Medical society engagement in contentious policy reform: the Ethiopian Society for Obstetricians and Gynecologists (ESOG) and Ethiopia's 2005 reform of its Penal Code on abortion.

Unsafe abortion is one of the three leading causes of maternal mortality in low-income countries; however, few countries have reformed their laws to permit safer, legal abortion, and professional medical associations have not tended to spearhead this type of reform. Support from a professional association typically carries more weight than does that from an individual medical professional. However, theory predicts and the empirical record largely reveals that medical associations shy from engagement in conflictual policymaking such as on abortion, except when professional autonomy or income is at stake. Using interviews with 10 obstetrician-gynaecologists and 44 other leaders familiar with Ethiopia's reproductive health policy context, as well as other primary and secondary sources, this research examines why, counter to theoretical expectations from the sociology of medical professions literature and experience elsewhere, the Ethiopian Society of Obstetricians & Gynecologists (ESOG) actively supported reform of national law on abortion. ESOG leadership participation was motivated by both individual and ESOG's organizational commitments to reducing maternal mortality and also by professional training and work experience. Further, typical constraints on medical society involvement in policymaking were relaxed or removed, including those related to ESOG's organizational structure and history, and to political environment. Findings do not contradict theory positing medical society avoidance of socially conflictual health policymaking, but rather identify how the expected restrictions were less present in Ethiopia, facilitating medical society participation. Results can inform efforts to encourage medical society participation in policy reform to improve women's health elsewhere in sub-Saharan Africa.

Key Messages

The ESOG’s reform contributions included building a research base, conducting evidence-based advocacy and framing the rationale for reform as grounded in public health and maternal mortality prevention.

Personal and organizational commitments to reducing maternal mortality, including that due to unsafe abortion, and experience with post-abortion care (PAC), motivated the participation of ESOG in the country’s 2005 reform of its law on abortion.

Political openings, including a period of relative democratic openness and the vehicle of Penal Code reform, facilitated ESOG’s participation in policymaking.

Civil society allies supported and enhanced ESOG’s contributions to reform efforts.

Introduction

What underlies the willingness of a professional medical society to actively support policy reform, particularly socially contentious reform? In 2005, Ethiopia enacted a broad reform of its Penal Code that liberalized law on abortion. Second only to South Africa and Mozambique, Ethiopia arguably now in practice has one of the more liberal laws on abortion in sub-Saharan Africa (SSA) (Singh ). The Ethiopian Society of Obstetricians–Gynecologists (ESOG) publicly supported this reform. Rather than avoiding policy engagement, ESOG leadership researched and articulated the problem of abortion-related maternal mortality; conveyed it to the elite public and to government policy makers; proposed policies; and later helped develop regulations that further expanded access by making new categories of health professionals eligible to provide services (Ministry of Health 2006, 2014). ESOG leadership’s active public role in support of reform runs counter to theoretical predictions and empirical evidence that professional societies only actively engage in policy making when it relates to their professional self-interest or when it does not threaten to generate conflict among members (Parsons 1951; Aries 2003; Halfmann 2003).

This study identifies three sets of explanations for ESOG’s surprising engagement in controversial policymaking. First, both individual obstetrician–gynaecologists (ob–gyns) and ESOG as an organization had strong, almost definitional commitments to maternal mortality prevention. Second, typical organizational constraints on medical society involvement in controversial policymaking were absent or relaxed at the time of the reform. Third, the political environment was conducive to ESOG policy engagement: governmental support, a propitious political moment, and NGO ally encouragement.

Background

Ethiopia’s Penal Code reform took place in the context of dauntingly high levels of maternal mortality: in 2005, 676 women died for every 100 000 live births, and unsafe abortion had long been one of the three leading causes of these deaths (WHO 2007; Åhman and Shah 2011; CSA and ICF 2012). However, Ethiopia’s cultural context would not lead one to expect the liberalization of abortion law. Major religious groups (Ethiopian Orthodox Christians, Muslims, Protestants) proscribe abortion except to save the life of the woman, and opposition to abortion among Ethiopians is also strong, as is common in non-industrialized, less secular countries (Inglehart and Norris 2017). Over 63% of the Ethiopian population views abortion as ‘never justifiable’ (World Values Survey 2007). However, as is typical globally, women continue to obtain abortions despite social censure and even at times their own misgivings (Dixon-Mueller 1995).

