Gender—the socially constructed roles, behaviours, activities and attributes that a given society considers appropriate for males, females and other genders—affects how people live, work and relate to each other at all levels, including in relation to the health system. Health systems research (HSR) aims to inform more strategic, effective and equitable health systems interventions, programs and policies; and the inclusion of gender analysis into HSR is a core part of that endeavour. We outline what gender analysis is and how gender analysis can be incorporated into HSR content, process and outcomes. Starting with HSR content, i.e. the substantive focus of HSR, we recommend exploring whether and how gender power relations affect females and males in health systems through the use of sex disaggregated data, gender frameworks and questions. Sex disaggregation flags female–male differences or similarities that warrant further analysis; and further analysis is guided by gender frameworks and questions to understand how gender power relations are constituted and negotiated in health systems. Critical aspects of understanding gender power relations include examining who has what (access to resources); who does what (the division of labour and everyday practices); how values are defined (social norms) and who decides (rules and decision-making). Secondly, we examine gender in HSR process by reflecting on how the research process itself is imbued with power relations. We focus on data collection and analysis by reviewing who participates as respondents; when data is collected and where; who is present; who collects data and who analyses data. Thirdly, we consider gender and HSR outcomes by considering who is empowered and disempowered as a result of HSR, including the extent to which HSR outcomes progressively transform gender power relations in health systems, or at least do not further exacerbate them.
Keywords: Framework, gender, gender analysis, health systems, health systems researchHealth systems are not gender neutral; gender is a key social stratifier, which affects health system needs, experiences and outcomes (Standing 1997; Jackson et al. 2006; Nowatzki and Grant 2011). Gender is defined as the ‘socially constructed roles, behaviours, activities and attributes that a given society considers appropriate for men and women’ (WHO 2015). As a social phenomenon, the meaning of gender is negotiated by individuals and societies and therefore varies over time and across contexts; in contrast to sex, which refers to the chromosomal characteristics that distinguish men, women and intersex people (Sen et al. 2007). Gender affects how females, males and people of other genders live, work and relate to each other at all levels, including in relation to the health system. As a power relation, it influences: vulnerability to ill- health; household decision-making and health seeking behaviour; access to and utilization of health services; the design and use of medical products and technology; the nature of the health labour force; the implications of health financing; what data is collected and how it is managed; and how health policies are developed and implemented (Standing 1997; Vlassoff and Moreno 2002; Sen et al. 2007; George 2008; Percival et al. 2014).
The World Health Organization (2007: 2) defines a health system as ‘all organizations, people and actions whose primary intent is to promote, restore or maintain health’. Health systems frameworks are evolving (van Olmen et al. 2012), with concepts related to systems thinking and people-centeredness most recently defining certain health systems characteristics. Systems thinking highlights the complex, dynamic, context-specific, multifaceted and interconnected nature of health systems and their components (Adam and de Savigny 2012). People-centeredness emphasises how health systems are ‘constituted by people and operate in social, political and economic contexts defined by people and groups’ with varying interests, values and power (Sheikh et al. 2014: 2). As gender influences how people interact dynamically in complex, multi-faceted and context-specific ways, reflecting varying interests, values and power, gender is at the core of health systems and hence should be within the heart of health systems research (HSR).
Yet, HSR often fails to sufficiently consider gender as a social relation (Standing 1997). When gender analysis is incorporated into HSR, it is often incomplete either focusing on females only (Percival et al. 2014), or not going beyond sex disaggregation (Johnson et al. 2009; Nowatzki and Grant 2011). Both of these approaches ignore the socially constructed power relations and gender norms that exist between and among males, females and other genders that can lead to vastly different health system needs, experiences and outcomes (Hunt 2004; Sen et al. 2007; Nowatzki and Grant 2011).
Moving forward, gender analysis in HSR entails researchers seeking to understand gender power relations and norms and their implications in health systems, including the nature of female’s and male’s lives, how their needs and experiences differ within the health system, the causes and consequences of these differences, and how ‘programs, services and policies might be better organized to ameliorate, accommodate or redress the differences among them’ (Jackson et al. 2006). As well as analysing differences between females and males, gender analysis, by focussing on the nature of power relations, also considers differences among females and among males. It includes examining gender in relation to other social stratifiers, such as class, race, education, ethnicity, age, geographic location, (dis)ability and sexuality, ideally from an intersectional perspective (Bottorff et al. 2011; Hankivsky 2012). An intersectional perspective examines how these markers dynamically interact, exploring how power plays out at multiple levels and through diverse pathways to frame how vulnerabilities are experienced.
