Health Policy

Generation of political priority for global health initiatives:

a framework and case study of maternal mortality
Jeremy Shiffman, Stephanie Smith

Lancet 2007; 370: 1370–79 Why do some global health initiatives receive priority from international and national political leaders whereas others
See Editorial page 1283 receive little attention? To analyse this question we propose a framework consisting of four categories: the strength of
Department of Public the actors involved in the initiative, the power of the ideas they use to portray the issue, the nature of the political
Administration, Maxwell contexts in which they operate, and characteristics of the issue itself. We apply this framework to the case of a global
School of Syracuse University,
initiative to reduce maternal mortality, which was launched in 1987. We undertook archival research and interviewed
Syracuse, NY, USA
(J Shiffman PhD, S Smith MPA) people connected with the initiative, using a process-tracing method that is commonly employed in qualitative
Correspondence to: research. We report that despite two decades of effort the initiative remains in an early phase of development,
Dr Jeremy Shiffman, Department hampered by difficulties in all these categories. However, the initiative’s 20th year, 2007, presents opportunities to
of Public Administration, The build political momentum. To generate political priority, advocates will need to address several challenges, including
Maxwell School of Syracuse
the creation of effective institutions to guide the initiative and the development of a public positioning of the issue to
University, 306 Eggers Hall,
Syracuse, NY 13244-1020, USA convince political leaders to act. We use the framework and case study to suggest areas for future research on the
jrshiffm@maxwell.syr.edu determinants of political priority for global health initiatives, which is a subject that has attracted much speculation
but little scholarship.

Introduction human rights. Global political priority is the degree to

Global health initiatives vary in the amount of political which international and national political leaders actively
priority they receive from international and national give attention to an issue, and back up that attention with
leaders. Child immunisation, family planning, and the provision of financial, technical, and human resources
HIV/AIDS, for instance, at some points have attracted that are commensurate with the severity of the issue. We
great resources, whereas malnutrition and pneumonia know that global political priority is present when: (1)
have received little attention despite also addressing international and national political leaders publicly and
high-burden disorders. We know little about the sources privately express sustained concern for the issue; (2) the
of variance in priority levels afforded to global health organisations and political systems they lead enact policies
initiatives, since there is an absence of systematic to address the problem; and (3) these organisations and
research into this subject. political systems provide levels of resources to the problem
We propose an initial framework for analysing the that are commensurate with its severity. These three
determinants of political priority for global health factors include not only international but also national
initiatives, and we hope future researchers will modify components, since global initiatives rarely aim to generate
and improve this framework. It consists of four categories: priority only among international organisations—they
the power of actors involved, the ideas they use to portray also seek political support from national political systems.
the issue, the nature of the political contexts in which Global political priority alone is not sufficient to address
they operate, and characteristics of the issue itself. We an international problem successfully. Effective policies,
apply this framework to the global safe motherhood technology, and implementation systems, among other
initiative, which was launched in 1987 to reduce levels of elements, are also crucial. However, global political priority
maternal mortality. This initiative reached its aids success, and therefore is essential to investigate.
20th anniversary in 2007, but despite determined efforts In this paper we present the framework, examine
by advocates, it has yet to attract the level of political determinants of global political priority for safe
attention its founders hoped it would receive. A recent motherhood with reference to this framework, point to
study on political priority for maternal mortality reduction challenges that the initiative might face in gaining
took the country as the unit of analysis and developed a priority over the coming decade, and identify questions
framework that sought to explain variance in levels of for future research into sources of political priority for
national priority for safe motherhood in Guatemala, global health initiatives.
pas suffisant que ça soit un problème de santé globale pour avoir des
Honduras, India, Indonesia, and Nigeria.1 Our study asks politiques efficaces
a similar question about variance in political priority Framework for determinants of political priority
levels, but the unit of analysis is the global health for global initiatives
initiative. It deepens and expands the first framework by Researchers have sought to understand why initiatives
grounding it in the extensive research on collective pursuing social and political change succeed or fail in
action. attracting political support. They have investigated several
A global initiative is an organised effort linking people types of collective action efforts, including international
and organisations across national borders to address an networks for issues such as climate change,2–6 social
issue of international concern, such as climate change or movements for causes such as civil rights,7–10 and policy

1370 www.thelancet.com Vol 370 October 13, 2007

 Health Policy

communities that aim to place particular issues on provides direction to the initiative. For example, the
national agendas.11–13 A central concern in collective action former director of the UN Children’s Fund (UNICEF),
research is the role of power: the power of actors James Grant, is often cited as an example of such a leader.
connected with the issue; the power of the ideas used to Strong guiding institutions (factor 3)—ie, organisations
define and describe the issue; the power of political or coordinating mechanisms with a mandate to lead the
contexts to inhibit or enhance political support; and the initiative—are also crucial. Initiatives might start through
power of some characteristics of the issue, such as the informal associations or as projects inside formal
number of deaths a particular disease causes, to inspire organisations, but they must build their own enduring
action.2–13 institutions if they are to survive.7 Continual competition
These four elements form the foundation of our among concerned organisations to control the issue
framework for the determinants of political priority for could hamper the creation of these structures. The Task
global initiatives (table). Initiatives are more likely to Force for Child Survival and Development (formerly led
attract political support if they share specific features in by Grant) has been noted as a guiding institution that is
all categories. = les 4 éléments ? particularly effective for the cause of child health. Finally,
The first element is the power of the actors involved in initiatives are more likely to generate political support if = organisations locales
the initiative. Initiatives differ in the strength of those they link with grassroots organisations in civil society
who participate in them, in the quality of linkages that are pushing for global attention to the issue (factor 4),
between these actors, and in their collective capacity to rather than remaining confined to select members of a
confront opponents.14–16 Among those who influence global policy community. Pressure from grassroots AIDS
initiative acquisition of political support are policy activists on national governments and international
communities (factor 1)—the network of individuals and organisations, for instance, has helped to increase donor
organisations who operate globally and who are linked by aid to address the disease in developing countries.17
a central concern for the issue. These communities Ideas also shape political support for initiatives. The
ex de include prominent leaders of non-governmental role of ideas in politics has inspired much research,4,18
commu organisations, government officials, bilateral donors, which is grounded in recognition that material influences
nauté
politiqu members of UN agencies, other international alone cannot explain all political behaviour and that
e organisations, and academics. Policy communities that people interpret the world around them very
agree on basic issues such as how the problem should be differently. The central ideational variable in collective
Les solved are more likely to acquire political support than action research is the frame—ie, the way in which an
communa are those that are divided by such issues, since politicians issue is understood and portrayed publicly.10 Any issue
uT politiq
sont + will be more likely to listen to those in agreement as can be framed in several ways. For example, HIV/AIDS
crédibles authoritative sources of knowledge. has been framed as a public-health problem, a
qd elles
sont
The emergence of respected leaders who are embraced development issue, a humanitarian crisis, and a threat to
d'accord by the community (factor 2) helps with coalescence and security.17,19 Some frames resonate more than others, and

