This chapter examines the drivers of school attendance problems using a bioecological framework that captures influences across process, person, context and time. Drawing on international research, the OECD Policy Survey on School Attendance Problems and PISA 2022, it shows that absences emerge from interacting influences rather than single causes. These influences operate across policy structures, community conditions, family-school relationships, school climate, peer dynamics, service capacity, family circumstances, health, well-being and student engagement. Some have intensified or changed since the COVID-19 pandemic. Absence patterns also show persistence over time and can intensify during key educational transitions.
Every Day Counts
Understanding, Preventing and Responding to School Attendance Problems
Abstract
Understanding the drivers of school attendance problems is essential for designing effective, preventive and context-sensitive policies. Absences are not caused by single factors but reflect interacting drivers across multiple levels over time.
Drivers of school attendance problems operate cumulatively across system levels
School attendance problems rarely result from a single cause. Instead, they emerge from the interaction of personal, relational and structural pressures across time, policy systems, communities, schools, families, peers and individual characteristics.
Absence trajectories often begin early and tend to persist
Prior absences are one of the strongest predictors of future absences across education systems. Patterns can form early and can stabilise or intensify during transitions between educational stages. These involve new academic expectations, social pressures and developmental changes, which can increase vulnerability to absences.
Structural conditions and access barriers can shape absence
Policy frameworks and community contexts can influence absences indirectly by shaping families’ resources, school practices and access to support services. Factors, such as neighbourhood safety and transport reliability can further constrain students’ attendance.
Changing norms and institutional co-ordination can influence absence
In some systems, parental tolerance of minor illness absences or term-time holidays seems to have increased, weakening shared expectations about daily attendance. At the same time, weak co‑ordination between schools and external services, along with gaps in mental-health and social support, can delay support and contribute to sustained absences.
Relationships, school environments and support capacity are central to absence patterns
Attendance is shaped by relationships and environments across schools, families and peer groups. Supportive school climates, strong school belonging, positive student-teacher relationships and stable peer networks are linked to better attendance, while bullying, weak school-family communication and fragmented support can contribute to more absences.
Family hardship, health-related challenges and disengagement remain critical
Material deprivation, unstable housing, caregiving responsibilities, parental health problems and family conflict can disrupt routines and reduce students’ capacity to attend school. Physical illness, mental-health difficulties, boredom, low motivation and perceived lack of relevance further influence individual attendance decisions.
Understanding the drivers1 of school attendance problems (SAP) is critical for designing effective responses. When policies and interventions are grounded in a clear understanding of what drives SAP, they are more likely to lead to meaningful improvements in attendance and, in turn, better educational and developmental outcomes. Analysing drivers enables more precise, context-sensitive strategies that move beyond generic or reactive approaches. It also strengthens the evidence base that informs future research, policy and practice.
Although attendance and absence are often treated as opposites, the factors that promote one do not necessarily prevent the other. For instance, students may attend school regularly due to strong peer ties but still face barriers, such as housing instability, which increase their risk of absence. In contrast, some may be absent due to illness or exclusion, despite feeling positively about school. Such examples underscore the importance of studying absence in its own right, not merely as the inverse of attendance.
Focusing on drivers of SAP can shift policymakers’ attention from symptom management to root cause intervention. Understanding the drivers might help move beyond responding to observable symptoms (such as repeated lateness) towards addressing the underlying causes. This is crucial for both effective policy design and supportive frontline practice. Analysing drivers can also enable cross-country comparisons by distinguishing between context-specific drivers and those that recur across national contexts, helping policymakers identify common patterns and potential system-level levers for change.
This chapter aims to provide a conceptually grounded, credible and practically useful overview of current knowledge on the drivers of SAP. It does not catalogue every driver as this is beyond the scope of this report. Indeed, absences can arise from many potential causes – 781 risk factors in one review – and are typically embedded in a web of personal and social problems (Gubbels, van der Put and Assink, 2019[1]). Because the COVID-19 pandemic might have fundamentally changed the role of some drivers, the primary focus of the chapter remains on publications released during or after the pandemic. While research often examines variables across multiple levels, the placement of each study in a specific sub-section reflects its primary focus. However, when interpreting the drivers of SAP, it is important to recognise that effects may result from interactions across levels (see the Interactions among levels section). Moreover, as also elaborated in Chapter 3, SAP should not be understood as the product of isolated drivers acting in a linear way. Rather, they often emerge through reciprocal and cumulative processes, in which academic difficulties, socio-emotional challenges and disengagement both contribute to, and are reinforced by, absences over time. The final section illustrates how drivers at different levels can interact and reinforce one another.
This literature is complemented with results from the OECD Policy Survey on School Attendance Problems (see report annex). The survey was completed by 45 education systems across the world (mostly OECD and EU countries) and collected results from national (or sub-national) sources on drivers of SAP.
The analysis of PISA 2022 data on long-term absence (Box 2.1) complements the overview of current knowledge on the drivers of SAP.2 It enables an examination of whether certain factors identified in the literature correlate with long-term absence and whether these associations vary among countries. Differences between students who reported long-term absence and those who did not, and differences adjusted for a range of student and school characteristics are shown.
In PISA 2022, students were asked whether they had ever missed school for more than three months in a row (OECD, 2021[2]). For each educational level (primary, lower secondary and upper secondary), they selected “No, never”; “Yes, once” or “Yes, twice or more” (ibid.). While PISA has traditionally also included an item on arriving late, skipping classes and skipping school in the two weeks before taking the assessment, this chapter focuses on the long-term absence measure for two reasons: (1) it has not been extensively explored in OECD reports, academic or other literature; and (2) compared to the item exploring arriving late and skipping, it is more closely aligned with measures of chronic absence that are of focus in many education systems.
In this chapter, two types of results are presented. First, students are categorised into two groups: those who reported missing school for more than three consecutive months at any educational level, and those who did not. Subsequently, differences between these two groups are reported on various measures (e.g. differences in bullying, school belonging, etc.).
Second, to account for the simultaneous impact of student and school factors on long-term absence, regression models are estimated that account for a range of student characteristics and school factors. The conclusions stemming from these models are reported in the relevant sections. Full model specifications are documented in Annex 2.B.
When interpreting these results, readers are cautioned that while long-term absence is reported retrospectively, the potential drivers of SAP are reported at the time of the assessment – there is thus a temporal inconsistency. Because the data cannot align these time frames, whether the measured factors are true drivers depends, among other things, on their stability over time. Moreover, other unmeasured factors can confound the relationship between the driver and SAP. Consequently, the estimates cannot be interpreted as causal (including in the conditional models). They describe differences in students’ home and school environments between those who were and were not long‑term absent, not the causal or driving effects of specific factors.
Moreover, the long-term absence measure itself is not without specific challenges. Readers are encouraged to consider at least these three:
The behaviour it captures is extreme: in most education systems, missing three or more consecutive months of school means missing more than a third of total school days in a year. National (or sub-national) measures of chronic absence are often defined below this threshold (see Chapter 1).
The measure may be biased due to sampled students who were absent: it is possible that students who are chronically absent missed PISA. While this cannot be ruled out, the association with sampling absence rates suggests that this does not present a significant challenge (Annex 2.A).1
COVID-related school closures may influence the measure: it is possible that some students interpreted the question of “missing school” not as missing instruction but as not being physically present in school. Typically, this would not present a problem, but the COVID-19 pandemic resulted in school closures in many education systems without missing instruction. While the correlation between school closures and long-term absence is moderate at the system level (Annex 2.A), including this variable in models predicting long-term absence does not alter the model coefficients or conclusions, and the variable itself is not statistically significant (Annex 2.B). Future releases of this item will shed more light on this issue.
1. Additional checks were performed to explore how the variable is clustered in schools. The majority of schools have very low shares of long-term absent students. On average across OECD (EU) countries, 45% (47%) of schools have fewer than 5% of long-term absent students. 21% (20%) of schools have 5%‑10% of long-term absent students, and 12% of schools have 10%-15% of long-term absent students. Only 2% of schools have 50% or more of long-term absent students.
As an initial descriptive illustration, PISA 2022 provides information on the reasons students report for long‑term absence (Figure 2.1). The reasons vary, but they are led by illness, followed at some distance by factors, such as feeling unsafe at school, boredom and closures due to natural disaster. This underlines that SAP are not associated with a single reported reason, but by a diverse set of health-related, relational, school-based and contextual factors.
Percentage of students who reported the following reasons for having missed school for more than three consecutive months at any education level
Source: OECD (2023[3]), PISA 2022 Results (Volume II): Learning During – and From – Disruption, Table II.B1.3.55, https://doi.org/10.1787/a97db61c-en.
Bronfenbrenner’s theoretical work offers a well-established framework for understanding how real-life processes and conditions shape human development (Bronfenbrenner, 1977[4]; Bronfenbrenner, 1981[5]; Bronfenbrenner, 1995[6]; Bronfenbrenner and Ceci, 1994[7]; Bronfenbrenner and Morris, 1998[8]). Originally developed to support research and theory on child development, the bioecological model has since been widely adopted in education-related research, including studies on SAP (Tong and An, 2024[9]).
Over time, the model was reformulated and introduced four interlinked components: process, person, context and time (PPCT) (Bronfenbrenner and Morris, 1998[8]; Bronfenbrenner and Morris, 2007[10]). The PPCT model emphasises how individuals engage dynamically with their environments across time. It supports both granular attention to immediate settings and reflection on broader influences over time, making it well-suited to research and policy analysis in the field of school attendance (Heyne, 2025[11]). In this chapter, the initial sections describe five nested environmental systems: the time (chronosystem) (Bronfenbrenner, 1977[4]; Bronfenbrenner, 1981[5]) and the context components (microsystem, mesosystem, exosystem and macrosystem). This is followed by the person characteristics of the PPCT model (Bronfenbrenner and Morris, 1998[8]; Bronfenbrenner and Morris, 2007[10]).3 An illustration of this framework is provided in Figure 2.2.
Note: Selected aspects of each of the components are displayed.
Chronosystem
The chronosystem captures the influence of time on attendance patterns, both through life transitions and broader historical trends (Heyne, 2025[11]). At the level of individual experience, transitions, such as school entry, adolescence or changes in family structure (e.g. divorce or separation) may alter a young person’s engagement with school. At the societal level, large-scale events, such as the COVID-19 pandemic, changes in education legislation, or shifting norms around digital learning can reshape school participation more broadly. Notably, the chronosystem includes not only periods of change but also forms of continuity, such as enduring cultural expectations about school or persistent structural inequalities, which may reinforce or counteract change.
Macrosystem
The macrosystem refers to the broader cultural, economic and ideological conditions that shape the values, norms and institutional structures across all other system levels (Heyne, 2025[11]). These include national policy frameworks for attendance, cultural expectations around school participation, societal attitudes toward disability or mental health, and economic conditions that influence inequality and educational investment. Such factors have an impact on attendance indirectly: by shaping how schools operate, how families prioritise education and what supports are made available. While macrosystem influences may resemble exosystem conditions, the distinction lies in the scale: the exosystem refers to specific external settings that indirectly influence the young person (such as a parent’s workplace or a local service), whereas the macrosystem reflects the broader societal blueprints – such as labour laws, educational ideologies or social policy priorities – that shape those settings in the first place.
Exosystem
The exosystem comprises settings or systems that the young person does not actively participate in, but which nonetheless exert indirect influence on their development and daily experiences (Heyne, 2025[11]). These influences are typically mediated through their effects on the microsystems surrounding the young person, especially the home and school. Common exosystemic factors include parental working conditions (such as irregular hours, job insecurity or high work-related stress), and the availability or accessibility of mental health and youth support services. Similarly, choices on resource allocation, placement practices or eligibility criteria for targeted programmes may influence how flexibly schools can respond to emerging attendance concerns. These indirect but specific influences underscore the importance of considering not only the young person’s immediate environment, but also the broader institutional and structural forces that shape their opportunities to attend and engage with school.
Mesosystem
The mesosystem encompasses the dynamic interactions between a young person’s immediate settings, i.e. between two or more microsystems in which the young person actively participates (Heyne, 2025[11]). This includes, for example, the relationship between home and school, as well as the connection between classroom and peer group experiences. These interconnections shape attendance by influencing how consistently expectations, values and supports are communicated and reinforced across contexts. Concrete examples of mesosystemic processes include parent participation in parent-teacher conferences, volunteering at school or maintaining regular contact with school staff. Co-ordination between school staff, such as teachers and counsellors, can also support timely responses to emerging attendance concerns. Peer relationships may function as a bridge or barrier across microsystems, depending on how they align with school norms, and how parents and educators interpret and respond to those dynamics. In some cases, neighbourhood environments, when experienced directly by the young person, interact with school or family life to influence their engagement with education. By contrast, interactions between the school and external support services (such as mental health or social work agencies) typically fall within the exosystem, unless the young person is directly engaged with both settings. Overall, when communication and co-ordination across microsystems are strong, they can help prevent SAP or facilitate re-engagement after a period of absence. Inconsistent or poorly aligned relationships, however, may contribute to misunderstandings, fragmentation of support and an increased risk of absences.
Microsystem
The microsystem refers to the settings in which the young person participates directly, i.e. through face‑to‑face interaction (Heyne, 2025[11]). These include the family, school, classroom and peer group. For students, the classroom and the school are primary microsystems. The relationships with teachers – developed daily through ongoing interactions – are particularly central influences, shaping not only engagement but also attendance patterns. Other attendance-relevant factors at this level include parent-child relationships, peer dynamics, classroom climate and the quality and accessibility of both digital and physical learning environments. Neighbourhood settings may also fall within the microsystem when the young person is directly engaged in local activities, such as volunteering, holding a part-time job, participating in sports clubs or attending community events. By contrast, neighbourhoods are treated as part of the exosystem when the influence occurs without the young person’s direct involvement, for example, when local crime rates, zoning regulations or neighbourhood associations shape parents’ decisions about safety or access to services. Whether a neighbourhood factor is microsystemic or exosystemic depends on the nature of the influence and the degree of the young person’s active engagement.
Person characteristics
In the PPCT model, the person characteristics refer to individual attributes that both influence, and are influenced by, interactions within the individual’s ecological systems (Heyne, 2025[11]). These attributes can be categorised into three types (Bronfenbrenner and Morris, 1998[8]; Bronfenbrenner and Morris, 2007[10]):
Demand characteristics, which invite or discourage reactions from the environment (e.g. age, gender, visible physical appearance or country of birth (through perceptions and interactions with others)).
Resource characteristics, encompassing abilities, skills and past experiences (e.g. cognitive aptitude and social competence).
Force characteristics, involving personality traits, motivation and temperament (e.g. curiosity, persistence, resilience and anxiety).
In the PPCT model, these characteristics are not static inputs but dynamically shape the nature and intensity of proximal processes a young person engages in, and are also transformed as outcomes. All these personal attributes interact dynamically with environmental factors to influence attendance and absence.
While demand characteristics, such as gender and country of birth, can shape how young people are perceived and responded to within their environments, they are not discussed in detail in this chapter as independent drivers of SAP. This is both to avoid duplication with the previous chapter and because these characteristics are better understood as correlates of absences rather than primary drivers. For example, gender differences in attendance patterns may be associated with higher prevalence of internalising difficulties among girls and behavioural difficulties among boys, leading to different pathways into absence. Differences by immigrant background may be associated with factors such socio-economic status or school belonging driving SAP.
Benefits of working with the bioecological model
The bioecological model offers distinctive value for understanding school absences as a complex, dynamic phenomenon, shaped by interacting influences across system levels and evolving over time (Heyne, 2025[11]). Rather than focusing solely on who or what influences SAP, the model encourages inquiry into how these drivers interact across various contexts and time periods. Specific benefits of applying the model include:
Integrated, multi-level thinking: The model supports holistic analysis across individual, relational, institutional and societal levels. This helps connect perspectives across policy, research and practice, reducing the risk of siloed interpretations. It responds to calls in the field of school attendance for more integrated approaches (e.g. Kearney (2021[12])). The model also highlights broader influences that are often overlooked, balancing the focus on person characteristics and microsystem drivers (Kearney, Childs and Burke, 2022[13]).
Potential for better-targeted interventions: By clarifying how different drivers operate at various levels, the model helps policymakers, researchers and practitioners identify intervention points that are appropriately aligned, thus potentially improving the fit between problem and response.
Context-sensitive use and comparison: The model can be adapted to local contexts (e.g. education systems, cultural expectations and resource constraints), while still allowing structured comparison and policy learning across national settings and institutional systems.
Shared conceptual language: It provides a common framework that facilitates cross-sectoral dialogue and knowledge exchange, making it easier to align efforts between policymakers, researchers, educators and practitioners.
The persistence of absences over time
Prior absences are a powerful predictor of subsequent absences across countries. Daily decisions and routines can form habits, those habits can settle into term-by-term patterns, and they can harden into trajectories across key developmental transitions. In New Zealand, chronically absent students are five times more likely to miss school again the following year, underlining the strength of year-over-year persistence (ERO, 2024[14]). Administrative data in England (United Kingdom) point to a similar pattern: over 80% of secondary students who missed more than 28 days in 2021/22 remained persistently or severely absent in 2022/23 (Department for Education, 2025[15]). Finnish registry evidence reveals that missing more than 20% of classes in grade 6 (the last year of primary education) strongly predicts remaining above this threshold throughout lower-secondary education (Hotulainen et al., 2024[16]).
