The government of Ontario provides public funding for two parallel school systems: separate (Catholic) and public (non-Catholic) – I'm going to resist the temptation to provide a paragraph here outlining the constitutional and political history that explains why this is the case. For what follows, all we need to know is that in Ontario, Catholics have the choice of sending their children either to separate or to public schools, and non-Catholics do not have this choice. Of course, both can send their children to private schools that do not receive government funding, but we're going to be comparing education outcomes in the two publicly-funded systems.
It's pretty easy to sum up the incentives facing the two sets of school administrators:
- Public schools have a monopoly on non-Catholics who can't afford to send their kids to private school.
- Separate schools face a clientèle that always has the option of switching to the public school system.
Of the two, the Catholic administrators have the greater incentive to provide higher-quality education: if the Catholic system were widely known to be dysfunctional, it would disappear.
Which brings us to this study by Wilfrid Laurier University economics professor David Johnson for the CD Howe Institute. He uses test score data to provide rankings for the various school boards, and he also corrects for observable socio-economic factors. He finds that
- Of the 13 'above-average' school boards, 11 are Catholic.
- Of the 10 'below-average' school boards, none are Catholic.
As natural experiments go, this is pretty clean: in a given community, both school boards are drawing from essentially identical populations. For example, York Region's public system does poorly, and it's separate system does well – and it's pretty much impossible to explain this fact by pointing to significant differences between Catholics and their non-Catholic neighbours.
Generally speaking, Catholic systems outperform the public systems with overlapping territories. And although it may be tempting for this former student in the Simcoe County Separate School Board (Guardian Angels and Monsignor Lee schools) to ascribe this difference to the preternatural abilities of Catholic students, a far more plausible explanation lies in the different sets of incentives facing the administrators of the two systems. The Catholic administrators are providing better education outcomes for no better reason than that they have the stronger incentives to do so.
Competition. It's a good thing.
Update: It turns out that this theme has in fact been developed more rigourously in a NBER working paper by David Card, Martin Dooley and McMaster's Abigail Payne, listed as a reference in David Johnson's note:
School Competition and Efficiency with Publicly Funded Catholic Schools: The province of Ontario has two publicly funded school systems: secular
schools (known as public schools) that are open to all students, and
separate schools that are open to children with Catholic backgrounds.
The systems are administered independently and receive equal funding
per student. In this paper we use detailed school and student-level
data to assess whether competition between the systems leads to
improved efficiency. Building on a simple model of school choice, we
argue that incentives for effort will be greater in areas where there
are more Catholic families, and where these families are less committed
to a particular system. To measure the local determinants of
cross-system competition we study the effects of school openings on
enrollment growth at nearby elementary schools. We find significant
cross-system responses to school openings, with a magnitude that is
proportional to the fraction of Catholics in the area, and is higher in
more rapidly growing areas. We then test whether schools that face
greater cross-system competition have higher productivity, as measured
by test score gains between 3rd and 6th grade. We estimate a
statistically significant but modest-sized impact of potential
competition on the growth rate of student achievement. The estimates
suggest that extending competition to all students would raise average
test scores in 6th grade by 6-8% of a standard deviation.
Worthwhile Canadian Initiative 
“Catholics have the choice of sending their children either to separate or to public schools, and non-Catholics do not have this choice.”
I’m sorry, but this premise simply is not the case. I attended Catholic school, and there were many students who were protestant, some evangelical, and even some jewish (leaving aside the agnostic). This, of course, was in the fairly white bread Niagara region. In Toronto and its suburbs, some schools might have Muslim populations challenging the size of actual Catholics. I’m mainly referring to secondary school, but I imagine this is likely the case for elementary school as well.
As I understand it, admitting any student, regardless of religion, was the price paid for full public funding of the separate school system back when Bill Davis was premier.
Makes me wonder, why stop with the school system? Why not intoduce competition into the public health sector. I’m not just suggesting more private clinics, but public hospitals competing against other public and private delievers of health services. Instead of the traditional block funding, why don’t we have hospitals generate revenue for services delievered? I know this is the one of the solutions Dr. Brain Day has put forward for improving our public healthcare system. By bringing in market forces, we might be able to deliver more efficent healthcare for Canadians.
If you standardize the data for average number of ESL students and average years of education of parents I think these differences would shrink dramatically.
