Valuing Time: Time Use Survey, the Capability Approach, and Gender Analysis

Abstract Time use survey is one of the fundamental, and most widely employed, research tools used to bring a gender perspective to project planning. However, narrow interpretations of time use data can distort the understanding of how project-induced time use changes affect women and men's well-being. This paper argues that the application of some of the central concepts of the capability approach can strengthen the scope of time use survey as a gendered planning tool, drawing on the example of the “Alliances” rural economic development project Georgia.

institutions, and is a high-profile policy objective of both multi-lateral bodies (such as UNDP, the World Bank or the more recently created UN WOMEN) and bilateral donor organizations, which frequently identify gender equality (or, increasingly, women's empowerment) as a cross-cutting theme that should be pursued through all sectors of intervention supported.
However, while progress on gender equality has been made in a number of areas, including political participation, or access to education, this progress is slow in many regions (UN WOMEN 2011). Other areas of development reflect mixed results on gender equality; for example, employment, where research into the "feminisation of employment" in the context of market globalization reveals increased gender balance in access to employment, but at the same time a deterioration of the terms of on which many women (and men) are employed (Koggel 2005;World Bank 2012). In other areas, such as the crisis in support to care roles in the context of the global recession (Esprey, Harper, and Jones 2010), gender inequalities are worsening.
This problematic progress on gender equality can be attributed a number of factors, including the impact of development processes such as economic globalization, and the global recession, or the lack of political will in the corridors of power. In addition, however, a large literature argues that development planning processes themselves, although intended to promote the well-being of women and men, often approach women and/or the pursuit of gender equality in ways that may fail to confront, or may even reinforce, gender inequalities (see, for example, Cornwall, Harrison, and Whitehead 2007). This literature also suggests that the institutional logic of development organizations, and their underpinning ideologies, often inhibit progress on gender equality (Levy 1998). Central to these critiques are the interrogation of the methodologies used for development planning, including research strategies and planning methodologies. This article argues that many of the feminist/gender critiques of development planning methodologies centre on the same concerns expressed by proponents of the capability approach (CA). Specifically they share a concern with: the distinction between achieved outcomes, and the freedoms or opportunities that underpin these; a concern with the institutional and relational spaces that determine these freedoms; and a concern with the autonomous agency of women and men. Furthermore, this article suggests that, at the project level, structuring the applied research that is used to inform project design around these concerns can help to improve a project's scope to promote gender equality.

The Capability Approach and Gender Policy and Planning: Shared Priorities
The CA has evolved from Sen's critique of understandings of development that focus primarily on the outcomes people achieve, and, in particular, on income as a material measure of these outcomes (Sen 1999). In contrast, Sen proposes that the focus of development interventions should be to expand people's freedoms to pursue goals that they value. In this light, a central pillar of the CA is the distinction between "functionings", which are women and men's actual achievement, and "capabilities" (or capability sets), which are the abilities and opportunities that women and men have to achieve the goals that they value (whether or not they do, in fact, act on these). Application of the CA would therefore change the goals of a development intervention, and how the impacts of a development intervention should be assessed. Rather than focusing on the outcomes (functionings) that are achieved, the focus is on expanding the capabilities or the freedoms to achieve valued functioning, which requires expanding what Frediani refers to as people's "capability space, which includes people's choice, ability, and opportunity to transform resources into achieved functionings" (2010,178).
This distinction between equality of outcomes and equality of freedoms, or opportunities, is also a central concern of gender policy and planning, and is reflected in most policy statements on gender equality. This is typified by the OECD DAC guidelines on gender equality and women's empowerment, which built on the Beijing Platform for Action to stress that "the aim is not that women and men become the same, but that their opportunities and life chances become and remain equal" (OECD/DAC 1999, 13).