Indeed, absence patterns consolidate early in schooling and then can become “sticky” across grades. In the United States, absence rates tend to stabilise after the third grade of primary education in a nationally representative cohort (Simon et al., 2020[17]). English (UK) longitudinal trajectory analyses from year 1 to year 11 in primary and secondary education indicate that absence patterns often stabilise or intensify over subsequent years; students in the moderate or increasing trajectory groups typically show sustained patterns across multiple stages of schooling rather than isolated fluctuations (Dräger, Klein and Sosu, 2024[18]). Finnish latent-class models similarly distinguish a majority with consistently few absences, a sizeable group whose absences rise at lower-secondary education, and a smaller group with abundant absences already visible in primary school (Hotulainen et al., 2024[16]). Evidence from Denmark and the Netherlands also reveals that extended absences visible in secondary education often have origins in primary years (Binsbergen et al., 2019[19]; Kristensen, Jensen and Krassel, 2020[20]).
In line with this evidence, PISA 2022 reveals that long-term absence is persistent, as self-reported by 15‑year-old students (Figure 2.3). On average across OECD countries, 36.1% of students who were long‑term absent in secondary education were also absent in primary education. In some countries (Bulgaria, Greece, Romania, the Slovak Republic and Thailand), more than half of students who were long‑term absent in secondary education were also absent in primary education. The association between long-term absence in primary and secondary education remains strong even after accounting for a range of other factors, such as socio-economic background (Annex Table 2.B.1).
Percentage of students who were long-term absent in secondary education, depending on whether they reported to have been long-term absent in primary education
Sorted in ascending order by long-term absent students in primary education.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Although attendance patterns are often persistent, they can change – usually gradually rather than through sudden, dramatic shifts. US analyses indicate that early non-attendance explains a meaningful, but far from total, share of later absences (Ansari and Pianta, 2019[22]). For example, absences in early childhood education and care (ECEC) explain 10%-20% of the variation in absences between first and third grade of primary education, and only 3%-4% of the variation between fifth and eighth grade. Similar patterns were observed across other grade levels (ibid.). This leaves a substantial portion of the variance to other influences, such as peer networks, bullying experiences and school climate (see below). Finnish transition mapping across seven years finds that students most often remain in the same absence category or move to a neighbouring one, with larger trajectory-jumps occurring but rarely (Hotulainen et al., 2024[16]). Moreover, truancy in grade 6 (the last year of primary education) predicts truancy in grades 7 and 9, but not uniformly so (ibid.).
Finally, although this report focuses on primary and secondary education, SAP can begin as early as in ECEC. In Chicago, Illinois (United States), for instance, nearly half of three-year-olds and one-third of four‑year-olds missed 10% of their pre-school years (Ehrlich et al., 2014[23]). Across the country, in any given year, 10% of children in ECEC and in grade 1 of primary education miss at least 10% of the school year (Attendance Works and Healthy Schools Campaign, 2015[24]). Moreover, 14% of children in ECEC are classified as “at-risk absentees”, missing one to six fewer days than the threshold for chronic absence, meaning that a substantial portion of children at the very start of primary education are chronically absent or close to it, at least in the United States (Romero and Lee, 2007[25]). In California (United States), being chronically absent (missing 10% or more days) in ECEC has a 0.5 probability of being chronically absent in primary education (Hough, 2019[26]).
Rising absences across years and transition pressures between educational levels
Student in New Zealand: Don’t like school. It is too repetitive across the years. You are learning the same thing year after year. Not learning new things. (ERO, 2022, p. 45[27])4
Absences typically rise as students progress through schooling, reflecting a pattern in which attendance behaviours intensify over time (Melvin et al., 2025[28]). Early transitions, especially the move into lower‑secondary education, often mark inflexion points where absences accelerate.
Absences, particularly unauthorised, generally increase with age, grade and year. In France in 2012, only 6% of students aged 12 or younger were absent compared to 36% among those aged 16 or older (Cristofoli, 2015[29]). In Sweden, 5% of primary-education students reported being absent every day or several times a week, compared to over 12% of secondary-education students (Swedish National Agency for Education, 2024[30]). In Chile, serious absence (students attending fewer than 85% of the total number of official school days in a year) stood at 23.7% in primary and lower-secondary education in 2025 and between 26.1% and 29.7% in upper secondary education (depending on the programme) (Government of Chile, 2025[31]). In Japan, the non-attendance rate rises steadily from 0.9% among first-grade primary school students to 3.7% among sixth graders (Ministry of Education, Culture, Sports, Science and Technology, 2026[32]). It then increases to 5.6% at the beginning of lower-secondary education and continues to rise throughout this educational level (ibid.). Figure 2.4 illustrates increases by grade in Croatia and the Slovak Republic. Similar patterns can be observed in Australia, Finland and Latvia (Hotulainen et al., 2024[16]; Hunter, Haywood and Chapman, 2025[33]; State Service of Education Quality, 2020[34]).
Average absence hours per student in Croatia (panel A) and the Slovak Republic (panel B)
Note: Measures in the two panels are not comparable. The panels display average absence (authorised and unauthorised) per student per grade. Data in panel B exclude special classes, special schools and second chance education study programmes.
Source: Ministry of Science, Education and Youth (n.d.[35]), Izostanci po razredima [Absences by grade], https://app.powerbi.com/view?r=eyJrIjoiM2Q1NjVmZDEtMGUyMy00MDBiLTkzYWItYjBhMTA3MDFlOWUxIiwidCI6IjJjMTFjYmNjLWI3NjEtNDVkYi1hOWY1LTRhYzc3ZTk0ZTFkNCIsImMiOjh9 (accessed on 17 February 2026); and OECD (2025[36]), OECD Policy Survey on School Attendance Problems.
Attendance pressures can intensify in the final year of primary education and through the move into secondary education (Ehrlich and Johnson, 2019[37]). Transitions between educational levels can be a challenging time for many students, particularly if it also involves changing schools and classmates (Varsik, 2025[38]). Between the primary and lower-secondary levels, the curriculum often undergoes significant changes, with the lower-secondary level dominated by more specialised subjects in larger schools and classes, and students are required to take on more responsibility (Beatson et al., 2023[39]; Howe, 2011[40]). In addition to challenges related to higher requirements of secondary schools, it occurs during the time of early adolescence and coincides with puberty, adding physical, intellectual, emotional and social changes (Bagnall, Fox and Skipper, 2021[41]; Patton and Viner, 2007[42]; Richards, 2011[43]; Short and Rosenthal, 2008[44]). Therefore, the transition from primary to lower-secondary education presents challenges for school systems, given that there can be a mismatch between the needs of early adolescents and the structure of lower-secondary education (OECD, 2018[45]).
Indeed, several education systems observe a marked increase in SAP at the beginning of lower-secondary education. Significant increases in absences between primary and secondary education are visible in Figure 2.4 above in Croatia and the Slovak Republic. Similarly, evidence from Denmark suggests that absences rise especially in the last year of primary education and often continues to climb thereafter (Kristensen, Jensen and Krassel, 2020[20]). This pattern is particularly visible for students with more than 10% of missed schooling and for unauthorised absences (ibid.). The Australian pattern of an “attendance cliff” at entry to secondary education is consistent with this dynamic, with the drop particularly pronounced for Indigenous students in remote areas and interpreted as arising from interacting emotional, cultural, structural and historical factors, including intergenerational trauma, racism and strained community-school relationships (Potia et al., 2025[46]). Some of this increase could also be explained by the fact that students change schools. In Finland, for instance, transitions to lower-secondary education are associated with an increase in SAP, possibly due to changes in classmates, familiar teachers and so on (Sergejeff, 2023[47]). Complementing this, research in the Netherlands indicates that students value the transparent social structure and mutual familiarity typical of primary schools as protective against disengagement, implying that the loss of these features at the transition may remove supports that help prevent SAP (Binsbergen et al., 2019[19]).
While the lower-secondary level seems to be where most absences are concentrated, SAP seem to decrease in upper secondary education in some education systems (Figure 2.4 above). This may be due to the self-selection mechanism, whereby students who are disengaged do not enter this educational level, as it may not be compulsory.
However, other education systems continue to observe increasing trends in absence at the upper secondary level. In Romania, for instance, the school participation indicator decreases from grade 9 to grade 12 (upper secondary education), reflecting increasing numbers of absences with each successive grade (Andrei, 2023[48]; Horga et al., 2024[49]). This trend may partly reflect students’ disengagement from programmes that were not their preferred option, alongside a growing focus on preparation for the upper secondary examination in a limited set of subjects (ibid.). In the United States, a chronically absent student in the last year of lower-secondary education has a 0.7 probability of being chronically absent the next year (Hough, 2019[26]).
International synthesis further suggests that less typical or repeated transitions, particularly in contexts of poverty or community violence, can undermine engagement and contribute to so-called “school refusal” (Leduc et al., 2022[50]). However, the direction of the relationship remains unclear: repeated transitions may contribute to the onset of “school refusal”, “school refusal” may lead to more frequent changes, or both may stem from shared underlying factors such as anxiety, exclusion or unmet needs. Difficulty adapting to less typical transitions, particularly in contexts of poverty or community violence, may undermine school engagement and heighten emotional distress. Leduc et al. (2022[50]) link this interpretation to the ecological principle of adaptation, defined as a young person’s capacity to evolve within changing environmental conditions. In this view, repeated school transitions may challenge that adaptive capacity and, in doing so, contribute to absences over time, particularly in the context of “school refusal” (ibid.). This point is related to another driver – peer consistency (see Peer and social influence).
Absence patterns are also heterogeneous. Research from England (United Kingdom), for instance, points out that while absences among students with special education needs (SEN) increase with age overall, this rise is steeper for some SEN groups. In particular, students with behavioural, emotional and social difficulties show the sharpest increases, becoming the highest-absence group by year 11 despite not starting with the highest rates in year 7 (Tanya Lereya et al., 2022[51]).
Policy design in absence patterns
School leader in Sweden: In nine cases out of ten, problematic absences are due to social factors outside the school. It then becomes a very strange situation when the responsibility is placed on the school… (Öhman, 2016, p. 27[52])
Macrosystem-level structures, such as policy frameworks, have a clear potential to shape absences. Legal attendance frameworks, legal provisions (e.g. punitive measures, incentives and rewards for attendance) and others can shape parents’ and students’ behaviours around absences, as well as school stakeholders’ responses to them. However, as is further elaborated in Chapters 4 and 5, such policies are rarely evaluated, and if they are, they often indicate insignificant results (Melvin et al., 2025[53]). Moreover, evidence from Texas (United States) suggests that attendance is driven far more by observed and unobserved student characteristics than by regional effects (Knight and Olofson, 2026[54]). This section focuses on policies that are not directly related to SAP (e.g. expenditure and early selection), while Chapter 4 discusses those that specifically aim to address absences.
Association between expenditure on educational institutions and absences
Lower levels of expenditure on educational institutions may contribute to higher rates of SAP by limiting the availability of academic support, student well-being services and inclusive learning environments that help students remain engaged in school. Under-resourced education systems may also face greater challenges in addressing barriers to attendance, particularly for students from disadvantaged backgrounds, increasing the risk of absences.
Per-student expenditure is correlated with long-term absence (Figure 2.5). Broadly speaking, countries with lower per-student expenditure report higher absence rates. Nevertheless, it is essential not to overemphasise the significance of this relationship. Outcomes do not only depend on the amount of expenditure, but also on the allocation and use of resources. This statistic can thus obscure other important factors that contribute to this association. Other macro-level influences that are correlated with per-student expenditure and long-term absence contribute to the association. Further research is needed to fully disentangle the role of macrosystem factors, including expenditure, in SAP.
Note: Long-term absence is measured as the percentage of students who reported missing school for more than three consecutive months at any educational level. Expenditure per full-time student refers to expenditure on primary and lower-secondary educational institutions per full‑time equivalent student (in 2022 USD PPP).
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025); and OECD (2026[55]), OECD Data Explorer: Expenditure on educational institutions per full-time equivalent student, https://data-explorer.oecd.org/s/2np (accessed on 9 March 2026).
Potential effects of early selection on school attendance problems
One policy that has long been a focus among OECD countries is early selection into educational tracks. Students in different education systems select programmes at various points in their academic path. On average across OECD countries, the first age of selection is 14.3 years, although in some education systems, students can select as early as at the age of 10 (OECD, 2023[3]). Earlier selection can create more space for specialisation. For instance, it can help students build a specialised set of skills that is in high demand and enter the labour market earlier (Stronati, 2023[56]). However, in many cases, early selection happens based on academic achievement, which can be problematic. Early student selection policies, which determine the grouping and separation of students based on their abilities, can exacerbate learning differences and educational inequities (European Commission: Directorate-General for Education, Youth, Sport and Culture, Public Policy and Management Institute, Downes, Nairz-Wirth and Rusinaitė, 2017[57]). Evidence linking early selection policies to SAP is, however, limited, so this section focuses mainly on engagement-related pathways that may precede or accompany absences.
In Germany, one study examines the national practice of early selection through students’ “stereotype awareness” (i.e. how they believe others perceive their track) (Bardach et al., 2023[58]). Stereotype awareness in grade 5 (first year of lower-secondary education) is significantly associated with lower school engagement at the same time point, across all tracks, with no significant differences in effect size between tracks. This supports the “stereotype awareness as harmful for all” hypothesis rather than the view that such awareness disproportionately harms students in specific tracks. No significant longitudinal associations are found between stereotype awareness and changes in engagement over time. The study accounts for socio-economic status, gender and immigrant background (ibid.). However, the results need to be interpreted with caution, given the reliance on engagement rather than SAP and the absence of Gymnasium students (the most academically oriented study programme), which limits generalisability to comprehensive systems or systems with later tracking.
Anticipatory effects of early selection can be observed before the transition itself. In the Flemish Community of Belgium, students who were expected to enter non-academic tracks reported lower behavioural and cognitive engagement and higher “sense of futility” while still in primary school (Boone and Demanet, 2020[59]). These differences persist, accounting for factors including gender, socio-economic background, ethnicity, grade retention and test scores, and are mediated by teachers’ subjective assessments of students’ competence rather than by students’ actual performance (ibid.). Although the study does not include attendance outcomes, the pattern suggests that a tracked secondary structure, operationalised via prospective track choice, can shape engagement precursors to absences before the formal transition occurs.
At the international level, evidence remains rare and mixed. One study finds that education systems that track students at a later age have higher levels of truancy based on PISA 2012, accounting for a range of student and country characteristics (Keppens and Spruyt, 2018[60]). Analysis based on more recent data that does not account for other factors suggests that the age of first stratification has a positive but small effect size on long-term absence (Figure 2.6).5 Education systems with similar levels of long-term absence exhibit a wide range of ages at which first stratification occurs. For instance, Czechia, Greece, Hungary, Latvia, Sweden and Türkiye have a long-term absence rate of around 7%, but their ages of first stratification range from 11 in Czechia to 16 in Latvia and Sweden.
Note: Long-term absence is measured as the percentage of students who reported missing school for more than three consecutive months at any educational level.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025); and OECD (2023[61]), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, Table B3.1.4, https://doi.org/10.1787/53f23881-en.
Other policies and practices potentially linked to absences
A further macrosystem lever with some evidence of relevance is class-size policy, though the causal literature is limited. Evidence from Tennessee’s (United States) randomised Project STAR experiment indicates that smaller classes in ECEC and primary education can improve attendance: experimental estimates suggest that a 10-student increase in class size raises the probability of chronic absence by around 3 percentage points and also increases total annual absence (Tran and Gershenson, 2021[62]). However, evidence is not uniformly positive across all forms of absences: using the same experimental setting, another study finds little evidence that smaller classes reduce infection-related absences specifically (von Hippel, 2024[63]).
Evidence also suggests that immigration enforcement practices can affect school attendance. An analysis in California’s Central Valley (United States) exploits day-level variation in immigration enforcement actions (“Operation Return to Sender”) and finds a 22% increase in daily student absences coinciding with these operations, with effects over three times larger for primary-education students than for secondary‑education students (Dee, 2025[64]). The authors attribute these patterns to heightened parental fear of separation (ibid.). Evidence from a school district in Florida (United States) similarly finds that heightened immigration enforcement leads to substantial and persistent increases in absences among foreign-born students (Camp et al., 2026[65]). The effects were particularly pronounced among older students, suggesting that attendance decisions may partly reflect students’ own responses to perceived enforcement risks (ibid.). In an earlier analysis that similarly focuses on immigration enforcement actions, school absences are the only outcome measure (from among academic and socio-emotional outcomes) that are significantly related to these practices (Sattin-Bajaj and Kirksey, 2019[66]).
Finally, two studies examining anti-discrimination laws for sexual and gender minority students report mixed results: one finds no significant association with attendance (Fields and Wotipka, 2020[67]), while another identifies a slight reduction in fear-based absences among gender and sexual minority students (Seelman and Walker, 2018[68]).
