Rebuilding health post-conflict: case studies,

reflections and a revised framework

Spencer Rutherford* and Shadi Saleh

Global Health Institute, American University of Beirut, Old Pharmacy Building, Room 202, PO Box 11-0236, Riad

El-Solh, Beirut, Lebanon

*Corresponding author. Global Health Institute, American University of Beirut, Old Pharmacy Building, Room 202, PO Box

11–0236, Riad El-Solh, Beirut, Lebanon. E-mail: spencer.rutherford52@gmail.com

Accepted on 21 February 2019

Abstract

War and conflict negatively impact all facets of a health system; services cease to function, resour-

ces become depleted and any semblance of governance is lost. Following cessation of conflict, the

rebuilding process includes a wide array of international and local actors. During this period, stake-

holders must contend with various trade-offs, including balancing sustainable outcomes with im-

mediate health needs, introducing health reform measures while also increasing local capacity,

and reconciling external assistance with indigenous legitimacy. Compounding these factors are

additional challenges, including co-ordination amongst stakeholders, the re-occurrence of conflict

and ulterior motives from donors and governments, to name a few. Due to these complexities, the

current literature on post-conflict health system development generally examines only one facet of

the health system, and only at one point in time. The health system as a whole, and its develop-

ment across a longer timeline, is rarely attended to. Given these considerations, the present article

aims to evaluate health system development in three post-conflict environments over a 12-year

timeline. Applying and adapting a framework from Waters et al. (2007, Rehabilitating Health

Systems in Post-Conflict Situations. WIDER Research Paper 2007/06. United Nations University.

http://hdl.handle.net/10419/63390, accessed 1 February 2018.), health policies and inputs from the

post-conflict periods of Afghanistan, Cambodia and Mozambique are assessed against health out-

puts and other measures. From these findings, we developed a revised framework, which is pre-

sented in this article. Overall, these findings contribute post-conflict health system development by

evaluating the process holistically and along a timeline, and can be of further use by healthcare

managers, policy-makers and other health professionals.

Keywords: Health system development, health system reconstruction, post-conflict, state-building, Cambodia, Afghanistan,

Mozambique

key Messages

• Applies a health system framework to three post-conflict countries.• Expands on the literature pertaining to health system development.• Presents a revised framework for post-conflict health system development.• Delineates sequencing of state-building activities across a 12-year period.

VC The Author(s) 2019. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved.

For permissions, please e-mail: journals.permissions@oup.com 230

Health Policy and Planning, 34, 2019, 230–245

doi: 10.1093/heapol/czz018

Advance Access Publication Date: 30 March 2019

Review

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

Introduction

War and conflict have changed dramatically over the course of the

last century; what was once a widespread, global occurrence has

now largely become concentrated to resource-poor countries in the

Global South (Waters et al., 2007; Brown et al., 2008; Kruk et al.,

2010). Indeed, of the 28 global conflicts currently being monitored

by the Council on Foreign Relations, all but one (the European refu-

gee crisis) are among low- and middle-income countries, with the

majority being concentrated to Sub-Saharan Africa and the Middle-

Eastern region (Council on Foreign Relations, 2018). While damag-

ing all facets of society, conflict inflicts particular devastation on the

health and health sector of these countries (Rubenstein, 2009;

Gordon, 2013). For example, a recent study from the World Bank

(The World Bank, 2017) estimating the economic damage caused by

the Syrian conflict found ‘more people may have been killed in Syria

due to a breakdown of the health system than due to direct fatalities

from the fighting’ (40). Therefore, investment and proper research

towards post-conflict healthcare reconstruction is imperative for

addressing these problems and establishing an effective response.

The present article will introduce and explore how conflict

impacts health and health systems and identify the main challenges

that must be overcome during the post-conflict period. Strategies

employed by international and local actors during the reconstruction

process are then evaluated across three post-conflict case studies,

and the strengths and weaknesses of the various approaches are

determined along a 12-year timeline. The findings from the case

studies are then presented in a revised framework for sector-wide

post-conflict development.

The conflict period: effects on health and health

systemsCivil and armed-conflict negatively impact both the health of the

population, and the health system of the region. The magnitude of

this devastation is extensive, with nearly all facets affected in some

capacity. Concerning the health of the population, morbidity and

mortality dramatically increase due to the immediate effects of con-

flict, and include (but are not limited to) war-related injuries, sexual

violence, forced labour, exploitation and unlawful detention

(Waters et al., 2007; ICRC, 2016). Additionally, the mass displace-

ment of populations during times of conflict not only intensifies

poor health conditions, but also serves to increase the prevalence of

communicable diseases such as malaria and respiratory infections

(Waters et al., 2007; Kruk et al., 2010). Finally, exposure to war-

related atrocities contributes to an increased prevalence of psycho-

logical disorders, including post-traumatic stress disorder, depres-

sion and anxiety (Panter-Brick, 2009; Abo-Hilal and Hoogstad,

2013).

Conflict has an equally destructive effect on the health system of

the region. Based on the World Health Organizations (WHO) ‘six

building blocks of a health system’ framework (WHO, 2007), all

modalities are negatively affected to some degree. As health profes-

sionals are increasingly being targeted during times of conflict, doc-

tors, nurses and pharmacists are fleeing conflict zones in rising

numbers (Fujita et al., 2011; Gordon, 2013; Fouad et al., 2017).

This not only serves to reduce the ‘health workforce’ of the region,

but it also prevents the proper provision of ‘service delivery’ follow-

ing the end of the conflict (Fujita et al., 2011; Gordon, 2013).

Additionally, the destruction of health infrastructure, including

health facilities (hospitals and clinics), monitoring systems and pub-

lic health institutions, decreases the amount of available ‘health in-

formation’ and the necessary ‘medical products, vaccines and

technologies’ required for a health system to function effectively

(The World Bank, 1998; Gordon, 2013). Finally, with the collapse

of ‘leadership and governance’, both state health policy and ‘financ-

ing mechanisms’ essentially become non-existent (Gordon, 2013).

The post-conflict period: definition and delineationDue to the unpredictable nature of post-conflict zones, a consider-

ation of the time-frame is paramount. Yet, the idea of ‘post-conflict’

is not easy to define (Haughton, 1998; Waters et al., 2007;

Ohiorhenuan and Stewart, 2008). This is especially true in modern

warfare, where conflicts are typically intra-state, protracted over

years, if not decades, and have multiple groups entering and exiting

the conflict at various points in time (Haughton, 1998;

Ohiorhenuan and Stewart, 2008). For general purposes, however,

the post-conflict period is usually defined by the ceasing of hostil-

ities, signing of a peace agreement or the demobilization, disarma-

ment and reintegration of forces (Ohiorhenuan and Stewart, 2008).

Moving into the post-conflict period, delineation of the rebuild-

ing process becomes somewhat vague and malleable. The first

year following successful conflict resolution is usually termed ‘tran-

sitional’, in that it involves the presence of an interim government

and the continuation of emergency health services (Haughton,

1998). With the establishment of a more permanent government

and increased local capacity, the next few years generally focus on

health sector ‘rehabilitation’ and ‘reconstruction’ (Haughton, 1998;

Csis, 2002). This eventually leads to ‘sustainable development’, and

at this point, the health sector should be indistinguishable from

those of other developing countries, irrespective of the damage from

the conflict (Haughton, 1998).

The end goal of this process is to return the health sector to a

pre-war functioning capacity (Haughton, 1998; The World Bank,

1998; Waters et al., 2007). However, this process can take years, if

not decades, and is generally delayed by unforeseen set-backs and

challenges (Haughton, 1998; The World Bank, 1998;

Ohiorhenuan and Stewart, 2008). Indeed, an analysis from

Haughton (1998) found that out of 15 countries experiencing civil

war prior to 1991, only 1 country achieved pre-war peak levels of

gross domestic product (GDP) per capita within a decade after

conflict. Therefore, when planning for the rebuilding process,

organizations should anticipate a slow recovery, and sequence

their activities accordingly.

The post-conflict period: development and

reconstructionGiven the challenges related to the timeline of post-conflict health

system development, what is the actual process, and what are some

of the main challenges? Concerning the process, there are essentially

two options for the health sector. The first is to simply maintain

the emergency relief programmes provided during the war; however,

this strategy is generally considered to be unsustainable, and further

risks undermining restoration of the national health sector

(Rubenstein, 2009). The second, more viable option, is the construc-

tion or reconstruction of the pre-conflict health system (Rubenstein,

2009).

Nonetheless, the ubiquitous destruction previously outlined

poses unique and significant challenges. Factors such as displace-

ment, trauma and political volatility contribute to problems not nor-

mally seen in regular development strategies (Waters et al., 2007;

Kruk et al., 2010). Additionally, the scale of reconstruction requires

significant donor input, and further necessitates a multi-faceted

rebuilding process that the state cannot support in isolation

Health Policy and Planning, 2019, Vol. 34, No. 3 231

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

(Gordon, 2013; Kruk et al., 2010). Consequently, there are a multi-

tude of international and local actors involved in the process, who

must act in concert to address an array of initiatives, including; the

provision of health services; the establishment of a disease preven-

tion and surveillance system; supply-chain management of pharma-

ceuticals, vaccines and medical equipment; and monitoring,

evaluations and financing of the health system, to name a few

(Rubenstein, 2009; Kruk et al., 2010).

