Future Agricultures blog

Opinion and comment from Future Agricultures researchers on agricultural politics, science and society in Africa.

Measuring women’s empowerment: A retrograde step?

Posted by on in Hot topic
  • Font size: Larger Smaller
  • Hits: 6033
  • Subscribe to this entry
  • Print

weaiThe Women’s Empowerment in Agriculture Index has just been launched with great fanfare. It intends to serve as a tool for measuring and monitoring women’s roles and their engagement in agriculture with the aim of closing identified “gender empowerment gaps”.  But does the WEA Index fall into the same trap of previous attempts, essentializing women’s roles and failing to get to grips with the social relations at the heart of gender dynamics in agriculture?

The FAC gender approach (see the discussion paper Gender and Other Social Differences: Implications for FAC?) argues that gender relations are dynamic. Women and men, as spouses, parents, siblings etc, seek both to maintain and change these roles to meet both individual and shared interests. We also acknowledge that changes in gender relations are intrinsically ambiguous and cannot simply be read off from sex-differentiated data. So is the WEA Index a step forward or actually a step backwards: does it learn the lessons from past attempts to highlight ‘women in development’?

As is noted in the material introducing the index, the WEA Index has been produced ‘because women play a prominent role in agriculture yet face economic constraints’ that reduce both their agricultural production and productivity, compared with that of their male counterparts.  In the realm of agriculture, the use of this kind of efficiency argument to justify policy support for women’s productive roles has been central to much policy and practice on gender, or more commonly women, over the last 40 years.  Today, this would be called ‘the business case for investing in women’.

Supporters of women in development (WID) approaches have provided considerable data over the years highlighting women’s labour contributions (‘time’ in the WEA Index) in various parts of the agricultural sector – perhaps suggesting implicitly that these role data reflect their interest in agricultural production – along with data on their (lack of) control over the use of assets such as land, capital, technical knowledge and income from production.  WID-influenced agricultural policy options have always been about achieving positive economic outcomes, with the assumption (too often unexamined) that positive social outcomes (such as empowerment or equity) would follow.  Associated programmes and projects have concentrated on filling the identified ‘gender asset gaps’, by targeting women with credit (usually microcredit) or seeking changes in women’s land or other natural resource rights.

As might be expected, criticisms of this policy approach abound, and success stories based on reliable evidence remain few and far between.  The approach to gender analysis based on sex-differentiated data (as in the WEA Index) is not new (remember the Harvard Gender Roles Framework introduced in the 1980s?) Moreover, its use has not improved our knowledge of social change processes, or of the way in which individuals, either separately or with others, use agriculture to get where they want to go. It also hasn’t resulted in significant changes in women’s position in society.  If this index helps provide greater insight into social change and decision-making processes within small-scale agricultural production units in sub-Saharan Africa, it must surely be welcomed.

Will the results of the WEAI initiative answer the question that is foremost in my mind: will our empowered women ‘go and dig’? What is their interest in agriculture? Are they simply using it as a means to an end – to get where they want to go?  Does this Index simply lead us down the same well-worn path of trying to fill a ‘gender empowerment gap’?

The claim for WEAI is that it is ‘grounded in evidence’ – supporting a view that equalising individual access to similar assets for women as for men results in better development outcomes.  Empowerment is presented as individual self-fulfilment, with individualised assets and clear ownership in the sense of freehold rights providing the necessary incentives to produce the economic outcomes that will justify a focus on (and investment in) women in agriculture.  Even if we choose to measure income increases, for instance, over which individual women and men have some decision-making power, can we really accept this as an indicator of empowerment?   Decision-making is not a process undertaken by men and women in isolation, with no reference to the interests and projects of others. How do we interpret responses like “it was his decision” or “I had no input into decisions about income”, knowing what we know about the complexity of ‘voice’ and ‘choice’, and bearing in mind that empowerment is an essentially relational concept? Clearly we should be very wary of accepting these responses at face value.

Is it not possible to view the income-earning decisions of individuals, households and families in the context of their economic and social lives as a whole? This would then shift the analysis towards the inter-dependency and linked lives of men and women, and towards a life course analysis that highlights shifts that have implications for the changing ways men and women engage in agriculture, rather than assuming that ‘one size fits all at all times’. It would also force us to agree that empowerment does not look the same for everyone, nor is it a fixed state. Surely a better understanding of the dynamics of decision-making, and therefore social change, would be a more satisfactory product for 2015 than an index that will only lead us back into an analytical, policy and programmatic cul-de-sac.

