Towards a framework for assessing empowerment
Sarah Mosedale, EDIAIS
Paper prepared for the international conference,
New Directions in Impact Assessment for Development: Methods and Practice,
Manchester UK, 24 and 25 November 2003
When policymakers and practitioners decide that “empowerment” – usually of women
or the poor – is a development goal what do they mean? And how do they determine the extent to which it has been achieved? Despite empowerment having become a widely used term in this context there is no accepted method for measuring and tracking changes.
Presumably if we want to see people empowered we consider them to be currently dis-empowered i.e. disadvantaged by the way power relations presently shape their choices, opportunities and well-being. If this is what we mean then we would benefit from being better informed about the debates which have shaped and refined the concept of power and its operation. Many sociologists, political scientists and anthropologists have discussed power but little of this debate appears to have percolated into development studies.
Therefore in this paper I briefly review how the concept of power was debated and refined during the second half of the twentieth century and discuss how power relations might be described and evaluated in a particular context. I then consider how the empowerment of women has been discussed within development studies and make some suggestions as to how it might be assessed.
This is very much work in progress and comments are very welcome.
Towards a framework for assessing empowerment
Sarah Mosedale, EDIAIS
1.1 What is empowerment?
„Empowering women‟ has become a frequently cited goal of development
interventions. However, while there is now a significant body of literature discussing how women‟s empowerment has been or might be evaluated, there are still major difficulties in so doing. Furthermore many projects and programmes which espouse the empowerment of women show little if any evidence of attempts even to define what this means in their own context let alone to assess whether and to what extent they have succeeded.
Different people use empowerment to mean different things. However there are four aspects which seem to be generally accepted in the literature on women‟s empowerment.
Firstly to be empowered one must have been disempowered. It is relevant to speak of empowering women, for example, because, as a group, they are disempowered relative to men.
Secondly empowerment cannot be bestowed by a third party. Rather those who would become empowered must claim it. Development agencies cannot therefore empower women – the most they can achieve is to facilitate women empowering themselves. They may be able to create conditions favourable to empowerment but they cannot make it happen.
Thirdly, definitions of empowerment usually include a sense of people making decisions on matters which are important in their lives and being able to carry them out. Reflection, analysis and action are involved in this process which may happen on an individual or a collective level. There is some evidence that while women‟s own struggles for empowerment have tended to be collective efforts, empowerment-orientated development interventions often focus more on the level of the individual.
Finally empowerment is an ongoing process rather than a product. There is no final goal. One does not arrive at a stage of being empowered in some absolute sense. People are empowered, or disempowered, relative to others or, importantly, relative to themselves at a previous time.
1.2 Women and empowerment
While the reasons for any particular woman‟s powerlessness (or power) are many and varied, considering women per se necessarily involves questioning what we/they
have in common in this respect. The common factor is that, as women, they are all constrained by “the norms, beliefs, customs and values through which societies differentiate between women and men” (Kabeer 2000, 22). The specific ways in
which this operates vary culturally and over time. In one situation it might reveal itself in women‟s lower incomes relative to men, in another it might be seen in the relative survival rates of girl and boy children and in a third by severe restrictions on women‟s
mobility. Virtually everywhere it can be seen in domestic violence, male-dominated decision fora and women‟s inferior access to assets of many kinds.
A woman‟s level of empowerment will vary, sometimes enormously, according to
other criteria such as her class or caste, ethnicity, relative wealth, age, family position etc and any analysis of women‟s power or lack of it must appreciate these other contributory dimensions. Nevertheless, focusing on the empowerment of women as a group requires an analysis of gender relations i.e. the ways in which power relations between the sexes are constructed and maintained.
Since gender relations vary both geographically and over time they always have to be investigated in context. It also follows that they are not immutable. At the same time particular manifestations of gender relations are often fiercely defended and regarded as “natural” or God-given. While many development interventions involve
challenges to existing power relations it tends to be those which challenge power relations between men and women which are most strongly contested.