Undermining all these challenges is a series of seemingly impos-

sible paradoxes that must be reconciled throughout the development

process. Health system programmes must be both not only effective

enough to address immediate health needs, but also efficient enough

to function in the long run (Rubenstein, 2009; USAID, 2009).

Balancing these short- and long-term goals is also divisive, as stake-

holders must attend to immediate health needs while also laying the

groundwork for addressing long-term health complications that

arise from conflict, such as mental disorders and physical rehabilita-

tion (Haughton, 1998; USAID, 2009). This raises additional debate

as to the level of reform measures that should be implemented, as

the post-conflict environment is sometimes seen as a ‘window of op-

portunity’ for reform (Haughton, 1998; USAID, 2009). However,

too many reform measures also risk overwhelming the already se-

verely weakened health sector. Further complicating matters are

issues related to ownership of the health system, as the international

community will generally assume responsibility for the health sys-

tem initially, as the state will not have the capacity to implement the

required measures (Haughton, 1998; USAID, 2009; Kruk et al.,

2010). Nevertheless, without proper attention given to the eventual

transfer of this ownership towards local institutions, the whole en-

deavour risks continued reliance on foreign aid, and may ultimately

prove unsustainable.

A final, key consideration is the necessity for contextualization,

as each post-conflict period will be unique, exposing specific geo-

graphic, demographic and political problems (Waters et al., 2007;

Brown et al., 2008; Kruk et al., 2010). While some of these context-

ual problems are more prevalent than others, such as gender dispar-

ities, refugee resettlement and the re-emergence of communicable

diseases (Waters et al., 2007; Brown et al., 2008; Rubenstein, 2009),

others are much more unique, such as the American military in-

volvement in Afghanistan (Gordon, 2014), the influence of

European, free-market ideologies in Kosovo (Shuey et al., 2003) and

persistent sectarian tensions in Lebanon (Traboulsi, 2012).

However, because integrating these contextual factors requires in-

sight from local organizations and time on behalf of the develop-

ment community, they are frequently overlooked, thereby

aggravating the situation.

Purpose of the present paperGiven these considerations, two overarching themes emerge regarding

post-conflict health system development, namely; it is a long process

that will take at least a decade, and; this process will require inputs at

all levels from multiple stakeholders. However, the literature adheres

disproportionately towards analyses that examine just one aspect of the

development process over the course of just 1 or 2years. While this ap-

proach is valuable in that it provides nuance and detail, it also silences

the long-term effects of both war and the actions taken by the inter-

national community, as well as the interconnectedness and multi-

sectorial nature of the development process.

Therefore, the present paper will attempt to address these gaps

by evaluating post-conflict health system development across a 12-

year timeline, and at all levels of the health system. This 12-year

timeline was chosen somewhat arbitrarily as the length of time

required for state-building is generally considered to be ‘at least a

decade’ (Haughton, 1998). By doing this, the aim is to understand

which areas of the health system should be prioritized at which

times following conflict, and ultimately, how various policies and

actions should be sequenced along the development process.

Drawing on three case studies, insights from academic publications,

government policies and demographic data are synthesized to create

a revised framework to demarcate the most effective and efficient

practices to be taken by the development community when rebuild-

ing health systems post-conflict.

Methodology

Selection of countries for analysisWhile there is a tendency to homogenize post-conflict countries, in real-

ity, they differ in many aspects, and require different policies to properly

contextualize each situation (Brown et al., 2008). Therefore, case studies

were selected based on economic similarities, using a typology of post-

conflict environments as posited by Brown et al. (2008). In brief, the

authors determined three major economic criteria as the basis for this cat-

egorization; the state of the economy, mainly regarding its level of devel-

opment (using the Human Development Index and GDP); the presence

of high-value natural resources; and the extent of horizontal inequalities

(both defined and measured previously by Brown et al. (2008). Using a

list of post-conflict countries from Ohiorhenuan and Stewart (2008: 7),

those with a low level of economic development, an absence of a high-

value natural resources and sharp horizontal inequities were identified.

These characteristics were selected as they are the slowest kinds of coun-

tries to recover from conflict, and therefore require the greatest amount

of external assistance post-conflict (Brown et al., 2008). From this,

Cambodia (1992–2003), Afghanistan (2001–2012) and Mozambique

(1993–2004) were selected, as all countries additionally experienced an

extremely protracted, civil conflict period, with a high number of civilian

deaths, and significant, post-conflict foreign intervention.

Data collectionThis was a non-systematic review that did not employ any inter-

views or observations to collect data. PubMed, Embase and Google

Scholar were used to search for published literature, while Google

and the WHO database were used to search for grey literature; add-

itionally, publications from the reference lists of identified literature

were also used. Search terms such as ‘health system’, ‘post-conflict’

and the names of countries chosen as the case studies were used as

search terms.

Studies which did not pertain to the health system specifically,

fell outside of the time-period chosen for analysis, or did not contain

information which fell into the framework chosen for analysis were

not included. While recognizing that varying observations and con-

clusions will be drawn from studies in post-conflict countries, neu-

trality and biases were minimized by referring back to the original

policy document when possible, including findings that were only

supported by evidence, and presenting discrepancies openly in the

case studies. Overall, case studies were built from a total of 33 publi-

cations; 15 for Cambodia, 13 for Afghanistan and 8 for

Mozambique (see Supplementary Data).

Framework for analysis and measuresTo systematically describe the development process, a framework

developed by Waters et al. (2007; Figure 1) was applied at three

‘segments’ of the development process. The framework outlines two

232 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

overarching needs for the rehabilitation of health systems in post-

conflict environments: policy issues and health system inputs.

Specifically, the authors identified five principal policy issues related

to rehabilitation of the health system, namely; co-ordination among

donors (such as sharing information and system-wide agreements);

political commitment by host governments (such as the presence of

a national strategy or policy); partnerships with non-governmental

organizations (NGOs; such as formalized agreements or effective

co-ordination); planning, prioritization and integration of health

services (again via national strategies and policies); and the sustain-

ability of the rehabilitation effort (the level of dependence on inter-

national assistance). Assessment of these policy issues were based on

definitions more extensively described by Waters et al. (2007). All

measures were assessed qualitatively from published studies and

grey literature, including government documents, organization

reports and demographic profiles.

Regarding health system inputs, the framework outlines a further

five needs, namely; financing, human resources, physical infrastruc-

ture, information systems and essential drugs. For financing, three

measures were evaluated using World Bank data; the percent of ex-

ternal financing directed towards the health sector; health expend-

iture per capita; and health expenditure as a percent of GDP.

Human resources, physical infrastructure and information systems

were all assessed qualitatively, using published and grey literature

(government documents, reports, policies, etc.). For essential medi-

cines, the average immunization rates of four commonly adminis-

tered vaccines, DTP3 (diphtheria, tetanus toxoid and pertussis),

Pol3 (polio), MCV1 (measles) and BCG (tuberculosis) were assessed

using WHO data. While the supply of vaccines is much simpler than

that of essential medicines (Foster et al., 2006), this measure was

chosen as no reliable data exists for the distribution of essential med-

icines during time periods evaluated. Furthermore, effective vaccin-

ation coverage still requires significant input from governments and

donors and is commonly used in the literature as an indicator of

supply-chain management in health systems research (Foster et al.,

2006; Perehudoff et al., 2018).

Independent of this framework, quality of governance was

chosen as an additional measurement, which included accountabil-

ity, monitoring and evaluation, and prevention of corruption. These

measures were chosen independently by the authors based on their

experience and knowledge of this field. Accountability and monitor-

ing of the health system was evaluated by clear responsibilities and

roles among stakeholders and the presence of health information

systems and monitoring and evaluation networks, respectively.

Corruption was defined as per the extent to which public power is

exercised for private gain (The World Bank, 2018) and measured

via World Bank data (Figure 2). All measures were assessed against

three commonly applied health outputs, specifically; average life ex-

pectancy, infant mortality, and under-five mortality (Wang et al.,

1999), again, using World Bank data. In order to gain a complete

picture of the development process and not just a snapshot at one

point in time, all measures were evaluated over a 12-year period,

and segmented into three distinct time-frames (years 1–4, years 5–8

and years 9–12).

LimitationsThere are a few notable limitations in this study. First, all of the

case studies are based off secondary data sources and findings

from the literature (some of which were in-turn based off of sec-

ondary data sources); consequently, they are quite prone to biases

and the overall quality of evidence is low. Secondly, as this is a

non-systematic review, it does not contain the methodological ro-

bustness, i.e. characteristic of these studies, including a formal as-

sessment of bias. However, there is criticism regarding the

application of systematic reviews within the field of international

development research, as they tend to value quantitative over

qualitative evidence, thereby overlooking the context and proc-

esses which are paramount to this kind of research (Mallett et al.,

2012). Based on this, the authors determined that a literature re-

view was a more suitable design for evaluating this research ques-

tion. Finally, while the authors attempted to ensure homogeneity

among case studies, this was severely limited by the non-

homogenous nature of post-conflict countries. For example, while

Afghanistan is broadly referenced by the literature as being in a

‘post-conflict’ period, in reality many areas of the country continue

to experience heavy fighting and a continued military presence

(Frost et al., 2016). Due to these differences, the reader should

keep in mind that comparisons between post-conflict countries are

not an exact science, and differential health outcomes may not ne-

cessarily be linked to the specific inputs or policies evaluated in

these case studies.

