Introduction: the ‘determined end’ to a 40-year decline

In 2017, one of Australia’s largest newspapers, the Sydney Morning Herald, announced the following headline: “Public schools increase share of enrolments, reversing 40 year trend”. The article goes on to assert that “public schools’ share of students has increased for the second year running nationwide, marking a determined end to a 40 year decline in government school enrolments” (see Munro 2017, my emphasis). The Educator Australia reported this as “good news for public schools and Indigenous students”, writing that Australian Bureau of Statistics (ABS) enrolment data indicate that

…the proportion of Australian children enrolled in government schools increased slightly over the past year. 2016 marks the second year in a row that public school enrolments have risen, bucking a downward trend that has troubled the sector for the last 40 years (Henebery 2017).

The newspaper reports were triggered by Australian Bureau of Statistics (ABS) data and an accompanied media release with the headline: “Government school enrolments on the rise”. An ABS spokesperson declared that Australia was experiencing a “reversal of the steady drift of students from government to non-government schools observed for much of the past 40 years” (ABS 2017). This was picked up by a broad range of online reports, magazines and newspapers, asserting that private school enrolment was “flatlining” (Koziol 2018), and “state schools more popular than private”, and “public school enrolments increased for the first time in seven years” (Daily Telegraph 2018).

The claims immediately caught my attention. Certainly, the “determined end” to a “40-year decline” of public school enrolment is a bold assertion and maintains clear implications for education policy reform. Is it really a “good news story” for public schools? Is privatisation not as far-reaching in Australia as previous research has indicated? Across OECD countries, the public school is undergoing significant reform and diversification, and research points to a consistent “increase in private school enrolment globally” (Verger et al. 2017, p. 757).

This paper will examine the counting of national school enrolment, and the publication of this count in the form of statistics, particularly as divided between ‘public’ and ‘private’ schools. For the international reader, public schools are also referred to as government schools, and are principally government-funded, secular and nominally free-of-charge to parents. Private schools are referred to as non-government schools, although they receive high levels of government funding. The non-government sector consists of the Independent and the Catholic sector, both of which charge direct fees to parents. Catholic schools maintain an emphasis on low-fees and accessibility; Independent private schools typically charge the highest tuition fees for parents, although there are exceptions. The majority of non-government schools are religiously affiliated.

National enrolment shares are often considered at face value, as reported by the “independent, distant, disinterested, external, expert consultants assumed to be neutral” (Gorur 2018, p. 90), such as the Australian Bureau of Statistics (ABS). In taking up Gorur’s (2014) “sociology of measurement”, and William Petty’s (1690) “political arithmetick”, this paper critiques the mobilisation of quantification within the national space and the way in which public statistics endeavour to equally inform and stabilise complex units of measurement into simplified narratives.

These narratives are dependent upon classifications and categorisations that the public are readily acquainted with, such as the public school and the private school. Therefore, the starting point is to acknowledge the deeply ideological and politicised positioning of the public school, not only in Australia but across OECD countries. The public school is commonly distinguished as the counterpoint of private schooling; free and accessible, and more readily associated with notions such as equity, social justice and inclusion. The public school has long been at the centre of political critiques and derided for economic inefficiency, and irresponsiveness to the market (e.g. Chubb and Moe 1985, 1990; Friedman and Friedman 1980). In Australia, the public school has been labelled as “failing” (Donnelly 2004, 2014), and disparaged by political leaders as being “too values neutral and politically correct”Footnote 1 with greater parent choice, diversification and privatisation predominantly regarded as panaceas for the alleged “crisis” of public schooling, referred to by Berliner and Biddle (1995) as a “manufactured crisis”.

Why do we count?

Counting is political for the questions it demands: why do we count, and for what purpose? Who do we count? Census institutions (such as the Australian Bureau of Statistics) are presented as a depoliticised and bipartisan apparatus of democratic governance, as a medium of rational calculability. I draw here on William Petty’s (1690) political arithmetick, as expanded upon by Alonso and Starr (1987). Political arithmetic points to the subjectivity of counting and the presentation of numbers for public accessibility and readability, described as “public statistics” (Prewitt 1987). Alonso and Starr (1987) write,

Political judgements are implicit in the choice of what to measure, how to measure it, how often to measure it, and how to present and interpret the results. These choices become embedded in the statistical systems of the modern state and the information they routinely produce (p. 3).

