Felix Hathaway
Published in post

The Case for Choice

Few subjects excite more anxiety among the UK’s chattering classes than education, and nowhere more so than among parents scrambling to get their children into a good school. In fact it can even start with nursery places. Parents worry over primary schooling and they worry over secondary schooling. The scale of the perceived issue can be seen in how much they are prepared to spend on it. Consider private schooling. According to the Independent Schools Council’s (ISC) annual census, the total in private fees paid at ISC schools in 2016/17 can be estimated at 8.7 billion pounds. If the same average costs per pupil are true of the other 20% of pupils in independent schools, the figure for total fees is over 10 billion pounds. The number of children in independents however is only 7% of the total. The bulk of education spending comes via the government, estimated by the OECD in 2014 to be 4.2% of GDP, excluding Universities.

However, this isn’t the only measure of educational insecurity among British people. The figures exclude spending on private tutoring for students, the average costs of which are £36 an hour for A-level tuition. Additionally, a house near a good state school commands an average premium of more than £18,000 in the UK (DfE, 2017). People are clearly prepared to move, spending tens of thousands of pounds, if there might be a possibility of a better education for their children. My own parents became regular church attendees for the sake of getting my youngest brother into their preferred primary school. Other parents take things completely into their own hands and homeschool, with local authority data showing that numbers have doubled since 2011. With so much effort and money going into securing an education, what are people worried about?

PISA, or the Program for International Student Assessment, is a triennial survey that aims to evaluate education systems by testing the skills and knowledge of 15-year olds, and is generally considered a reliable measure of attainment. These tests, measuring science, maths and reading, show the United Kingdom consistently in the middle of the pack since 2006. However this basic analysis obscures the concerning disparities within the UK. Digging further into the results, it emerges that the gap between the best and worst performers in the UK is higher than in almost every comparable nation, with the least able more than 8 years behind by the time they’re fifteen. Equally concerning is the finding by the OECD in 2013 that the UK was the only advanced country in which literacy and numeracy among 16 – 24 year olds was lower than that among those between the ages of fifty-five and  sixty-five. Furthermore, there is a litany of other issues from bullying to the content of the curriculum that can cause parents to shy away from many schools in Britain.

Whatever the concerns however, politicians have got the message: education matters. Glancing through the manifestos at the last election, “Towards a National Education Service” is a headline in the Labour manifesto with promises to boost spending, cut class sizes and invest in new buildings. The Conservatives promised an extra 4 billion over the next parliament, and an ending of the ban on new state selective schools. Such promises aren’t particularly new. Government has been reforming education for as long as many people can remember. Real education spending has increased fivefold between 1960 and 2010. Future governments will doubtless continue to prioritise education and try to improve things. The question is how. What are the alternatives, and what works?

Since the introduction of mandatory schooling in 1870, education policy has been characterised by a series of increasingly top down reforms. The definition of what constitutes an education has been made stricter, mandatory schooling has been extended to 16, education enrollment to 18, and the national curriculum has been in place since 1988. Someone has to choose what policies to enact, so why not the experts? After all, we consult doctors on our health, architects on our homes and bankers on our savings. The idea that education is too important not to be centrally managed by the state has become ingrained in both main political parties, as well as among aid and development agencies.

Recently however, the dominance of this model has increasingly been challenged by evidence from developing nations. James Tooley, in his 2009 book the Beautiful Tree, describes networks of low cost private schools serving the very poorest slum dwellers across the developing world, and challenged the narrative among development agencies that education had to be provided by the state to be both universal and of a good standard. In 2017 he revisited the subject in “Education: War and Peace.” The results from war torn countries found that children in these low cost private schools typically outperformed those in government schools, and that private schools were notably more efficient as measured by cost effectiveness. Nor were they particularly more expensive for the parents than the state alternatives: findings from Liberia demonstrated that, when the full range of costs including transport and uniforms were included, government schooling was only 25% percent cheaper than the low cost private alternative. Although low cost private schooling is not widespread across the developed world, the same findings of increased efficiency and higher attainment are born out both by detailed studies, such as those examining the Swedish education system (Sahlgren, 2010), and the annual UK league tables.

