Docstoc

Improving the quality of data on early childhood education

Document Sample
Improving the quality of data on early childhood education Powered By Docstoc
					 
 
 




     Improving the quality of data on early
             childhood education


 
                              
                              
                              
                              
                              
                              
                              
                              
                              


    A report for the Department of Education, Employment
                   and Workplace Relations




                            by



               Marion McEwin and Matthew Ryan




                       September 2008
                              




                             1
Executive Summary.................................................................... 4
  Background ................................................................................................................4
  Data sources ...............................................................................................................4
  Statistical limitations..................................................................................................4
  Institutional factors ....................................................................................................5
  Common reporting .....................................................................................................5
  Review of Government Service Provision (RoGS) ...................................................6
  Improving data on early childhood education ...........................................................6
Introduction ................................................................................ 7
  Project Brief ...............................................................................................................7
  Background ................................................................................................................7
  The importance of data ..............................................................................................7
  Key information needs...............................................................................................7
  Some definitional issues ............................................................................................8
  Data items and indicators...........................................................................................9
Data sources............................................................................. 10
  Types of data sources...............................................................................................10
  Assessing data quality..............................................................................................11
Strengths and weaknesses of available datasets ................... 12
  (i) ABS Census of Population and Housing ............................................................13
  (ii) ABS Child Care Survey (ABS CCS) .................................................................14
  (iii) State/territory administrative data.....................................................................16
  (iv) National Preschool Census (NPC) ....................................................................17
  (v) Child Care Supplementary Data Collection (CCSDC) ......................................20
  (vi) Child Care Benefit payment administrative data ..............................................22
  (vii) Longitudinal Study of Australian Children (LSAC)........................................23
  (viii) The Household, Income and Labour Dynamics in Australia Survey
  (HILDA) ..................................................................................................................24
  (ix) Community Services Industry Survey ABS......................................................24
Data Map: from data item to data source ................................ 26
Issues and strategies for improving data ................................ 28
  Problems, challenges and issues for data development ...........................................28
  Some facts and figures .............................................................................................29
  Importance of common reporting ............................................................................31
  Achieving common reporting ..................................................................................32
  Future directions for early childhood education data ..............................................35
  Next Steps ................................................................................................................40
Appendix 1: Identified Data Deficiencies ................................ 41
  The size of the ‘gap group’ ......................................................................................41
  Hourly costs of preschool ........................................................................................42
  Four-year qualified preschool teachers....................................................................42
  Disadvantaged groups..............................................................................................43
Appendix 2: Project Description ‘Improving the quality of data
on early childhood education’ .................................................. 46
Appendix 3: Template for ‘starting point’ analysis .................. 47
Appendix 4: Census measures of disadvantage ...................... 50
Appendix 5: Bibliography ......................................................... 53
 
                                                             2
ACRONYMNS

ABS         Australian Bureau of Statistics
AEDI        Australian Early Development Index
AGCCCS      Australian Government Census of Child Care Services
AIHW        Australian Institute of Health and Welfare Services
AVETMISS    Australian Vocational Education and Training Management
            Information Statistical Standard
COAG        Council of Australian Governments
CCB         Child Care Benefit
CCMS        Child Care Management System
CCS         Child Care Survey
CCSDC       Child Care Support Data Collection
CEACS       Childhood Education and Care Survey
CSIS        Community Services Industry Survey
CSDWG       Children’s Services Data Working Group
CSNMDS      Children’s Services National Minimum Data Standard
DEEWR       Department of Education, Employment and Workplace
            Relations
ECE         Early Childhood Education
FACSIA      Department of Family and Community Services and
            Indigenous Affairs
HILDA       Household Income and Labour Dynamics Australia
LSAC        Longitudinal Survey of Australian Children
LDC         Long Day Care
MCEETYA     Ministerial Council on Education, Employment and Youth
            Affairs
NPC         National Preschool Census
NSSC        National Schools Statistics Collection
OECECC      Office of Early Childhood Education and Child Care
RoGS        Report on Government Services
SEIFA       Socio-Economic Indexes for Areas
VET         Vocational Education and Training




                             3
Executive Summary
Background

This project is a broad landscape study of the data on early childhood education
available to support program development, performance monitoring and funding
allocation. It discusses the strengths and weaknesses of various data sources and
comments on future possibilities for data improvement.

The impetus for the project is a policy focus on early childhood education and, in
particular, the Australian Government commitment to providing universal access to
quality preschool education for all children in the year before formal school.

Data sources

There are quite a number of collections that provide information on early childhood
education including:

   •   DEEWR National Preschool Census
   •   Australian Government Census of Child Care Services
   •   ABS Census of Population and Housing
   •   ABS Child Care Survey
   •   State/Territory Preschool Censuses and Administrative Data
   •   Child Care Benefit Payment Administrative Data
   •   Longitudinal Survey of Australian Children

All of these collections have their own specific objectives. Consequently even
though the individual datasets are quite rich there are many limitations of the
existing data when it is assessed against the information needed to support program
development and performance monitoring of the universal preschool access
commitment.

When data items drawn from a preliminary list of performance indicators are
mapped against possible data sources, many gaps are evident. These gaps are in
part a consequence of an ambitious performance indicator list. This list may be
refined to focus on a subset of the most important and useful.

Nevertheless it is clear that even the most fundamental measures of preschool
attendance and participation cannot currently be reliably produced in a form that is
comparable across the jurisdictions and consistent over time.

Statistical limitations
The major weaknesses of the various data sources can be summarised as:

       •   Partial coverage of most collections
       •   Inconsistent definitions across jurisdictions
       •   Duplication and double counting
       •   Difficulties integrating preschool and child care statistics


                                          4
Institutional factors

The observed data limitations need to be seen in the context of the diverse
arrangements for delivering preschool education across the country. These make
the compilation of statistics that describe and monitor the sector a very complex
task. Problems and challenges for data development include:

         •   The split of policy and program responsibilities between Commonwealth
             and state/territory governments.
         •   The different mix of early childhood provision across the jurisdictions
             between government, private and community.
         •   The variety of settings in which preschool may take place and the
             imprecise boundary between child care, preschool and early education.
         •   The possibility that children and teachers may participate in more than
             one program and service.
         •   Different school starting ages across jurisdictions hence different
             definitions of the “year before formal schooling” and associated
             differences in the age composition of the preschool population.


Common reporting
Common reporting is a means by which statistical outputs can be brought together
on a consistent basis and the limitations described above are addressed. Features
and arrangements which underpin successful examples of common reporting
include:
   (1)       A resolve on the part of some auspicing entity (e.g. COAG) that
             common reporting is an important goal that is inextricably linked to
             efficient and effective service delivery.
   (2)       A corresponding commitment of the key players across the jurisdictions
             to make it happen.
   (3)       Adequate priority, resources, effort, attention and time to undertake the
             necessary development work.
   (4)       A willingness to change existing systems and approaches to meet
             agreed statistical goals.
   (5)       Development and implementation of national uniform data standards
             that provide for consistency within a collection and allow for data to be
             integrated across collections.
   (6)       Oversight or coordination by an agency with the necessary expertise and
             credibility with both the jurisdictions and the Commonwealth, and
             which takes responsibility for the statistical output and its quality.
   (7)       An advisory and/or consultative group (and associated working groups)
             with representation of all stakeholders to work with the coordinating
             agency.




                                            5
Review of Government Service Provision (RoGS)

Common reporting across a wide range of government functions including
children’s services has been the aspiration of the COAG Report on Government
Service activities for over a decade. Progress has been made in the early childhood
area despite still being well short of this goal.

RoGS has been hampered in its task by the absence of an integrated national
collection from which to draw early childhood statistics. In comparison, RoGS is
much better served in the health, schools and vocational educational areas.

The definition of a government service used by RoGS in the preschool and
childcare area is currently limited to government funded and/or provided activities.
That is, the RoGS definition does not include preschools which are both not
government funded and which are not provided by government. The Starting Point
Analysis may reveal to what extent the RoGS definition is currently being
measured. The RoGS definition may need to be broadened if it is to remain
relevant in monitoring provision of universal access to preschool in the changed
policy environment

Improving data on early childhood education
The new Council of Australian Governments (COAG) framework for federal
financial relations makes governments responsible and publicly accountable for the
money spent and the outcomes delivered. Good statistics are essential to assess
these outcomes and monitor performance towards achieving national goals.

Changes needed to ensure that statistics are adequate for these purposes include:

   (1)      The implementation of common reporting requirements described
            above.

   (2)      A national collection which covers all preschools across the country
            with common definitions. This would either be an enhanced national
            preschool census or a replacement for it.

   (3)      Child care data that can be added to provide a complete picture. This
            would be a replacement for the former Australian Government Census
            of Child Services.

In addition, consideration should be given to constructing a linked dataset that
charts children’s activities in their early years and their transition to school. Such a
data set would overcome many of the difficult methodological issues that confront
statistical collections in the early childhood area. It would also support the analysis
of school outcomes and children’s well-being in the context of experiences in their
early years.

Marion McEwin
Matthew Ryan
15 September 2008



                                           6
Introduction
Project Brief
The project is aimed at improving the quality of data on early childhood education.

This report has been commissioned to:
   • Undertake a landscape study of information requirements.
   • Describe and analyse the advantages, accuracy and limitations of various
       sources of data on early childhood education and related activity.
   • Scope the possibilities for shorter and longer term data improvements.

Background
Early childhood education is a high priority policy issue. The Australian
Government’s agenda for early childhood education and childcare focuses on
providing Australia families with high-quality, accessible and affordable integrated
early childhood education and childcare.

Specifically, there is a commitment that all children, in the year before formal
schooling, will have access to:
   • 15 hours of government-funded early childhood education, for a minimum
       of 40 weeks per year.
   • Through programs delivered by degree-qualified early childhood teachers.
   • In public, private and community-based preschools and child care.

Universal access is to be achieved within five years, with the Australian
Government working together with State and Territory governments.

The importance of data
Data is needed to ensure that decisions and strategies are soundly based on good
evidence. More specifically it is required to:
    • Monitor progress towards stated goals and targets.
    • Assist in program development and implementation.
    • Support research, evaluation and review.
    • Underpin costing models and funding allocations.

At its meeting of 26 March 2008, the Council of Australian Governments (COAG)
agreed to develop implementation plans for improving national data to guide
planning and monitor effectiveness of the COAG reforms on early childhood.
COAG recognised in particular the need for a collaborative whole-of-government
approach to improve the evidence base on early childhood education. The newly
formed Office of Early Childhood Education and Child Care (OECECC) are
seeking to provide leadership for this task.

Key information needs
A range of data will in many cases jointly address requirements to monitor
performance towards goals, support funding models and program evaluation. The
key questions that this data should address include:



                                         7
   1. How many and what proportion of children are participating in preschool
      programs in the year preceding formal schooling? What are the
      characteristics of these children and their families? How does this
      participation vary across states and areas within state?
   2. How many and what proportion of children are participating in some form
      of early education that does not meet the government commitment? In what
      ways does this fall short? How does this vary across states and areas within
      state?
   3. To what extent are children enrolled in preschool programs taking full
      advantage of them in terms of their attendance during the week and across
      the year?
   4. What are the barriers to participation? What are the factors affecting non-
      participation including socio-demographic, geographic and contextual
      factors?
   5. What is the nature and characteristics of preschool provision? What are the
      types of providers of preschool programs? What are the hours and patterns
      of participation in preschool programs, on a weekly basis and across the
      year?
   6. What are the characteristics of preschool program delivery? What are the
      qualifications of staff delivering preschool programs across all settings?
   7. What funding is needed to support delivery of the universal access
      commitment?

Some definitional issues
A full canvassing of all definitional issues in the early childhood area is beyond the
scope of this project. However, it is useful to draw attention to some of these
definitional issues as it points to areas where further thinking and discussion will be
needed.