I’m not saying that these incentives don’t have an effect, just that it may not be as large as the effect observed here.
Can the Catholic schools expell unruly students? I know of teachers in Australia who thought this was the most important difference between private and public schools.
If the Catholic schools can select the students that they educate, then they can effect the outcomes in other ways than performance of the school. Having low performing students bounced to the public schools can affect the results quite a bit.
The balance of evidence suggests that school choice has a positive effect on school performance.
However this example suffers from selection bias and can tell us nothing about whether school choice increases school effectiveness. This is not a natural experiment. The model will suffer from selection of student into schools based on their unobserved characteristics.
In particular parents have an large incentive to fake being catholic, the cost of 14 years of private education. Those non-Catholics who do apply to catholic schools will be more able and from a higher socio-economic backgrounds.
Additionally catholic schools have an incentive to cream skim the most able students even if there are doubts about their parents faith. This is illustrated above by Andrew F.
To answer the question does competition between school increase exam results you must have a valid IV, see Hoxby 2000 for a disputed example.
I don’t think there is any difference between the Catholic and public schools’ ability to expel students.
Lots of ignorance here.
1: There is no shortage of non-catholics at catholic schools. So the basic premise of this post is a lie. Both boards are in competition. [I so enjoy being called a liar for repeating verifiable facts. So I’m going to respond to this point by quoting from the Simcoe Muskoka Catholic School Board’s website: “Children attending one of our elementary schools must be either baptized Catholic or have a parent who is Catholic. In certain circumstances, at discretion of the school superintendent, some exceptions may apply.” – ed]
2: The basic premise of the study was adjusting for ESL, poverty and such. So some of the comments here are misplaced.
3: Catholic schools get more money. [No, they don’t. – ed]
It is in no way surprising that of the two school systems, the wealthier one produces better results.
Well-funded education – it’s a GOOD thing.
I think it is very hard to make this a properly controlled experiment. I’m not sure if this is the case, but imagine two parallel systems, one with a more limited capacity that people (regardless of reality) think is better. Then that system will be tend to choose better students (not necessarily observable SE factors) because of selection and self-selection (those queuing for the “better” school may be more motivated). The only way for the experiment to be controlled is for people to be randomly assigned to the schools, but then the premise was that choice was good! I’m not necessarily against choice (I’m generally for it, but concerned that self-selection may fragment society, and that it may also self-select the better teachers for the better students, when maybe the weaker students need the better teachers). This is a tricky issue, I’m not sure economist are wise to get mixed up in it.
Andrew F…
Where would students go that are expelled from public schools (given that education is compulsory)? Surely, there are some statistics on this?
In Alberta miscreants go from Separate to Public for one last chance on the way out the door. Is there enough of these to drag down the statistics? Unlikely.
Maybe the Separate boards are smaller and it is easier to manage smaller school systems?
Is there any control in the paper for the differences in the curriculum? Could it be possible that the required religion and morality classes, and the attempt of schools to instill Catholic ideals into all aspects of student life, might – just possibly – have a beneficial effect on student academic outcomes?
Or maybe it’s the uniforms.
Both systems follow the same curriculum, and receive the same funding per student.
Another highly speculative theory: parents that take a more active role in their child’s academic life may be more likely to send their child to Catholic school (due to perceived stricter behavioural rules, discipline, uniforms, etc.). Children of these kinds of parents are more likely to have been read to as a child, strictly observed homework routines, etc.
I have a feeling that the performance disparity is the result of a range of factors.
Stephen, why no response to the point made about Catholic school being open to students of any religious background? It makes a material difference to your argument in your first post. I’m curious where the authors of these papers are getting this info, for that matter. Maybe I should actually read them ;).
Actually, I did respond. From the Simcoe Muskoka Catholic District School Board website:
Children attending one of our elementary schools must be either baptized Catholic or have a parent who is Catholic. In certain circumstances, at discretion of the school superintendent, some exceptions may apply.
And as a student who went through the system, tradition – or call it inertia – is pretty important. Parents who went to Catholic schools send their kids to Catholic schools pretty much out of habit.
I think the source of the confusion is that the policy is different for elementary and secondary school. For instance, the Dufferin-Peel CDSB policy is as follows:
“Who can register at DPCDSB schools?
