• Full Report version

Schools and other education and training institutions are formally responsible to governments and informally to parents and students. Many countries devolve decision-making to regional and local school authorities, encouraging both bottom-up and top-down accountability. Emphasis on accountability poses several challenges for schools.

GOVERNMENT REGULATIONS HELP IN MONITORING SCHOOL QUALITY

Government education regulations vary worldwide. For example, while nearly all 71 systems reviewed for the GEM Report had teacher qualification regulations, less than 40% had maximum pupil/teacher ratios (Figure 4). Regulations can hold education providers accountable but may not be effective in practice. In poorer countries, many schools did not comply with existing regulations for reasons outside their control. For instance, underfunding means many schools in Tajikistan are not properly heated in winter despite regulations.

Traditionally, school inspections monitored regulatory compliance, with effectiveness depending on inspectors’ skills. Some research indicated that principals who felt strong accountability pressure from inspections acted to improve performance.

Many private schools in poorer countries are not regulated

Private schooling has expanded. The number of countries with over 20% private enrolment increased between 2005 and 2015 (Figure 5). In some countries in sub-Saharan Africa and Southern Asia, many private schools are unregulated, especially low-fee schools serving poor populations, which have grown faster than governments can handle. Some schools remain unregistered to avoid overly restrictive regulations. Regulating private schools to improve equity requires concerted action.

Weak regulatory environments are especially problematic when powerful private chains expand rapidly. Bridge International Academies operates more than 500 schools in 5 countries. Inspections in Kenya and Uganda reported unqualified teachers, inadequate infrastructure and unauthorized curricula and courts have upheld ministries’ moves to close some schools.

An inspection focus on education quality is welcome but difficult to implement

Increasingly, especially in richer countries, inspection is shifting away from compliance with regulatory standards and towards evaluating the quality of teaching and learning that takes place in schools. However, this supportive function is difficult to carry out well. Inspection systems in poorer countries face resource and capacity constraints. In South Africa, supervisors resisted inspection reform, partly from memories of apartheid inspections. In many contexts, improving inspection takes time. By 2015, only 45% of inspectors in Angola had received training in reforms begun in 2010.

Quality assurance in early childhood education focuses on easily observable aspects

Despite the importance of early childhood education to children’s holistic development, the World Bank’s Systems Approach for Better Education Results between 2010 and 2015 showed that only 14 of 34 low and middle income countries had established early childhood education standards and compliance monitoring systems.

In quality assurance, countries often favour easily measurable and observable operational characteristics, such as infrastructure and pupil/teacher ratios. Even so, countries often struggle to monitor compliance systematically, as examples from Belize, Indonesia, Nepal and Swaziland suggest.

Other systems try to assess more nuanced aspects of teaching. In Chile, educators at all municipal schools are evaluated every four years against the Good Teaching Framework standards, a process involving self-evaluation, external observations, peer assessment and a portfolio. Teachers rated ‘unsatisfactory’ are re-evaluated the following year and
barred from teaching if no progress has been made.

Some instruments, such as the Early Childhood Environment Rating Scale, help assess the quality of interactions between teachers and children. Developed and widely used in the United States, the scale has also been adapted for use in other high income countries, including Germany and Italy.

Direct measures of early childhood development can support quality assurance processes. The longitudinal study Growing Up in Scotland aims to link early experiences with later outcomes among 14,000 children in three cohorts, with results feeding back into redesign of early childhood care and education policy.

Community contributions, particularly by parents, are crucial for ensuring quality of early childhood care. In France, the National Family Allocation Fund produces a regular barometer based on parent satisfaction surveys, and elected parent representatives provide input to the Early Childhood Commission of the General Council.

Only 14 of 34 low and middle income countries had established early childhood education standards and compliance monitoring systems

Quality assurance mechanisms in higher education reflect varying objectives

Countries’ legal frameworks provide for single or multiple national agencies responsible for quality assurance in higher education, although many low income countries have not yet established national systems. Regional arrangements, such as the Lisbon Convention, have spurred development of national quality assurance systems, with countries incorporating regional standards into national law.

Quality assurance assessment involves standard setting, institutional self-assessment, external expert and peer review, evaluation reports and appeal processes. Standards, either prescriptive or advisory, cover higher education inputs, activities and outputs. China’s Quality Assessment of Undergraduate Education standards encompass 19 indicators in 8 major areas: university mission, teaching staff, facilities, academic curriculum, management, atmosphere, learning outlines and feature programmes.

Quality assurance agencies hold themselves accountable via annual reports, databases, regional and international agency registers, and national information centres. The International Network for Quality Assurance Agencies in Higher Education publishes a manual of best practices to encourage accountability and transparency; 18 members in countries from Costa Rica to the United Arab Emirates have been aligned with its Guidelines of Good Practice. However, much of the information in accountability reports is not widely disseminated beyond higher education experts. Regulatory frameworks covering cross-national higher education largely focus on supporting institutions.

International students are often unaware of their rights, and information may be difficult to access. Countries should prioritize identifying and raising awareness about disreputable providers and encouraging student bodies to disseminate information on good-quality providers.