Case studies and reflections

Overview of health system developmentAn overview of the conflicts in each country can be found in Box 1.

One- to four-years post-conflict

Following the end of hostilities, Cambodia, Afghanistan and

Mozambique each had to contend with similar challenges, includ-

ing; extreme reductions in health outcomes, sometimes stratified

along certain demographics such as gender or class; an influx of

NGOs that needed to be co-ordinated and managed; a severely

weakened Ministry of Health, with minimal capacity and legitim-

acy; a power vacuum, or a weakened and conflicted governing

body; widespread destruction of health infrastructure, particularly

in rural areas; the collapse of the health workforce, sometimes due

to purposeful targeting of said workforce; and finally, simmering

tensions leftover from the conflict.

The health sector in post-conflict Cambodia had to deal with a

rapid influx of international and local NGOs, all under the highly

contentious and fragmented governing body of the United Nations

Transitional Authority in Cambodia (UNTAC; UNRISD, 1994;

Figure 1. Framework from Waters et al. (2007).

Health Policy and Planning, 2019, Vol. 34, No. 3 233

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

Box 1. Background of case studies

Case 1. Cambodia (1970–1991)

History: As per the United Nations Research Institute for Social Development (UNRISD, 1994), Cambodia has been subject

to decades of protracted conflicts and brutal political regimes. Following a US supported military coup in 1970, the country

became substantially involved in the Vietnam War. Extensive carpet-bombing by US forces on the Vietnamese–Cambodian

border caused widespread devastation and internal displacement. This instability allowed the Khmer Rouge political party

to gain power in 1975, and a radical return to Cambodia’s agricultural and traditional roots was implemented. During this

period, an estimated half a million to 2 million people, mainly intellectuals and minorities, were systematically detained

and killed in the now infamous ‘killing fields’.

The Khmer Rouge was overthrown in 1979 with support from the Vietnamese, who then established their own republic.

Initially, the regime received substantial emergency support and aid relief from Western powers; however, upon declaration

by the UN of an ‘end to the emergency’ in 1982, an aid embargo was imposed on the country in an effort to stop the

Vietnamese occupation (UNRISD, 1994). This drastically reduced the amount of foreign funding assistance, and further

aggravated internal conflict within the country between various power factions. It was not until the establishment of the

State of Cambodia and signing of the Paris Peace Accord in 1991 that negotiations to end the conflict were successful,

allowing for Cambodia’s re-entry into the international community (UNRISD, 1994).

Key challenges: Entering this post-conflict period, Cambodia had some of the worst health indicators in Asia; the Ministry of

Health in Cambodia [MoH(C)] estimated life expectancy to be 47 years for men and 49 years for women, and infant mortality

was 120 per 1000 births (Heng and Key, 1995). Public health infrastructure was equally devastated; under the Khmer Rouge

the number of health professionals was reduced to 25 doctors and 3 members of staff at the MoH(C) (Heng and Key, 1995). A

two-tiered health system also emerged under the Vietnamese-backed regime, with the public sector being severely under-

funded and many health facilities falling into extreme disrepair, particularly in rural areas (Heng and Key, 1995).

Case 2. Afghanistan (1979–2001)

History: Prior to 1979, Afghanistan had a relatively sizeable and functioning health sector, albeit with less services in rural

areas. However, much of this infrastructure was destroyed following the Soviet invasion of 1979. During the invasion, the

Soviets attempted to depopulate rural areas in order to reduce support for the indigenous ‘mujaheddin’; this included the

systematic destruction of healthcare facilities. Consequently, when Soviet troops withdrew in 1988, over 60% of rural health

facilities had been destroyed (Cook, 2003).

Following Soviet withdrawal, Afghanistan plunged into a civil war as infighting broke out between different ‘mujaheddin’

groups. Gradually, the Taliban began to take power, and by 2001, they controlled roughly 90% of the country. During this

time, the limited capacity of the government to provide proper services resulted in NGOs playing a crucial role in health

service delivery. Prior to the US-led occupation invasion in 2001, an estimated 20 International NGOs (INGOs), and 200

Local NGOs were providing 80% of the available healthcare in Afghanistan (John, 2001).

Key challenges: With the collapse of the Taliban in 2001, US forces established an interim government, thereby bringing

about the ‘post-conflict’ period. The Ministry of Public Health in Afghanistan [MoH(A)] was faced with some of the worst

health outcomes in the world; estimated life expectancy was 54 years for men and 56 years for women, and in some

regions the infant mortality rate was 165/1000 births—the highest ever recorded (Newbrander et al., 2014). There was

roughly one physician for every 50 000 people (Cook, 2003), and only 10% of the population lived within an hour walking

distance from a health facility, largely in rural areas (Newbrander et al., 2014). Additionally, because of the restrictions

placed on women under the Taliban, there were substantial gender inequities, including a deficiency of midwives and fe-

male doctors (Cook, 2003).

Case 3. Mozambique (1977–1992)

History: The Mozambican Civil War lasted from 1977 to 1992, and is commonly characterized as a proxy war to the Cold

War (Robinson, 2006). Following independence from Portugal in 1975, Mozambique’s ruling party, the Mozambican

Liberation Front (FRELIMO), began to implement extreme authoritarian socialist reforms. These measures created some

backlash, and resulted in the creation of the militant/political movement, the Mozambican National Resistance (RENAMO)

group. Hostilities broke out into full-blown conflict in 1977, with FRELIMO backed by the Soviet Union, and RENAMO

backed by white-ruled Rhodesia and the South African apartheid regime (Robinson, 2006).

RENAMO employed many guerrilla war techniques and carried out raids against rural towns and important infrastructure,

including much of the rural health network. FRELIMO was unable to properly defend against these attacks, and relocated

much of its population and health sector to urban areas. A military stalemate, combined with the collapse of the Soviet

Union and the ending of apartheid in South Africa, contributed to peace-talks, and culminated with the signing of the Rome

Peace Accords in 1992 (Robinson, 2006).

Key challenges: Entering the post-conflict period, Mozambique had the worst health indicators of the three countries that

were included, with an estimated life expectancy of 42 for men and 45?years for women, and an infant mortality rate of 155/

1000 births. Additionally, the Ministry of Health in Mozambique [MOH(M)] faced challenges mainly relating to the severe rural

and urban inequalities, persistent tensions between RENAMO and FRELIMO groups, and high number of amputations and

mutilations resulting from the extensive use of landmines during the conflict (Pavignani and Colombo, 2001). Notably, the situ-

ation in Mozambique is unique in that, with support from the WHO and the World Bank, the MoH(M) started planning for

health system reconstruction in 1989, 3 years before the end of the conflict (Pavignani and Colombo, 2001).

234 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

Lanjouw et al., 1999; Gellman, 2010). While many attempts were

made to co-ordinate these stakeholders, the MoH(C) lacked the ne-

cessary capacity and authority to properly implement these meas-

ures (UNRISD, 1994; Lanjouw et al., 1999). Consequently, the

initial reconstruction process was largely driven by the agenda of

donors via allocation of funds towards international and local

NGOs (Lanjouw et al., 1999; Gellman, 2010).

A similar narrative occurred in Mozambique, where the post-

conflict governing body, the United Nations Operation in

Mozambique (UNOMOZ), was unable to properly manage and co-

ordinate the large number of NGOs who had entered the country

(Barnes, 1998; Pavignani and Colombo, 2001). Indeed, the co-ordi-

nating arm of the operation, the UN Office for Humanitarian

Assistance Coordination (UNOHAC), was met with competitiveness

and resistance, even within UN departments (Pavignani and

Colombo, 2001). Accordingly, NGOs mainly worked and were

influenced by donors with minimal government input (Barnes,

1998). This was despite extensive planning for health system devel-

opment conducted by the MoH(M) prior to the end of the conflict

(Pavignani and Colombo, 2001).

The post-conflict period in Afghanistan was able to resolve these

co-ordination problems by introducing a contracting-out model

through performance-based partnership agreements (PPAs; Akashi

et al., 2006). Under this scheme, NGOs would bid for regional con-

tracts via the MoH(A) (with support from the World Bank and

WHO) to implement a Basic Package of Health Services (BPHS;

(Ameli and Newbrander, 2008; Fujita et al., 2011). Designed by the

MoH(A) with funding from the major donors (World Bank, WHO,

and USAID), the BPHS focused on addressing immediate health

needs through the delivery of primary health care (PHC) services

(Michael et al., 2013). While this approach succeeded in co-ordinat-

ing the NGO sector and rapidly scaling up health services, it had

some notable disadvantages. Primarily, NGOs were less willing to

bid for more remote contracts due to insecurity, creating geograph-

ical inequalities (Michael et al., 2013). Furthermore, the autonomy

and stewardship of the MoH(A) in deciding which activities to fund

was minimal, as the MoH(A) only truly managed three provincial

contracts, while the World Bank, the European Union, and USAID

were responsible for managing all others (Michael et al., 2013;

Newbrander et al., 2014).