Incorporating gender analysis into HSR should ideally be done at all stages of the research process. It includes considering gender when defining HSR aims, objectives, or questions; within the development of study designs and data collection tools; during the process of data collection; and in the interpretation and dissemination of results (Ravindran and Kelkar-Khambete 2008; Johnson et al. 2009). In this article, we outline how gender analysis can be incorporated into HSR content, process and outcomes. While we review gender within each HSR area separately, we recognize that they interact, overlap and reinforce one another; and an approach that takes forward gender within these areas is mutually reinforcing. We also note that while other genders need to be considered within gender analyses, this article deals primarily with relations between males and females.
HSR content refers to the substantive focus of HSR, whether it be financing, or human resources for health, etc. A starting point for understanding how gender power relations substantively affects males and females differently in health systems, entails applying sex-disaggregation, gender frameworks and gender analysis questions into HSR content.
To incorporate gender analysis into HSR, data and information must first be disaggregated by sex. Collecting sex-disaggregated data means distinguishing between males and females when gathering information, and ensuring that this information is recorded and maintained (Hunt 2004; Nowatzki and Grant 2011). In some contexts, it will also be important to go beyond the male-female binary and include options for other groups, such as transgender and intersex populations. Disaggregating data by sex is critical as aggregated datasets can mask differences between males and females, leading to assumptions that males and females share the same experiences—a bias which negatively affects the validity and reliability of research evidence (Johnson et al. 2009; Nowatzki and Grant 2011).
For example, not reporting sex as a variable in health labour force surveys or human resources for health studies conceals the gendered composition of the health workforce (George 2008). Looking across occupations, those at the lower tiers of the health workforce, which require less education and have less employment security and earning potential, have higher proportions of women. In addition, within the same occupation, and at times even in female dominated health cadres, women are often promoted less frequently and earn less than men (George 2008; Newman 2014). Research that does not disaggregate data by sex could therefore generate evidence that fails to adequately portray the true nature of the health workforce, leading to policy which fails to consider gendered drivers of inequality. For example, during the development of lay health worker policy in South Africa, while poor working conditions were recognised, the gendered basis of that inequality (assumptions that women can volunteer their time, structural and labour market discrimination that left women with few employment options, failure to prioritize women’s career pathways) failed to gain policy attention (Daniels et al. 2012). Similarly, not disaggregating data by sex fails to recognize how men’s higher disease burdens for certain conditions are linked to gender-related norms which promote high risk behaviour such as smoking and alcohol consumption (Sen et al. 2002; Hawkes and Buse 2013).
In addition to sex, data should also be disaggregated by other social stratifiers, such as age, race, ethnicity, disability, socioeconomic status and geographic location, depending on the context and issues under consideration. Gender analysis, however, goes beyond merely disaggregating by sex and other social stratifiers. Researchers must use such disaggregation as a trigger that, with the appropriate frameworks and questions, can spark further investigation of how social inequalities lead to different experiences within health systems, and/or can be further retrenched or reversed by health systems (Gilson et al. 2007). As collecting and analysing additional data within health information systems can meet with resistance due to the extra human and financial resources, as well as time required for the large sample sizes needed to significantly detect social biases across different axes, there is a clear need to demonstrate the importance of collecting this information and using it to promote change.
In further examining gender inequities, gender frameworks can help researchers further organise their thinking, research questions, data collection, and analysis. Within the context of HSR, researchers new to gender may find gender frameworks particularly useful in helping to focus their thinking on key aspects of gender power relations which are most relevant to their study. With the diversity and dynamic nature of substantive issues addressed by HSR, any framework used may need further adaptation to be fully effective in understanding the complex power relations that characterise gender within health systems. When choosing or adapting a framework it is important to understand its underlying principles or theoretical underpinnings, as each framework will differ in its assumption of what needs to be analysed and addressed, some of which are discussed below (Warren 2007).
In developing this article, we found 42 gender frameworks, guidelines and tools developed by a variety of agencies and organisations. Focussing on frameworks only, we found 15 that focussed on health, health systems and development ( Table 1 ). Many of these earlier documents focus on development, reflecting how gender analysis has been debated and reviewed in development studies since the 1970s (Boserup 1970), in contrast to its relatively nascent recognition in health systems (Standing 1997). Earlier frameworks guided analysis that highlighted women’s instrumental role in development, i.e. how projects would be more efficient after considering women’s contributions (Harvard Analytical Framework in March et al. 1999). Others helped distinguish between women’s reproductive, productive and community roles, as well as how women had strategic needs linked to the distribution of power that were different from immediate practical needs (Moser 1993). Some focussed on the social relations and institutions that govern gender relations inhibiting broader wellbeing and empowerment (Kabeer 1994). Overtime frameworks have identified how gender norms, beliefs, roles, time allocation, division of labour, access to resources, and rules and decision making constitute gender power relations.