Chinese Statistical Systems as Evidenced by the Tianjin Project

Soon after the founding of the People's Republic of China, the government established a centralized and unified statistics system in the National Bureau of Statistics (NBS) under the State Council. The NBS, according to the Statistics Law of the People's Republic of China (1983, last revised 2009), is the main provider of national statistical data and is responsible for organizing, directing and coordinating statistics work throughout the country. This agency formulates and supervises the implementation of statistics policy, and plans the national statistics system based on unified standards and methods for both surveys and reporting-based collections of data – two of the principal components of the Chinese statistics system. These components make up the government's comprehensive statistics structure (also known as the statistics bureau structure), and the line ministries' (or sector ministries) statistics structure.

According to the Statistics Law, over 40 statistics departments are located in the different line ministries and agencies.Footnote ⁹ These serve as major channels for the compilation of data from reporting structured upwards from county to national levels in each domain. The structure is hierarchical. The statistics department in a ministry of the central government, such as education, health or family planning, compiles statistics submitted by departmental offices in local governments at and/or above the county level. There are, therefore, many different levels within the structure responsible for the collection, production, analysis and dissemination of data. Other domains that follow this structure include traffic, tourism, finance, customs (import and export), foreign capital utilization, culture, railways, science and technology, public security and registered permanent residency, and social development, including registration of non-profit organizations, social assistance, and social security.

Often, the statistics gathered reveal differences in a single indicator, such as the number of new births in a given year. Differences exist because, in this case, family planning agencies might collect one number based on contacts among women in general and pregnant women specifically, while hospitals might collect another based on the number of deliveries in hospital (China claims to have reached a hospital delivery rate of over 98 per cent), and police in the public security system might collect yet another number based on household registrations. The Chinese government addresses such differences by submitting the data from each domain to the NBS for final corroboration and compilation into the national statistical yearbook or other collections for public use. According to the stipulations of the Statistics Law, each ministry must develop its own reference points and supplementary standards that govern its statistics work and submit these to the NBS for examination and approval. The NBS then acts as a coordinating body for the centralized statistics system and holds final authority over the publication of data, and at the same time, organizes and operates inter-ministerial relations to develop reference points and the necessary standards to integrate data across domains. In both the processes of reporting within each domain and corroboration by the NBS, the quality of data is examined and questioned with respect to accuracy.

Moreover, line ministries are required to submit their surveys to the State Council through the NBS for examination and approval. Surveys conducted by departments of lower level governments, for example the health bureau of a city, are administered by the department in question and are reported to the statistics agency at the same level of government (the statistics bureau of the city), as well as the health bureau of the senior government. Local surveys are administered by the statistics agencies of local governments at or above the level of county, or in conjunction with the relevant department or departments of the same level of government, and are reported for approval and record keeping by the statistics agency at the next highest level of government. In short, all surveys must adopt a framework appropriate to the rules and regulations and be submitted to a higher level of government for approval and corroboration. The framework stipulates survey objectives, contents, methods and respondents, organizing forms and questionnaires as well as the final publication of data. Any changes to the framework must receive the approval of the higher level of government that issued the original permission.Footnote ¹⁰

The structure of the NBS, then, consists of agencies in a hierarchy that operate under the leadership of both local governments, which are responsible for personnel, and senior levels of the statistics agencies responsible for technical work and funding. These agencies are responsible for carrying out the tasks of reporting to the NBS through unified reporting tables designed by the NBS based on data submitted from townships and towns. Governments of townships and towns below the county level, according to the “Detailed rules for the implementation of the Statistics Law” (1987, last revised 2006), are staffed with full-time or part-time statisticians responsible for the specific coordination and administration of the statistics work in their respective areas, including managing the basic statistics data and improving systems of accounting and the statistics files of townships and towns in accordance with relevant state stipulations.Footnote ¹¹ At the bottom of the hierarchy are village accountants in the countryside or accountants in urban neighbourhood committees. They are responsible for keeping accounts and records about the households in their area for their local use as well as for township statisticians to complete reporting tables. Within this structure, local records serve the purposes of statistics reporting, including administrative reporting and surveys, required by senior levels of government.