Climate-related and natural shocks disrupt schooling through closures and access barriers
Climate-related hazards are documented as a direct source of schooling disruption. UNICEF (2025[69]) estimates that at least 242 million students (from ECEC to upper secondary education) in 85 countries or territories experienced school disruptions in 2024 due to heatwaves, tropical cyclones, storms, floods and droughts. More specifically, evidence from Jakarta (Indonesia) indicates that floods disrupted children’s access to education during the 2013 flood emergency and early recovery period, with surveyed schools reporting disruption to education services, damage to school facilities (Lassa, Petal and Surjan, 2022[70]). Other barriers included blocked roads, flooded homes and classrooms, which prevented students or teachers from going to school (ibid.). However, these events are not limited to the Global South. For instance, hot days in England (United Kingdom) can increase absences, with effects particularly observed for illness-related absences and authorised holidays (Conte Keivabu, 2024[71]). Cold exposure had no substantive overall effect, except for illness-related absences in energy-poor neighbourhoods (ibid.).
Natural disasters can also affect schooling through longer-term enrolment, dropout and attainment pathways, rather than only through short-term absences or school closures. For example, volcanic eruptions in Java (Indonesia) reduced the likelihood that children were enrolled in school, with effects worsening over time and operating partly through household spending reallocation, earlier transitions to work, school infrastructure, pupil-teacher ratios and telecommunications networks (Bimardhika and Moorena, 2024[72]).
Neighbourhood and school environments associated with absence
Student in the Netherlands: Our neighbours argue almost every day. Today, they were throwing bricks. They also … play hardcore music almost every day, so I can't study very well because of that. (Children's Ombudsman, 2017, p. 31[73])
Neighbourhood conditions surrounding schools operate as exosystem influences on attendance, primarily through safety, stress and access-related mechanisms, with evidence from the United States linking both safety exposures and the built environment to daily participation. In New York City, day-to-day variation in arrests and reported violent crimes within 1 000 metres of a school is associated with higher absences, with a one-standard-deviation increase in nearby arrests predicting a 0.7% rise in absences, and a comparable increase in violent crime predicting a 0.4% rise in absences (Vachuska, 2025[74]). Once interactions are included, the arrest effect becomes contingent on school composition, rising to a statistically significant 2.4% increase in schools where 80% of students were Black, but remaining insignificant in schools with an average Black enrolment. Thus, conditions around the school can potentially have an impact on absences, likely operating through heightened fear, reduced institutional trust or disrupted routines (ibid.). In Flint, Michigan (United States), systematic field assessments of neighbourhood physical disorder (e.g. vacant lots, boarded buildings, broken windows) reveal a curvilinear association in which both low and high levels of disorder are linked to higher attendance, while moderate levels correspond to lower attendance (Smart et al., 2020[75]). This pattern suggests complex dynamics, possibly linked to the targeting of interventions in the most and least distressed areas (ibid.). The models account for grade level but lack individual-level covariates, and generalisability beyond Flint’s specific context of economic decline and racial segregation might be limited. Similarly, in Baltimore (United States), students whose estimated routes to school require walking along streets with higher violent-crime rates have higher rates of absence (Burdick-Will, Stein and Grigg, 2019[76]). In general, however, evidence on neighbourhood and community factors is limited (Melvin et al., 2025[28]).
Schools serving concentrated disadvantage often experience systematically higher SAP, reflecting both intake composition and compounding resource pressures. In Ireland, DEIS (Delivering Equality of Opportunity In Schools) schools, which receive extra resources because they have the highest concentrations of students at risk of educational disadvantage, have consistently higher absence rates across urban and rural settings in both primary and secondary schools (TESS, 2022[77]; TESS, 2023[78]; TESS, 2024[79]; TESS, 2025[80]). For instance, in 2023/24, 28% of students in DEIS secondary schools lost 20 or more days of schooling, while only 19% of students in non-DEIS schools did so (TESS, 2025[80]). In New Zealand, attendance declines are sharpest in low-decile schools, mirroring socio-economic stratification (ERO, 2022[27]; ERO, 2025[81]). In France, a similar pattern is visible in lower-secondary education. In 2023/24, 9.6% of students in lower-secondary schools addressing areas facing the greatest social difficulties (in the priority education network (réseau d’éducation prioritaire)) were absent without authorisation for four half-days or more during a month, compared to 3.7% in schools outside of the network (Cristofoli, 2026[82]). The rate was highest in the most intensive priority-education category, at 12.8%. Absences were also strongly patterned by school social composition: the quarter of lower-secondary schools with the lowest social-position index had an average absence rate of around 13%, compared with 4% among the most socially advantaged quarter (ibid.).
Travel time, cost and reliability
Secondary-education student in Romania: I live 40km from the city... It was terrible: I wouldn't make it to the first class and had to leave about half an hour before the penultimate lesson. I wouldn't make it to the last class at all, so I could get home. (Horga et al., 2024, p. 58[49])
As mentioned in Chapter 1, patterns of absence vary by school location in several education systems. One potential driver of these location-based differences is transportation availability, affordability and reliability, which can translate distance and infrastructure gaps into daily attendance barriers by shaping families’ capacity to get students to school. In New Zealand, 10% of parents reported being likely or very likely to keep a child home when transport is challenging, and 5% kept their child at home due to transport issues, with weather-related disruptions and difficulties getting children to school cited in surveys and focus groups as the main challenges (ERO, 2022[27]). In Romania, distance and transport problems are associated with lateness and absences: nearly one quarter of secondary students are often late or absent due to transport difficulties, almost 40% reported this sometimes, and over 20% miss classes due to transport costs (Plăeșu et al., 2024[83]). Qualitative accounts highlight congested traffic, poor infrastructure, timetable misalignment and adverse weather as salient mechanisms (Horga et al., 2024[49]). Similarly, in Baltimore (United States), more difficult commutes, either in the form of increased travel time or complexity, leads to students missing more days of school (Stein and Grigg, 2019[84]).
Transport barriers can constrain attendance particularly for marginalised and remote communities. In the Slovak Republic, transport barriers are identified among the challenges faced by marginalised Roma communities that can undermine regular participation in schooling (OECD, 2025[36]). Consistent with these patterns, analysis from New South Wales (Australia) indicates that students attending remote schools have lower attendance levels and larger year-to-year declines than students in urban and regional schools – even accounting for socio-economic status, well-being and engagement – situating transport, service access and local resource constraints as structural pressures on daily attendance (CESE, 2024[85]).
Links between home and school shaping absences
Parent in Australia: They were putting [it] back onto the parents, it’s our fault … instead of ‘school attendance is tanking, so is numeracy and literacy’ and the department isn’t prepared to go, ‘Maybe it’s a problem with the system’. (Amin and Ettinger-Epstein, 2024[86])
Stronger school-home relationships have the potential to improve absence patterns, indicating the importance of co-ordinated expectations and communication across school and home. In a meta-analytic review conducted before the COVID-19 pandemic, low parental school involvement has a large effect on school absences (Gubbels, van der Put and Assink, 2019[1]). Furthermore, in Illinois (United States), schools with stronger pre-pandemic family-school connections recorded better-than-expected attendance in 2021/22 even after accounting for prior attendance, school characteristics and community variables (Learning Heroes and TNTP, 2023[87]). In the Flemish Community of Belgium and the Dutch-speaking part of Brussels, lower parental school involvement – captured by student-reported participation in school activities, informal contact with teachers and involvement in school councils – is associated with both passive school withdrawal (parent aware but not supportive of an absence) and active school withdrawal (parental consent to an absence) (Kruithof and Keppens, 2024[88]).
Breakdowns in school-home relationships can also precipitate disengagement pathways that ultimately lead to withdrawal from education. In the United Kingdom, qualitative research explores how unmet needs, failed support meetings and deteriorating home-school relationships contribute to students’ withdrawal from education (Gillie, 2025[89]). Families described anxiety-related school avoidance, sanctions for lateness and escalating distress as culminating in deregistration. A central insight is that despite sustained parental attempts to collaborate with schools, communication frequently breaks down, leaving families without viable in-school support (ibid.). Echoing this pattern, New Zealand’s reports indicate that action is often too slow (ERO, 2024[14]). Only 43% of parents of chronically absent students meet with school staff, 18% of school leaders refer only after more than 21 consecutive days absent, 68% of attendance service staff say referrals are never or only sometimes made at the right time, and approximately half of schools make no referrals to attendance services (ibid.). Similarly, in Ireland, only 13% of parents speak to school about attendance, and 6% look for guidance or support on school attendance (RED C, 2025[90]).
While internationally comparable data on school-home relationships are scarce, the available evidence suggests that, in most countries, parental engagement does not differ significantly between students who were long-term absent and those who were not. Two measures are selected for this chapter: school leaders’ reports on whether parents initiated discussion about their child’s progress or behaviour (panel A in Figure 2.7), and school leaders’ reports on parents’ volunteering in physical or extra-curricular activities at school (panel B in Figure 2.7). In both cases, among students whose school leaders reported that such parental activities took place, differences between long-term absent students and those who were not long‑term absent are not statistically significant. This finding should be interpreted with caution, as parental involvement may increase in response to children’s difficulties at school, which can blur the direction of the relationship. More complex models that account for a range of other factors suggest that parental participation in such school-related activities is not significantly predictive of long-term absence (Annex Table 2.B.1).
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. Panel A of the figure displays the difference, in percentage points, between students who were long-term absent and those who were not long-term absent in the share of students attending schools where the school leader reported that at least half of parents discussed their child’s progress or behaviour with a teacher on their own initiative. Positive values mean a higher prevalence of parent-initiated discussions among long-term absent students. Panel B of the figure displays the difference, in percentage points, between students who were long-term absent and those who were not long-term absent in the share of students attending schools where the school leader reported that at least half of the parents volunteered in physical or extra-curricular activities. Positive values mean a higher prevalence of parental volunteering among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Every day at school does not count the same anymore
Parent in Wales (United Kingdom): Children who have overall good attendance should be allowed to take term time holidays without a fine. Holidays and trips are educational. Persistent poor attendance should be fined, not one-off holidays. (Parentkind, 2024, p. 57[91])
As mentioned in Chapter 1, there are indications that some parents altered their attitudes towards (compulsory) schooling after the COVID-19 pandemic. These shifts can have an impact on how families and schools align expectations about everyday attendance. They also shape the threshold for keeping children at home. While evidence is still scarce and mostly qualitative in nature, it likely presents a new challenge for policymakers.
Parental tolerance for non-essential term-time absences seems to have increased in several education systems. In Norway, school leaders reported that an increasing number of parents are taking their children out for extended weekends or holidays (Bergene et al., 2023[92]). In Wales (United Kingdom), 71% of surveyed parents said it is acceptable to miss at least a day for a holiday (Parentkind, 2024[91]). In England (United Kingdom), more than half of surveyed parents across socio-economic groups reported it is acceptable for a child not to attend school for a day to go on a holiday when it is more affordable (Carr et al., 2026[93]). In New Zealand, 76% of parents are very likely or likely to keep their child out of school for a cultural or special event, 47% for a holiday of one or two days, 58% for their child to participate in an out‑of-school sporting event, and 9% for their child’s birthday (panel B in Figure 2.8). In Ireland, a market research poll found that 98% of parents believe that attending school every day is important, but 65% of parents also think that occasional absences are not a serious concern (RED C, 2025[90]). Indeed, 63% of adults think that religious events are an acceptable reason for missing school, and 39% think that family events are an acceptable reason (panel A in Figure 2.8). In addition, 44% agree that taking a week-long holiday is acceptable, 34% agree that taking a day off for a long weekend is acceptable, and 32% agree that taking a day out is acceptable. Early research from England (United Kingdom) suggests that many of these opinions are consistently shared across educational levels, gender and socio-economic groups (Carr, Whitehead and Burtonshaw, 2025[94]).
Likewise, in England and Scotland (United Kingdom), several local authorities and schools cited changing parental attitudes and term-time holidays as emerging post-pandemic issues, possibly reflecting increased financial pressures felt by families and the related lower costs of taking holidays during the school year (Education Scotland, 2023[95]; Moore and Walker, 2025[96]). Some education systems, where granular data collection is available (e.g. France and the Slovak Republic), observe higher rates of absence around school vacation times (Cristofoli, 2026[82]; OECD, 2025[36]). Signals of this tolerance towards absences also predate the pandemic: in Iceland, most administrators already observed a rise in requests for holiday leave during school hours, and nearly half judged the impact on learning to be substantial (Ministry of Social Affairs and Housing, 2019[97]). In Sweden, some school leaders observed more tolerant parental attitudes towards their children’s absences already in 2016 (Öhman, 2016[98]).
Note: Panel A displays adult (18+) answers to the following item: “Below are a number of circumstances that parents/guardians/caregivers might experience in any given day. For each of these, can you please indicate the degree to which it is acceptable or not to take a full day off school?”. Data were collected in September 2025 from a nationally representative sample of 1 267 adults aged 18 and over in Ireland, supplemented by an oversample of 500 parents of school-going children aged 5-18. The oversample was weighted down in the analysis to reflect parents’ correct population share. Panel B reports responses to the question: “How likely would you be to keep your child out of school for the following reasons?”
Source: ERO (2025[99]), Back to class: How are attitudes to attendance changing? Technical appendix, https://evidence.ero.govt.nz/documents/back-to-class-how-are-attitudes-to-attendance-changing-technical-appendix (accessed on 19 December 2025); and RED C (2025[90]), Unpublished market research poll provided to the OECD by Tusla Education Support Service (TESS).
Expectations for attending school during a minor illness seem to have also softened. In Norway, school leaders reported that students are more often kept home for mild symptoms than before, though views differ on whether this persists (Vennerød-Diesen et al., 2024[100]). School leaders also observe this pattern across social groups, suggesting a broader cultural shift rather than a narrow socio-economic effect (ibid.). School staff in England (United Kingdom) and Sweden describe similar notions of staying home with mild illness (Moore and Walker, 2025[96]; Swedish National Agency for Education, 2023[101]). In Ireland, 75% of parents agree that mild sickness (cough, cold, runny nose and no temperature) is an acceptable reason for missing a school day (panel A in Figure 2.8 above). Some English (UK) and Norwegian school staff also link lower thresholds to more flexible parental work, noting that home-office arrangements make supervision at home easier (Bergene et al., 2023[92]; Moore and Walker, 2025[96]). These health-related and work-related shifts can interact to normalise short, discretionary absences.
Confusion about rules can weaken compliance with attendance expectations. In Norway, some parents misunderstand national absence regulations (e.g. on the maximum number of absences for parents to authorise) as automatic entitlements (Bergene et al., 2023[92]). In England (United Kingdom), there are indications that pandemic-era messaging about preventing the spread of illness continues to shape parents’ and students’ attitudes toward attendance (Moore and Walker, 2025[96]). Similarly, the Children’s Commissioner for Wales (United Kingdom) highlights that families may nowadays feel they should keep children home with mild symptoms even when the child is well enough to attend school, echoing pandemic messaging (Children, Young People and Education Committee, 2022[102]). Such non‑attendance patterns have been complex to reverse (Rowlands, 2022[103]). Where information is unclear, parents may underestimate the cumulative impact of absences. In New Zealand, for instance, when parents and students do not understand the implications of non-attendance, predicted chronic absence rises from 7% to 9% (ERO, 2024[14]).
Co-ordinated support between schools and other institutions
Teacher in Germany: We are not trained to handle very serious personal problems, and it would be nice to have someone nearby in situations like that. (Enderle et al., 2025, p. 11[104])
When support for addressing students’ needs does not transfer smoothly between institutions, students are more likely to miss school. Yet, co-ordinated support across schools and external services is often weak, with consequences for attendance. In the United Kingdom, for instance, schools’ delayed responses to mental-health needs can slow access to Child and Adolescent Mental Health Services and erode belonging, which can, in turn, contribute to extended non-attendance (Corcoran and Kelly, 2022[105]; Moore and Walker, 2025[96]). More specifically, in England (United Kingdom), a lack of information-sharing among education, health and social care (e.g. due to different information management systems) can be a barrier to timely support (Ofsted and Care Quality Commission, 2025[106]).
Evidence from New Zealand echoes these co-ordination challenges. Information is sometimes or never shared across agencies, schools and attendance services (ERO, 2024[14]). Attendance Services also reported spending excessive time locating students, with weak enforcement levers and unclear accountability compounding delays (ibid.). As a result, many school leaders seek more explicit role definitions. Resourcing constraints exacerbate these systemic frictions – adviser caseloads range from approximately 30 to over 500 cases, and per-student funding varies significantly – so capacity does not always align with need (ibid.). In Denmark, some parents step in as de facto case managers to ensure coherence and progress, proactively co-ordinating meetings and action plans amid reported discontinuity in processes and turnover among case handlers and professional staff (Børns Vilkår and the Egmont Foundation, n.d.[107]). In Sweden, some school leaders complained of lack of guidelines, active leadership and co-operation between their schools and municipalities (Öhman, 2016[98]).