In attempting to understand the differences in women and men's opportunities, gender policy and planning approaches have increasingly recognized that it is problematic to focus on equality between women and men as two distinct social/interest groups. Such a binary division fails to recognize the diversity of the experiences and interest of different groups of women and men, as gendered identities intersect with other sources of social identity, such as class, race, or sexuality. For example, research and development interventions focused on masculinities have highlighted that "men do not form a singular interest group, as gendered norms may favour the interests of some men, but limit the opportunities and freedoms of men and boys who do not conform to hegemonic notions of masculinity" (Connell and Messerschmidt 2005). This implies that, in addition to targeting women and men as beneficiaries or client groups, planning for gender equality requires interventions that address "institutional practices and social relations through which disparities are reinforced and sustained" (OECD/DAC 1999, 13).
Such gender norms and institutions are critical determinants of the opportunities that different people have to realize their aspirations, and are therefore clearly an important determinant of different groups of women and men's respective capability spaces.
Another concern shared by gender planning and the CA is a focus on the expansion of agency, and the need to distinguish between interrelated but different goals of the expansion of agency and the expansion of well-being (Sen 1990). Specific definitions of agency are contested, but agency can be defined broadly as the ability for women and men to freely make choices and pursue goals that reflect their priorities. A concern with supporting the autonomous agency of women and men is reflected in the CA's emphasis on promoting the freedoms for people to achieve their own valued priorities. This has required, however, proponents of the CA to reflect on the factors that can distort or limit "free" choices, with some debate over the extent to which factors such as adaptive preference (Sen 2002) or internalized social control, reflected in notions such as Gramsci's hegemony or Bourdieu's habitus, can limit subaltern groups' ability to articulate and pursue their interests. Thus, in using the concept of adaptive preference, Sen engages with how social norms and conditions can preclude an awareness of inequality, and on this basis he advocates the importance of collective action and democratic value formation (such as feminist or class-based movements) to question such norms (Sen 2002).
This concern with agency overlaps with a focus on the social "position" of women and men (Batliwala 1994), and social norms, as determinants of the opportunities that are available to different groups of women and men. This is because, as Sen puts it, "No individual can think, choose or act without being influenced in one way or another by the nature and working of the society around him or her" (2002,80).
A preoccupation with agency is also shared by advocates of gender equality and women's empowerment. A critical task for those working to promote gender equality is to ensure that policy and planning engages with women and men's own values and aspirations, rather than instrumentalizing them for agendas that they do not identify with. A cornerstone of gender planning, therefore, is identifying and acting on the needs and priorities expressed by women and men themselves, using conceptual tools such as Moser's (1993) Practical and Strategic Gender Needs or Molyneux's (1995) Practical and Strategic Gender Interests.
As with the CA, work on gender equality is concerned with hegemonic structures that may distort value formation and the definition of gender interests. In fact a gendered analysis of agency has added much to debates on the CA, building on Sen's own work on gender and women's agency and well-being (Sen 1990(Sen , 1999, including how the CA deals with questions of individual versus group and relational capabilities. One complicating issue here is that agency cannot be understood on purely individual terms because "Taking people serious as agents means taking their relationships and commitments to other people seriously as well" (Peter 2005, 25). For example, Uyan Semerci's study of women migrants in lowincome settlements in Istanbul suggests that, for many of them, one of their most important sources of well-being is the achievements of their children, and that having children is a source of freedom and empowerment for them as "their status in society depended on their children" and "they were proud to have their children achieve higher education, to have jobs and so forth" (2007,210). On the other hand, research also suggests that women's decisions to prioritize their needs of their families over their own needs is often not a free choice, but is the result of social norms and pressures leading to practices of "compulsory altruism" (Land and Rose 1985).

Valuing time 49
Unpacking the extent of women, or men's, autonomous agency is therefore further complicated by the fact that gender norms are not usually enforced through overt mechanisms of social control. Rather, research suggests that unequal gender norms are often promulgated through hegemonic value systems that subaltern women and men themselves internalize (including norms such as "compulsory altruism"). This can lead to the apparent satisfaction of women and men with social structures that, to an outsider, appear not to be in their interests. In addition, it has been argued that another way in which unequal power structures may gain the consent of the very groups that they oppress is by ensuring that these subaltern groups have some stake in the maintenance of these structures. Kandiyoti (1998) makes this point in her discussion of how women "bargain with patriarchy". On the other hand, however, the idea that subaltern women (and men) may have "true interests" (Lukes 1974) that are hidden from them by invisible, hegemonic power structures, but visible to development professionals as outsiders, is problematic, and appears to contradict the principles of participatory development. In this light, one of the critical challenges for gender analysis is to try and discover the extent to which women and men are in the position to make free, and informed, choices about whether to challenge, or adhere to, the existing gender division of labour and gender norms in the societies in which they live.