Elementary
Elementary admissions are open to Catholic children and children of Catholic parents. Junior Kindergarten students must be 4 years of age, and Senior Kindergarten students must be 5 years of age, by December 31 of the school year for which they are registering.
Secondary
Secondary school admissions are under the provincially legislated Open Access and no child should be refused on religious grounds unless they are unwilling to respect the rites and practices of the Catholic school system.”
http://www.dpcdsb.org/CEC/Schools/Admissions.htm
What I find surprising is that there haven’t been successful challenges of the restriction of primary school admissions to Catholics.
I think the source of the confusion is that the policy is different for elementary and secondary school.
Perhaps – but the test results we’re talking about are for kids in grades 3 and 6. So the elementary school eligibility rules are what matter here.
What might be interesting to study is university test results for students from each school system. I know the Faculty of Mathematics at Waterloo tracks students’ first year math grades versus high school entrance averages to normalize for grading variations in the admissions process.
There must be millions of possible reasons why catholic schools might do better on average. Seems to me that the evidence for assuming it is because the Separate School board administration is worried about losing students to the public system is far too weak to come to any sort of conclusion that that is what accounts for the difference.
Or, in other words, what nmd said.
Millions? Indeed – but I think we can rule out most of them (eg: Catholics are inherently smarter; Catholics’ guardian angels’ guide their pencils, etc) out of hand.
The authors of the NBER study control for what they can observe about families and their neighbourhoods. If there’s a better explanation based on what we can observe, then I’d like to see it.
There is another explanation: teachers at Catholic schools are more likely to be motivated by their religion than teachers at other schools. Doing God’s work, fear of God, raising good Catholics, etc. There are many reminders of religion (crucifixes, priests, nuns, co-religious colleagues and parents) in Catholic schools which might add this motivation.
(I believe the school choice argument – and regardless of their character two public systems are better than one – but I’m surprised no one has pointed out that religion might actually matter).
Alright, this Laurier study looks… questionable (though, one should expect bad stats if they’re from WLU rather than UW). School boards, separate or public, are heterogeneous with respect to socio-economic determinants, numbers of schools, size of schools, proportions of recent immigrants, and a variety of other factor. The coin-flipping example as the basis for providing background about the methodology makes this statistician think that this non-peer-reviewed think tank article should not be taken seriously in the slightest.
Anyhow, if this is true:
Parents who went to Catholic schools send their kids to Catholic schools pretty much out of habit.
Then the “competition” argument gets destroyed rather easily – you cannot assert an effect of competition if that’s not how parents are actually making their schooling choices and, in any case, a “natural” experiment is just an observational study which can provide descriptive interpretation only.
Fair enough – although the force-of-habit story could be used to explain why the effect found in the NBER study was so small. But still positive.
Well,
We estimate a statistically significant but modest-sized impact of potential competition on the growth rate of student achievement.
Why don’t they report the p-values and confidence intervals in the abstract?
The estimates suggest that extending competition to all students would raise average test scores in 6th grade by 6-8% of a standard deviation.
This isn’t true – while you might be able to say that “competition” is significantly associated with an increase in average test scores, making this a causal claim is impossible in light of the non-experimental nature of the study. How did they deal with self-selection bias? One thing I might assert is that Catholic schools a smaller less heterogeneous population, so that the corresponding variances in test score improvement are also smaller. This has rather crucial bearing on what kind of stats are appropriate for an analysis, and it doesn’t seem to be mentioned here.
I suggest that you read the paper.
@Andrew F
The reason for the difference is likely that elementary school funding is constitutionally grounded and therefore discrimination is protected. High school funding (post-Davis) is legislatively grounded and therefore Charter challenges could be mounted if it were not Open Access.
It’s worth noting that there may be systemic differences in the Catholic and public costs, especially in the area of salaries.
We are discussing elementary here, not high school. I could only easily find numbers for OSSTF (public secondary school teachers) and OECTA (English Catholic teachers) in a single area (Toronto).
Click to access 2004-2008_coll_agree.pdf
Click to access toetfo.pdf
These both overlap 2004, and they show differences between the two systems. Given that all teachers in a system belong to the union corresponding to that system, any study should control for these systemic differences.
If a major spending component (teachers’ salaries) is different (higher or lower) it either frees up or consumes funds that might go to other parts of the system.