Should public schools be privatised? (Full discussion in one thread)

If you’d prefer to read the discussion with Andrew Norton in one go, here it is in chronological order.

(Thanks to JG for the suggestion.)


Andrew Norton is a researcher with the Centre for Independent Studies, and works in the office of the Vice-Chancellor of the University of Melbourne. He has formerly worked as an adviser to a Liberal Minister. Andrew Leigh is an economist at the Australian National University, and has formerly worked as an adviser to a Labor Shadow Minister.

Despite hailing from opposite ends of the political spectrum, the two Andrews link to and comment on each another’s blogs on a regular basis. So we thought it might be enjoyable for us (and perhaps interesting for others) to debate one another over a topic that we disagree about.

The idea of the discussion is that it’s a bit like the two of us sitting down to breakfast. So each morning, one of us will put up a post. At the end of the conversation (probably six days), we’ll open it up to comments. Think of it as a radio chat with talkback at the end, if you like. All posts will appear on both our blogs.

We decided to kick things off with the topic “Should public schools be privatised?”. Andrew Norton begins with the case for the affirmative.

Day 1

Hi A.L,

According to Australian Education Union election advertising, we need a federal government that will put public education first. But do we need public education at all? Would there be anyone calling for it, if we did not have it already?

People are used to the idea of state schools, so they don’t think about how uneasily government-controlled education fits with liberal democracy. If someone said that Australia’s media should be owned by the state, with journalists told by the state what they should say, with media audiences examined to make sure they had absorbed the official line, there would be predictable and justifiable outrage.

Yet public education means essentially that for Australia’s young people. The government owns most schools, employs most teachers, tells them what to teach through state-set curricula, and examines students to make sure they have it right—even kids escaping to private schools can’t avoid these last two aspects of state-run education. And unlike state-owned media, there are severe consequences for ignoring state education.

Across the political spectrum, activists want to use public education to influence young minds. In his book Dumbing Down, Kevin Donnelly documents how left-wing academics and teachers shape curricula to fit their political agenda. In government, the Liberal Party proposed a national history curriculum, which was widely seen as another front in the so-called ‘culture wars’.

Rather than fostering social unity, as some of its supporters claim, state-controlled education is a source of division and nastiness. Instead of allowing different groups to devise their own curriculum, and letting parents choose between them, we fight over a common curriculum. The public education lobby stirs class and sectarian resentment in its attempts to take funding from private schools.

And what is it, can you remind me, that makes public education worth preserving?



Day 2

Dear A.N,

I’ve enjoyed your writings on education for some time, so am chuffed to be discussing perhaps the biggest issue in education policy: should we have public schools at all? I’m also pleased for another reason – after a year of getting beaten up by the AEU and ACER for my work on teacher quality, I’m finally lining up with the comrades on an education policy issue.

Our question concerns one of the great puzzles of public finance. Across the globe, there is huge variation in whether governments play an active role in banking, airlines, pensions, and even health. But so far as I am aware, every government in the world runs a large share of the schools in that country. As Julius Sumner Miller (a privately-paid educator, you’ll point out) used to say on Australian television, “Why is it so?”

You suggest one answer: governments use public education as a means of indoctrinating their citizenry. I don’t deny that this can be important. My mother – an educational anthropologist – wrote her PhD thesis on the way in which the Indonesian government used schools in Aceh to indoctrinate young Acehnese minds into the belief that their identity was as Indonesians first, and Acehnese second.

Still, indoctrination isn’t all bad. I’d like to live in an Australia where children shared a basic understanding of democratic values, and understood our geography and our history. I’m more confident that public schools will achieve this than I am about private schools. Sure, lots of people have been fighting over what should be in the school curriculum, but we’ve also been hotly arguing about refugee policy and water policy. Sometimes conflict is a sign that an issue matters.

You seem to be worried that the agenda is being hijacked by a particular interest group. It’s true that the kinds of people who self-select into the teaching profession are not politically representative of society as a whole. But in some sense that’s inevitable (beyond politicians themselves, it’s hard to think of an occupation whose members are politically representative of the broader community). Moreover, when I’ve looked at surveys that measure the attitudes of teachers in public and private schools, I can’t tell the two groups apart – so it’s not clear that scrapping public schools would leave you with a more politically centrist group of teachers.

What’s interesting to me about your position on this issue is that you’re not arguing that the government should stop funding schools. Implicitly, you accept that there are sound economic reasons for governments to pay for schooling (as James Heckman likes to say, the biggest credit constraint is the inability of a bright child to buy good parents). So you’re either arguing (1) for the private school funding to be raised to the same level as public school funding, or (2) for the government to get out of the schooling game entirely.

