Sunday, December 2, 2012


School choice and accountability: caveat scholacticus

The choice debate
In the English city of Guildford in 2011, every final-year student in the Royal Grammar School earned at least three A-levels, the highest secondary-school subject qualification. The equivalent figure for the city’s Kings College for the Arts and Technology was just 69%. Neither figure was a surprise, nor is such variation exceptional. In most places, it is simply accepted that specific schools – like individual teachers – have different results which tend to persist over time. A natural conclusion is that giving parents, and through them students, the ability to choose better performing schools should lead to better outcomes.
Unfortunately, this issue is far more complex and not just because of the range of systems through which choice operates across the world – including both publicly and privately funded options. Whatever their specific strong and weak points, all these arrangements need accurate information. Getting it wrong can be harmful. A study of Beijing parental selection of primary schools found that excessive optimism about place availability at better schools led parents to use up application choices on schools that were already full. Less optimistic parents snapped up places at the next tier of schools, leaving only markedly worse ones for the children of those making the initial mistake.[10]On the other hand, researchers in North Carolina found that better, clearer information on local schools increased the number of low-income parents taking advantage of school choice, and that the children so placed performed better.[11]  As in any quasi-market, for choice to work, schools have to reveal how well they are doing: choice and accountability must go hand in hand.
Chart 5: School choice in selected countries, aggregated score, 2009
Chart 5: School choice in selected countries, aggregated score, 2009

Note: The score, which is on a scale of 0 to 1, is an aggregate of the following indicators: enrolment choices (freedom of enrolment choice at primary and lower secondary education), the level of school choice (% of pupils living in an area with more than two schools), parental expectations, and financial choice and information (availability of school vouchers and government responsibility for informing parents on school choices (primary and lower secondary).
Sources: Economist Intelligence Unit and OECD.
Any accountability system, however, requires some decision on what should be measured. Demographic differences between the children in the two Guildford schools above might explain the gap in results far better than the education provided. Mr Cappon notes of Canada: “Social class and school choice tend to go together.” Indeed, much of the choice and accountability debate continues because such other issues cloud the picture.
Recent research suggests that, at the system-wide level, the potential for informed choice helps raise educational outcomes and reduces costs. In particular, a cross-country comparison of the number of private, often faith-based, schools – an indication of the degree of choice – with the 2003 PISA results found that, even after controlling for other factors, “the share of schools that are privately operated has an economically and statistically significant positive effect on student achievement in mathematics, science, and reading.”[12] The benefits were greater than average for students with a lower socio-economic status where such private schools were publicly funded, as in Belgium and the Netherlands. Professor Woessmann, one of the authors, explains: “If there is more choice for parents, and more non-governmental school operators so that schools are not managed by one big state monopoly, countries perform much better.”
How this choice drives the system to better results in practice, however, is a matter of no little debate. Indeed, any discussion involving market-like mechanisms and education inevitably leads to contentious, often politicised, debate. Unfortunately, the resultant heat has shed little consistent light.
Vouchers and charter schools
Some of the most investigated choice initiatives operate in the US. Voucher programmes provide funding – generally assigned by lottery as the programmes are almost invariably oversubscribed – that pay for the private education of underprivileged children. A 2008 review by Patrick Wolf, Professor of School Choice at the University of Arkansas, looked at the 10 best studies of these programmes and found widely varying results.[13] In general, all or some students who used vouchers did better academically in certain fields, especially maths. A more recent study by Mr Wolf of the long-standing Milwaukee voucher system brought further variability: voucher students there outdid peers in reading but underperformed in maths.[14]
The impact of such programs on abilities tends to be unpredictable, but that may not be the point. Parents almost invariably are satisfied with them, although perhaps for reasons quite apart from grades. Given the public options available to some of these students, physical safety is an issue: one study found no academic differences for voucher users, but they did have lower arrest rates.[15] 
Another possible impact of choice is to create competition so that all schools improve, especially where they are made to give data on results. Debate on the extent to which this has taken place and whether competition was the driver of perceived change is also on-going.[16] The one clear point is that vouchers, and choice, do not seem to hurt existing school systems.[17]
A more widespread US experiment in using choice and accountability to improve education has been the growth of charter schools. These autonomous, privately-run but publicly-funded schools open to all students – capacity permitting – exist in 41 states. In return for autonomy, these institutions are made accountable. Charters are granted with binding requirements to achieve certain levels of academic success among students.
As with vouchers, the success of charter schools as a whole is the focus of intense debate. The largest review to date of research presents a mixed picture. The Center for Research on Education Outcomes looked at research from 15 American states and the District of Columbia. It found that, on average, students in these schools tended to do slightly worse than those in nearby public schools. But the broader message was variety: 17% of charter schools do better, 46% are just as good, and 37% do worse. Moreover, the success of the schools depends on the way they are regulated. Roughly even numbers of states had schools where students on average did better than in traditional schools and schools where students did worse.[18]
Chart 6: School responsibility and autonomy, average score, 2009
Chart 6: School responsibility and autonomy average score, 2009