According to the regulatory system governing China's statistics – comprising the Statistics Law, administrative regulations (including regulations for censuses and for customs statistics), and rules on surveysFootnote ¹² – the NBS is responsible for drawing up plans for important national surveys, including the Population Census, Economic Census, Agricultural Census, and over 27 other statistics survey schemes. For the important general surveys as well as for the sample surveys, the NBS utilizes its survey offices in each province (municipality and autonomous region), and survey teams in each prefecture (city, autonomous prefecture and league) and in one-third of the counties (urban districts and banners in rural autonomous regions). These survey offices and teams carry out census and sample surveys on key statistical data required for macroeconomic control and national accounting. Such sample survey schemes cover areas of business (or the entities that have registered as “legal persons”), agricultural output and price statistics, industrial statistics, labour wages, basic social and economic conditions at and above the county and city levels, social development statistics, and statistics about women and children, such as the National Survey on the Status of Chinese Women.Footnote ¹³ Sample surveys on rural households and statistics on agriculture, forestry, husbandry and fisheries, for instance, are undertaken yearly by the survey offices. This collection of data is sometimes based on individual households undertaking interviews and keeping journals of information, including household income from all sources and household expenditures. At other times, it is based on information provided by village records, such as statistics on basic conditions in rural communities. These records serve as the basis for completing survey forms, which are ultimately submitted to the NBS via the provincial survey offices. Monthly labour force participation rates are based on sampled household members over the age of 16 who are registered on urban neighbourhood records. Wage statistics are based on all “legal persons,” that is, entities who submit completed unified reporting forms to statistics offices in rural townships and towns, which are then submitted to senior governments from county to the provincial levels in the 31 provinces, autonomous regions and municipalities of China. These provincial-level statistics bureaus then submit their final compilations to the NBS.Footnote ¹⁴

The Chinese centralized statistics system is highly controlled and hierarchical, and based on approval of reporting through administrative tables and survey forms. Any changes to the administrative tables or a questionnaire – for instance, to incorporate sex-disaggregated indicators – would require approval from a more senior level of management in the system. It is, therefore, extremely important for the NBS to coordinate with line ministries and lower levels of government to create sex-disaggregated indicators, for example by encouraging local statistics agencies to take initiatives to gather non-conventional data. The findings of the Tianjin Project are examined next in order to understand how some current reporting tables are used and to assess existing gaps in gender statistics in the Chinese system.

Findings of the Tianjin Project

In 2008, a group of scholars in Tianjin initiated a research project to discover the importance of gender statistics to local policymaking in support of the Chinese commitment to UN Millennium Development Goals (MDGs) and popular demands for the government to expand its service functions.Footnote ¹⁵ The project received financial support from the United Nations Gender Facility Fund in China, and carried out field investigations and analysis of the use of reporting tables in three local districts: a rural county making its way out poverty; a street in a city centre district; and a rural township becoming an urban street in a rapidly urbanizing district near the booming port of Tanggu 塘沽, an economic zone in Tianjin.Footnote ¹⁶ While preparing research reports on the three districts in 2009 and writing their final report in 2010, the researchers held consultations with participating agencies in local government and community management committees. Those invited to participate included local experts on statistics and members of the local communities.Footnote ¹⁷ Central to their findings was the identification of gaps in the use of gender statistics in reporting tables, especially those used for employment, family planning, education, community management and political participation. Action and inaction in these areas suggest how well government serves society. Improving local governments' use of gender statistics in these areas, therefore, would help to transform government from an authority to a service in society, particularly in areas of production and reproduction. Advancements in gender statistics at the local level would especially improve services as a result of an increased understanding of the different experiences of local women and men. The reporting tables that follow – employment in industrial production, family planning, education, community management and political participation – are reproductions of the original tables reviewed by the report's authors.Footnote ¹⁸

1) A table on industrial production and employment shows a lack of gender statistics

