The Cost of Education · Question 03
01 / The answer
Do more expensive schools actually produce better results?
Higher Cost
More Benefit
A higher cost is not proof that my child will benefit more.
Lower Cost
Less Benefit
A lower cost is not proof that my child will benefit less.
Cost alone does not tell us how much the school adds for this child.
02 / Why results alone can mislead
Three influences. One result we can see.
Child
- earlier learning
- family circumstances / support
- previous opportunities
Peers
- classmates
- expectations
- learning culture
School
- teaching
- feedback
- support
- curriculum / practices
What we observe
- achievement
- progress
- experience
A strong final result does not tell us how much came from each part.
03 / Results are not the same as progress
Do not confuse the finish with the distance.
Result
Who finishes with the highest result?
Progress
Who made more progress from a similar starting point?
Researchers try to make that comparison fairer by looking at children with more similar earlier results and backgrounds.
Australian studies do not all reach the same answer.
04 / A parent decision path
What should I ask about my child?
Starting group
Were these students already doing well?
What were they achieving before?Change
What changed while they were here?
Did similar children make more progress?This child
Will my child actually receive the benefit?
Is it relevant, accessible and suitable compared with our realistic alternatives?
How did we get here?
Seven discoveries behind the answer
The seven Discoveries below separate the result we can see from what students brought, what peers and schools may add, and what the evidence still cannot settle.
Discovery · EvidenceGreat results—but what do they tell us?
A school’s results are worth celebrating.Evidence — Q03-P04-S19
A school’s results are worth celebrating. But before we use them to judge the school, we need to know whose results they are.
Are we looking at the whole graduating year, one subject, or a selected group studying ahead of their age?
Clyde’s early VCE mathematics results are a useful example. They show what that group achieved. They are not the results of a complete graduating Year 12 class.19
That does not take anything away from the students or their teachers. It helps us understand what we are comparing.
“The students did well” and “the school helped them do better than they otherwise would have” are different claims.
We need more than the final result to know how much the school helped.
Discovery · EvidenceWere these children already doing well?
Children do not all arrive at secondary school with the same learning behind them.Evidence — Q03-P04-S01 · Q03-P04-S20 · Q03-P04-S21 · Q03-P04-S14 · Q03-P04-S16 · Q03-P04-S17
Children do not all arrive at secondary school with the same learning behind them.
Their earlier schooling, family circumstances and opportunities may differ. Some schools select students by academic performance. Some neighbourhood schools also select students for particular accelerated programmes.012021
We should therefore ask what students were achieving near the beginning—not only what they achieved at the end.
Glen Waverley’s reports of families moving into its school area are relevant here. But movement alone does not tell us why every family moved or how much it changed later results. Our work on Clyde, Glen Waverley and East Doncaster has not established that parents explain their success.141617
A school may attract children who are already doing well, help them make further progress, or do both.
The task is to find out which explanation the evidence supports, rather than assume one.
Discovery · EvidenceDo similar children make more progress?
A fairer question is:Evidence — Q03-P04-S01 · Q03-P04-S02 · Q03-P04-S03 · Q03-P04-S04
A fairer question is:
When children start at a similar level, do those at one school go on to learn more than those elsewhere?
Researchers try to make that comparison fairer by looking at children with more similar earlier results and backgrounds.
The Australian studies we reviewed do not all give the same answer. Some find no clear extra advantage for Catholic or independent schooling in the learning they measured. Others still find an advantage, including in senior-secondary results.01020304
They study different children, ages and outcomes. We have not fully resolved why some of the overlapping studies disagree.
Making the comparison fairer also does not settle every question. Researchers cannot measure everything. A remaining difference might reflect the school, something else they have not captured, or both. And not finding a clear advantage does not prove that every school is the same.
The useful question is not just “Who gets the highest marks?” It is “Who makes more progress from a similar starting point?”
Even then, we need to ask what caused that progress.
Discovery · Peers · bounded hypothesisCan the classmates help my child do better?
A parent might reasonably think:Evidence — Q03-P04-S12 · Q03-P04-S06 · Q03-P04-S07 · Q03-P04-S08 · Q03-P04-S16
A parent might reasonably think:
“I want my child around others who take learning seriously.”
That is an illustrative argument, not a quotation from a research participant.
The reviewed Victorian peer research gives us a reason to take this seriously: it found benefits from higher-achieving classmates in the setting it studied.12
But two things need to be kept separate.