The managerial climate of these post-conflict environments had

an impact on the progression of infrastructure rehabilitation and

human resource mobilization. As development in Cambodia and

Mozambique was undertaken without a clear national strategy, it

was largely donor-driven and projects with immediately visible out-

puts were highly favoured (Martınez, 2006; Lanjouw et al., 1999).

Consequently, funding directed towards health infrastructure re-

habilitation was prioritized, although some critics characterized it as

disproportionate to other health system inputs in Cambodia

(Lanjouw et al., 1999), and slow and expensive in the rural areas in

Mozambique (Pavignani and Colombo, 2001). Conversely, because

the BPHS mainly focused on delivering PHC services, infrastructure

development in Afghanistan focused on expanding access to these

services through the construction of PHC and Community Health

Care Centres (Ameli and Newbrander, 2008).

Regarding the mobilization of health workers, there are a few

things the development community must consider, including; how

to increase the quantity of the health workforce while also ensuring

both the quality and diversity of delivered services; guaranteeing

equal coverage; and addressing any country-specific, contextual bar-

riers (Roome et al., 2014). In Cambodia and Afghanistan, a national

health workforce assessment was conducted early-on, revealing a

lack of healthcare workers across all professions in both countries,

and a lack of female health professionals in Afghanistan (Fujita

et al., 2011). Both countries focused on increasing the quantity of

the healthcare workforce, with Afghanistan specifically prioritizing

the development of midwives and increasing the number of female

health workers (Fujita et al., 2011). In Mozambique, as would even-

tually be the case in Cambodia, the high number of NGO-provided

services combined with a drop in the salary of national health work-

ers from austerity measures imposed by international funding agen-

cies created an ‘internal brain-drain’ and the emergence of a two-

tiered health system (Pfeiffer, 2003). On a more innovative note

however, the MoH(M) integrated FRELIMO and RENAMO health

Figure 2. Control of Government Corruption. From the World Bank: ‘Control of Corruption captures perceptions of the extent to which public power is exercised

for private gain, including both petty and grand forms of corruption, as well as “capture” of the state by elites and private interests. Estimate gives the country’s

score on the aggregate indicator, in units of a standard normal distribution, i.e. ranging from approximately �2.5 to 2.5’ (The World Bank 2018)..

Health Policy and Planning, 2019, Vol. 34, No. 3 235

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

workers to promote social cohesion and diffuse underlying political

tensions (Pavignani and Colombo, 2001).

Five- to eight-years post-conflict

At this point in the narrative, we begin to see somewhat of a diver-

gence in the approaches taken towards health system development.

The consequences of specific actions and policies begin to manifest

themselves, contextual factors become more pronounced, and the

political climate accentuates considerably. However, there persists

key similarities during this period, namely; a priority to claim na-

tional ownership of the health system and reduce reliance on exter-

nal aid; a Ministry of Health that continues to contend with limited

capacity and legitimacy; a more pronounced division between local

institutions and the NGO-donor sector; the persistence or re-

occurrence of conflict; and continued differences between rural and

urban healthcare coverage.

In Cambodia, adoption of ‘The Health Coverage Plan (1995)’

framework by the MoH(C) represented an attempt to shift to-

wards government ownership of the health system (Hill and Eang,

2007). The framework outlined plans to decentralize the health

sector, with the introduction of Operational Districts, and expand

health-care facilities to the entire population (Annear et al., 2015).

However, due to continued fighting with remnants of the Khmer

Rouge, in reality operations were limited at the district level, and

the health sector remained fairly centralized throughout this

period (Annear et al., 2015; Witter et al., 2016). Development of

the national health system was further challenged by a sector

dominated by NGO-driven projects and services (Godfrey et al.,

2002; Pfeiffer, 2003; Martınez, 2006). The MoH(C) attempted to

address this through the ‘Health Financing Charter (1996)’ which

approved the right for public health facilities to levy official user-

fees in order to generate additional revenue to become more com-

petitive (Hill and Eang, 2007; Annear et al., 2015; Witter et al.,

2016).

Similarly, the health sector in Mozambique became increasingly

fragmented; however, the government lacked the capacity to both

compete with NGO services and implement planned initiatives

(Martınez, 2006). Taking a somewhat different approach than

Cambodia, the MoH(M) focused on aligning the NGO-donor sector

with the national health system (Pavignani and Durao, 1999;

Pfeiffer, 2003; Martınez, 2006). They established multiple, cross-

cutting co-ordination mechanisms, including a Pooling Arrangement

for Technical Assistance in 1996 (Pavignani and Hauck, 2002), and

a Sector Coordination Committee in 1998 (Martınez, 2006). Both

of these were formed with the intention of moving towards a Sector

Wide Approach (SWAp) to management of the health sector

(Ministry of Health, 2001; Martınez, 2006; Negin and Hort, 2010).

By integrating a purchaser-provider split from the outset, the

PPA mechanism in Afghanistan allowed the MoH(A) to by-pass

many of the challenges faced by Cambodia and Mozambique.

Moving into this period, the MoH(A) managed to revise the BPHS

to include services for mental health and disabilities, and introduced

the Essential Package of Hospital Services (EPHS) to expand services

to secondary and tertiary care (Frost et al., 2016). However, while

the MoH(A) made attempts to increase coverage in rural areas

through monetary incentives, their success was minimal at best

(Michael et al., 2013; Frost et al., 2016). Attempts to reconcile fi-

nancial challenges were also undertaken, including the removal of

user-fees at PHC centres to promote equity (Steinhardt et al., 2013),

and the piloting of community funds to decrease reliance on foreign

aid (Rao et al., 2009).

With support from the Asian Development Bank (ADB), the

MoH(C) made significant advances in infrastructure development

(Asian Development Bank, 2004). Through the Basic Package of

Health Services ‘Project (1998)’, community health centres were ren-

ovated and provided with equipment and essential drugs, and dis-

trict hospitals were strengthened (Asian Development Bank, 2004).

Afghanistan continued to strengthen its’ PHC network; however,

inequalities continued to persist in rural areas (Ameli and

Newbrander, 2008; Frost et al., 2016). Additionally, the first signs

of corruption began to manifest, with reports that health facilities

were mainly concentrated in the villages of provincial health direc-

tors (see Box 2; Newbrander et al., 2014; Frost et al., 2016). In

Mozambique, the MoH(M) made attempts to strengthen their refer-

ral/tertiary care system by investing in urban infrastructure; how-

ever, rural areas continued to remain in a poor state of repair, with

a general lack of basic equipment (Ministry of Health, 2001).

During this time, both Cambodia and Mozambique shifted focus

from increasing the quantity of healthcare workers to improving the

quality-delivered services. This strategy proved to be unsustainable

in Cambodia, as staffing standards were not met for nurses and mid-

wives, particularly in rural areas (Fujita et al., 2011; So and Witter,

2016). Although the same did not occur in Mozambique, the quality

of training was criticized by some as being inadequate and financial-

ly unsustainable (Pavignani and Colombo, 2001), and specialized

staff, such as doctors for tertiary care, continued to be recruited

through expatriate technical assistance (Ministry of Health, 2001).

However, significant gains were made in expanding the workforce

Box 2: Health information systems: issues with monitoring

and evaluation

A key requirement for effective health system develop-

ment is the presence of reliable health indicators, report-

ing mechanisms, and accountability measures (Kruk

et al., 2010). Without these, the process risks the misallo-

cation of funds towards ineffective programmes and

projects, the development of inefficient policies and ini-

tiatives, and further long-term consequences concerning

sustainability and stewardship. However, post-conflict

environments are extremely susceptible to these prob-

lems, as the climate is generally very volatile, lacks a le-

gitimate government structure, and is particularly vulner-

able to ulterior motives on the part of donors, political

parties, or militia factions (Gordon, 2013).

Indeed, some form of corruption or inconsistencies in

reporting were observed across all case studies.

Afghanistan, scoring the lowest on control of govern-

ment corruption (Figure 2) was particularly adept to this,

with many researchers questioning the reliability and ac-

curacy of official reports documenting health outcomes,

coverage, and accessibility (Michael et al., 2013; Frost

et al., 2016). Even in Mozambique, which scored the

highest on the same measure (Figure 2), there were dis-

crepancies and reports of purposeful misreporting in

rural areas (Ministry of Health, 2001; Martınez, 2006).

These examples illustrate the need for effective and im-

partial monitoring and evaluation mechanisms which

can be implemented throughout the entirety of the de-

velopment and reconstruction process.

236 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

to rural areas through resource re-allocation (Pavignani and

Colombo, 2001, Ministry of Health, 2001). Afghanistan continued

to invest in midwifery education, and ultimately established an ac-

creditation system to ensure service quality (Fujita et al., 2011;

Turkmani et al., 2013), while also diversifying services by investing

in community healthcare workers (Fujita et al., 2011; Najafizada

et al., 2014). Despite significant efforts, however, rural inequalities

in workforce distribution and quality continued to persist (Fujita

et al., 2011; Turkmani et al., 2013).