Many scholarship programmes regularly account to donors for resources spent, but their reports would be more useful if they also provided timely information to students, families and universities. Longitudinal studies measuring programme impact and capturing university and alumni feedback are useful. The Commonwealth Scholarship Commission in the United Kingdom, for example, surveys award holders and uses the results to inform future programme design.

Quality assurance in higher education involves standard setting, institutional selfassessment, external expert and peer review, evaluation reports and appeal processes

Governments need to be held accountable for ensuring affordable access to higher education

Higher education enrolment has been growing steadily, driven by improved student progression rates and higher numbers of part-time students.Governments use national legislative frameworks to foster equity and affordability in higher education, but few countries guarantee universal access. Those that do include Ecuador, Greece and Tunisia. Many laws guaranteeing access to higher education, including those in Brazil and the Lao People’s Democratic Republic, prohibit discrimination and encourage access for minorities and disadvantaged groups.

As demand for higher education has grown, governments have shifted some of the cost burden onto individuals, either by increasing tuition and fees or encouraging private sector provision. Even where there are no fees, however, this by itself is not sufficient to ensure affordability. Without additional support, free universal access can still end up subsidizing the rich. The Philippines, for example, abolished public colleges’ fees in 2016, but these were already attracting students from richer backgrounds.

Tuition fees should be combined with financial aid programmes, which may include grants, loans and tax benefits. Loan repayment assistance for students with low income can help increase affordable options. Targeting low income populations is critical, but means testing can be difficult in countries with less reliable measures of household finances, as in many low income countries.

Tuition fees should be combined with financial aid programmes, which may include grants, loans and tax benefits

Skills providers and certifiers are accountable to trainees and employers

A robust quality assurance system for professional skills development helps hold authorities and service providers accountable to beneficiaries, such as workers and employers, and to each other.

Skills development qualification systems need coherent governance, with a common framework that outlines clear aims. One way to link labour market demand and provider supply is to involve employers and social partners in developing frameworks, although this has not always proved easy, for instance in Poland and Tunisia.

The increasing number of non-government training providers should comply with regulatory standards and procedures to be accredited and operate. As with higher education, accreditation is a quality assurance process whereby usually external government or professional authorities confirm providers have met set standards.

Aiming to train 400 million people by 2022, India’s ambitious skills development programme has to ensure that certification is transparent, trainees receive the full benefits, candidates register using a unique identification number and no illegal subcontracting to non-accredited providers occurs. The government needs to protect trainees from false claims promising them employment and requesting fees in return. Similarly, in Australia, a Senate inquiry examined whether private training providers’ marketing misled candidates, especially disadvantaged ones, about the value of qualifications to be earned.

All of the more than 200 adult literacy and numeracy programmes in a UNESCO database have carried out some monitoring and evaluation

Monitoring can improve accountability in adult literacy programmes

Accountability in adult literacy and numeracy programmes is complicated by the wide range of programmes, providers, funding streams and perceived aims. Even so, countries are increasingly setting quality standards and expectations for results. Monitoring systems are becoming common: all of the more than 200 adult literacy and numeracy programmes in the UNESCO Effective Literacy and Numeracy Practices Database have carried out some monitoring and evaluation, usually as part of programme management and implementation.

Collecting financial data helps governments hold non-government providers accountable for quality and resource allocation. South Africa’s Kha Ri Gude (Let Us Learn) literacy programme contracts with a private company for financial accounting and reporting, and to update the learner and educator databases of a management information system. Teacher payments depend on expenditure and attendance data submission. An audit in 2016 found that volunteers had been allegedly paid stipends for more learners than indicated on their claims.

Monitoring literacy programme results can help ensure accountability. Field visit assessments are one method, used for example in Pakistan. Others include formative and summative strategies, such as tests, oral presentations and self-assessments.

High income countries often assess achievement using standardized national assessment frameworks and tools, sometimes linked to public funding, as in the United States. Some middle income countries, including the Islamic Republic of Iran and Mexico, give online final exams automatically generated for each district. Other countries rely more on class facilitators to generate formative and summative assessment and do not systematically collate data for analysis. And some programmes go beyond narrowly construed literacy skills in assessing learning achievement. France’s Fight Against Illiteracy programme evaluates participants on autonomy, confidence, motivation, daily life interactions and cognitive development.

USING STUDENT-LEVEL LEARNING DATA TO HOLD SCHOOLS TO ACCOUNT IS COSTLY…

Governments are increasingly interested in collecting data on school and student learning outcomes. In principle, this information should enable education leaders at the national, subnational and school levels to make evidence-based decisions, as long as the information is of good quality and they have decision-making power independent of political interests.

Summative assessment results are used at the individual level to make student admission and progression decisions and at the institutional level to position schools relative to one another to identify areas for improvement. At the system level, summative assessment results can help monitor whether standards are met.

Countries differ in how they use individual learning data. Some systems, such as Japan’s, focus on national examinations, which determine progression between levels in a given year but do not enable comparisons of learning over time.

Other systems define standards for expected learning and organize assessments to report against them. England (United Kingdom)’s complex and comprehensive system on learning outcomes is based on national standards, an elaborate student assessment mechanism and an external evaluation system. The data collected are used to prepare inspectors prior to school visits, inform parents, help school leaders set targets, identify pupils in need of additional support, and support local and national authorities in monitoring performance for accountability purposes.