To summarize, therefore, the CA and approaches to gender policy and planning share a number of preoccupations, three of which will be explored further in this paper. Firstly, they share a focus on the need to distinguish equality of outcomes (or functionings) from equality of opportunities (or capabilities). Secondly, they both highlight the importance of focusing on social relations and institutions as determinants the "capability spaces" that allow for equal freedoms for different groups of women and men, rather than simply focusing on women and men as social groups, or categories. Finally, they both stress the importance of agency, both as a valued goal in itself and as instrumental to articulating and expanding other capabilities that are crucial for well-being, but highlight the difficulty of unpacking what "free choice" about gender norms means in a given context.

Time Use Survey as an Entry Point for Bringing a Gender Perspective to Development Planning
As discussed above, therefore, policy discourses on gender equality emphasize a number of principles in common with the CA. However, for these principles to be reflected in interventions that are implemented in practice, they need to be reflected in the planning tools that determine the design of interventions (such as baseline research) and the processes of design (i.e. planning activities that give space for the inclusion of the expressed needs of different groups of women and men, and girls and boys).
A number of gender analysis methods are central to this endeavour, including, for example, the Moser/Development Planning Unit (DPU) Gender Analysis methodology (Levy 1998;Moser 1993) or the Harvard Gender Roles framework. These frameworks create the scope to bring a gender perspective to development planning activities. The DPU methodology, for example, sets out to challenge common planning assumptions about the lives and needs of women and men in projects' target populations by: disaggregating units of intervention (such as "the household," "the poor," or "communities"); by exploring how gender relations determine the gender division of labour, and differentiate access to and control over resources; and by recognizing the different practical and strategic gender needs expressed by different groups of women and men.
Clearly, empirical evidence is a crucial input to such analysis. Collecting evidence about the lived realities of women and men requires research tools that, on the one hand, are informed by gender analysis frameworks, and, on the other, provide relevant and disaggregated data which provide a sound basis for gender analysis. Given the earlier discussion, therefore, it important that gender analysis frameworks and the data collection tools that support them are properly able to reflect the core principles common to the CA and gender planning outlined above. In practice, however, there is often a tendency for tools to be used in ways which are mechanistic and narrow, thereby limiting their scope to reveal the extent to which gender relations determine women and men's different freedoms, opportunities and space to enact autonomous agency. It is therefore useful to revisit such tools, and reconsider how some of the CA principles could be better integrated into their design and application.
One set of research tools that is routinely used as a basis for gender planning clearly reflects this need. These are the tools designed to collect data on women and men's daily time uses. Such tools are fundamental to the cannon of gender planning as they offer a very immediate insight into the gender division of labour, by highlighting the different patterns of women and men's routine activities. These data highlight the kinds of support (i.e. practical gender needs) that women and men need to conduct their existing gender roles. It can also reveal how much free time women and men have to participate in development interventions, and when this time is available; or (as is frequently the case with poor women and men) if they do not have free, time use analysis can highlight the existing tasks that will need to be sacrificed for them to participate in interventions.
There has also been an increasing focus on time use research as an input to gender analysis, as a result of the growing recognition of the importance of time as a feature of poverty, and in particular as a gendered feature of poverty. This is based on the recognition that income deprivation is only one aspect of ill-being, and that, as Chant (2006) points out, the current focus on the "feminization of poverty" might be better understood in terms of the feminization of responsibilities for care, and the associated time demands that this places on women in the context of poverty. The understanding of time as an important gendered aspect of poverty also resonates with relational conceptualizations of poverty (Mosse 2010) because a critical issue is the way that gender relations differently affect the space for women and men to negotiate and control their own time uses.