A class of high-achieving children can have a high average simply because those children were already doing well. Showing that they help another child learn more requires different evidence.
We are also exploring whether a school’s reputation attracts academically focused families, whose children may help reinforce its learning culture and results. That is a possible explanation—not something we have demonstrated for our three schools.06070816
The school does not have to have created every part of a good peer environment for a child to benefit from joining it.
But we still need to ask whether that benefit is real for the child, and whether an accessible lower-cost school could offer it too.
Discovery · EvidenceWhat is the school doing that helps?
None of this means schools do not matter.Evidence — Q03-P04-S13 · Q03-P04-S09 · Q03-P04-S10 · Q03-P04-S11
None of this means schools do not matter.
The research includes evidence that particular school practices can improve learning. One Australian study tested a defined teacher-development programme and found improved student results.13
That directs us towards more useful questions. What feedback does the child receive? What happens when they fall behind? What happens when they are ready for harder work? Is the promised support actually available?
Resources may help a school provide these things. But the amount spent, how it is used and what changes for students are separate questions. Evidence about school funding is not automatically evidence about what an extra dollar of private tuition buys.091011
Ask what the school does that helps children—not only what it charges or what it owns.
A tested teaching programme does not prove that every school using similar language delivers the same results. We still need evidence about what happens in the school being considered.
Discovery · Safety · evidence boundaryWill my child feel safe—and get help?
A parent may care about more than marks:Evidence — Q03-P04-S25 · Q03-P04-S22 · Q03-P04-S23
A parent may care about more than marks:
“Will my child feel safe? Will they fit in? Will someone listen and act if something goes wrong?”
Those are illustrative questions. They cannot be answered by looking at the sports grounds or examination results.
Facilities, classroom behaviour, bullying, belonging and the way complaints are handled are different parts of school life.
One Australian study, using older student data, found lower average reported bullying in independent schools. It did not show that paying higher fees caused that difference or that a particular school would be safer for a particular child.25
The named-school cases need care too. A serious failure at an expensive school challenges any promise that fees guarantee safety. It does not tell us whether private or government schools are safer overall. A school’s willingness to investigate, acknowledge problems and take corrective action also matters.2223
For parents, the questions become concrete: Can a child report a problem safely? Who helps? What protection follows? Does someone check that the harm has stopped?
“We take wellbeing seriously” is a starting promise. What matters is what the child actually experiences.
Discovery · FindingDoes paying more mean my child will do better?
This brings us back to the bill.Evidence — Q03-P04-S01 · Q03-P04-S02 · Q03-P04-S03 · Q03-P04-S04 · Q03-P04-S09 · Q03-P04-S11 · Q03-P04-S25
This brings us back to the bill.
A higher-fee school might offer something genuinely useful for a child. A lower-cost school might already offer the support, challenge or community that child needs. Neither possibility can be decided by the price alone.
The research we reviewed does not establish a general rule that higher fees buy more learning or greater safety.01020304091125
Sometimes the more expensive school may be the better choice for a child. The price itself does not tell us that.
That does not mean the extra money buys nothing. It means we need to identify the particular benefit rather than assume it comes with the fee.
A higher price doesn’t automatically mean more benefit for my child. A lower price doesn’t automatically mean less benefit.
Strongest parent counterargument
The experience can matter even without a guarantee.
An illustrative parent’s argument—not a research-participant quotation—is:
“I’m not expecting a perfect school or a guaranteed future. I want somewhere my child feels supported, has the right opportunities and enjoys learning. The cheaper schools we can actually choose may not offer the same experience.”
Response
That can be a reasonable reason to choose the school.
A child feeling supported or finding a community they value does not have to lead to a higher salary to matter. Nor must a parent wait for perfect research before making a decision.
But valuing the experience and proving that it produces better future results are different things. We should ask whether the child will receive what the family is choosing, whether the other options differ in the way claimed, and what evidence supports any promised improvement.
Where this leaves us
Strong results and helping a child make progress are not always the same thing.
Some schools get better results. Some schools help children make more progress. Those are not always the same thing.
A school may teach well, attract students who were already doing well, bring together helpful classmates—or combine all three. Our research does not give us one formula for separating those contributions at every school.
Paying more does not settle the question. Neither does choosing the school at the top of a results table.
The useful question is:
The useful question
Would my child do better here than at the other schools we could choose—and what makes us think so?