Preschool– In this report the term “preschool” is used to denote the early childhood
education program described in the government commitment. Statistical collections
use the term variously. Some are circular e.g. “preschool comprises services
usually provided by a qualified teacher on a sessional basis in dedicated preschools”
(RoGS). The national preschool census adopts the definitions used across states
and territories and notes that these are different.

Access – This is a fundamental aspect of the government commitment. Yet it is not
straightforward to define. Access is not the same as enrolment or attendance.
Rather, it encompasses availability of places, across a range of locations, with
convenient session times that suit the needs and circumstances of families. Even if
these conditions are met, there will be families who choose for various reasons not
to take up (non-compulsory) early education in a preschool setting.

Region – It is obvious that some form of spatial information is important below the
jurisdiction level. Whether this is for small areas such as LGA or for broader
groupings (such as the remoteness classification) needs to be determined. Careful
thought is needed on how fine geographic data could be used (assuming it could be
reliably generated).

                                           8
Data items and indicators
Based on the questions above and OECECC identified performance indicators in
the following table, it is possible to identify a list of specific measures and data
items. Further optional data items are listed in Appendix 3.

Category        Quantitative Information
                set against the population, demographic and geographic profile of each state and
                territory e.g. against ABS data/estimates

Participation   • Enrolment and attendance rates in early childhood education across the
and Equity        jurisdiction
                    o against eligible population of children
                    o by service type
                    o at a regional level

                • Rates of participation by children from disadvantaged groups (Indigenous;
                  children of newly arrived migrants; low SES; children with disabilities),
                  including at a regional level (where available).
                • Number of children who currently miss out on early childhood education by
                  demographic profile (where available)

Accessibility   • Number of early childhood education providers by location (across preschool
and Utility       and child care settings),
                • Number of places, existence of waiting lists, hours of operation by service type
                  (where available)
                • Provision of integrated child care and preschool services
                • Broader on-site service – e.g. maternal and child health, parent support
                • Fees charged by service type and fee reduction for low income users

Quality         • Workforce qualifications (care staff; education/teaching staff) by service type
                • Child/staff ratio by service type
                • State licensing/registration requirements by service type
                • Teachers registration (where applicable)
                • Number of workforce exemptions granted (where applicable)

Quantity        • Average hours of attendance in a preschool program per week by service type
                • Average number of weeks per year a preschool program is provided by service
                  type


 
The next section on ‘Data sources’ identifies the issues in selecting a data source
and the actual data sources available for informing these indicators. The strengths
and weaknesses of the data sources are identified.

The following section ‘Data map’, maps from required data indicator to data source.
It aims to answer the question – given the indicator I want to quantify, what data
item from which data source can I use?
 


                                           9
Data sources
Types of data sources
In common with any subject field, there are various sources of data relating to early
childhood education and these have varying properties. There is no single,
definitive source.

Data may come:
   • From a self-contained statistical collection designed specifically to gather,
      process and combine information on a particular topic (e.g. the preschool
      census or the survey of early childhood education and child care).
   •     From a more general collection (e.g. the census of population and housing
         or the longitudinal survey of Australian children).
   •     As the by-product of an administrative process set up for regulatory or
         service delivery purposes (e.g. preschools, customs, taxation, health
         benefits).
   •     As the result of an analytic or modelling process which brings together data
         from one or more sources (e.g. SEIFA, RIM, STINMOD).
Potentially all of these have a part to play in the information system put in place to
underpin decision making, performance monitoring, and evaluation of service
delivery.


Activity                   Strengths and weaknesses
Census                     o Can provide a high level of geographic and cross
                             classified detail
                           o Usually takes a while to process
                           o May be slow to incorporate new developments
Survey                     o More cost effective than a census & better scope for
                             error control
                           o Can provide an integrated national picture
                           o Results for smaller states subject to higher sampling
                             errors
                           o Longitudinal surveys can provide better
                             understanding of cause and effect but are costly and
                             may be affected by attrition biases
                           o One off surveys may be better for specific
                             investigations while repeated surveys are needed to
                             chart change over time
Administrative process     o Can be highly cost effective as some data flows are
                             generated as a by-product of government transactions
                           o Statistics are comprehensive for a given
                             administrative function
                           o Coverage can be partial if some components outside
                             the system
                           o Inconsistency between jurisdictions when systems
                             differ
                           o Are amenable to national data standards initiatives to

                                          10
                               align concepts, methods etc to produce uniform
                               national statistics
Analytic and/or            o   Can bring data together from a number of sources and
modelling system               produce outputs that cannot be directly collected
                           o   Can adjust for missing data, errors and other
                               inaccuracies
                           o   Can present scenarios and answer what-if questions
                           o   Highly dependent on assumptions which can be made
                               transparent and allow for sensitivity analyses
 
In the education and child care area, statistics can be collected from the supply
(service delivery) side or from the demand (client) side. Typically supply side
statistics are obtained by surveying the institution or business (that is, school,
preschool or child care centre) while demand side statistics are obtained by surveys
of the population. The respective counting units in these collections are then
institutions and children. This is not to suggest that information on children (i.e.
clients) cannot be obtained from surveys of institutions but rather that compiling
statistics based on them is not always straightforward and some important indicators
such as participation rates can become problematic.

Assessing data quality
Almost without exception all statistical data is subject to error and/or quality issues.
These are a function of how the data is gathered and compiled. The key issue is to
understand this error and take steps to minimise errors that impact seriously on the
intended use of the statistics. Errors can arise at the collection stage through:
o Incomplete coverage – where only a part of the population of interest is
    included.
o Reporting errors – where the respondent is unable to provide information
    exactly as requested or does not understand what is required.
o Sampling error – associated with surveys – this can be controlled and estimated
    through good sample design but will have most impact on finely disaggregated
    data and estimates for small states and territories.
o Non-response errors – arising when those units which do not respond have
    different characteristics from the responding units and distort the inferences
    made.
o Comparability errors – caused by differences in timing, concepts, definitions,
    compilation methods across time or between states/territories.

Errors can also arise at the compilation stage. Statistics derived as the difference or
ratio of other estimates may be affected as the errors in the base figures can be
magnified. Ratios can be particularly problematic when the numerator and
denominator are drawn from different estimation systems and are not congruent –
that is, they are measuring different things.

Finally, perceptions of data quality may be captured under the heading of
credibility. Data that can be shown to have been generated by a disinterested
agency using a nationally consistent methodology minimises the risk that data may
be manipulated or ‘gamed’ by one or more stakeholders in order to achieve narrow
goals that may not be in the national interest.  



                                          11
Strengths and weaknesses of available datasets

There are eight data sets of potential relevance in the analysis of early childhood
education. The table summarises the more detailed discussion that follows it.

Collection/Source          Strengths and weaknesses
Census of Population       o identifies whether children attend preschool but the
and Housing (ABS)            variable is deficient due to the range of preschool
                             settings
                           o provides the population baseline for up to date
                             estimates of resident population by age (which are
                             used to construct rates)
                           o provides a wide array of local area contextual data
                           o provides an area based socio-economic index of
                             disadvantage
Child Care Survey          o as a supplement to the monthly labour force survey
(ABS)                        provides national and state/territory statistics every
                             three years
                           o provides good contextual data on parents and family
                             circumstances including work, income, indigenous
                             status etc
                           o data on preschool attendance is collected but early
                             learning has not been a focus of this survey prior to
                             2008
Report On Government       o produced annually to allow for regular monitoring
Service Provision          o has a wide array of performance measures, including
(RoGS)                       costs, participation and demand indicators
                           o draws on ABS surveys and relevant administrative
                             collections managed by a variety of agencies within
                             each jurisdiction.
                           o underpinned by a collaborative process which is
                             looking to progressively improve data quality and
                             comparability
                           o significant areas of inconsistency in data between
                             jurisdictions
                           o preschool participation rates subject to double
                             counting and other conceptual problems and lack face
                             validity
National Preschool         o prime purpose is to determine Indigenous education
Census (DEEWR)               program funding.
                           o also produces counts of all children enrolled in
State/Territory
                             preschool by age and sex derived from state/territory
Preschool Censuses
                             preschool censuses
                           o major limitation is the different definitions of
                             preschool across the jurisdictions
Australian Government      o covers all child care providers in receipt of
Census of Child Care         Commonwealth funding (although about 10% do not
Services (FaCSIA)            respond)
                           o includes some information on preschool programs

                                          12
                              and staff characteristics such as qualifications
                          o   last run in 2006 - undergoing redevelopment
Longitudinal Survey of    o   includes data on child care and early education
Australian Children           arrangements
                          o   follows children through time for 7 years from
                              2003/04
                          o   only 2 cohorts chosen and will not be refreshed
Community Services        o   child care services are included in this irregular ABS
Industry Survey (ABS)         survey last run in 1999/2000
                          o   information collected relates to the business operation
                              including employment and financial data
                          o   information is not collected on the nature of the
                              service delivered or its clients (ie the children)


(i) ABS Census of Population and Housing

The ABS Census of Population and Housing is conducted every five years, most
recently on 8 August 2006. It is the most comprehensive statistical stock take of the
nation obtaining a wide range of characteristics of people, families and dwellings.
All persons resident in Australia on Census night are included. Data is obtained
through self-enumeration by the householder.

The Census contains virtually no sampling errors, although it may have other types
of errors including respondent error, processing error and item non-response.
(These errors occur in surveys as well)

A wide range of data is readily available from the ABS in many different forms
including tabulations, community profiles, and confidentialised unit record sample
files. The Census also provides the base for other statistical compilations and
derivations including population estimates, population projections and socio-
economic indexes for areas (SEIFA).

Information on the number of children attending preschool is obtained in the
Census from a question on type of educational institution attending within which
there is a response category – ‘preschool’. Children in Long Day Care are not
identified. An instruction on the census form states mark “no” for children enrolled
only at child care centres.

Strengths
   • Contains a wide array of contextual data on the social, demographic and
      economic circumstances of families and children.
   • The most reliable source of information on small areas (e.g. LGA, suburb
      postcode) and small populations (e.g. Indigenous, immigrant groups, etc).
   • Provide the benchmark for population estimates and projections.
   • A key Census output, socio-economic indexes for areas (SEIFA) can be
      used to identify areas of relative disadvantage.
   • Does collect some information on preschool attendance.
   • Data is readily available and accessible.

Weaknesses
                                         13
      •    Data of this degree of comprehensiveness is inevitably infrequent (5 yearly).
      •    The generality of the Census means that education attendance data in respect
           of preschools can only be collected for a very broad concept.
      •    It is not clear how parents would report when children attend a preschool
           program within a child care centre.
      •    It is not possible to identify overlaps between the preschool and childcare
           sectors.

Overall

The Population Census is not sufficiently frequent or focussed enough to provide
performance monitoring information. However it can provide extensive contextual
information that might be used in conjunction with other collections. Furthermore,
the annual age specific population estimates for States and territories which are
based on the Census are a crucial input to funding allocations. Population
projections are important for estimating future costs of programs. Finally, the
SEIFA indexes can provide valuable geographic information on the distribution of
disadvantage across the country. See Appendix 4.


(ii) ABS Child Care Survey (ABS CCS)

ABS Child Care Survey (Cat. No 4402.0) is conducted throughout Australia every
three years as a supplement to the ABS Labour Force Survey (LFS). The most
recent survey was conducted in June 2005 with results released 11 months later.
The 2008 Survey is currently underway.

The survey collects data on the usage of various forms of child care and preschool
for children aged 0–12 years, and some aspects of families' requirements for formal
care. Information is also collected on the cost of child care as well as receipt and
non-receipt of the Child Care Benefit (CCB), and the income and working
arrangements of parents.

Information is obtained by interviewing the child’s parent or guardian and relates to
the child care attendance in the week preceding the interview. The survey is timed
to avoid school holidays.