As a result, time use survey has been increasingly employed as a tool for monitoring gendered aspects of poverty and well-being at the national scale, with many countries now engaging in periodic national time use surveys, which supplement other national poverty and development indicators (Esquivel et al. 2008). However, time use surveys with women and men are also a basic staple of the gender studies that are often conducted to provide baseline information for the design of development projects and programmes. Time use surveys at the project level are usually based on quantitative questionnaires, and/ or Participatory Rapid Appraisal (PRA) tools such as the Daily Activities chart used by the DPU gender planning programme and organizations such as the SDC with whom the DPU programme works (SDC 2005) or the "24 hour clock" used by organization such as the FAO and IFAD (International Fund for Agricultural Development 2000).
However, the collection and interpretation of time data is very strongly influenced by the underpinning ideas about some of the issues related to capabilities discussed earlier. For example, a standard definition of time-poverty is when "some individuals do not have enough time for rest and leisure after taking into account the time spent working, whether in the labour market, for domestic work, or for other activities such as fetching water and wood" (Blackden and Wodon 2006, 6). This definition implies that the way to address time poverty is to expand women and men's free time (for rest and leisure). However, from a capability perspective this is too narrow. Firstly, a lack of time is not necessarily indicative of poverty-a well-paid, but workaholic, lawyer for example, could not be considered time poor as he/she is in a position to spend less time on work but has chosen not to do so. Secondly, where this element of choice is not present (i.e. where the time-poor are poor are also income-poor), more free time does not alleviate poverty unless there are opportunities for women and men to use this time in ways which they value. In this light, Burchardt (2008) approaches time poverty as something that is experienced by those who could avoid income poverty only by incurring time poverty, and proposes an understanding of "time capability", as the various time and income combinations that people could have, in the context of the policy environments that affect these choices.
Often, however, both the collection and analysis of time use data is conducted in ways that do not reflect this, more nuanced, CA. The data collected may therefore be used to justify interventions that are not in fact appropriate to the needs of poor women and men. Application of some of the core elements of CA thinking, discussed earlier, would help guard against this. The following section describes a case study that shows how these CA principles can be integrated into time use survey.

Valuing time 51
Mercy Corps' Alliances Project Mercy Corps' "Market Alliances against Poverty" project is a rural economic development project working in the South Caucasus nation of Georgia. As part of the baseline study for this project, a gender survey was undertaken, which included the collection of time use data as an input for gender analysis. What is interesting about this survey is the way that it attempted to go beyond only quantifying and accounting for changes in women and men's time uses, to exploring how project activities affected women and men's experience and control of time, and how, or whether, the project expanded the opportunities for women and men to pursue meaningful alternative uses for their time. In this light, the project's analysis of time uses has, in practice, incorporated some of the core principles of the CA approach outlined earlier in this paper.
The Market Alliances against Poverty project has been running since 2008 in the region of Samstkhe Javakheti (Georgia), implemented by the international non-governmental organization Mercy Corps, with funding from the Swiss bilateral development agency (SDC/DEZA). Samstkhe Javakheti, like other remote rural regions of Georgia and its neighbouring South Caucasus countries, is characterized by a high level of rural poverty, with 64% of residents in the region falling beneath the national poverty line. It is also characterized by ethnic diversity, with a high proportion of ethnic Armenia residents, as well as Ajarans (an ethnic minority group from the west of Georgia).
The project is working to address rural poverty by improving linkages between farm households and markets for their produce, using a strategy developed through the M4P (Making Markets Work for the Poor) framework (DFID/SDC 2008). In particular, given that (according the project's inception survey) 90% of rural households in the targeted areas own cattle, and that there is market demand for meat and dairy products in Georgia, the project has focused on supporting market access for livestock producing farm households, with a focus on both meat and dairy value chains.