The act of counting is a metric of social inclusion and defines who is in or who is out. A clear example of this is the counting of Indigenous people in Australia. Indigenous and Torres Strait Islander people were not counted or included within the Australian Census until the late 1960s, because Indigenous people were not included or counted as Australian citizens.

In taking up Gorur’s (2014) “sociology of measurement”, the very act of counting, ordering and measuring alters our conception concerning who matters, and what matters, and ultimately underpins important reference points around society, culture and politics. Economies are essentially undergirded by counting, but publics are equally made informed of the counts through contoured statistics; such as the counting and measuring of poverty, for example, or the counting of Gross Domestic Product, the counting of unemployment, or the measurement of educational equity (see Gorur 2014). Thus, there are two distinct features of the modern census: first, the modern census is a means for acquiring a sense of democratic legitimacy, or as Rose (1991) points out, ‘governing by numbers’. The ‘collection of numbers, statistics, ranking, comparisons, data is central to state functioning’ (Lingard 2011, p. 356). Second, the modern census is published for polity consumption, accessibility and readability (as ‘public statistics’, see Prewitt 1987). Statistics are made available to the public as a means of modern governance in the bid for political accountability and transparency. But the way in which numbers are presented to the public should not be assumed to be neutral. There are a number of choices made in how numbers and statistics are presented and published, involving a multitude of technical decisions, and categorisations. Large-scale comparisons and indicators, including the counting of national enrolment shares, are reliant on ‘the mobilisation of a huge machinery of expertise, trust, pragmatism’ in order to define categories, stabilise classifications, and establish methodologies and protocols (Gorur 2015, p. 578). In other words, these are complex processes which require a great amount of simplification and translation, to render these indicators and statistics for public consumption. This is substantiated in their study of early school leaver indicators, a strategic indicator as employed by the European Commission, with the authors noting that ‘institutional differences become flattened in the indicator data’ (Kuusipalo and Alastalo 2019, p. 16). These published statistics are ‘lenses through which we form images of our society’ (Alonso and Starr 1987, p. 3), constructing a type of public yet depoliticised narrative around schooling.

A further example of the selectivity involved in counting, and rendering this count as a public statistic, is evident in how national school enrolment is reported. The exact proportion of the increase in public school enrolment, which led to the proclamation that there was a ‘determined end’ of public school decline was, in actuality, an improvement in decimal places. This is further explained in the ABS media release:

An ABS spokesperson… said that government schools now educated 65.4 per cent of all Australian school students (2,483,802), rising slightly from 65.2 per cent in 2015 and 65.1 per cent in 2014 (ABS 2017).

Arguably this constitutes what Morgenstern (1950) refers to as ‘specious accuracy’, as expanded upon by Rose (1991) ‘in which figures are reported to several decimal places’ (p. 680). The rapid expansion of counting is evident in census institutions and the following section will focus on the counting of schools in further detail.

Counting schools through bipartisan institutes: customising data

The Australian Bureau of Statistics (ABS) collects data pertaining to schools under the banner ‘4221.0—Schools, Australia’ and the data are freely available on the internet (see ABS 2019). An extremely comprehensive range of data can be downloaded in the form of an Excel Spreadsheet or PDF, but this has not always been the case. The ABS data have continually grown more sophisticated, complex and rigorous in scope and scale. The first year from which data are available dates to the year 1960 (at the time of writing) available via their website. A user can download a static six-page PDF document containing a simple dataset: five tables including number of schools, number of teachers, number of pupils and age of pupils (by nation and state). It is included here to point to the stark differences (see Fig. 1).