Fortunately, we have an interesting natural experiment of reform (1). In 1981, the Chilean military government passed a law (or whatever dictatorships do to put things into effect) that gave the same per-child funding to non-government and government schools. Fifteen years later, 62% of kids were still in government schools. This suggests to me that when they can vote with their feet, most parents will still choose government schools. Perhaps many Chilean parents didn’t even have a choice. In remote areas, government schools have distinct economies of scale, since they can rely on the central bureaucracy for administrative support. In the Australian context, would the private sector really set up a public school in the Cape York community of Arukun in exchange for $11,000 per child?

Public education is worth preserving because it helps engender shared knowledge and values; because a public system guarantees access for all children; and because its economies of scale will often make the public sector more efficient than the private sector.

Yours, publicly,


Day 3

Dear A.L.,

Periodically, our politicians rediscover civics in schools. In 1994, the Keating-appointed Civics Expert Group released a report supporting civics and citizenship education. One of their reasons was that a survey, which they had commissioned, revealed widespread ignorance of our political institutions.

Two things are important about the report’s date. The first is that it was published about 120 years after education became free and compulsory, yet neither attribute has yielded much political knowledge. The second is that was published about 140 years after democratic institutions were established in Australia, and 90 years after universal suffrage. Australia is a long-term successful democracy despite the absence or failure of civics education in our public schools. 

While understanding political institutions can do no harm and would probably do some good, you don’t need it to acquire democratic values. The citizens of stable democracies like the US and the UK show similar levels of ignorance to our own. Democracy is so deep in the culture as to not need teaching or defending in principle, even while we argue endlessly over the detail. Children much too young to participate in elections intuitively understand that voting is a fair way of deciding things.

There is no reason to believe that private schools would challenge this democratic ethos. Surveys show that current ex-private school students are more active in political affairs and more strongly in favour of democratic rights than those who went to government schools. (Though generally there are only minor opinion differences regardless of school background.) In my view, preserving public education to teach civics is a non-solution to a non-problem.

Moving from general principles to practical details, I don’t favour your option (1); the conventional voucher scheme. Liberalising 100% of demand but only 30% of supply would only deliver a faster version of what we have now, of government schools losing 0.3% or 0.4% of market share a year. We need to liberalise supply as well, by making the current public schools independent. To achieve economies of scale schools could form associations or join chains; the Catholic systemic schools already have lower per student costs than governments schools. The important thing is accountability to parents rather than to central bureaucracies and politicians.

You wonder whether the private sector would set up a school in Cape York, which seems to be your feasibility objection to option (2), the government getting out of educational delivery. There is charity in the private school sector—indeed, donated money is already rescuing some Cape York Indigenous students from their abysmal public schools—but I agree that’s not enough. We need the right price signals.

Which brings us to the tricky issue of finance. As you say, some parents cannot afford proper schooling for their kids. On the other hand, lots of parents can, and do  already, spend significant dollars on their children’s education. The latter group count against conventional voucher systems. It would cost $5 billion in additional taxation to fund them on the same basis as government schools.

Instead, we could fund all schools on a similar basis to private schools now, according to parental SES background.  That would lead to reduced rather than greater government expenditure, with tax cuts helping parents finance higher private outlays. Schools servicing the most disadvantaged areas would get the most money, providing what was necessary to make private schools viable. Schools in the most affluent areas could be taken completely out of the public funding system; this if nothing else about my proposal would please the AEU.


Day 4

Dear A.N.,

Let’s be careful about what we claim for private schools. The fact that private school attendance is positively correlated with civic activity doesn’t tell us anything about the causal impact. It could be that private school kids have rich parents, don’t move house as often, or any of a dozen unobservables. Just as the positive correlation between AWAs and wages doesn’t provide any useful causal evidence, so too correlation doesn’t mean causation in the case of public and private schools. My guess is that if we take a given child, she would be no more civically active if she attended a private school than if she attended a public school.

While I don’t think private schools have a special gift for turning kids into what Robert Putnam would call “social capitalists”, neither do I think that the typical private schools does a worse job of teaching tolerance than the typical public school. That’s basically why I support private schools getting government dollars: the per-child government funding to private schools is only about 55% of the per-child funding to a public school, so the taxpayer saves 45%. It may be the case that private schools have some positive externality (Caroline Hoxby argues that competition from private schools has the potential to ‘lift all boats’), or negative externality (if private schools skim the cream, the peer effects in public schools may go down), but I basically think that the best thing about private schools is that every kid who goes there saves you and me about $5000 ($11,000-$6000).