ESOG and ob–gyns in Ethiopia

Ethiopian ob–gyns are a small and nascent group: fewer than 250 ob–gyns served between the 1950s and 2007 in Ethiopia, a country now of almost 100 million, and the first graduate training program was only established in 1980 (Gaym 2010). Prior to this, Ethiopians received ob–gyn training on an ad hoc basis from foreign ob–gyns inside the country or travelled abroad for training. Between 1974 and 1990, many Ethiopians went to Eastern European countries for graduate medical education (Gaym 2007). Thus their training was in countries where abortion services are integrated into the national health systems, and where a Marxist and more receptive, perspective on abortion and contraception prevailed (Tietze and Lehfeldt 1961). The ob-gyn professional society—ESOG—is also recent, founded in 1992 with an explicit mission to address Ethiopia’s elevated maternal mortality (Gaym 2007; ESOG 2010). Although ESOG grew from 75 founding members to 234 in 2012 (ESOG 2012), representing the 176 ob–gyns in practice in the country in 2007 (Gaym 2010), it remains a small organization with leadership historically serving on a volunteer basis.

Globally, as well-educated, more affluent, socially connected, and most often male health professionals, ob–gyns have significant social capital, and as such have been influential contributors to national reproductive health policy (Gasman ; Briozzo and Faundes 2008; Shaw 2014). Ethiopia is no exception. Ethiopian ob–gyns play at least a triple role: they provide medical care, train other providers and contribute to development of government guidelines and policies. They are the providers who address the most difficult cases, including those of women with severe complications of unsafe abortion who make it to medical facilities. They have high clinical autonomy due not only to their medical training, but also to national shortages of medical professionals (Kinfu ). As is the case in most SSA countries, the Ministry of Health relies on ob–gyns for technical advice and policy input.

Reform of national laws on abortion in Ethiopia and SSA

Liberalization of national abortion law, with its promise of access to legal and safer services, remains an infrequent event. Nonetheless, the global trend has been one of liberalization, even in SSA where laws are largely restrictive inheritances from colonial rulers (Cook and Dickens 1981; Finer and Fine 2013). However, reform of abortion law in SSA has been largely on limited grounds (to permit abortion to save the life of the woman, in cases of rape or incest, or to preserve the physical health of the woman). The region has the highest proportion of countries permitting abortion only to save the life of the woman (Singh ; Finer and Fine 2013). Only South Africa and Mozambique (as of 1996 and 2014, respectively) allow women access to abortion in the first trimester without legal restrictions.

Prior to Ethiopia’s 2005 reform, abortion was permitted only to ‘save the pregnant woman from grave and permanent danger to life or health’, and required approval by two doctors (Empire of Ethiopia 1957).

The final version of the revised 2005 Penal Code was not the complete decriminalization of abortion law initially proposed in Parliament, due to late-breaking public opposition from the Patriarch of the Ethiopian Orthodox Church (Wada 2008). Nonetheless, the revised Penal Code did include a lengthy list of exceptions, permitting abortion in the cases of rape, incest or foetal impairment; if pregnancy continuation or birth would endanger the health or life of the woman or foetus; if the woman had physical or mental disabilities; or if the woman was a minor physically or mentally unprepared for childbirth. Further, ‘mere statement from the woman’ was deemed sufficient proof of eligibility (Federal Democratic Republic of Ethiopia (FDRE) 2005), and regulations have further simplified access (Ministry of Health (MOH), FDRE 2006, 2014).

Theory and empirical evidence

On balance, both theory and historical experience suggest that medical societies would avoid involvement in socially contentious policy reform, in order to avoid conflict as well as to protect professional and economic interests and standing.

Theory on medical society behaviour with regard to sensitive policymaking would predict disengagement. Sociological study of the professions describes physicians as having a monopoly on a technical body of knowledge that then accords them professional autonomy, public respect and authority (Parsons 1951). The medical profession is seen as distinguished by its collective (rather than self-interested) orientation, and is expected to have a professional commitment to prioritize patient well-being over personal gain, thus creating trust and legitimizing physician authority (Parsons 1951, 1975). However, medical societies’ incentives are understood to be different from those of individual physicians, with societies having an incentive to avoid engaging in public policymaking if this risks provoking conflict for members (Parsons 1951, 1975).