The NBS designed this table, which the Tianjin municipal government distributes to its townships. Like all townships, F Township collects the information called for in the table and submits it to the county government where it is entered into the statistics system network to be amalgamated at the municipal level every month. The amalgamated information is then submitted to the NBS. The table collects information on the number of employees, as well as output, income, profit, taxes, and so on, of businesses in the township. The form as written makes no requirement to disaggregate data on the number of employees in each enterprise by sex. There is, therefore, no way of knowing differences between women and men in each category. For example, large-scale industries tend to be state-owned and offer higher wages and better benefits; however, there is no way of knowing if primarily men or women are employed in these companies. There is, moreover, no way of knowing the contributions made by men versus women to taxes and output, or the scale of changes in these indicators from year to year. Without the appropriate information about women's contributions and any segregation they may experience in employment, governments might feel less pressure to provide women with the necessary training to improve their employment and remuneration. The researchers suggest replacing the existing table with Table 1, which adds sex-disaggregated information about employees.Footnote ¹⁹

Table 1: Industrial Production in F Township in 2006 (10,000 yuan/per person)

Notes:

* The designated size is defined by the government in 2011 to be any enterprises with revenue of 20 million yuan per annum. Before 2010, it was demarcated at 5 million yuan per annum. See http://www.stats.gov.cn/tjzs/cjwtjd/201311/t20131105_455942.html.

Revising the table in this way is not a huge task, but requires gender awareness. If collected, information that measures male and female employment in each enterprise could provide evidence of the status of women and hence their need for technical training to improve the quality of their employment. It could also make women's contributions visible. Finally, over time, the collection of this information would suggest trends in men and women's employment and thus assist with recruitment and promotion.

The conventional employment rate currently used in China does not disaggregate by sex and provides no indication of the quality of employment and types of professions women and men undertake. On the rare occasion that the rate is disaggregated by sex, however, the data show that women and men are “segregated” in terms of positions and professions, and that women experience discrimination based on their sex. According to the Third National Survey on the Status of Chinese Women,Footnote ²⁰ the majority of women work in private and individually-owned sectors, while men work in state-owned sectors where salaries and benefits are higher and protected.Footnote ²¹ Expanding the collection and use of gender statistics in employment is a first step in meeting the requirements of the Law on the Protection of Women's Rights and Interests of the People's Republic of China that guarantees women the right to work and to benefits including welfare, social protection and training.Footnote ²² Even a sex-disaggregated employment rate, however, will only just begin to reveal the difficulties women experience in employment choices, type of work and profession, number of employed years, and employment quality. Other indicators, including separate employment rates for women and men in different fields, sex ratios in certain types of professions, sex ratios of the employed in different age groups, sex ratios of the employed in different income groups, and sex ratios of the employed in access to social protection including maternal/paternal benefits, would further reveal these differences and challenges.

2) A reporting table on family planning reveals an absence of gender-sensitive indicators

The National Commission on Population and Family Planning, which merged with the Ministry of Health in 2013 to become the National Commission on Health and Family Planning, designed this table to collect data across China's countryside on indicators including the number of women of childbearing age, the number of births, and contraception use. None of the indicators are, however, disaggregated by sex. As a result, the statistics fail to show the sex ratio of births either inside or outside the one-child quota, whether the card issued to couples with one child (to acknowledge their entitlement to benefits under the one-child policy) has been given to a couple with a girl or a boy, and whether the husband or the wife was sterilized. Although the Chinese government replaced the one-child quota with the two-children quota in 2015, the sex ratio of births (inside or outside the new quota) remains an important indicator to inform efforts not only to improve men's participation in family planning but also to address the imbalanced sex ratio at birth.Footnote ²³

Elsewhere, I have shown that the Chinese government's approach to fertility control has historically focused on women, and that women have, as a result, shouldered up to 85 per cent of family planning responsibility since the inception of the one-child policy in the 1980s.Footnote ²⁴ Although Article 17 of the 2002 Population and Family Planning Law of China states that husband and wife together share family planning responsibilities,Footnote ²⁵ national statistics show that, in 2011, men's overall contraception use still totalled only 14.7 per cent, an estimate based on local reporting of contraception use among married women.Footnote ²⁶ Without a sex-disaggregated indicator of contraception use, however, this table fails to promote men's participation, as the law requires. Moreover, the preference for sons in Chinese culture highlights the importance of the first-born child's sex on parents' decision whether or not to have another child. Son preference has also contributed to an imbalance in sex ratios at birth owing to sex-selective abortions. Given these considerations, this table not only skews responsibility for contraception use towards women but also overlooks the widespread sex-selection practices that result in an imbalanced sex ratio at birth.