Fig. 1
figure 1

Screen grab of Australian Bureau of Statistics data from 1956 to 1960. It was first published in 1961. The download is a static 6-page PDF document

Full size image

The extent by which the scale and typology of counting has evolved and expanded is significant. In the year 2000, the ABS first began releasing multiple datasets which could be downloaded as an Excel Spreadsheet (as opposed to the single PDF), referred to as ‘data cubes’, with each data cube available as a mobile document which can be customised and tailored for the individual data user. The following figure demonstrates the data available in 2017 (see Fig. 2):

Fig. 2
figure 2

Data cubes available via the Australian Bureau of Statistics 4221.0-Schools, Australia website. The data cubes are mobile documents and offer customizable and comprehensive data for the individual user

Full size image

The data are customisable for the user’s requirements. If they are willing, a user is able to tally, calculate and account for a wide range of numerical data particular to their varied interest and need; such as the count of schools nation-wide or state; the count of non-special schools, or government schools, Independent schools or Catholic schools, depending on nation or state; the counting of primary enrolment; the counting of secondary enrolment, the counting of staff; the counting of Indigenous and Torres Strait Islander students; the counting of full-time students; the counting of part-time students; or a number of combinations that can be customised using the ABS Excel Pivot Table. In current iterations of the data, a user can divide enrolment shares between sectors: Government (public), Independent (private) and Catholic (private). In the earlier records, the data are simply divided between government and non-government (incorporating a number of private school sectors), and moreover, it states that the enrolment data are not available or complete for all states (see ABS 1960).

The way in which counting takes places indicates social and cultural values around schooling at different points of time, but also the influence of marketisation policies. As Arreman and Holm (2011) point out, upper secondary education has clearly been subjected to the effects of privatisation and market-based principles in a more rigorous way, in comparison to the earlier years of schooling. This is evident in the counting of secondary schools, and the sub-divisions-in-divisions. In 2011, the ABS first made sub-categorical distinctions to secondary school, dividing this category into two separate modes of schooling—junior secondary and senior secondary. In this parameter-modification, the unit of analysis (secondary school) is further divided into two different units (junior secondary and senior secondary) in addition to sector divisions (Government/Independent/Catholic). This enables a user to determine the level of student mobility and flow between junior and senior secondary school (see Forsey et al. 2017). This is an example of parameter-modification, which alters the view and conception of secondary schooling, thereby becoming two separate categorisations. The market is constantly rendering products as more unique, measurable and distinct.

Clearly, the act of counting is subject to a variety of fluctuations and vicissitudes in constructing stable categories and methodologies. But the public reading of national school enrolment share frequently overlooks the numerous typologies involved in shaping and presenting statistics. This is pointed out by Gorur and Wu (2015) who write, in their critique of OECD PISA data, that “the unit of analysis is an important consideration in making comparisons” (p. 653, emphasis in original). The reading of national school enrolment share compares very different units of analysis. Comparing primary school enrolment to secondary school enrolment elides the fundamentally different opportunities and mobilities which these schooling sectors offer students. This is not to discount the important role of primary schools in counteracting inequality, particularly when considering the building blocks of literacy and numeracy, but in modern capitalist economies, the secondary school is an essential component for future pathways and capitalist-oriented mobilities for students. It is also, as pointed out earlier, far more marketised within the global purview. Secondary schools are more likely to be involved in ‘tracking’ or streaming their students into pathways; and when students ‘drop out’ or end their schooling as early leavers, it is predominantly in their secondary school years (see Kuusipalo and Alastalo 2019). While both “OECD and emerging economy countries have achieved universal or near universal primary education, curbing inequalities in secondary and higher education remains a concern” (OECD 2015, p. 44). The secondary school enables further pathways into prestigious universities, vocational study or careers, and studies have shown that parents are well-aware and engaged in “network building” and “parental strategising” in secondary schools (Rowe and Windle 2012; Stacey 2015). Secondary schools offer high-tiered mathematics and STEM courses for access into prestigious universities (Perry and Southwell 2014; Schmidt et al. 2015). School choice or “shopping for schools” is more apparent in the upper secondary school years (Rowe and Lubienski 2017).

When counting and comparing enrolment shares across schools, context matters and attentiveness to units of analysis matters. A reliance on broad categorisations, or misidentification of demographic and geographical variations, simplifies and elides important differences, in addition to masking privatisation within the public sector. Counting and comparing school enrolment requires further attentiveness to context and units of analysis. The following section briefly discusses method and analysis, before critiquing ABS data.

The study and analysis

This paper draws on Australian Bureau of Statistics (ABS) data to examine between-sector differences in enrolment, but the key aim of this analysis is to critique statistical representations and highlight the discrepancies. The sources of data are included in the table below (see Table 1). I draw on the most current ABS school enrolment data available, which is 2018 at time of writing. The data are published on the ABS website and can be downloaded as a range of spreadsheets (see, 4221.0—Schools, Australia, ABS 2019). Data were collected for the period 1996–2018 (see Table 1), and for each year of data collection, data were collected from two separate ABS sources to be verified and cross-checked (see Table 1).

