1.1 THE IMPORTANCE OF GENDER IN STATISTICS
1.2 PRODUCING GENDER-SPECIFIC STATISTICS
1.3 MEETING USERS' NEEDS
1.4 THE PRIORITY DOMAINS OF GENDER-SPECIFIC STATISTICS
1.5 THE PRIORITY DOMAINS FOR THE AGRICULTURAL SECTOR
1.6 THE MAIN INFORMATION SOURCES ON AGRICULTURE AND RURAL DEVELOPMENT
1.7 FAO ACTION ON GENDER STATISTICS
1.8 THE ROAD AHEAD
Women and men do not play identical roles in any society; nor do they have equal access to education, work, career opportunities and economic resources. This means that political and economic leadership is also unequally shared, which leads to gender disparities in the enjoyment of benefits from economic and social development. In recent decades, advocates of women's rights have drawn attention to these facts and the need to consider them in policy and programme formulation.
Since the First World Conference on Women, held in Mexico City in 1975, approaches to "women's issues" have undergone considerable change. The original strategy approach was to treat women as a separate, homogeneous entity in isolation from global policies, and this often aggravated existing forms of discrimination. Projects designed specifically with women in mind were underfunded. Specialized "women and development" units were allocated few resources, so had little say at the policy level. Awareness of this led to a reorientation of approaches and the vision expanded from an exclusive focus on women's concerns to a more holistic view of gender interaction within the full social context - the gender perspective.
This new approach focuses on gender disparities in the impacts of economic and social policies, and the fact that men, women and their interactions affect every aspect of the development process. The gender perspective pays close attention to the mechanisms that regulate gender interactions and their impact on men and women, by making reference to gender-based socio-economic characteristics.
Nowadays, international organizations and governments give greater recognition to the need to strengthen the participation of women in order to achieve sustainable development. However, although the contribution of women is rather more visible now than it was 25 years ago, there is still a long way to go. The lack of adequate data on true gender disparities in everyday life, as well as in the economic, social and political spheres, has frequently given rise to inappropriate policies, plans and projects. The issue can only be resolved by a carefully planned approach to statistics production.
In many developing countries today, much of the rural sector, especially women, live in poverty. Despite the fact that sustainable agricultural development aims at balancing greater productivity and better yields with natural resource conservation, enhanced incomes, job creation and improved levels of food and nutritional security, many development programmes and policies have actually exacerbated poverty or done nothing to improve local standards of living, especially those of women.
Development plans are formulated primarily in terms of economic criteria, while social and human parameters are seen mostly as justifications for economic decisions. When the human factor is given as much importance as the economic aspects, planning exercises become very complex; introducing a gender perspective complicates the issue even more. Planners rarely see the relevance of the gender perspective, partly because they lack reliable, impartial data on the type and extent of men's and women's separate contributions.
In a world in which economic value is reckoned in purely monetary terms, women's work, which is often unpaid, is not considered to be productive work. So, although women are the pillars of subsistence economies and pivotal to food security, their activities tend to be excluded from economic accounts. Agricultural statistics therefore tend to under-represent, or even omit, variables that are essential to a clear understanding of rural sector activities and rural development. This severely limits planners' grasp of the real situation in rural economies which, in turn, constrains their potential to act.
Until a few years ago, the demand for specific data and indicators incorporating a gender perspective was limited to advocates of the rights of women and disadvantaged groups. Nowadays, the user audience has expanded to include decision-makers at every level and in every area of social and economic development. There is greater general awareness of the need for a gender perspective in development policy formulation, and of the corresponding need for pertinent statistics. At the same time, as reliable data become available, they help to promote and justify change and to dissipate doubts and scepticism with respect to the relevance of innovative approaches such as the gender perspective.
In short, statistics incorporating a gender perspective are now essential for:
· Advocates of gender equity, who want them to boost awareness of their concerns;The concepts and methods proposed and adopted by the many countries who seek a more faithful mirror of reality have begun to bear fruit, and new strategies have been developed to improve data presentation and dissemination, incorporating a gender perspective into statistics production.