More recent theoretical research on the role of the medical professions with respect to policy, largely in more affluent countries, has focused less on physicians’ collective social contract orientation, and has predicted rather that organized medicine will pursue professional self-interest through public policy (Frenk and Donabedian 1987; Duran-Arenas and Kennedy 1991; Harrison 1994). Professional self-interest is defined as providers’ economic interests and clinical autonomy, including control over how, when, and if services are delivered, and as a consequence, control over patients. Accordingly, theories of the professions would posit that a medical society would both resist expansion of the scope of practice of other medical professions (e.g. midwives, health officers or nurses) and try to retain authority over whether and when to provide abortion services to women (Mohr 1979; Aries 2003). Halfmann (2003) also finds that organizational priorities and historical experience shape medical societies’ policy participation and focus during abortion law reform.

While individual ob–gyns have been strong and effective voices in many national efforts to liberalize laws on abortion (Halfmann 2003; Hessini 2005; Gasman ; Briozzo and Faundes 2008), empirical research in both industrialized and low-income countries shows their medical societies not to have been at the forefront of socially contentious abortion law reform. Prior to the 1960s, physician’s medical associations in Western countries often sought to curtail women’s legal access to abortion in order to defend or build physician professional power vis-à-vis other medical cadres (Mohr 1979; Morland 1993; Scheuermann 1995; Joffe 2009). During the wave of abortion law reforms that took place in OECD countries starting in the late 1960s, few medical societies either called publicly for legal reform or were lead reform actors, instead many opposed reform as a challenge to physician power (Aries 2003; Halfmann 2003; Gasman ). If medical associations entered into the policy debates, they did so after recognizing the inevitability of reform, and engaged to preserve physician discretion over services (Aries 2003; Halfmann 2003). In Poland and Brazil, where public policy battles over abortion law have resurfaced, medical societies have largely remained on the sidelines (Faúndes ; Nowicka 2007; Chełstowska 2011; De Zordo and Mishtal 2011; de Assis Machado and Maciel 2017). Even in settings with high maternal mortality due to unsafe abortion, medical societies have largely been absent or more peripheral to reform efforts, such as in South Africa (Guttmacher ; Klugman 2008); Guyana (Nunes and Delph 1995); Nigeria (Oye-Adeniran ) and/or have largely only contributed to regulatory guidance after reform, as in Nepal (Shakya ). Similarly, none of the five West African ob–gyn societies participating in a FIGO initiative to prevent unsafe abortion included reform of their national laws and policies on abortion as part of their action plans (Leke 2014).

In the few cases where ob–gyns societies have publicly supported reform early on, there has either already been clear government support for reform and/or the reform has left authority over the procedure in the hands of physicians, as in India (Jesani and Iyer 1993; Hirve 2004) and Uruguay (Briozzo 2016; Wood ); or where prominent ob–gyns have come under legal attack, as in Kenya (Mbugua 2004; Jaldesa ). On balance, historical experience and the theoretical literature do not predict medical society engagement with or support for abortion law reform.

Materials and methods

This retrospective case study examines the factors behind professional medical society support for abortion law liberalization. It used observational data (in-depth interviews); as well as ESOG and government documents; opinion data; and secondary research. The University of California, Berkeley Committee for Protection of Human Subjects provided ethical approval (Protocol ID: 2011-03010).

The 54 people interviewed are a purposive sampling of individuals familiar with the 2005 reform of the Penal Code in Ethiopia, selected using a combination of reputational and positional criteria (Tansey 2007). Men were 60% of all informants; all but two informants were Ethiopian nationals. Interviewees included 10 Ethiopian ob–gyns, 6 of ESOG’s 10 past and current Presidents and 5 of ESOG’s nine Secretaries. All were men affiliated with maternity or university hospitals, Ethiopian and international NGOs, and/or the Ethiopian federal government. All ob–gyns contacted were interviewed (Table 1).

Table 1.

Interviewee backgrounds (2012, 2007)

Primary affiliation only
Government	6
Women’s rights NGOs	4
Ethiopian (reproductive) health NGOs	9
Reproductive health medical professionals	10
Researchers	6
International NGOs	12
Media	1
Donors	4
Religious leaders (Ethiopian Orthodox Church, Supreme Council of Islamic Affairs)a	2

Fifty-two (52) individuals interviewed in 2012; two (2) people interviewed in 2007 (Ferede Alemu 2010).