Table 1 The sources of data for each year (1996–2018) from Australian Bureau of Statistics 4221.0-Schools, Australia (ABS 2019)
Full size table

The raw data were first downloaded through the ABS website, and the name of the spreadsheets in which data are acquired and verified is provided (see Table 1). The ABS raw data are available as a numerical count, rather than a percentage. Data were analysed by converting the numerical count into a percentage, as based on the number of full-time students enrolled (part-time students excluded). The percentage was then calculated according to a number of different variables including:

  1. (1)

    Affiliation (government and non-government);

  2. (2)

    Affiliation (government/Catholic/Independent);

  3. (3)

    Year of enrolment;

  4. (4)

    National Report on Schooling (ANR) School Level, or School Level (depending on the year of publication).Footnote 2

To be clear, the percentage of student enrolment within different sectors of schooling is based on the total numerical count of full-time students enrolled, and part-time students are excluded. This is attentive to the ongoing population growth and enrolment increase in all sectors of schooling. The labels above indicate the precise terms used within ABS data.

This paper critiques the role of counting and the publication of national enrolment shares, from the period 1996 to 2018. This period of time was selected to generate up-to-date and current ABS data, and to reflect political machinations. Research indicates that the drift to private schooling was first evident in the mid-1970s, following the introduction of recurrent grants for private schools (Ryan and Watson 2004). Watson and Ryan (2010) show that by the mid-1990s there was an upward trajectory in terms of growing enrolment share in private schools. The year 1996 marks the beginning of the federal Howard Government, a period of time which introduced policies aimed at privatising and marketising the school sector, such as the removal of the New Schools Policy in 1996 (which restricted competition between public and private schools) (see Windle 2009). Policies aimed at increasing competition and marketisation in the schooling sectors have been ongoing since, despite changes in federal government factions (Forsey et al. 2017). Research has consistently demonstrated that enrolment in different sectors of schooling in Australia is patterned by socio-economic status (SES), and public schools tend to serve the majority of students from lower SES backgrounds (Perry et al. 2016; Watson and Ryan 2010).

Discussing the data: the private and public sector in Australia

The table below shows the nation-wide percentage of full-time students enrolled in secondary schools, according to the sector of school (see Table 2).Footnote 3 The emboldened text shows the first year in which a change of pattern was recorded (1996–2018).

Table 2 Enrolment in Secondary Schools in Australia, by sector: government (public); Catholic (private); Independent (private)
Full size table

As Table 2 shows, enrolment in public secondary schools has consistently declined since 1996, with an overall loss of 6.8% enrolment share. The first year which demonstrates an increase of 0.1% is 2016 to 2017. In 2018, the proportion has returned to 2014 levels (59.2%), and evidently fluctuations have occurred within this period between 2014 and 2018. There are also other year-to-year decreases which are very marginal, although an annual decrease within the public secondary school sector is ongoing during the period 1996–2016. It has not returned to the proportional share evident in the mid-1970s, when the majority of the population (approximately 75%) were enrolled in public secondary schools (see Forsey et al. 2017; Watson and Ryan 2010). The increase of 0.2% is so marginal that it is difficult to consider in any conclusive way as being a means to predict future enrolment. It would also be necessary to consider where this marginal increase has occurred, in which spaces—metropolitan, regional or rural—and in respect to significant gaps between public schools, not only in terms of achievement gaps, but also material gaps (Stacey 2019).

The gradual decline in public school enrolment in the secondary years is shown on the following figure (see Fig. 3).

Fig. 3
figure 3

A visual of enrolment shares in secondary schools in Australia, according to sector: government (public); Catholic (private); Independent (private). A gradual decline in public school enrolment is evident, whereas there is a gradual increase in Independent school enrolment, in the secondary school sector

Full size image

The enrolment trend for the Independent secondary school sector shows a reverse pattern, demonstrating an annual increase in enrolment share (5.3% proportional increase from 1996 to 2018). This sector has not experienced any proportional declines since 1996, and the steady growth is robust and consistent, with the only stagnant period of growth occurring between 2009 and 2011.