While there has been criticism of attempts to “import” Northern feminisms to the South it is patronising and incorrect to assume that feminism is a Northern concept. Women of the South have their own history of organisation and struggle against gender-based injustices. Also, gender analysis arising from the second wave of feminism in the North has benefited from extensive criticism of its initial lack of attention to class and ethnicity and its Eurocentricity and there has now been some twenty years of dialogue and joint action between Northern and Southern feminists.
1.3 Problems for agencies
There are those who argue that empowerment lies beyond the sphere of what can be measured. Others consider attempts to do so to be dangerous in terms of the centre asserting control of the periphery. Certainly measuring processes is more complicated than measuring products and there are obvious contradictions inherent in any attempt to prescribe empowerment. However I would argue that, though difficult, measurement must be undertaken for there can be little point in funding an activity if it is impossible to tell whether or not it has been successful.
Although I argue that the measurement of empowerment has to be undertaken I do not assume that it will be non-problematic. Indeed it is not difficult to see difficulties for agencies which wish to facilitate empowerment and to measure impact. Some of these are practical – if women themselves are to determine what they wish to change about their situation and how they wish to do it then how are agencies to plan, budget for and monitor activities? This is not a new problem however. It has been extensively explored by those pursuing “participatory” models of development who have tested and refined iterative methods of planning, implementing and evaluating interventions which focus on the knowledge and preferences of the intended “beneficiaries”.
Such participatory methods have been important within development discourse and practice for some thirty years and there is a sizable body of literature and experience detailing and criticising them. A good deal of refinement has been carried out, of methodologies such as participatory rural appraisal (PRA) and participatory learning and action (PLA), in response to criticisms that earlier versions tended to reproduce existing local power relations and to fail to engage the most disadvantaged, particularly women whose access to public space is often the most circumscribed. Participatory methodologies for agreeing locally relevant indicators of poverty, including wealth ranking, have informed participatory methodologies for agreeing indicators of empowerment.
However difficulties for agencies seeking to facilitate empowerment go beyond the “merely” practical. Funding agencies are necessarily in a position of power in relation
to activities which they fund. How does this power relationship affect agencies‟ ability to facilitate the empowerment of women? If participants themselves will largely determine aims, objectives and means the agency has no guarantee that it will like the results.
That women‟s empowerment can threaten state interests is illustrated by an example from India where women, tired of abuse from drunken husbands and the loss of much of their meagre household income organised to close down all the liquor shops in Andhra Pradesh. They raided them and poured away the alcohol, hijacked delivery trucks, burned down shops and humiliated shop owners and drunken men. The movement was catalysed by a story used in a literacy programme that described a young heroine who did this. The government responded by removing the story from the literacy programme (Stein 1997; 36).
When planning projects and programmes with a view to working towards women‟s empowerment, agencies need to consider “the extent to which the agency itself is able to accommodate the empowerment of women and to what extent such empowerment is actually threatening to the state and/or the agency” (Mosedale 1998, 52). Again this problem is not limited to agencies seeking to foster empowerment; much development practice seeks to improve the lot of the poor and has the potential to come into conflict with the interests of the relatively privileged.
2. Models of power
It is often remarked that more effort should be made, particularly by those promoting empowerment, to understand the concept of power. Many sociologists, political scientists and anthropologists have discussed power but little of these debates appears to have percolated into development studies. Jo Rowlands, in Questioning
Empowerment, notes that “it is in its avoidance of discussing power that the fundamental weakness of the literature on women and development lies” (Rowlands
1997; v). Her intention is to “encourage more precise usage and to explore how a more disciplined use of the concept of empowerment might provide a useful tool for activism, gender planning, project planning and evaluation” (Rowlands 1997; vi).
Here I attempt therefore to introduce what seem to me to be some of the most pertinent debates.
2.1 Three faces of power
Within the social sciences power was first typified as power over. As Robert Dahl
defined it “A has power over B to the extent that he can get B to do something that B would not otherwise do” (Dahl, 1957, 202-203). “In this approach, power is
understood as a product of conflicts between actors to determine who wins and who loses on key, clearly recognised issues, in a relatively open system in which there are established decision-making arenas” (Gaventa and Cormwall, 2001).
