1/
This document has been prepared by the Australian
Bureau of Statistics. It has been issued as
submitted.
Introduction
1.
Gender statistics has been described
as 'today's approach' to obtaining adequate information
on all individuals, taking into consideration their
differences in all spheres of society (Hedman et
al., 1996). As such, it is an important development
affecting the field of social statistics, which
are concerned with the experiences of individuals:
their personal
characteristics, how they relate to their various
social groupings, such as family or household, and
how they interact with the society in which they
live. Social statistics are a major source of information
for governments and others in developing social
policies and monitoring social progress.
2.
This paper outlines the ABS experience
with various gender statistics issues in the context
of its activities in the field of social statistics.
In particular the ABS experience with Time Use surveys,
the Women's Safety Survey and social reporting publications
is considered.
The
international context
3.
In the 1970s the demand was for
statistics on women to support the activists promoting
change to address perceived inequalities and disadvantages
in the status of women. International statements
resulting from International Women's Year and the
UN Decade for Women (1975-1985) stressed the importance
of data collection and analysis relating to the
situation of women. (The problem was seen as a lack
of data by sex; however, it has been argued that
the problem was rather that such data was collected
but not analysed adequately and made available to
users (Hedman et al., 1996)). By 1995, when the
4th UN World Conference on Women was held in Beijing,
the emphasis had changed. In the Platform for Action,
the main document from that conference, national,
regional and international statistical services
were asked to take action to 'ensure that statistics
related to individuals are collected, compiled,
analysed and presented by sex and age, and reflect
problems, issues and questions related to women
and men in society' (Strategic objective H.3). This
was associated with other moves to mainstream gender
issues in wider policy and program development (UN,
1996).
ABS
Framework for Social Statistics
4.
During the past two decades the
ABS has developed a substantial program of social
surveys, much of which is concerned with the social
well-being of the population. The OECD model
of areas of social concern has been followed, where
a range of fundamental human needs and aspirations
are seen as underlying an overall notion of social
well-being. These needs include:
a healthy life
a job providing income
and a sense of self worth
command over economic resources
individual development
through education
shelter and security through
housing
personal safety and protection
from crime
social contacts, opportunity
and participation
pursuit of culture/leisure
activities
satisfactory physical environment
5.
These needs form the basis for
a number of different areas of social concern, which
also correspond in broad terms to some of the major
portfolio areas of government.
6.
In its social statistics work
the ABS recognises the following areas of social
concern:
population
family
health
education and training
labour
income and expenditure
housing
crime and criminal justice
culture-leisure
7.
The ABS also produces statistics
relating to population groups which have special
needs or which may be disadvantaged or underprivileged.
These groups have included the aged, youth and children,
lone parents, people with disabilities, carers,
Aboriginal and Torres Strait Islander people, the
overseas born, and women. A number of surveys with
a particular population group focus have been undertaken,
and social reports drawing data together from a
range of sources to describe the social conditions
of particular groups have been produced. The gender
statistics approach is also important in this area
of reporting; it has been shown that these social
conditions may have very different outcomes
for women and men within the population group, which
need to be identified and incorporated in policy
and program development.
Australian
data sources for gender statistics
8.
Most of the major issues generally
associated with the need for gender statistics are
already covered through existing ABS data sources.
These include the household survey program, the
population census and the collections based on administrative
data.
Household survey
program
9.
In the past 15 years the ABS has
developed an extensive program of household surveys
which provide a vast array of social and labour
related data. The program provides detailed and
in depth information about particular areas of social
concern (eg health, income, housing) or population
group (eg aged, people with disabilities). The program
is also developed to allow connections across surveys
through use of standard classifications of sociodemographic
variables and data modules. This allows analysis
of population groups across areas of social concern.
The household survey program is currently undergoing
review (see paragraphs 36-40). Key elements of the
current ABS household surveys program are:
monthly labour force
survey providing regular measures of employed
and unemployed people and their characteristics;
it also includes standard geographic and demographic
items including sex, marital status, age, relationship
in household, family type, birthplace and period
of arrival for those born outside Australia,
for the whole population.
monthly supplementary
surveys run in conjunction with the labour
force survey with topics varying from month
to month. A number are labour market related,
eg employment benefits, job search experience,
and are collected on a regular basis annually
or biennially. Other supplementary surveys deal
with a broader range of social topics, eg voluntary
work, child care, health insurance.
special social surveys
conducted each year which are large scale detailed
surveys with the capacity to collect complex
data on a specific group in the population or
issue of social concern. Surveys covering the
areas of health, household expenditure, disability
and ageing, crime and safety and time use have
been repeated at longer intervals of five to
seven years. In addition there have been regular
population survey collections of data on income
and housing costs. Other surveys which have
been run from time to time include housing,
literacy, families and education and training.