This survey defines preschool as “Educational and developmental programs for
children in the year (or in some jurisdictions, two years) before they begin full-time
primary education.” In June 2005, Preschool was excluded from the definition of
formal child care on the grounds that preschool is preparatory to education rather
than childcare as such1. Nonetheless, preschool statistics were separately compiled
and presented as part of the Child Care Survey2.

Hierarchical data structure

Demographic                                   Service
Characteristics of family& parents            Child care and preschool arrangements,

1
    ABS cat. 4402.0, explanatory note 13.
2
    ABS cat. 4402.0, Tables 24-26.
                                            14
   Type and composition                     Type of care arrangements used
   Location                                 Days and hours of attendance
   Employment and labour force status       Cost of care
   Work hours and arrangements              Whether CCB received
   Income                                   Reasons need additional formal care
Characteristics of child                       Work-related
   Age and sex                                 Personal-related
   Whether attended school                     Child-related
   Main language child speaks at home       Reasons additional care not taken up
   Country of Birth                            None exist/don’t know where they
                                            are
                                               Cost too much
                                               Booked out


Over time, the survey has attempted to grapple with the ambiguity between early
education and child care inherent in the provision of these services to children. The
2008 survey has been expanded to include new content on early childhood learning
and has been retitled the Childhood Education and Care Survey.

Strengths
   • A comprehensive national ABS survey providing good comparisons over
      time.
   • ABS has been prepared to evolve the survey content over time to better meet
      the needs for policy development and evaluation.
   • Estimates for Australia and large states subject to small sampling error.
   • Provides good contextual information on parents and families.
   • Provides information on reasons for use and barriers to access.
   • Collects information about day care, and preschool attendance at the same
      time so that overlaps between the sectors can be identified.
   • Can identify the proportion of 4 year olds who are not attending preschool
      and/or long day care and how this varies by socio-economic status.
   • Information on indigenous children is available from unpublished data
      although it would only be accurate at a highly aggregated level.

Weakness
  • Figures for small states and territories are subject to high sampling errors
  • Small area data (i.e. LGA, postcode or suburb) is not feasible.
  • The survey does not include remote and sparsely settled areas – the effect
     will be negligible for most states but will have most impact on NT
     estimates.
  • Children who usually attend preschool/childcare but were away in that week
     will be missed.
  • The key focus of the survey has been childcare and the survey has been less
     concerned with preschool and early learning.
  • Information collected on preschool attendance relies on parent knowledge
     and perception of what constitutes preschool in an area where definitions
     and terminology can be somewhat imprecise.
  • Preschool attendance within a child care centre will probably not be
     distinguished.

                                         15
Overall

This survey can provide a big picture understanding of the child care patterns across
the country and particularly how they related to the labour force participation of
parents. It is one of the few vehicles that can provide congruent measures of non
participation and the characteristics of children (and their families) who are not
attending preschool. Due to its frequency and nature this survey could not play a
central role in monitoring progress of the jurisdictions towards the universal
preschool access goal. Nevertheless it can provide valuable supplementary and
support information.


    (iii) State/territory administrative data
 
Administrative data collections in the area of early childhood education are
managed by a variety of government agencies responsible for education and/or
children’s services functions within each state and territory. Data are collected by
these agencies and published in the annual Report on Government Services (RoGS),
most recently on 31 January 2008.
 
RoGS, first published in December 1995, is produced by the Steering Committee
for the Review of Government Service Provision (SCRGSP) which comprises
representatives from central agencies across Australia. The Secretariat for the
SCRGSP sits inside the Productivity Commission. The early childhood data is
assembled within the Children’s Services Working Group and is not formally
audited by the Secretariat to the SCRGSP.

Hierarchical data structure
State Government preschool spending        Preschool staff
(annual)                                   Number of staff employed by State
Govt expenditure on preschool services     Government funded and/or managed
  Recurrent expenditure ($)                child care and preschool service
  Administration expenditure ($)           providers
  Other expenditure ($)
  Family assistance ($)                    Primary contact staff
  Net capital expenditure on preschools        With a relevant formal qualification
                                               Without relevant formal
Preschool children                                qualification
Preschool places                               3+ years of relevant experience
                                               < 3 years of relevant experience
Number of children using State             Administrative staff
Government funded and/or provided          Other staff
child care services by age
                                           Preschool providers
Number and share of the following          Number of licensed and/or registered
categories of children s in preschools     preschool providers
o from non-English speaking                       Community managed
    background                                    Privately managed
o with Indigenous backgrounds                     Government managed
o with disabilities

                                          16
o       from regional areas                   Number of preschool service providers
o       from remote areas                     that provide non-standard hours of
                                              service
Average hours per week of preschool
attendance                                    Substantiated breaches arising from
                                              complaints about State Government
                                              registered or licensed preschool service
                                              providers



Strengths

    •     It is frequent (annual) and timely by comparison to other ECE data.
    •     It provides a range of data not otherwise accessible.
    •     Provides insights into what data is available at the jurisdiction level and how
          that data is presently organised.
    •     The report contains good contextual and descriptive data and explanation of
          the rationale for the performance measures
    •     Has some jurisdictional ‘buy-in’.

Weaknesses

    •     The data lacks national consistency and there are many qualifications to it
          (as indicated by the numerous footnotes to most of the tables).
    •     The outputs have not been subject to a quality assurance framework needed
          if data is to be used for performance monitoring.
    •     Data is pre-digested and reported in summary form by the jurisdictions.
          Therefore there is little scope to re-organise data to address specific policy
          issues. Unit record data is held by the jurisdictions.
    •     Data is not comprehensive and excludes non government preschools.

Overall

Over the last 10 years there has been considerable expansion of the Children’s
Services chapter in RoGS with increases in the number of indicators and discussion
of them. However the statistics are not underpinned by national uniform data
standards. If RoGS was to fulfil a longer term role and to produce performance
measures in line with the new framework for federal financial relations, then current
processes would need to be strengthened considerably.


(iv) National Preschool Census (NPC)

The NPC is conducted annually by DEEWR to determine Indigenous Education
Program (IEP) funding for institutions in scope of the collection. While the NPC
focuses on Indigenous children, it also collects data on the age, sex and
state/territory of all children enrolled in preschools. It has a 99% respondent rate
from about 3,500 contacted preschools.



                                            17
The collection has been conducted since 1983 with comparable data (other than for
Queensland) available from 2001 onwards. Prior to 2005 the NPC was known as
the National Indigenous Preschool Census. Data is collected with respect to a
reference week in August.

Data on government preschools are compiled from state government departments
using their existing census arrangements. Data on non-government institutions and
organisations offering an educational program are obtained by DEEWR through the
Non-Government Supplementary Census, on contract.

The collection covers government preschools which are included on each
jurisdiction's census list. Non-government establishments involved in the provision
of preschool education, registered preschools and centres offering an educational
program are also included.

The collection counts students as enrolled if they were on the roll during the census
week, and had attended a preschool education program in the last month. The data
therefore reflects the flow of Preschool students in the month leading up to the
census.

There are two parts to the collection. First, forms are despatched to registered non-
government preschools using the list provided by state and territory registration
authorities. All preschools are asked to return the form even if they did not have
any Indigenous enrolments in the census week. If the preschools do have
Indigenous enrolments they were required to complete an extra questionnaire.
Schools are given approximately four weeks to return the questionnaire. Telephone
follow-up of non-respondents is used.

Secondly, government data (and data in respect of Victorian non-government
preschools) are requested directly from the relevant state or territory government
authority.

Hierarchical data structure

Preschool details                           Indigenous students
State or territory                          Date of birth
Statistical District of preschool           Number of sessions
Metropolitan, Provincial, or Remote         Total hours enrolled
(MCEETYA classification)                    Whether expects to enrol in school next
Government or non-government                year
Number of Indigenous students enrolled
                                            Students
Number of all students enrolled             Age
                                            Sex
Number of Indigenous children               Indigenous identifier
currently on the waiting list               Government or non-government
                                            preschool
Number of all children currently on the     State or territory
waiting list                                Metropolitan, Provincial, or Remote
                                            Statistical District of preschool
Whether also offered a childcare

                                          18
program                                            Staff
                                                   Teaching Staff, Teaching Aides, or
Number of staff                                    Other
                                                   Indigenous identifier
Number of Indigenous staff



Note the following qualifications to the definitions used in the NPC:

    •   ‘Preschool’. The definition used includes all State and Territory specific
        terminology, such as kindergarten, pre-primary or Child Parent Centre.
    •   ‘Location’. The NPC uses the MCEETYA3 Classification of Geographic
        location of ‘metropolitan’, ‘provincial’ and ‘remote’. This classification
        differs from the more widely used ABS classification of geographic
        locations4.
    •   ‘Indigenous’. Every attempt was made to use the definition of an
        Indigenous child as “Any child who is identified as Aboriginal or Torres
        Strait Islander by the parent or guardian and is recognised as such by the
        community.” However, sometimes this is settled by the director of the
        Preschool filling out the Census form.
    •   Consistency between jurisdictions. From 2001, children in the “year
        before Year 1” were excluded from the NPC. Queensland has been steadily
        introducing a year before year 1 (‘Prep’) into its formal school system – in
        line with the rest of Australia. This implementation was completed in 2007.
        Accordingly, the estimates of Queensland Preschoolers will be understated
        in the published NPC reports from 2001 through to 20075.

Strengths

    •   The NPC has a particular focus on indigenous preschoolers.
    •   The NPC is potentially rich in regional detail.
    •   The 2008 edition of the NPC will have Queensland preschooler estimates
        that are no longer inherently downwards biased.

Weaknesses
  • Does not use a standard definition of preschool across the country.
  • No cost/price data.
  • No staff qualifications data.
  • The filter on indigenous status may be partial.
  • There has been little testing of the questionnaire and certainly not recently.
  • The problem of interpretation of "Preschool" by respondents, results in
     participation numbers differing in all censuses [see ABS SCH in
     bibliography].


3
  Ministerial Council on Education, Training and Youth Affairs
4
  Accessibility/Remoteness Indicator of Australia (ARIA)
5
  As the NPC in any particular year refers to data from the previous year, the Queensland reforms
will not be fully reflected in the NPC until 2008.

                                                19
   •   Outputs are not visible and accessible – they appear to be on limited
       circulation.

Overall

This is seen as a collection with considerable potential. While improvements to
address the weaknesses identified above are highly desirable, its focus on
Indigenous children and its comprehensiveness and potential for regional data are
very important.


(v) Child Care Supplementary Data Collection (CCSDC)

The Australian Government Census of Child Care Services (AGCCCS) was
conducted regularly since 1986 to 2006, and may not be conducted again in its
current form. The bulk of the child based information that had previously been
collected as part of the Census will now be collected by the Child Care
Management System (CCMS). The Australian Government is currently
investigating data collection mechanisms to supplement these data, such as the
CCSDC.

The information provided in the Census was the most comprehensive data available
on Australian Government approved and funded child care. It was used extensively
for monitoring growth, operation of services and assisting in policy formulation and
planning. Externally, summary data was made available to State and local
governments, peak child care organisations through the internet and the Report on
Government Services (RoGS), and for research into child care.

The Census examined information about child care users, staff and carers, and
operational details of child care services for nine service types. In particular, the
Census collected information on the number of services that offered an early
childhood education program and was conducted by a qualified early childhood
teacher. The Census also collected detailed information about staff and carers, in
particular their child care related qualifications and experience in the industry.

Move to the Child Care Supplementary Data Collection (CCSDC)

As a replacement to the census, the CCSDC is being conducted during September
and October 2008, to be reported in December 2008. This aims to deliver the
information that used to be available in the census but is not currently provided by
CCMS. The supplementary data is mainly about the staff, care recipient activities
and service's operation.