Beginning in the 1980s, the sex ratio at birth in China started to widen, and since 1985, has persistently remained over 110, higher than the normal sex ratio of between 102 and 107 boys born for every 100 girls born.Footnote ²⁷ According to the sixth census in 2010, the sex ratio at birth remained highly abnormal at 118.1.Footnote ²⁸

This increasingly imbalanced sex ratio at birth results from a number of factors related to sex-selective abortions. First, local governments conceal, omit and misstate the number of newborns and the real birth rate, thus contributing to an imbalanced sex ratio in Chinese statistics.Footnote ²⁹ Second, the family planning policy has been unevenly implemented at the provincial level because of different emphases, levels of resources, and efforts put into statistical reporting, all of which significantly affect the collection of true and effective data. Nevertheless, the increased sex ratio has affected marriages and families, reproduction, employment, social protection, and economic and social development as a whole. The negative effects on marriage and family relationships are substantial, counteracting the building of a harmonious, well-off society that the Chinese government has set as a national goal. It follows that both contraception use and the sex ratio among newborns are important indicators for measuring the work of family planning and the effectiveness of services, including family planning and health services provided by local governments. It is, therefore, desirable to revise the existing table by adding sex-disaggregation as shown in Table 2.Footnote ³⁰

Table 2: Family Planning Situations in Rural F Township in 2007 (Revised)

The revised table shows the following indicators disaggregated by sex: number of births inside and outside the quota, the sex ratio at birth in single-child families, and most importantly, the contraception responsibilities borne by women and men in the rate of hysterectomies versus vasectomies. If adopted, these sex-disaggregated indicators would begin to provide much-needed information to improve family planning policy and implementation, especially in the countryside where the sex of newborns inside the quota often influences whether pregnancies occur outside the quota. This information would help to implement more effective family planning policies as the number and the ratio of newborn boys inside the quota would probably influence the number of cards issued to couples with one child. An accurate understanding of these indicators would help to predict population growth and its connection to sustainable development and investment in education and social services.

3) An education reporting table misses girls

The S Street unit uses this table to collect information about the number of students in each grade. Without disaggregation by sex, however, it is impossible to know the sex ratio of students in each grade, or that of those attending schools inside versus outside the neighbourhood. The latter may not be a strong issue in urban areas because girls do not have a higher dropout rate than boys; however, this is an issue in the countryside, and particularly in the schooling of children of urban migrant workers who are denied urban hukou 户口 status and are often prevented from attending schools in the urban neighbourhoods where their parents work and live. Disaggregation by sex of those who attend schools inside and outside the neighbourhood would likely show whether boys and girls are treated differently in such scenarios, and whether boys are kept with parents while girls are left at home in the countryside – one possible factor explaining the higher number of boys than girls in urban schools.

Education is an important means of increasing human capital and advancing skills, knowledge, values and attitudes towards equality and justice. The UN's fourth Millennium Development Goal requires the elimination of gender differences in primary and secondary education. Sex-disaggregated data on students, especially those who attend schools inside and outside the district, would help to monitor the achievement of this goal by revealing differences in gendered needs among students in different schools, and the sex ratios of those advancing to higher grades and completing primary and secondary schooling. A re-designed table (see Table 3) might work by adding a single line to provide sex-disaggregated information on students inside and outside the neighbourhood. This re-designed table would help local governments to track the numbers of boys and girls in different grades, and thus help them to meet the different needs for school facilities – for instance, sports, toilets, and clinics – in order to improve the quality of learning and promote equality in education.

Table 3: Primary and Middle School Students in the Neighbourhoods of S Street (Revised)

Source:

Chen et al Reference Chen, Chen and Wang2011, 68.

4) A community management table on dispute resolution is only concerned with incidents and records but not about the people involved.Footnote ³¹

The S Street unit designed this table as a trial to gather information about different types of disputes. As there is no indicator to suggest the kinds of people involved in the disputes, however, it is impossible to identify the vulnerable parties and, therefore, to provide services targeting the vulnerable. Without sex-disaggregated information on the individuals involved in disputes, it is difficult to protect the rights of women who are often in the weaker position in marital and family conflicts, family planning disagreements and labour wrangles.