Note: The score is the average of 'index of responsibility for resources allocation' and 'index of responsibility for curriculum and assessment'. These indexes have an OECD mean of 0 and a standard deviation of 1. Positive values on these indexes indicate relatively more responsibility for schools than the local, regional or national education authority.
Sources: Economist Intelligence Unit and OECD.
Dr Finn believes that greater autonomy and accountability are needed within US schools, but he also remarks that “one of the sobering lessons of the last 15 years is that hanging a sign with the word 'charter' in it on the front door does not make it a better school. In any state, some of best and worst schools are charter schools, except perhaps in Massachusetts because it only gave charters to people who knew what they were doing.”
Indeed, the wide variety is probably a predictable result of how these schools provide value. According to Professor Stecher, “the strength of charter schools seems to be that they permit innovation outside of bureaucracy, for good or for ill. The movement needs to be accompanied by careful monitoring to protect the welfare of kids, but it is leading to some really interesting opportunities and models of reform.” He cites Aspire Public Schools, a California non-profit charter school system that, even though three-quarters of students come from impoverished families, had average scores that exceeded the state’s overall mean by more than 5%.
The broader lesson seems to be an obvious one. In the words of a study by Harvard academics, “school choice can improve students’ longer-term life chances when they can gain access to schools that are better....”[19] The key, as in any market situation, is deciding which ones are: sometimes choice means opting for existing provision, but this does not negate its value.
School choice in developing countries
Where such provision is poor, however, choice and accountability can be essential. James Tooley, Professor of Education Policy at Newcastle University, has done extensive research into the huge number of unofficial private schools used by economically underprivileged students in developing countries. In many cases, rather than trusting state provision, families are willing to spend often a substantial part of their income to send children to these unregistered schools. The reason is simple: parents know that education is important but public provision is sub-standard or illusory. Professor Tooley ascribes parents' decisions in this area to their mistrust of state-school teachers, who are accused of absenteeism, poor teaching habits and poor attitudes toward students themselves.
As with any unofficial activity, it is hard to assess its full scope. Professor Tooley notes that the best data from India shows around a quarter attending private schools in rural areas, and other research indicates around 65-70% do so in urban areas. He therefore estimates the overall total at around 40% or more – a figure consistent with his own, less detailed research in communities in Ghana, Kenya, and Nigeria.[20]
These schools exist because they provide results: Professor Tooley’s research in a variety of locations has found significantly better reading, mathematics, and English skills. Similarly, World Bank-supported researchers from the Learning and Educational Attainment in Punjab Schools (LEAPS) project found that in that Pakistani state, students in such private schools were on average 1.5 to 2.5 years ahead of counterparts in government schools, even though the latter spent three times as much per pupil.[21]
What makes these private schools so much more effective is not immediately clear, says Professor Tooley. They typically have fewer resources, class sizes vary widely and often the teachers are not as well trained or do not have as much teaching experience. He concludes that “there is a missing ingredient [from public schools that exists] in private schools. It must be accountability. The teachers have to teach, otherwise they get removed; the schools need to please parents.”
The extreme situation faced by these parents gives the same message as the correlation between PISA outcomes and private-school numbers: choice and accountability can have an important impact on results. On the other hand, the experience of school choice in the US shows that the way these mechanisms work are complex, require parents to have as much information as possible and can penalise wrong choices as much as reward right ones. Rio de Janeiro’s Ms Costin points out, however, that the effort needed to bring in parents is worth it even in the poorest areas: “They are not second-class citizens. Their opinion is important. Parents know which school is a good school. Social pressure for quality can be exerted even by illiterate parents."