Description Factors shaping political priority

Actor power The strength of the 1 Policy community cohesion: the degree of coalescence among the network of individuals and organisations
individuals and that are centrally involved with the issue at the global level
organisations concerned 2 Leadership: the presence of individuals capable of uniting the policy community and acknowledged as
with the issue particularly strong champions for the cause
3 Guiding institutions: the effectiveness of organisations or coordinating mechanisms with a mandate to lead
the initiative
4 Civil society mobilisation: the extent to which grassroots organisations have mobilised to press international
and national political authorities to address the issue at the global level
Ideas The ways in which those 5 Internal frame: the degree to which the policy community agrees on the definition of, causes of, and
involved with the issue solutions to the problem
understand and portray it 6 External frame: public portrayals of the issue in ways that resonate with external audiences, especially the
political leaders who control resources
Political The environments in 7 Policy windows: political moments when global conditions align favourably for an issue, presenting
contexts which actors operate opportunities for advocates to influence decisionmakers
8 Global governance structure: the degree to which norms and institutions operating in a sector provide a
platform for effective collective action
Issue Features of the problem 9 Credible indicators: clear measures that show the severity of the problem and that can be used to monitor
characteristics progress
10 Severity: the size of the burden relative to other problems, as indicated by objective measures such as
mortality levels
11 Effective interventions: the extent to which proposed means of addressing the problem are clearly explained,
cost effective, backed by scientific evidence, simple to implement, and inexpensive

Table: The four categories for the framework on determinants of political priority for global initiatives

www.thelancet.com Vol 370 October 13, 2007 1371

 Health Policy

different frames appeal to different audiences. Finance policymakers will perceive harmful problems as more
ministers, for instance, might be more likely to respond serious (factor 10). Problems with fairly simple,
to frames that emphasise the economic costs of a health inexpensive, evidence-based solutions will be easier to
problem than are health ministers, who might pay more promote than will those without these features, since
attention to frames that focus on public-health benefits policymakers prefer to devote resources to issues that
Unifie la communauté politique sur and losses. Frames that resonate internally (factor 5) they think they can address effectively and cheaply
une même définition unify policy communities by providing a common (factor 11). Cheap and effective vaccines, for example,
understanding of the definition of, causes of, and have helped to generate political attention for child
Donne une direction au problème, solutions to the problem. Frames that resonate externally immunisation policies.
informe sur son impact (factor 6) move essential individuals and organisations to No one factor is necessary or sufficient for political
action, especially the political leaders who control the support: some initiatives that have not attracted political
resources that initiatives need. priority possess several of these characteristics and some
The political contexts in which actors operate also exert initiatives that have received political attention are
substantial influence over political support levels.20–23 without several. For example, HIV/AIDS presently
Donc la communauT politique a Those involved in the initiative might have little control attracts great political support despite extensive
une influence sur le contexte
over these contextual factors, but they should take them contention within its global policy community, and polio
into account if they wish to develop effective strategies. eradication continues to receive substantial resources
Many elements of political context matter, but two are despite the disease’s small global burden compared with
essential. First are policy windows (factor 7), which are many other disorders. Conversely, chronic diseases such
moments in time when worldwide conditions align as cardiovascular disorders, cancer and diabetes, some
favourably for an issue, presenting advocates with communicable diseases such as pneumonia, and a
especially strong opportunities to reach international number of risk factors such as malnutrition attract few
and national political leaders.12 Policy windows often worldwide resources despite causing high morbidity and
open after major disasters (such as a tsunami), mortality. These cases suggest the need for continual
discoveries (a new vaccine), or forums (global UN research into the determinants of political priority for
conferences). For example, the Millennium Development global health initiatives, including assessment of the
Goals (MDGs) have helped with the opening of policy relative causal weights of the factors, their interactive
windows for some of the causes included. A second effects, and whether different combinations of factors
crucial element of context is the global governance could raise the issue in global health.
structure for the sector (factor 8)—ie, the set of norms However, existing research into collective action
(shared beliefs on appropriate behaviour) and the provides evidence that, other things being equal, every
institutions that negotiate and enforce these norms. factor enhances the likelihood that an initiative will
International treaties, laws, and declarations exist for receive priority. A global policy community is more likely En
many sectors, including trade, the environment, and to generate political support for its concern if it is résumé

health, usually with an international organisation or set cohesive, well-led, guided by strong institutions, and
Multitude d'instances of organisations in charge of their enforcement. In some backed by mobilised civil societies; if it agrees on
non-coordonnées et avec des
priorités différentes sectors, these structures work well; in others, they are solutions to the problem and has developed frames for
dysfunctional. Several studies have expressed concern the issue that resonate with political leaders; if it takes
about the increasingly fragmented structure of global advantage of policy windows and is situated in a sector
governance for health, with many organisations with a strong global governance structure; and if it
competing for power, constantly shifting priorities, and addresses an issue that is easily measured, is high in
no one organisation or set of organisations with the severity, and has effective interventions available. In such
power to coordinate.24–25 a situation, the power of those involved with the initiative,
Finally, the nature of the issue itself shapes political ideas, political contexts, and issue characteristics all work
priority. Some issues are intrinsically easier to promote in favour of the initiative.
than others.2,26 Problems that are easily measured are
more likely to gain political support than are those that Methods used to apply framework
are not, since policymakers and advocates will have To examine the global safe motherhood initiative we
information to confirm the severity and monitor progress used process-tracing, which is a method that is
(factor 9). For instance, studies providing credible commonly used in qualitative social science inquiry and
evidence of high population growth and fertility rates involves analysis of several sources of information to
interacted with other factors in the 1970s and 1980s to uncover social processes and assess causality.27
convince political leaders in many developing countries In 2005–07, we conducted 23 interviews, lasting on
that they had population problems that needed attention. average 1·5 h, with individuals centrally involved in the
Problems that cause substantial harm, as indicated by development of the global initiative, including most of
objective measures such as numbers of deaths, are more its founders. All individuals had worked on safe
likely to attract resources than are those that do not, since motherhood with a UN agency, multilateral financial

