Identification of major information gaps.
As a general comment, the degree of sophistication in data collection, compilation and publication in PNG can be described as modest. Reliable national and provincial data are not readily available. Most government
departments concerned with social conditions do not collect data on a regular basis. Whilst health and education are exceptions to this, the data that they produce are of limited scope and reliability. National censuses and surveys suffer from difficult field operations, both logistically and with respect to the response error.
Perhaps the most important national source of recent data is the 1990 Census of Population and Housing. However, due to real and perceived difficulties in field operations, PNG censuses do not provide as comprehensive data as do other Pacific Island censuses. Employment and demographic measures of fertility and mortality are areas on which limited data are collected.
4.2 Economic Activity
Data on economic activity from the 1990 census includes only the question on economic activity itself. This question is less than adequate in capturing women's economic activity and hence the data is of limited use for gender analyses. This is in large partly due to the use of the category 'housework'.
Whilst this category may have some relevance for the urban population, its use in rural areas is of questionable validity. Gender bias in field operations is clearly evident in the results. Amongst rural females aged 10 and over 187,000 were reported as engaged in housework, compared to 18,000 males. This represents 18 percent of the female rural population aged 10 plus, stretching the imagination in a subsistence economy.
By way of 'compensation', rather fewer females than males are reported as being engaged in purely subsistence activities (277,000 females and 314,000 males) and as 'unemployed' (22,000 females and 45,000 males). This latter category is of doubtful validity, especially in rural areas. Even in urban areas a pilot question about whether the respondent was looking for, or available for, work fn the last seven days was found to be 'not meaningful' (National Statistics Office: 1994, p. 14).
For both sexes census data on economic activity are divided into two main groups, 'monetary' and 'non-monetary'. This represents a departure from the more usual grouping, 'economically active' and 'not economically active'. Subsistence activity is aggregated with the 'not economically active' that includes students, those doing housework and the sick, elderly and disabled. This implicitly places value on monetary activity while devaluing subsistence activity in which women are particularly are engaged. The 'unemployed' is also included in this non-monetary grouping.
There is need for further research to be done in aggregating women's subsistence and other unpaid work. Current data in this area is misleading.
Other census information on economic activity is not available at the national level. Occupation is available for the urban population only with the assumption that all rural residents are engaged in agriculture. Whilst this may be largely true the assumption precludes knowledge of the development of small-scale rural economic activities such as food processing, in which women are increasingly becoming engaged.
For the data on occupation that are gathered for the urban population, available tabulations are not adequately disaggregated by sex. Detailed occupation by sex is tabulated by citizen/non-citizen status only, omitting tabulations such as occupation by sex and by educational attainment which are important for any gender analysis of the labor market.
Information on industry is not available in census data. A1989 pilot question on industry in urban areas was omitted from the census because it was 'found not meaningful' (National Statistics Office: 1994, p. 14).
Other sources of data on employment in urban areas have not been available in recent years. For example, the most recent data on urban employment refers to 1989.
Administrative records of employees, such as kept by the Public Service, do not usually provide useful data on employment because the sex of employee is often not recorded. In rural areas, within the overall classification of subsistence activity, data are lacking on the nature of that activity.
Women's involvement in agriculture, fishing, forest resource harvesting and management of the environment are all areas where detailed information is lacking. Mining, employment in mining and its impact on communities, families and women and men is another area where information is deficient. Clearly, small-scale studies are appropriate in these areas.
Additional work is needed in gender disaggregating occupational statistics. Substantial questions on industry need to be included in the census form.
Although extensive studies have been done in industry (particularly forestry), there is scant literature in this area that addresses gender issues. There is a need for a gender study in forestry to be conducted using case studies from a variety of regions in PNG.
Women's involvement in credit schemes has not been documented extensively although this is a burgeoning area for development assistance. It is essential that training is provided to enable accurate records to be taken on women and credit facilities and repayment rates.
* It is recommended that donors and other funders of credit schemes work together on ensuring consistent data is collected on credit schemes throughout PNG. The Investment Promotion Authority has a registry of Women in Business and it may be possible for organizations running credit programs to feed information into this central national database.
4.3 Demographic Indices
Knowledge of current demographic trends in PNG is either lacking or unreliable. The 1990 census did not include questions designed to obtain information on which to base demographic estimates of fertility and mortality. Where estimates of life expectancy have been made available they are valid only for females since the data on which these estimates were made was collected only for females. Sex differentials in mortality are thus unknown.
The widely used estimate of male life expectancy in 1980 was assumed to be approximately two years less then female life expectancy, based on demographic models of sex differences (of doubtfui applicability to PNG). The 1990 census does not provide estimates of life expectancy for either sex, though data to estimate female adult mortality were collected while child mortality was not. Estimates of child mortality were obtained in the 1991 Demographic Sample Survey (commonly called DHS) so that life expectancy estimates have been made available. This 1991 survey is of questionable reliability, however, and covered only a selection of provinces due to financial constraints. A 1996 Demographic and Health Survey has been conducted which should fill many of the gaps in demographic and reproductive health knowledge.
Further gender-disaggregated data needs to be consistently collected throughout different PNG regions. The standardizing of questions between departments will enable regional studies to contribute towards national data.
4.4 Social Conditions
Data on health from service delivery records are significantly under-reported. Service delivery records at health posts are often not compiled at higher levels. This appears to be due to inadequate communication between the different levels of service management. The annual Handbook of Papua New Guinea Health Statistics contains data provided by provincial offices. Significant information gaps hamper effective health planning. Health data are, however, the most useful source of information addressing the social impact of economic policies.
In education, discrepancies exist between enrollment data from schools and attendance data from census reports, casting doubt on the reliability of both.