The Law for the Protection of the Rights and Interests of Women in China outlaws any gender-based violence at home or in society as a whole.Footnote ³² Better information collection and monitoring would help to implement this law and promote women's rights and equality, especially to a safe and healthy life, and would improve government services. It is, therefore, important to revise the table so that it identifies vulnerable groups and individuals in society (see Table 4, to assist discussion later). To enhance the table further, community workers and statisticians for S Street should come together to discuss the needs, design and use of gender-sensitive indicators. The table that would result from such collaboration would help to improve community services by providing a more accurate reflection of overall trends and developments in gender inequalities in society. Moreover, an improved table would save the local community the time and effort involved in completing tables meant more for reporting to higher levels of government than for local policymaking.

Table 4: Bi-weekly Inspection among Neighbourhood Committees in S Street (Revised)

5) A table showing political participation and local governance fails to show female involvement

F Township uses a table designed by the county to collect information on village governance with respect to village cadres, members of the village Party committees, members of the village council, Party members and village heads.Footnote ³³ These data are reported to the provincial government and, ultimately, the central government. Since this table seeks information without disaggregation by sex, it cannot provide an accurate basis for government decision making in terms of understanding women's contributions to, and participation in, village governance. Political participation is the most direct means for any class or individual to advance claims for gender equality and to exercise the political rights guaranteed by the Law on the Protection of the Rights and Interests of Women in China.Footnote ³⁴ It is, therefore, an important measure of democratization and social progress in society.

Nevertheless, women's political participation in China remains highly unsatisfactory. Just 637 female delegates participated in the National People's Congress in 2008, while only 29 members of the standing committee were women, respectively just 21.3 per cent and 16.6 per cent of the totals. At the People's Political Consultative Conference in 2008, there were 395 women present and 30 members of the standing committee were female, respectively just 17.7 per cent and 10.1 per cent of the totals. In government, women are mostly concentrated in the lower levels of the hierarchy and represent an extreme minority at senior levels. Only 23.3 per cent of members of the Communist Party overall are women.Footnote ³⁵

Because women's participation in all levels of government, as indicated by sex ratio, is of central importance for the quality of treatment women receive in other fields of life, including employment, family planning, education and dispute resolution, it is necessary to revise the table used by F Township to provide the statistics needed to reflect sex ratios in village government (see Table 5).Footnote ³⁶

Table 5: F Township's Rural Economy Statistic Collection, Village Administration (Revised)

A table revised in this way would provide the necessary information to help the township and its senior administrators make decisions to meet the needs of women leaders, including training and childcare. Moreover, over time, this table would provide an understanding of the trends and changes in women's participation in local governance. If all townships were to adopt such a table and incorporate sex-disaggregated information into their statistical gathering, the projection of regional and national trends and changes in women's participation in governance would become possible. This information in turn would help to monitor progress on China's Law on the Protection of the Rights and Interests of Women, and on the implementation of the national action plan, the Programme for the Development of Chinese Women, 2011–2020.Footnote ³⁷

Analysis of the above tables pursues two tasks: to disaggregate existing indicators by sex in order to identify gender-based differences, and to enter new indicators reflective of inequalities in order to make the unknown known. The study in Tianjin clearly reveals a need to incorporate gender-sensitive indicators into statistical reporting, especially through the design of surveys and the collection and evaluation of data. Doing so is necessary to identify gender-based inequalities and their implications for local policymaking, and to overcome inequalities and achieve the goal of promoting production and reproduction in one integral process of creating a just and equitable environment for women and men. In the next section, I show why current “state feminist” approaches to gender statistics fall short of achieving these goals.

Gender Statistics: Feminist Political Economy versus “State Feminism”

Gender statistics, according to a recent UN discussion, are more than simply a disaggregation of data by sex;Footnote ³⁸ rather, they form part of a policy-oriented approach to governance. As demonstrated in the analysis of the above tables, data must be disaggregated by sex in order to reveal gender issues that might otherwise remain hidden. Sex disaggregation provides evidence of gender differences and strengthens and improves the entire statistical system simply because it allows for a close reading of the real lives of women and men. When all relevant indicators and variables in state-organized data collection are gender-aware they reflect the different needs of women and the root causes of differences with men. They also reflect change over time and, therefore, help to monitor and evaluate progress once a baseline has been established. As a result, a focus on sex disaggregation effects changes in the quality of governance because policymakers, better informed of gender-based inequalities, can work towards policy solutions to overcome these inequalities. Yet, if the data are disseminated even further and made available to local governments, men and women, improvements in the effectiveness of government service will be even more profound.