All but two interviews were conducted by the author in English in 2012 (7 years post reform), as all interviewees had tertiary-level education for which English is the medium of instruction. Two other interviews of religious leaders were conducted in Amharic and summarized in English (Ferede Alemu 2010). All but 10 interviews were recorded and transcribed verbatim; unrecorded interviews were summarized shortly after completion. Interviews averaged 55 min in length, with a range from 20 min to 4 h. All interview transcripts/write-ups were then coded using HyperRESEARCH 3.5.2 qualitative data analysis software (ResearchWare 2015). They were first descriptively coded for professional affiliation and gender, then to identify actors, roles, interests and timing of key events. Quotations here are all from Ethiopian informants.

Questions in the 2012 semi-structured interviews were on the reform sequence and actors; the framing of the debate; the reform roles and actions of the informant’s organization; and summative questions on perceived causes of reform. Phenomena of interest are ESOG’s position on liberalization of the Penal Code on abortion; its motivations, roles, and actions related to reform; the timing of key events; and the factors governing active medical society engagement in reform. As the argument here is that this Ethiopian case does not match empirical evidence or theoretical predictions related to medical society involvement in contentious policy reform, I first produced a chronology of the reform with special attention to the contributions and motivations of the ob–gyn society (ESOG), and then looked for factors that might mitigate forces discouraging ESOG policy participation predicted by theory and historical experience. Secondary and other data sources helped to validate statements made by interviewees and fill any gaps. If events were described in similar ways by multiple respondents from diverse types of organizations, all involved in the reform, saturation was deemed to have been reached.

As with any retrospective research, this study has limitations, primarily recall and social desirability bias. The reform’s prominence may reduce recall bias. I tried to mitigate social desirability bias by interviewing a wide variety of informants; comparing individuals’ statements with what they might have an institutional or reputational incentive to say, inquiring about the roles of other actors; asking informants to critique their own assessments; and saving questions about impact for a later stage (Peabody ; Berry 2002). Further, bias resulting from respondents’ desire to claim credit for reform is likely counterbalanced by discomfort with the idea of ‘abortion upon request’ and desire to avoid broader public attention around abortion.

Results and discussion

Review of the three sets of factors associated with ESOG policy engagement follows below, covering ESOG organizational and individual member commitments; organizational structure and experience; and the receptive political environment.

ESOG support for reform grounded in commitment to reducing maternal mortality

Ethiopia’s ob–gyns had deep concerns about maternal mortality as well as broad support for liberalization of the law on abortion. Before the reform, ESOG’s 2000 nationally representative survey of medical providers showed that 80% of ob–gyns viewed the existing (1957) law as overly restrictive, and that over three quarters believed that preventing the interruption of schooling was an acceptable rationale for legal abortion (Lakew ). Interviews suggest that individual ob–gyns’ motivations sprang from their direct experience caring for women dying from unsafe abortions and their belief that they and their profession were the ones with the knowledge, skills, and experience to prevent these deaths.

The magnitude and the mortality, the mortality. And that it’s affecting mainly the young people. You see, the eyes of these young people remain in your eyes, when they are almost dying. They see you, they say, “please save me.” This is almost a daily practice in our Gyne ward when we are residents, in the so-called ‘septic room.’ There are three or four people, going into septic shock, very eager to survive. But you can’t save them, because they came so late. And there were no proper antibiotics at that time. For an ob-gyn, it was the kind of thing that you start to hate the … We used to operate at least one patient a day, at least every day, with abortion complications.   Interview 12 (Ob-gyn)

Several ESOG informants described this experience of being unable to save young women with their lives ahead of them as motivating their support for legal reform. Ob–gyns knew of the impacts of unsafe abortion not only due to their work, but also as a result of their research on maternal mortality and its causes in Ethiopia, which was shared at ESOG and other meetings (ESOG 1999; ESOG 2017). Ob–gyns interviewees spoke of their special responsibility, even moral obligation, to speak out for policy change, as those who best knew the scope and impacts of unsafe abortion.

