Gender Analysis

Whether you are an evaluator or someone commissioning evaluation, any intervention to be evaluated that takes place within human society and involves human interactions will have gendered dimensions. 

And that means that you as an evaluator should be able to identify and analyse those gendered dimensions.

But the way in which this analysis is done will depend on how the evaluator (and the intervention being evaluated) thinks about gender in the first place.

Note: Gender is one of many ‘markers of difference’ along which humans tend to judge each other. Other such markers include disability, ethnicity, gender identity, race, sexual orientation, sexuality, or socio-economic status. The ways in which these markers of difference intersect to increase marginalisation, exclusion and inequity is also vitally important for evaluators to keep in mind.

What do we mean by gender?

‘Gender affects everyone, all of the time. Gender affects the way we see each other, the way we interact, the institutions we create, the ways in which those institutions operate, and who benefits or suffers as a result of this.’ (Fletcher, 2015: Addressing Gender In Impact Evaluation: What Should Be Considered?)

The importance of considering gender is widely acknowledged in evaluation Terms of Reference, training curricula, evaluation-related publications and evaluation reports. But these documents often fail to clearly define what they mean by the term. There can be an assumption that the word is clearly understood, but in reality there is no one accepted way to understand what gender is. And that means there is no one accepted way of doing ‘gender analysis’. Similarly, there is no one way of doing feminist evaluation (as noted on the BetterEvaluation feminist evaluation theme page). This is about your way of thinking.

With that in mind, this theme page will begin by exploring the meaning of gender before moving on to issues of gender analysis.

There are many different definitions of gender, but the majority focus on unfair differences in the ways that women and men (categories of people) are treated in our societies. For example:

‘Gender refers to the roles, behaviours, activities, and attributes that a given society at a given time considers appropriate for women and men … In most societies there are differences and inequalities between women and men in responsibilities assigned, activities undertaken, access to and control over resources, as well as decision-making opportunities.’ 

(UN Women Gender Equality Glossary)

Alternatively gender can be defined as:

‘a process of judgement and value … related to stereotypes and norms of what it is to be masculine or feminine, regardless of your born sex category … certain forms of femininity and masculinity are given greater value than others (with particular forms of dominant masculinity usually having the greatest access to power and resources).’ 

(Fletcher, 2015: Addressing Gender In Impact Evaluation: What Should Be Considered?)

What do we mean by gender analysis?

If one takes the first definition provided above as a starting point, then analysing the gendered aspect of an intervention would involve focusing on the differences between men and women within that intervention. For example, are there equal numbers of men and women involved in the intervention and, if not, why not.

If one takes the second definition as a starting point, then analysing the gendered aspect of an intervention would involve examining the judgements, stereotypes and norms related to masculinity and femininity that occur in the intervention’s context and, from there, exploring the effect these stereotypes and norms had on the intended intervention outcomes.

The difference between these two definitions is a crucial one when it comes to evaluating change. The first definition is looking for change in ‘differences and inequalities between women and men’ while the second one is looking for change in the ‘process of judgement and value’ that rewards certain forms of masculinity and femininity while punishing others.

This is an important difference. Gender analysis that is based on the seemingly binary categories of women/men looks for changes in the numbers, and ways, that men and women are engaged in an intervention.

Gender analysis that is based on processes of judgment, norms and stereotypes looks for whether, and how, changes have occurred during the intervention in judgements, stereotypes and norms related to masculinity and femininity, and the effects of those changes. 

Example

Taken from a Guidance Note on impact evaluation (Fletcher, 2015: Addressing Gender In Impact Evaluation: What Should Be Considered?), the following table demonstrates the gendered dimensions of an intervention, when analysed for its effect on (a)gender as process of judgement, and (b) gender as category (men/women). This work draws on Patricia Rogers’ (2012) key evaluation questions for impact evaluation:

Table 1: Gender as a process vs. gender as a category in a road-building intervention

Important issues and considerations

It is worth repeating that every evaluation of an intervention that occurs within human society and involves human interactions will have gendered dimensions, whether or not these gendered dimensions are acknowledged within the intervention design, explicit or implicit theory of change (an implicit theory of change can also be understood as a ‘mental model’) and/or intervention implementation.

If, having examined an intervention’s documentation and having spoken to key stakeholders you can find no reference at all to gender in any form (gender as man/woman or gender as process of judgements, stereotypes and norms on masculinity and femininity), it is fair to identify the intervention as ‘gender blind’. In other words, it ‘ignores gender considerations all together’ (Interagency Gender Working Group, undated).

The implications of this would then need to be explored; and the depth of the analysis undertaken in exploring this will depend in large part on the resources available for the evaluation. It may well be that your gender analysis of such an intervention cannot go much beyond saying ‘this intervention is gender blind, and as such falls below the minimum standards expected’.

An intervention that ‘examines and addresses … gender considerations’ can be defined as ‘gender aware’ (Interagency Gender Working Group, undated) but then needs to be further analysed to see where it lies on what has been called ‘the gender continuum’:

An ‘exploitative’ intervention with regards to HIV prevention would be one that takes cultural judgements around gender and sexuality then uses them to promote ‘good’ behaviour: telling young women they will be ‘spoilt’ if they have sex before marriage, or saying that men who have sex with men (and who are often judged within society as ‘not real men’) are vectors of disease, when the risk lies in unprotected anal sex regardless of the sex of the bodies involved.