In 2011, recognizing the need to have a stronger understanding of the extent to which the project strategy was able to increase the economic opportunities of both women and men in low-income farm households, the Alliances project conducted a gender survey as a part of the project's wider baseline study. The gender survey was based on the gender analysis framework outlined in the SDC/DEZA Gender Toolkit, which is coherent with the Moser/DPU gender analysis methodology, and also included a component of time use survey with women and men.
An initial gender study had already been undertaken, which was focused mainly upon intra-household dynamics and labour division, but which did not analyse constraints and opportunities in the market for women. The 2011 study went further, to consider not only "horizontal" dynamics (at household level) but also "vertical" dynamics, to answer questions such as who benefits along the dairy and meat value chain, what power do women or men gain/lose as a result of value chain changes, and what other options do different groups of women and men have to participate in or benefit from value chains? This gender survey was based on: desk review of documents; field focus group discussions; a client questionnaire; interviews with local government officials; interviews with finance institutions and a local university/professional training centre; and review of the "Alliances" project monitoring and evaluation documents.
The focus group component of the research involved a series of discussions held in the three municipalities (Adigeni, Aspindza and Akhaltsikhe) of the Samtskhe-Javakheti region. Of the eight villages targeted, three were Armenian villages (to try and capture any potential differences between Georgian and Armenian populations). In each village, separate male and female focus group discussions were organized, reaching a total of around 120 participants, with between 50 and 60 attending the women's focus group discussions and between 50 and 60 the men's focus group discussions. An open-ended questionnaire was used to structure the discussion, which included questions related to time use.
One of the key time use changes identified during the gender survey was the impact of the project's support to dairy processing enterprises. As a part of the strategy to increase market access for farm households in the target area, the project has been working to support dairy processing businesses to collect raw milk from milk collection centres (MCCs), thereby increasing the quality of dairy products produced and increasing access to urban markets. This involves a move away from the artisanal processing of dairy products (cheese, butter, yoghurt) in farms, which is almost always done by women. Currently for many households, in the absence of refrigeration technologies, such home processing is the only way of preserving perishable milk (particularly in summer months when temperatures in the region are extremely high). Farm-processed dairy products are used both for household consumption and/or for sale in local markets. According to the findings of the gender time survey, artisanal processing of dairy products in farms by women typically takes two hours a day. In this light, the project support the sale of raw milk through MCCs would save women a considerable amount of time on a day-to-day basis.
If we take Blackden and Wodon's (2006) definition of time poverty, discussed earlier, this expansion of free time would represent a straightforward reduction in women's time poverty. However, if we apply some of the CA principles to interpret these data, this might not necessarily constitute a straightforward step forward in women's well-being, and the findings become more complex to analyse, with some contradictory tendencies. For example, the project's gender analysis also highlighted that, in the context of social norms that limit women's mobility, sale of cheese provides an opportunity for some women to visit the towns that act as market centres. This includes some women farmers, as well as intermediaries who buy cheese from other women and then take the cheese to market. According to the survey: Some women like to go to market for selling produce and for many, in addition to buying products needed at home, it is the only chance to "see the town" and visit some shops. This trend varies in different villages. In the highland communities, populated mainly by those from Ajara, it is still a problem for a woman to go to the market because of the very traditional and conservative culture still existing in these communities. Therein, the women either do not express an interest or are limited in their ability to access markets because of the perception that "a woman should sit at home and take care of the family and the children".
On the other hand, according to another informal economy study undertaken for the Alliances programme by Mercy Corp, some women who sell their cheese via intermediaries rather than going to market themselves consider this an additional advantage, in terms of time saved.
In addition, in terms of decision-making on the use of household budgets, the survey found that women had some more discretion over income from the sale of cheese that they had processed than over other sources of household income (although this discretion remains within certain parameters, i.e. that expenditures are for "household needs"): When a woman sells produce in the market … she can buy the goods for family independent of her husband in that she knows best what is required for the household.
What was not clear from the survey findings to date, and is a topic for further research, is whether or not women have similar discretion over income from the sale of raw milk at MCCs.