Hierarchical data structure

 Children                                    Service
   Numbers attending by age group               Type of management
   Numbers by age group for the                 Fees charged by age
      following categories :                    Numbers of children participating in
   Indigenous origin                                in-house preschool program
   With non English speaking parents            Numbers of children taken out to

                                          20
    Needing assistance due to disability            government preschool program

 Staff
    Weekly hours worked
    Type of work
    Employment status
    Years of experience
    Qualifications (type and level)
    In-service training in last year


Strengths

   •   The former AGCCCS national uniform data standards will be maintained.
   •   Results will be presented and analysed in a comprehensive and thorough
       statistical report, including RoGS.
   •   The CCSDC will also identify access to preschool programs (either in-house
       or take out) within child care centres.
   •   Reporting child and service level profiles with data collected through CCMS
       in future should provide a rich dataset.
   •   The CCSDC will minimise administrative impost on providers at a lower
       cost than the AGCCCS.


Weaknesses
  • Response rates have declined in the last 10 years (89% in 2006 for LDC)
     and will need to be monitored.
  • Questions on preschool programs do not include information on number and
     length of sessions to measure against the 15 hours per week commitment.
  • It does not appear possible to distinguish whether staff with early childhood
     qualifications is four year trained.
  • Further work to explore the timing, definitions and potential for correlation
     with the National Pre-school Census may be required.

Overall

It is expected the CCSDC will be repeated until an alternative can be put in place.
Something along the lines of the former AGCCCS will be needed to capture those
preschool services delivered in long day care settings.

The CCSDC will be conducted using a targeted questionnaire to minimise
administrative burden on providers by only collecting what is required to
complement CCMS data. The CCSDC is structured to minimise administrative
impost on providers at a lower cost than the AGCCCS.

Information will be made available through the RoGS process and summary reports
published to the DEEWR Internet site. Release of data will be required to meet the
same protocols as applied to the release of the previous AGCCCS particularly in
regard to unit record level and commercial-in-confidence data. More specific
requests for data for research activities will be assessed on a case-by-case basis with
approval subject to the Department’s assessment of the request.
                                           21
(vi) Child Care Benefit payment administrative data

The Child Care Benefit (CCB) is a payment offered to parents to offset the costs of
using approved child care. Additionally, CCB is offered at a minimum rate to
parents whose children use registered care. The CCB payment for approved child
care is income tested and tapers to a zero rate. Most families eligible for CCB
choose to have it paid directly to their child care service as reduced fees, but a small
proportion of families receive CCB as a lump sum at the end of the financial year.

Child care services are approved by DEEWR, and identified through a unique
identifier in the Child Care Management System (CCMS). There is a one year
process to transition all child care services to CCMS – the first services were
transitioned in February 2008. Therefore until full transition occurs there will be
two sources of CCB administrative data – CCMS for transitioned services; and
Centrelink’s Child Care Operator System (COS) for services pending transition.

The CCB payment system identifies the large proportion of the population of
children whose families claim CCB for attending Australian government approved
child care services. CCMS therefore captures information at child, family and
service provider level, for each attendance at approved child care.


Hierarchical data structure
  Children                                   Parent/guardian
      Date of birth                                        Date of birth
      Sex                                                  Sex
      State/Territory                                      Marital status
      Remoteness area                                      Australian residence
      Residential address                                  Relationship to child
      CCB paid hours of weekly care                        Labour force status
      Actual hours of approved care                        Income
      per week/day (derived from start        Service provider
      time/finish time)                                    Service type
      School attendance                                    Approved places
                                                           Centre based/home based

Strengths

   •   The CCB data warehouse will contain a complete record of attendance at
       approved child care services at the individual child level.
   •   Data can be produced at smaller geographical areas and by type of care.
   •   With the unique identifiers for parents/guardians, children and service
       providers, there is enormous potential to link these data to other rich data
       sources, and across time.

Weaknesses

   •   The collection does not obtain information specifically focussed on
       children’s involvement in preschool programs associated with long day care
       centres.

                                          22
   •   Limited information is available for children whose parents do not claim
       CCB.
   •   Access to data is restricted in accordance with legislative requirements.

Overall

This data source is far from complete. Other, supplementary and complementary
data sets would be needed to build a coherent national picture. As such, it is not
possible to primarily rely on this data set.


(vii) Longitudinal Study of Australian Children (LSAC)

The Longitudinal Study of Australian Children is the first national longitudinal
survey of children and their families and is sponsored by FaCSIA. That is, rather
than ‘point-in-time’ estimates, LSAC follows the same children and their families
‘through time’. Detailed information is collected on early child care and education
arrangements

Data are being collected over seven years from two cohorts every two years. The
first cohort of about 5000 children aged less than 12 months in 2003/4 will be
followed until they reach 6 to 7 years of age, and the second cohort comprising
about 5000 children aged 4 years in 2003/4 will be followed until they reach 10 or
11 years of age. Study informants include the child and their parents, carers and
teachers.

In the first wave of data collection, many of the older cohort were in preschool.
The main data collection for LSAC Wave 1 was from March through to November
2004.

The second wave of interviews began in late 2005 with most families being
contacted between March and November 2006.

Strengths
   • The survey has a specific focus on children and collects an extensive range
      of relevant and relatable data on children, their families and social
      circumstances.
   • Good coverage of preschool and long day care attendance and their
      overlaps.
   • The longitudinal nature of the survey makes it possible to analyse the impact
      of early experiences on later outcomes.

Weaknesses
  • The survey panel is not being updated with new cohorts of children and both
     cohorts in the survey will be beyond the preschool years when the early
     education initiative commences.
  • The sample size would support only limited state based information.

Overall



                                         23
LSAC has little to offer future performance monitoring of universal access to
preschool education. But it is nevertheless an important part of the statistical
landscape on children. The richness of the dataset and the nature of research and
analysis that can be undertaken with such data make it important data to keep in
mind.


(viii) The Household, Income and Labour Dynamics in Australia Survey
       (HILDA)

The Household, Income and Labour Dynamics in Australia Survey is a household-
based longitudinal study which began in 2001.

The wave 1 panel consisted of 7,682 households and 19,914 individuals. Interviews
are conducted annually with all adult members of each household. By Wave 4, a
little less than 80% of the original individuals had been maintained in the data set 6.
Funding has been guaranteed for eight waves.

Data releases occur in January. The release in January 2007 will be for data
collected from 2001 (wave 1) to 2005 (wave 5). Data collection has been sub-
contracted to ACNielsen, a private market research company.

Detailed information is collected about early child care and education arrangements
split according to whether the care is sought for work-related or non-work related
reasons. The longitudinal feature of HILDA relates to the households and adults in
the study. It may be possible to construct longitudinal data from the child’s
perspective, although this is not a purpose of the study.

HILDA has similar features to LSAC. However, whereas LSAC is conducted from
the child’s perspective, HILDA is conducted from the parent’s perspective. Given
its particular focus on early childhood experiences of care and education, LSAC is
the preferred longitudinal data set. HILDA is not examined any further in this
report.



(ix) Community Services Industry Survey ABS

The Community Services Industry Survey was last run in by the ABS with respect to
1999-2000 and is next scheduled in respect of 2008-09. Data would be collected
later in 2009 with results released in 2010. It is part of an annual program of
economic surveys targeting the services sector of the economy. The survey covers
all employing businesses and organisations providing community services including
child care services.

The main focus of the survey is the business operation, particularly the financial
aspects of the business. Information was also collected on the levels of activity and
services provided by these businesses/organisations. ABS may be prepared to add


6
    HILDA Survey Annual Report 2005, p10.
                                            24
questions to the next survey which could obtain more information on aspects of the
service delivered, children attending and staff qualifications


Strengths
   • An ABS economic survey focussing on child care organisations/businesses.
   • Provides interesting information on the cost structure of child care
      businesses, the sources of income and general profitability.
   • Possibility to influence some content of next survey.

Weaknesses
  • Run infrequently.
  • Survey unit is the management unit not the individual centre itself.
  • Has no current focus on preschool.

Overall

This infrequently run survey with its economic focus, could only ever be something
on the edge of the statistical landscape. However the cost information obtained is
of interest. ABS willingness to expand the content and relevance of the survey to
better suit DEEWR needs and purposes should be worth pursing.




                                        25
Data Map: from data item to data source
This section presents the data items needed to support the indicators identified in
the statement of requirement for the starting point analysis as critical (see appendix
3), which are mapped to a data source. Data sources are ranked in order of their
merit.

Indicator                                                 Source
PARTICIPATION AND EQUITY
1. Enrolment rates by jurisdiction                        RoGS; ABS CCS
2. Attendance rates by jurisdiction                       No identified source.
3. Enrolment rates by jurisdiction by service type        RoGS Table 3A.37
4. Attendance rates by jurisdiction by service type       No identified source.
5. Enrolment rates by jurisdiction at a regional level    NPC
6. Attendance rates by jurisdiction at a regional level   No identified source.
7. Rates of participation by indigenous children by       NPC
jurisdiction                                              RoGS
8. Rates of participation by indigenous children by       NPC possibly.
jurisdiction by region
9. Rates of participation by children of newly arrived    RoGS, e.g. Tab 3A.37 Non-
migrants by jurisdiction                                  English speaking background
                                                          concept
10. Rates of participation by children of newly           No identified source.
arrived migrants by jurisdiction by region
11. Rates of participation by children from low SES       RoGS, e.g. Tab 3A.37
families by jurisdiction
12. Rates of participation by children from low SES       No identified source. The
families by jurisdiction by region                        number of children (but not
                                                          those attending preschool) in
                                                          low SES regions can be
                                                          estimated from Census/SEIFA.
13. Rates of participation by children with               RoGS, e.g. Tab 3A.37
disabilities by jurisdiction
14. Rates of participation by children with               No identified source -
disabilities by jurisdiction by region
15. Number of children who currently miss out on          RoGS, Table 3A.11
ECE                                                       ABS CCS and Ryan/DEEWR
16. Number of children who currently miss out on          CCS – large states only
ECE by demographic profile
ACCESSIBILITY AND UTILITY
17. Number of early childhood education providers         RoGS, e.g. Tab 3A.36
by location by setting
18. Number of places by service type                      RoGS, e.g. Tab 3A.32
19. Waiting lists by service type                         NPC
20. Hours of operation by service type                    [ABS CCS] ; NPC
21. Availability of integrated child care and             [ABS CCS]
preschool services                                        NPC

                                          26
23. Fees charged by service type                         ABS CCS
24. Fee reduction for low income users by service        RoGS, e.g. Tab 3A.31
type
QUALITY
25. Workforce qualifications by education/teaching       RoGS, e.g. Tab 3A.35
staff by service type
26. Workforce qualifications by care staff by service    RoGS, e.g. Tab 3A.35
type
27. Child/staff ratio by service type                    RoGS, e.g. Tab 3A.33
                                                         RoGS, e.g. Tab 3A.34
28. State licensing /registration requirements by        No identified source.
service type
29. Teachers registration                                No identified source.
30. Number of workforce exemptions granted               No identified source.
31. Average hours of attendance in a preschool           RoGS, e.g. Tab 3A.32
program per week by service type
32. Average number of weeks per year a preschool         No identified source.
program is provided by service type
33. Parental participation/engagement                    No identified source.
34. Community engagement                                 No identified source.
35. Existence of specialist services (disability etc.)   No identified source.
and innovative integration models
36. Availability of vacation care                        No identified source.
37. Early intervention identification
38. Indigenous specific ECE programs                     No identified source.
39. Indigenous staff                                     NPC
40. Parental satisfaction                                RoGS, e.g. Tab 3A.38
41. Average hours of preschool attendance in per         No identified source.
week at a regional level
42. Average weeks per year a preschool program is        No identified source.
provided at a regional level
43. [Breakdown of total jurisdiction government          RoGS, e.g. Tab 3A.31
spending on preschools]

The data items here drawn from the OECECC performance indicator table on page
9 need substantially more refinement and thinking through before they will form a
useful national data set. The measures implied go way beyond currently available
outputs and as this report shows many indicators that might be considered core have
substantial limitations.