Population census
10.
The five yearly population census
is also a major source of social data with its prime
benefit being the availability of detailed data
for small geographic regions. It also provides detailed
information about small groups in the population,
eg those defined by religion, language etc.
Administrative
statistics
11.
Administrative statistics provide
an important range of social statistics including:
registration of births,
deaths (including cause of death), marriages,
divorces
overseas movement to and
from Australia
staff and students at primary
and secondary schools
reported crime
persons in prison
12.
In addition, a range of official
administrative by-product collections from government
departments (eg recipients of pensions and benefits,
students at various types of tertiary educational
institutions) which generally include some sociodemographic
data, at least age and sex, are available.
Examples
of issues relevant to gender statistics already
addressed by the ABS
13.
In his report on the 1911 Census,
the Commonwealth Statistician of the time, G.H.
Knibbs, recognised the value of work in the home.
He wrote: 'The terms . . "breadwinners" and "dependent"
require to be used with care as certain of those
classed as "dependents" contribute much more largely
to the upkeep of the community than certain others
who are classed as "breadwinners". For example,
all married women engaged in domestic duties for
which direct payment is not made, are classed as
"dependents", while all pensioners are classed as
"breadwinners". There can be no question that as
regards the current life of the community the former
group contributes the more largely.' Unfortunately
no further action was taken in relation to this
issue at that time. (Incidentally, it is clear that the term 'breadwinner'
was used to mean being in paid employment, with
no connotation as to household hierarchy.)
14.
In recent years a number
of gender statistics issues have been addressed
in the development of the ABS survey program, through
the introduction of new survey topics. These
were identified through the regular program of consultations
with major users and reviews of the forward work
program.
In 1996 the ABS conducted
the Women's Safety Survey, with funding support
from the Federal Government's Office of the
Status of Women and the (then) Commonwealth
Department of Health and Family Services. The
survey provides information on women's safety
at home and in the community and, in particular,
on the nature and extent of violence against
women in Australia. The ABS was responsible
for the development and conduct of the survey
and the usual extensive testing was carried
out to ensure that the survey would collect
objective and high quality data. The survey
was conducted under the authority of the Census
and Statistics Act 1905, as are all ABS
surveys. However, women were advised they were
not obliged to participate. Nationally, a response
rate of 78% was achieved. Some special
procedures were introduced to ensure safety
of participants and reliability of data provided.
All interviews were conducted in private; only
female interviewers from the ABS panel were
used, and they were provided with special training
to enable them to deal with issues associated
with violence against women (ABS, 1996).
Use and cost of child-care
and participation in voluntary work through
an organisation or group have been included
as supplementary surveys.
Several topics identified
in the National Women's Health Policy, such
as contraceptive use and participation in cancer
screening programs, were included in a special
supplement to the most recent National Health
Survey.
Special surveys of two
key population groups have been conducted on
a regular basis - women and men with disabilities,
and older women and men. The role of carers
in relation to these groups was also investigated
(ABS, 1999).
Data from several regular
supplementary surveys have been of increasing
concern for gender statistics; eg earnings and
earnings differentials, and associated sociodemographic
factors; participation in education; and educational
qualifications.
In 1978 the labour force
survey changed to collection on a monthly basis.
In the first publication from the new system,
23 of the 25 tables were disaggregated by sex,
and a number by age as well. The monthly labour
force publication continues to provide a core
set of tables relating to the labour force status
of the population, with 32 out of 33 tables
in current output disaggregated by sex. In addition
detailed analyses of the data disaggregated
by sex are available electronically. There
is therefore a large bank of accessible information
relating to Australian women and men, with time
series data for most important items.
15.
The ABS has introduced several
major revisions of standard data classifications.