1372 www.thelancet.com Vol 370 October 13, 2007

 Health Policy

institution, university, or international non-governmental The case of safe motherhood

organisation. We also analysed data from a In 1987, the World Bank, WHO, and the UN Population = 3 communautés politiques
complementary study on political priority for maternal Fund (UNFPA) sponsored a conference in Nairobi, Kenya, conférence qui a pu mettre
mortality reduction in five developing countries, which which launched the Global Safe Motherhood Initiative.37 Its l'attention sur cette problématique
included interviews with bilateral donors, national aims were to raise awareness of the roughly half a million
political officials, and national non-governmental yearly maternal deaths worldwide, nearly all of which
organisation leaders.1 Additionally, we undertook occurred in developing countries, to spark efforts to
archival research on the history of the initiative, address this problem, and to reduce maternal mortality
collecting and reviewing more than 70 documents from levels by half by the year 2000. After the conference, an
coordinating organisations for the initiative, donors, Inter-Agency Group for Safe Motherhood (IAG) formed
UN agencies, non-governmental organisations, and to focus global attention on the issue, bringing together
other actors. Furthermore, we consulted published work several international agencies and non-governmental
on maternal mortality and the safe motherhood organisations associated with the launch. Since this
initiative. Once we had collected the information, we conference two decades ago, safe motherhood advocates
reviewed the interview transcripts, documents, and have engaged in a sustained effort to convince
published work to analyse factors that shaped political international organisations and national political leaders
priority for the initiative. One of our aims was to capture to prioritise maternal mortality reduction.
the perspectives of global safe motherhood actors Nearly all respondents perceive the initiative’s results
themselves on the state of political priority for the in terms of production of political support to have been
initiative. We therefore draw heavily on excerpts from disappointing. Many published accounts on the initiative
the interviews in presentation of our data. We provided also reached this same conclusion.29,38,39 However, several
a draft of the paper to many of the interviewees, and respondents believe that the 20th anniversary year of the Anniv qui permet de remettre en
incorporated feedback that they provided. We also initiative will offer an opportunity to generate political avant la problématique

distributed and presented the paper at a consultation momentum for safe motherhood.
organised by the MacArthur Foundation in May, 2007,
on global safe motherhood strategy, which was attended Actor power
by 24 individuals connected to initiative. We revised the Since the start of the initiative, the policy community has commu politiques ne se sont pas
draft on the basis of feedback from several participants been divided over intervention strategy, which has affected mise d'accord sur les solutions
who attended the meeting. its credibility with international and national political
We do not take a position on the technical debates leaders (interview [I] 3 [May, 2006]; I9 [June, 2006];
surrounding appropriate intervention and measurement I15 [June, 2006]). Throughout the 1970–80s antenatal risk
strategies that have emerged in the initiative.28–30 Instead, screening and the training of traditional birth attendants
our aim was to examine the content of these debates and formed the core strategies for maternal survival.28 An
how they were understood by the participants themselves influential article in The Lancet in 198540 expressed concern Article (dans journal connu) a
to assess the effect the debates had on political support about these strategies, arguing that most maternal deaths influencé dans le mauvais sens ->
pas en faveur de résolution du
for the initiative. Our focus is on the initiative’s global could not be prevented and that women needed access to problème
level rather than national or grassroots level actors and emergency obstetric care in the event of complications at
debates, except in instances when national and grassroots childbirth. Although they do not deny the need for
organisations have affected or been affected by the global emergency obstetric care, other advocates emphasised
safe motherhood initiative. National and community that its importance had been exaggerated (I2 [June, 2005])
experiences, and the perspectives and voices of actors and that community-level and preventative activities had
from developing countries, are crucial to the history of crucial roles in the prevention of maternal deaths. Some
safe motherhood and deserve research attention.31–36 This advocates supported what has been termed the skilled
case study, however, limits itself to examining attendance approach. Its core message, expressed at safe
developments surrounding the global safe motherhood motherhood’s tenth anniversary conference in 1997, in
initiative. Sri Lanka, was to ensure skilled attendance at delivery,
A limitation of the method of our case study was the defined as “having a health worker with midwifery skills
difficulty in controlling for confounding variables of present at childbirth, backed up by transport in case
influence, and in assessment of the relative causal weight emergency referral is required”.41
of factors that we identified as shaping political priority. Participants in the debates suggest that the
As we note above, this study should be taken as an initial disagreements were more than technical; they took on a
exploration of the complex question of issue ascendance personal quality. One respondent commented:
and neglect in global health. Substantially more research, “[People became] extremely defensive about their ideas…
ideally comparing global health initiatives varying in If you didn’t agree with the idea you were bad and
political support levels, will be necessary to establish wrong…It was kind of like President Bush. If you are
which factors are most and least influential in shaping against this idea then you are a traitor”
political priority. I2 [June, 2005]