Nine- to 12-years post-conflict

Despite the ever-increasing distance we are now reaching from the

conflict period, its effects and subsequent response from the devel-

opment community continued to reverberate throughout the health

sector. This period is uniformly characterized by a transition to-

wards some form of a SWAp to health, with all countries presenting

some kind of initiative or agreement outlining this. Additionally, it

is during this period in which we finally begin to see increased na-

tional ownership of the health system, with a move towards decen-

tralization, either through the strengthening of rural networks, or

through the use of contracting-out agreements. Finally, based on the

case studies, the limitations of assistance begin to become apparent,

with countries either overcoming key challenges or being unable to

properly reconcile them.

The 1998 general elections in Cambodia, which served to incorp-

orate Khmer Rouge guerrillas into government forces, ushered in a

period of social and economic stability (Gollogly, 2002). Following

the successful piloting of an external contracting model (Bhushan

et al., 2002; Asian Development Bank, 2004; Akashi et al., 2006),

where, similar to Afghanistan, international NGOs were sub-

contracted by the MoH(C) to deliver health services, the MoH(C)

adopted a SWAp initiative (Annear et al., 2015). As per this agree-

ment, all stakeholders would work under the MoH(C) to deliver an

Box 3: Immunization campaigns: vertical vs horizontal programmes

The combined average immunization rates for DTP, Pol3, MCV1 and BCG in each country are presented in Figure 3

(detailed rates per vaccine found in the Supplementary Data). While all countries initially had relatively low vaccination

rates post-conflict, only Afghanistan followed a linear, upward trend; immunization rates in Cambodia dipped, surged

quickly and then tapered off; and immunization rates in Mozambique steadily improved, then plateaued. The trend seen in

Cambodia may be attributed to the National Immunization Programme (NIP), a centralized, vertical programme managed

by the MoH(C). While the campaign was successfully expanded in the early- to mid-years post-conflict, there remained

shortages of drugs and other programmes at the district and local levels (Egami et al., 2012). This created large gaps in

vaccination coverage between rural and urban areas, and was unsustainable in that it contributed to the declining immun-

ization rates seen in Figure 3 (Soeung et al., 2007; Egami et al. 2012). Indeed, as evidenced by Afghanistan and

Mozambique, horizontal immunization programmes with greater integration at the district and local levels, while not as im-

mediately effective, are more sustainable and efficient in the long run (Jani et al., 2008; Mbaeyi et al., 2017).

Although data from the WHO shows significant gains in immunization coverage for Afghanistan and Mozambique, these

numbers have been questioned and refuted by other researchers. Specifically, they have pointed to numerous problems,

including; discrepancies between individual survey results, reports from the government, and reports from multilateral

organizations (Mavimbe et al., 2005; Michael et al., 2013; Mugali et al., 2017); the quality of data and the administrative cap-

acity required to record this data (Mavimbe et al., 2005; Michael et al., 2013); and inconsistencies relating to outbreaks of

diseases and high reported vaccination coverage (Jani et al., 2008; Mugali et al., 2017). This further exemplifies a larger

problem within post-conflict countries; namely, high levels of corruption, low accountability and the need for effective mon-

itoring and evaluation mechanisms (see Box 2).

Figure 3. Average immunization rate by country.

Health Policy and Planning, 2019, Vol. 34, No. 3 237

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

agreed-upon framework of health services (Annear et al., 2015). In a

further effort to increase ownership of the health system, there was

also increased focus on prioritizing MoH(C)-delivered services as

opposed to INGO-delivered services (Annear et al., 2015).

In Afghanistan, the continued reliance on NGOs for health serv-

ices under the PPA mechanism created a strained relationship with

the Afghan government (Michael et al., 2013). The MoH(A)

expressed concerns, mainly citing the need to maintain state

Box 4: Financial resources

External resources for health. All three countries substantially relied on external financial assistance at varying degrees

throughout the 12-year post-conflict period (Figure 4). Cambodia received large amounts of financial assistance in the first

4 years following the end of the conflict, but this eventually tapered off, and the country became the least-reliant on exter-

nal aid of all the case studies. Afghanistan received low-levels of external aid initially; however, this increased substantially

at around 4-years post-conflict, most likely influenced by the election of a donor-friendly, US backed administration in 2004

(Newbrander et al., 2014). Mozambique received the highest percentage of external aid throughout all 12 years. In terms of

absolute amounts, Afghanistan received far more external assistance than both Cambodia and Mozambique, equal to

roughly 30USD per capita in the last year of the period evaluated (compared with 10USD per capita in Cambodia, and

16USD per capita in Mozambique).

Health expenditure. Two measures of health expenditure were assessed; per capita health expenditure (Figure 5), and

health expenditure as a percent of GDP (Figure 6). Cambodia had mid-levels for both measures compared with the other

case studies; Afghanistan spent the most on healthcare per capita, but the least as a percent of GDP; conversely,

Mozambique spent the least on healthcare per capita, but the most as a percent of GDP.

Figure 5 Health expenditure per capita by country.

Figure 4. External resources for health by country.

238 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

legitimacy, as well as the higher costs of implementing the PPA

mechanism for service delivery (Palmer et al., 2006; Michael et al.,

2013). These tensions, as well as continued reliance on foreign aid,

were addressed by the System Enhancement for Health Action

Transition (SEHAT) fund initiative—a World Bank proposal consid-

ered to be a precursor to SWAp (Michael et al., 2013; The World

Bank, 2013). This initiative aimed to increase national ownership of

the health sector by ensuring long-term sustainability through the

strengthening of national health institutions and decreasing reliance

on external aid (The World Bank, 2013); however, based on data

from Figure 4 (external resources for health by country) there is no

concrete data to support the success of this policy.

Mozambique also saw a move towards a SWAp with the signing

of the ‘Kaya Kwanga’ agreement (Martınez, 2006; Negin and Hort,

2010). This was a non-binding agreement between all development

partners outlining a shared vision for health between the MoH(M)

and donor community, and prioritizing that technical assistance be

driven by national, and not donor, priorities (Pfeiffer, 2003;

Martınez, 2006; Negin and Hort, 2010). These initiatives were fur-

ther established through the ‘Health Sector Strategic Plan’, a nation-

al document which prioritized a SWAp to policymaking,

decentralization and the correction of rural and urban inequalities

(Ministry of Health, 2001; Negin and Hort, 2010).

In Cambodia, development of health infrastructure continued to

focus on strengthening community health services as per the Basic

‘Package of Health Services Project (1998)’ (Asian Development

Bank, 2004; So and Witter, 2016). In alignment with the decentral-

ization reforms, measures were also taken to enhance the manage-

ment of district health centres (Asian Development Bank, 2004).

Regarding human resources, local recruitment initiatives were

undertaken to address the continued gaps in rural health services

(Fujita et al., 2011; Witter et al., 2015; So and Witter, 2016), while

the integration of Khmer Rouge staff served to dissipate tensions (Ui

et al., 2007; So and Witter, 2016).

Afghanistan continued to face barriers in increasing the number

of healthcare workers in rural areas, with a report finding there to be

17 public health workers per 10 000 people in rural areas, compared

with 36 per 10 000 in urban areas (Frost et al., 2016). While the

MoH(A) attempted to address these inequalities with the deployment

of mobile health teams, persistent conflict with the Taliban, particularly

in the south, blocked these initiatives (Frost et al., 2016). Mozambique

saw strides in human resource mobilization resulting from the

implementation of their ‘Human Resources Development Plan 1999’

(Ministry of Health, 2001). This successfully increased the number

of university-trained personnel working within the health sector, there-

by reducing workforce imbalances; however, these professionals mostly

remained concentrated in urban areas (Ministry of Health, 2001).

Health outcomesHealth outcomes

As per percent change, Mozambique saw the greatest gains across

all health outcomes, followed by Cambodia, then Afghanistan

(Table 1). While these results are impressive, Mozambique still had

the worst relative health outcomes of all three countries, with a life

expectancy of 51 years in the final year assessed; comparatively

Cambodia had the best at 61 years, with Afghanistan at a close se-

cond with 60 years (Table 1; details in Figures 7–9). However, it

should be noted that the improvement or worsening of health out-

comes does not always occur linearly (Wang et al., 1999). Indeed,

Figure 6. Public health expenditure as percent of GDP by country.

Table 1. Percent changes and health outcomes 12-years post-conflict by country

Country

Health outputs Cambodia, n (%) Afghanistan, n (%) Mozambique, n (%)

Life expectancy 61 (13.0)a 60 (7.1) 51 (15.9)

Infant mortalityb 62 (27.9) 70 (23.1) 95 (37.5)

Under-5 mortalityb 79 (33.1) 97 (25.4) 141 (38.2)

aHealth outcome 12-years post-conflict (percent change).bPer 1000 births.

Health Policy and Planning, 2019, Vol. 34, No. 3 239

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

when observing infant and under-5 mortality rates in Cambodia,

one can clearly see a plateau, almost worsening, in the first 8 years,

followed by a sharp improvement in the last 4 years (Figures 8 and

9). Therefore, these results should be interpreted with some degree

of caution.