It can therefore be seen that, rather than justifying an assumption that the reduction in time spent on dairy processing represents a straightforward reduction in women's poverty through time savings, a more capability-influenced analysis of the impact of the move to sale of raw milk would also need to factor in a possible reduction in some women's control over income from the sale of dairy produce, and/or the lost opportunities for visiting market centres.
In this light, if the project had only quantified and analysed raw data on singular issues, such as changes in time use, mobility or control over income, they could have come to quite contradictory conclusions. The project approach, therefore, has been to use these data as a basis for discussion with women and men in farm households that, in line with the CA principles, focuses on how the women in question value and prioritize outcomes such as time saving, mobility, or control over income, and looks not only at the free time created, but also at the opportunities for using this free time, and the changing attitudes (of women and men) about women's control over their time. The findings based on these qualitative discussions can give a more nuanced understanding of the project's impact on the alleviation Valuing time 53 of women's poverty than a mechanistic interpretation of time use changes. These qualitative discussions, and the associated findings, are revealing in light of the CA principles discussed earlier.

Time use, Functionings and Capabilities
As discussed above, an impact of the project's support to MCC and the collection of raw milk is a reduction of the time that women spend on milk processing, and the project has taken on board this finding in a number of ways. From a capability perspective, however, the question is not only how much "free" time is created, but what women are able to do with this time, and whether they can use it to pursue goals of value to them.
While not explicitly using a capability framework, the gender survey has acted as the basis for research that employs the CA principles. The project is currently undertaking ongoing monitoring on "how women spend time saved as a result of programme intervention and in what way it contributes to improvement of women's lives", which focuses on the alternative uses that women are putting the extra time to in practice (i.e. the new time use functionings that women are able to realize) and on exploring the additional activities that women are in a position to use this time for, regardless of whether or not they actually choose to pursue these activities (i.e. the expanded capabilities that this extra time creates for women). As will be explored below, the central concepts of the CA can help to guide this research, and give a clearer understanding of how the project is affecting gender equality.

Time and Capability Spaces
To better understand the second question, of how the additional time has expanded women's capabilities to pursue goals of important to them, it is important to explore the "capability space" that women have to use this time.
The gender survey set out to explore the constraints and opportunities that women have in using their time, and in making decisions about their own time uses, thus shedding light on changes in women's capability spaces.
What is clear from the findings of the gender survey is that while time is one constraint to women's capabilities, there are also other contextual constraints, including factors that affect women's scope to be economically active, as well as social attitudes about gender roles. In this light, while the free time generated by sale of raw milk has expanded women's capability spaces to a certain extent, this expansion in turn might be limited by norms about women's use of and control over their own time. Some responses from the gender survey are quite revealing of attitudes that women's chief focus should be on domestic work and caring for the family (apparently prioritized over the expansion of economic activities that is envisaged by the project strategy). One of the male respondents from Tskruti village stated: I am all for selling raw milk. If you make the calculations, you get the same income through selling raw milk in the summer time and my wife gets free time to spend on the family. Women can knit socks using that time and make sure that their husbands' feet are kept warm.
On the other hand, there is clearly a level of debate and change in gendered attitudes of women's time use, at least as far as women's right to leisure is concerned, as demonstrated by the contrasting point of view from another male respondent: She is also a human being and she also needs time to sit with the neighboring women, drink coffee and chat. (Male respondent, 32 years old) Women's responses also give some insights into how the "capability spaces" in which they operate determine how they are able to use the extra time generated by the move to sale of raw milk. Importantly: The majority of the women interviewed stated that there are no alternative employment opportunities within their communities where they could spend the two hours they saved from making cheese. Although: Some of the women see the profitability of the time saved as it would be spent in their gardens to grow vegetables. This time saved provides them with an opportunity to properly irrigate and weed vegetable plots which leads to increased productivity and, accordingly, increased incomes after they harvest and sell the produce.
Furthermore, some women see the time created as a means to increase their ability to address some of the barriers that currently inhibit their earning potential and economic participation: Armenian village women prioritized the learning of the Georgian language which, in many cases, prevents them from accessing trainings or using demonstration plots which could also add to their knowledge.