As I understand it, you’d now like to fund private schools on the same basis as public schools. The problem with this is that it takes away the best argument for private school funding. You now have to debate in an area where we have very little evidence either way. So far as I’m aware, there’s not a single Australian study that estimates the competitive benefit of private schools on public schools; not a skerrick of research that tries to value the lost peer effects for public schools. So all one can then do is to fall back on general principles, claiming that the private sector works better than the public sector. Well, maybe it does, but if you’re arguing that we should transform school funding, it would be nice to have more than theory on your side. Given the current level of private sector investment in Cape York, it’s hard for me to have much faith in private school entrepreneurs to set up and run schools there.

But let me return to the one point on which I agree with you: making the SES funding formula more progressive. At present, private schools are funded based on the average income of the postcodes where the parents come from. But there are plenty of rich parents living in poor neighbourhoods. So in Imagining Australia, we argued that a better-targeted scheme would be to ask parents for their taxfile numbers, and target private school funding based on a five-year average of parental income. We could make the ‘voucher amount’ transparent to parents (as we don’t at the moment), so as to encourage schools to charge differential prices to low-income parents. Your proposal would take private schools in affluent areas out of the government funding system. Mine would take schools with rich parents out of the government funding system. The difference is subtle, but important, methinks.

My own work has been pretty critical of what economists call ‘school productivity’. It worries me that education policymakers in Australia tend not to report outcomes, are reluctant to deviate from uniform salary schedules, don’t look at empirical evidence when it comes to choosing the right curriculum, and in many cases seem wedded to class size cuts (a policy outcome I think has minimal benefit once class sizes get below 30). I suspect you would agree with me on the diagnosis, but unlike you, I’m pretty optimistic about the ability of public schools to reinvigorate themselves. When it comes to delivering educational opportunity to the most disadvantaged, the private sector doesn’t have much of a track record.

Yours, in solidarity,


Day 5

Dear A.L,

As you suggest, former private school students might show more civic attitudes and behaviour than former government school students, but that doesn’t show that the school is the cause. Identifying causes requires much more research than I have done. I mentioned this finding not as an argument for private schools (it may or may not be that), but because you suggested that public schools provide a shared ‘basic understanding of democratic values’. I say, consistent with the evidence I think, that we will get that either way. Civics is a neutral factor in this debate.

If you are overstating the importance of civics in choosing a school system, I think you are understating how radical my proposal is. I don’t want to fund private schools on the same basis as public schools; quite the reverse I want to fund what are currently public schools on the same in-principle basis as private schools are now, according to their students’ socio-economic background.

I’m open to argument on the best measure of socioeconomic background. The current system uses proxies based on where the student lives, you favour more accurate ATO income statistics to create—if I read you correctly—personalised, income-tested vouchers.

The ATO may be the best original data source, but I prefer funding schools on an average basis, rather than via vouchers. I have the usual concerns about the work disincentives linked to means testing, but from an educational perspective individualised funding would encourage parents who must pay high fees anyway to send their kids to a high-fee school in an affluent area. That would replicate the lost peer effect problem that you mention.

Unlike the public school lobby, I don’t think kids should be conscripted into providing peer effects for their classmates. But if they voluntarily attend a school with a lower average SES rating their own family’s, everyone may end up better off. Each group takes advantage of the other group’s socioeconomic status.

Consequently my proposal, as you say, wouldn’t take all rich parents out of the government funding system. But it is still likely to take a lot more out than your plan to keep the public school system. Extrapolating from ABS estimates of school attendance by family income, about 180,000 students from high-income families attend government schools, not so far behind the 223,000 students from these families in private schools. Under your system, high income families in government schools get much higher educational subsidies than low SES families in private schools—about 186,000 of them. Unless there is some very compelling reason for encouraging public schools, this arrangement seems regressive and inequitable.
Obviously, I don’t think there is any such compelling reason. I think my system could capture whatever positive features public schools have. I would privatise all existing government schools, but there would be no need to abandon positive aspects of the public school ethos, such as a special commitment to the disadvantaged. Schools that did that would get the highest funding rates, enough to keep education free. But no schools would be stuck with the bad aspects of public education, such as protection of incompetent or unsuitable staff, poor curriculum, and excessive centralised control.To achieve economies of scale, as I noted earlier in the week, schools could form chains or join associations of like schools. National chains or associations could create national curricula, without going down the monopoly curriculum path that, alas, both political parties are taking us. With competition between them for students, they would want to prove that their curriculum is the best—perhaps they would, as you hope, look at the empirical evidence. If some chains or associations could convince parents than small classes are ineffective, they would claim market share by offering lower fees or spending fee income on something else.