“Better” needs to mean something specific: more learning, more suitable support, a safer experience or another benefit the family can identify. Research cannot predict one child’s future with certainty. It can help us ask better questions about the promises we are relying on.
Sources and evidence boundaries
What this investigation relies on.
Academic learning remains primary and safety remains one supporting Discovery. Housing-cost estimates, degree value, lifetime wages, networks, a school ranking, a value-for-money calculation and a personalised school recommendation remain outside scope.
Source 19 · Q03-P04-S19
VCE Results 2025
Clyde Secondary College
15 December 2025; 2025 examination outcomes
Evidence boundary: Full HTML rechecked. Only Year 10 General Mathematics Units 3/4; no complete graduating cohort. Unknown scored denominator; inconsistent above-40/40+ wording. Do not infer total cohort size from public honour-roll names.
Source 01 · Q03-P04-S01
School Sector Differences in Student Achievement in Australian Primary and Secondary Schools: A Longitudinal Analysis
Gary N. Marks · Journal of School Choice 9(2)
LSAC; Years 3, 5 and 7
Evidence boundary: Publisher abstract/preview and notes; full result tables unavailable. Keep quantitative table claims provisional; exact sample exclusions, baseline dates, switching treatment and intervals not fully audited.
Source 20 · Q03-P04-S20
High Achievers Program
Clyde Secondary College
Undated current webpage; accessed September 2026
Evidence boundary: Inherited primary-page reading from Pass 02; no new outcome claim. Current programme criteria do not establish membership or historical selection for the 2025 VCE class.
Source 21 · Q03-P04-S21
ALPHA — Advanced Learning Program for High Achievers
East Doncaster Secondary College
Current webpage; accessed September 2026
Evidence boundary: Inherited primary-page reading from Pass 02; no new outcome claim. Provision and selection mechanism only, not causal programme benefit.
Source 14 · Q03-P04-S14
2024 Annual Report to the School Community
Glen Waverley Secondary College / Victorian Department of Education
2024 outcomes; report approved subsequently
Evidence boundary: PDF text rechecked; numerical page rendering unsuccessful. Year-level aggregates are different cohorts. Numbers below are research transcriptions, not approved chart input.
Source 16 · Q03-P04-S16
Enrolment Information
Glen Waverley Secondary College
Undated school statement; accessed September 2026
Evidence boundary: Full HTML rechecked. The school reports residential movement; individual motives, intake change and causal result effects remain unlinked.
Source 17 · Q03-P04-S17
2024 Annual Report to School Community
East Doncaster Secondary College / Department of Education
2024; attested in 2025
Evidence boundary: PDF text rechecked; targeted NAPLAN-table screenshot failed. Same-cohort growth cannot be calculated by subtracting Year 7 and Year 9 aggregates from the same annual report.
Source 02 · Q03-P04-S02
The public-private debate: school sector differences in academic achievement from Year 3 to Year 9?
Sally A. Larsen; Alexander Q. Forbes; Callie W. Little; Simon H. Alaba; William L. Coventry · The Australian Educational Researcher 50, 275–306
Years 3–9; historical cohort
Evidence boundary: Publisher abstract, metadata, references and ethics/funding statement read; subscription full text not obtained. Do not infer complete sampling or attrition rules from acknowledgments or references. Online 2022; issue 2023.
Source 03 · Q03-P04-S03
Does school type affect cognitive and non-cognitive development in children? Evidence from Australian primary schools
Hong Son Nghiem; Ha Trong Nguyen; Rasheda Khanam; Luke B. Connelly · Labour Economics 33, 55–65
Primary-school LSAC cohort
Evidence boundary: Publisher abstract/introductory preview and institutional metadata; full paper not obtained. Bologna lists an open version, but its file was not retrievable in this review. Do not invent estimator details or sample size.
Source 04 · Q03-P04-S04
Do Catholic and Independent schools “add-value” to students’ Tertiary Entrance Performance? Evidence from longitudinal population data
Gary N. Marks · Australian Journal of Education 59(2), 133–157
2011 tertiary-entrance cohort
Evidence boundary: Publisher abstract and metadata read; full model tables not obtained. Tertiary entrance aggregate standard deviations and ATAR ranks are different quantities. Full uncertainty and cohort-selection audit remains open.
Source 12 · Q03-P04-S12
Achievement effects from new peers: Who matters to whom?