Subsequently a second dimension or face of power was recognised - the ability to prevent certain people or issues from getting to the decision-making arena in the first place. Bachrach and Baratz argued that political scientists must focus “both on who gets what, when and how and who gets left out and how” (Bachrach and Baratz 1970, 105). This dimension of power is concerned with the rules and methods of legitimising some voices and discrediting others.
Stephen Lukes then suggested that perhaps “the most effective and insidious use of power is to prevent such conflict arising in the first place” (Lukes 1974, 24). From this perspective the powerful may win conflicts not only by doing so in open conflict or by preventing opposing voices from being heard. They may also get their own way by so
manipulating the consciousnesses of the less powerful as to make them incapable of seeing that a conflict exists. As Sen observes “There is much evidence in history that acute inequalities often survive precisely by making allies out of the deprived. The underdog comes to accept the legitimacy of the unequal order and becomes an implicit accomplice” (Sen 1990, 26).
These three dimensions (or faces) of power over therefore consist of one party
getting their own way against the interests of another party either by winning in open conflict, preventing their opponent being heard or preventing their potential opponent from even realising that there is a conflict of interests. These are all examples of a zero sum game i.e. by definition one person‟s gain is another‟s loss (even if, as in the third dimension above, the loser may not even be aware of her loss).
2.2 Non zero-sum models of power
Other forms of power also appear in the literature where one person‟s gain is not necessarily another‟s loss. These tend to be referred to as power within, power to
and power with.
Power within, for example, refers to assets such as self-esteem and self-confidence. In a sense all power starts from here – such assets are necessary before anything
else can be achieved. “[A] woman who is subjected to violent abuse when she expresses her own opinions may start to withhold her opinions and eventually come to believe she has no opinions of her own. When control becomes internalised in this way, overt use of power over is no longer necessary” (Rowlands 1998; 12). The internalisation of such feelings of worthlessness is a well-recognised feature of women‟s oppression and therefore many development interventions seek to bring about changes at this level.
Joke Schrijvers uses the term “autonomy” and defines it to mean, “a fundamental criticism of the existing social, economic and political order…an anti-hierarchical
concept, which stimulates critical and creative thinking and action… transformation which comes from within, which springs from inner resources of one‟s own as an individual or a collectivity” (Scrijvers, 1991, 5-6 quoted in Stromquist, 1995, 15-16)
Power to is defined as “generative or productive power (sometimes incorporating or manifesting as forms of resistance or manipulation) which creates new possibilities and actions without domination” (Rowlands 1997; 13). In other words this is power
which increases the boundaries of what is achievable for one person without necessarily tightening the boundaries of what is achievable for another party. For example if you learn to read it makes many more things possible for you. It does not restrict me (except, I suppose, from using your illiteracy to benefit myself).
Power with refers to collective action, recognising that more can be achieved by a group acting together than by individuals alone. Many interventions aiming to empower women note the importance of creating opportunities for women to spend time with other women reflecting on their situation, recognising the strengths they do posses and devising strategies to achieve positive change.
To develop critical minds women need a place where new ideas can be discussed and new demands arise. For Sara Evans, the prerequisites for developing an “insurgent collective identity” are:
; Social spaces where people can develop an independent sense of worth as
opposed to their usual status as second-class or inferior citizens
; Role models – seeing people breaking out of patterns of passivity
; An ideology that explains the sources of oppression, justifies revolt, and
imagines a qualitatively different future
; A threat to the newfound sense of self which forces the individual to confront
inherited cultural definitions
; A network through which a new interpretation can spread, activating a social
(Evans 1979, 219-220).
2.3 De-facing power
Clarissa Rile Hayward, writing in 1998, criticises the fundamental choice of question which theorists on the nature of power have sought to answer i.e. the question “What does it mean to say that A has power over B?” She points out some problematic assumptions that this approach involves – not least the distinction between free
action and action shaped by the action of others. In developing their description of how A prevails over B through the mechanisms of first, second and third-dimensional power, theorists have maintained this distinction as central to their understanding of how power shapes human freedom.