If improvements are to be made to early childhood education data it is essential that
attention be focussed and priorities set in terms of what is most important and why.
Clear links between the indicators chosen and the purpose they serve can be helpful
with this process. We draw attention to the articulation of indicators contained in
RoGS. While we are critical of various aspects of RoGS statistics the report
provides a good model in other areas.




                                           27
Issues and strategies for improving data

The new Council of Australian Governments (COAG) framework for federal
financial relations makes governments responsible and publicly accountable for the
money spent and the outcomes delivered. However it leaves it up to the individual
jurisdictions the means by which they achieve this. The statistics and associated
information needed to support this accountability are crucial to the efficacy of the
whole system. This has particular pertinence in the early childhood education area.

Problems, challenges and issues for data development
(1)    The split of policy and program responsibilities between Commonwealth
       and state/territory governments.
(2)    The different mix of early childhood provision across the jurisdictions.
(3)    The variety of settings and arrangements in which preschool education can
       take place and the imprecise boundary between child care, preschool and
       early education.
(4)    The number of programs and services in which a child and teachers may
       participate.
(5)    Different school starting ages across jurisdictions hence different definitions
       of the “year before formal schooling” and variation in the age composition
       of the preschool population.
(6)    The lack of integration between early childhood and primary education.
(7)    The wide range of terminology used by the sector to describe its activity
       including: - preschool, kindergarten, play school, early learning centre and
       early childhood education centre. Thus when data is collected from parents,
       they may not be reporting according to intended preschool definitions.




Source: Report on Government Service Provision, 2008

                                         28
As can be seen in the previous diagram there are substantial differences between the
states and territories in the arrangements for preschools. This reflects the diversity
of jurisdictional policy and practice relating to the provision of early childhood
services. But even this diagram does not do full justice to the diversity of
arrangements. For example, in recognition of the education continuum, some
preschools are a part of (or are associated with) primary schools while others are
stand-alone. Within the government sector, the role played by the parent
community in assisting with preschool management may vary.

While more traditionally, preschool programs have been delivered through separate
preschools, a growing number are delivered in association with long day care.
Within child care centres some children are taken out during the day to attend
nearby government sponsored preschools while others may attend a preschool
program within the centre. It is also possible and quite legitimate for children to be
enrolled in and attend more than one preschool program across the week.


                               Preschool arrangements




                                          Preschool
                                          Programs




                                                             Long Day
                        Preschool                              Care
                                                               Centre




        Government     Community        Private       Community       Private




       Extracted from Report on Government Service Provision, 2008 Table 3.1, P 3.9



Some facts and figures

It is valuable to look at figures drawn from a range of collections to illustrate the
types of data available to build up a picture of the preschool age target group of
children:

   •    At 30 June 2006 there were 260,000 children aged 4 years (ABS cat. no.
        3201.0).


                                           29
       •   Of the 213,000 children counted in preschool in 2006, 76% attended
           government preschools and 24% non-government preschools (NPC).

       •   Of the 420,000 children attending long day care centres in 2006, about a
           quarter were aged 4 years or more (AGCCCS).

       •   In 2006, 48% of long day care centres provided access to a preschool
           program, either in-house (34%) or through arrangement with a local
           preschool (16%) (AGCCCS).

       •   In 2005-06, 244,000 children were identified as using state/territory funded
           preschools – 89% of these children were in their year prior to full time
           school (RoGS).

   The figures above and below are drawn from available collections that are as close
   as possible to the same time point in an attempt to limit discrepancies that would
   result through time differences. (In some cases later data is available – this later
   data is not shown in the table below.)

                            Children attending preschool

                           NSW        VIC     QLD        SA      WA     TAS       NT    ACT      AUST
2006 Popn Census         114,023   78,096   58,748   18,577   27,174   4,023   2,811   4,319   307,820
2005 Child Care Survey    77,500   74,800   55,700   19,000   20,200   4,200   1,900   3,800   257,100
2006 Preschool Census     63,799   76,333   14,962   19,784   24,841   5,859   3,485   3,935   212,998
2007 ROGS                 61,080   58,397   63,710   21,120   26,291   6,165   3,467   3,327   243,557

   While the various collections from which these figures are drawn are a potentially
   rich source of information, all have their weaknesses. These are apparent in the
   inconsistencies that appear as soon as attempts are made to bring figures together –
   as for example in the table above.

   There are reasons for the differences observed in this table. In particular there are
   scope differences and specific anomalies. Some of these are well documented and
   understood while others are more perplexing. A detailed analysis and reconciliation
   to be undertaken by Boston Consulting Group as part of a Starting Point Analysis
   will shed further on this. Our purpose for including these numbers is to illustrate
   that there are issues in all collections that need to be addressed.

   The major problems arise through:

       •   Incomplete coverage (most evident in the NPC and RoGS).
       •   Inconsistent definitions of preschool across the jurisdictions (NPC and
           RoGS) or interpretations on the part of respondents (ABS Population
           Census and Child Care Survey).
       •   Potential for duplication when counting preschool attendance both within a
           collection (e.g. RoGS participation rates over 100%) or when attempting to
           add data across collections (e.g. AGCCCS to NPC or RoGS).




                                             30
Importance of common reporting
Common reporting is a means by which statistical outputs can be brought together
on a consistent basis. While common reporting is necessary to underpin the federal
financial relations framework, it is also desirable in order to inform best practice
about the nation’s early child hood education programs. The new Commonwealth
Government early childhood initiative provides both the imperative and opportunity
to enhance the existing framework of collections and improve the evidence base for
early childhood education.

Education Ministers have identified early child hood education and care as an
important strategic priority for the nation. Indeed “early childhood reform is seen
as essential for Australia’s future prosperity” (Communiqué: Joint
MCEETYA/MCVTE meeting – 17 April 2008). The priority given by Australian
Governments to ECE underlines the need for common reporting.

 Proportion of children attending State and Territory government funded
 and/or provided preschool services in the year immediately before the
 commencement of full time schooling a, b, c, d, e, f




 a The figure shows the proportion of 4 year old children in preschool services (a proxy for ‘year before fulltime
 school’) using data collected from State and Territory enrolment figures. The enrolment figures are divided by
 the number of 4 year olds in each jurisdiction, using ABS population projections. The two datasets are
 estimated at different times of the year, and are up to six months out of sequence with each other. Some non-
 4 year olds may also be included in the enrolment figures. b There is some double counting of children in
 NSW, Qld, WA and NT because some children moved in and out of the preschool system throughout the year
 and some children accessed more than one sessional program. As a result, the number of children reported in
 preschool exceeds the number of children in the target population. There is no double counting for Victoria,
 SA, Tasmania and the ACT because a snapshot is used for each year’s data collection. c NSW data for 2006-
 07 include for the first time preschools managed by the NSW Department of Education. NSW data do not
 include the non-government school sector in any of the years. The count for preschool attendance includes
 children aged from 4 to 5 years, 11 months attending child care services. d Victorian data include some
 children attending funded preschool services conducted in centre-based long day care centres and
 independent schools. e WA data for 2002-03 exclude the non-government sector. f Data for SA include all


Source: Report on Government Services 2008

First and foremost we need to acknowledge and accept that we do not have
common reporting. The diagram above (Figure 3.5 from the 2008 Report on
Government Services) illustrates this point. Most of the figures are not comparable
and the presence of percentages over 100% highlights problems with data quality.




                                                        31
Achieving common reporting

The new framework for federal financial relations, stipulates that working groups
should develop performance indicators that are meaningful, understandable, timely,
comparable and accurate. Common reporting has as its goal the production of such
statistics. It would be wrong, however, to suggest that common reporting is a totally
new goal. On the contrary, common reporting across a wide range of functions has
been the aspiration of the COAG Report on Government Service activities for over
a decade. Given this a number of questions are relevant:

   •   What has been achieved to date in the early childhood area?
   •   What are the impediments to achieving common reporting?
   •   Where has common reporting been successful and what are some of the
       critical success factors?

Progress to date

The first RoGS in 1995 did not include a Children’s Services chapter as measures
were not sufficiently developed. This chapter first appeared in 1997 and was 20
pages in length. By 2008 the chapter had grown to 80 pages. More performance
measures, together with more descriptive and useful contextual material have been
added.

Perhaps the most significant expansion has been the number and detail of the
footnotes. Many point to the lack of comparability of statistics across the
jurisdictions. This is not to suggest that data has become less rather than more
comparable over time. Rather, the footnotes reflect an increased understanding on
the part of the jurisdictions of the features and characteristics of their data.

Much has therefore been achieved, despite not having attained common reporting.

Impediments to common reporting through RoGS

The diverse arrangements for delivering preschool education across the country
make the compilation of statistics that describe and monitor the sector a very
complex task.

The absence of standard definitions so each jurisdiction reports according to its own
definition and differences are handled through footnotes. This can affect the most
important indicators such as preschool participation rates.

There are differences between jurisdictions in the extent to which preschools
operating within non-government schools are reflected in the figures.

In the early childhood area, unlike many other fields (eg health, schools and
vocational education), RoGS is not able to draw on a national collection which
operates within a performance reporting framework specifically requiring that the
statistics and indicators meet a national quality standard before publication. This
places RoGS at a serious disadvantage.



                                         32
The interpretation of a “government service” has been quite narrowly based
although varying across jurisdictions. Thus some report only those services which
are government managed or run while others include services which are
government funded. It is worth highlighting that the RoGS estimates for
preschoolers conceptually covers “State and Territory Government funded and/or
provided preschool services”. Note that this definition mixes the funding role and
provider roles of State and Territory Governments in preschools. The table below
attempts to draw out all of the possible categories of funded preschool provider that
the RoGS definition needs to be reconciled with. The ‘ticks’ in the table below
explicitly identify the different categories of State-government funded and/or
provided preschools that would seem to be captured by the RoGS definition. The
‘crosses’ are not covered by RoGS.
                                                           Preschool Provider
                                           Government Community Long Day     Private 
                                                                  Care    (Independent 
                                                                            & Catholic)
State &                Fully funded

Territory             Partly funded

Funding                 Not funded


Some of the boxes in the table are in the same category as ‘unicorns’. For example
“unfunded Government preschools” and “fully funded preschoolers in non-
government schools”.
    •    However, because some of the ‘unicorns’ may emerge as vehicles for
         preschool delivery, it is useful to identify now how we think they would be
         handled by RoGS.
As seen in the above table, there are further complications in whether government is
limited to the Commonwealth or includes state/territory governments7.

Finally it is also possible to view a government service as encompassing the setting
of standards and associated licensing.

Where is common reporting successful?

Common reporting is most easily achieved when a single agency has responsibility
for a national collection. In its simplest form this would involve the agency dealing
directly with individual collection units. But it could also involve collecting data
through an intermediate source (such as the jurisdiction) from the administrative
systems in place to deliver services to people (such as health, education, justice,
community services). These are the areas to which we need to look to for models.
Positive examples can be seen with health, school education, and vocational
education and training.

The National Schools Statistics Collection (NSSC) is an example of a successful
model in this area, although some issues still impact on the usefulness of this
collection. The responsible agency is the ABS in conjunction with education
7
  The RoGS preschooler estimates are based on the fact that preschools have been a state/territory
responsibility.
                                                 33
departments of the various jurisdictions. Proposed changes to the NSSC by the ABS
to compile data at unit record level will further increase the utility of this collection.

Another effective model is the Australian Vocational Education and Training
Management Information Statistical Standard (AVETMISS). The AVETMISS
offers a nationally consistent standard for the collection, analysis and reporting of
vocational education and training (VET) information throughout Australia. These
standards have been developed and refined over a number of years through
consultations with jurisdictions, and are the basis for reporting key performance
measures for the VET sector.