But even if top down state provision is flawed, does it give equal opportunities to all participants? Is it fair? This is the other half of the case made against the unorganized provision of education as a service provided through a market. What happens to those children whose parents can’t afford, or aren’t motivated enough, to send them to the better schools? Speaking to a teacher at a well-regarded London comprehensive, her primary objection was summed up as a concern that differentiation between schools, particularly allowing them to control their own admissions, would simply lead to ‘cream skimming.’ This is where those students with strong academic skills, or motivated parents, the pupils who would ‘do well anyway’ are creamed off by selective schools, which then get better results. The rest of the pupils are then ‘left behind’ in schools which struggle to maintain discipline and where they are consigned to underperformance and failure. The archetypal example of this, ironically enough, is the top down, tripartite system of grammars and secondary moderns (alongside secondary technical schools) which operated in England and Wales between 1945 and its formal abolishment in 1976.

The exact effects and impacts of the tripartite system are still being debated today, but the question of whether the state can effectively institute educational equality of opportunity can be answered relatively simply. It cannot. As touched upon above, there is already a huge amount of differentiation in the state system based upon the income and motivation levels of parents. Just as with rationing in the second World War, or prohibition in the 1920’s, one cannot simply abolish a market without creating incentives for black markets or other alternative provisions of the good or service in question. In the case of education, much of the displacement has been towards housing and private tuition. Motivated parents still want a say in their children’s education and are still prepared to pay for it. The enduring popularity of private schools demonstrates the same fact, with the percentage of parents who would send their children to private school if they could afford it consistently polling at 50% or above. This is despite having to effectively pay twice for education, not to mention many people’s aversion to it on grounds of ‘equality.’ The net result, is that not only is it morally wrong for experts in government, local or central, to dictate where parents are to send their children; it is impracticable. It is only the poor parents who can be dictated to effectively, as those can afford to will opt out where they believe their children’s interests would otherwise be damaged. The problem of ‘cream skimming’ is still a real one, and we will return to it, but there is no reason to believe that a top down solution could solve it without having to massively curtail the freedoms of parents to choose where they live or to opt for private education.

If centralized control by educational authorities isn’t the answer to our problems, then what is? Someone still has to choose where and how children are to be educated. If not the ‘experts,’ what about the parents? As shown above, when people are prepared to pay for something, then it’s almost impossible to stop some sort of market being created for it’s provision. But should market involvement be kept to a strict minimum, or should we give it free reign? Obviously, there isn’t a strict dichotomy here, but what is the case in favour?

The concept of spontaneous order was one of the most significant ideas to emerge from the Enlightenment. The argument was that top down control and coordination were not in fact necessary for the creation of an ordered system; structure, with all it’s appearance of design could arise naturally from a process of evolution by selective experimentation. This idea is broadly accepted in sciences such as biology, but it also applies to institutions such as language, the UK constitution and the market economy. A concrete example is in the provision of food. No one person plans the distribution of sandwich outlets in London, or the items on their menus. Instead, where there is a perceived opening in the market, a vendor will set up shop. If their offering is attractive enough, they succeed and can potentially expand. As no stockist can compel customers to buy from them, the only way that another sandwich vendor can survive and compete is by offering a better, cheaper or somehow differentiated product. This way if I desire club sandwiches and you desire prawn mayonnaise, there is an incentive for sandwich makers to provide for both of our desires. No one person’s tastes dictate the entire supply. People get what they want, provided they are prepared to pay for it, and there is a constant incentive to improve both the cost and quality of provision, as doing so provides the opportunity to make a profit.