Hayward argues that “any definition of the line dividing free action form action that is in part the product of power‟s exercise itself serves the political function of privileging as natural, chosen or true some realm of social action” (Hayward 1998, 26). She
illustrates this with reference to a scenario outlined by Jeffrey Isaac (Isaac, 1987). He argues that a teacher going to school to give a lecture exercises the power of a teacher. Yet tomorrow the students might boycott class and conduct their own teach-in. Isaac offers this as an example of the fact that, although the structure of education requires teachers to be dominant and students subordinate, the way this relationship is worked out in concrete practice is contingent and dependent on the way particular groups and individuals choose to deal with their circumstances. This view, in considering the boycott as “free choice” removes it from the category of what may be considered interesting and worth analysing in terms of power relations.
Hayward however contests this view. She argues that whether or not a given student takes part in the boycott probably “depends upon social influences that are themselves unchosen: her parents‟ childrearing practices for example, or norms
prevailing among her peers at school” (Hayward, 1998, 26). “Once one
acknowledges that identity itself is in part a product of power relations, that fields of action are necessarily bound, for example, through processes of acculturation and identity formation, it becomes necessary to reject a view of power that presupposes the possibility of distinguishing free action from action shaped by the action of others” (Hayward, 1998, 26).
Hayward suggests that power might be more usefully thought of “not as instruments
powerful agents use to prevent the powerless from acting freely, but rather as social boundaries that, together, define fields of action for all actors” Hayward, 1998, 27). Mechanisms include laws, rules, norms, customs, social identities and standards that both constrain and enable action. Whereas those using the three faces of power model think of freedom as independent action, using Hayward‟s model, freedom is defined as “the capacity to act upon the boundaries that constrain and enable social
action by, for example, changing their shape or direction” (Hayward, 1998, 27).
Under this view the appropriate focus of the study of power remains the same – i.e.
“patterned asymmetries in the ways in which power – that is, the network of social
limits which defines fields of action – shapes freedom” (Hayward 1998, 28). But
instead of asking how power is distributed and whether A has power over B the
question is “How do power‟s mechanisms define the (im)possible, the (im)probable, the natural, the normal, what counts as a problem?… Do fields of social possibility vary systematically, for example, among groups or across social settings?” (Hayward, 1998, 28).
Any limit to action that is at least partly the product of human action then becomes a valid subject for critical analysis. This obviously includes limits to action that no particular agent A might be plausibly held to “have” or to “use”. In her example of American public education Hayward would include such limits to action as “the local
school district, municipal boundaries, zooming regulations, tax and housing policies, a firmly entrenched tradition of local control over public schooling and a decidedly narrow constitutional interpretation of the role of states in providing education”
(Hayward, 1998, 30).
By this view power relations are not only inescapable but necessary for promoting a range of social goods. Rather than asking whether the actions of some are constrained by the action of others we should look for significant differences in social entitlement and constraint and consider how entrenched or mutable such differences might be. The greater and more asymmetrical are the social limits that define what is possible within a given power relation then the closer that relation approximates a state of domination. “Critical questions about how power shapes freedom are not, then, reduced to questions about distribution and individual choice. Rather, they are questions about the differential impact of social limits to human action on people‟s
capacities to participate in shaping their lives and in shaping the conditions of their collective existence” (Hayward 1998, 32).
Hayward takes pains to explain that she does not reject the critical project that has driven the power debate. She sees her contribution as mainly methodological and offers defacing power as a contribution to the project of identifying and criticising differential forms of social constraint on freedom in order to inform strategies for changing them.
2.4 Foucault and feminism
“Few thinkers have influenced contemporary feminist scholarship on the themes of power, sexuality and the subject to the extent that Michel Foucault has” (Deveaux 1996, 211). Foucault‟s development of his model of power involves recognising the existence of multiple power relations. Power is considered to circulate and to be
exercised rather than possessed. Resistance, where individuals contest fixed
identities and relations in ways which may be subtle, is seen as an inevitable companion of power.