www.thelancet.com Vol 370 October 13, 2007 1373

 Health Policy

Another participant described the whole history of the WHO, respectively). However, such agency leadership
initiative to be “one of competing camps” (I3 [May, in the UN system never developed for safe
2006]). A third participant believed that positions became motherhood.30 UNICEF, UNFPA, WHO, and the UN
Difficulté de consensus entre deeply entrenched and nearly immutable: Development Program (UNDP) all developed safe
parties prenantes motherhood activities, which were often run inde-
“We always know the answer. First traditional birth
attendants and antenatal care, then that doesn’t work so pendently of one another. At some points, the agencies
skilled attendants and emergency obstetric care… There were antagonistic, differing on intervention approach
is view bias. [You must] take the company line” and competing for scarce safe motherhood resources
I6 [May, 2006] (I7 [June, 2006]; I10 [June, 2006]; I11 [June, 2006]; I17 [July,
2006]). One respondent from the UN involved in safe
A 2006 Lancet series on maternal survival called for motherhood believed that the core problem was an
deliveries to be attended by midwives in health centres, absence of UN agency ownership:
with other medical professionals present and higher “UNICEF was involved but children are its bread and
levels of care available if needed.42 The series indicated a butter…UNFPA was neither here nor there…It had
consensus that had been building gradually among some advocacy and policy but not programs. The WHO
members of the policy community about the need to balances between norms and standards and
have both skilled attendants at birth and emergency implementation—back and forth—it deals with many
obstetric care if needed. Many respondents noted a things. So safe motherhood doesn’t have a home in the
United Nations and that’s a big problem”
substantial decrease in tension in the policy community,
I10 [June, 2006] des communauté politiques influentes étaient
partly because of this emerging consensus (I4 [May, censées prendre la parole pour cette problématique
mais ne l'ont pas fait.
2006]; I5 [May, 2006]; I6 [May, 2006]; I7 [June, 2006]; Between 2002 and 2005, pressured by donors who did
I8 [June, 2006]; I15 [June, 2006]). However, this agreement not want to fund separate efforts (I7 [June, 2006]), the modificati
on de la
did not encompass all members of the policy Safe Motherhood Initiative and the IAG gradually présentati
on du
community. Some expressed strong concern about what merged with other initiatives to become a broader problème
to do in the interim, before such facilities could be partnership for maternal, newborn, and child health. : la santé
des mère
established, in view of resource scarcity and the difficulty The idea of continuum of care was intuitively appealing n'était un
that poor countries faced in expanding care.43 to some organisations and actors involved, since it problème
différent
Weak guiding institutions hindered the acquisition of sought to ensure that the health of newborns, children, de la
santé
political support. The IAG grouped effective individual and mothers would be promoted in a synergistic way mais l'un
advocates for safe motherhood and well-respected (I18 [July, 2006]; I23 [September, 2006]). The idea and influence
l'autre ->
researchers. However, it included technical officials in birth of the partnership were contentious, however; its ce qui a
inscité
the represented agencies rather than their senior leaders, leaders have had to manage tensions between its des
hampering its ability to develop global political support members since its launch (I3 [May, 2006]; I5 [May, 2006]; commu
politique
for maternal mortality reduction.39 Furthermore, I7 [June, 2006]; I9 [June, 2006]; I15 [June, 2006]; I18 [July, à
s'investir
controversy surrounded membership—IAG members 2006]; I23 [September, 2006]). The alliance between safe ++
made a deliberate decision initially to remain small. One motherhood and newborn survival has been uneasy, and
founding member explained the rationale for this: that between safe motherhood and child survival even
“A lot of groups wanted to be in and the IAG was not
more so. One point of contention concerns different
perfect but worked well together…The group felt strongly perspectives on facility versus community or home-based
that smaller was better. It was easier when dealing with delivery. Additionally, many advocates of child survival
difficult issues [such as] abortion” have been wary of politically contentious discussions
I21 [August, 2006] surrounding unsafe abortion, which is a leading cause
of maternal mortality. Underlying these tensions has
Another initial IAG member commented that this been a concern over the division and control of scarce
decision created difficulties, noting that, “we were resources. Commenting on the place of safe motherhood
accused, rightly so, of being a small inside group” amidst these partnership tensions, one respondent said:
(I23 [September, 2006]). A non-member articulated one “There are three siblings. Child survival is older, richer,
of the consequences of this decision: more resourceful. The newborn is weak, small, but got a
“The IAG was not and never was perceived as a strong new grant from Gates for US$60 million. It is the small
mouthpiece for safe motherhood until much later” child in the family that everyone looks to. Safe
I9 [June, 2006] motherhood is the middle child; it doesn’t know exactly
where to be. We need a good parent to take care of the
three equally, or unequally—safe motherhood needs
Another issue for guiding institutions, one that the
more vigorous opportunities”
IAG was not able to resolve, concerned relations between I10 [June 2006] Explique que les soins à la maternité sont liés aux
UN agencies. For other health issues such as child problématiques liées à l'enfance , mais est pourtant oublié

survival, family planning, and technical advice, a clear Supporters of the partnership argue that the cause of
UN agency took the lead (UNICEF, UNFPA, and maternal survival rightly belongs under its fold, integrated