The ABS uses ILO standard
definitions for classifying the economically
active population. A number of issues relating
to women's different pattern of labour force
participation have been addressed in recent
years at the international level, with similar
action at national level: underemployment (of
part-time workers who would prefer to work more
hours); marginal attachment to the labour force
(where strict job-search criteria are not met
but the person is wanting work and is available
to start work within a reasonable time); and
contributing family workers, where labour force
activity is now assessed on the same time criterion
as for paid employment.
The second edition of the
Australian Standard Classification of Occupations
(ASCO) was introduced in 1996 for use in all
statistical collections including the Census.
The revised classification now better reflects
the recent structural changes in the Australian
economy, in particular the increasing importance
of service industries as compared to manufacturing.
As a result the skill levels associated with
many occupations stereotyped as 'female' and
with the new information technology-based
occupations are now recognised in the classification
(ABS, 1997b).
The identification of 'household
head' with the male member of a married couple
is no longer accepted. It is necessary, however,
to identify a reference person for determining
relationships between household and family members.
During the interview a respondent is now asked
to identify someone in the household as the
appropriate reference person. If the person
nominated during interview is inappropriate
for applying the standard classification of
relationships used in all ABS surveys, a hierarchic
set of criteria can be applied at coding stage
to re-allocate reference person status. The
first criterion is that the person is one of
the partners in a registered or de-facto marriage.
Either partner can therefore be used as reference
person. In other contexts this status is classified
as 'member of a couple'.
Dissemination
of gender statistics
16.
The ABS series of social report
publications has contributed to developing gender
statistics. In 1991 An Index of Information about
Women was published, which provided a guide
to the scope and availability of ABS data on women.
This was followed in 1993 by a social report Women
in Australia. For the first time, a range of
relevant statistics about women, from ABS and other
sources, were brought together, presenting where
possible comparable data for men and trends over
10-20 years. A similar report on health issues was
published in 1994, covering the seven major priority
areas identified in the Federal Government's National
Women's Health Policy. In 1994, 1995 and 1997 the
Australian Women's Year Book was produced,
jointly funded with the Office of the Status of
Women. Data from both ABS and external sources were
used. The basic structure for all three editions
included a number of core topics such as population
and demographic data; education and training;
working life; economic resources; health. New or
periodic data, particularly data from ABS special
social surveys, were incorporated as these became
available. Comparative data for women and men were
presented both in tables and graphs. Sets of summary
indicators, covering a ten year time series, were
also developed for the core chapters.
17.
The ABS publishes an annual series,
Australian Social Trends, presenting information
on contemporary social issues and areas of public
policy concern, and intended for use by a wide audience.
A number of articles relating to gender issues have
been published, eg trends in fertility; looking
after the children; gender differences in
educational achievement; trends in women's employment;
differences in men's and women's earnings. Articles
on other issues may also include relevant disaggregated
data.
18.
The ABS is currently preparing
a statistical appendix for Women in Australia,
1999 to be published by the Office of the Status
of Women. The appendix presents a series of summary
tables showing the circumstances of women in Australia
in the late 1990s with comparable data for men,
and including comparisons over time where data are
available.
The
major conceptual issue of recognition and valuation
of unpaid work
19.
Because many unpaid household
activities were classified as falling outside the
production boundaries as defined by the System of
National Accounts, the activities were not included
in calculations of the national product. As many
of the activities were performed by women, this
was often interpreted as gender bias, which, it
was suggested, had the important consequence in
this instance of producing an inaccurate estimate
of the national economy. The question of recognising
the value of unpaid work was addressed in the review
of the SNA, which proposed the introduction
of separate satellite accounts. It has also
been generally recognised that regular time use
data are necessary for such accounts, and that an
international classification of activities for analysis
of time-use statistics should be developed.
20.
The ABS has now had experience
of three Time Use surveys: a pilot survey in the
largest city, Sydney, in 1987, and two national
surveys in 1992 and 1997.
21.
The planning and development of
the 1987 pilot survey benefited from the wide-spread
international interest in time use studies in the
post-war period. It included the daily activity
patterns of women, in particular in relation to
care of children and the domestic care of the frail,
sick and disabled, and the effects of labour force
participation on activity patterns. The valuation
of unpaid work, and its inclusion in estimates of
GDP, was also an issue for certain women's lobby
groups. More generally, the methodological issues
being considered in the review of the 1968 UN SNA
were of interest within the ABS and government.