Feminists have both critically analysed and built on this model. The exploration of women‟s day-to-day experience of and resistance to power relations has been productive both in demonstrating the diverse sources of women‟s subjugation and in celebrating and spreading resistance. Indeed consciousness-raising groups in the UK, from the 1970s on, used the sharing of exactly such lived experiences to great effect in changing awareness and motivating collective action for change. The feminist assertion that “the personal is political” was part of the process of
recognising that power was exercised in personal relationships (and not just between men and women) as well as in more public arenas.
Indeed the recognition that women‟s day-to-day struggles and cooperation with the
men in their lives (not just husbands and lovers but fathers, brothers, sons etc) involved power relations challenged an important aspect of the agenda-setting aspect of male power i.e. that which defined what went on in the family as “private”.
This breaking down of women‟s isolation within the family and the concomitant taboos against breaking this imposed silence were an important constituent of “second wave Western feminism” and can be observed in many different cultural settings today.
Foucault described the modern paradigm of power as having two axes. One, which emphasises a disciplined, useful body, is the “anatamo-politics of the human body”.
In the other, the “bio-politics of the population”, the state‟s attention turns to reproduction i.e. health, birth and mortality. Foucault‟s “docile bodies” thesis is taken up by some feminists who find his account of self-surveillance useful in describing women‟s internalisation of patriarchal standards of femininity. As Foucault describes the process “There is no need for arms, physical violence, material constraints. Just a gaze. An inspecting gaze, a gaze which each individual under its weight will end by interiorising to the point that he is his own overseer, each individual thus exercising this surveillance over, and against, himself. A superb formula: power exercised continuously and for what turns out to be at minimal cost” (Foucault 1980, 96).
Foucault‟s own analysis did not extend to recognising that bodily experiences are different for men and women or to considering that the (usually female) problematic conditions of anorexia and bulimia might be “located on a continuum with feminine normalising phenomena such as the use of makeup, fashion and dieting, all of which contribute to the construction of docile, feminine bodies” (Bordo 1989, 23
paraphrased in Deveaux, 216). And, since he did not focus on women, Foucault failed to recognise “the subjective and deeply interiorised effects upon women ourselves both of the emotional care we give and of the care we fail to get in return”
(Bartky 1991, 111).
Foucault‟s model has been criticised for overlooking the broader, structural aspects of power and resistance. By refusing to engage in normative discussions Foucault‟s theory “undermines attempts at social change by obscuring the systematic nature of gender oppression”…“power, being everywhere, is ultimately nowhere” (Hartsock, N. 1990, 170).
Collins argues that these two conceptions of power – the dialectical relationship
linking oppression and activism where groups with more power oppress those with less and the concept of power as an intangible entity which circulates and to which individuals stand in varying relationships - are more usefully regarded as complementary than competing (Collins 2000, 275). Dialectical approaches indicate the need to develop group-based identity and strategies whereas the circulation model directs attention to “how domination and resistance shape and are shaped by individual agency” (Collins 2000, 275). Collins proposes a model of power (and
resistance) in four domains – the structural, disciplinary, hegemonic and
The structural domain of power is where social organisations are organised to reproduce women‟s subordination over time. This includes the police, the legal
system, schools, labour markets, banking, insurance and the media. Within this domain power cannot accrue to individuals without transforming the social institutions that foster their exclusion. In the disciplinary domain power relations are managed through bureaucratic hierarchies and techniques of surveillance. “Bureaucracies, regardless of the policies they promote, remain dedicated to disciplining and controlling their workforce and clientele” (Collins 2000, 281). The hegemonic domain
of power deals with ideology, culture and consciousness and is important in involving women in supporting the own subordination and that of other groups. The interpersonal domain of power “functions through routinized day-to-day practices of
how people treat one another… such practices are systematic, recurrent and so familiar that they often go unnoticed” (Collins 2000, 287).