In many ways, the challenges affecting the VET system in Australia parallel those
affecting early childhood services. VET activity may take place through a complex
array of providers, including providers which are completely privately funded, and
those whose main activity is something other than the provision of VET (such as
higher education providers). The AVETMISS might therefore be an appropriate
vehicle to form the basis for overhauling standards, definitions and data
management in the area of early childhood.

A framework for common reporting

We can identify certain features and arrangements which underpin successful
examples of common reporting? These include:

       (4) A resolve on the part of some auspicing entity (e.g. COAG) that
           common reporting is an important goal that is inextricably linked to
           efficient and effective service delivery.
       (5) A corresponding commitment of the key players across the jurisdictions
           to make it happen.
       (6) Adequate priority, resources, effort, attention and time to undertake the
           necessary development work.
       (7) A willingness to change existing systems and approaches to meet
           agreed statistical goals.
       (8) Development and implementation of national uniform data standards
           that provide for consistency within a collection and allow for data to be
           integrated across collections.
       (9) Oversight or coordination by an agency with the necessary expertise and
           credibility with both the jurisdictions and the Commonwealth, and
           which takes responsibility for the statistical output and its quality.
       (10) An advisory and/or consultative group (and associated working
          groups) with representation of all stakeholders to work with the
          coordinating agency
Under this framework, one agency has responsibility and is held accountable for
assembling and maintaining the data set to agreed quality standards. This agency
works within a consultative framework that ensures that the key stakeholders take
an active role in identifying and solving data quality problems and issues. Data is
available not only for summary reports but is also accessible for supplementary
analysis.

                                           34
Future directions for early childhood education data

The foregoing discussion of limitations of the existing statistics and performance
measures points to the importance of setting up the arrangements for generating the
necessary statistics at the outset when program delivery is being formulated. This
has both plusses and minuses. On the positive side decisions can be taken that
ensure that needs for data will not be overlooked. On the negative side, as the final
shape of the program is still to be determined, administrative systems that might
give rise to valuable data are yet to be defined.

The policy directions endorsed by Ministers emphasise the need for reform in the
area of early childhood. This implies changes in the shape or form in which these
services are constituted. If these changes are towards a more integrated and
convergent approach then better statistics are much more likely.

A vision for the longer term

The changed performance reporting requirements in accordance with the new
framework for federal financial relations provide the opportunity to build on
existing RoGS processes to provide the national uniform early childhood statistics
needed to properly monitor performance and evaluate practice. Corresponding
changes are needed both on the part of the jurisdictions and the RoGS reporting
process if common reporting is to be seriously pursued. Most important of these are
expanding the collections so they provide a more comprehensive picture of
preschool provision, adopting common definitions and removing double counting.

A number of collections could be considered as the basis for expansion and
improvement to achieve common ECE reporting in the short or longer term. These
are the National Preschool Census (NPC), the Australian Government Census of
Child Care Services and the National Schools Statistics Collection (NSSC).

In the NPC, data on government preschools are compiled from state governments
using their existing census arrangements. Data on non-government institutions and
organisations offering a preschool program are obtained by DEEWR through the
Non-Government Supplementary Census, on contract. The NPC is currently the
only collection which obtains any data on non-government preschool programs,
although this information is compiled from state/territory data holdings and is
unlikely to be comprehensive at this stage.

The AGCCCS was previously managed by FaCSIA (now part of OECECC) which
has played a central role in collecting, compiling, analysing and disseminating the
data. While some data requirements pertinent to the new universal access
commitment (particularly relating to preschool hours and the identification of 4 year
training early childhood teachers) were not previously included in the former
AGCCCS, or the CCSDC, their inclusion in a future collection would be a logical
extension. Certainly expanding the collection to encompass preschools would be a
major change. However, it is consistent with the moves towards integration of the
child care and preschool sectors. Furthermore it would capitalise on the statistical
strengths underpinning this collection.



                                         35
The NSSC is managed and published by the ABS from statistics on schools
compiled from collections conducted in cooperation with the Ministerial Council on
Education, Employment, Training and Youth Affairs (MCEETYA), by the state and
territory Departments of Education (government series), and by the Australian
Government Department of Education, Employment and Workplace Relations
(DEEWR) (non-government series). Information on early childhood education
providers has the potential to be included as part of this collection although it could
be several years before this change is fully implemented. A major advantage of
building on the NSSC would be the ability to analyse the transition to school and to
manage statistically the preschool to primary school interface in the context of
different school starting ages across the jurisdictions.

An expanded role for the NSSC or NPC would draw on the administrative systems
of the jurisdictions that currently feed into RoGS but would need to go further to
ensure comprehensive coverage of the preschool sector, including both preschool
programs not provided by governments (either Commonwealth or state/territory),
and those not in receipt of government funding. The statistics would be presented in
such a way that they could be brought together with statistics compiled on
preschool activity occurring within a child care setting, while still enabling
providers to be uniquely identified and counted.

The statistics would be compiled according to national data standards developed
and agreed to by the jurisdictions under the RoGS reporting framework. This would
be a major change from current arrangements, with jurisdictions committing to
change their administrative data collections to align with the agreed standards,
classifications and definitions. We recognise that this would involve extra effort on
the part of jurisdictions but see this as consistent with the heightened priority now
afforded to early childhood education by the Australian Government.

While an expanded role for the NPC, AGCCCS or NSSC may be suggested in the
short or longer term, a number of collections are, or have been, important in
providing data on early childhood education, including
   • ABS Childhood Education and Care Survey (CEACS) (3 yearly).
   • Child Care Supplementary Data Collection (CCSDC) (interim collection
        only at this stage).
   • Child Care Management System (CCMS) (annual).
   • ABS Population Census (5 yearly).
   • ABS Resident Population Estimates (annual).
   • ABS Community Services Industry Survey (CSIS) (irregular).

Some of these data sets are updated annually, while others are less frequent but, as
is currently the case, retain their place under the spotlight as the latest available
information.



                 Participation &
                                     Providers &
                      non-                           Workforce       Costs & funding
                                      programs
                  participation

Currently      NPC                 AGCCCS            NPC             AGCCCS
available      CCCS                CCS               AGCCCS          Centrelink

                                          36
sources         CCS                 ROGS              ROGS             ROGS
                ROGS
                Population
                Census
Planned &       CCSDC               CCSDC             CCSDC            CCSDC
future
sources         CCMS/Data           CCMS/Data         CCMS/Data        CCMS/Data
                warehouse           warehouse         warehouse        warehouse
                NSSC
                CEACS               CSIS              CSIS             CSIS
Possible        NECDC               NECDC             NECDC            NECDC
longer term     Enhanced            Enhanced          Enhanced         Enhanced
sources         ROGS                ROGS              ROGS             ROGS
                CEACS               NSSC              NSSC             NSSC
                NSSC

In the table above, an illustrative collection, the National Early Childhood Data
Collection (NECDC) is used to denote a future source that would embody the
features of common reporting described in the earlier aspirational statement. It
would correspond to a revamped and expanded NPC, AGCCCS or NSSC.

In the short to medium term

National data standards are an essential foundation for common reporting. A
working group should be established to review existing definitions etc and develop
this draft standard. This can build on the existing RoGS experience. A key input to
this work should be the Children’s Services National Minimum Data Set
(CSNMDS) which has been developed by the AIHW, under the guidance of the
Children’s Services Data Working Group (CSDWG). The CSNMDS provides a
framework for collecting a set of nationally comparable data for child care and
preschool services.

In the light of the Commonwealth’s new election commitments, however, a
thorough review of the CSNMDS in relation to performance information
requirements is warranted. The Performance Information Management Group
(PIMG) was established to coordinate the provision of data for this purpose and to
set in place the arrangements for data standards to support the development of
performance information. Governance arrangements need to be established between
the PIMG and other committees as to their respective responsibilities for data
standards into the future.

Collect administrative data – The “universal” access element of the new ECE
initiative brings with it a requirement to collect data on the totality of the preschool
sector regardless of the nature of the service delivery. Formation of a
comprehensive list (or register) of all licensed services provides the logical start
point for an administrative collection. Such lists exist nationally in the child care
sector as they are triggered by the CCB requirements. Preschool licensing is a
state/territory function which could form the basis for the generation of
comprehensive lists.

Explore linked data - In the longer term there is value in designing a life cycle
approach to education and to monitor student outcomes and progress. From the time

                                           37
of birth the activities of children are charted in an array of administrative systems
often linked to the needs and circumstances of their families. Following their
official registration and eligibility for the baby bonus and family tax benefit (from
Centrelink), through interaction with the health system (Medicare and HIC),
attendance at child care and preschool and eventually through school, children exist
in various administrative systems.

On the face of it, it would make sense to use such systems to track children
participating in preschool programs, children participating in child care with no
preschool programs, as well as to track those who do not interact at all with formal
settings before school. Tracking would be facilitated with the use of a unique child
identifier which would greatly reduce the prospect of confusing multiple enrolments
(which have the effect of inflating measures of ECE enrolment) with single
enrolments. The unknown extent of multiple enrolments is a major source of
uncertainty in much of the RoGS data – even in those jurisdictions which conduct a
‘snapshot’ of their preschool enrolments. Unique child identifiers could then be
linked to unique student identifiers which are already a feature of most (perhaps all)
government school systems around the country. This would also facilitate the
collection of information on the transitions occurring prior to school, as well as to
monitor key outcomes once students enter based on literacy and numeracy
performance testing.

Linking this data in accordance with appropriate privacy and access protocols has
the potential to unlock a valuable integrated information system in a far more cost
effective way than commissioning separate surveys. Feasibility studies to scope and
explore this further would be valuable, particularly to evaluate the use of unique
identification of providers, as well as those participating in early childhood
programs.

The National Preschool Census conducted annually by DEEWR, has
comprehensive coverage of the preschool sector and currently provides a more
complete picture than jurisdictional administrative systems. It also has a particular
focus on Indigenous children and so is important for this reason alone. Further
thinking is needed to determine how data from this collection should be utilised for
universal access performance monitoring in the longer term. In the short term, data
quality analysis which involves confrontation of 2006 NPC data with other sources
would provide valuable insight into establishing the quality of the NPC, particularly
in obtaining data on children in private preschool programs, and in assessing access
to education of Indigenous children in remote areas.

The ABS has already indicated its interest in extending the National Schools
Statistics Collection (NSSC) to collect preschool data, although this option is likely
only to be realised in the medium to longer term. What is not clear is whether this
could encompass the range of data needed for effective performance monitoring of
early childhood education. It would be valuable to engage with ABS to establish
whether in the longer term this information could be made available.

Data on Indigenous children and those in remote areas - Given the importance
that both the Australian Government and COAG place on outcomes for Indigenous
children, key measures need to be presented, where possible, by Indigenous status.
Currently, however, data from most sources do not support disaggregations for

                                          38
Indigenous and non-Indigenous sub-populations although measures for these
distinct groups would be considered ideal. There are also specific issues affecting
data on remoteness areas, and there is more than one classification on remoteness in
use across collections.

A possible option to draw together a number of these elements in the shorter term
would be to establish a National Early Childhood Data Collection (NECDC), as in
Figure 1.




The NECDC could draw on the model provided by the National Schools Statistics
Collection (NSSC)8 and would be a collaborative arrangement between state,
territory and Australian Government education and child care authorities and a
collection/compiling agency such as the Australian Bureau of Statistics (ABS)
and/or the Australian Institute of Health and Welfare (AIHW).

The ABS and/or the AIHW would provide statistical oversight, guidance and
independence for the collection. The NECDC would apply a set of concepts,
definitions and classifications developed jointly by these agencies and agreed to by
a steering committee.

In this model, the States and Territories would continue to compile data on
preschools that they provide and/or fund. However, there would need to be a
supplementary collection of non-government providers. Possibly, this may involve
DEEWR integrating the National Preschool Census (NPC) and the Child Care
Management System (CCMS)9 with a view to filling in the “non-government” gaps

8
  However the specific vehicle would correspond to a revamped and expanded National Preschool
Census (NPC), Australian Government Census of Child Care Services (AGCCCS) or NSSC.
9
  Although the CCMS currently has data gaps on child care providers with preschool programs.
                                              39
not covered by state and territory administration data. The collections that form the
NECDC would aim to be conducted as close as possible to the schools census (1st
Friday in August).