1374 www.thelancet.com Vol 370 October 13, 2007

 Health Policy

with child and newborn health (I23 [September, children.45,46 Despite these efforts, no frame has convinced
2006]). However, several safe motherhood advocates are many political leaders, which is a situation that continues
suspicious, wondering whether the partnership ultimately to puzzle several members of the policy community. As
will serve the cause of maternal survival (I3 [May, 2006]; one respondent states:
I9 [June, 2006]; I16 [June, 2006]; I17 [July, 2006]). One “Why is it like this? Why have we not seen the flow of
advocate notes that with the emergence of the partnership, resources to do something about it? The issue has all the
many members of the safe motherhood community are no emotional appeal on so many different levels. The case
longer sure if an initiative for safe motherhood still exists. can be made economically, in terms of household
Weak mobilisation of civil society organisations has also productivity, the next generation of children. What more
hindered the acquisition of political support. In 1999, the do you need to capture the imagination? I don’t
understand it”
White Ribbon Alliance formed with the aim of promoting
I15 [June, 2006]
cross-national advocacy for safe motherhood, linking civil
society institutions with donor and other organisations. The initiative’s initial positioning could have contributed
However, few grassroots organisations concerned with to difficulties in attracting political support from one group
the global dimensions of the issue have emerged. One that might otherwise have lent powerful backing to the
reason could be the absence of access to the political cause: the women’s movement. When World Bank officials
process on the part of those most directly affected by this first came up with the idea for an international conference
issue—ie, poor women with little education, who face for maternal mortality in 1985, they were conscious of the
substantial gender discrimination in many poor countries. unwillingness of the US administration to support family
Several respondents note also the initiative’s detachment planning internationally. Sensitive to the US adminis-
from grassroots activities that do exist (I9 [June, 2006]; tration’s conservatism regarding reproductive issues, Bank
I10 [June, 2006]; I15 [June, 2006]). One respondent, officials wanted to find, in the words of one of the
commenting on how far removed policy community organisers, “a concept that is politically unassailable, a
debates were from local realities, said: name that brings in money, that makes a lot of people
“There’s a huge disjuncture. 35 000 feet discussions. And
heroes, something the American administration cannot
I’m worried the gap is getting bigger. International oppose” (I22 [August, 2006]). From these deliberations, the
discussions are devoid of reality on the ground” term safe motherhood was borne. The organisers coalesced Mot qui devait ê porteur de sens pour la
cause
I15 [June 2006] surrounding the term, partly because they thought that it
was unlikely to incite active opposition from the
A widely embraced leader could have helped surmount administration. However, as one respondent put it:
historical difficulties of policy community fragmentation, “The feminists didn’t like the term ‘safe motherhood’ so Mot qui ne fait pas consensus, donc qui
weak guiding institutions, and little civil society [the issue] was never picked up by women’s groups” n'est pas fédérateur
mobilisation. However, no such figure emerged (I5 [May, I7 [June, 2006]
2006]; I15 [June, 2006]). Several individuals associated
with the initiative at particular junctures were perceived The same respondent notes that many men, too, may Beaucoup d'hommes (et notamment
les hommes qui font les politiques) ne
potentially to have this capacity, but they did not take up have been reluctant to engage the issue, but for a different sont pas touché par cette
the opportunity. Referring to child survival’s best-known reason, arguing: problématique. Notamment, car ça
touche à la vie sexuelle de la femme.
leader, one respondent surmised: “The neglect of women’s issues…does reflect some level of
“Safe motherhood doesn’t have a Jim Grant. Where’s the unconscious bias against women at every level, from the
ambassador?” Manque un leader, un porte-parole community to high-level decisionmakers…While we may
I15 [June, 2006] ignore it, maternal health does involve sex and sexuality; it
is bloody and messy; and I think many men (not all, of
Ideas course) have a visceral antipathy for dealing with it”
Finding a resonating set of ideas—ie, positioning the issue
publicly in ways that attract political support—has been a The framing of the issue inside the policy community internal frame
persistent challenge for the initiative. Since the initiative’s also has posed difficulties. Fundamentally, the community
launch, several political leaders in developing countries has united with a shared belief that maternal mortality is
have come to perceive maternal mortality as an issue that a neglected tragedy that demands redress. This idea was
is deserving of attention and resources.33,36,44 However, few the source for an article that helped bring international
leaders have prioritised maternal mortality, especially attention to the issue. The 1985 piece in The Lancet by
compared with the many national leaders that have Allan Rosenfield, regarded by many to be the cause’s
prioritised issues such as child survival and HIV/AIDS. most effective champion, and Deborah Maine,40
Safe motherhood advocates have made concerted emphasised the neglect of maternal survival in favour of
efforts to develop frames for the issue that might child health. However, beyond this core point of
resonate. They have emphasised the severity of the agreement, the policy community until recently has had
issue, made rights-based arguments, connected the issue difficulty identifying common ideas. An internally
to economic outcomes, and noted the effects on resonating frame would need clear answers to several

www.thelancet.com Vol 370 October 13, 2007 1375

 Health Policy

issues, none of which the policy community has yet to within the UN system) has hampered the capacity of the
resolve in full. These issues include whether maternal initiative to create and take advantage of opportunities.
mortality or maternal health more broadly is the focal On the positive side, the UK Department for International
concern; how progress should be measured; whether the Development (DFID), influenced by the MDGs, has
continuum of care idea is embraced as the core increased maternal and newborn health funding from
positioning of the issue; the precise strategies to address GB£0·9 million in 2001–02 to £16·2 million
the problem; and the relation of the initiative to other in 2005–06.48 Other donors also increased funding for
health concerns, including family planning, the broader maternal survival during this time.48 Furthermore, the
reproductive health agenda, and health systems MDG commitment stood behind several substantial
development. global calls for action, including a declaration in 2005 in
New Delhi, India, from UN agency heads and many
Political contexts developing country governments calling for global
Advocates have sought to build a favourable global progress on maternal, newborn, and child health.49
les commu politique ont political environment for safe motherhood by organising Additionally, the leaders of the countries that rank
créées un environnement
politique propice
international meetings and events for maternal mortality number one and two in terms of numbers of maternal
reduction, seeking inclusion of the issue in other global death, India’s Prime Minister Manmohan Singh and
meetings, and aiming to take advantage of policy windows Nigeria’s former President Olusegun Obasanjo,
such as those associated with the MDGs (I7 [June, 2006]; commented publicly on the maternal health MDG. They
I18 [July, 2006]; I19 [August, 2006]; I21 [August, 2006]; I22 expressed concern about their countries’ high levels of
[August, 2006]; I23 [September, 2006]). The effect of their maternal death, and demanded that their governments
efforts is uncertain. AbouZahr has argued in a review of act to address the issue.35–36
Malgré effort pour créer the history of the initiative that “these efforts have lacked Conversely, several studies show continuing large gaps Donc, il y
fenêtre politique ceci n'a a eu des
pas fonctionné conviction”, noting that, “safe motherhood meetings in global resources for maternal health.50–52 One study efforts
tend not to attract the most senior decisionmakers.”39 estimated that an additional US$1 billion was needed to mais pas
suffisame
The first efforts to promote maternal mortality reduction meet maternal and newborn health needs in 2006,53 and nent
took place in 1985, before the launch of the initiative, at another identified a need for a minimum yearly average
the end of the UN Decade for Women, when advocates increase of $3·9 billion over 10 years to meet combined
for women’s rights identified maternal mortality reduction maternal and newborn health needs.52 Furthermore, only
as one of several issues that might sustain the women’s a few developing countries have made maternal mortality
agenda (I21 [August, 2006]). The 1987 Nairobi conference reduction a political priority since the initiative’s launch.1
was the first major international event for safe The initiative’s 20th anniversary year, 2007, could
motherhood. Safe motherhood also made the agenda of present new opportunities for generation of political
the Third International Conference on Population and priority for safe motherhood. In February, 2007,
Development in Cairo in 1994, and the Fourth World IMMPACT (a maternal mortality research initiative)
Conference on Women in Beijing in 1995.47 In 1997, a disseminated results from several years of studies on
conference in Colombo, Sri Lanka, marked the tenth measurement strategies and programme assessment. In
anniversary of the initiative, and in 1999, new initiatives September, 2007, an initiative led by the Norwegian
and programmes formed for maternal mortality government to accelerate progress towards the
reduction, including Columbia University’s Averting achievement of the child and maternal survival MDGs
Maternal Death and Disability Program. This programme was launched. The culminating event is the Women
received US$50 million from the Gates Foundation, Deliver conference, which will be held in October,
which is the largest grant so far for safe motherhood. 2007. Heads of state, ministers of planning and finance,
A policy window—ie, a favourable confluence of events and other senior political officials have made
providing an opportunity for advocates to press political commitments to attend this event, creating the potential
leaders—opened for safe motherhood in 2000. In that for the meeting to bring about the high-level political
year, UN member states announced the MDGs, a set of support for the issue which has previously been lacking.
poverty alleviation objectives for the year 2015. Maternal
health was one of a select group of health goals to make Issue characteristics
the MDGs, with goal number five being the reduction of Three characteristics of the issue itself have made
the global maternal mortality ratio by 75% over 1990 levels attracting political support for maternal mortality
by the year 2015. difficult. First, maternal deaths are not as common as are
Whether the maternal health MDG, efforts by advocates those caused by several other high-burden disorders (eg,
to take advantage of the MDG, and continual efforts by HIV/AIDS, malaria); second, accurate measurement of
global safe motherhood advocates have had substantial maternal mortality is technically difficult; and third, the
effect on political support and resources is uncertain. A interventions to avert maternal death are not as simple as
weak global governance structure for health (including are those for some other disorders (such as several
absence of leadership on the issue of safe motherhood children’s diseases that are preventable by vaccine).