In the realm of education, there is a serious reluctance to concede that education provision could function in the same way as the sandwich market or the supermarket system. People are concerned about whether poorer parents could afford to send their children to schools that were entirely private, and there is an almost intuitive distrust of the profit motive. However there have been proposals in the UK, and around the world that move education in a more market-oriented direction. One of these is the idea of vouchers, where the state ensures that there is a certain mount of money available per child to spend on education but, within certain strictures, it is the parents of the children who decide where and how it is spent. The idea is to restore the link between consumer and producer that is broken when the government both provides and allocates schooling. In the UK, the creation of academies and free schools has attempted to reestablish those links. These have been largely modelled on the highly successful Swedish system (Sahlgren, 2010).  The government provides the bulk of the funding in the same way as for a state school, but the schools control their own admissions, hiring, teaching and curricula. The aim is that by forcing the schools to respond to what parents, not politicians, want, there will be greater innovation in the schools and a drive for higher standards. The question is: do they work?

Maybe. In the UK there have been two major rounds of work that are methodologically robust. The first, by researchers from the LSE, examined the converter academies created by Labour between 2000-01 and 2008-09 (Eyles and Machin, 2015). These researchers used the pupils at schools which later converted to form a control group with similar characteristics and tracked the performance of the two cohorts. The study found that these early converter academies had a significant impact on pupil attainment that increased over time as the pupils remained in academies. A later study and its follow up work (Andrews, Perera et al., 2017) examined the impact of the academies set up under the coalition government between 2010-11 and 2013-14 under the same methodology. Similar work was done on primaries that converted during the same period (Eyles, et al., 2016). The conclusion of these studies was that academisation made no noticeable difference to schools in the long run, other than those already graded outstanding which showed continual improvement. Several explanations for the discrepancy in impact have been suggested, from problems with sample selection to the initial batch of academies taking greater advantage of the increased flexibility.

One thing that is clear is that while some academies and chains have struggled others, such as the Harris Federation and Ark Trust, have been remarkably successful. The promise of giving schools more autonomy while allowing parents more say in where they send their children was that there would be greater experimentation, and that good schools would drive out the bad. The experimentation is taking place and so perhaps it is just a matter of time before standards consistently start rising. However, progress is slow. One of the potential issues behind this is that much of the operations of free schools and academies are still tied to central government. Visiting Fulham Boys School, a free school in west London, the founder told me that his biggest headache was dealing with the Department for Education to get approval for what buildings they could use. Paradoxically, this is also something that opponents of free schools complain about when they talk about the ‘extra resources’ given over to them, which almost inevitably go into acquiring buildings to create new school places and would be a fixed upfront cost for opening any new school. A solution both to the problem and the complaint would be allow the entry of for-profit groups into the market. Because of their ability to access equity markets, these would have a much easier time raising funds, and so would not have to depend on government patronage for their facilities. Additionally, the drive for profits to satisfy shareholders could potentially turbo charge the set up of new schools and chains, as appears to have been the case in the Swedish system. Some of these would inevitably fail, however there would be real incentives for the successful ones to expand and to adopt the best practices of the schools that succeed.

A further move that could increase the pace and scope of the reforms would be to extend the program into a full voucher system. This is the idea, that parents should be allowed to spend the state allocated money for their education in any school of their choice, even an independent school, where they might also pay top up fees. Programs of this type exist in numerous countries around the world, with successful to mixed results which researchers agree largely depend upon the structuring of the system. Successful examples are the program in Columbia, where those selected by lottery to enter the program were 10% more likely to have finished 8th grade (Angrist et al., 2002), and systems in Europe such as the program in Estonia. Researcher Patrick Wolf concludes from his US studies that “More gold standard studies have validated the benefit of private school choice than any other major education reform.”

Some programs however appear to have done little good and possibly positive harm. The best-known example is the Louisiana Program where researchers found that students who transferred scored 24 percentile points lower on maths tests after their first year at a new private school (Mills and Wolf, 2016), similar studies on other US programs also suggest drops in math scores on switching to a new school at least during primary grades. There is some recent evidence that the schools are adjusting and that pupils are making up lost progress in later years (Mills and Wolf, 2017), as well as debate over the extent to which the differences are due to the stresses of switching schools.