The steering committee would also develop protocols for eliminating double
counting of providers and/or preschoolers – perhaps through the use of unique
identifiers. The NECDC could provide the basis for RoGS preschool reporting and
may also have other uses, for example, to supplement the annual ABS ‘Schools’
publication.


Next Steps
Further refinement of the performance monitoring requirements will greatly sharpen
the focus for deciding on the best options for data collection.

Priorities for data quality improvement can then be established having regard to
relative importance of different indicators.

Data quality studies including the Starting Point Analysis being undertaken by
Boston Consulting Group will provide further insight into the costs and benefits of
addressing particular collection deficiencies.




                                         40
Appendix 1: Identified Data Deficiencies
We have identified the following preliminary list of data deficiencies. The list is
not exhaustive as data issues will continue to arise depending on what questions the
steering group thinks may need to be answered. The issues below will need to be
resolved at some point to accommodate the development and implementation of the
Australian Government’s ECE commitments. The items are listed in rough order
of their judged significance in terms of developing ECE policy.


The size of the ‘gap group’
The ‘gap group’ are the children in the year preceding formal schooling who
receive no preschool whatsoever.

This may be inferred from the proportion of children who receive preschool in the
year prior to formal school. So to take an example, NSW reported 54,181 children
in preschool in 2006-07 in the year prior to formal school in 2006-07 (RoGS Table
3A.
11). As this is 64.6% of the estimated residential population of NSW four year olds
this might suggest a ‘gap group’ of 100% - 64.6% = 36.4%.

However there are a large number of qualifications to the 64.6% estimate. Some of
the qualifications (as identified in footnotes to the RoGS tables) to the 64.6%
estimate are listed below. Similar concerns arise with the other jurisdictions; we
focus on NSW because it is the largest.

“Data for estimated residential population are six months out of sequence with the
data for children using State or Territory government funded and/or provided
preschool services in year before full time school.” [Table 3A.11]

“There is some double counting of children in NSW, Qld, WA and NT because
some children moved in and out of the preschool system throughout the year and
some children accessed more than one sessional program.” [Table 3A.11]

“In 2006-07, data includes information from both [NSW Departments of] DoCS
and DET. DoCS count of children attending preschool the year prior to full time
schooling is the count of children 4 to 5 years, 11 months, attending childcare
services.” [Table 3A.33]

“Preschool programs are provided for children who are both in designated
preschool services and other childcare services above the age of 4 years. DET
counts all children in Preschools except younger children (< 5yrs old) in Wilcannia
Centra, John Brotchie and Moama PS.” [Table 3A.33]

In addition, we understand that Australian-government funded child care centres are
required to provide preschool programs in NSW. This is not included in the 64.6%
estimate.



                                         41
Finally, there are an additional 81 private preschool providers in NSW (Table
3A.36) for which no other data appears to be reported in RoGS.

This issue was examined in some detail in work that was commissioned by DEST
in 2006. That work extracted estimates of the number of children in the ‘gap group’
from the ABS Child Care Survey (ABS cat. 4402.0) and could be updated
following the release of the next Child Care Survey [in 2009]. There are two
concerns related to this work: (1) the method for extracting estimates of the gap
group are less reliable for the smaller states (because of their smaller sample size);
and (2) not everyone will follow its fairly sophisticated methodology


Hourly costs of preschool

Estimates of hourly costs of preschool are critical for analysing ECE policy options.

Preschool costs are shared between the Australian Government (presently minimal,
although some child care subsidies may be used to fund preschool programmes),
State and Territory Governments (mainly preschool subsidies), local government
(which often provide buildings for ‘free’) and parents (‘out-of-pocket’ costs).

There is little to base estimates of the Australian Government and local
government’s contributions to preschool.

Average hourly private preschool costs by jurisdiction are identified in the ABS
Child Care Survey and presented at RoGS Table 3A.28 for 2004-05.

Average hourly State government costs may be inferred to an extent from data on
total preschool attendance and total state funding of preschools. The total number
of hours of preschool attendance further requires:
A: estimates of the average hours per week that children attend preschool; and
B: the number of weeks per year that preschoolers attend.

States, other than NSW, provide estimates for A in RoGS jurisdictional tables. We
note that Victoria’s estimate was compiled from a one-off research project.

States do not presently provide an estimate of B.

Reporting of A and B is central to delivery of the Australian Government’s
commitments to providing 15 hours of preschool over 40 weeks.


Four-year qualified preschool teachers

In meeting the Australian Government’s commitment that preschool be provided by
teachers with a relevant four-year degree, more comprehensive information will be
required as to the qualifications of preschool teachers and the proportion of
preschool students that they are responsible for.

Presently, jurisdictions report to RoGS their preschool primary contact staff
numbers “with a relevant formal qualification”. To meet the Australian
                                         42
Government’s commitment, “relevant formal qualification” should be altered to
“relevant four year formal qualification” so as to avoid any ambiguity.

We don’t know what fraction of children receive their preschool from these
qualified teachers – that is another aspect that would need to be considered.

In terms of costing the Australian Government’s commitment, it might be useful to
know the wage premium that four year teachers attract compared to three year
teachers. This data might be compiled from DEEWR’s internal data sources.


Disadvantaged groups

This relates directly to the ‘gap group’ issue. Given our difficulties in measuring,
even in the aggregate, the size of this group, we cannot be optimistic of there being
data that provides the family context of the ‘gap group’. If the ‘gap group’ were
from high SES backgrounds, we would be more relaxed. But there are clues in the
ABS child care survey that the ‘gap group’ – if we were able to identify them –
would most likely come from the ranks of the less well off. The chart below shows
how children in neither preschool nor Long Day Care (where they might access
preschool) are over-represented amongst low income families.

             Figure 1: The ‘gap group’ are over-represented in low income families
            100%



            90%



 Share of   80%

 children
 3-5        70%
 years
 old not
            60%                                       Neither LDC nor preschool
 in                                                   Long Day Care only
 school                                               Preschool (total)
            50%



            40%



            30%                                                                                Source: ABS child
                                                                                               care survey (2005)
                                                                                               (unpublished data)

            20%



            10%
                                                           Parental Income

             0%
                   Less than $400 pw   $400-$799 pw          $800-$1,199 pw       $1,200-$1,999 pw       $2,000 or more




This conclusion is reinforced for the three larger states (for which the data in the
ABS child care survey is relatively robust).




                                                      43
        Figure 2: The gap group are over-represented in low income families (NSW)

                               100%



                               90%



                    Share of   80%

                    children
                    3-5        70%
                    years
                    old not
                               60%                                              Neither LD C nor preschool
                    in                                                          Lo ng D ay Care only
                    school                                                      Preschool (to tal)
                               50%



                               40%



                               30%                                                                                         Source: ABS child
                                                                                                                           care survey (2005)
                                                                                                                           (unpublis hed data)

                               20%



                               10%
                                                                                    Parental Income

                                   0%
                                           Less th an $400 pw    $400-$799 pw           $800-$1,199 pw       $1,200-$1,999 pw         $2,000 or mo re



      Figure 3: The gap group are over-represented in low income families (Victoria)

                            100%



                            90%



                 Share of   80%

                 children
                 3-5        70%
                 years
                 old not
                            60%                                                 Neither LDC no r preschool
                 in                                                             Lo ng Day Care on ly
                 school                                                         Prescho ol (total)
                            50%



                            40%



                            30%                                                                                                 Source: ABS child
                                                                                                                                care survey (2005)
                                                                                                                                (unpublished data)

                            20%



                            10%
                                                                                    Parental Income

                             0%
                                        Less th an $400 pw      $400-$799 pw            $800-$1,199 pw        $1,200-$1,999 pw            $2,000 or mo re



     Figure 4: The gap group are over-represented in low income families (Queensland)

                            100%



                            90%



                 Share of   80%

                 children
                 3-5        70%
                 years
                 old not
                            60%
                 in
                                                                        N either LDC no r preschool
                 school
                                                                        L on g Day Care only
                            50%
                                                                        Preschool (total)


                            40%



                            30%                                                                                                 Source: ABS child
                                                                                                                                care survey (2005)
                                                                                                                                (unpublished data)

                            20%



                            10%
                                                                                    Parental Income

                             0%
                                        Less th an $400 pw      $400-$799 pw            $800-$1,199 pw        $1,200-$1,999 pw            $2,000 or mo re




The gap group children are more likely to have sole parents than coupled parents –
as shown in the next two charts.




                                                                                  44
.
                             Figure 5 Gap group by parental income – couples.
               100%



               90%



    Share of   80%

    children
    3-5        70%
    years
    old not
               60%
    in
                                                         Neither LDC nor preschool
    school
                                                         Long Day Care only
               50%
                                                         Preschool (total)


               40%



               30%                                                                                   Source: ABS child
                                                                                                     care survey (2005)
                                                                                                     (unpublished data)

               20%



               10%
                                                                     Couple parent income

                0%
                      Less than $400 pw          $400-$799 pw          $800-$1,199 pw   $1,200-$1,999 pw         $2,000 or more


                         Figure 6: Gap group by parental income – sole parents.
               100%



               90%



    Share of   80%

    children
    3-5        70%
    years
    old not
               60%
    in
                                                         Neither LDC nor preschool
    school
                                                         Long Day Care only
               50%
                                                         Preschool (total)


               40%



               30%                                                                                   Source: ABS child
                                                                                                     care survey (2005)
                                                                                                     (unpublished data)

               20%

                                                                 Sole parent income
               10%



                0%
                             Less than $400 pw                          $400-$799 pw                       $800 or more




                                                                45
Appendix 2: Project Description ‘Improving the quality of
    data on early childhood education’
C.2 Overview: There are many sources of survey and administrative data relating to
    early childhood education, but not all data are readily available, timely,
    comprehensive or of good quality, and comparisons across jurisdictions for key
    sub-groups in the population are difficult. The current project is to analyse
    information requirements and data sources, and to advise the Department on
    what needs to be done to improve data collection and capacity.
C.3 Project objectives: The objectives of the project are to:
     •   analyse information requirements in relation to the universal access
         commitments as outlined above, and map these to sources of data, both
         existing and potential
     •   describe and analyse the advantages, accuracy and limitations of various
         sources of data on early childhood education and related activity, in the light
         of information required for potential performance indicators
     •   consider options for shorter and longer term data improvements, and
         preferred data sources for achieving possible improvements to data collection
         and reporting. This will include consideration of whether to develop a
         centralised data set or agreement to a common reporting standard.
C.4 Information requirements: Major information requirements include, but are
    not limited to:
     •   participation in, and access to, early childhood education programs in the year
         before full-time schooling
     •   non-participation in early childhood education programs, and factors affecting
         non-participation
     •   the nature and characteristics of early childhood education program provision
         (including the hours and patterns of enrolment and/or attendance in
         preschool programs, and the types and characteristics of preschool program
         providers)
     •   the characteristics of early childhood education program delivery (including
         the qualifications of staff delivering preschool programs across all settings)
     •   the issues related to costs and/or funding of early childhood education
         programs, particularly what funding is needed to support delivery of the
         universal access commitment
     •   the socio-demographic and contextual factors affecting the above (including
         Indigenous status and spatial/regional characteristics).
     In order to define performance information requirements, the Consultant will
     work with the Department to map these requirements against data sources, in
     order to suggest improvements to data according to the following four broad
     categories:
     • Participation and Equity
     • Accessibility and Utility
     • Quality; Quantity.