In the end, you take a conservative approach to schooling. Like a conservative, you adopt the circular argument that if it has never been done before, and so has no historical evidence, it should never be done at all. Like a conservative, you are confident that existing institutions can adapt. But where is the evidence for that? Public schools have been in ‘crisis’ for as along as I can remember, despite every Labor government coming to power promising to make a difference in public education. Another did so last month. Yet already we can see the seeds of another failure, with most of the money labelled as ‘education’ being blown on more family welfare through the education tax refund, and not a word on teacher quality.



Day 6

Dear A.N.,

What a brutal final paragraph! So if I don’t support your plan, I guess that makes me a conservative who doesn’t care about teacher quality. Few arrows could have better found their target.

I’m pleased that you and I have found common ground on civics, at least. I began by claiming that public schools provided a common crucible. You said that public school civics was in crisis. But since neither of us could provide empirical evidence, we weren’t willing to stand fiercely by our claims (incidentally, I think this is why people like John Quiggin and I enjoy arguing with you so much more than we enjoy arguing with many people on the right of the political spectrum). Perhaps one day, policymakers will have some believable causal estimates of the impact of school type on some set of ‘good’ civic indicators. In the absence of that, only the firebrands will be able to stay passionate about that one.

So now we’re left with the more traditional economic arguments. You take the view that privatised public schools will be more efficient and more equitable. I think the opposite is true on both counts. Whether it’s for political or economic reasons, companies like Edison Schools that have set up to run large numbers of schools have done very badly. According to their Wikipedia entry, their costs are higher than the public system, and they have still failed to make a profit in all except one quarter of their decade-long existence. A result like this makes me concerned about the viability of such a model in Australia. If private schools have one-third of the market when they get 55% of the government funding, can we really be confident that they could make a viable go of it in the rest of the market with 100% of the government funding? Shocking as it may sound coming from an economist, I think there may be some things that governments do better than markets, and one of them is running schools.

On the equity argument, I don’t think that the high use of a public service by the rich automatically suggests we should privatise it. My guess is that roads are used more by the rich than the poor, but I think the government has a core function in roadbuilding. National defence and policing are also of more benefit to the rich, since they have more to lose from civil unrest. But that doesn’t make me think that privatising them would help the poor. The good burghers of Rose Bay would find it much easier to set up their own private police force than those in Rooty Hill.

Related to the equity argument is the question of peer effects. You say that “Unlike the public school lobby, I don’t think kids should be conscripted into providing peer effects for their classmates.” Unlike either of you, I think the answer depends on whether peer effects are nonlinear. If the benefit that high-ability kids bring to low-ability kids is much larger than the cost that low-ability kids impose on high-ability kids, then perhaps we would want to dragoon high-ability kids into underperforming schools. (I’m reminded here of a comment that my thesis supervisor Christopher Jencks once made about neighbourhood effects: ‘it’s odd that policymakers spend a lot of time thinking about how to subsidise poor people to live in middle-income suburbs, when it might be cheaper to pay a few rich people to live in poor suburbs’.)

But I still haven’t removed your most wounding barb – that my nihilistic position leaves me as a conservative who doesn’t care about teacher quality. And I guess that here I can only fall back on my faith that governments eventually do the right thing. It took most of the twentieth century before the progressive side of Australian politics finally realised that surrounding ourselves with high tariff walls hurt the poor more than it helped them. Part of the challenge was that organised labour represented producer interests, while consumer interests were disparate and disorganised. It took a succession of tariff board reports, industry restructuring plans, and a lot of talking before a Labor government finally took action. When it comes to education policy, I think the same is true today. The voices of teachers are much more forceful than those that advocate for children’s interests. Sometimes, what’s good for teachers is good for kids. But not always.

You’re right that those looking for bold ideas on teacher quality have not exactly been deluged by innovative proposals in this campaign. Nonetheless, to simply sell off all schools would effectively give up on education policymakers. By contrast, I’m still optimistic enough to believe that just as economists’ ideas prevailed in the domain of trade policy, the same can happen in education policy. As Keynes famously wrote, “the ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else.”

Thanks for the discussion. It’s been a pleasure.

Yours, idealistically,


If you would like to comment on the discussion, please go to this thread.

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