The Catholic secondary school has increased its enrolment share since 1996, although only by a fractional proportion (+ 1.5%). The Catholic secondary school sector recorded its first enrolment decline in 2015 and this decline has been ongoing to date. This decline was recorded during the Royal Commission into Institutional Responses to Child Sexual Abuse, following widespread complaints of child sexual abuse in the Catholic Church. It is possible that the decline in Catholic school enrolment is linked to a wider social and cultural context and the associated stigma surrounding the Catholic Church in Australia.

The rise of enrolment share in public schooling is largely attributable to the primary school years, rather than secondary school years. This is demonstrated in finer detail below, which indicates the importance of being attentive to units of analysis (see Table 3).

Table 3 The percentage of full-time students enrolled in primary schools in Australia, according to the sector of schooling: Government (public); Catholic (private); Independent (private)
Full size table

This articulates with previous studies which demonstrate the greater levels of marketisation and privatisation evident in the secondary school years (Arreman and Holm 2011; Rowe and Perry 2019; Perry and Southwell 2014). A significantly greater proportion of full-time students attend public schools in Australia, in the primary school years (70.3% compared to 59.2%). Therefore, in pointing out privatisation, it is important to identify systemic meta-privatisation, that is, explicit forms and contours occurring “in” and “of” the public sector (see Ball and Youdell 2008).

Table 3 demonstrates that enrolment share in public schools has consistently declined since 1996.Footnote 4 However, there is a change in 2014 and this year marks a gradual increase in enrolment shares in the primary and public sector demonstrating an increase that is relatively more significant (1.3%). The Independent school sector has recorded a continued growth in enrolment share since 1996 but, in a contra pattern, the Independent primary school sector recorded a proportional decrease in enrolment share in 2015 for the first time. The Catholic school sector has recorded declining enrolment since 2012.

These increases are shown in a graph (Fig. 4).

Fig. 4
figure 4

A visual of enrolment shares in primary schools in Australia, according to sector: government (public); Catholic (private); Independent (private). A gradual decline in public school enrolment is evident, but there is a slight increase in enrolment share in 2014

Full size image

Overall, when amalgamating both primary and secondary school sectors, the public government school has declined in enrolment share since 1996, whereas the private sector has increased its enrolment share during this same period of time, with the most significant enrolment increase within the Independent private school sector. Certainly, in the primary school sector, enrolment share in the Catholic school has declined since 1996 (− 0.9%) and increased only marginally in the secondary school sector (+ 1.5%).

But it begs the question as to where these increases are most evident, in terms of demographics and spatiality. In critiquing an “average performance score” and mean scores in PISA, Gorur and Wu (2015) write that “the country average often masks wide variations between one state or province within the country” (p. 653). This is also evident in counting nation-wide enrolment share. Thereby, it is useful to consider enrolment shares according to state/territory within the secondary school sector (see Table 4).

Table 4 The following table shows 2018 secondary school enrolment shares as divided between sectors (Government; Independent and Catholic), and state/territory in Australia
Full size table

The table is arranged from lowest to highest enrolment share, as based on the government school sector. The table also shows the enrolment proportion change from 2017 to 2018, albeit predictions as based on this fraction of time should be made tentatively.

Overwhelmingly, the Catholic secondary school sector has experienced a decline in enrolment share across all jurisdictions, except for the Northern Territory (0.3% marginal increase). The Independent school sector is more mixed, with three jurisdictions experiencing very small declines. The public school sector has achieved enrolment share growth in the majority of jurisdictions, except for the Northern Territory and New South Wales (although these are marginal declines, at 0.1%).

The Australian Capital Territory (ACT) records the lowest percentage of full-time students enrolled in public schools, whereas the Northern Territory records the highest percentage, closely followed by Tasmania, reflecting socio-economic status, social composition and demographics. Western Australia has the highest proportion of students enrolled in Independent secondary schools (21.2%), followed by South Australia (20.5%). These states would also be described as more politically conservative and socially homogenous, compared with New South Wales and Victoria. Tasmania, followed by the Australian Capital Territory (ACT), has the lowest proportion of students enrolled in Independent secondary schools. Although, the ACT has the highest proportion of full-time students enrolled in private schools overall, because it records the highest percentage of students enrolled in Catholic schools (28.1%), followed by New South Wales (24.4%).