School
School climate, belonging and student-teacher relationships
Student in New Zealand: Teachers; some don’t connect with learners; some seem like they hate kids; they target you; makes you ditch or want to ditch certain periods. (ERO, 2022, p. 45[27])
Student-teacher relationships, school climate and belonging are central features of the microsystem. Evidence from several sources suggests a link between these factors and SAP. While most studies are observational or cross-sectional, lacking the ability to estimate causal relationships, the direction of associations remains consistent across settings. In a meta-analytic review conducted before the COVID‑19 pandemic as well as a more recent rapid literature review, having a negative school attitude, poor student‑teacher relationships and negative school or class climates are significantly associated with absences (Gubbels, van der Put and Assink, 2019[1]; Melvin et al., 2025[28]).
National (sub-national) research confirms that more positive climates and relationships are associated with lower levels of SAP. In Ohio (United States), students with higher starting levels of school belonging and gains during the year miss about three fewer days and are seven percentage points less likely to be chronically absent, accounting for demographic and school factors (Ansari et al., 2025[108]). German evidence supports these results by looking at PISA and accounting for a range of other background characteristics (Feldhaus et al., 2025[109]). In Sweden, positive school climate (as reported by students in lower-secondary education) is associated with lower absences, and the “protective” effect of a favourable school climate is larger for students with tertiary-educated parents, illustrating how family resources can amplify microsystem supports (Karlberg et al., 2020[110]). In Denmark, accounting for a range of other factors, poorer teacher-student relationships are predictive of absences (Kristensen, Jensen and Krassel, 2020[20]). In the Slovak Republic, teachers in classrooms with a friendly climate reported roughly 5% fewer student hours missed, though the data are not representative (OECD, 2025[36]). Student voices in the Netherlands underscore what a favourable climate looks like in practice: a safe school culture, adults not looking away from problems, fair treatment, and avoiding practices that over-examine or “question” students (Binsbergen et al., 2019[19]). Indeed, in England (United Kingdom), many children with SEN reported that they are no longer in education because they feel that school staff does not care about them, want them or understand their needs (Ofsted and Care Quality Commission, 2025[106]).
At an internationally comparable level, students who reported being long-term absent feel a lower sense of belonging at school in all countries (panel A in Figure 2.9). The predictive power of school belonging remains significant even in models that account for a range of other factors (Annex Table 2.B.1). The index of school belonging is created by combining responses to statements such as “I feel like I belong at school” or “I feel lonely at school”. Moreover, in most countries, students who reported they were long-term absent attended schools whose school leaders indicated more negative school climates (panel B in Figure 2.9). School climate is estimated based on school leaders’ accounts of, e.g. profanity and vandalism.
Note: Statistically significant differences are marked in darker colours. Panel A of the figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of sense of belonging. The index scales students’ ratings of their agreement with six statements (e.g. “I feel like I belong at school.”, “I feel lonely at school.”). Positive values mean a greater sense of belonging at school among long-term absent students. Panel B of the figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of negative school climate. The index scales school leaders’ answers about the extent of problem behaviours that contribute to a negative school climate in their school (e.g. “Profanity”, “Vandalism”). Positive values mean a more negative school climate among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Relationships with teachers can also shape attendance. In the United States, teacher-reported closeness is linked to lower absences, with modest effects that grow when supportive relationships persist across years (Ansari et al., 2024[111]). Moreover, foster-involved youth in the United States who perceived a caring school adult have significantly lower odds of chronic absence, accounting for a range of demographic factors (Lamb et al., 2022[112]). In Ireland, students’ feelings toward their teacher predicts disaffection, and supportive expectations are associated with lower truancy (Darmody and Thornton, 2014[113]; Darmody, Smyth and McCoy, 2008[114]; McCoy et al., 2007[115]). In New Zealand, not liking at least one teacher is a commonly cited barrier to attendance and is associated with a 19-point drop in regular attendance among secondary learners (ERO, 2022[27]; ERO, 2023[116]). In Italy, weaker student-teacher relationships are significantly associated with “school refusal”, particularly through feelings of alienation from teachers and classmates (Sorrenti et al., 2025[117]). Finnish evidence likewise highlights that warm, caring and proactive outreach by teachers to students who are absent or returning from a period of absence can prevent and reduce absences (Hotulainen et al., 2024[16]). Finally, personal connections between students and teachers, the importance of personalised attention and positive school climate were also highlighted by stakeholders in Chile (Education Quality Agency, 2019[118]). Internationally comparative evidence supports these results: in all but one country (Denmark), long-term absent students reported lower-quality student‑teacher relationships (Figure 2.10). Student-teacher relationships are estimated based on students’ responses to statements such as: “The teachers at my school are respectful towards me”, “When my teachers ask how I am doing, they are really interested in my answer”. The association between long‑term absence and the quality of student-teacher relationships disappears, however, once other factors are accounted for (e.g. socio-economic background, see Annex Table 2.B.1), except for the model that only includes EU countries. While further investigation is needed, this variable is likely related to factors (e.g. school belonging) that mediate or confound the uncontrolled relationships.
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. The figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of quality of student-teacher relationships. The index scales students’ ratings of their agreement with eight statements (e.g. “The teachers at my school are respectful towards me.”, “When my teachers ask how I am doing, they are really interested in my answer.”). Positive values mean a higher quality of student-teacher relationships among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Proximal processes within the school environment, such as daily opportunities for engagement and relationship-building, also play an important role in shaping attendance. In England (United Kingdom), students, teachers and other stakeholders reported that enrichment activities build confidence, friendships and better student-teacher relationships; and these benefits may be strongest for children in poverty (Centre for Young Lives, Leeds Beckett University and YMCA George Williams College, 2025[119]). Several case studies also indicate that teachers’ involvement in enrichment activities (including volunteering off‑site) improves relationships and supports attendance (ibid.). Reports from New Zealand point in the same direction: secondary learners who could not have taken part in certain school activities are 33 percentage points less likely to attend regularly (ERO, 2023[116]). The importance of engaging extra‑curricular activities driving attendance is also highlighted by students in Chile (Education Quality Agency, 2019[118]).
Teacher absences
Parent in France: Then it was the substitute [teacher] for the substitute [teacher] who did not show up... (Battaglia, 2021[120])
Teacher and student absences appear linked, although causal evidence is missing. When regular teachers are absent, schools often restrict attendance, merge classes or rely on short notice substitutes, signalling that the school day will offer fewer routines, weaker relationships and lower instructional value, which can encourage families (and older students) to keep their children at home. In France in 2023/24, 8% of teaching hours in lower-secondary education was not covered due to teachers being absent and not replaced (Court of Auditors, 2025[121]). In England (United Kingdom), during the national teacher strike in 2023, 43% of students attended school (58% in primary schools and 24% in secondary schools), indicating that mass teacher absences can coincide with drops in student attendance (Department for Education, 2023[122]). Most of the missed sessions during this time were “planned”: knowing staff absences would be high, some schools restricted attendance. In Chile, stability of the teaching staff is highlighted as a key factor driving attendance (Education Quality Agency, 2019[118]).
Internationally comparable evidence related to teacher absences and shortage is inconclusive. In most countries, there do not seem to be significant differences in shortages of teaching staff between students who were long-term absent and those who were not (panel A in Figure 2.11). While on average across OECD countries long-term absent students seem to be in schools where school leaders more often reported teacher shortages, this difference is significant in only six countries (in two, the difference points in the opposite direction). Similarly, teacher absences are not significantly related to students’ long-term absence (panel B in Figure 2.11). However, these findings should be interpreted with caution, as the country totals may mask important variation, given that teacher shortages and absences may be unevenly distributed across regions and schools within countries. However, the association between long-term absence and these two variables remains insignificant even in more complex models that account for a range of other factors (Annex Table 2.B.1).
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. Panel A of the figure displays the difference, in percentage points, between students who were long-term absent and those who were not, in the share of students attending schools where the school leader reported that instruction was hindered by a lack of teaching staff to some extent or a lot. Positive values mean a lack of teaching staff acting as a hindrance to learning (according to school leaders) more among long-term absent students. Panel B of the figure displays the difference, in percentage points, between students who were long-term absent and those who were not, in the share of students attending schools where the school leader reported that instruction was hindered by teacher absenteeism to some extent or a lot. Positive values mean teacher absenteeism acts as a hindrance to learning (according to school leaders) more among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Non-instructional service gaps driving absences
School professional in Germany: I think schools often have to find their own ways and look for or even start projects themselves. But it’s probably not really their job; it would be actually better if external organizations took care of that (…) if there were more opportunities and points of contact. But out of necessity, schools are probably just tinkering and putting projects together on their own. (Enderle et al., 2025, p. 12[104])
In some cases, the needs underlying school attendance cases have become more complex since the COVID-19 pandemic, while non-instructional support services have not expanded accordingly. Families sometimes struggle to access specialist guidance, counselling is misunderstood or underused, and school health capacity is limited. Evidence from New Jersey (United States), for instance, indicates that higher student-to-nurse ratios are associated with higher chronic absence, with the strongest effects for Black students, socio-economically disadvantaged students and those in foster care (Markus-Rodden et al., 2025[123]). In Norway, some practitioners reported that the complexity of attendance cases has increased, especially since the pandemic (Bergene et al., 2023[92]). As needs have multiplied and overlapped – physical health, mental health, stress, etc. – frontline staff often lack the non‑instructional supports to respond. These capacity limits mean that complex cases often receive generalist responses, which can prolong absences.
Bottlenecks and weak service design, particularly around identifying students’ needs, deepen the problem. Already before the pandemic, Danish families and schools described long waits for assessments and external help, with most teachers (77%) reporting delays for services like social workers, leading to fragmented, stop-start support (Børns Vilkår and the Egmont Foundation, n.d.[107]). Some Norwegian school leaders likewise noted that many mental-health symptoms remain undiagnosed (Bergene et al., 2023[92]). In France, mental-health-related “school refusal” requires specific mental-health care, but among students with transcultural backgrounds, staff reported cultural taboos and uncertainty that complicate identification and referral (Martin et al., 2020[124]). In England (United Kingdom), some school leaders reported that their schools lack the required capacity and resources to meet the needs (Ofsted and Care Quality Commission, 2025[106]).
Echoing this from the family perspective, roughly one in five parents in Wales (United Kingdom) cited unmet needs as barriers to attendance, underscoring the importance of timely assessment and co‑ordinated support (Parentkind, 2024[91]). In England (United Kingdom), some parents agree, linking attendance difficulties to wider system pressures, including child mental health and SEN backlogs, and express low confidence that schools can secure timely support (Burtonshaw and Dorrell, 2023[125]). Moreover, parents can find the support systems too complicated to navigate, particularly for those who are new to the country (Ofsted and Care Quality Commission, 2025[106]). Similarly, in Denmark, parents said already pre-pandemic that they struggle to find competent guidance on options and on evidence-based responses to anxiety and developmental disorders; teachers echoed this gap, with nearly half (49%) feeling they have little or no opportunity to provide sufficient help to students with worrying absences (Børns Vilkår and the Egmont Foundation, n.d.[107]). As a result, in England (United Kingdom), over half of surveyed children with SEN who are out of school reported that they did not get the required support from practitioners to help them to attend school (ibid.).
Peers
Peer and social influence
Student in the United States: I was with other people. I guess it was kind of like if I’m in class and they’re out there, I’m missing out on something more important with my friends. I was never by myself when I ditched – ever. I think that’s how I thought about it then. I was like, ‘Well, what are they doing when I’m sitting here doing nothing?’ (Dahl, 2015, p. 130[126])
Attendance decisions are often social: many students watch who shows up, absorb the norm and the disruptions in peer presence can nudge them to stay home. In Texas (United States), higher levels of in‑person classmate absence increases the likelihood of individual student absences the following day: a ten percentage point rise in peer absences is associated with a 2.6 point increase in the probability of being absent, even after accounting for student, classroom and date fixed effects (Kirksey et al., 2024[127]). These spill-over effects extend up to two days later but diminish beyond that window (ibid.). Similarly, in Spain, higher peer truancy is associated with both the likelihood of an adolescent engaging in truancy and the number of days they were truant (Escario, Giménez-Nadal and Wilkinson, 2022[128]). In Germany, research based on PISA 2012 finds that students report more truancy when they perceive more classmates in their class as truant, accounting for a range of background characteristics and school factors (Saelzer and Lenski, 2016[129]). New Zealand data reveal that secondary learners who dislike the people in their class are 20 percentage points less likely to attend regularly (30% vs. 50%), and a national literature review identifies “relationships”, including peer relationships, as the most critical theme underpinning absences (ERO, 2023[116]; Richards and Clark-Howard, 2023[130]). Qualitative results from Chile highlight that respect and camaraderie among students are linked to good attendance rates (Education Quality Agency, 2019[118]). These findings support the idea that truancy may be socially contagious. When skipping school becomes normative among peers, it increases both the likelihood that others will follow suit and the frequency of their absences.
More broadly, reports in several education systems indicate that problems with school friendships or peer integration (or, in contrast, alienation) can be triggers for SAP: highlighted in Finland (friendship problems as a recurring cause), Ontario (Canada) (feeling lost and invisible in large schools), Scotland (United Kingdom) (poor peer relationships cited by multiple schools in one sample), and Romania (low attendance linked to difficulties integrating socially and low sense of belonging) (Brown and Birioukov-Brant, 2021[131]; Education Scotland, 2023[95]; Horga et al., 2024[49]; Määttä et al., 2020[132]). These can also be connected to other drivers, e.g. school belonging and bullying.
Peers can also buffer risks inasmuch as classroom composition fosters stability and connection. In California (United States), for instance, familiar faces – a higher share of classmates from the prior year – are linked to lower absences (Jacob Kirksey and Elefante, 2022[133]). A 25-percentage-point increase in peer familiarity is associated with a roughly 1-percentage-point decrease in overall absence, even after accounting for a range of background factors (ibid.).
Bullying, cyberbullying and safety
Student in Australia: It wasn’t safe for me. I got bullied every day. Made me feel... like I was locked up in a cage. (Amin and Ettinger-Epstein, 2024[86])
Bullying is a strong, cross-national predictor of SAP. It can undermine students’ sense of safety and belonging, trigger distress and psychosomatic complaints, and drive avoidance of hostile peer groups. Because bullying can be hidden from school staff and can follow students outside school hours, students may respond by staying home and families may reclassify emotionally driven absences as “illness”, producing increases in both authorised and unauthorised absences. In a recent rapid literature review, being bullied is significantly associated with school absences (Melvin et al., 2025[28]).
Other research confirms that bullying is related to SAP (OECD, 2026[134]). In Ireland, bullying is found to have a negative impact on school attendance, with disaffection often linked to peer relationships (Darmody and Thornton, 2014[113]; Thornton, Darmody and McCoy, 2013[135]). In Finland, adolescents who reported experiencing bullying several times a week had 45% higher odds of illness-related absences and truancy (Alanko et al., 2023[136]). Similarly, being bullied at school is identified as an important factor related to SAP in Germany, New Zealand and Norway (Bergene et al., 2023[92]; ERO, 2022[27]; ERO, 2025[81]; Feldhaus et al., 2025[109]). From the perspective of parents in Finland, factors related to the safety of the school day and bullying emerge as drivers behind absences (Markkanen et al., 2022[137]). In New Zealand, non‑causal evidence suggests that parents who would keep a child home because of bullying are 23 percentage points more likely to have a child who does not attend regularly (ERO, 2023[116]).
Victimisation happening online also increases the risk of SAP. In Spain, the experience of cyberbullying is associated with an increased likelihood of engaging in truancy (Escario, Giménez-Nadal and Wilkinson, 2022[128]). For instance, students who reported being cyberbullied very frequently are nearly three times more likely to be truant than those who did not report so (ibid.). In the United States, accounting for a range of factors, relational victimisation is consistently associated with increased absences (Williford et al., 2020[138]). Cyber victimisation is not directly associated with absences but shows complex interactions when considered alongside perceived teacher attachment, a result possibly shaped by out-of-school dynamics (ibid.).
Limited evidence also suggests that it is not only bullied students who are more likely to be absent, but also those who perpetrate such behaviours (or are both victims and perpetrators). Research from Finland reveals that the odds of truancy increase 1.5 times with frequent bullying victimisation, 8.8 times with frequent bullying perpetration, and 4.5 times with bullying victimisation and perpetration (Alanko et al., 2023[136]).
Internationally comparative data reveal that in all countries, students who reported that they were long‑term absent are more likely to be bullied (Figure 2.12). Being bullied is estimated based on students’ accounts of, e.g. “Other students left me out of things on purpose” or “Other students made fun of me”. This index remains predictive of long-term absence in models that account for a range of other factors (Annex Table 2.B.1).
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. The figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of being bullied. The index scales students’ ratings of how often they had a range of experiences at school that are indicative of being bullied during the past 12 months (e.g. “Other students left me out of things on purpose.”, “Other students made fun of me.”). Positive values mean a greater likelihood of being bullied among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Safety concerns extend beyond bullying to gendered harassment, sexual harassment and broader feelings of unsafety, which can also translate into missed school. In the United Kingdom, the 2025 Girlguiding Girls’ Attitudes Survey found that one in ten girls aged 11-16 said they had missed school to reduce the risk of sexual harassment, with higher shares among disabled girls, neurodivergent girls and LGBTQ+ girls (Sinmaz, 2025[139]). Evidence from Sweden also links gendered victimisation to absences: in a study of grade-nine lower-secondary students, cyber sexual harassment was associated with unauthorised repeated absences among both boys and girls, while both cyber and in-person sexual harassment were significantly associated with unauthorised repeated absences among girls in adjusted models (Dahlqvist, 2026[140]). This suggests that SAP can be driven not only by bullying in the narrow sense, but also by wider safety concerns, such as sexualised forms of victimisation that disproportionately affect girls’ attendance.