                                           46
Appendix 3: Template for ‘starting point’ analysis
There are three elements to the starting point analysis: a contextual picture of the
provision of early childhood education and related early childhood services in each
State or Territory; critical information; and optional information. For both the
critical and optional information, sources would need to be referenced (i.e.
administrative data, department; survey data). Data mapping will take into account
information already collected through the Early Childhood Development Subgroup
working parties and existing data sources.

A contextual picture

States and Territories will be asked to provide an overview of how early childhood
education is delivered in their jurisdiction, covering both the traditional preschool
sector and early childhood education delivered in child care. The type of contextual
information that would be very useful is outlined in the following table. States and
Territories would be encouraged to provide any further information that they
consider relevant in building a picture of provision in their state – including
identifying specific constraints.

Governance                A broad description of state governance and regulatory
                          arrangements for early childhood education, for example:
                          • Portfolio responsibilities for early childhood education and
                              child care, including around quality and legislation
                          • Representation on Ministerial Councils
                          • Interagency governance arrangements
Quality Systems           • Quality systems and regulation for early learning programs
                              in different settings
                          • Early learning curriculum frameworks against evidence
                              based best practice in quality
Expenditure10             • A breakdown of actual or estimated costs/jurisdiction’s
                              expenditure on early childhood education e.g. per child, by
                              disadvantage e.g. children with disabilities, Indigenous, by
                              location, in different settings (e.g. LDC compared to
                              preschools), subsidies to parents/providers
                          • Non-government expenditure (if available)
Public/Private            • the predominant model of delivery/access to ECE programs
Mix                           in either ‘preschool’ or long day care
                          • the share of public, community not-for-profit, private-for-
                              profit (small and corporate), catholic and independent
                              schools of ECE across the preschool and long day care
                              sectors
Integration/              • extent of integration that currently occurs by location (or in
innovative                    development) between preschool and child care, including
models                        co-location, full integration, cooperative arrangements
                              between services
                          • integrated services on school sites including government
                              and non government schools
10
     Noting this may entail preliminary mapping to inform future work.
                                                  47
                       •   different kinds of provision according to local needs and
                           circumstances
                       •   innovative approaches that currently work well, especially
                           for hard to reach groups, remote communities etc (case
                           studies if time allows)
Workforce              •   early childhood education and care workforce supply issues
                           by location, sector
Planning               •   planning processes for service delivery – including a range
                           of early childhood and parent support services



Critical information

The following table outlines the information that will be sought from each
jurisdiction, recognising that data will be patchy and incomplete in some areas, and
will come from a mix of information sources (noting that one of the objectives of
this gap analysis is to identify and recommend core performance indicators based
on information that is currently available). It is grouped across four broad data
categories: participation and equity, accessibility and utility, quality, and quantity.
As there will be regional differences within a state for these variables, it would be
ideal if some disaggregation beyond a whole-of-state level is possible (for instance
providing data by LGA/SLA or according to metro, non-metro, remote etc
classifications).

Category        Quantitative information
                Set against the population, demographic and geographic profile of each state and
                territory e.g. against ABS data/estimates

Participation   • Enrolment and attendance rates in early childhood education across the
and Equity        jurisdiction
                    o against eligible population of children
                    o by service type
                    o at a regional level

                • Rates of participation by children from disadvantaged groups (Indigenous;
                  children of newly arrived migrants; low SES; children with disabilities),
                  including at a regional level (where available).
                • Number of children who currently miss out on early childhood education by
                  demographic profile (where available)

Accessibility   • Number of early childhood education providers by location (across preschool
and Utility       and child care settings),
                • Number of places, existence of waiting lists, hours of operation by service type
                  (where available)
                • Provision of integrated child care and preschool services
                • Broader on-site service – e.g. maternal and child health, parent support
                • Fees charged by service type and fee reduction for low income users



                                           48
Category        Quantitative information
                Set against the population, demographic and geographic profile of each state and
                territory e.g. against ABS data/estimates

Quality         • Workforce qualifications (care staff; education/teaching staff) by service type
                • Child/staff ratio by service type
                • State licensing/registration requirements by service type
                • Teachers registration (where applicable)
                • Number of workforce exemptions granted (where applicable)

Quantity        • Average hours of attendance in a preschool program per week by service type
                • Average number of weeks per year a preschool program is provided by service
                  type



Optional information

The suggestions presented below (again according to the six categories) represent
information that would be very useful, but that may not be available in each state.
Ideally, this information would be provided where it was available, recognising that
it may need to come from qualitative, rather quantitative, sources.

Outcomes        Broad Indicators
Participation   • parental participation/engagement, community engagement
and
Equity
Accessibility   • existence of specialist services (disability etc) and innovative integration models
and Utility     • availability of vacation care
                • use of services as an opportunity to identify children / families needing early
                  intervention

Quality         • whether many services offer Indigenous specific programs / employ Indigenous
                  staff
                • any indications of parental satisfaction

Quantity        • average hours of attendance in a preschool program per week at a regional level
                • average number of weeks per year a preschool program is provided, a regional
                  level




                                          49
Appendix 4: Census measures of disadvantage
DEEWR commissioned a special cut of the 2006 census which classified the
population of 0-4 year olds by Census District and by their:
   • Indigenous status (Indigenous, Non-Indigenous and Unstated);
   • Remoteness category (major city, inner regional, outer regional, remote and
       very remote); and
   • Whether they were in the bottom 20% of Census Districts by a broad
       measure of social disadvantage (the Socio-Economic Index for Areas
       (SEIFA) measure of disadvantage) or the top 80%.

Table 1 summarises this data. ‘SEIFA Low’ refers to the children in Census
Districts that were ranked in the bottom 20% of areas in terms of disadvantage.
‘SEIFA High’ refers to the top 80%. ‘Unstated’ refers to children whose
Indigenous/Non-Indigenous status is unknown.

            Table 1 Distribution of 0-4 year old population (2006, exc. o/s territories
                      Indigenous          Unstated              Non-Indigenous             Total
                      SEIFA Low SEIFA High SEIFA Low SEIFA High SEIFA Low SEIFA High
Aust Major Cities           6,714  11,309      12,003    40,091    118,200     669,171      857,488
Aust Inner Regional         6,194   6,289       3,245     8,595     45,798     170,606      240,727
Aust Outer Regional         6,653   5,303       2,396     4,914     26,738      75,131      121,135
Aust. Remote                2,891   1,543         596     1,308       3,182     13,012       22,532
Aust Very Remote            6,420     404         297       362       1,370      3,751       12,604
Australia                  28,872  24,848      18,537    55,270    195,288     931,671    1,254,486


Table 2 presents the information in Table 1 as a share of the total 0-4 year old
population. For example, the proportion of the total population of 0-4 year olds
who live in Australian major cities AND who are Indigenous AND who sit in the
bottom 20% of the SEIFA distribution was 0.5%.

Table 2: Distribution of 0-4 year old population (2006, excl. o/s territories).
                      Indigenous          Unstated              Non-Indigenous             Total
                      SEIFA Low SEIFA High SEIFA Low SEIFA High SEIFA Low SEIFA High
Aust Major Cities            0.5%    0.9%        1.0%      3.2%        9.4%    53.3%         68.4%
Aust Inner Regional          0.5%    0.5%        0.3%      0.7%        3.7%    13.6%         19.2%
Aust Outer Regional          0.5%    0.4%        0.2%      0.4%        2.1%     6.0%          9.7%
Aust. Remote                 0.2%    0.1%        0.0%      0.1%        0.3%     1.0%          1.8%
Aust Very Remote             0.5%    0.0%        0.0%      0.0%        0.1%     0.3%          1.0%
Australia                    2.3%    2.0%        1.5%      4.4%       15.6%    74.3%        100.0%




Note that:
   • There is a significant pool of Indigenous children (2% of the total
       population) who do not live in the most disadvantaged areas. Nearly half of
       these (0.9%) are in major cities.
   • Of the remote and very-remote children, a significant group of Non-
       Indigenous children are from the higher SEIFA group.
   • Indigenous status is less well defined in more urban areas – the ratio of
       ‘Unstated’ to ‘Indigenous’ declines as locations become more remote.

It is useful to examine further the inter-relationship between disadvantage,
remoteness and Indigenous status.



                                                50
Summary measures and indicators by remoteness area

                           Major         Inner         Outer         Remote          Very      Australia
                           Cities      regional      Regional                      Remote
Area (‘000 square
kilometres)
                                14           220           803         1,021          5,646       7,704
Population (’000)        14,167.7       4,084.2        1,965.7         314.8          169.1     20,701.5
Population share
(%)
                              68.4          19.7            9.5           1.5            0.8        100
Indigenous people
(‘000)
                            164.3          108.2         113.3           49.5          81.9       517.2
Proportion
Indigenous (%)
                               1.1           2.7            5.8          15.7          48.4          2.5
Children 0-4 years
(‘000)
                            857.5          240.7         121.1           22.5          12.6      1,254.5
Proportion aged 0-4
years (%)
                               6.1           5.9            6.2           7.3            8.2         6.1
Proportion aged 0-4
years living in
disadvantaged
                              16.0          22.9          29.5           29.6          64.2         19.3
areas (1) (%)
With post school
qualification (%)
                              48.9          42.8          39.8           37.7          28.8         46.6
Median personal
income ($/week)
                              495            406           419            484           347         466
Source: ABS, 2006 Population Census
(1) Defined as the bottom 20% of areas based on the ABS SEIFA index of relative disadvantage

Remote areas of Australia account for a large proportion (86%) of the country but
are sparsely populated containing only 2.3% of the people.

Indigenous people who make up 2.5% of the population are much more likely than
the rest of the population to live in remote areas. They account for 15.7% of people
living in areas classified as remote and almost 50% of people living in very remote
areas.

The characteristics of people living in very remote parts of the country are therefore
influenced by their Indigenous population. For example incomes tend to be lower
and people are less likely to have educational qualifications beyond school.

There is a strong relationship between disadvantage, remoteness and Indigenous
status. People living in very remote areas are four times as likely, to live in a
disadvantaged area than those living in the major cities. Again this is largely driven
by the concentrations of Indigenous people.




                                              51
REMOTENESS




INDIGENOUS POPULATION DISTRIBUTION




                           52
Appendix 5: Bibliography
DEECD (2008), Victorian Government Department of Education and Early
Childhood Development, Victorian kindergarten policy, procedures and funding
criteria, 2008 update, 2008.

Australian Bureau of Statistics, 2005 Child Care Australia, ABS cat. no 4402.0.

Australian Bureau of Statistics, 2007, Measuring Learning in Australia: Concepts
and Directions in Early Childhood Learning, ABS cat. no 4232.0.

Australian Bureau of Statistics, 2007 Directory of Education and Training
Statistics, ABS cat no 1136.0

Australian Bureau of Statistics Statistical Clearing House (ABS SCH)
http://www.nss.gov.au/nss/home.NSF/84c014dd96ddf6cbca257118001dbbee/730c3
b870f9ef60cca256e8d0017fe69?OpenDocument

Australian Bureau of Statistics, Experimental Projections of the Aboriginal and
Torres Strait Islander Population 1996-2006, ABS cat. 3231.0.

Australian Bureau of Statistics, Population Projections Australia 2004 to 2101,
ABS cat.3222.0.

ACT Education and Training, Government Preschool Bulletin, February 2006
Census.

Department of Family and Community Services, 2004 Census of Child Care
Services.

Strategic Information Consultants, National Preschool Census 2005, Aboriginal
and Torres Strait Islanders and All Students, Summary Report, June 2006.

Harris L. and Ungerer J., What can the Longitudinal Study of Australian Children
tell us about infants’ and 4-5 year olds’ experiences of early childhood education
and care?, Family Matters No. 72, Summer 2005, Australian Institute of Family
Studies.

NSW Department of Community Services, NSW State Budget 2006-07.

Productivity Commission (2006), Report on Government Services, (Chapter 14,
Childrens Services).




                                         53

				
DOCUMENT INFO