While the collection and analysis of gender statistics in China have progressed under Chinese “state feminism,” they still lag far behind current demands for information, largely because “state feminism” focuses on high-level government and ignores evidence-based policymaking and improved governance at the local level. “State feminism” relies on women's official organizations to promote the use of gender statistics in support of government action plans to enhance women's development and gender equality. This women's machinery is represented in China by the Working Committee on Women and Children under the State Council and its powerful ally, the All China Women's Federation (ACWF), which is not a part of the government but still falls under the rubric of the Communist Party and, therefore, forms part of the ruling establishment.Footnote ³⁹ These women's agencies are responsible for designing, implementing, evaluating and reporting on the progress of the national action plan, the Programme for the Development of Chinese Women. China is, at the moment, implementing its third Programme for the Development of Chinese Women, 2011–2020.Footnote ⁴⁰

An examination of all three Programmes for the Development of Chinese Women, beginning in 1995, shows that the ACWF has campaigned for mainstreaming equality between women and men in policymaking and has demanded the use of gender-sensitive indicators and the creation of a women's database in statistics gathering. Article 17 of the first Programme, which covered the years 1995–2000, demanded a strengthening of national statistics on women's development through the establishment of a women's database and increased use of gender-sensitive indicators in information collected about women. The second Programme (2001–2010) emphasized the promotion of overall statistics on women's development, the use of gender indicators and the development of a sex-disaggregated database, instead of a women's database. This programme also called for the incorporation of gender indicators into the general statistics system and statistical investigations. The most recent programme reiterates the earlier demand for a database on women at the prefecture, provincial and national levels, along with the need to incorporate more gender indicators into the general statistics system and investigations.Footnote ⁴¹

For these women's agencies, gender statistics offer the means to analyse and study changes in social and economic development from a gendered perspective, especially focusing on differences embedded in various issues and conditions relating to the status, roles, and distribution of power and resources of and between men and women. Analysis of difference helps governments to devise ways to overcome inequalities and to promote the equitable development of everyone.Footnote ⁴² Clearly, the linkage between gender statistics and broad policymaking is established in this approach to information gathering, but the equality of governance still remains largely unchanged. The lack of linkages between gender statistics and equality of governance in Chinese official discussions, however, reflects debates among Chinese researchers, particularly those who serve the women's agencies, between those more oriented towards the evaluation of national action plans and those who desire to influence policymaking overall.

These debates have shaped current approaches to gender statistics in China. On the one hand, collaboration between the NBS and the ACWF since 1995 has resulted in three national surveys on the status of Chinese women, in 1990, 2000 and 2010, which have generated data in all fields of concern to Chinese women and provided evidence for the design of the national action plans.Footnote ⁴³ On the other hand, the NBS booklets, Women and Men in China: Facts and Figures, published every four years since 1995, are supported by international organizations and follow international models for setting indicators that are broadly related to the Beijing goals (1995), the MDGs, and the concerns of funding agencies including UNFPA and UNICEF. The data are compiled from the administrative statistics collected by the NBS annually, the census taken every ten years, the one per cent population surveys taken every five years, and labour statistics. The 2012 edition includes a new section called “Gender outlook,” which draws on data from the 2000 and 2010 national surveys to reflect changes in gender awareness, gender identity and attitudes towards gender roles.

The recent NBS “Guidelines on the plans for statistical development and reform in the 12th Five-Year Plan” demands a strengthening of statistics on population, social environment and resources. It also calls for raising the level of social statistics collection and dissemination, and improving statistical monitoring of the implementation of the national action plans for women and children.Footnote ⁴⁴ These demands have resulted in the implementation of a number of sex-disaggregated indicators in the national statistics reporting system (214 in 2014).Footnote ⁴⁵ Recently added sex-disaggregated indicators in the areas of health, employment and political participation include death rates by breast cancer and cervical cancer, suicide rates by sex, the current employment rate by sex, non-agricultural employment rate by sex, employment in the public services by sex, women's share in leadership positions at different levels of government, and women's share of chief positions in different departments of senior government.