Home and family
Socio-economic conditions at home
Student in the United States: My motivation for ditching was, uh, working at Burger King and making money and, uh, you know, having a job and, uh, helping my mom basically, uh, pay bills, putting food on the table. I’d help her pay, you know, different bills, water bills, whatever I could find to help her with. (Dahl, 2015, p. 129[126])
Socio-economic conditions at home can influence absences through distinct but interrelated pathways. On the one hand, financial strain within the home environment can constrain families’ capacity to support regular attendance. For instance, material hardship can affect access to transport, school uniforms, learning materials and digital tools. On the other hand, broader economic conditions, such as labour market opportunities, can incentivise students to be absent from school in favour of immediate earnings. Taken together, these mechanisms help explain why socio-economic background is consistently associated with absences across countries, even though its effect is not necessarily large (Gubbels, van der Put and Assink, 2019[1]; Melvin et al., 2025[28]).
Financial strain is consistently associated with higher levels of absences, although its strength varies across contexts. Multiple systematic/meta-analytic reviews of studies, predominantly from the United States, suggest that family-level measures, such as disadvantaged socio-economic background (e.g. financial hardship, parental education and family structure (e.g. single-parent families)), are reliably associated with higher rates of absence (Gubbels, van der Put and Assink, 2019[1]; Sosu et al., 2021[141]). Similarly, comparative PISA analyses indicate that socio-economic background is the most consistent correlate of skipping school in Sweden and the United Kingdom, with directionally similar but weaker effects in Germany and Japan (Fredriksson et al., 2024[142]). Other national evidence aligns. In Scotland (United Kingdom), all measures of socio-economic background (eligibility for free school meals, parental education, parental class, housing tenure and neighbourhood deprivation) are independently associated with absences (Klein, Sosu and Dare, 2020[143]). In Wales (United Kingdom), persistent absence is framed in the context of the cost of schooling (e.g. school uniforms, supplies, trips, after-school clubs, transport) (Children, Young People and Education Committee, 2022[102]). In the Netherlands, qualitative work with students who were long-term absent highlights serious home problems (e.g. housing and financial stress) as one of the prominent drivers (Binsbergen et al., 2019[19]). Indeed, the effects of neighbourhood conditions and financial strain often compound. For instance, families in poverty are more likely to live in neighbourhoods with multiple social problems (see section Neighbourhood and school environments associated with absence).
More recent trends from some countries reveal a changing pattern. In Germany, analyses of PISA 2012‑22 indicate the expected relationship between socio-economic background and skipping school up to 2018 (Broschinski et al., 2025[144]). However, this pattern reversed in 2022, when students from advantaged families reported higher levels of skipping school alongside a general increase in truancy, even after accounting for gender, age, school type, prior achievement and school belonging (ibid.). In Chile, average annual attendance as well as serious absence rates vary little by socio-economic background of students (Ministry of Education and Centre for Studies, 2025[145]). This suggests that, while socio-economic disadvantage remains a major risk factor for SAP, the strength and pattern of its association can vary across countries and over time, reflecting wider cultural and institutional contexts and broader societal changes.
In addition to these home-based pressures, labour market opportunities can act as a distinct “pull” factor by encouraging students to prioritise short-term earnings over school attendance. Evidence from several countries suggests that students may take on paid work alongside or instead of schooling, particularly in contexts of financial pressure. For example, reports from Romania highlight the need for students to engage in paid work alongside other responsibilities, such as sibling care and support needed from children for family work (Andrei, 2023[48]; Horga et al., 2024[49]; Bÿlÿÿoiu et al., 2024[146]; National Center for Education Policy and Evaluation, 2023[147]). Education Scotland (2023[95]) similarly notes cases of students working regular hours (e.g. holding Friday jobs). Evidence from Extremadura (Spain) suggests higher absences among students in families with unstable or demanding work patterns (General Inspectorate of Education and Evaluation, n.d.[148]). Many students can be attracted to opportunities (whether legitimate or not) presented online for earning money, e.g. through gaming, content creation and social media, often without the need for formal qualifications (Carr, Whitehead and Burtonshaw, 2025[94]). In Argentina, adolescents conducting productive activities have more absences, late arrivals, repetition and vulnerability to dropout (UNICEF and Argentine Society of Pediatrics, n.d.[149]). While these dynamics often overlap with financial strain at home, they point to a separate mechanism whereby the availability of income-earning opportunities can directly compete with school attendance.
Internationally comparative evidence in this regard is presented in Chapter 1 and below. Chapter 1 suggests that students with tertiary-educated parents have lower rates of truancy in most education systems at the primary and secondary levels. For long-term absences, however, the differences are not significant in most countries. Additional angles on the socio-economic conditions of students who reported being long-term absent are provided in Figure 2.13. Panel A reports differences in the index of economic, social and cultural status that accounts for parents’ occupation and home possessions, in addition to parental educational level. Based on this measure, long-term absent students are more likely to be socio‑economically disadvantaged in most countries. The index is also a significant predictor of long-term absence in models that account for a range of other factors, though not in the model that only includes EU countries (Annex Table 2.B.1). Panel B displays a statistic that is more closely related to poverty: not eating at least once a week in the past month. On average across OECD countries, 18.4% long-term absent students reported they had not eaten at least once a week in the past 30 days because there had not been enough money to buy food. This share stands at 7.3% among students who were not long-term absent, yielding a difference of 11.1 percentage points shown in the figure.
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. Panel A of the figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of economic, social and cultural status. This is a composite index that accounts for parents’ occupation, educational level and home possessions. Positive values mean that a more advantaged socio-economic background was more prevalent among long-term absent students. Panel B of the figure displays the difference, in percentage points, between students who were long-term absent and those who were not, in terms of whether they had not eaten at least once a week in the past 30 days because there was not enough money to buy food. Positive values mean that not eating was more common among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Parental health and family functioning
Family circumstances can be a core driver of SAP. Parental mental and physical health difficulties, ineffective family systems, weak parental control and child abuse victimisation can all elevate absence risk (Gubbels, van der Put and Assink, 2019[1]; Melvin et al., 2025[28]). These factors can heighten stress and conflict, disrupt routines, and dilute supervision and expectations, pathways that can lead to more absences. In contrast, positive parent-child interactions are associated with fewer absences (Melvin et al., 2025[28]).
Parental health problems and emotional dysregulation are linked with higher absences and “school refusal”. South Australian data identify parental emotion dysregulation as being associated with “school refusal” (Chen et al., 2024[150]). Echoing these patterns, some Scottish (United Kingdom) local authorities and schools cited parental mental health and anxiety as drivers of non-attendance, with several noting increases since the COVID-19 pandemic (Education Scotland, 2023[95]).
Where parents are unwell, students may also assume caregiving roles that directly pull them from school, as reported by 8% of students in New Zealand (ERO, 2022[27]). In Norway, SAP are more common in families where one or both parents stayed at home due to long-term sick-leave or unemployment (Hysing et al., 2017[151]). Reports from Romania similarly point to a parent’s health as a reason for absences (Horga et al., 2024[49]). In the United States, children with a parent with a history of cancer are more likely to miss school even after accounting for extensive child, parent and family covariates, underscoring how chronic parental illness can constrain caregiving and destabilise routines (Zheng et al., 2022[152]). Another US study reveals that having a sibling with chronic absence is a predictor of experiencing SAP in the following school year (Chu et al., 2019[153]).
Moreover, adverse or unsafe home environments, especially exposure to violence and abuse, can be strong predictors of SAP. In Scotland (United Kingdom), some schools flagged domestic violence as an attendance barrier (Education Scotland, 2023[95]). A Minnesota (United States) study places adverse childhood experiences (e.g. household violence, abuse and substance use) among the top predictors of unauthorised absences, highlighting the disruptive role of instability and threat in the home (Lee et al., 2023[154]). In Iceland, school leaders reported that difficult home circumstances are a reason for school avoidance among 29% of primary-education students (Ministry of Social Affairs and Housing, 2019[97]). Reports from Denmark also identify complex social problems at home as primary causes of SAP and document more absences for children in vulnerable positions (Ilsvard et al., 2024[155]; Rambøll, 2018[156]). Research from Finland similarly notes recurring home-condition problems as reasons for absences (Määttä et al., 2020[132]). Evidence from a meta-analysis further indicates that a history of child abuse victimisation elevates absence risk (Gubbels, van der Put and Assink, 2019[1]).
Family functioning and engagement also matter. A meta-analytic review links low parental support/acceptance, ineffective family systems and low parental control with more absences. German and Scottish (UK) reports similarly cite low parental interest/support as reasons for SAP (Education Scotland, 2023[95]; Feldhaus et al., 2025[109]; Gubbels, van der Put and Assink, 2019[1]). Romanian responses additionally point to family relationship problems as a contributor to absences (Horga et al., 2024[49]). Similarly, in Extremadura (Spain), constant conflict in families and disruption within the family, as well as family neglect, are identified as drivers of SAP (General Inspectorate of Education and Evaluation, n.d.[148]).
At an internationally comparable level, students who reported they had been long-term absent are more likely to live in less supportive families (Figure 2.14). The index of family support scales students’ ratings of how often their parents or someone else in their family engaged in behaviours indicative of family support (e.g. “Discuss how well you are doing at school”, “Spend time just talking with you”). Higher family support is significantly predictive of lower long-term absence also in models that account for a range of other factors (Annex Table 2.B.1).
Difference between students who reported being long-term absent and those who did not
Note: Statistically significant differences are marked in darker colours. The figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of family support. The index scales students’ ratings of how often their parents or someone else in their family engaged in a range of behaviours indicative of family support (e.g. “Discuss how well you are doing at school”, “Spend time just talking with you”). Positive values mean greater family support among long-term absent students.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
Physical health
Parent in Finland: She is under such extreme stress and strain that it paralyses her. It may also show up as physical symptoms, such as stomach pain. (Kaipainen, 2023[157])
Physical health affects whether students can attend school in multiple ways, through illness, pain and fatigue, as well as through sleep and nutrition that shape daily readiness. When sickness is common, poorly managed, or perceived by families as a reason to keep children home, absences can rise and learning time can be lost. Poor physical health was identified as a significant absence risk factor in a meta‑analysis and rapid literature review (Gubbels, van der Put and Assink, 2019[1]; Melvin et al., 2025[28]). Students who experience ill health, both in terms of physical impairments and mental disorders, are often considered as having SEN. This contributes to the fact that, as elaborated in Chapter 1, students with SEN often experience more SAP.
Across several education systems, data indicate that illness is a major driver of missed days. In Finland, illness is the most common reason for absences, with over 70% of primary and lower-secondary students in the (limited) sample reporting at least 1-2 days off for illness in the three months before the data collection (Hotulainen et al., 2024[16]). Ireland sees illness-related absences climb post-pandemic, from 26.9% in 2020/21 to 39.9% in 2023/24 in primary schools, and from 18.4% to 28.1% in secondary schools (TESS, 2024[79]; TESS, 2025[80]). Almost all parents in Ireland also believe that sickness that includes temperature and medical appointments are acceptable reasons for missing school (RED C, 2025[90]). In Australia, while almost all types of absences have increased, illness-related SAP have almost doubled, from 6.6 days absent per student in 2017 to 11.6 in 2024 (Hunter, Haywood and Chapman, 2025[33]). Similarly, school leaders in Norway reported more severe cases of students being bedridden after the COVID-19 pandemic (Bergene et al., 2023[92]). In New Zealand, health is the most frequently named barrier to attendance by parents: two-thirds said they are likely to keep a child home for a minor or chronic condition, 76% actually did so for a minor illness in 2022, and 27% of chronically absent students themselves cited physical health as a reason for absences (ERO, 2022[27]; ERO, 2024[14]). Some Romanian secondary-level voices echo these realities: students cited illness and medical appointments during school hours, while teachers reported frequent medical exemptions to justify accumulated absences (Horga et al., 2024[49]). Finally, evidence from Denmark provides a more nuanced picture. While most physical impairments and learning disabilities do not correlate with absences, students with autism spectrum disorders and behavioural disorders experience more periods of high absence, longer periods of absences and higher total absence (Kristensen, Jensen and Krassel, 2020[20]).
Health-related absences can reinforce educational disadvantage, particularly when recurrent sickness absences cluster with other risks and gender-specific health needs, such as menstrual problems. In a regional study of primary education in the Netherlands, for instance, most students are absent at least once during the school year (85% in mainstream and 79% in special schools), with sickness absences the most common form of absence, affecting 75% and 71% of students, respectively (Pijl et al., 2021[158]). Extensive sickness absences affect 13% of students in mainstream education and 23% in special education, and cluster with other markers of vulnerability, including low parental education, more doctor visits, tardiness and unauthorised absences (ibid.). Menstrual problems may similarly contribute to absences. In one systematic review, 20% of women are absent from educational institutions due to dysmenorrhea6, with an estimated 12% in high-income countries (Armour et al., 2019[159]). Qualitative syntheses further indicate that menstrual pain, heavy bleeding, fear of leakage, stigma, inadequate toilet access and limited school support can shape girls’ attendance, concentration and participation at school (Barrington et al., 2021[160]; Thomas and Melendez‐Torres, 2024[161]).
Chronic and complex health needs can increase the risk of absences throughout adolescence and contribute to exclusion from learning. In England (United Kingdom), adolescents with diagnosed chronic health conditions – covering physical illnesses (e.g. asthma, epilepsy and diabetes) – are at substantially greater risk of persistent absence (Jay et al., 2025[162]). Elevated risks of exclusion and non-enrolment are also observed, underscoring that absences are a key mechanism by which health-related disadvantage becomes educational disadvantage, notably when schools cannot adequately support complex needs (ibid.). Similarly, children with special health care needs due to a chronic health condition in Germany have higher absence rates (Schlecht et al., 2023[163]). They are particularly elevated among those with functional limitations, treatment or counselling for emotional, behavioural or developmental problems, and those who experience two or more special health care consequences (ibid.).
A separate strand of research also looks at the connections between sleep (or the lack of) and SAP. Sleep difficulties are rising. For instance, in 2014, 17% of 15-year-old boys across multiple OECD and EU countries reported experiencing difficulties getting to sleep more than once a week (Health Behaviour in School-aged Children study, 2023[164]). In 2022, the share rose to 23%. For girls, the shares stood at 27% in 2014 and 37% in 2022 (ibid.). A meta-analysis synthesises that most sleep parameters (bedtime, duration, efficiency, insomnia symptoms, fatigue and sleepiness) are associated with absences (Kjeøy and Lysvik, 2023[165]). Among primary and secondary education students in Japan, irregular sleep patterns are associated with school non-attendance, with students experiencing sleep rhythm problems (e.g. difficulty waking or falling asleep) having substantially higher odds of absences (the relationship likely being bidirectional) (Hirata et al., 2026[166]). This is echoed in interviews with Romanian students experiencing SAP, who cited difficulties in adapting to the school schedule, including waking up in the morning (Horga et al., 2024[49]). Finally, qualitative research from students in England (United Kingdom) indicates that exhaustion, which can be driven by spending time on digital devices, is a key reason for missing school (Carr, Whitehead and Burtonshaw, 2025[94]).
Nutritional diseases have also been linked to SAP, even though the evidence is older and scarcer. In several primary and secondary education schools in rural Colombia, stunted growth and overweight students have significantly more absence days compared to students with adequate nutritional status (Vargas cruz et al., 2016[167]). Similarly, research among 12-19-year-olds in Mexico indicates that anemia can be a significant determinant of school attendance (Mosiño, Villagómez-Estrada and Prieto-Patrón, 2017[168]). In the Netherlands, evidence links being overweight with a higher risk of long‑term absence and staying at home (Binsbergen et al., 2019[19]).
Evidence from PISA confirms that sickness is a key reason for students’ long-term absence. Figure 2.1 above demonstrates that it is by far the most commonly cited factor. Figure 2.15 displays that, on average across OECD countries, 70.5% of 15-year-old students cited this reason, with shares rising to above 80% of students in many countries.
Percentage of students who reported sickness as a reason for having missed school for more than three consecutive months
Note: * Caution is required when interpreting estimates because one or more PISA sampling standards were not met (see Reader’s Guide, Annexes A2 and A4 in OECD (2023[61])).
Source: OECD (2023[3]), PISA 2022 Results (Volume II): Learning During – and From – Disruption, Table II.B1.3.55, https://doi.org/10.1787/a97db61c-en.