Revised framework

OverviewThe revised framework depicted (Figure 10) is constructed purely

from the application of the framework to the case studies; it expands

on the framework originally presented by Waters et al. (2007) to re-

flect best-practice development processes across a 12-year timeline

from the post-conflict year. Within this process, specific inputs are

delineated, and the organizations that should be responsible for said

inputs are also included. New inputs were also added, namely pre-

vention of government and donor corruption, accountability among

stakeholders, and monitoring and evaluation of the health system,

all falling under the section of ‘Governance’.

For organizational purposes, the development process is sepa-

rated into four phases, each characterized by a certain end goal or

trait, specifically:

1. The Response Phase: This refers to the first year after the con-

flict; it is characterized by an absence of functioning local institu-

tions and the beginning of the international community’s

involvement in the reconstruction process.

2. The Transitional Phase: This phase is characterized by the re-

emergence of local institutions, such as a Ministry of Health and

a national government. Development of the health sector is pri-

marily managed by the international community. It is transition-

al in that the main challenge during this time is moving from an

emergency-based health system to a more permanent one.

3. The Reconstruction Phase: This is the main development phase,

with large-scale reconstruction and development initiatives

Figure 7. Life expectancy by country.

Figure 8. Infant mortality by country.

240 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

undertaken within the health sector. This period also sees the

transition of stewardship from the international community to-

wards the local institutions and organizations.

4. The Sustainable Development Phase: This phase is mainly char-

acterized by promoting sustainable development, namely; ensur-

ing national ownership and financing of the health system. It

should be noted that none of the countries selected in the case

studies reached this phase within 12 years as is characteristic of

the vast majority of post-conflict countries (Haughton, 1998).

Therefore, this phase is an ‘ideal’ endpoint, and may not neces-

sarily be achievable in practice.

The framework further depicts how specific policies and inputs

should be sequenced throughout these phases, as well as which

stakeholders should be responsible for said policies and inputs. The

sequencing and responsibility of specific policies and inputs are out-

lined below.

GovernanceCorruption

Preventing corruption and promoting stability within the health sys-

tem requires the presence of an impartial auditing body that has the

Figure 9. Under-five mortality by country.

Figure 10. Adapted framework for post-conflict health system development.

Health Policy and Planning, 2019, Vol. 34, No. 3 241

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

necessary authority and capacity to control and prevent corruption

across all levels of governance (Mackey and Liang, 2012). This ap-

proach has seen some success in Afghanistan, which established a

High Office of Oversight and Anti-Corruption (HOO) body to im-

plement anti-corruption strategies (Spector, 2012). However, this

again requires the presence of a functioning local government. The

framework therefore proposes that the primary multilateral organ-

ization handling the reconstruction process (the WHO, or a major

donor such as the World Bank) assume this responsibility through-

out most of the development process, as they can develop and imple-

ment monitoring tools to prevent corruption such as health

expenditure auditing and tracking, adoption of a code of conduct

and ethics, and earmarking of funds (Mackey and Liang, 2012).

This system should be implemented by this organization throughout

the ‘Transitional and Reconstruction Phases’, and will ultimately be

transferred over to a government institution during the ‘Sustainable

Development Phase’.

Monitoring and evaluation

This should be undertaken by donors, and indeed, as the primary

source of funding for most of the development process, it would be

in their best interest to do so. The main concern here is the need for

‘impartial’ system, and the influence of donors must not interfere

with the priorities of the national health system. Therefore, to

achieve this separation between funding and implementation, there

needs to exist a legitimate state institution and a clear national

health strategy (as opposed to donors channelling funds through

third parties, such as NGOs). This monitoring system should be

managed by the donor community until the ‘Sustainable

Development Phase’, at which point responsibilities can be trans-

ferred to a government institution. Again, this approach saw some

success in Afghanistan with the establishment of the HOO body

(Spector, 2012).

Accountability

Lastly, a fundamental yet rarely asked question concerns responsi-

bility; namely, who is accountable for the failure or success of proj-

ects, and the outcome of the health system as a whole? Put simply,

there are no easy answers to this question; post-conflict health sys-

tem development involves an array of different stakeholders with

overlapping responsibilities and obscure boundaries, a major chal-

lenge for establishing good governance. Indeed, ascribing account-

ability can be somewhat of a meaningless process if one begins to

think about who these institutions might be accountable to. SWAp

can be a useful tool for establishing this; however, due to the lack of

functioning local institutions this is difficult to implement during the

early stages of post-conflict reconstruction. Therefore, throughout

the ‘Response and Transitional Phases’, accountability should rest

mainly with donors and the international community; this should

shift during the ‘Reconstruction Phase’ with adoption of a SWAp

initiative and increased government accountability; and ideally,

local governments being completely accountable by the end of the

‘Sustainable Development Phase’.

PoliciesFollowing the end of hostilities, marked by either a peace agreement

or UN-sponsored resolution, the ‘Response Phase’ will see minimal

national involvement in the reconstruction process, and instead,

there will be an influx of international NGOs, foreign donors and

multilateral organizations. During this brief period, the primary

multilateral organization (such as the UN, the World Bank or the

WHO) handling the reconstruction process should begin conducting

a situational health needs assessment for the purposes of developing

a national health policy.

Moving on to the ‘Transitional Phase’, the primary multilateral

organization should work closely with the emerging national or

transitional government to re-establish an effective and visible

Ministry of Health. This can be done by supporting the Ministry via

the implementation of the national health policy, the development

of new policies to expand health services and address contextual

problems as they arise throughout the process, and the co-ordination

of the NGO-donor community. The latter can be achieved through

the use of a contracting-out PPA mechanism, as it proved successful

in both Cambodia and Afghanistan. The primary multilateral organ-

ization should be responsible for managing these contracts initially;

however, the long-term goal during this phase is to transfer manage-

ment of these contracts to the Ministry of Health. At the end of this

period, a system similar to what was seen in Afghanistan should be

in place, with funding being directed through the Ministry towards

NGOs who have been contracted-out to deliver a package of health

services as outlined in the national health policy.

The ‘Reconstruction Phase’ should prioritize expanding on the

national health policy to begin encompassing secondary and tertiary

health services, as well as contextual factors, such as urban and rural

inequalities, gender disparities and specific health disorders. Based

on the case studies, planning during this phase should prioritize the

adoption of SWAp, which saw success in co-ordinating stakeholders

when implemented earlier. This usually involves the pooling of

donor-funds towards a national health policy, and prioritizes local

capacity building by mandating sectorial leadership to the MoH

through partnerships with donors and other organizations (Peters

and Chao, 1998; Hutton and Tanner, 2004; Peters et al., 2013).

Essentially, these reform measures aim to promote a sustainable,

government-led partnership with donor agencies (Peters and Chao,

1998; Peters et al., 2013), and have been characterized as a suitable

option for countries emerging from conflict (Peters and Chao,

1998).

Moving to the ‘Sustainable Development Phase’, policies should

focus on the continued implementation of SWAp reform measures.

This phase should prepare for the eventual exiting of the inter-

national community, and ultimately, complete national ownership

of the health system. Specifically, this involves the gradual phasing

out, integration or replacement of management responsibilities; co-

ordination mechanisms; and the replacement of international NGOs

with local institutions.

InputsBased on the case studies, there are very few inputs that the develop-

ment community will realistically be able to implement during the

‘Response Phase’, and services during this time will mainly be main-

tained via the emergency services that were provided by NGOs and

other organizations during the conflict period. However, donors

and NGOs can still use this time to develop and implement a NIP as

soon as possible.

Moving to the ‘Transitional Phase’, the donor community should

focus on funding projects that facilitate implementation of the na-

tional health policy via NGOs. This will be somewhat contextual

and will ultimately be determined by the health needs of the popula-

tion, the geography of the region and the level of destruction caused

by the conflict. However, all things considered, in order to achieve a

rapid response, the construction of PHC centres is a smart option;

this can also be done in tandem with initiatives to rehabilitate

242 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

hospitals and other health centres. Additionally, the MoH should

work to increase the number of national health workers, as there

will most likely be an overall deficiency due to the conflict. These

health workers can be employed within the NGO contracting-out

system through a national hiring strategy, in order to promote na-

tional ownership and sustainability. Due to the likely reduced cap-

acity of the national health system during this time, high-level,

technical assistance, such as doctors and other specialists, will have

to be recruited through expatriates and the INGO system. It should

be noted that this model only functions with sufficient support to

national institutions and a co-ordinated health strategy, again

appealing to the importance of SWAp initiatives in post-conflict

countries.

During the ‘Reconstruction Phase’, the MoH and donor commu-

nity should begin to pilot and develop strategies for decreasing reli-

ance on external aid. This is easier said than done, and many post-

conflict countries are unable to ever fully sustain themselves finan-

cially. Potential solutions include limiting prior external assistance

to an amount that the country can realistically support, raising com-

munity health funds and introducing user-fees. However, this last

strategy should be used sparingly as it may serve to increase inequal-

ities. Additionally, there should be a focus on expanding reconstruc-

tion activities to more specialized facilities, such as hospitals and

out-patient clinics. Regarding health workers, training should also

be expanded to more specialized services, such as doctors and sur-

geons, while also maintaining a sufficient quantity of other health

workers. Lastly, immunization programmes previously administered

by NGOs should start to be transferred to the MoH, with planning

for integration at the local level.