[10]Fang Lai, Elisabeth Sadoulet, Alain de Janvry, “The Adverse Effects of Parents' School Selection Errors on Academic Achievement: Evidence from the Beijing Open Enrollment Program”, Economics of Education Review(2009) v28 n4: 485-496
[11]Justine S. Hastings and Jeffrey M. Weinstein, “Information, School Choice, and Academic Achievement: Evidence from Two Experiments”, The Quarterly Journal of Economics, (2008): 1373-1414.
[12]Ludger Woessmann and Martin West, “Competition from private schools boosts performance system-wide”,Vox
[13]“School Voucher Programs: What the Research Says About Parental School Choice”, Brigham Young University Law Review, (2008): 415-446.
[14]The Comprehensive Longitudinal Evaluation of the Milwaukee Parental Choice Program: Summary of Final Reports, February 2012,
[15]Julie Berry Cullen, Brian A. Jacob, and Steven Levitt, “The Effect of School Choice on Participants: Evidence From Randomized Lotteries”, Econometrica, (2006), 74: 1191–1230.
[16]See: Caroline Hoxby, School Choice and School Productivity: (Or Could School Choice Be A Tide That Lifts All Boats?), 2002, NBER Working Paper 8873, an influential, article advocating this argument,; Greg Forster, A Win-Win Solution: The Empirical Evidence on School Vouchers,2011; David N. Figlio and Cecilia Elena Rouse, Do Accountability and Voucher Threats Improve Low-performing Schools?, 2005, NBER Working Paper 11597.
[17]Research on voucher programmes in Chile have produced similarly contrasting results to those in America (Francisco Gallego, “School Choice, Incentives, and Academic Outcomes: Evidence for Chile”, paper 39, Econometric Society 2004 Latin American Meetings; Chang-Tai Hsieh and Miguel Urquiola, “The effects of generalized school choice on achievement and stratification: Evidence from Chile’s voucher program”, Journal of Public Economics (2006) 90: 1477–1503).
[18]Center for Research on Education Outcomes, Multiple Choice: Charter School Performance in 16 States, June 2009.
[19]David Deming, Justine Hastings, Thomas Kane, and Douglas Staiger, School Choice, School Quality and Postsecondary Attainment, 2011, NBER Working Paper 17438.
[20] It should be noted for disclosure purposes that Pearson, who commissioned this report from the Economist Intelligence Unit, is a minority investor in a chain of schools in Ghana co-founded by James Tooley.
[21]James Tooley, Yong Bao, Pauline Dixon, John Merrifield, “School Choice and Academic Performance: Some Evidence From Developing Countries,” Journal of School Choice, 2011, 5: 1–39; Baladevan Rangaraju, James Tooley, Pauline Dixon, The Private School Revolution in Bihar: Findings from a survey in Patna Urban, 2012; World Bank, Learning and Educational Achievement in Punjab Schools Report Summary, 2008.