Substance use also intersects with absences. A meta-analytic review finds that smoking, drug abuse and alcohol abuse are significant risk factors for absences (Gubbels, van der Put and Assink, 2019[1]). Country evidence echoes this: in Finland, substance use in the sixth grade (the last year of primary education) is most typical among surveyed students with early-onset SAP (Hotulainen et al., 2024[16]). In the Netherlands, among students missing three or more months, problematic cannabis use is linked to a greater risk of long-term absence and staying at home (Binsbergen et al., 2019[19]). However, these associations likely reflect, at least in part, simultaneity and reverse causality, whereby absences may also increase the risk of substance use, and both may stem from shared underlying factors, such as mental‑health difficulties and adverse environments (see also Chapter 3). As a result, substance use is best understood as part of a broader constellation of interrelated risks (e.g. sleep disruption, mental-health strain) that can compound and sustain patterns of missed learning.
Mental health and well-being
Parent in Wales (United Kingdom): COVID completely ruined her. Locked in, she couldn’t continue with her sports … She became depressed and self harmed. She could not face school. She did not attend … Loss of peers through lack of contact. Loss of muscle tone and fitness because of loss of sports. Made a major impact. (Children, Young People and Education Committee, 2022, p. 17[102])
Mental-health difficulties can undermine students’ ability to attend school by driving anxiety-based avoidance, low mood and greater use of clinical appointments. They also interact with school conditions (e.g. peer dynamics, academic stress and transitions) to turn short spells of absences into patterns of persistent non-attendance. A meta-analytic review identifies psychiatric symptoms and disorders, especially depression and anxiety, as significant risk factors for absences (Gubbels, van der Put and Assink, 2019[1]). Indeed, across numerous studies, poor mental health – including emotional difficulties, internalising and externalising symptoms, and diagnosed anxiety or depression – is consistently linked with higher rates of SAP (Melvin et al., 2025[28]).
Mental health challenges as a driver of absences are not a new factor. In Iceland, for instance, reports from 2019 estimate that around 2.2% of primary-education students struggled with school avoidance, with anxiety and depression cited by roughly three-quarters of school leaders as the main reasons (Ministry of Social Affairs and Housing, 2019[97]). More recent causal analyses using pre-pandemic longitudinal data from England (United Kingdom) suggest that poorer mental health contributes to absences (especially authorised absences) (Arnot, 2025[169]). HeadStart’s7 findings align: as mental health difficulties increase, absences increase, with COVID-19 exacerbating absences among young people who already have anxiety (Lereya and Deighton, 2019[170]; McDonald, Lester and Michelson, 2022[171]). Mental disorders also appear to play a significant role in Australian secondary schools (Lawrence et al., 2019[172]). New Zealand brings the family lens into focus. Nearly half of parents reported they would keep a child home for mental health reasons, and over half of chronically absent students cited mental health as a driver of their absences (ERO, 2024[14]; ERO, 2025[81]).
Evidence from other education systems suggests that the situation is more nuanced and not uniformly worsening. In Norway, interviews with school leaders describe an increase in worrying absences linked to mental-health problems (anxiety and restlessness) and concerns about rising autism-spectrum diagnoses outpacing capacities in mainstream education (Bergene et al., 2023[92]). Yet, national survey data indicate no continuing rise in self-reported mental-health problems in recent years and a decline among secondary‑education girls, even as contact with psychologists has grown (Bakken, 2024[173]). Portugal’s report indicates that psychological vulnerability is easing, from about one-third of students in 2022 to about one-quarter in 2024 (Matos et al., 2024[174]). Another slightly contrasting view comes from some school leaders in England (United Kingdom) who observe a growing tendency for students (or their parents) to label everyday challenges as mental-health problems, raising concerns that this shift in language may blur the line between typical stressors and clinically significant conditions (Moore and Walker, 2025[96]).
Nevertheless, risk profiles for SAP consistently point to internalising problems and co-occurring conditions. Research in the Netherlands (pre-COVID) associates long-term absence and staying at home with internalising difficulties and excessive stress across life domains (Binsbergen et al., 2019[19]). Reports in Denmark similarly link psychological challenges, such as anxiety, with more absences (Kristensen, Jensen and Krassel, 2020[20]; National Agency for Education and Quality, 2024[175]; Rambøll, 2018[156]). In Wales (United Kingdom), studies indicate that students with neurodevelopmental or mental-health disorders, or with a record of self-harm, are more likely to miss school than peers, even after accounting for age, sex and deprivation (John, 2021[176]). Policy reviews echo that anxiety and well-being issues, both pre-existing and COVID-19 amplified, are central to understanding recent shifts in attendance (Rowlands, 2022[103]). Notably, anxiety and anxiety-related symptoms are the most prominent drivers of “school refusal” (Leduc et al., 2022[50]). England’s (United Kingdom) pre-COVID longitudinal research adds that children with long-term physical conditions have poorer mental health and more absences (including persistent absence), with strong associations for neurodevelopmental conditions, migraines and atopic conditions, reinforcing the need for routine mental-health and attendance enquiry in clinical and school settings (Finning et al., 2021[177]). More recently, anxiety, overwhelm, burnout and emotional exhaustion were cited by some students as reasons for their absence (Carr, Whitehead and Burtonshaw, 2025[94]). Researchers in Finland underline how shocks in a student’s immediate circle or school (e.g. trauma or serious illness/death of a loved one) can heighten susceptibility to SAP (Sergejeff, 2023[47]). In the Slovak Republic, unpublished research indicates adolescents indifferent to schooling are more likely than satisfied peers to feel hopeless, report multiple weekly health complaints, engage in fighting, and skip school, suggesting a clustering of disengagement, distress and absences (OECD, 2025[36]).
Worries about schoolwork and examinations, as well as the reaction of other students and school staff upon return after a period of absence, are cited as pressures that make coming back to school more difficult for some students in England (United Kingdom) (Moore and Walker, 2025[96]). This is particularly true if the reasons for absences are personal or sensitive to discuss with others (ibid.). Similarly, some Welsh (United Kingdom) learners develop anxieties about returning to large, busy schools or have preferences for studying digitally from home (Rowlands, 2022[103]).
Disengagement, boredom and motivation
Student in Romania: I don't do anything at certain times, so I prefer to rest for my after-school activities. Because it's frustrating to sit in a classroom for 6-7 hours without learning anything that really helps you in real life or at least that is something interesting. (Horga et al., 2024, p. 57[49])
Disengagement – whether emotional, behavioural, social or cognitive – helps explain why SAP rarely have a single cause, as school experiences, relationships, perceived relevance and individual motivation can reinforce one another over time. One way to understand this web of relationships is through the lens of attachment theory. In this view, SAP arise when the bonds between students and their schools weaken. Small pressures (feeling unsafe, low belonging, strained relationships, etc.) first erode emotional and cognitive engagement, then show up as lateness, selective absences and ultimately truancy (Keppens and Spruyt, 2019[178]). In England (United Kingdom), for instance, students in the top quarter for engagement are ten percentage points less likely to be persistently absent than those in the bottom quarter of engagement (Jerrim, 2025[179]). Since the COVID-19 pandemic, students’ emotional engagement has dropped in many countries, with secondary-education students reporting big falls in feeling safe and proud of their school, and girls’ engagement often falling more than boys’ (Jerrim and Kaye, 2025[180]).
Digital pull factors can also weaken students’ engagement with schooling by competing for attention, reducing motivation to learn and increasing the risk of SAP, particularly when digital leisure use becomes excessive (OECD, 2025[181]). Digital technologies may also affect attendance more indirectly through their effects on well-being. Higher screen time is associated with anxiety and depression and may displace sleep, exercise and in-person social interaction, all of which can make sustained engagement with school more difficult (OECD, 2024[182]). In parallel, research on digital distraction indicates that social networking, messages and alerts can interrupt concentration, reduce attention and leave students less engaged in learning (Martin et al., 2025[183]). Indeed, Norwegian stakeholders connect disengagement to digital pull factors (e.g. social media and gaming), which can compete with school for time and interest, especially among students already ambivalent about the classroom (Bergene et al., 2023[92]).
Disinterest and the lack of motivation are also linked to school attendance. In Norway, from 2018 to 2023, fewer students reported being motivated, interested in learning or looking forward to school (Norwegian Directorate for Education and Training, 2024[184]). In Denmark, Extremadura (Spain) and Scotland (United Kingdom) stakeholders linked a lack of interest or motivation to more SAP (Education Scotland, 2023[95]; General Inspectorate of Education and Evaluation, n.d.[148]; Rambøll, 2018[156]). Related evidence from the Slovak Republic associates indifference to schooling with higher risk skipping school (OECD, 2025[36]). In Finland, based on a limited sample, low-absence students reported lower amounts of cynicism, whereas early-absence students showed the highest rates of questioning of the value of studying (Hotulainen et al., 2024[16]). New Zealand’s findings about participation barriers and perceived relevance reinforce how these attitudes translate directly into day-to-day attendance decisions (ERO, 2023[116]). Secondary students who are interested in lessons are 18 percentage points more likely to attend them (57% vs. 39%), and seeing school as important for the future adds 26 points (52% vs. 26%). Learners who think going to school every day matters are 23 points more likely to attend regularly (60% vs 37%) (ibid.). Some students’ accounts from England (United Kingdom) suggest that the motivation for attending is closely linked to obtaining qualifications (and, by extension, improved outcomes in life), but school is not viewed as a formative experience itself (Carr, Whitehead and Burtonshaw, 2025[94]).
Finally, disengagement can be triggered by boredom in school, particularly if boredom is chronic. While definitions are difficult to pin down, and the relationship between disengagement and boredom is likely bidirectional, the two concepts are linked (Macklem, 2015[185]). It is, therefore, worrying that boredom is increasingly widespread in some countries (Bakken, 2024[173]). Internationally comparable evidence confirms that boredom is one of the key reasons for long-term absence (Figure 2.1 above). Panel A in Figure 2.16 further displays that in some countries, such as in Bulgaria, Cyprus, Greece, Romania and Thailand, more than a third of students cited boredom as a factor.
Even though conceptually more related to the school microsystem, evidence from PISA suggests that students who reported being long-term absent are more likely to be in classroom environments that are noisy, where students do not listen to the teacher, or where students are distracted by digital resources (panel B in Figure 2.16). These behaviours may contribute to, reflect or result from student disengagement, demotivation or boredom. These and other aspects of classroom behaviour are captured in the index of disciplinary climate, which is based only on mathematics lessons. In all countries, disciplinary climate was reported to be less positive among long-term absent students. Disciplinary climate remains a significant predictor of long-term absence even in models that account for a range of other factors (Annex Table 2.B.1).
Percentage of students who reported boredom as a reason for long-term absence (panel A) and difference in disciplinary climate between students who reported being long-term absent and those who did not (panel B)
Note: * Caution is required when interpreting estimates because one or more PISA sampling standards were not met (see Reader’s Guide, Annexes A2 and A4 in OECD (2023[61])). Panel A of the figure displays the percentage of students who reported boredom as a reason for having missed school for more than three consecutive months. Panel B of the figure displays the difference, in index points, between students who were long-term absent and those who were not in the index of disciplinary climate in mathematics. The index scales a range of situations that students reported occurring in their mathematics lessons (e.g. “Students do not listen to what the teacher said.”, “Students get distracted by using digital resources”). Positive values mean a more positive disciplinary climate among long-term absent students. Statistically significant differences are marked in darker colours.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
In some education systems, concerns about the relevance of the curriculum and the nature of classroom instruction have emerged as important factors in understanding student disengagement and SAP. In Illinois (United States), higher academic rigour observed in classrooms is associated with higher attendance over the school year. The association points to the role of well-structured, challenging instruction and day-to‑day teacher-student interaction in supporting regular presence (Basileo, Lyons and Toth, 2024[186]). Stakeholders’ accounts in Romania suggest that low student motivation, weak alignment between interests and school pathways, and limited applied learning are all being linked to absences (Horga et al., 2024[49]; National Center for Education Policy and Evaluation, 2023[147]). In Wales (United Kingdom), many parents question curriculum relevance and demand more life skills to be taught, greater diversity of vocational options and stronger preparation for the future (Parentkind, 2024[91]). While not necessarily related to the curriculum, many students in England (United Kingdom) complained of limited space for autonomy, self‑expression and too much rigidity (Carr, Whitehead and Burtonshaw, 2025[94]).
As noted in previous sections, SAP drivers rarely act alone. This section provides a few examples of how chronosystem, macrosystem, exosystem, mesosystem, microsystem and person characteristics can interact. Rather than providing an exhaustive overview of such interactions, it aims to convey the message that the different levels do not act in isolation – a consideration that policymakers might make when designing policies and practices to tackle SAP (see also Chapter 4) (Heyne, 2025[11]).
Based on quantitative analyses of linguistically diverse caregivers and their children’s attendance patterns in a school district in the United States, Jones et al. (2024[187]) reveal that Spanish-speaking caregivers are significantly more likely than English-speaking caregivers to report limited or no access to computers, the Internet and printers as barriers to participation. They are also more likely to endorse the need for school‑based supports, such as academic help, healthcare access, transportation and meals. These patterns suggest that systemic inequities (macrosystem) can interact with linguistic background to generate unequal access to key learning resources (exosystem). Qualitative responses deepened this picture, highlighting how inflexible caregiver work schedules – more common among linguistically diverse families – limits their capacity to support remote learning, further straining home-school communication (mesosystem). At the microsystem level, students’ daily learning environments are shaped by disrupted routines and limited academic feedback, while personal experiences, such as anxiety, stress and diminished motivation, further impede engagement. Several language groups, particularly Spanish- and Russian-speaking, are more likely than English-speaking caregivers to report that falling behind in schoolwork is a reason for their child’s absences before the COVID-19 pandemic. Spanish-speaking caregivers are also more likely to cite factors such as suspension or not feeling safe at school. The pandemic intensified these challenges by disrupting established structures and amplifying pre-existing vulnerabilities. Although the study does not formally test cross-level interactions, its integration of comparative statistical results and caregiver narratives illustrates how absences and disengagement may stem from converging pressures across multiple ecological layers.
Another example comes from Detroit, Michigan (United States). Singer et al. (2021[188]) analyse how person characteristics and factors at different ecological levels simultaneously influence absences: person (including economic disadvantage, special education status and ethnicity), microsystem (school-level conditions, such as student stability, school sector, discipline rate and the concentration of disadvantaged students), exosystem (neighbourhood-level violent crime and residential vacancy), and chronosystem (prior-year absences and mobility across schools or residences). The study reveals, for instance, that even independent of individual disadvantage, the composition of the school matters: a one standard deviation increase in the percentage of economically disadvantaged students at a school is associated with 46% higher odds of chronic absence, and this association remains statistically significant, though reduced to 26%, after accounting for prior-year absences. Indeed, when prior absences are included, the predictive strength of several socio-economic indicators declines, including student-level economic disadvantage, school-level disadvantage and neighbourhood-level residential vacancy. This attenuation suggests that persistent absence may reflect entrenched disadvantage rather than additive or escalating effects over time. Although the models do not include formal statistical interaction terms, the simultaneous modelling of variables across multiple ecological levels illustrates that factors at the individual, school and neighbourhood levels are each associated with absences. In particular, the findings show that school composition matters, in addition to individual disadvantage.
Finally, Enderle et al. (2025[104]) conducted a qualitative study in Germany that provides a conceptually grounded account of how young people with social, emotional and behavioural difficulties describe the factors they perceived as helpful in preventing or addressing SAP. They identify a range of interacting influences across system levels that students associated with both the onset and resolution of attendance difficulties. Students described how a lack of co-ordinated communication between school staff, families and external services left them feeling unsupported and confused about expectations – particularly during key transition points – which intensified their disengagement from school. Some also recounted being excluded from decisions about their own reintegration plans, which undermined their sense of agency and trust, especially when family-school co-ordination occurred without their input. In other cases, legal or disciplinary responses were perceived as punitive and emotionally disconnected, making students more reluctant to return to school. While students did not explicitly name broader systems, the authors interpret these accounts as reflecting the compounding influence of macrosystem-level pressures and emotionally distant microsystem relationships. The absence of consistent peer belonging and tailored teaching practices further amplified these challenges, particularly when students felt that their needs were not understood or their progress was not recognised. Together, these accounts illustrate how fragmented or poorly aligned actions across systems can function as drivers of absences, either by escalating distress or by weakening students’ confidence in the value and accessibility of education. While the sample is small and locally specific, the study offers valuable insights into how students with social, emotional and behavioural difficulties experience the intersection of personal, relational and structural influences on both their challenges with attendance and their efforts to return to school.
References
[136] Alanko, K. et al. (2023), “Time Trends in Adolescent School Absences and Associated Bullying Involvement Between 2000 and 2019: A Nationwide Study”, Child Psychiatry & Human Development, Vol. 56/3, pp. 640-647, https://doi.org/10.1007/s10578-023-01601-1.
[86] Amin, M. and S. Ettinger-Epstein (2024), The kids who fear school, https://www.abc.net.au/news/2024-04-29/school-refusal-cant-australia-education-four-corners/103669970 (accessed on 22 August 2025).
[48] Andrei, R. (2023), Studiu privind cauzele abandonului școlar [Study on the Causes of School Dropout], GO-AHEAD Association, https://go-ahead.ro/wp-content/uploads/2024/02/STUDIU-Cauzele-abandonului-scolar.pdf (accessed on 11 August 2025).