In the final ‘Sustainable Development Phase’, the development

community should continue to work towards supporting special-

ized equipment and care but shift priorities towards ensuring sta-

ble national ownership of the health sector. While this will be

facilitated via the aforementioned SWAp reform measures, the

international community can take further steps to ease this transi-

tion by continuing to train and support health workers, provide

technical assistance for the use of specialized equipment and facili-

tate the transfer of management responsibilities to government

institutions.

Conclusion

To summarize, this article applied a health systems framework from

Waters et al. (2007) to three post-conflict case studies, with add-

itional measures taken for levels of corruption and specific health

outcomes. Measures of governance, inputs and policies were then

evaluated over a 12-year period following the end of conflict in

Cambodia, Afghanistan and Mozambique. From these case studies,

the Waters et al. (2007) framework was revised to reflect the most

effective development practices along this 12-year timeline.

The revised framework identifies key activities and their proper

sequencing throughout the development process. Acting as a road-

map, this framework can be used as a tool throughout the recon-

struction process and can be further expanded upon through contin-

ued in-depth analyses of post-conflict countries. By identifying

which stakeholders should conduct certain activities and when, this

framework also advocates for a co-ordinated approach, and hopes

to promote effective and efficient policy design and implementation

among all members of the development community throughout the

entirety of the reconstruction process.

Lastly, there are some findings and observations from the case

studies that warrant further discussion or research. Primarily, this

study further highlighted the need to develop sustainable financing

mechanisms (Witter, 2012) and effective governance measures

(Mackey and Liang, 2012) within post-conflict countries.

Additionally, while many individuals in the development commu-

nity contribute the successful outcomes in Afghanistan to the PPA

model (Newbrander et al., 2014), a simpler explanation may be the

high proportion of funding received by Afghanistan compared with

other countries (see Box 3). The extent to which this funding is polit-

ically determined should not be discounted, and this begs the ques-

tion; is financial aid to post-conflict countries determined purely by

the needs of the health sector, or by other factors? Future research

should aim to evaluate this in order to hold donors and the inter-

national community accountable to equitable funding among post-

conflict countries.

Supplementary data

Supplementary data are available at Health Policy and Planning online.

Ethical approval

None required.

Conflict of interest statement. None declared.

References

Abo-Hilal M, Hoogstad M. 2013. Syrian mental health professionals as refu-

gees in Jordan: establishing mental health services for fellow refugees.

Intervention 1: 89–93.

Akashi H, Fujita N, Akashi RK. 2006. Aid coordination mechanisms

for reconstructing the health sector of post-conflict countries. JMAJ 49:

251–9.

Ameli O, Newbrander W. 2008. Contracting for health services: effects of util-

ization and quality on the costs of the Basic Package of Health Services in

Afghanistan. Bulletin of the World Health Organization 86: 920–8.

Annear P, Grundy J, Ir P et al. 2015. The kingdom of Cambodia health system

review. Health Systems in Transition 5: 178.

Asian Development Bank. 2004. Project Completion Report on the Basic

Health Services Project in Cambodia. https://www.adb.org/sites/default/

files/project-document/70061/pcr-cam-27410.pdf, accessed 27 November

2018.

Barnes S. 1998. NGOs in peace-keeping operations: their role in

Mozambique. Development in Practice 8: 309–22.

Bhushan I, Keller S, Schwartz B. 2002. Achieving the Twin Objectives of

Efficiency and Equity: Contracting Health Services in Cambodia. Manila,

Philippines: ERD Policy Brief Series, Economic and Research Department,

Vol 6, Asian Development Bank. http://www.adb.org, accessed 27

November 2018.

Brown G, Langer A, Stewart F. 2008. A Typology of Post-Conflict

Environments: An Overview. http://www.crise.ox.ac.uk/, accessed 27

November 2018.

Cook J. 2003. Post-conflict reconstruction of the health system of

Afghanistan: assisting in the rehabilitation of a provincial hospital–context

and experience. Medicine, Conflict and Survival 19: 128–41.

Council on Foreign Relations. 2018. Global Conflict Tracker. Council on

Foreign Relations. https://www.cfr.org/interactives/global-conflict-track

er#!/global-conflict-tracker, accessed 27 November 2018.

Csis A. 2002. Post-Conflict Reconstruction: Task Framework. https://csis-

prod.s3.amazonaws.com/s3fs-public/legacy_files/files/media/csis/pubs/

framework.pdf, accessed 27 November 2018.

Egami Y, Fujita N, Akashi H et al. 2012. Can health systems be enhanced for

optimal health services through disease-specific programs?—results of field

studies in Viet Nam and Cambodia. Bioscience Trends 6: 1–6.

Health Policy and Planning, 2019, Vol. 34, No. 3 243

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

Foster S, Laing R, Melgaard B, Zaffran M. 2006. Ensuring Supplies of

Appropriate Drugs and Vaccines. The International Bank for

Reconstruction and Development/The World Bank. http://www.ncbi.nlm.

nih.gov/pubmed/21250304, accessed 27 November 2018.

Fouad FM, Sparrow A, Tarakji A et al. 2017. Health workers and the weapon-

isation of health care in Syria: a preliminary inquiry for The Lancet

American University of Beirut Commission on Syria. Lancet 390: 2516–26.

Frost A, Wilkinson M, Boyle P, Patel P, Sullivan R. 2016. An assessment of the

barriers to accessing the Basic Package of Health Services (BPHS) in

Afghanistan: was the BPHS a success? Globalization and Health 12: 71.

Fujita N, Zwi AB, Nagai M, Akashi H. 2011. A comprehensive framework for

human resources for health system development in fragile and post-conflict

states. PLoS Medicine 8: e1001146.

Gellman M. 2010. World views in peace building: a post-conflict reconstruc-

tion challenge in Cambodia. Development Practice 20: 85–98.

Godfrey M, Sophal C, Kato T et al. 2002. Technical assistance and capacity

development in an aid-dependent economy: the experience of Cambodia.

World Development 30: 355–73.

Gollogly L. 2002. The dilemmas of aid: Cambodia 1992-2002. Lancet 360:

793–8.

Gordon S. 2013. Health, conflict, stability, and statebuilding: a house built on

sand? Journal of Intervention and Statebuilding 7: 29–44.

Gordon S. 2014. The military physician and contested medical humanitarian-

ism: a dueling identity? Social Science & Medicine 120: 421–9.

Haughton J. 1998. The Reconstruction of War-Torn Economies. http://cite

seerx.ist.psu.edu/viewdoc/download; jsessionid¼59C5CD08D7E44C7704

4F6B34C71A5C83?doi¼10.1.1.537.9648&rep¼rep1&type¼pdf, accessed

27 November 2018.

Heng MB, Key PJ. 1995. Cambodian health in transition. BMJ (Clinical

Research ed.) 311: 435–7.

Hill PS, Eang MT. 2007. Resistance and renewal: health sector reform and

Cambodia’s national tuberculosis programme. Bulletin of the World Health

Organization 85: 631–6.

Hutton G, Tanner M. 2004. The sector-wide approach: a blessing for public

health? Bulletin of the World Health Organization 82: 893.

ICRC. 2016. Protracted Conflict and Humanitarian Action: Some Recent

ICRC Experiences. https://www.icrc.org/sites/default/files/document/file_

list/protracted_conflict_and_humanitarian_action_icrc_report_lr_29.08.16.

pdf, accessed 27 November 2018.

Jani JV, De Schacht C, Jani IV, Bjune G. 2008. Risk factors for incomplete vac-

cination and missed opportunity for immunization in rural Mozambique.

BMC Public Health 8: 161.

John E. 2001. No NGOs and the economic recovery of Afghanistan.

Development Practice 11: 633–6.

Kruk ME, Freedman LP, Anglin GA, Waldman RJ. 2010. Rebuilding health

systems to improve health and promote statebuilding in post-conflict coun-

tries: a theoretical framework and research agenda. Social Science &

Medicine 70: 89–97.

Lanjouw S, Macrae J, Zwi AB. 1999. Rehabilitating health services in

Cambodia: the challenge of coordination in chronic political emergencies.

Health Policy and Planning 14: 229–42.

Mackey TK, Liang BA. 2012. Combating healthcare corruption and fraud

with improved global health governance. BMC International Health and

Human Rights 12: 23.

Mallett R, Hagen-Zanker J, Slater R, Duvendack M. 2012. The benefits and

challenges of using systematic reviews in international development re-

search. Journal of Development Effectiveness 4: 445–55.

Martınez J. 2006. Implementing a Sector Wide Approach in Health: The Case

of Mozambique. London, UK. http://sergiorosendo.pbworks.com/f/

Martinezþ2006_mozambique_swap_review.pdf, accessed 27 November

2018.

Mavimbe JC, Braa J, Bjune G. 2005. Assessing immunization data quality

from routine reports in Mozambique. BMC Public Health 5: 108.

Mbaeyi C, Kamawal NS, Porter KA et al. 2017. Routine immunization service

delivery through the Basic Package of Health Services program in

Afghanistan: gaps, challenges, and opportunities. Journal of Infectious

Diseases 216: 273–9.