A suggestion that might avoid some of the dangers of lost performance for children in the early grades (where the effects appear strongest) would be the introduction of vouchers initially at the college and 6th form level only. Beyond anything else, this would simply be honest; currently the Department for Education only provides funding for 13 ½ hours a week of contact time for A and AS level students. The head of Harris Westminster Sixth Form, a high achieving selective free school in London, described this as “the biggest lie in education.” His school offers 23 hours a week of guided learning on a budget only 20% higher per pupil, with the extra money largely coming from private sponsors. Acknowledging that the government funding in reality only covers part time 6th form education and allowing it to be topped up privately would allow more pupils access to the quality of learning enjoyed by many of their peers. The money could also be placed directly into the hands of the students, who can already make significant decisions about their lives including getting married or enlisting in the armed forces, provided there’s parental assent. Current pupils could instead choose to spend their voucher on out of hours education that could fit in around a job.  Rather than parents or the state performing a duty of care, could they not be made more responsible for their own education?

Finally, the argument that private and selective education undermine education at comprehensives through “cream skimming” is largely defused at this age. Those pupils ‘left behind’ are no longer being forced to remain in school, and laws forcing students to be in some sort of training up until 18 could also be rescinded. Rather than looking at vouchers for primary and preschool education, as many programs do, potential students could be placed in charge of their own education as young adults. If they make the wrong choices that would still be tragic, but their decisions and the consequences would be their own. That link between choice and responsibility is the key element in a well-functioning market; allowing students to take responsibility, subject to basic parental consent, would allow an expansion of choice with providers encouraged to serve the real customers: pupils. Such an approach would be “bottom up” and could harness the power of the market system to meet the needs and wants of the next generation.



The author would like to acknowledge Alex Wade of Fulham Boys School and James Handscombe of Harris Westminster Sixth Form for agreeing to be interviewed, as well as Toby Young for providing me with up to date data on academy performance.

Felix Hathaway


  1. March 22, 2018Reply
    Peter Sullivan

    Great article! Really interesting read!

    In a token system would they be divvied out equally or for equity?
    How does the education system work in those countries with the highest rankings?

    • March 22, 2018Reply
      Felix Hathaway


      My preference in any sort of voucher system is for equality (everyone gets the same voucher) rather than aiming at some sort of equity (by which I suppose you mean some groups getting more or less so that the end amount of money they have to spend is more even). This is based off of a general dislike of governments using policy to favour certain groups at the expense of others, as well as a preference for simplicity. It’s worth noting however that in terms of ‘equity’ independents already offer a huge number of scholarships and supported places, so I would also question whether a differentiated voucher system is neccesssary.

      As to countries with high ranking education systems, I think there are a few things that are fairly common. I can’t pretend to be an expert on education globally, but to put it simply:

      – Choice / differentiation (many of these countries have voucher systems, or features that drive competition between schools e.g. being very high density) See the Estonian voucher program, or Direct Subsidy Scheme schools in Hong Kong. This often extends to selective schooling.

      – Widespread instance of parents spending thier own money. Often through private tuition, as another way of ‘topping up’ school fees. Also though direct payments to schools, for example in Singapore, where primary education is free, but parents commonly spend money at (heavily subsidised) secondary schools.

      – ‘Whole class teaching’ or ‘Direct instruction’ as opposed to ‘child centred learning.’ That is, they often have slightly higher student to teacher (and teaching assistant) ratios, and more time is spent learning from the teacher at the front of the class in rigidly planned lessons. Outside the scope of this blog, but has been convingly shown by US research in ‘Project Follow Through’ to be the most effective teaching method.

      There are almost certainly a bevy of other considerations including culture and relative wealth, however they cannot readily be controlled by policymakers.

Leave a Reply

Your email address will not be published. Required fields are marked *