These (obstetrician–gynaecologists) are the learned people […] They know the problem, they know much of the causes of the problem, and they know the solutions. […] So again, crossing their arms and sitting cannot be a way out, they have to do something, right? These people have a conscience. If they don’t do it for women who are suffering or dying …. There is this moral issue, right?             Interview 14 (Ob-gyn)

Ob–gyns also knew of safe alternatives to the unsafe abortions that sent injured women to the overflowing labour and delivery wards of the capital’s hospitals. Through the training of many senior ob–gyns in Soviet bloc countries, as well as contact with providers of safe abortion services in Addis Ababa, ESOG leaders knew this maternal mortality could be reduced through access to safe abortion services (Gaym 2010). Since at least the early 1990s, a significant proportion of the services in the capital, Addis Ababa, had been safe abortion services, as a number of medical professionals, including those at the Dejazmach Balcha Russian Red Cross Hospital and at Marie Stopes International, Ethiopia, provided safe abortion services (Mgbako ). Further, much of ESOG’s leadership, including at least three past ESOG Presidents, had been key movers in the introduction of post-abortion care (PAC), a comprehensive set of clinical and health interventions to preserve the life and health of women presenting with complications of unsafe abortion. The introduction of PAC not only brought clinical skills to save lives (the same used for safe abortion care), but it offered an opportunity for providers to clarify and articulate their values and also their roles and professional responsibilities regarding unsafe abortion and abortion more generally (Turner ).

The introduction of post-abortion care (PAC) services here in Ethiopia […] has given us actually the competence to familiarize everybody with the topic and with the quality of care. So if that was not there, definitely the groundwork would have been missing, so it would have been difficult to jump from nowhere to policy formulation, no way!        Interview 33 (Ob-gyn)

Further, the way that ob–gyns had discussed the rationale for PAC—in terms of saving women’s lives—foreshadowed how ESOG leaders framed discussion later during the 2005 reform process.

ESOG founding rationale and mission

When discussing ESOG’s involvement in reform, all ob–gyns interviewees began by pointing to ESOG’s founding motivation and explicit mission of addressing Ethiopia’s elevated levels of maternal morbidity and mortality (Gaym 2010; ESOG 2012). They viewed this mission, as well as the research and practice of ESOG’s founders, as directly consonant with support for reform of the country’s laws on abortion. They noted that ESOG’s founding was in part sparked by the 1987 inaugural Safe Motherhood conference.

There was consensus among interviewees that ESOG had taken a clear position that reform was needed, and that internal opposition was absent or muted. Several referenced ESOG’s 2000 nationally representative provider survey as revealing the profession’s strong support for liberalization of the law (ESOG 2002). ESOG’s President in 2001 articulated the position to the media that: ‘pregnancies which occur through rape, inefficacy of contraceptives as well as pregnancies that endanger the life of the mother should be given free access to safe abortion’ (German-Ethiopian Association Newsletter 2001). The one dissenting ob–gyns interviewed was more troubled by how ESOG’s focused on advocacy to the exclusion of professional development activities. A former President explained that while there might have been discomfort among some ESOG members about the reform or about ESOG’s public stance, they would not have voiced opposition explicitly or publicly due to loyalty to the profession.

Yes, ESOG has always been positive, I mean even if there are people who are opposed to [abortion reform …] probably they know that they are not many, or probably they don’t want as a person to be included, but as an institution, it is okay. I don’t know how they see it, but there has never been, or I have never witnessed, anybody who has raised his or her hand and said this is immoral, this is negative, this is … you know. I have never heard that in the meetings of ESOG that I have attended.

Interview 14 (Ob-gyn)

At the same time, ESOG’s leadership recognized the organization’s role as a professional association representing membership, rather than an advocacy group with a specific policy platform, and the constraints this placed on advocacy. They also noted that ESOG had neither the clout of a trade union nor the strength or inclination to have an adversarial relationship with government. They emphasized ESOG’s close relationship with government, and saw their role as serving as a resource for government.

This is just a professional society, it’s not like a labour union whereby it imposes a new policy. But we closely cooperate and collaborate with the Ministry of Health in terms of formulating policies or act as resource person for whatever activities the Federal Ministry of Health undertakes.

Interview 33 (Ob-gyn)

Reduced constraints on ESOG’s policy engagement

Interviews as well as review of ESOG documents and secondary data suggest that ESOG’s active role in reform is also explained by the newness of the organization and its streamlined structure, by a favourable political landscape, and by the high value reform allies placed on ESOG’s contributions.