An ‘accommodating’ intervention would be one that ‘does not rock the boat’, for instance by claiming it would be ‘culturally inappropriate’ to involve sex workers or trans women in ‘women’s projects’ or accepting that, in an intervention working with civil society organisations, all those organisations are run by men (who will no doubt be from the majority religious and ethnic group in the intervention site).

A ‘transformative’ intervention would be one that works on gender processes; in other words, one that goes beyond the men/women categorical approach and that looks at judgements, stereotypes and norms of masculinity and femininity, and how they are applied regardless of the sex assigned to a body at birth. Referring back to Addressing Gender in Impact Evaluation:

Focusing on the stereotypes, norms and judgements related to masculinity and femininity (rather than on male/female) frees us up to think about the processes through which certain forms of femininity and masculinity are given greater value than others (with particular forms of dominant masculinity usually having the greatest access to power and resources).

Examples would include an intervention that shifted people’s ideas of what it takes to be ‘a leader’ (often considered to be something that requires certain traits seen as being dominantly masculine, such as lack of compassion and willingness to engage in conflict over conciliation; take for example the former UK Prime Minister Margaret Thatcher). An intervention that supported women’s right to contraception could also be gender transformative, if it did so by challenging the view (often held by women, as well as by men) that women who do not have children are somehow ‘lesser’ or that sex for pleasure is incompatible with ‘femininity’.

It is important to know which category the intervention to be evaluated falls into because this will determine, in part, what you can expect to achieve with using a gender analysis as part of your evaluation. In all instances, however, the role of the evaluator is to bring to light the positive and negative effects of the intervention in relation to gender. Sometimes these effects will be intended; often they will be unintended.

How to do a gender analysis

There are many different tools promoted for undertaking gender analysis. However many of these are based on a purely categorical understanding of gender, where ‘men’ and ‘women’ are seen as homogenous groups that are somehow in opposition to each other. The dynamics of gender—the ways in which we all participate in (re)creating norms and stereotypes against which we judge ourselves and others—are usually lost.

The reality is that there is no single, sure fire, A-Z way of doing gender analysis. But there are key steps to follow:

1. Question yourself: How do you think about gender? (How aware are you of the intersections between gender and other markers of difference that feed inequity?

2. Identify whether the intervention to be evaluated is gender blind or gender aware: Start with a desk review of key documents, such as funding applications/design documents; monitoring reports; theory of change or other logic model if available; but supplement this using different methods: see following section. N.B.: doing a word search on ‘gender’ in key documents is not the same as analysing the documents for gendered content. It is entirely possible for an intervention to not actually use the word ‘gender’ but to still challenge stereotypes and norms. Equally, there are many interventions where the documents use the word ‘gender’ widely but the intervention itself is gender blind.

3. Gender blind interventions: If, based on this first stage, the intervention is gender blind (in that it pays no attention to gender, either in categorical terms or in terms of processes of judgement), then say so! If resources allow, it would also be appropriate to draw out some of the consequences of this gender blindness. There are many resources to draw on that clearly define gender as a required component of evaluation of interventions that take place within human society and involve human interactions. (This page is one such resource; others include the Australian Department of Foreign Affairs and Trade Monitoring and Evaluation Guidelines, the Australasian Evaluation Society Guidelines for the Ethical Conduct of Evaluations, which deserves praise for also requiring account to be taken of other markers of difference, not just gender; and the United Nations Evaluation Group, whose latest resource is listed below.)

4. Gender aware interventions: If, based on stage two above, you determine that the intervention is gender aware, how would you describe its understanding of gender? There may not be an explicit explanation—this is often missing—but you will be able to tell by what is done, with whom, and how. Refer back to Tables 1 and 2 for examples of how the different types of gender might play out in the same intervention.  Easy questions to keep in mind are: ‘does this intervention engage differently with men and women?’ or ‘does this intervention engage with judgements, stereotypes and norms about masculinity and femininity?’. As we are talking about evaluation, your over-arching questions for either understanding of gender should be: What gendered changes have occurred as the intended, or unintended, results of the intervention? (descriptive question) What explains those changes? (causal question) and, if you are undertaking an impact evaluation, ‘What does this say about the value of the intervention?’ (evaluative question) . If the evaluation is a process evaluation, then you would need to think about ‘How can the intervention be adapted to reduce any negative changes, increase positive changes, and ensure that the intervention’s theory of change or mental model is appropriate?

Methods

There is a large range of methods available to the evaluator as they move through the steps above. As with all forms of social enquiry, the method adopted should be coherent with your evaluation logic. What do you want to know, who can help you find that out, and finally which method(s) are most appropriate? Selection of method cannot be your starting point in gender analysis; thinking your logic through has to come first.

And remember; as with feminist evaluation, gender analysis is, after all, an exercise in exploring inequities and promoting social justice. As such it sits most comfortably with ‘participatory, empowering’ methods (Podems, 2016).

The Institute of Development Studies manages a Participatory Methods website that has a specific section on participatory monitoring and evaluation, as well as providing details on a wide range of methods that can be used to support gender analysis.

The World Bank Group also has a list of methods (archived link) ranging from desk review to stakeholder workshops and social mapping.