There is a need to develop of a comprehensive national study of gender equity barriers in PNG educational institutions (schools and tertiary education) focusing on issues such as harassment, bullying and sexual and physical assault. The study should also include a review of campus security and safety mechanisms for women.
The first step towards the development of a coherent approach to the issue of gender equity in education is to collate information on current knowledge. Baseline data are needed for many of the issues and patterns that influence the lack of access, low enrollment, poor retention and weak performance of females in education in PNG.
Although extensive anecdotal evidence appears in different studies there has to date been no extensive or comprehensive national study on issues of sexual discrimination in PNG educational institutions. The study could identify major indicators and patterns and trends in the occurrence of barriers to gender equity:
* Discrimination (both covert and overt to include harassment, bullying, teasing, sexual assault);
* Violence in educational institutions (a gendered analysis of the effects on students and teachers);
* Cost of education;
* Lack of access to private counseling;
* High rate of teenage pregnancy and marriage;
* Parental/spousal influence;
* Lack of relevance of curriculum and teaching methods;
* Accommodation issues for women and girls in boarding situations (student and teachers);
* Regional and clan variance;
* Safety for women on campus: review the need for additional facilities on tertiary campuses;
* Critique of the effectiveness of current of codes of conduct, policies or guidelines being used in educational institutions for preventative behaviors and disciplinary actions.
This study could also investigate how Recommendation 29 and 30 from the Final Report on Domestic Violence (Law Reform Commission: 1992) have been implemented:
Recommendation 29: That anger management and conflict resolution skills be included in the curriculum at both primary and high school levels and that in-service training on these topics be provided for all teachers. Recommendation 30: That the Education Department and the Teaching Services Commission implement the recommendations from the Gibson Report (Education Research Division Research report No. 65). Also that clear policies be developed with respect to spouse beating, so that a teacher who physically assaults his or her spouse will be charged and disciplined. Repeat offenders will be dismissed from the teaching service and teachers who are victims of domestic violence will be supported and not penalized.
A key area of concern in contemporary PNG, especially for women, is the increasing level of violence. Information on its impact is lacking. Some studies have been conducted on domestic violence and violence towards women and attitudes to such violence, however, the general issue of violence is under-researched.
Until recently no data was available on poverty. This situation had been improved by the recent poverty assessment that involved a 1996 household survey of income, consumption and expenditure. This survey will provide valuable data on gender dimensions of poverty.
Male Gender Studies
There is general lack, as far as gender analyses are concerned, in data and studies addressing males. Male attitudes have been identified as a major impediment to the progress of women in many areas of life, including education, employment and credit. There is a lack of behavioral and attitudinal studies addressing males and on the impact of male attitudes and behavior on communities and development.
4.5 Availability of Data
One of the major problems in national data availability is decentralization. The extent and quality of data vary considerably among provinces, and national compilations are of patchy reliability. Administrative records are often created and/or stored in ways that does not easily allow tabulation or gender-disaggregation. In addition difficult field operations reduce the usefulness of national data collection exercises. Whilst censuses and national surveys are necessary to provide population counts and overview information, such exercises do not provide much by way of data that are useful for analyzing gender dimensions. National overview statistics do not identify how interventions can be made to assist women in tangible ways.
Given the difficulties involved and the limited usefulness of aggregate data, it would be both unrealistic and misguided to look to improved national or provincial data collection systems as the main source of relevant gender-disaggregated data. Local level data are, in any case, often the most appropriate for examining the gender and social effects of development processes and economic policies, particularly on subsistence communities and on the poorest sections of communities. However, the current availability of local level data is also poor. Efforts to improve the availability of information and data on women in PNG should concentrate on studies at the local level.
4.6 Updating and Building on Recent Studies
Statistical Country Profile
There has been an extremely timely updating of the seminal work by Heather Booth of the statistical profile of gender in PNG. The Gender Desk in the National Statistics Office produced a draft of this updated profile in 1997. Although the profile incorporates data from recent studies, much of the comprehensive statistical data is drawn from the 1990 census. This is because there has not been any comprehensive collection of data since this census.
Law Reform Commission
Strategies to address systemic violence and discrimination in PNG were identified through the Law Reform Commission in the Final Report on Domestic Violence in 1992. However, many of the recommendations were never implemented and the Commission no longer exists. The recommendations were comprehensive in their coverage of legal measures including the present legal provisions and their deficiencies, the police, criminal law, district courts and protection orders, and village courts. It seems unnecessary to recommend any other options for addressing issues of violence against women, other than to reiterate the importance of implementing the strategies already identified from this seminal report. Possible intervention strategies in this area:
* Support and strengthening of ICRAF paralegal training for women
* Implementation of the Law Reform Commission report 14 on Domestic Violence.
There are many areas where further study and research would be useful in developing gender-disaggregated data that would assist a comprehensive gender analysis within PNG.
It is important that studies that are conducted are made public and shared with participants and stakeholders. There is a great deal of anecdotal evidence to suggest that people in PNG, particularly women, are tired of 'being studied'. It is also essential that scarce resources are not wasted on unnecessary or inaccurate studies.
It is recommended that major donors in PNG hold an annual meeting with UPNG and the National Research Institute to discuss upcoming studies. This would enable collaboration between academics and donors and others and avoid possible duplication in research.
AusAID Baseline Study on Violence
In 1997 AusAID commissioned a study of the work NGOs in PNG are doing to address issues of violence, particularly against women and girls. Although this report has not yet been made public, the results of the survey and fieldwork illustrate that very few NGOs are working on violence issues. The key organizations active in the area are: the Individual Community and Rights Advocacy Forum (ICRAF), East New Britain Social Action Group (ENBSAG), Simbu Women's Resource Center (SWRC), and the East Sepik Council of Women (ESCOW).