Different countries assess different learning outcomes, with some focusing exclusively on language and mathematics and others assessing a broader range. Countries also differ in the kind of school and student background information they collect to enable contextual comparison. In Australia, context information on schools, including finance, demographic structure and socio-educational advantage, is made available through the My School website. In Denmark, the Agency for Education and Quality introduced a student well-being indicator.

Managing the information from learning outcomes can be challenging for education systems. Even high income countries need to work hard to avoid simplistic interpretations

But managing all this information can be challenging for education systems. Even high income countries need to work hard to avoid simplistic interpretations by adjusting for school and student socioeconomic information and for whether schools and students improve over time. Countries are increasingly introducing such value-added measures, but they can be insufficiently precise, and the conclusions drawn from them need to be tempered.

These problems are exacerbated in middle and low income countries. Information on outcomes that would allow reliable comparisons is costly to produce, and the necessary investment in capacity may be prohibitive. These countries often focus more on final examination results than on comparisons against standards. Jordan’s National Test, for example, assesses each grade every three years, but the results are not comparable over time as test items change regularly. Published reports consist mostly of descriptive tables with no policy-related analysis, and teachers receive no support to understand the results despite the objective to provide pedagogical assistance.

…AND THE EVIDENCE THAT PERFORMANCE-BASED ACCOUNTABILITY DELIVERS GOOD-QUALITY EDUCATION IS MIXED

Of 101 education systems reviewed, 51 make student test scores publicly available, including 17 which use them to sanction and reward schools and educators. Test scores, however, are heavily determined by factors outside school control.

There is no clear evidence that sanctioning schools for test scores improves learning: Statistics typically show no or marginally positive gains. The United States No Child Left Behind Act threatened low-performing schools with closure. It had marginal positive effects on student performance, widened the black–white achievement gap and exposed students, especially in lowperforming schools, to narrower curricula as schools prioritized tested subjects.

Schools may adjust to performance-based accountability systems in negative ways, gaming the system and avoiding sanctions to the exclusion of longer-term reforms. Reshaping the testing pool, narrowing curriculum, teaching to the test and cheating were found in Australia, Chile, the Republic of Korea and elsewhere, disproportionately affecting disadvantaged schools and students.

Performance-based accountability may result in schools adjusting in negative ways, gaming the system and avoiding sanctions to the exclusion of longer-term reforms

MARKET COMPETITION IN EDUCATION CAN DEEPEN SOCIAL DIVISION

One potential accountability mechanism is competition. The idea is that, if parents can choose their children’s school, it puts schools under pressure to perform better to attract students.
Making school information publicly available and understandable is a prerequisite for parents to choose and for a market to function. In many middle and high income country education systems, schoollevel test results are posted publicly. However, in poorer countries, information is not easily accessible or understandable for intended users.

For instance, online report cards are rarely accessed in the United Republic of Tanzania, where internet access is low. In Kenya, 72% of parents did not know how to use literacy and numeracy information. Some middle and high income countries have been proactive in creating a market for schools. School choice policies have increased in over two-thirds of member countries of the Organisation for Economic Co-operation and Development (OECD) in the last 25 years. However, evidence suggests that school choice policies benefit more advantaged populations. Parents frequently base choice on factors such as demographic composition, which can lead to diminished diversity and reinforce socio-economic divisions. In Finland, school choice was primarily exercised by educated families whose children excelled academically. In Santiago, Chile, only one in four parents of grade one students chose the highest performing school from their shortlists, and almost 70% looked at schools only in terms of religious affiliation.

Voucher programmes can equalize school choice, but their impact on education is mixed. Colombia’s programme targeting low income neighbourhoods increased private school enrolment, as well as voucher recipients’ achievement levels and graduation rates. However, making vouchers universally available and allowing schools to raise their fees may increase inequality in access without improving student performance. Sweden’s universal voucher program has been associated with growing segregation. Chile has a highly stratified system. Its voucher programme has encouraged selective admission of high-achieving or high-income students. Reforms to improve targeting in 2008 did little to improve equity.

COMMUNITIES CAN HELP SHAPE AND MONITOR SCHOOL POLICIES AND PRACTICES

Social accountability by communities can improve school responsiveness and efficiency. Community monitoring often focuses on infrastructure, staff attendance and budgeting, but the impact of one-time interventions can be unsustainable. In Ethiopia, community partnering with government to collect school data and increase community dialogue had positive results. However, lack of resourcing can threaten such projects’ sustainability.

Community stakeholders participate in school-based management (SBM), which sees decision-making authority and responsibility transferred to local actors. SBM has improved student achievement and attendance in countries including Indonesia and Mexico. However, unwillingness to share responsibility with community members has stymied some SBM efforts, as in Hong Kong, China. Community representation sometimes excludes marginalized groups. Elite capture was a problem for some SBM committees in Nepal.

In school choice systems, parents frequently base choice on factors such as demographic composition, which can lead to diminished diversity and reinforce socio-economic divisions