In addition, the survey found that the opportunity to save time through raw milk sales did not create the same opportunities for all women, due, again, to differences in women's capability spaces. Switching from cheese making to raw milk sales, however, did not seem to be profitable for those women having less than three milking cows, who stated that their milk volume was insufficient to supply to a MCC and, at the same time, produce cheese for their own consumption:. I wish to have more time to spend on other things rather than making cheese. But I have no other choice. I have only one milking cow and I make cheese for my family. Whatever is surplus, I am taking to sell at the Akhaltsikhe market. (Female respondent, 54 years old) The situation also differs in the villages located far from Akhaltsikhe (the main market centre) where milk collection and raw milk sale practices are not known. After considering the benefits of raw milk sale, some of the women in these communities are willing to adopt this practice whilst others still think that it is more profitable to make cheese, as they always make and have cheese for family consumption which they can store until late autumn when the price is significantly higher in comparison with the summer. Additionally, most of the rural families keep pigs, and the whey that is left over from the cheese-making process is used to feed them.
Generally, therefore, while some women can generate economic opportunities from the time saved through raw milk sale, the environment in which they live limits the (economic) opportunities that this additional time can generate for most of them. This does not, nonetheless, imply that this additional time has no value for them. From their responses, it is clear that women also see value in the opportunities that additional times gives them for pursuing activities other spheres (primary related to their domestic and caring roles), as discussed during the survey: . Women see the opportunity to deliver "quality service" to their families. Two hours saved enables them to serve their family members "without running", which results in the provision of quality service and, accordingly, the "satisfaction" of their family members. . Some of the women see the benefit of extra time spent on their children's education. Because of their workload, women pay less attention to their children in doing their homework. The quality of education is a precondition for good farming practices or employment in the private or government sectors that can lead rural families on the path to increased incomes.
Valuing time 55

Time use and Autonomous Agency
Another question that the Alliances gender survey touches on is the extent to which, given the social context in which they operate, women are able to make "free" choices about what goals they wish to pursue and how they want spend their time. This is something that the Alliances project's gender survey has begun to explore, thereby creating a number of interesting entry points to build on in their ongoing research.
As discussed above, the responses from women and men during the gender survey suggest that there are strong social norms about appropriate time uses for women in Samstkhe Javakheti, and furthermore that the nature of these norms differs in intensity in Ajaran, Georgian and Armenian ethnic communities. Such norms include issues around women's mobility in public spaces outside the home, about women's responsibility for domestic tasks, and about their decision-making roles, all of which affect decisions about how, where, and to what ends women spend their time. These norms are largely accepted by both women and men, although with some instances of women and men questioning them. Such norms are clearly relevant to the Alliances project's central focus on increasing opportunities for the economic participation of women and men in rural households in Samstkhe Javakheti.
According to the survey findings, "existing inequalities in the division of domestic tasks are perceived as a natural order". It is interesting to note in this light that many of the opportunities for the use of extra time generated by raw milk sales mentioned by women during the research (discussed above) relate to women's customary reproductive role (care and domestic work). As discussed earlier in relation to autonomous agency, norms that support women's predominant role in domestic work, or women's altruism, are not inherently bad, as long as this is a division of labour that women actively choose and support. To understand what this means for women's autonomous agency it is important to understand the extent to which these norms are voluntarily adopted by women, or imposed on them by (hegemonic) social norms, or their acceptance of their circumstances (i.e. lack of opportunities for alternative time uses, such as economic activities, in their capability spaces, creating some form of "adaptive preference"). However, as discussed earlier, the possibility of hegemonic influence over value formation makes this a problematic area to analyse. So what does this mean for research, and specifically for research into gender relations in the project area?
Firstly, in order to understand whether decisions about time use are made as "free choices", gender research into issues such as time use needs to go deeper in exploring the trade-offs that women and men make, the priorities that they have, and what motivates these. Simply asking questions such as "would you like more free time?" is not useful, unless, as was the case with the Mercy Corps study: . the trade-offs that this free time involves for women (in terms of, for example, access to markets and increased control over earnings) are also discussed; . the value, and source of value (e.g. the symbolic and social status of different time uses for women and men) of these time uses is reflected on; and . the scope that women have to decide how they will use this additional free time is researched.