· Planners, who want them for economic and social policy formulation, implementation and monitoring;
· Development experts, who want to review and analyse gender aspects and interactions;
· International, government and non-governmental organizations (NGOs), who use them in project and programme design, implementation and evaluation;
· The general public, who wants them for a better understanding of society.
Gender-specific statistics constitute a new field. Unlike the simple disaggregation of data by gender, gender-specific statistics are built on concepts and definitions designed to detect gender-differentiated conditions and characteristics and gender interactions. Appropriate data collection methods that capture and reflect the real situations of various groups, while avoiding the distortion that can arise from cultural factors and stereotypes, are an additional prerequisite. The subsequent stages of data processing and dissemination are also of great importance.
Gender-specific statistics must be presented in a form that allows easy access to a wide range of users, many of whom will have no special expertise in this area. Statistics must be simple, plain and easy to understand. Particular care must be taken with the compilation of tables, graphs and maps and the preparation of these for publication because the data cover a range of fields and use material from many different sources.
Gender-specific statistics cannot be produced independently of global statistical systems - in fact, these systems need to incorporate a gender perspective. The national statistical bureaux responsible for the production and dissemination of official data should consider gender-specific data collection, compilation, analysis and presentation as an integral part of their work, not a separate task. The production and improvement of gender-specific indicators should be written into existing data collection programmes, censuses, periodic surveys and sampling, in close collaboration with users, in order to make the best use possible of existing statistical systems and data.
In this context, statistical data producers need to understand the concept of gender in five stages of data production.
Stage 1: Identification of gender issues and their implications for social improvement
The identification of gender issues for statistical treatment requires ongoing and permanent dialogue between statistical data producers and users such as policy-makers or planners.
One major gender issue is women's lack of access to inputs. Commonly held prejudices result in procedures that bar women's access to land and credit. Discrimination, caused by prevailing social and cultural patterns, also affects the division of household labour and the levels of education. The lack of equality in these domains has major implications for women, including greater poverty, precarious nutritional status, heavier workloads and reduced bargaining power.
An in-depth review of the whole issue is needed to identify and analyse causes, effects and interactions. Gender experts, leaders and other users need reliable, appropriate statistics and indicators from which they can reach conclusions, tackle the issues and adopt the appropriate measures to mitigate, if not solve them.
Stage 2: Identification of gender relevant data
The second phase identifies the data and indicators that are required. Statistical data producers and users work together to determine what data and indicators are needed for policy formulation, monitoring and evaluation from a gender perspective.
Stage 3: Reviewing existing sources
Stage 3 assumes that the available data are periodically reviewed by users to evaluate whether their expectations are being met. Such reviews can be used to upgrade existing methods of data collection, develop new programmes and rectify problems concerning omissions, quality and relevance.
Implicit in any review of existing sources is an analysis of questionnaire concepts and definitions and consideration of the criteria for coding and grouping the data to ensure that they mirror gender realities. Reorientation requires an in-depth review of data collection methods, from the design of questionnaires and sampling to the selection and training of enumerators.
Stage 4: Improving existing sources and developing new data collection programmes
Reorienting existing sources automatically reveals any gaps. The concepts and methods used in periodic programmes to acquire, reprocess and disseminate data can then be rectified, and a decision made as to whether or not additional data need to be collected. At this stage, statistical data producers can determine future needs with regard to providing reliable data to meet users' requirements.
Discrimination in employment, to mention only one significant aspect of gender equity, often calls for highly detailed tabulation (usually in three figures) to avoid the grouping of different remuneration levels and categories under a single heading. Labour force surveys and censuses provide a wealth of data on employment, but some data are either too generalized or else are not brought into the equation at all. New data may emerge if the old data are regrouped. Fresh data may need to be collected, but the gaps can usually be filled by simply adding a few new questions to existing surveys.
The dynamics of this process inevitably lead to the search for new programmes. In this context, the introduction of time-use studies, which only a few countries include in their regular data collection programmes, would be a welcome innovation for the global statistical programme.
