Do school teachers send their children to government schools?

I’ve often wondered whether teachers more or less likely to send their children to private schools, and in surfing through back issues of the Australian Education Researcher, I found at least a partial answer. A 2008 paper by Helen Proctor, titled “School Teacher Parents and the Retreat From Public Secondary Schooling: A View from the Australian Census, 1976-2001”, uses custom-made census cross-tabulations. The results are only for Sydney, but make interesting reading nonetheless.

First, here’s the overall pattern, where the columns are the type of school the child attended.

And now, let’s look at school teacher fathers and mothers (rows) and the type of school their children attended (columns).


In both censuses, Sydney school teachers were less likely to send their children to government schools than the Sydney average parent, and they have shifted towards non-government schools at about the same rate as everyone else. About half of all teachers now send their children to a non-government school.

One way to explain this is that teachers parents value education more highly than the typical parent, so spend more on it. But you might also think that part of the shift is due to teacher parents sending the children to their own school (to make drop-offs easier, or maybe to keep an eye on junior in the playground). Just as the share of private school children has grown, so has the share of private school teachers. Usefully, Proctor addresses this by showing a breakdown by school type:


So it is true that government school teachers are more likely to send their children to a government school than are Sydney parents in general. Overall, 58% of Sydney children attend government schools, but among the children of female teachers, the share is 63%, and among the children of male teachers, the share is 75% (we know that men are in general more likely to join a union, so I wonder if this is partly male teachers being more ideological about where their children attend school?). Nonetheless, with about one-third of government school teachers choosing private schools for their own children, you have to wonder how well these ‘public is best’ advertisements play with public school teachers.

Of course, these data are 2001 Sydney figures, so it’d be neat to have some national analysis using HILDA or the 2006 Australian census. It would also be useful to know what happens when you use individual-level data and hold constant parental education, household income, region, etc.

This entry was posted in Economics of Education. Bookmark the permalink.

5 Responses to Do school teachers send their children to government schools?

  1. John Quiggin says:

    Again I think you are misreading the results, or at least that your interpretation reflects your priors. As a general statement of the results how about Private school teachers are about as likely to send their kids to public school as vice versa

    On the analysis, I think income/SES would explain a lot
    (i) Teachers are more likely than average to send kids to private school because their incomes are above average
    (ii) Families where the mother is a teacher probably have higher incomes than families where the father is a teacher (more likely to be two-income for one thing).

  2. Andrew Leigh says:

    John, your italicised statement is certainly correct, but I’m not sure where you think I’ve misread the results.

    And yes, I too would like to see some regression analysis, though my priors are weaker than yours.

  3. John Quiggin says:

    “Misreading” is too strong, but I think the discussion reflects the way the debate has been framed. After all, isn’t it equally interesting that a substantial proportion of private school teachers use the public system for their own kids?

  4. Andrew Leigh says:

    Yep, I agree that’s interesting. Another one that one would want to pursue with microdata is the combination of declining Catholic allegiance on the Census religion question, alongside rising use of Catholic schools.

  5. Nick Biddle says:

    This is an interesting topic for me (both my parents were teachers and I went to a comprehensive public school in Western Sydney).

    I did some reasonably simple regressions using the 5% Census sample with the probability of attending a non-government school (conditional on being a school student) as the dependent variable.

    After controlling for age, gender, geography, Indigenous status, mobility, English proficiency, family composition, household employment and home ownership (but not income):

    – Kids who lived in a household that contains a person working in the ‘Preschool and School Education’ industry and the government sector are LESS likely to attend a non-government school;
    – Kids who lived in a household that contains a person working in the ‘Preschool and School Education’ industry and the private sector are MORE likely to attend a non-government school;
    – The base case is a kid who lives in a household where no one works in the ‘Preschool and School Education’ industry.

    I only did household level as opposed to family level analysis as the latter would be it’s own research project. But the back-of-the-envelope calculations seem to support John’s guess that SES is a key part of the story.

Comments are closed.