Gender statistics, nevertheless, remain underdeveloped. By one account, there is only one junior staff member in the division of monitoring and evaluation within the department of social, science and cultural statistics at the NBS who is responsible for gender statistics. Her responsibilities include the statistical monitoring of the national action plans for women and children, the collection of data for, and publication of, the Women and Men in China booklets, and coordinating surveys on the status of Chinese women. She has only a very limited ability to coordinate efforts to expand gender statistics gathering with other departments, including labour and economic statistics, and other ministries, and has limited access to the policymaking level of the NBS.Footnote ⁴⁶ This understaffing demonstrates just one way that the demand to mainstream gender statistics in the Chinese statistical system is not yet fully reflected in all legislation and policy.

There is, therefore, a shortage of sex-disaggregated data reflecting gender gaps in China in employment, in participation at all levels of decision making, access to services, protection of rights, population, marriage and family, income and wealth, education, health care, community participation and management, crime and assault, time-use in household-based work and care, as well as social security and recreational activities. In particular, there is an absence of data regarding vulnerable groups, such as poorer rural women, female migrants, adolescent girls and youth, elderly women, people living with disabilities, and ethnic minority women, that might help to identify existing inequities and expose discrimination and violence.

This lack of gender statistics in the Chinese statistics system suggests a continuing failure of the NBS to gather enough credible, useful and accurate data to fulfil its needs for informed policymaking. It may also be read as a failure of the “state feminist” approach of the official women's organizations, which focuses mostly on the macro-statistical system and overlooks the abilities of grassroots women and community activists to collect data and promote equality of governance, especially at the local level.

The downside of the “state feminist” approach to gender statistics is its failure to mobilize grassroots activists as producers and users of statistics, resulting in a need to produce useful gender statistics outside of, or on the fringe of, the official statistics system. In either the legislation governing the operation of the statistics system or in the design and conduct of general censuses and surveys, gender statistics remains an outside/alien concept.

A feminist political economy approach, however, defines policymaking as a process of contestation whereby women and community activists can exercise citizenship rights by participating in, and influencing, the making of decisions at all levels so as to effect changes in the equality of governance. It argues that while mainstreaming gender statistics is necessary throughout senior levels of government to inform evidence-based policymaking, it is equally important to mobilize community and individual social activists to keep gender-sensitive records for villages and urban neighbourhoods in order to hold government accountable, to effect changes locally, and to achieve equality and well-being for all. The Tianjin Project has helped to illustrate this last point with two examples. In the first, village communities involved in the Tianjin Project keep records of local households with sex-disaggregated information about the members of the households and their social and economic conditions. This information is, however, overlooked at higher levels largely because government reporting tables do not often ask for gender-sensitive data. In the second example, the community management table mentioned above designed by the S Street unit, and upon which the revised Table 4 is based, shows that the levels of Chinese government closest to urban communities do design and adopt statistics tables on their own to help record and analyse events that affect the lives of local people. It also demonstrates that it is necessary to adopt more tables like this one in the effort to raise gender awareness among local government officials. Given the nature of local administrative systems of statistics collection, it is possible for grassroots data collectors in urban streets and rural townships, and for the village and urban neighbourhood accountants who keep households records, to initiate changes and even to begin new instruments such as time-use studies or gender-based violence surveys. Such new instruments may require the approval of senior management but may nevertheless be implemented by local statistics officers at the township or county level.

A feminist political economy perspective is thus more comprehensive and effective than a “state feminism” perspective because it promotes the use of gender statistics to reveal women's subordination as well as needs through grassroots mobilization and not just through statisticians and policymakers at the national level. The recognition of local women, community activists and policymakers as both producers and users of gender statistics overcomes the shortcomings of state-based feminism. The advantage of empowering women and social activists as producers and users of gender statistics is found in equipping them to express their demands for, and interests in, advancing more evidence-based policymaking and to monitor the implementation of policies to overcome inequalities. The result is balanced, scientific and coordinated management and development through policies that require the government to be more transparent and to consider the interconnectedness of political, economic and social life nationally and locally, as well as the balance between productive and reproductive activities. Perhaps as a first step towards scientific management through policymaking, the government of China could synchronize community statistics with the official statistics system to allow for the gathering of gender statistics that are meaningful for the expansion of government services.