in particular settings, but which

also can be compared across settings in service of a broader theory of health systems. It seems to

me that the authors are in fact offering such a framework, at least in the case of investigating P4P. It

might be worth a short discussion at the end of the paper: i.e., "we need better health systems

research; here is one way forward."

The Discussion section at the end of the paper includes reference to this and other papers and reads

as follows:

p.26 “This approach is in line with that advocated by Mills et al (2008), Bradley et al (2011), Gilson et al (2011), Bennet et al (2011) and others in a recent series of articles on health systems research. Bradley et al (2011) suggest that renewed interest in health systems research from WHO and other funding agencies is related to the shift from an ‘international health’ to a broader ‘global health’ perspective, a shift that has been accompanied by involvement of a wider range of disciplines in health research, including anthropology, sociology and psychology. Gilson et al (2011, 1) take up the theme of the role of social science arguing that: “As health policies and systems are themselves social and political constructions, it is important to acknowledge the particular value of social science perspectives in the field.” These authors also argue for more systematic and rigorous approaches to

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

social science research into health services and health systems, including the development of case study methodologies linking processes to outcomes across different settings (Mills et al, 2008). This paper is intended as a first step in moving this agenda forward in relation to P4P.

Finally, a very minor correction is needed: the sentence beginning on line 19, p. 7, needs to be fixed.

It currently reads: "In another positive evaluation, a P4P for public health providers in China helped

realign incentives away from a profit motive that had encouraged over treatment, towards patient

oriented care." I think there is a missing word.

The final phrase has been deleted.

p.7 “In another positive evaluation, a P4P for public health providers in China helped realign incentives away from a profit motive that had encouraged over treatment (Yip et al., 2010).

In summary, this is a strong paper, providing an excellent contribution to the literature on health

systems performance, and, most importantly, bringing the study of health systems to the attention

of medical anthropologists.

Reviewer #3:

This paper deals with an interesting and highly relevant topic within health systems research and

health policy in general - the good and the bad of using financial incentives to boost performance in

the health sector in developing countries. The topic is all the more relevant in the context of the

current 'neoliberal' global health discourse and funding mechanisms that are dominated by the

World Bank, the Global Fund, and new public health philanthropists like Bill and Melinda Gates and

Bill Clinton.

As the authors state, this is a think piece, or more accurately, a systematic review of the current

knowledge and issues surrounding Paying for Performance in the health sector, especially in

developing countries. The authors have done an excellent job of reviewing the literature that deals

with a number of key concerns (both positive and negative (including unintended effects) with the

P4P strategy in streamlining and improving health workers' performance and programmatic

outcomes. They cite examples from Rwanda, Haiti, Ghana and Cameroon to illustrate key points.

The authors conclude that health systems are best seen as changing social systems, where issues of

social hierarchy, social status and different forms of capital (a la Bourdieu) affect health workers'

morale and performance, and ultimately the sustainability of any externally-funded P4P schemes.

Therefore, anthropological theory, practice and ethnographic studies in particular, have an

important role in making

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

such schemes relevant to the local context. In sum, there is a lot that one can learn from reading this

paper. That said, I do have a few concerns that I hope the authors will address.

1. This paper belongs in the Health Systems Research domain and should be read as a systematic

review rather than a "think piece." The authors have cited Nichter's work and alluded to Pfeiffer's

work as well, but in my reading, this paper has little to do with anthropology, even very broadly

speaking, and definitely not medical anthropology. In its present format, this paper is unlikely to

attract the medical anthropology readership. Anthropology is a very broad discipline and it seems to

me that the authors are conflating anthropology with ethnography or ethnographies.

We have removed reference to this paper as a “think piece”. We present it as a review.

We have clarified the role of anthropology:

p.25-26 “Anthropology can provide a bridge between social theory and practical health policy work.

Anthropology’s strengths in this regard include the methodology of sustained participant-

observation in the research setting, the validation of local perspectives, and a tradition of drawing

on a wide range of theoretical foundations (Introduction, Hahn and Inhorn, 2009).”

2. If this paper were to be accepted for publication in Social Science and Medicine, as a first step, the

manuscript needs to be put on a strict diet; it is overly long, repetitive and has sections that deserve

a good bit of trimming so that the clutter is removed.

We have removed repetition through edits to and reorganization of Sections 3 , 4 and 5.

Although the authors have demonstrated that they have an excellent grasp of the most recent

literature on the subject of P4P, there is immense scope them to rewrite the paper in a more lucid

manner so that the paper attracts a wider audience. For a start, the practice of including a citation

after every sentence may be more appropriate for a "review paper" but not for a "think piece."I

personally found this style to be distracting and annoying when going over some sections, such as,

for example, page 17. There are four different ideas crammed in one paragraph. Readers could get

lost trying to figure out what this paper is all about, or what are the key ideas that are holding the

paper together!

The paragraph on p.17 has been deleted as part of the reorganization of Section 4, which now

presents a more focused and coherent theoretical framework drawing primarily on Bourdieu.