2.5 Understanding power relations – an example
The model of power towards which I am moving is complex and fluid. It includes structural faultlines based on, for example, sex and class where membership of a particular group (women, peasants) has significant implications for the shape of the power structure within which an individual operates. This is not to say that membership of such a group is either a necessary or a sufficient condition for the particular geometry of any individual‟s position in this power structure. This would be to deny people any individual or collective agency which is obviously absurd. However it is to accept that group membership constrains a person‟s possibilities and defines some boundaries, which being socially constructed can therefore be changed. The extent to which an individual presses against, or accepts, these boundaries and the extent to which change is opposed (and the power of those opposing it) all contribute to the shape and durability of these boundaries.
I also recognise that people have more or less power depending on their specific situation and that they can be relatively powerless in one situation and relatively powerful in another. On a micro level I see each person at the centre of their own space of freedom, a space defined by, and defining, the shifting contours of the multiple containers which circumscribe their lives.
The part of the empowerment debate which I intend to focus on, as both the most interesting and the most under-researched, is that model of women‟s empowerment
which asserts that its function is to radically change oppressive gender relations. In other words the model of power which I wish to use is strongly influenced by Hayward in that what is of interest is how women can build “the capacity to participate effectively in shaping the social limits that define what is possible” (Hayward 1998, 32). The question which I wish to consider particularly closely is how we may know whether or not such changes are being achieved in the field by development interventions which seek to empower women in this way.
Assessing empowerment then requires identifying and mapping power relations i.e. the social constraints to action, including the amount of freedom of action or room to manoeuvre contained within each boundary and the strength of resistance to change of the boundaries. Changes over time could then be identified. But how is this to be done?
Attempting to map the entire network of constraints to action in any situation would be a horribly complex task and one probably best not attempted. Instead assessment would focus on an action or group of actions identified as most significant by those constrained.
Let us consider education as an example. Firstly we could discuss whether this is an area in which the constraints to action are significantly different for women and men. For any particular situation (country, region, village, family) we could compare primary school enrolment or completion rates for boys and girls, look at literacy levels for each group, consider relative rates of entry to secondary or tertiary education and so on. We would of course find that in many contexts girls were significantly disadvantaged relative to boys.
I now consider the situation where a girl wishes to expand her own freedom of action in order to start going to school. Starting from a “three faces of power” perspective questions relating to open, suppressed or avoided conflict can be asked. Informed by
the “de-faced” model of course we can also consider constraints which are not consciously imposed by any identifiable agent. We can also consider non-zero sum power relations.
Is there an open conflict? In this case a girl wants to go to school but other, more powerful, people or social mores act so as to prevent her. Who? Why? For example – her parents (mother or father or both?)
; because her labour is needed at home
; because they cannot afford books/uniform
; because they can see no benefit or consider that costs outweigh benefits
; because any economic benefits from her education will benefit her future
husband‟s family not her birth family
; because it is not socially acceptable for girls to be educated
; because they fear losing her or her rejecting them if she becomes educated
For each of these (or other) reasons, questions could be asked as to how immutable such opposition is. For example: How much is her labour needed at home? What extra input would be needed from elsewhere to allow her to be freed from this requirement?
Is there a suppressed conflict? Is it impossible for the girl to say what she wants? Why?
; because she is afraid of being punished
; because she is afraid of being mocked
; because, believing they are too poor, she does not wish to embarrass her
; because she is afraid of causing conflict between her parents
; because she knows it is not considered appropriate for girls to express their
; because she wants to be good
Is it impossible for the girl to even develop the desire to go to school? Why?
; because she has never heard of a girl going to school
; because she cannot conceive of herself as being someone who could learn to
read and write
; because she cannot imagine any benefits of being educated
; because she has been socialised so as to have very little perception of
herself as an individual
; because she entirely models herself on her (uneducated) mother
; because she has internalised her community‟s beliefs that this is
inappropriate for girls
; because she believes it will damage her marriage prospects
What does a girl need in order to pursue education? For example:
; ability to analyse her situation and think of improvements
; belief that her actions can have effects
; curiosity about the wider world
; confidence that she could learn
; some consciousness of the benefits of education