ESOG’s organizational structure and stage of development

Several features of ESOG’s organizational structure gave leadership more latitude to engage in policymaking than might otherwise have been expected. It was at an early stage of organizational development, one where it was highly motivated by mission, but relatively lacking in systems and resources (Whetten 1987). While ESOG had management and governance systems, they remained relatively flexible, allowing volunteer leadership to pursue policy goals with fewer impediments. Further, as ESOG’s leadership worked on a volunteer basis, it tended to attract those motivated to work on topics and activities to which they were personally committed (Andrews and Boyne 2010). ESOG’s leaders during this period of reform included several who led the introduction of PAC to Ethiopia, and thus were more accustomed to dealing with the negative societal views on abortion.

ESOG leadership had further latitude because of how responsibility for work was allocated. At ESOG’s annual meeting, the General Assembly typically delegated broad responsibility for project implementation during the upcoming year to ESOG’s President and Executive Board or, in this case, to the volunteers on the sub-committee working on reform. This allowed those working on a particular project or policy reform to operate with substantial autonomy. The limited formal contact between leaders and membership also reduced potential obstacles for ESOG leaders’ public contributions to reform.

Further, ESOG’s organizational youth meant that it lacked precedents prescribing or proscribing particular types of policy activity. Previous ESOG policy work had not been on socially contentious issues such as abortion, nor had it been on high profile legislative reform. Given the absence of such precedents, ESOG leaders had more room to contribute actively and publicly to policymaking.

Finally, ESOG was also relatively less susceptible to donor pressure, as it had few external sources of financial support prior to 2006 when the final Ministry of Health regulation on abortion was issued (Table 2). Significantly, ESOG had no direct U.S. government funding until after the reform. During Ethiopia’s reform period, U.S. international reproductive health assistance was governed by the Mexico City policy (known also as the ‘Global Gag Rule’) barring non-U.S. organizations receiving U.S. assistance from providing abortion care services or referrals or speaking publicly on abortion legal reform, even if financed by other sources (Cohen 2000; Executive Office of the President 2001).

Table 2.

ESOG’s externally funded projects and consultancies (1992–2012)

Period	Number of projects & consultancies	Number with US government funding
1992–2006	4	1 (indirectly)
2006–08	10	2 (1 direct; 1 indirectly)
2008–10	12	1 (direct)
2010–12	14	1 (direct)

Indirect US government assistance is channelled through a US international NGO (ESOG website, 2012). Periods correspond to ESOG Presidencies. The first Presidency after the legal reform began in 2006

There was the Gag Rule and so on at that time. Even ESOG didn’t have, didn’t get a lot of support in terms of projects at that time, because ESOG was considered as an institution that promoted abortion. At that time, we were having very much difficulty in getting projects and so on, research support. Because USAID and everybody considered it as an abortion organization.                  Interview 12 (Ob-gyn)

By involving the organization in policy on abortion during a U.S. Republican administration (2000–08), ESOG’s leaders knew they were forgoing potential U.S. government funding. Thus ESOG’s policy work occurred counter to its financial interests.

Favourable political environment

Virtually all interviewees described a political environment favourable to reform. In particular, the government’s well-known ideological commitment to improving women’s status and enactment of progressive policies that were foundations for later reform, a moment of relative democratic opening for civil society advocacy, and the active support from other advocates, all reduced the risks of participating for ESOG.

Ethiopia’s reform took place under the aegis of the current political regime, the Ethiopian People’s Revolutionary Democratic Front (EPRDF), a coalition of groups led by the Marxist Tigray People’s Liberation Front (TPLF). The ruling party’s history of policies enacted to improve women’s status, of proud secularism (Haustein and Ostebo 2011; Abbink 2014), and specific senior leadership guidance to advance liberalization of the law (Wada 2008; Holcombe 2014), created a political environment friendly to reform.

The government has a good and strong position on promoting women’s rights, starting right from the bush. There were many women soldiers involved in the fighting.