Returns to schooling: education, labour and social outcomes

The individual benefits
On a personal level, education is good for you – literally. In most countries, levels of academic attainment correlate with life expectancy, and some research suggests that this link is causal rather than coincidental.[22] Other apparent personal benefits statistically related to time spent in education include, according to one extensive literature review, promoting better decisions on “marriage, and parenting. It also improves patience, making individuals more goal-oriented and less likely to engage in risky behaviour.”[23] For some, learning itself is fun.
The most researched aspect of personal gains from education is the economic one, referred to as the returns to schooling. Since Gary Becker published Human Capital in the mid-1960s, a host of studies have calculated the financial benefit in various countries of time spent in school. These typically reveal a gain in annual earnings of between 8% and 10% for every additional year of education.
It is not, however, straightforward to use such insight in order to improve a country’s average earning potential. Education may not even be the cause of individual higher wages: instead it could be that educational success signals to employers the presence of other valuable qualities. Moreover, returns to education vary, on occasion widely, in a number of ways. For one thing they tend to be higher in less developed countries. In wealthier nations benefits tend to accrue more at the tertiary level, while in poorer ones they have been shifting from the primary to the secondary level. Time in school beyond that required for the occupation which the student eventually takes up – known as “over-education” – yields substantially lower returns. Results also differ by geography, or even city, within countries, and often also between gender.
Just why these differences appear is not always clear, but simply keeping everyone in the black box of education a few years longer will not yield magic results. Above all, the quality of education matters: one World Bank study suggested that the apparent decline in the returns to primary education in developing countries may arise from the length of time it takes to teach even basic literacy and numeracy in a number of those countries.[24] 
Getting the best at national level
Good education may, in most cases, help the individuals being educated, but does it help their society as well? A substantial literature sees behavioural impacts on educated individuals that have positive societal impact – for liberal democracies at least – including, to name just a few, better health for the relatives of those educated, lower arrest rates, higher voter participation and even a greater tendency to support free speech.[25]
In considering country-level benefits, the more common area of study has also been economic. On a basic level, education helps. Our correlation analysis shows a strong link between average years in school – or school life expectancy – and labour productivity. This does not surprise Namibia's Mr Angula: “A well-educated nation is likely to be innovative. I don’t think that you have to go to the statistical evidence to find that. People are able to use knowledge for economic development.” It is not simply that better educated people themselves are more productive. Extensive research has found a spill-over effect from education, with benefits arising both from how the educated share their knowledge with others and how they are better able to pick up new skills themselves by building on their existing education.
Chart 7a: Relationship between school life expectancy and labour productivity, 1990-2011
Chart 7a: Relationship between school life expectancy and labour productivity, 1990-2011
Note: The scatter matrix shows the correlation of school life expectancy for all years against all possible future years for overall productivity of labour.  The correlation in each set of years is well above our threshold for "strong" correlations of 0.65.
Sources: Economist Intelligence Unit and UNESCO.