Stage 5: Data compilation, analysis, presentation and dissemination by specific user groups
Data compilation, analysis, presentation and dissemination are crucial to obtaining gender-specific data, because such data often touch on sensitive issues about which little is known or which call for new thinking. When data are well presented, they reach a wider audience of specialized and unspecialized users, while the correct and appropriate analysis of available statistics can help to avoid user bias concerning gender.
During this stage, producers and users need to take a joint look at strategies for disseminating findings. This means tailoring the product specifically in terms of needs, demand and level of expertise (written, visual and electronic publications), and putting together data dissemination campaigns for both established and potential markets.
The users of gender statistics cover a broad spectrum that includes policy-makers, planners, gender experts, the general public, national and international development agencies, NGOs, research institutes and the media. Each of these categories has its own way of reasoning and its own conceptual and technical expertise. Each uses statistics in accordance with its own particular familiarity with statistical analysis and its own understanding of gender considerations. In this context, permanent and effective cooperation between statistics producers and users is needed to produce appropriate data. The following need to be known:
· Target groups - ministries, development agencies, mass media, research workers, gender equity advocates;Data users and producers need to work closely together to incorporate a gender perspective in all stages of the data production process. Data producers, who have the technical expertise and the tools to meet users' needs, bear the major responsibility.
· Specific needs of different groups of users;
· Desired product formats for each target group (publication/diskette/CD);
· Resources and strategies available for product dissemination, by statistics product and audience characteristics.
Statistics producers need to broaden their horizons to encompass social concerns regarding gender issues, and link them to the technical and methodological aspects of statistics production. They must be able to discern which data and analyses users want, even when users may have difficulty in expressing their needs or are unfamiliar with data handling. In addition, since this effort benefits everyone, statisticians should work to broaden their audiences and reach potential users of statistics. Brochures should be attractively designed to hold readers' attention (including those who are normally put off by statistics) and to generate a demand for more detailed statistics. The support of top management is crucial in securing the necessary human and financial resources to carry out these activities.
For their part, users need to develop a basic set of skills so that they can read, interpret and use statistics correctly. They need to have a clear idea of the prerequisites of statistics production to grasp the work that statisticians do. This will enable users to frame sensible requests and provide feedback to statisticians on how they use statistics in their everyday work and what progress they have achieved.
Effective ways of strengthening the collaboration between users and producers to improve statistics from a gender perspective include joint workshops, the inclusion of users and statistics experts on all technical committees, and the participation of statistics producers on social and economic policy development bodies.
Dissemination strategies need to be developed very early in the production process, with input from representatives of the target audience. Ways of improving the dissemination of published data include: workshops1 to launch new products; the involvement of academic bodies; use of the media (press releases for TV, newspapers, magazines, brochures, etc); school and university presentations; and high-level meetings.
¹ The UN's International Research and Training Institute for the Advancement of Women (INSTRAW) has held a number of workshops for the users and procedures of statistics to improve statistics on women (e.g. Costa Rica, Pakistan, China and Ecuador).
In recent years, there have been a number of efforts to produce gender-sensitive statistics. Users are becoming clearer about gender considerations and demanding more sophisticated statistics. Statistics producers, who are increasingly aware of socio-economic issues, now tend to include such data in their work.
In the interests of balanced, sustainable socio-economic development, domains that are still insufficiently documented and require more detailed data (and statistics) have been identified. These include food security and the eradication of poverty, for which there is an increasingly urgent need for gender-specific indicators concerning:
· The extent of men's and women's unpaid work, particularly for food and agricultural production;New concepts, definitions and data collection methods have been adopted by statisticians and various international specialized agencies to ensure better coverage of these aspects. This has led to time-use studies that generate more reliable data on unpaid male and female work, and to the subsequent reorientation of the UN System of National Accounts so that it targets the economically active population more accurately; all of which illustrate the current commitment to improving information and incorporating a gender perspective into official statistics. However, other topics, such as household resource allocation, have barely been touched on.
· Male and female access to resources, inputs and employment;
· A clearer understanding of household characteristics, especially factors in poverty and household resource access and allocation;
· The extent of male and female participation in economic, social and political decision-making.