Tagged in: gender


  • Guest
    Ruth Meinzen-Dick Monday, 29 April 2013

    From Ruth Meinzen-Dick, IFPRI

    A recent reposting drew my attention to this blog, which I had missed last year when it came out. Because I was involved in the development of the index, I thought I would reply to provide more information.
    For those who would like to assess the methodology themselves, there are now a lot more resources available in a WEAI Resource Center: http://www.ifpri.org/ourwork/program/weai-resource-center. This includes a Discussion paper that provides more on the methods (including the qualitative work that went in to validating it and the survey itself), and an Instructional Guide that provides more detail on exactly what goes into each indicator. Those who are concerned that this is cutting off research may be interested in the PhD dissertations that we are supporting to explore and validate particular dimensions of the index by comparing results using the index to in-depth analysis using qualitative and other techniques. The availability of do files and instructional materials on the website also makes it possible for program managers to incorporate indicators measuring women’s empowerment directly into their M&E efforts.
    While I don’t think anyone would argue that this is a perfect way to measure women’s empowerment (it focuses on agriculture, which excludes many other aspects of empowerment, and even within agriculture, cannot be comprehensive), the question is whether it is useful to have any measure by which to track changes over time, and especially to think through how agricultural development interventions affect women.

  • Guest
    Prof.(Mrs)Vijayakhader Friday, 26 April 2013

    Socioeconomic Empowerment of women is the need of the day.To see the development/ improvement at community level , village level,state level ,National level and country level the empowerment of women is very very essential.


  • Guest
    Cecile Jackson Wednesday, 23 May 2012

    I have no objection to bringing sex disaggregation into the survey instruments used by economists at IFPRI - this seems to me a worthwhile exercise, and there is a useful role played by these analyses. Although I rather prefer the simplicity and modest intentions of the Harvard Framework which only ever claimed to be a tool for busy jobbing gender analyst, and not a definitive method for measuring empowerment. My concern, like many of the other commentators here, is that this is certainly not what it claims to be (ie a breakthrough in measuring empowerment), and it would be most unfortunate if it comes to define research and analysis on gender and empowerment. The slippery term empowerment has been hijacked and turned into a narrow and degraded weasel word, which serves many purposes and interests (Andrea Cornwall and others have written on this) - and we can object to reductionist uses on analytical grounds, as some of the commentators here do. I agree with the objections, and will not repeat them. But I am also reminded of an earlier debate in the 1990s of how gender terms are simplified and bandied around in development agencies for all manner of purposes, and Shahra Razavi's point that we need to judge these on the discursive good or ill that they do, rather than simply saying that they are misunderstood or wrong. The ill that I see from this kind of exercise is that generations of students will continue to think that they can research 'empowerment' with standardised questionnaire surveys, with questions about 'decision making', and discover the hard way the limitations of such data. By all means disaggregate all the data collected in surveys by sex - excellent! - but let us not mislabel it gender analysis, or claim to to be a metric of empowerment.

  • Guest

    Whilst there is of course a valid argument for trying to understand and address the barriers to women's greater participation in agriculture, I agree with Christine Okali's concerns about the extent to which this index will address women's empowerment. Andrea Cornwall makes the helpful point that definitions of empowerment vary and that many see it going far beyond individual productive ability.

    Given the dynamism of gender relationships, as Christine notes, I would also question the extent to which social relationships and in particular social processes can be quantified through this type of data collection. Whilst the index does attempt to compare women's 'empowerment' with that of the main adult male in the household, I have doubts about the validity of rating these relationships with numbers, particularly within a relatively narrow sphere (agricultural production).

    Another aspect of concern relates to the control and 'ownership' of certain agricultural assets - in particular livestock. In most pastoral societies livestock are rarely 'owned' by a single person, but rather each animal often represents a complex set of rights and responsibilities, based on where the animal came from, and from whom, its uses and control over its products and disposal, and pastoral women often play a key (if not always the dominant) role in this process. Each pastoral herd also often includes livestock 'belonging' to outsiders - relatives, friends or through commercial agreements - over which the household members may exercise certain rights and responsibilities. Livestock ownership in such contexts is therefore highly complex, and not a simple absolute. My (albeit initial) reading of the index questionnaire does not lead me to feel confident that these complexities will be encapsulated in the scores. Failure to recognise the role that pastoral women can play in such situations may in fact further disempower them.

  • Guest
    Mohamadou Sall Sunday, 13 May 2012

    First, as the WEAI is calculated, it is assumed that all the five domains reflect the women empowerment. This must be proved. A strong method to get this proof is to calculate a statistical index called Cronbach Alpha. According to his average (fom 0 to 1 or from 0 % to 100 %), the Cronbach Alpha will show the homogeneity of five domains (while calculating, each domain is considered as an item and quoted j. When is average is at least 0.80 or 80 %, we can consider that all the five domains reflect the women empowerment.