Michael M, Pavignani E, Hill PS. 2013. Too good to be true? An assessment of

health system progress in Afghanistan. Medicine Conflict and Survival 29:

322–45.

Ministry of Health. 2001. Strategic Plan for the Health Sector (PESS). http://

www.usaid.gov/mz/doc/plan/mh_stratplan_2010.pdf, accessed 27

November 2018.

Mugali RR, Mansoor F, Parwiz S et al. 2017. Improving immunization in

Afghanistan: results from a cross-sectional community-based survey to as-

sess routine immunization coverage. BMC Public Health 17: 290.

Najafizada S, Labonte R, Bourgeault I. 2014. Community health workers of

Afghanistan: a qualitative study of a national program. Conflict and Health

8: 26.

Negin J, Hort K. 2010. Governance and Management Arrangements for

Health Sector-Wide Approaches (SWAps): Examples from Africa, Asia and

the Pacific. http://ni.unimelb.edu.au/__data/assets/pdf_file/0003/542451/

governance-management-arrangements-health-SWAps.pdf, accessed 27

November 2018.

Newbrander W, Ickx P, Ferouxuddin F, Stanekxai H. 2014. Afghanistan’s

Basic Package of Health Services: its development and effects on rebuilding

the health system. Global Public Health 9: S6–28.

Ohiorhenuan J, Stewart F. 2008. Post-Conflict Economic Recovery Enabling

Local Ingenuity. New York. http://www.undp.org/content/dam/undp/li

brary/crisisprevention/undp-cpr-post-conflict-economic-recovery-enable-

local-ingenuity-report-2008.pdf, accessed 27 November 2018.

Palmer N, Strong L, Wali A, Sondorp E. 2006. Contracting out health services

in fragile states. British Medical Journal 332: 718–21.

Panter-Brick C. 2009. Conflict, violence, and health: setting a new interdiscip-

linary agenda. Social Science & Medicine 70: 1–6.

Pavignani E, Colombo A. 2001. Providing Health Services in Countries Disrupted

by Civil Wars: A Comparative Analysis of Mozambique and Angola. http://apps.

who.int/disasters/repo/14052.pdf, accessed 27 November 2018.

Pavignani E, Durao JR. 1999. Managing external resources in Mozambique:

building new aid relationships on shifting sands? Health Policy and

Planning 14: 243–53.

Pavignani E, Hauck V. 2002. Pooling of Technical Assistance in Mozambique:

Innovative Practices and Challenges. http://ecdpm.org/wp-content/uploads/

2013/11/DP-39-Technical-Assistance-Mozambique-2002.pdf, accessed 27

November 2018.

Perehudoff SK, Alexandrov NV, Hogerzeil HV. 2018. Access to essential med-

icines in 195 countries: a human rights approach to sustainable develop-

ment. Global Public Health 14: 421–44.

Peters D, Chao S. 1998. The sector-wide approach in health: what is it? Where

is it leading? International Journal of Health Planning and Management 13:

177–90.

Peters DH, Paina L, Schleimann F. 2013. Sector-wide approaches (SWAps) in

health: what have we learned? Health Policy and Planning 28: 884–90.

Pfeiffer J. 2003. International NGOs and primary health care in Mozambique:

the need for a new model of collaboration. Social Science & Medicine 56:

725–38.

Rao KD, Waters H, Steinhardt L, Alam S, Hansen P, Naeem AJ. 2009. An ex-

periment with community health funds in Afghanistan. Health Policy and

Planning 24: 301–11.

Robinson DA. 2006. Curse on the Land: A History of the Mozambican Civil

War. http://www.adelinotorres.info/africa/David Robinson, Curse on the

Land-History of the Mozambican Civil War.pdf, accessed 27 November

2018.

Roome E, Raven J, Martineau T. 2014. Human resource management in

post-conflict health systems: review of research and knowledge gaps.

Conflict and Health 8: 18.

Rubenstein L. 2009. Post-Conflict Health Reconstruction. www.usip.org,

accessed 27 November 2018.

Shuey D, Qosaj F, Schouten E, Zwi A. 2003. Planning for health sector reform

in post-conflict situations: Kosovo 1999–2000. Health Policy (New York)

63: 299–310.

So S, Witter S. 2016. The Evolution of Human Resources for Health Policies

in Post-Conflict Cambodia: Findings from Key Informant Interviews and

Document Reviews. https://rebuildconsortium.com/media/1372/wp19-evo

244 Health Policy and Planning, 2019, Vol. 34, No. 3

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019

lution-of-hrh-policies-in-post-conflict-cambodia-_-kii-and-document-

reviews.pdf, accessed 27 November 2018.

Soeung S, Grundy J, Kamara L, McArthur A, Samnang C. 2007.

Developments in immunization planning in Cambodia—rethinking the cul-

ture and organization of national program planning. Rural and Remote

Health 7: 630.

Spector BI. 2012. Negotiating Anti-Corruption Reforms in Post-Conflict

Societies: The Case of Afghanistan. Vol. 18. https://www-jstor-org.lib

proxy.ucl.ac.uk/stable/pdf/24590862.pdf? refreqid¼excelsior%3Adf6cb41

e9368b505310bb4d844732947, accessed 27 November 2018.

Steinhardt LC, Rao KD, Hansen PM, Alam S, Peters DH. 2013. The effects of

user fees on quality and utilization of primary health-care services in

Afghanistan: a quasi-experimental health financing pilot study in a

post-conflict setting. The International Journal of Health Planning and

Management 28: e280–97.

The World Bank. 1998. Post-Conflict Reconstruction: The Role of the World

Bank. http://documents.worldbank.org/curated/en/175771468198561613/

pdf/multi-page.pdf, accessed 27 November 2018.

The World Bank. 2013. Emergency Project Paper on a Proposed Grant in the

Amount of SDR 65.1 Million to the Islamic Republic of Afghanistan for the

System Enhancement for Health Action in Transition Project. Washington,

DC. http://documents.worldbank.org/curated/en/371911468197965410/pdf/

736450PJPR0P120Official0Use0Only090.pdf, accessed 27 November 2018.

The World Bank. 2017. The Toll of War: The Economic and Social

Consequences of the Conflict in Syria. Washington, DC. http://www.world

bank.org/en/country/syria/publication/the-toll-of-war-the-economic-and-so

cial-consequences-of-the-conflict-in-syria, accessed 27 November 2018.

The World Bank. 2018. Metadata Glossary: Control of Corruption: Estimate.

http://databank.worldbank.org/data/glossarymetadata/source/1181/con

cepts/series, accessed 27 November 2018.

Traboulsi F. 2012. A History of Modern Lebanon, 2nd edn. London: Pluto

Press.

Turkmani S, Currie S, Mungia J et al. 2013. “Midwives are the backbone of

our health system”: lessons from Afghanistan to guide expansion of midwif-

ery in challenging settings. Midwifery 29: 1166–72.

Ui S, Leng K, Aoyama A. 2007. Building peace through participatory health

training: a case from Cambodia. Global Public Health 2: 281–93.

UNRISD. 1994. Between Hope and Insecurity: The Social Consequences of

the Cambodian Peace Process. http://www.unrisd.org/80256B3C005BCC

F9/httpNetITFramePDF? ReadForm& parentunid¼0989F68532E21DA58

0256B6500558BEB & parentdoctype ¼book &netitpath¼80256B3C005

BCCF9/(httpAuxPages)/0989F68532E21DA580256B6500558BEB/$file/

beet.pdf, accessed 27 November 2018.

USAID. 2009. A Guide to Economic Growth in Post-Conflict Countries.

http://pdf.usaid.gov/pdf_docs/Pnado408.pdf, accessed 27 November 2018.

Wang J, Jamison D, Bos E, Preker A, Peabody J. 1999. Measuring Country

Performance on Health: Selected Indicators for 115 Countries. Washington,

DC. http://documents.worldbank.org/curated/en/754391468762326013/

pdf/multi-page.pdf, accessed 27 November 2018.

Waters H, Garrett B, Burnham G. 2007. Rehabilitating Health Systems in

Post-Conflict Situations. WIDER Research Paper 2007/06. United Nations

University. http://hdl.handle.net/10419/63390, accessed 27 November 2018.

WHO. 2007. Strengthening Health Systems to Improve Health Outcomes:

WHO’s Framework for Action. http://www.who.int/healthsystems/strategy/

everybodys_business.pdf, accessed 27 November 2018.

Witter S, Bertone MP, Chirwa Y et al. 2016. Evolution of policies on human

resources for health: opportunities and constraints in four post-conflict and

post-crisis settings. Conflict and Health 10: 1–18.

Witter S, Falisse J-B, Bertone MP. 2015. State-building and human resources

for health in fragile and conflict-affected states: exploring the linkages.

Human Resources for Health 13: 33.

Witter S. 2012. Health financing in fragile and post-conflict states:

What do we know and what are the gaps? Social Science & Medicine 75:

2370–7.

Health Policy and Planning, 2019, Vol. 34, No. 3 245

Dow

nloaded from https://academ

ic.oup.com/heapol/article-abstract/34/3/230/5423845 by Am

erican University of Beirut user on 27 M

ay 2019