Poverty, food insecurity and environmental degradation are three key issues in the achievement of sustainable development, especially rural development. Policy-makers must remember that rural men and women are stakeholders with key roles to play in food production and natural resource management, and that any measure adopted to solve these problems will, therefore, have an impact on rural men's and women's living conditions and gender interactions. There is a clear need to base policy and action on adequate statistical data.
The FAO Programme for the World Census of Agriculture 2000 highlights the priority domains that need to be explored in order to detect production patterns, identify the population segments involved and devise items that will facilitate the inclusion of a gender perspective in analyses of the situation.
A broad range of data on the human and production aspects of the rural and agricultural sectors are available in such sources as agricultural, population and housing censuses, as well as in farm, rural labour force and employment surveys. Other useful sources, which do not necessarily target only the rural sector, include food consumption, household income and expenditure surveys and time-use studies. The scope and limitations of these sources are reviewed in Chapter 4.
At FAO, there is widespread recognition that the statistical coverage of human resources in the rural and agricultural development context is weak. Various workshops and seminars have been held to discuss possible orientations and resources to solve this problem. In 1991, FAO organized an interagency consultation on the enhancement of gender-disaggregated databases and statistics. On that occasion, the data sources were reviewed and the following specific recommendations were made:2
· Define "head of household" accurately;
· Define the economically active population appropriately;
· Promote the measurement of secondary activities;
· Ensure that holdings are identified through the households concerned;
· Improve linkages between the population census and the agricultural census;
· Develop standard criteria to determine the minimum size of holdings covered;
· Ensure that multiple holdings are not amalgamated into one holding and that joint holders are identified;
· Make subsistence agriculture more visible;
· In the agricultural census, include indicative questions that provide a framework for sub-sampling in further investigations.2 FAO, 1994. Improving gender-disaggregated data on human resources though agricultural censuses. Women in Agriculture No. 8, Annexe B, p. 4. Rome.
At the First World Conference on Women in Mexico City in 1975, it was stressed that better statistics were urgently needed to provide decision-makers with the necessary background data on women. Since that time, many international and national organizations have opened up dialogue between policy-makers and statistics producers, and there is growing awareness of the importance of adequate statistics at all levels of policy formulation and decision-making.
The general approach to development has undergone a profound modification in recent years. It has moved from viewing women as separate entities outside the social and family context (women and development) to embracing a more holistic view that includes gender considerations and gender interaction (gender and development). At the same time, the working hypotheses used in statistics have been revamped, and now consider gender roles and how they complement one another, as well as seeking to avoid the biases and prejudices that creep into the design of statistical methods. The gender perspective is becoming an integral part of statistical systems.
Traditionally, agricultural censuses and surveys focused more on production than on human resources, and on large holdings rather than small production units. The most difficult aspect of obtaining gender-specific statistics in agricultural censuses is the quantification of women's work. Women tend to combine productive and household tasks, switching between the two or doing both at the same time, so the relevant analytical units (household and holding) become interlinked and often overlap in a way that cannot be differentiated.
The statistical evaluation of women's contribution in the agricultural sector and rural development goes beyond the mere measurement of their work. While quantification is obviously essential, other equally basic factors are men's and women's different access to inputs and resources and the amount of time each has available. The fundamental gender considerations in agriculture still tend to be overlooked in permanent and periodic data collection efforts. Gender-differentiated land use and landownership, as well as access to sources of credit, training and extension services, technology and other essential resources, are still not receiving due consideration.
The main constraints on data collection concern the inadequate concepts and definitions used and methodological weaknesses arising from sexist biases. These can be overcome by reorienting, adapting and improving concepts and methodology at all stages of the data generation process - questionnaire content and formulation; preparation and implementation of fieldwork; coding and grouping; presentation of results (tables or more sophisticated indicators); and dissemination.
Before tackling these conceptual and methodological aspects as a whole, a series of considerations concerning gender need to be looked at. It is vitally important that statistical data users and producers have a clear idea of the significance, scope and implications of the concept of gender.