1376 www.thelancet.com Vol 370 October 13, 2007

 Health Policy

The most recent estimate of the number of annual These problems notwithstanding, safe motherhood
maternal deaths globally is 529 000 for the year 2000.54 advocates have used estimates, however imprecise, to
Although this figure is high, it is much lower than the generate attention for the cause. For example, in the
annual number of deaths from HIV/AIDS (2·9 million), mid-1980s, Robert Cook from WHO sponsored studies Communauté politique
tuberculosis (1·6 million), and malaria (1 million), as well that produced the first estimate of global maternal Mesure chiffrée
as the number of deaths to children younger than 5 years mortality levels: half a million maternal deaths annually.65
(10·6 million), and neonates (4 million).55–59 Many This figure drew the attention of international agency focus par +ieurs communauté
politique
advocates for safe motherhood are acutely aware of these heads and others key to the initiative’s inception.30,39
figures (I5 [May, 2006]; I15 [June, 2006]) and have debated Revised estimates from UN agencies in the mid-1990s
whether maternal health advocacy should instead provoked discussion between national leaders and UN suscite intérêt des leaders
emphasise other indicators of severity, such as morbidity, agencies surrounding the issue, which might have
lifetime risk of death, or combined maternal and newborn increased maternal mortality’s profile with donor
deaths (which total 4·5 million). institutions.39
Ci-dessus,
dit que c'est Additionally, maternal mortality is more difficult to
pas si
fréquent
measure than are other health outcomes such as infant Conclusions
lais en mortality, child mortality, and fertility.60 The fairly low Factors shaping global political priority for safe
même
temps numbers of maternal deaths in any specific geographic motherhood
difficulté de locality mean that confidence intervals for estimates If we consider all the four categories of factors that affect
le mesurer.
Donc pas sifrom most survey methodologies are large, making the acquisition of global political support, we see that the
sûr que si
peu
actual levels difficult to ascertain and whether change safe motherhood initiative has had many difficulties. With
fréquent occurred across time difficult to establish. These respect to actors, the global policy community has been
par rapport
à d'autres challenges have led some researchers and programmers fragmented, no powerful institutions have emerged to
m+ to turn to process indicators to assess effect and measure guide the initiative, and organisational rivalries have
progress. Other researchers, although not denying the persisted throughout its history. Additionally, although
need for such indicators, emphasise the importance of the initiative included highly capable individuals, it never
continuing to find better ways to measure maternal found a recognised leader. Furthermore, the primary
mortality itself.61 victims of maternal mortality (poor women in the
Another difficulty is that the interventions necessary to developing world) have little political power and are
prevent maternal death are not as simple as are those for disadvantaged by gender inequalities, and civil society
other disorders, such as specific diseases that are mobilisation to make this cause a global priority has been
preventable by vaccine. Few of the leading maternal weak.
health epidemiologists believe that a simple solution is With regard to ideas, the global policy community has
available, and most argue that functioning health systems not yet established an internally resonating frame, and
are crucial.42,62 Disagreement exists about the actual still struggles to find external frames that will move
degree of complexity of the necessary interventions, the political leaders to action. With respect to political
strength of the evidence base for these interventions, and context, global policy windows have opened, but how
their cost (I2 [June, 2005]; I3 [May, 2006]; I6 [May, 2006]; well the policy community has taken advantage of these
I7 [June, 2006]; I12 [June, 2006]; I15 [June, 2006]).63,64 opportunities is unclear. The fragmented global structure
Several respondents suggest that the sometimes of governance for health has made an institutional home
contentious public discussions surrounding measurement for safe motherhood difficult to find. With respect to
and evidence have had adverse effects on the initiative’s issue characteristics, the severity of the problem is low
ability to acquire political support and resources, and have compared with other conditions if indicated by mortality
contributed to policy community fragmentation: levels, hampering resource acquisition and mobilisation
“We hang out our dirty washing. Other people are more
efforts. Measurement is a continuing problem, and no
discreet” widely accepted and simple way to monitor progress has
I15 [June, 2006] emerged. Interventions are not simple, the evidence
base for these interventions is weaker than it is for some
“We don’t know what’s effective. We can’t measure other issues, and the policy community has had trouble
outcomes very well”
developing consensus on which interventions should be
I3 [May, 2006]
prioritised. Because of these difficulties caused by both
“We focus on uncertainties. That is the truth but it will the nature of the issue and the decisions of actors, the
not convince the Minister of Finance” safe motherhood initiative remains in a state of infancy
I15 [June, 2006] even after 20 years.
“I would go with my ideas [to a donor] and [X] would go
with hers and who was to say who was correct” Building global political priority for safe motherhood
I9 [June, 2006] 2007 could present a window of opportunity to generate
political support for the cause. Cohesion is building in
Montre le manque de consensus