[111] Ansari, A. et al. (2024), “The cumulative, timing‐specific, and enduring associations between student–teacher relationships and early elementary outcomes”, Child Development, Vol. 96/2, pp. 475-491, https://doi.org/10.1111/cdev.14177.
[108] Ansari, A. et al. (2025), Rebuilding Connections Post-COVID: The Importance of School Belonging for Addressing Absenteeism, https://crane.osu.edu/our-work/rebuilding-connections-post-covid-the-importance-of-school-belonging-for-addressing-absenteeism/ (accessed on 18 August 2025).
[22] Ansari, A. and R. Pianta (2019), “School absenteeism in the first decade of education and outcomes in adolescence”, Journal of School Psychology, Vol. 76, pp. 48-61, https://doi.org/10.1016/j.jsp.2019.07.010.
[159] Armour, M. et al. (2019), “The Prevalence and Academic Impact of Dysmenorrhea in 21,573 Young Women: A Systematic Review and Meta-Analysis”, Journal of Women’s Health, Vol. 28/8, pp. 1161-1171, https://doi.org/10.1089/jwh.2018.7615.
[169] Arnot, M. (2025), The relationship between mental ill health and absence in students aged 13 to 16: Results from the longitudinal study of young people – cohort 2, https://assets.publishing.service.gov.uk/media/681b676c9ef97b58cce3e518/The_relationship_between_mental_ill_health_and_absence_in_students_aged_13_to_16.pdf (accessed on 14 August 2025).
[24] Attendance Works and Healthy Schools Campaign (2015), Mapping the Early Attendance Gap: Charting a Course for Student Success, https://www.attendanceworks.org/mapping-the-early-attendance-gap/ (accessed on 19 December 2025).
[41] Bagnall, C., C. Fox and Y. Skipper (2021), “When is the ‘optimal’ time for school transition? An insight into provision in the US”, Pastoral Care in Education, Vol. 39/4, pp. 348-376, https://doi.org/10.1080/02643944.2020.1855669.
[173] Bakken, A. (2024), Ungdata 2024: Nasjonale resultater [Youth data 2024: National results], NOVA, OsloMet.
[58] Bardach, L. et al. (2023), “Does students’ awareness of school-track-related stereotypes exacerbate inequalities in education?”, npj Science of Learning, Vol. 8/1, https://doi.org/10.1038/s41539-023-00203-9.
[186] Basileo, L., M. Lyons and M. Toth (2024), “Leading Indicators of Academic Achievement: Investigating the Predictive Validity of an Observation Instrument in a Large District”, Sage Open, Vol. 14/2, https://doi.org/10.1177/21582440241261119.
[120] Battaglia, M. (2021), Dans les établissements scolaires, recherche remplaçants désespérément [In schools, desperate search for replacements], https://www.lemonde.fr/societe/article/2021/11/23/dans-les-etablissements-scolaires-recherche-remplacants-desesperement_6103247_3224.html (accessed on 22 August 2025).
[39] Beatson, R. et al. (2023), “Improving primary to secondary school transitions: A systematic review of school-based interventions to prepare and support student social-emotional and educational outcomes”, Educational Research Review, Vol. 40, p. 100553, https://doi.org/10.1016/j.edurev.2023.100553.
[92] Bergene, A. et al. (2023), Fravær som vekker bekymring [Absences that raise concerns], https://hdl.handle.net/11250/3087762 (accessed on 16 June 2025).
[72] Bimardhika, E. and L. Moorena (2024), Disruption to Schooling: Evidence from Volcano Eruptions on Java Island, Indonesia, Asian Development Bank Institute, https://doi.org/10.56506/jvvo2998.
[19] Binsbergen, M. et al. (2019), Niet thuisgeven: Schooluitval vanuit het perspectief van thuiszitters [Do not give up: School dropout from the perspective of home sitters], Kohnstamm Instituut, https://www.nro.nl/onderzoeksprojecten/niet-thuisgeven (accessed on 11 August 2025).
[198] Bond, C. et al. (2024), “School non-attendance”, Educational and Child Psychology, Vol. 41/1, pp. 5-8, https://doi.org/10.53841/bpsecp.2024.41.1.5.
[59] Boone, S. and J. Demanet (2020), “Track choice, school engagement and feelings of perceived control at the transition from primary to secondary school”, British Educational Research Journal, Vol. 46/5, pp. 929-948, https://doi.org/10.1002/berj.3606.
[107] Børns Vilkår and the Egmont Foundation (n.d.), Håndtering af Bekymrende Fravær i Folkeskolen [Handling Worrying Absences in Primary School], https://bornsvilkar.dk/fravaer-i-folkeskolen/ (accessed on 12 August 2025).
[6] Bronfenbrenner, U. (1995), Developmental ecology through space and time: A future perspective, American Psychological Association, Washington, https://doi.org/10.1037/10176-018.
[5] Bronfenbrenner, U. (1981), The Ecology of Human Development, Harvard University Press, https://doi.org/10.2307/j.ctv26071r6.
[4] Bronfenbrenner, U. (1977), “Toward an experimental ecology of human development.”, American Psychologist, Vol. 32/7, pp. 513-531, https://doi.org/10.1037/0003-066x.32.7.513.
[7] Bronfenbrenner, U. and S. Ceci (1994), “Nature-nurture reconceptualized in developmental perspective: A bioecological model.”, Psychological Review, Vol. 101/4, pp. 568-586, https://doi.org/10.1037/0033-295x.101.4.568.
[10] Bronfenbrenner, U. and P. Morris (2007), The Bioecological Model of Human Development, Wiley, https://doi.org/10.1002/9780470147658.chpsy0114.
[144] Broschinski, S. et al. (2025), “Forschungsnotiz zu PISA 2012–2022: Analysen zum Einfluss des familialen Hintergrunds auf den Schulabsentismus, Research Note on PISA 2012–2022: Analyses on the Influence of Family Background on School Absenteeism”, KZfSS Kölner Zeitschrift für Soziologie und Sozialpsychologie, Vol. 77/2, pp. 247-269, https://doi.org/10.1007/s11577-025-01008-x.
[76] Burdick-Will, J., M. Stein and J. Grigg (2019), “Danger on the Way to School: Exposure to Violent Crime, Public Transportation, and Absenteeism”, Sociological Science, Vol. 6, pp. 118-142, https://doi.org/10.15195/v6.a5.
[142] Burhan, N. (ed.) (2024), “Which students skip school? A comparative study of sociodemographic factors and student absenteeism using PISA data”, PLOS ONE, Vol. 19/5, p. e0300537, https://doi.org/10.1371/journal.pone.0300537.
[125] Burtonshaw, S. and E. Dorrell (2023), Listening to, and learning from, parents in the attendance crisis, Public First, https://www.publicfirst.co.uk/wp-content/uploads/2023/09/ATTENDANCE-REPORT-V02.pdf (accessed on 16 June 2025).
[146] Bÿlÿÿoiu, C. et al. (2024), Liceeni din România. Implicarea și autonomia liceenilor [High school students from Romania. Involvement and autonomy of high school students], National Center for Educational Policies and Evaluation, https://www.ise.ro/liceeni-din-romania-implicarea-si-autonomia-liceenilor (accessed on 14 August 2025).
[65] Camp, A. et al. (2026), Immigration Enforcement Actions and Empty Desks: Persistent and Acute Attendance Effects, https://doi.org/10.26300/bz9z-w627.
[93] Carr, K. et al. (2026), How Schools Work for Every Child, https://theparentvoiceproject.org.uk/wp-content/uploads/Parent%20Voice%20Project%20Report%202.pdf (accessed on 7 May 2026).
[94] Carr, K., P. Whitehead and S. Burtonshaw (2025), Listening to, and learning from, young people in the attendance crisis, https://impetus-org.files.svdcdn.com/production/assets/publications/Report/listening-learning-young-people-attendance-crisis.pdf?dm=1763151286 (accessed on 17 April 2026).
[119] Centre for Young Lives, Leeds Beckett University and YMCA George Williams College (2025), Beyond the classroom: The role of enrichment in tackling the school absence crisis, https://cdn.prod.website-files.com/659fd56cbd8d3f4a80aaac76/67a614f50d1d776cdb401fe9_e395f5cde76251aafb2430a5c27847d9_24_13_031%20-%20Report%20-%20Centre%20for%20Young%20Lives_FINAL.pdf (accessed on 18 August 2025).
[85] CESE (2024), Understanding attendance - An updated research review on school attendance, Centre for Education Statistics and Evaluation, https://education.nsw.gov.au/about-us/education-data-and-research/cese/publications/research-reports/understanding-attendance (accessed on 12 August 2025).
[150] Chen, J. et al. (2024), “School Refusal Behaviors: The Roles of Adolescent and Parental Factors”, Behavior Modification, Vol. 48/5-6, pp. 561-580, https://doi.org/10.1177/01454455241276414.
[102] Children, Young People and Education Committee (2022), Pupil absence, Welsh Parliament, https://senedd.wales/media/jcvef0z5/cr-ld15456-e.pdf (accessed on 12 August 2025).
[73] Children’s Ombudsman (2017), Alle kinderen kansrijk: Het verbeteren van de ontwikkelingskansen van kinderen in armoede [All children have a chance: Improving the development opportunities of children in poverty], https://www.kinderombudsman.nl/publicaties/alle-kinderen-kansrijk (accessed on 14 August 2025).
[200] Childs, J. and C. Scanlon (2022), “Coordinating the Mesosystem: An Ecological Approach to Addressing Chronic Absenteeism”, Peabody Journal of Education, Vol. 97/1, pp. 74-86, https://doi.org/10.1080/0161956x.2022.2026722.
[153] Chu, B. et al. (2019), “Developing an Online Early Detection System for School Attendance Problems: Results From a Research-Community Partnership”, Cognitive and Behavioral Practice, Vol. 26/1, pp. 35-45, https://doi.org/10.1016/j.cbpra.2018.09.001.
[189] Cohen, J. (1988), Statistical Power Analysis for the Behavioral Sciences, Lawrence Erlbaum Associates.
[71] Conte Keivabu, R. (2024), “Temperature and school absences: evidence from England”, Population and Environment, Vol. 46/1, https://doi.org/10.1007/s11111-024-00448-5.
[105] Corcoran, S. and C. Kelly (2022), “A meta‐ethnographic understanding of children and young people’s experiences of extended school non‐attendance”, Journal of Research in Special Educational Needs, Vol. 23/1, pp. 24-37, https://doi.org/10.1111/1471-3802.12577.
[121] Court of Auditors (2025), Le temps d’enseignement perdu par les élèves au collège [The teaching time lost by students in middle school], https://www.ccomptes.fr/fr/publications/le-temps-denseignement-perdu-par-les-eleves-au-college (accessed on 19 December 2025).
[82] Cristofoli, S. (2026), En 2024-2025, l’absentéisme touche en moyenne 7 % des élèves du second degré public [In 2024-2025, absenteeism affects on average 7% of students in public secondary schools], https://doi.org/10.48464/ni-26-11.
[29] Cristofoli, S. (2015), “L’absentéisme des élèves soumis à l’obligation scolaire”, Éducation & formations, Vol. 88-89/06, pp. 101-121, https://doi.org/10.48464/ef-88-89-06.
[126] Dahl, P. (2015), “Factors Associated With Truancy”, Journal of Adolescent Research, Vol. 31/1, pp. 119-138, https://doi.org/10.1177/0743558415587324.
[140] Dahlqvist, H. (2026), “Unexcused Repeated Absence in Grade Nine in Sweden: The Role of Victimization, Academic Factors, and Depressive Symptoms”, European Journal of Educational Research, Vol. 15/2, pp. 497-512, https://doi.org/10.12973/eu-jer.15.2.497.
[8] Damon, W. and R. Lerner (eds.) (1998), The ecology of developmental processes, John Wiley & Sons, Inc.
[114] Darmody, M., E. Smyth and S. McCoy (2008), “Acting up or opting out? Truancy in Irish secondary schools”, Educational Review, Vol. 60/4, pp. 359-373, https://doi.org/10.1080/00131910802393399.
[113] Darmody, M. and M. Thornton (2014), “What predicts disaffection in Irish primary schools?”, Educational Studies, Vol. 41/1-2, pp. 29-46, https://doi.org/10.1080/03055698.2014.955729.
[64] Dee, T. (2025), Recent Immigration Raids Increased Student Absences, https://doi.org/10.26300/a62e-h526.
[15] Department for Education (2025), Year-on-year change in pupil attendance, https://www.gov.uk/government/publications/year-on-year-change-in-pupil-attendance (accessed on 16 June 2025).
[122] Department for Education (2023), Pupil attendance in schools: Week 6 2023, https://explore-education-statistics.service.gov.uk/find-statistics/pupil-attendance-in-schools/2023-week-6 (accessed on 22 August 2025).
[118] Education Quality Agency (2019), Buenas prácticas para disminuir la inasistencia escolar [Good practices to reduce school absence], https://hdl.handle.net/20.500.12365/4535 (accessed on 14 November 2025).
[95] Education Scotland (2023), Improving Attendance: Understanding the Issues, https://education.gov.scot/resources/improving-attendance-in-scotland/improving-attendance-understanding-the-issues-report/ (accessed on 11 August 2025).
[23] Ehrlich, S. et al. (2014), Preschool Attendance in Chicago Public Schools, https://consortium.uchicago.edu/publications/preschool-attendance-chicago-public-schools-relationships-learning-outcomes-and-reaso-0 (accessed on 19 December 2025).
[199] Enderle, C. (2025), ““So That You Can Feel Well”: Perspectives of Students with Social, Emotional and Behavioral Difficulties on Factors that Support School Attendance”, School Mental Health, https://doi.org/10.1007/s12310-025-09764-z.
[104] Enderle, C. et al. (2025), “Their voices matter: student and professional perspectives on overcoming school attendance problems in the context of social, emotional, and behavioral difficulties”, Frontiers in Education, Vol. 10, https://doi.org/10.3389/feduc.2025.1627098.
[81] ERO (2025), Back to class: How are attitudes to attendance changing?, Education Review Office, https://evidence.ero.govt.nz/documents/back-to-class-how-are-attitudes-to-attendance-changing-research-report (accessed on 19 December 2025).
[99] ERO (2025), Back to class: How are attitudes to attendance changing? Technical appendix, https://evidence.ero.govt.nz/documents/back-to-class-how-are-attitudes-to-attendance-changing-technical-appendix (accessed on 19 December 2025).
[14] ERO (2024), Left behind: How do we get our chronically absent students back to school?, Education Review Office, https://evidence.ero.govt.nz/media/notdbxih/left-behind-how-do-we-get-our-chronically-absent-students-back-to-school.pdf (accessed on 11 August 2025).
[116] ERO (2023), Attendance: Getting Back to School, Education Review Office, https://evidence.ero.govt.nz/media/w4qji31h/attendance-getting-back-to-school.pdf (accessed on 12 August 2025).
[27] ERO (2022), Missing Out: Why Aren’t Our Children Going to School?, Education Review Office, https://evidence.ero.govt.nz/documents/missing-out-why-aren-t-our-children-going-to-school (accessed on 16 June 2025).
[128] Escario, J., J. Giménez-Nadal and A. Wilkinson (2022), “Predictors of adolescent truancy: The importance of cyberbullying, peer behavior, and parenting style”, Children and Youth Services Review, Vol. 143, p. 106692, https://doi.org/10.1016/j.childyouth.2022.106692.
[57] European Commission: Directorate-General for Education, Youth, Sport and Culture, Public Policy and Management Institute, Downes, Nairz-Wirth and Rusinaitė (2017), Structural indicators for inclusive systems in and around schools – Analytical report, Publications Office of the European Union, https://doi.org/10.2766/200506.
[109] Feldhaus, M. et al. (2025), “Schulabsentismus aus einer humanökologischen Perspektive – aktuelle Analysen mit PISA 2022, School absenteeism from a human-ecological perspective—current analyses with PISA 2022”, Zeitschrift für Erziehungswissenschaft, Vol. 28/3, pp. 781-804, https://doi.org/10.1007/s11618-025-01292-5.
[67] Fields, X. and C. Wotipka (2020), “Effect of LGBT anti-discrimination laws on school climate and outcomes for lesbian, gay, and bisexual high school students”, Journal of LGBT Youth, Vol. 19/3, pp. 307-329, https://doi.org/10.1080/19361653.2020.1821276.
[177] Finning, K. et al. (2021), “Mental health and school absenteeism in children with long‐term physical conditions: A secondary analysis of the British Child and Adolescent Mental Health Surveys 2004 and 2007”, Child: Care, Health and Development, Vol. 48/1, pp. 110-119, https://doi.org/10.1111/cch.12910.
[148] General Inspectorate of Education and Evaluation (n.d.), Memoria Curso 2022/2023 [Academic Year Report 2022/2023], https://www.educarex.es/pub/cont/com/0060/documentos/Memoria_Absentismo.pdf (accessed on 19 December 2025).
[89] Gillie, S. (2025), “‘It feels like we’re out of the rat race’: Family reflections on traumatic school experiences leading to home education”, British Journal of Special Education, https://doi.org/10.1111/1467-8578.70004.