Chart 7b:  Relationship between school life expectancy (in 1995) and labour productivity (in 2010)
Chart 7b:  Relationship between school life expectancy (in 1995) and labour productivity (in 2010)
Note: The scatter chart shows the correlation of school life expectancy in 1995 against overall productivity of labour in 2010 for 37 countries.  The correlation is 0.817, well above our threshold for "strong" correlations of 0.65.
Sources: Economist Intelligence Unit and UNESCO.
The difficulty for policymakers, though, is deciding what sort of education works best when so many factors affect the economy. Predictably, quality appears to be more important than duration. In one analysis, Professors Hanushek and Woessmann found that when cognitive skills, as measured by PISA scores, are correlated with GDP, then the impact of total years of schooling becomes irrelevant. In other words, how long it took to learn was less important than that learning had occurred.[26] This may seem obvious, but it is directly applicable to decisions such as starting primary education a year earlier or using the same resources for teacher training.
More complicated than quality is the question of what sort of content in an education system will yield the best labour market and economic outcomes. For example, some countries prize strong vocational school programmes while others prefer more unified systems. One advocate of vocational education is Professor Schwartz, who says of the US that “having a system focussed entirely on preparing students for four-year colleges and universities is a major problem. Only 30% of young Americans actually get a four-year degree by their mid-twenties, and many of those wind up in jobs that didn't require a degree. The consequence of not having a strong post-secondary vocational system is that most young Americans reach their mid-twenties without the skills and credentials needed for success in a technology-driven economy.”
Mr Angula, whose country is looking to bolster its vocational education system, adds that systems “need to create linkages between the school and the community, and the school and the economy, so that education should have a meaning in the context that it is practised. Sometimes it is hard for students to apply their knowledge or skills.” Without seeing any relevance, they might simply leave education.
Softer skills
The questions of the appropriate education content to best ensure future economic growth and how best to equip students to face an uncertain future are also at the core of reforms in some of the more successful school systems, particularly in Asia. Singapore’s Professor Lee explains that “of today’s job titles compared to those of 1995, many are very new; the skills are very new. We anticipate that evolution will be fast into the future.” For over a decade, his country’s Ministry of Education has engaged in future scanning to identify the likely skills needed in the coming years, and adjusted its offerings to students accordingly. More important, since 1997, says Professor Lee, Singapore has shifted away from teaching rote knowledge to a firm foundation in the basics of maths, science, and literacy combined with an inculcation of how to understand and apply information. “We feel it contributes toward the students acquiring knowledge and skills of cognition and creativity attributes which are very important in the 21st century landscape.”
Both of these developments reflect an attitude that education systems need to be prepared for ongoing change rather than seek a single, best end state. “No education system can remain static,” writes Singapore’s Prime Minister, Lee Hsien Loong, in the foreword to a recent report on education and geopolitics in the 21st century. “The world is changing rapidly. Technology is transforming our lives. The skills needed in the future will be very different from those needed today.”[27]
Singapore is not alone. Shanghai students finished first in the latest PISA tests, but China is also shifting toward a much greater emphasis on creativity. Professor Zhao explains that the country’s leadership believes “the economy is moving quickly from a labour-intensive one to a knowledge economy. It needs creative talent.” Indeed, he finds it ironic that China is moving more in the direction of Western models even while politicians in those countries sometimes praise that of traditional Asian education. South Korean schools, meanwhile, are now being encouraged to develop "creativity, character and collaboration".
Chart 8: Percentage of labour force reaching secondary and tertiary attainment, selected countries, 2008 (%)
Chart 8: Percentage of labour force reaching secondary and tertiary attainment, selected countries, 2008 (%)
 Source: International Labour Organization.
Teaching people how to work together is indeed of growing relevance to the economy. According to Ms Parthasarathi, “A lot of education in the second half of the 20th century has made children fiercely individualistic, not good in a team, but these team skills – an ability to interact with respect with people; to empathise; to be innovatively adventurous – are essential for certain types of creativity.” In order to drive the teaching of collaborative skills, the Assessment and Teaching of 21st-Century Skills project – a multi-stakeholder group that includes the education ministries of the US, Australia, Singapore, Finland, the Netherlands and Costa Rica – has been seeking to develop metrics to test such abilities. These will be integrated into the PISA 2015 tests – a sign, Professor Schleicher says, that “the kinds of skills that matter in life are changing.”
Education can clearly deliver substantial social and economic outcomes. Understanding how it does so, however, and maximising those results are still works in progress for educational leaders. Says Mr Mackay, Chair of the Australian Institute for Teaching and School Leadership: “None of the countries you might think would be complacent are complacent at all: they are investing in new metrics.”

[22]Hans van Kippersluis, Owen O’Donnell, and Eddy van Doorslaer, “Long Run Returns to Education: Does Schooling Lead to an Extended Old Age?”, Journal of Human Resources (2009): 1–33.
[23]Philip Oreopoulos, Kjell G. Salvanes, “How large are returns to schooling? Hint: Money isn’t everything”, National Bureau of Economic Research Working Paper 15339, September 2009.
[24]Tazeen Fasih, Linking Education Policy to Labor Market Outcomes, World Bank, 2008, see also Tazeen Fasih, et al, Heterogeneous Returns to Education in the Labor Market, World Bank Policy Research Working Paper 6170, August 2012.
[25]Thomas S. Dee, Are There Civic Returns to Education? National Bureau of Economic Research Working Paper 9588, March 2003; Craig Riddell, “The Impact of Education on Economic and Social Outcomes: An Overview of Recent Advances in Economics”, Canadian Policy Research Network, 2006.
[26]Eric A. Hanushek and Ludger Woessmann, “Education and Economic Growth”, in Dominic J. Brewer and Patrick J. McEwan, eds. Economics of Education (2010).
[27]Foreword to Michael Barber, Katelyn Donnelly and Saad Rizvi, Oceans of innovation: The Atlantic, the Pacific, global leadership and the future of education, 2012.


Post a Comment