    Secondly, the scoring is statically a weak method because it makes the assumption that all the five domains have the same weight in the women empowerment. It means the five domains have the same contribution to this empowerment. Empirically, it is not true because political, economic, social and cultural characteristics in each society and in each country make this the contributions of these five domains different from:
    - one point (space) to another ;
    - one period to another ;
    - one category of women to another.
    Then, it is important to assume the five domains have not the same contribution in the WEAI. To measure the contribution in each domain, we propose to run a factor analysis. This method will show us which domains (factors) contribute to the women empowerment.

  • Guest
    Andrea Cornwall Friday, 11 May 2012

    The vast body of evidence from decades of research on gender in development ought to be enough to caution against such a reductionist approach to conceptualising and measuring empowerment. Christine Okali is entirely right to highlight the limitations of the approach taken by this index, and to foreground the need for gender analysis.

    And yet, it seems, there's dwindling interest amongst those driving the development agenda in dealing with any such complexity. The 'smart economics' mantra of 'investing in girls and women' offers donors a chain of causality that is as wishful a fiction as the idea that women and men can be neatly divided into two categories and the relationship between them recalibrated by donors/philanthrocapitalists making women (and girls) objects of their 'investment'.

    Within this frame of reference, this kind of index makes sense. But it is not a framing that is shared by many who work with empowerment in practice. Srilatha Batliwala offers a very different way of thinking about empowerment:

    "Empowerment is not a goal, but a foundational process that enables marginalized women to construct their own political agendas and form movements and struggles for achieving fundamental and lasting transformation in gender and social power structures."

    It would be a fine thing for an index to be developed that sought to assess the extent to which development intentions to promote women's empowerment are supporting this 'foundational process'. To do this, the measurers would need to shift their focus away from individual women and towards an approach that could better capture the dynamics of power and change involved in that process.

  • Guest
    Mohamadou SALL Tuesday, 08 May 2012

    Christine raises up critical questions about the "Women's Empowerment in Agriculture Index" (WEAI).
    Our first reflexion is related to the methodology of WEAI :
    - From which score, can we assume the adequate achievement in a domain ?
    - In which extent do the 3 case studies (Bengladesh, Guatemala, Uganda) reflect the role of women empowerment in agriculture ?

    This second question is from my point of view, very critical in a sense that the decisionmaking and its use in the household, in the household economy and in the public area depends probably on several factors : historical, economic, social and cultural. It is very important to adress the following questions :
    - Are all women interested in getting empowerment ?
    - In which spaces, women are very interested in getting empowerment (management of their relations within the household, management of household economy, getting empowerment in public space e.g improving their position and their status in local bodies, in local economic bodies and in local NGO,)
    - For the empowered women, is working in agriculture a priority ?
    - How do the historical, social and cultural factors affect these goals and choices ?

  • Nathan Oxley
    Nathan Oxley Thursday, 03 May 2012

    The following comment was submitted by Dr Paula Kantor, ICRW, Washington

    While the index may help in getting more programs to focus on women, it is unlikely to move the 'gender agenda' forward. What do we think about the use of a man in the household as the benchmark? This has implications of ignoring the interlocking inequalities … are these men ‘empowered’? It is also part of the individualization problem – this woman compared to this man. But what about the bigger picture - the importance of those ‘other social relations’, in other institutional contexts, and the wider context overall, that are removed from the analysis? Another concern relates to the way empowerment is presented categorically – empowered/not empowered – that disguises the complicated process that it is. In our analyses we also need to be aware of the other forms of inequality that may operate.

  • Nathan Oxley
    Nathan Oxley Thursday, 03 May 2012

    The following is a comment submitted by Janet Seeley PhD. Professor of International Development, School of International Development, University of East Anglia:

    Christine makes a valuable point about taking a life-course approach to looking at women's empowerment (and gender relations) in agriculture -- time is important, as are changing relationships in what men and women, girls and boys do and can be expected to do. A cross-sectional `one size fits all' approach misses that.

  • Nathan Oxley
    Nathan Oxley Monday, 30 April 2012

    This comment is from Cathy Farnworth, by email:

    This is an interesting contribution which will take a long time to unpack, so lets hope people have a go. One key area to consider is the relationship between asset ownership (whether individualised or clearly shared) and increased decision-making power, as well as empowerment in domains beyond the household. The evidence points all ways. It is patently obvious that that wealthy women in some societies face severe restrictions on their movement, and that they lack voice in many forums. On the other hand, there is also strong evidence that women with no or few productive assets also face serious restrictions on their ability to take decisions, for example on how to best deploy farm implements on the farm, make investments etc. This can be directly related to their lack of control over assets in some cases. What is certain is that there is no linear relationship between asset control and voice/ empowerment, but there is certainly a relationship at times - which requires further investigation into what these enabling conditions are. I would have several theories to put forward here, but for now hope others may contribute.

Leave your comment

Guest Friday, 28 July 2017