www.thelancet.com Vol 370 October 13, 2007 1377

 Health Policy

the policy community as proponents of skilled attendance 2 Keck ME, Sikkink K. Activists beyond borders: advocacy networks
and emergency obstetric care bridge their differences. A in international politics. Ithaca and London: Cornell University
Press, 1998.
major policy window has opened, prompted by the MDGs 3 Finnemore M, Sikkink K. International norm dynamics and
and the 20th anniversary of the initiative. Additionally, political change. Int Organ 1998; 52: 887–917.
the Partnership for Maternal, Newborn and Child Health 4 Finnemore M, Sikkink K. Taking stock: the constructivist research
program in international relations and comparative politics.
could provide a new coordinating mechanism for global Annu Rev Polit Sci 2001; 4: 391–416.
leadership on the issue. 5 Haas PM. Introduction: epistemic communities and international
Creation of political momentum will need four policy coordination. Int Organ 1992; 46: 1–35.
connected political challenges to be addressed, in addition 6 Florini AM. The third force: the rise of transnational civil society.
Washington, DC: Carnegie Endowment for International Peace,
to continuing technical challenges surrounding 2000.
intervention and measurement. First is building on the 7 McAdam D, McCarthy JD, Zald MN. In: McAdam D, McCarthy JD,
growing cohesion in the policy community so that it can Zald MN, eds. Preface. Comparative perspectives on social
movements: political opportunities, mobilizing structures, and
speak with authority and unity to international and cultural framings. Cambridge: Cambridge University Press, 1996.
national political leaders. Second is the creation of 8 Tarrow S. Power in movement: social movements, collective action,
enduring guiding institutions to sustain the initiative. The and politics. 2nd edn. New York: Cambridge University Press, 1998.
partnership might provide a platform, but in view of its 9 Johnston H, Noakes JA. Frames of protest: social movements and
the framing perspective. Lanham, MD: Rowman & Littlefield
recent creation and tensions in the policy communities Publishers, 2005.
that compose it, assessment cannot yet be made. Third is 10 Snow DA, Rochford EB Jr, Worden SK, Benford RD. Frame
finding external frames that resonate and will convince alignment processes, micromobilization, and movement
participation. Am Sociol Rev 1986; 51: 464–81
political leaders that they should be concerned about the 11 Sabatier PA. The advocacy coalition framework: revisions and
issue. Policy community members have offered several relevance for Europe. J Eur Public Pol 1998; 5: 98–130.
ideas on framing, but none has taken hold widely. Fourth 12 Kingdon JW. Agendas, alternatives and public policies. Boston:
Little, Brown and Company, 1984.
is building stronger links with national initiatives and
13 Baumgartner FR, Jones BD. Agendas and instability in American
mobilising country-level civil society organisations. The politics. Chicago: The University of Chicago Press, 1993.
weakness of such links and minimal social mobilisation 14 Buse K, Mays N, Walt G. Making health policy. Maidenhead: Open
for the cause in countries with high maternal mortality University Press, 2005.
15 Reich, MR. The politics of health sector reform in developing
has hampered the acquisition of global political support. countries: three cases of pharmaceutical policy. Health Policy 1995;
32: 47–77.
Future research on political priority for global health 16 Brugha R, Varvasovszky Z. Stakeholder analysis: a review.
Health Policy Plan 2000; 15: 239–46.
initiatives
17 Harris PG, Siplon PD. The global politics of AIDS. Colorado: Lynne
The main question is why do some global health initiatives Rienner Publishers, 2007.
attract political priority whereas others remain neglected? 18 Berman S. Review: ideas, norms and culture in political analysis.
Our investigation, grounded in a synthesis of research on Comp Polit 2001; 33: 231–50.
collective action and one case study, represents only a 19 Prins G. AIDS and global security. Int Aff 2004; 5: 931–52.
20 Leichter HM. A comparative approach to policy analysis: health care
start in examination of this complex issue. Further study policy in four nations. Cambridge: Cambridge University Press,
and refinement of the framework is needed, ideally 1979.
through comparisons across global health initiatives that 21 Grindle M, Thomas J. Public choices and policy change: the
political economy of reform in developing countries. Baltimore:
vary in levels of political support. Such studies would have Johns Hopkins University Press, 1991.
great value both for theory development and for practice. 22 Walt G, Gilson L. Reforming the health sector in developing
Empirically-grounded explanations on issue ascendance countries: the central role of policy analysis. Health Policy Plan 1994;
9: 353–70.
and neglect would advance our knowledge of dynamics
23 Gourevitch, P. The second image reversed: the international
for agenda setting in global health. Equally importantly, sources of domestic politics. Int Organ 1978; 32: 881–912.
they would offer guidance to struggling global health 24 Walt, G. Global cooperation in international public health. In:
initiatives such as safe motherhood on how to generate Merson MH, Black RE, Mills AJ, eds. International public health.
Gaithersberg, MD: Aspen Institute, 2001.
much needed political support. 25 Lee K, Collinson S, Walt G, Gilson L. Who should be doing what in
Acknowledgments international health: a confusion of mandates in the
We thank the MacArthur Foundation for funding this study; the Center United Nations? BMJ 1996; 312: 302–07.
for Global Development for providing a visiting fellow position during 26 Stone, DA. Causal stories and the formation of policy agendas.
which research was undertaken and results written up; the many Polit Sci Q 1989; 104: 281–300.
individuals who agreed to be interviewed for this research, who were 27 Yin R. Case study research: design and methods. 2nd edn.
generous and candid with their insights, information, and time, and Thousand Oaks, California: Sage, 1994.
without whose openness this study would not have been possible; and 28 De Brouwere V, Tonglet R, Van Lerberghe W. Strategies for
Ben Elberger for his thoughtful comments on drafts, and his research reducing maternal mortality in developing countries: what can we
support for this work. The views presented represent those of the learn from the history of the industrialized West?
authors alone, who take responsibility for all errors. Trop Med Int Health 1998; 3: 771–82.
29 Maine D, Rosenfield A. The safe motherhood initiative: why has it
References stalled? Am J Public Health 1999; 89: 480–02.
1 Shiffman J. Generating political priority for maternal mortality 30 Starrs AM. Safe motherhood initiative: 20 years and counting.
reduction in 5 developing countries. Am J Public Health 2007; Lancet 2006; 368: 1130–32.
97: 796–803.