Interview 44 (Ob-gyn)

The government came to power in 1991 with a record and a clear agenda to advance progressive policies related to women. Their swift adoption of a progressive National Women’s Policy (1993), a National Population Policy (1993), and particularly the new Constitution (1994), made it clear there would be a complete overhaul of the 1957 Penal Code, with an explicit eye for expanding the ‘democratic rights and freedoms’ of women, as well as reducing the ‘grave injuries and sufferings caused to women and children by reason of harmful traditional practices’ (Constitution, Article 35, Rights of Women). ESOG leaders and other reform supporters all alluded to the Constitution and allied policies as explaining why reform of the law on abortion later took place.

The Constitution was very specific on the importance of protecting and promoting women’s rights, and against harmful traditional practices. With a favourable Constitution, it is very likely that there would be reforms aligning policy with the Constitution.            Interview 22 (Ob-gyn)

However, although interviewees saw the TPLF’s progressive ideology and track record as promising for reform of existing laws related to women, including abortion, they noted that high-level political leaders did not want to take action without public demonstration of support. The government saw it as the responsibility of civil society to educate the public and galvanize support, and permitted civil society policy advocacy insofar as it was ‘harmonized’ with the government’s developmental agenda (de Waal 2013). This approach to policy change grew from its experience as an insurgency movement in rural Tigray fostering social change (Hammond and Druce 1990; Barnabas and Zwi 1997; Segers ).

Ok, you bring it forward, he [the government] is not going to cook it and finish everything and give it to you. But you have to bring it up. You need initiators to take it up to the top of the ladder, to make sure it'll be blessed by them. It’s progressive, they have been for it, and they were for it, and they encouraged the movement to take its course. They facilitated everything, took it up to the Parliament and had it passed through. Who else could do it for you, unless you have got a government that is very cooperative and progressive?       Interview 33 (Ob-gyn)

The government formally channelled civil society participation, including that of ESOG, through the Reproductive Health Working Group (RHWG). It convened the RHWG in 2002 to educate the public about maternal mortality and unsafe abortion, to mobilize support for reform, to offer technical assistance to government leaders, and later to develop the law’s regulations (Wada 2008). Members of the RHWG, including ESOG representatives, presented in public forums, on the radio, and at national and regional government hearings on the proposed Penal Code reforms. ESOG, and NGO leadership generally, saw a government expectation of reform, and of civil society contribution to this reform.

Reform supporters also made use of the historic moment of relative democratic openness in Ethiopia’s political system prior to a 2005 crackdown (Hagmann and Abbink 2011). Informants saw the period as one of openness and policy possibility.

Yes, that period was a very good time for civil society, for the blooming of civil society, for creation of new NGOs and everything. I think it was very conducive for bringing about new ideas and for supporting family planning and a lot of different issues. So in terms of the environment, not in terms of the issue, it was more conducive than what we have now. Because it has a lot of space.          Interview 25 (Women’s rights leader)

In addition, the decision to overhaul the entire Penal Code provided the structured opportunity for advocates to push for inclusion of reform of the law on abortion. As noted above, during the Penal Code reform, the national government left open space for NGOs to present their experience, research, and perspectives; and to inform policymakers and the elite public.

Support for ESOG contributions from other advocates

ESOG’s reform collaborators saw ESOG and obstetrician-gynaecologists as essential contributors to policy discussions—‘their absence would have been fatal’ (Interview 6, Ethiopian International NGO leader)—further smoothing the way for ESOG participation. Civil society support for reform sprang from two sources. The first was women’s organizations, most notably the Ethiopian Women Lawyers’ Association (EWLA), championing a broad set of reforms to improve women’s social and economic status (EWLA 2001; Ashenafi 2004). The second was a network of reproductive health NGOs seeking to limit maternal mortality, particularly that due to unsafe abortion, and to expand women’s access to reproductive health services. ESOG’s involvement was with this second group. Reform allies ensured that ESOG leaders were prominent if not lead speakers at public events. While interviewees didn’t necessarily point to ESOG as the most proactive reform supporter, they uniformly attested to the high value of ESOG’s participation.

It’s true. We want them [obstetrician-gynaecologists] on the forefront because our people look up to them on these issues. And also, you know, one voice of an ob-gyn is worth a million of ours. And it was so important to make sure that we are working with them.            Interview 6 (NGO leader)

Other advocates emphasized the special value of ESOG and ob–gyns as messengers for reform. They brought credibility to their policy advocacy, particularly in the eyes of more traditional representatives in Parliament and the regional governments. One women’s rights advocate emphasized not only the complementarity of the strategies of women’s advocates and those of ESOG, but also the special value of testimony from male ob–gyns, particularly with more conservative people.