Data on the patterns of time use were published
and the ABS prepared experimental estimates of the
value of unpaid household work, and reviewed the
relationship with estimates of GDP (ABS, 1990).
The pilot survey was also used to test thoroughly
the collection methodology, including use of time
diaries and optimum reporting period, and to further
develop a classification of activities relevant
to Australian time use patterns.
22.
With wide-spread interest continuing
in time use statistics, the ABS has now followed
up the pilot survey with two national Time Use Surveys,
in 1992 and 1997. In each survey year there
were four collection periods of one fortnight in
each of the four quarters of the year, thus covering
any seasonal differences in patterns of activity.
All days of the week were surveyed in equal proportions.
Individual time diaries were completed for two specified
days, by all members of the selected households.
In addition, a household form and individual questionnaires
were completed by personal interview. The individual
questionnaire covered demographic and socio-economic
information; in 1997 ABS standard collection modules
were used for topics such as income, education and
the labour force.
23.
The classification of time use
activity was further developed for use in 1997,
following a detailed review of the 1992 classification.
The conceptual basis was taken from work done by
Dagfinn Aas (1982). In this typology, there are
three dimensions: what the activity is, where the
activity takes place, and with whom the activity
takes place; and four categories of time: necessary,
contracted, committed and free time. The 1997 Time
Use Activity Classification used nine major categories,
arranged to relate to the typology:
Necessary time
1. Personal care activities
Contracted time
2. Employment activities
3. Education activities
Committed time
4. Domestic activities
5. Child care activities
6. Purchasing activities
7. Voluntary work and care activities
Free time
8. Social and Community
interaction
9. Recreation and leisure
24.
The overall patterns of time use
activity are of continuing interest to government
and non-government organisations, with relevance
to policy development in diverse areas, including
community services, employment, women's affairs,
families, education.
25.
The 1992 data were used to update
the estimated value of unpaid household work and
volunteer and community work, generally termed 'total
unpaid work', and further work was undertaken on
issues of definition, the theoretical framework
for the valuation of unpaid work and actual estimation
methods (ABS, 1994b). A similar update is being
prepared using the 1997 data set, and comparisons
between the two time points will be made.
The
emerging issue of changing work patterns
26.
The data from time use diaries
are also being used to investigate the specific
topic of changing patterns of paid work - both length
of working days and the time of day/day of the week
the activity takes place.
27.
Standard labour force data have
already indicated that there is a shift away from
the norm of a standard full-time job of 35-40 hours
a week. This trend has been observed in other OECD
countries as well as Australia. The proportions
of the working population with very long hours or
with short hours are both increasing. This is a
major issue for the well-being both of individual
women and men, and of their families, and has wider
implications for, eg, provision of welfare services.
Time use diaries provide more detailed data for
extending the analysis to look at prevalence of
'unsociable working hours', ie work that takes place
outside the (former) 'standard' work day of 9am
to 5pm on weekdays, and to relate this to other
individual characteristics such as occupation, family
relationships.
Continuing
issues - individual access to and control of economic
resources
28.
The Beijing Platform for Action
identified several continuing issues relating to
women's poverty and access to economic resources.
29.
Many forms of income can be measured
at the individual unit level, and the ABS has such
data from surveys of employee earnings and from
various Income and Housing Costs Surveys. These
provide appropriate data for eg. comparison of female
and male earnings; sociodemographic data are also
available to extend the scope of the analysis.
For many measures of economic well-being, however,
the preferred statistical unit is the income unit,
ie the group that forms a single spending unit and
which (it is assumed) shares economic resources.
30.
For analysing distribution of
income, and more general economic well-being, the
household is the smallest unit for which the necessary
measures of consumption, saving and net worth can
be constructed (ABS, 1995). Thus assumptions have
to be made about the extent of income sharing within
the income unit, and the unit measures of
economic well-being have to be considered only as
proxies of individual economic well-being.
31.
With the development of a standard
income module suitable for use in various types
of survey instruments, it will be possible in the
future to link income data with a wide range of
other sociodemographic variables, and extend the
scope of analysis relating to economic well-being.
Continuing
issues - changing definitions to accommodate new
concerns
32.