Secondly, in line with the principles of participatory, or action research, the research process itself can be a means of expanding women's (and men's) agency. Freire's concept of conscientization, as a means for women and men to critically reflect on gender relations, is as relevant for research as it is for education (Freire 1997). Processes of gender research, such as Mercy Corp's time use survey, by making visible differences in women and men's time uses, and decision-making about how they use their own time, can in themselves begin to change attitudes about the gendered nature of time uses and control over time, through creating spaces (e.g. in focus group discussions) for critical reflection and exposure to alternative practices and norms.
This could be an interesting entry point for the project to document and explore changing attitudes about women and men's gender roles and time uses. While such research has not yet formally been factored into the Alliances gender survey, members of the survey team shared anecdotal memories of the reactions of women and men during the focus groups, as they publicly discussed and realized how unevenly women's and men's time is used. A further reflection that this raises is that, if research into time use attempts to explore issues of agency by engaging in critical reflection on time use opportunities, it becomes impossible to distinguish the research findings (on expansion of agency) from the research process itself-as the research becomes one of the conversion factors that may expand the agency of the participants.
To date, the gender survey component of the baseline study has informed the Alliances project in a number of ways. Specifically, the findings acted as an information resource to bring a gender perspective to the development of the project strategy. The findings of the gender survey showed that in order to increase both women and men's access to markets, and increase their incomes (the stated goal of the project), it would be necessary to modify project interventions to address some of the specific barriers to women's participation in, and control over, agricultural value chains.
The specific findings on the time use impacts of selling raw milk rather than in-farm processing of dairy products are important for the programme, since women in the project area were identified to be time poor during the gender survey. This has acted as a basis for further monitoring, and monthly data are now collected on a monthly basis to understand how much time was saved by women. The project is now working with a research institution to do an impact assessment and examination of "how women spend time saved as a result of programme intervention and in what way it contributes to improvement of women's lives", which will be used as a basis to steer future project phases.
In addition, while time surveys showed that women are the main labour force in livestock farming, and especially in the dairy sub-sector, the gender survey also showed that they have very limited access to agricultural services and markets. Accordingly, the project has applied mechanisms that seek to promote women's direct access to agricultural services (such as artificial insemination or veterinary services) and to income generated through sales of farm produce, with a distinction drawn between women in mixed households and independent women farmers in female-headed households.
The gender survey also highlighted women's limited access to public decision-making, as well as a lack of awareness of gender equality legislation within municipalities, and has therefore developed a focus on promoting women's role in the governance of agricultural markets. The project has therefore followed up with in-depth interviews with local government staff in the project area about the level of compliance with the national Gender Law and Action plan, and is exploring the scope for Gender Sensitive Budgeting by local governments to better incorporate women's and men's needs into planning processes. The findings are now being used to work with local governments and increase their application of state gender policies, with a view to the increased participation of women in decision-making.

Conclusions
If the research tools used to collect data for gender analysis are both to reveal how planned interventions affect gender equality and to give women and men space to voice their interests in the planning process, then the design and application of such tools need to fulfil a number of criteria. As has been illustrated by the case of the Alliances project, in the case of time use surveys this implies more than quantifying changes in the time uses of women and men, and making assumptions about the impact of these changes on their well-being. This requires using time use surveys as a basis for examining what time use changes mean for women and men themselves. This paper has argued that some of the central tenets of the CA can be used to add these elements to time use research. In particular, the distinction between functionings and capabilities, the examination of capability spaces, and engaging with issues of autonomous agency all add crucial depth to the collection and analysis of time use data. As shown by the experience of the Mercy Corps Alliances project, undertaking qualitative research that engages with these issues is an entry point which can begin to unpack the complexities of how changes in time use affect women and men's well-being and opportunities. Such qualitative research both creates an entry point for further quantitative research into time use and well-being, and can also act a means to interrogate quantitative datasets on changing time use patterns.
Valuing time 57