What really complemented our story is the involvement of the obstetrician-gynaecologists, because they are mostly men. There are women, but the majority of them are men. It is not westernized women shouting and saying ‘you have to do this and this’ on television and camera. It was these doctors, men, saying that on camera and in the meetings. […] So they brought the numbers; we brought the stories. Those complemented each other very well in Parliament, for dealing with religious groups, and for other reluctant people.    Interview 25 (Women’s rights group leader)

Most advocates saw ESOG member involvement as critical to legitimizing and detoxifying public discussion of unsafe abortion as a policy problem.

Conclusions

This study documents what underpinned the active involvement of a medical society in supporting the 2005 reform of Ethiopia’s Penal Code with respect to abortion, counter to theoretical and empirical expectations of self-interested behaviour. Drawing on their own and others’ research to highlight evidence of the problem and the need for change, ESOG communicated with policy makers and the informed public. They opened and legitimized public discussion of maternal mortality and unsafe abortion by framing it as an issue of saving lives. In their own view, as well as that of other reform supporters, the high social status of the well educated and mostly male ob–gyns, as well as how they talked about abortion, made them more palatable messengers on the sensitive issues of maternal mortality, unsafe abortion, and legal reform, particularly for older and more traditional leaders.

Although this research cannot provide a definitive answer to why ESOG was able to engage in contentious policymaking, it offers some promising explanations. It finds that an explicit organizational mission focused on maternal mortality prevention, combined with relaxed organizational and political constraints to participation, left a larger space for ESOG’s policy engagement. In particular, ESOG’s social mission, its young and lean organizational structure, and the lack of organizational precedents or funding constraints to bar action; a government receptive to progressive reform and to civil society contributions; a moment of historical opening; and the high value other advocates placed on ESOG contributions, all facilitated ESOG’s participation in controversial policymaking. While Ethiopia’s government exerts tight control over civil society policy advocacy (Østebø ), this does not negate the lesson that ob–gyn societies elsewhere can capitalize on political openings. Abortion law reform was not a first level priority of the government, and without activism by civil society actors (albeit carefully government sanctioned activism), the reform would likely not have taken place. Activism necessarily here took the form of collaborative policy advocacy, given the requirements of Ethiopia’s political regime. Further, there does appear to be more latitude for more recently formed and mission-driven medical societies such as ESOG to contribute to reform efforts, as they may not confront the internal constraints faced by more established medical societies.

In sum, this case does not lead us to revise or reject existing theories about the conditions deterring medical associations from participating in controversial reforms. However, counter to theoretical predictions from the professions literature about protecting professional prerogatives from other medical cadres and clinical autonomy with respect to clients, ESOG neither opposed expanding the scope of practice of midwives, health officers and clinical nurses to include provision of abortion services, nor insisted that medical necessity be the condition for women’s eligibility for services (i.e. therapeutic abortion), but supported a law that de facto enables women to decide whether they are eligible for services. It may be that ESOG and medical societies in contexts with acute provider shortages are not as preoccupied as those in more affluent contexts with preserving clinical discretion and preserving scope of practice vis-à-vis other medical cadres.

By examining how an ob–gyn society contributed to reform in Ethiopia, this research can help advance understanding of the circumstances in which professional medical societies in the region can advance socially contentious yet life-saving policy reforms. Ethiopia has features common to other countries in SSA: elevated levels of maternal mortality including that due to unsafe abortion, weak health infrastructure, chronic shortages of ob–gyns and other medical professionals, and young ob–gyn societies. ESOG’s successful engagement in reform suggests that even with contentious policies, medical societies can be politically relevant in part due to their scientific credibility and technical expertise, as well as their organizational commitment to maternal mortality reduction. Further, motivated leadership can take advantage of political openings to help advance policy benefiting the nation as a whole, rather than the narrow material interests of the profession. Civil society partners can also amplify medical society impacts by strategically facilitating their participation in targeted activities. Future comparative analysis could help assess whether ob–gyn societies in SSA, with their relatively tighter ties to government and often regular contributions to national administrative policymaking and training in reproductive health, are likely to be more ready to contribute to reform than is the case in other regions.