1378 www.thelancet.com Vol 370 October 13, 2007

 Health Policy

31 Hay C. Dying mothers: Maternal mortality in rural Indonesia. 49 The Delhi Declaration on Maternal, Newborn, and Child Health.
Med Anthropol 2000; 18: 243–79. Delhi, 2005. http://www.internationalmidwives.org/modules/
32 Shiffman J. Generating political will for safe motherhood in ContentExpress/img_repository/Delhi_Declaration_FINAL_10_
Indonesia. Soc Sci Med 2003; 56: 1197–207. April_05.pdf (accessed Dec 1, 2006).
33 Shiffman J, Stanton C, Salazar AP. The emergence of political 50 Borghi J, Ensor T, Somanathan A, Lissner C, Mills A, on behalf of
priority for safe motherhood in Honduras. Health Policy Plan 2004; the Lancet Maternal Survival Series steering group. Mobilising
19: 380–90. financial resources for maternal health. Lancet 2006; 368: 1457–65.
34 Shiffman J, Garcés Del Valle AL. Political history and disparities in 51 Powell-Jackson T, Borghi J, Mueller DH, Patouillard E, Mills A.
safe motherhood between Guatemala and Honduras. Popul Dev Rev Countdown to 2015: tracking donor assistance to maternal,
2006; 32: 53–80. newborn, and child health. Lancet 2006; 368: 1077–87.
35 Shiffman J, Okonofua FE. The state of political priority for safe 52 Johns B, Sigurbjörnsdóttir K, Fogstad H, Zupan J, Mathai M.
motherhood in Nigeria. Br J Obstet Gynaecol 2007; 114: 127–33. Estimated global resources needed to attain universal coverage of
36 Shiffman J, Ved RR. The state of political priority for safe maternal and newborn health services. Bull World Health Organ
motherhood in India. Br J Obstet Gynaecol 2007; 114: 785–90. 2007; 85: 256–63.
37 Starrs, A. Preventing the tragedy of maternal deaths: a report on the 53 WHO. World health report 2005: make every mother and child
International Safe Motherhood Conference. New York: Family Care count. Geneva: World Health Organization, 2005. http://www.who.
International, 1987. int/whr/2005/en/index.html (accessed Feb 1, 2007).
38 Weil O, Fernandez H. Is safe motherhood an orphan initiative? 54 WHO. Maternal mortality in 2000: estimates developed by WHO,
Lancet 1999; 354: 940–43. UNICEF, and UNFPA. Geneva: World Health Organization, 2004.
39 AbouZahr C. Cautious champions: international agency efforts to 55 UNAIDS. AIDS epidemic update: December, 2006. Geneva:
get safe motherhood onto the agenda. Stud Health Serv Organ Pol UNAIDS, 2006.
2001; 17: 384–411. 56 WHO. Fact sheet number 104: tuberculosis. Geneva: World Health
40 Rosenfield A, Maine D. Maternal mortality—a neglected tragedy. Organization, 2007.
Where is the M in MCH? Lancet 1985; 2: 83–85. 57 Roll Back Malaria. World Malaria Report 2005. Geneva: Roll Back
41 Starrs A, on behalf of the Safe Motherhood Inter-Agency Group. Malaria, WHO, UNICEF, 2005. http://www.rbm.who.int/wmr2005/
The safe motherhood action agenda: priorities for the next decade. (accessed Feb 1, 2007).
New York: Family Care International, 1998. 58 Bryce J, Boschi-Pinto C, Shibuya K, Black RE, the WHO Child
42 Campbell OM, Graham WJ, on behalf of The Lancet Maternal Epidemiology Reference Group. WHO estimates of the causes of
Survival Series steering group. Strategies for reducing maternal death in children. Lancet 2005; 365: 1147–52.
mortality: getting on with what works. Lancet 2006; 368: 1284–99. 59 Lawn JE, Cousens S, Zupan J. 4 million neonatal deaths: When?
43 Costello A, Azad K, Barnett S. An alternative strategy to reduce Where? Why? Lancet 2005; 365: 891–900.
maternal mortality. Lancet 2006; 368: 1477–79. 60 Stanton C, Abderrahim N, Hill K. DHS maternal mortality
44 Campbell O, Gipson R, Issa AH, et al. National maternal mortality indicators: an assessment of data quality and implications for data
ratio in Egypt halved between 1992–93 and 2000. use. Calverton: Macro International, 1997.
Bull World Health Organ 2005; 83: 462–71. 61 Graham WJ. Now or never: the case for measuring maternal
45 Freedman LP. Using human rights in maternal mortality programs: mortality. Lancet 2002; 359: 701–04.
from analysis to strategy. Int J Gynaecol Obstet 2001; 75: 51–60. 62 Ronsmans C, Graham WJ, on behalf of The Lancet Maternal
46 Filippi V, Ronsmans C, Campbell OM, et al. Maternal health in poor Survival Series steering group. Maternal mortality: who, when,
countries: the broader context and a call for action. Lancet 2006; where, and why. Lancet 2006; 368: 1189–200.
368: 1535–41. 63 Miller S, Sloan NL, Winikoff B, Langer A, Fikree FF. Where is the
47 Campbell O. What are maternal health policies in developing “E” in MCH? The need for an evidence-based approach in safe
countries and who drives them? A review of the last half-century. motherhood. J Midwifery Womens Health 2003; 48: 10–18.
Stud Health Serv Organ Pol 2001; 17: 415–45. 64 Paxton A, Maine D, Freedman L, Fry D, Lobis S. The evidence for
48 Department for International Development. DFID’s maternal emergency obstetric care. Int J Gynaecol Obstet 2005; 88: 181–93.
health strategy—reducing maternal deaths: evidence and action, 65 WHO. Maternal health and safe motherhood programme: progress
second progress report. London: Department for International report 1987–1990. Geneva: World Health Organization, 1990.
Development, 2007.

www.thelancet.com Vol 370 October 13, 2007 1379