Since we now argue, following this reviewer’s advice, that this is a review, not a think piece, we have

not adopted a conscious policy of reducing the number of references. However, this has occurred as

a consequence of the reorganization which we have undertaken.

3. Section 5 discusses how ethnographies undertaken with stakeholders could contribute to the

design and monitoring of programs that are more sensitive to potential unintended effects on the

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

wider health system and over the long term" (page 5) In reading this, and the paper's title, I was

hoping to see some ethnographic data, or a case study, but alas... I suggest the authors remove the

"What can Anthropology Contribute?" tag question from the title and rephrase it as How can

Anthropology Contribute - and then go on to actually spell out with examples (other than Nichter's

dated work conducted in the context of India's primary health care program in the 1970s), the 'how'

of anthropology's contribution, rather than simply stating the obvious that detailed ethnographies

are needed to demonstrate how local reward systems are enmeshed in local social networks, and so

forth. How can ethnographic research/studies contribute to a better understanding of "how these

process work, and

how they vary across social and cultural contexts?" If the authors can grapple with this question,

that would be their original contribution in this paper. Ideally, this paper should have had a separate

section on empirical data highlighting the performance of performance or in Foucault's words, the

conduct of conduct.

We have changed the title which now reads:

Paying for Performance and the social relations of health care provision: an anthropological

perspective

We feel that this title is more descriptive of the paper, without making unrealistic claims.

We have changed the sentence referred to on p.5 which now reads:

“Section 5 suggests how theoretically informed anthropological research can help

critically assess P4P and adapt schemes to particular settings.”

The ‘how’ of anthropology’s contribution is illustrated by references incorporated into various

sections, for example George p.11, Mosse p.15, Nichter, p.15, Pfeiffer p.19,

We are unable to provide case study ethographies of P4P beyond those referenced in Section 2 such

as the qualitative study of the Rwanda scheme by Paul (2009). The purpose of this paper is to

highlight the need for more ethnograhies of P4P and to point to issues which could provide a focus

for such case studies.

4. Finally, the authors seem to suggest that detailed ethnographies are some sort of a panacea to

deal with the problems associated with P4P; most ethnographers will disagree, given that all

ethnographies (or ethnographic research in general) are partial, committed and constructed,

depending on the ethnographer's positionality, theoretical orientation, communicative competence,

and so forth. Ethnographic research can highlight what works in one context and what does not and

why. The suggestion that "P4P should always be adapted to particular contexts in consultation with

local stakeholders" is a sincere proposition, but in reality there is no guarantee that P4P will always

work and will have no unintended effects just because you have consulted local stakeholders.

Anthropologists and health economists have amply demonstrated the power of 'corruption' in

health programs even in contexts where local stakeholders have been consulted.

We certainly recognize that ethnographies are no panacea but suggest that they provide a starting

point for critically assessing impacts of P4P on social relations of health care provision which have

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

generally been missed by policy makers and reseachers perhaps partly due to their disciplinary

orientation.

5. Minor points:

Remove the Eldridge and Palmer (2009) reference at the end of the abstract on page 2.

[EDITOR'S NOTE - Disregard; any quotations in the Abstract must be referenced immediately below

it, as you have done]

The section on The Performance of Performance:

If this is a think piece, what is the quote (empirical data) on page 16 doing? "As one midwife

explained (AUTHOR, .) "Even if someone ..we have to visit at her house"?

This has been deleted.

6. The authors may find these two sources useful in reframing the conclusion.

1. Pfeiffer, James. International NGOs in the Mozambique Health Sector: The "Velvet Glove" of

Privatization. In. Arachu Castro and Merill Singer (eds) Unhealthy Health Policy: A Critical

Anthropological Examination. Altamira. James .2004.

Thank you for this suggestion. We have included this reference on p. 19

“Pfeiffer’s (2004) ethnography of the impact of health policy in Mozambique illustrates one way in

which this can occur. Pfeiffer describes how international donor policy to support NGOs in health

service delivery in Mozambique not only undermined the public health system but also increased

socio-economic inequality within communities served by the NGOs. The recruitment and funding of

a new cadre of NGO staff drained the public service of skilled human resources while at the same

time creating a new elite enclave class.”

2. Introductory chapter from Anthropology and Public Health: Bridging Differences in Culture and

Society (2nd ed). Robert Hahn and Marcia Inhorn (eds). Oxford. 2009

Thank you for this suggestion. We have included this reference on p.25-26 as mentioned above:

“Anthropology can provide a bridge between social theory and practical health policy work.

Anthropology’s strengths in this regard include the methodology of sustained participant-

observation in the research setting, the validation of local perspectives, and a tradition of drawing

on a wide range of theoretical foundations (Introduction, Hahn and Inhorn, 2009).”

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Hope this helps.

Reviewer #4:

I think this is an excellent paper that provides both interesting and practical insights of significant

policy relevance - and also adds to understanding of the role of anthropology in health policy and

systems research

Reply

Reply to all

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Forward