Duncan McVicar; Julie Moschion; Chris Ryan · Economics of Education Review 66, 154–166
Years 7 and 9 across cohorts
Evidence boundary: Final-publication institutional abstract rechecked. No full-table re-audit; use final 2018 subgroup wording, not an earlier working-paper summary.
Source 06 · Q03-P04-S06
The substantiveness of socioeconomic school compositional effects in Australia: measurement error and the relationship with academic composition
Michael G. Sciffer; Laura B. Perry; Andrew McConney · Large-scale Assessments in Education 10, article 21
2015–2017 NAPLAN, Years 3–5 and 7–9
Evidence boundary: Open original full HTML methods, results and discussion read. Read alongside both 2024 debate papers. Statistical mediation is not independent proof of causal ordering.
Source 07 · Q03-P04-S07
No substantive effects of school socioeconomic composition on student achievement in Australia: a response to Sciffer, Perry and McConney
Gary N. Marks · Large-scale Assessments in Education 12, 8 (2024)
2017 NAPLAN with earlier scores; published 14 March 2024
Evidence boundary: Open original full HTML read. Observational model comparison; read together with the October 2024 reply. No ethnic or genetic interpretation is adopted.
Source 08 · Q03-P04-S08
The substantiveness of school socioeconomic compositional effects: a response to Marks
Michael G. Sciffer · Large-scale Assessments in Education 12, 35 (2024)
Published 22 October 2024
Evidence boundary: Open original full HTML read. A methodological response, not a new independent experiment. Disagreement concerns mediation, specification and effect interpretation.
Source 13 · Q03-P04-S13
Improving student achievement through professional development: Results from a randomised controlled trial of Quality Teaching Rounds
Jennifer Gore; Andrew Miller; Leanne Fray; Jess Harris; Elena Prieto · Teaching and Teacher Education 101, 103297
Primary-school trial with 8-month follow-up
Evidence boundary: Original journal and author-university indexed abstract rechecked. Intervention and measured outcome specific; no sector or fee comparison.
Source 09 · Q03-P04-S09
Educational Achievement and the Allocation of School Resources
Deborah A. Cobb-Clark; Nikhil Jha · Australian Economic Review 49(3), 251–271
Final article, September 2016
Evidence boundary: Publisher abstract and author-centre record read. Final article wording differs from the 2013 working paper. Full final tables not obtained; do not silently substitute working-paper estimates.
Source 10 · Q03-P04-S10
Educational Achievement and the Allocation of School Resources — working paper
Deborah A. Cobb-Clark; Nikhil Jha · IZA Discussion Paper 7551
August 2013 version; Victorian administrative data 2008–2011
Evidence boundary: Full PDF text read; targeted screenshot attempts unsuccessful. Pre-publication version only. Within-school variation in public-school SRP spending, not tuition; exact table cells not promoted as final-journal results.
Source 11 · Q03-P04-S11
What Impacts Can We Expect from School Spending Policy? Evidence from Evaluations in the United States
C. Kirabo Jackson; Claire L. Mackevicius · American Economic Journal: Applied Economics 16(1), 412–446
Historical policy-evaluation studies
Evidence boundary: Final publisher abstract and metadata rechecked. US public-spending policy evidence, not Australian fees. No currency conversion or tuition-return extrapolation.
Source 25 · Q03-P04-S25
School climate, student engagement and academic achievement across school sectors in Australia
Wojtek Tomaszewski; Ning Xiang; Yangtao Huang · The Australian Educational Researcher 51, 667–695
LSAC / historical school-age cohort
Evidence boundary: Open original article rechecked at publisher. 2012 LSAC climate observations; publication online 2023/issue 2024. Descriptive bullying means and adjusted academic-outcome models are different analyses.
Source 22 · Q03-P04-S22
Cultural Review Report — school response hub
Ballarat Grammar
Page checked 14 September 2026; review released July 2026
Evidence boundary: Full HTML rechecked. No separate completed-outcome verification identified on the checked hub. This is not proof of no implementation or a missed deadline.
Source 23 · Q03-P04-S23
Independent Cultural Review of Ballarat Grammar
Intersection
Review released 23 July 2026
Evidence boundary: Full original reviewer HTML read. Systemic cultural review, not incident adjudication or a representative sector-risk estimate. Participant totals differ in reviewer and school summaries; no total is used as a prevalence denominator.