Conclusion

In relation to OECD PISA data, Gorur (2015) writes that “these indicators seek to describe education systems not only precisely but also comparatively and predictively” (p. 578 emphasis in original). The media headlines proclaiming an “end” to the private school drift in Australia endeavour to draw on statistics to confidently compare and predict future enrolment trends. However, when subjecting the count to finer scrutiny based on different units of analysis, it is apparent that a confident prediction regarding the death of the private school drift is highly problematic. Clearly, this obfuscates the extent of privatisation and rising enrolment in private schools, in Australia. Certainly, Australia is well above the average proportion of students enrolled in private schools (the OECD average is reported to be 18%) (see OECD 2011). In comparison to other OECD countries, such as the United States, New Zealand or Canada, Australia maintains an exceedingly high proportion of students enrolled in private secondary schools (41%). This number is annually increasing with a steady rise within the Independent secondary school sector. Overall, the Independent private school has experienced continued growth in enrolment share, whereas overall the public school has recorded an ongoing decline. It would be simplistic to attribute this decline exclusively to rational parental choice, thereby omitting the raft of government incentives and social drivers around private school choice.

This paper has not sufficiently explored the reasons for the decline, nor the reasons for the growth in the public primary school sector. It is beyond the scope of this paper to explore this in great depth, however, an analysis of these figures in conjunct with political, economic and social conditions would be useful in future work (see Forsey et al. 2017). As a preliminary suggestion, the reasons for the growth in enrolment in the public primary school sector are most likely attributable to economic and market conditions in Australia, rather than ideological. In the 2018 OECD Economic Indicators Report, it shows that Australia stands out from other OECD countries for its continued economic growth, but also for its rising costs. Capital cities such as Sydney and Melbourne are regarded as the most expensive in the world, due to house prices (OECD 2018); and when it comes to the cost of living, Australia maintains some of the world’s highest costs per-student for higher education, and private secondary school education—which is unregulated, uncapped and rising at a faster rate than the Consumer Price Index (CPI). When faced with these economic challenges, parents living in major cities are likely to seek to avoid additional costs incurred at primary school level by enrolling their children in public primary schools.

The aim of this paper has been to point to the political arithmetic involved in the counting of national school enrolment, and the publication of school enrolment as contoured statistics and media headlines. The publication of school enrolment and the celebration of a “good news story for public schools” is mutually politicised and performative in the sense that it is instrumental for democratic governance, and the construction of “informed publics” (Gorur 2018). These claims around public school enrolment may influence funding debates and be utilised in the outsourcing of government reviews to think-tanks and the like, and for these reasons they are a core part of the circulation of a particular discourse around public schooling. Public statistics, as attached to an expert bipartisan and depoliticised institution, convey authority and legitimacy in numbers, assembling a narrative around public schooling. The narrative suggests a strong public schooling system, rather than a weakening system experiencing consistent decline. When this fails to be illuminated in the political arithmetic of public statistics, this essentially masks the meta-privatisation occurring within public schools and obfuscates the marketisation and material gaps between public secondary schools. Further, when we neglect to highlight the substantial differences in enrolment share between primary and secondary schools within the public sector (70.3% compared to 59.2%), and group this as one broad category, we are eliding and flattening key differences which are associated with capitalist-oriented mobility, equity and opportunity.

“Informed publics” writes Gorur (2018), is a process in which “previously distant actors are drawn into new relations with “settled” accounts and summary calculations” (p. 90). Whereas once the publication of school data, in terms of ‘counting’ and measuring enrolment, teacher counts and so forth, was largely non-existent and thin, it now constitutes a thorough and comprehensive array of measurements that are published and presented to the public. Counting is by no means a-political, nor does it occur in a cultural vacuum; counting and public statistics is underpinned by social, cultural and economic incentives, characterised by intent, selectivity and politicised judgements around what to count, who to count and for what purpose. The public is informed in order to leverage their support but also to stabilise and give authority to the numbers (Gorur 2014). The counting to one or two decimal places indicates the “specious accuracy” and political arithmetic of counting, purposefully taken up to lend authority, accountability and stability to the numbers and statistics.