Changing concepts, item definitions
etc, particularly those used in regular repeating
surveys such as the ABS monthly labour force survey,
will disrupt time series for the data items affected.
33.
Classification changes agreed
in international fora, such as ILO standards, have
to be accepted if data sets are to be internationally
comparable. The ABS plays an active part in such
international discussions and accepts the need for
comparability. As an example, however, following
adoption of the changed standard for classifying
unpaid family workers in 1986, estimates of employment
and unemployment varied somewhat compared with the
previous period.
34.
It may be possible to collect
data so that sub-items can be combined in different
ways, making possible the construction of, and therefore
comparisons with, previous measures. The ABS has
tried this approach eg. with the definitions used
in the 1997 Survey of Disability and the new classification
of time use activity.
35.
However, there will be continuing
tensions between the preference for comprehensive
data appropriate to current conditions and the need
for essential time series data.
Review
of the ABS Household Survey Program
36.
Household surveys play a crucial
role in providing the range of statistics required
for monitoring the social and economic well-being
of the population (and inter alia for developing
gender statistics) and for informing the development
and evaluation of relevant government programs.
However, the resources available for such programs
are rarely elastic. As is generally recognised,
social surveys, especially interview-based national
population surveys, are expensive to run and analyse.
37.
In 1998/99 the ABS established
a review of its household survey program.
The impetus for the review came from a recognition
that the current level of demand for information
from household surveys exceeded the capacity of
the survey vehicles and resources available.
It was also recognised that there was a lack of
flexibility resulting either from commitments to
regular surveys in particular topic areas or from
constraints imposed by the survey vehicles.
There were extensive discussions with users to establish
the highest priority information needs from the
household survey program.
38.
At the time of writing this paper,
final recommendations are still in preparation.
However, from the earlier rounds of consultations
it was apparent that for every survey topic that
had been included in the program in recent years
there was a user or group of users giving it high
priority. Overall a strong case was made for regular
data collection across all areas of social concern.
39.
There was considerable interest
in a general social survey which would regularly
obtain broad information across a range of areas
of social concern. This reflects the increasing
recognition that many social phenomena are inter-related
and that social policy is becoming less sectoral
as a consequence. The interest in gender statistics
also can be related to this change in focus. Users
also reported the need for longitudinal data, to
reflect the dynamics of an increasingly diverse
society, in particular changes in behaviour and
personal circumstances over time.
40.
The timing and content of the
data collections relating to health, disability,
household expenditure, and income and housing costs
are currently under review. The ABS is also
examining information needs relating to education
and training, literacy, assets and wealth, and violence
and safety. Finally, there are substantial
needs for information to adequately address the
social and economic well-being of the Aboriginal
and Torres Strait Islander people of Australia.
References
ABS (1990)
Measuring Unpaid Household
Work: Issues and Experimental Estimates
ABS (1993)
Time Use Survey 1992
- User's Guide
ABS (1994a)
How Australians Use Their
Time, 1992
ABS (1994b)
Unpaid Work and the Australian
Economy, 1992
ABS (1995)
A Provisional Framework
for Household Income, Consumption, Saving
and Wealth
ABS (1996)
Women's Safety, Australia
1996
ABS (1997a)
Women's Safety Australia
1996: User Guide
ABS (1997b)
ASCO - Australian Standard
Classification of Occupations, Second Edition
ABS (1998a)
How Australians Use Their
Time, 1997
ABS (1998b)
Time Use Survey 1997
- User's Guide
ABS (1999)
Disability, Ageing and
Carers: Summary of Findings
ABS (1994 on)
Australian Social Trends
ABS (1994/5/7)
Australian Women's Year
Book
Aas, D. (1982)
'Designs for Large Scale Time Use Studies
of the 24 Hour Day', in It's About Time:
Proceedings of the International Research
Group on Time Budgets and Social Activities,
ed. Z. Staikov, Printer, Sofia, Bulgaria.
McEwin, M. (1995)
Social Indicators and Social Statistics
in Australia. Paper presented to Siena
Group Seminar, Oslo.
Hedman, B.,
Perucci, F. and Sundstrom, P (1996) Engendering
Statistics: a Tool for Change, Statistics
Sweden.
UN (1996) The
Beijing Declaration and Platform for Action,
United Nations, New York.