[196] Gottfried, M. and K. Gee (2017), “Identifying the Determinants of Chronic Absenteeism: A Bioecological Systems Approach”, Teachers College Record: The Voice of Scholarship in Education, Vol. 119/7, pp. 1-34, https://doi.org/10.1177/016146811711900704.
[37] Gottfried, M. and E. Hutt (eds.) (2019), Reinforcing Student Attendance: Shifting Mind-Sets and Implementing Data-Driven Improvement Strategies During School Transitions, Harvard Education Press.
[26] Gottfried, M. and E. Hutt (eds.) (2019), Roll Call: Describing Chronically Absent Students, the Schools They Attend, and Implications for Accountability, Harvard Education Press.
[66] Gottfried, M. and E. Hutt (eds.) (2019), Schools as Sanctuaries? Examining the Relationship Between Immigration Enforcement and Absenteeism Rates for Immigrant-Origin Children, Harvard Education Press.
[31] Government of Chile (2025), CEM Interactive: Asistencia acumulada mensual [CEM Interactive: Monthly Cumulative Attendance], https://app.powerbi.com/view?r=eyJrIjoiYjJjZTY1YzEtZmQ3MS00OTAzLThjZTYtMjYwYTIxMmFmNmFhIiwidCI6IjJlNGNmZTUwLTA1ODAtNDE0MC05Mzg3LTRlY2RlMzlkZWY2MCIsImMiOjR9 (accessed on 13 November 2025).
[1] Gubbels, J., C. van der Put and M. Assink (2019), “Risk Factors for School Absenteeism and Dropout: A Meta-Analytic Review”, Journal of Youth and Adolescence, Vol. 48/9, pp. 1637-1667, https://doi.org/10.1007/s10964-019-01072-5.
[164] Health Behaviour in School-aged Children study (2023), Data browser (findings from the 2021/22 international HBSC survey), https://data-browser.hbsc.org (accessed on 15 April 2026).
[11] Heyne, D. (2025), The Drivers of Absence: A Background Note for the OECD Project on School Attendance Problems.
[166] Hirata, I. et al. (2026), “Relationship Between Sleep Irregularity and School Non-Attendance Among Japanese Elementary and Junior High School Students”, Children, Vol. 13/1, p. 80, https://doi.org/10.3390/children13010080.
[49] Horga, I. et al. (2024), Liceenii din România. Acces și participare [High school students in Romania. Access and participation], National Center for Educational Policies and Evaluation - Education Research Unit, https://www.ise.ro/liceenii-din-romania-acces-si-participare (accessed on 11 August 2025).
[16] Hotulainen, R. et al. (2024), Perusopetuksen poissaolojen ja niihin annetun tuen vaikutukset koulunkäyntiin: Tutkimushankkeen loppuraportti [The effects of absences from basic education and the support provided for them on school attendance: Final report of the research project], HEA, Finnish Centre for Educational Evaluation, Faculty of Education, University of Helsinki, http://hdl.handle.net/10138/578593 (accessed on 11 August 2025).
[43] Howe, A. and V. Richards (eds.) (2011), Adolescent development, Routledge, https://doi.org/10.4324/9780203818718.
[40] Howe, A. and V. Richards (eds.) (2011), Introduction: Exploring the great divide, Routledge, https://doi.org/10.4324/9780203818718.
[33] Hunter, J., A. Haywood and M. Chapman (2025), Every day matters: Improving school attendance in Australia, Grattan Institute, https://grattan.edu.au/wp-content/uploads/2025/12/Every-day-matters-policy-brief-Grattan-Institute-2025.pdf (accessed on 11 December 2025).
[151] Hysing, M. et al. (2017), “Parental work absenteeism is associated with increased symptom complaints and school absence in adolescent children”, BMC Public Health, Vol. 17/1, https://doi.org/10.1186/s12889-017-4368-7.
[155] Ilsvard, S. et al. (2024), U-start - En tværprofessionel taskforce mod skolefravær: Evaluering [U-start - An interprofessional taskforce against school absenteeism: Evaluation], https://viden.sl.dk/artikler/boern-og-unge/socialt-udsatte/evaluering-af-u-start-en-tvaerprofessionel-taskforce-mod-skolefravaer/ (accessed on 14 August 2025).
[133] Jacob Kirksey, J. and J. Elefante (2022), “Familiar Faces in High School: How Having the Same Peers from Year-to-Year Links to Student Absenteeism”, Journal of Education for Students Placed at Risk (JESPAR), Vol. 29/1, pp. 23-42, https://doi.org/10.1080/10824669.2022.2102982.
[162] Jay, M. et al. (2025), “Chronic health conditions and school absence, exclusions, and non-enrolment: a cohort study using the Education and Child Health Insights from Linked Data database”, Journal of Public Health, https://doi.org/10.1093/pubmed/fdaf064.
[179] Jerrim, J. (2025), Mind the Engagement Gap: A National Study of Pupil Engagement in England’s Schools, ImpactEd Group, https://www.impactedgroup.uk/resources/report-mind-the-engagement-gap (accessed on 18 August 2025).
[180] Jerrim, J. and N. Kaye (2025), The decline in pupils’ emotional engagement with school. How does England compare to other countries?, UCL Centre for Education Policy and Equalising Opportunities, https://econpapers.repec.org/RePEc:ucl:cepeow:25-03 (accessed on 18 August 2025).
[176] John, A. (2021), How school attendance data could play a crucial role in targeting mental health support for pupils, https://www.adruk.org/news-publications/news-blogs/how-school-attendance-data-could-play-a-crucial-role-in-targeting-mental-health-support-for-pupils-493/ (accessed on 14 August 2025).
[187] Jones, T. et al. (2024), “Linguistically Diverse Caregiver Perspectives on Barriers and Facilitators of School Attendance and Engagement During the COVID-19 Pandemic”, Contemporary School Psychology, Vol. 29/1, pp. 94-112, https://doi.org/10.1007/s40688-024-00493-9.
[157] Kaipainen, M. (2023), Schools will soon have to take a tougher approach to student absences – two mothers tell us why their children don’t want to go to school, https://yle.fi/a/74-20028813 (accessed on 22 August 2025).
[110] Karlberg, M. et al. (2020), “The Importance of School Pedagogical and Social Climate to Students’ Unauthorized Absenteeism – a Multilevel Study of 101 Swedish Schools”, Scandinavian Journal of Educational Research, Vol. 66/1, pp. 88-104, https://doi.org/10.1080/00313831.2020.1833244.
[12] Kearney, C. (2021), “Integrating Systemic and Analytic Approaches to School Attendance Problems: Synergistic Frameworks for Research and Policy Directions”, Child & Youth Care Forum, Vol. 50/4, pp. 701-742, https://doi.org/10.1007/s10566-020-09591-0.
[13] Kearney, C., J. Childs and S. Burke (2022), “Social Forces, Social Justice, and School Attendance Problems in Youth”, Contemporary School Psychology, https://doi.org/10.1007/s40688-022-00425-5.
[178] Keppens, G. and B. Spruyt (2019), Preventie van spijbelen en vroegtijdig schoolverlaten: Eindrapport [Preventing truancy and early school leaving: Final report], https://researchportal.vub.be/en/publications/preventie-van-spijbelen-en-vroegtijdig-schoolverlaten-eindrapport (accessed on 18 August 2025).
[60] Keppens, G. and B. Spruyt (2018), “Truancy in Europe: Does the type of educational system matter?”, European Journal of Education, Vol. 53/3, pp. 414-426, https://doi.org/10.1111/ejed.12282.
[127] Kirksey, J. et al. (2024), Absent Peers, Present Challenges: The Differential Impact of In-Person and Virtual Classmate Absences on Future Attendance, Center for Innovative Research in Change, Leadership, and Education, https://ttu-ir.tdl.org/server/api/core/bitstreams/87f6f12d-a77a-4b78-acb2-0f44e58384fb/content (accessed on 13 August 2025).
[165] Kjeøy, I. and R. Lysvik (2023), Kunnskapsstatus om bekymringsfullt fravær i skolen [Worrying school absence levels: A Scandinavian literature review], https://www.fafo.no/zoo-publikasjoner/kunnskapsstatus-om-bekymringsfullt-fravaer-i-skolen (accessed on 21 August 2025).
[143] Klein, M., E. Sosu and S. Dare (2020), “Mapping inequalities in school attendance: The relationship between dimensions of socioeconomic status and forms of school absence”, Children and Youth Services Review, Vol. 118, p. 105432, https://doi.org/10.1016/j.childyouth.2020.105432.
[54] Knight, D. and M. Olofson (2026), How Large are District Effects on Student Attendance? Implications for School Funding Based on Average Daily Attendance, Annenberg Institute at Brown University, https://doi.org/10.26300/55r5-q644.
[20] Kristensen, N., V. Jensen and K. Krassel (2020), Panelanalyse af bekymrende skolefravær [Panel analysis of worrying school absences], VIVE, https://www.vive.dk/da/udgivelser/panelanalyse-af-bekymrende-skolefravaer-dx3ynqvb/ (accessed on 7 May 2026).
[88] Kruithof, E. and G. Keppens (2024), “Who is allowed to stay at home during school hours? An analysis of school withdrawal”, Educational Studies, pp. 1-20, https://doi.org/10.1080/03055698.2024.2369859.
[112] Lamb, B. et al. (2022), “The power of connectedness: Associations between caring non-parental adult relationships, school attendance, and discipline among foster-involved youth”, Children and Youth Services Review, Vol. 142, p. 106633, https://doi.org/10.1016/j.childyouth.2022.106633.
[131] Landell, M. (ed.) (2021), Exploring the voices of young people in school absenteeism: what schools need to know, Jerringfonden.
[70] Lassa, J., M. Petal and A. Surjan (2022), “Understanding the impacts of floods on learning quality, school facilities, and educational recovery in Indonesia”, Disasters, Vol. 47/2, pp. 412-436, https://doi.org/10.1111/disa.12543.
[172] Lawrence, D. et al. (2019), “Impact of mental disorders on attendance at school”, Australian Journal of Education, Vol. 63/1, pp. 5-21, https://doi.org/10.1177/0004944118823576.
[87] Learning Heroes and TNTP (2023), Investigating the relationship between pre-pandemic family engagement and student and school outcomes, https://edredesign.org/resources/investigating-relationship-between-pre-pandemic-family-engagement-and-student-and-school (accessed on 12 August 2025).
[50] Leduc, K. et al. (2022), “School Refusal in Youth: A Systematic Review of Ecological Factors”, Child Psychiatry & Human Development, Vol. 55/4, pp. 1044-1062, https://doi.org/10.1007/s10578-022-01469-7.
[154] Lee, K. et al. (2023), “Using data-driven analytics and ecological systems theory to identify risk and protective factors for school absenteeism among secondary students”, Journal of School Psychology, Vol. 98, pp. 148-180, https://doi.org/10.1016/j.jsp.2023.03.002.
[170] Lereya, T. and J. Deighton (2019), Learning from HeadStart: the relationship between mental health and school attainment, attendance and exclusions in young people aged 11 to 14, https://www.tnlcommunityfund.org.uk/media/headstart_evidence_briefing_3.pdf?mtime=20190226100632&focal=none (accessed on 14 August 2025).
[132] Määttä, S. et al. (2020), Kouluakäymättömyys Suomessa [School absenteeism in Finland], Finnish National Agency for Education, https://www.oph.fi/fi/tilastot-ja-julkaisut/julkaisut/kouluakaymattomyys-suomessa (accessed on 14 August 2025).
[185] Macklem, G. (2015), Boredom in the Classroom, Springer International Publishing, Cham, https://doi.org/10.1007/978-3-319-13120-7.
[137] Markkanen, E. et al. (2022), Sitouttavan kouluyhteisötyön arviointi – arvioinnin väliraportti [Evaluation of Engaging School Community Work - interim evaluation report], The Finnish Education Evaluation Centre (FINEEC), https://www.karvi.fi/fi/julkaisut/sitouttavan-kouluyhteisotyon-arviointi-arvioinnin-valiraportti (accessed on 13 August 2025).
[123] Markus-Rodden, M. et al. (2025), “The relationship of social determinants of health and school nurse presence on chronic absenteeism”, Nursing Outlook, Vol. 73/5, p. 102459, https://doi.org/10.1016/j.outlook.2025.102459.
[183] Martin, F. et al. (2025), “Digital distractions in education: a systematic review of research on causes, consequences and prevention strategies”, Educational technology research and development, Vol. 73/6, pp. 3423-3451, https://doi.org/10.1007/s11423-025-10550-6.
[124] Martin, R. et al. (2020), “School Refusal or Truancy? A Qualitative Study of Misconceptions Among School Personnel About Absenteeism of Children From Immigrant Families”, Frontiers in Psychiatry, Vol. 11, https://doi.org/10.3389/fpsyt.2020.00202.
[174] Matos, M. et al. (2024), Relatório Saúde Psicológica e Bem-estar [Psychological Health and Well-being Report], https://info.dgeec.medu.pt/saude-psicologica-e-bem-estar-2024/ (accessed on 14 August 2025).
[160] Mavhu, W. (ed.) (2021), “Experiences of menstruation in high income countries: A systematic review, qualitative evidence synthesis and comparison to low- and middle-income countries”, PLOS ONE, Vol. 16/7, p. e0255001, https://doi.org/10.1371/journal.pone.0255001.
[115] McCoy, S. et al. (2007), Attendance and Students’ School Experience, ESRI and the National Educational Welfare Board.
[171] McDonald, B., K. Lester and D. Michelson (2022), “‘She didn’t know how to go back’: School attendance problems in the context of the COVID‐19 pandemic—A multiple stakeholder qualitative study with parents and professionals”, British Journal of Educational Psychology, Vol. 93/1, pp. 386-401, https://doi.org/10.1111/bjep.12562.
[197] Melvin, G. et al. (2019), “The Kids and Teens at School (KiTeS) Framework: An Inclusive Bioecological Systems Approach to Understanding School Absenteeism and School Attendance Problems”, Frontiers in Education, Vol. 4, https://doi.org/10.3389/feduc.2019.00061.
[28] Melvin, G. et al. (2025), Barriers to school attendance and reasons for student absence: Rapid literature review, Australian Education Research Organisation, https://www.edresearch.edu.au/research/research-reports/barriers-school-attendance-and-reasons-student-absence (accessed on 8 December 2025).
[53] Melvin, G. et al. (2025), “Interventions to promote school attendance and address student absence: Rapid literature review”, Australian Education Research Organisation, https://eric.ed.gov/?id=ED673994 (accessed on 4 December 2025).
[145] Ministry of Education and Centre for Studies (2025), Reactivación del Sistema Educativo Chileno a 3 años del retorno a la presencialidad post COVID-19: Asistencia, inasistencia grave/crítica, desvinculación y resultados de aprendizaje SIMCE, [Re-activation of the Chilean Education System 3 years after the return to in-person learning post-COVID-19: Attendance, serious/critical absenteeism, disengagement and SIMCE learning outcomes], https://hdl.handle.net/20.500.12365/21705 (accessed on 14 November 2025).
[32] Ministry of Education, Culture, Sports, Science and Technology (2026), 令和６年度 児童生徒の問題行動・不登校等生徒指導上の諸課題に関する調査結果について [Results of the 2024 Survey on Problem Behaviour, School Absenteeism, and Other Issues Related to Student Guidance], https://www.mext.go.jp/content/20260116-mxt_jidou02-100002753_1_3.pdf (accessed on 4 May 2026).
[35] Ministry of Science, Education and Youth (n.d.), Izostanci po razredima [Absences by grade], https://app.powerbi.com/view?r=eyJrIjoiM2Q1NjVmZDEtMGUyMy00MDBiLTkzYWItYjBhMTA3MDFlOWUxIiwidCI6IjJjMTFjYmNjLWI3NjEtNDVkYi1hOWY1LTRhYzc3ZTk0ZTFkNCIsImMiOjh9 (accessed on 17 February 2026).
[97] Ministry of Social Affairs and Housing (2019), Þörf á opinberum viðmiðum um skólasókn [Need for official standards for school attendance], https://www.stjornarradid.is/verkefni/felags-og-fjolskyldumal/velferdarvaktin/frettir/stok-frett-velferdarvaktar/2019/03/12/Thorf-a-opinberum-vidmidum-um-skolasokn/ (accessed on 12 August 2025).
[96] Moore, E. and M. Walker (2025), Voices from the classroom: Understanding how secondary schools support pupils returning from absence, NFER, https://www.nfer.ac.uk/publications/voices-from-the-classroom-understanding-how-secondary-schools-support-pupils-returning-from-absence/ (accessed on 8 December 2025).
[168] Mosiño, A., K. Villagómez-Estrada and A. Prieto-Patrón (2017), Asociación entre el desempeño escolar y la anemia en adolescentes en México [Association between school performance and anemia in adolescents in Mexico], https://mpra.ub.uni-muenchen.de/79385/1/MPRA_paper_79385.pdf (accessed on 16 April 2026).
[18] Moulin, L. (ed.) (2024), “Trajectories of school absences across compulsory schooling and their impact on children’s academic achievement: An analysis based on linked longitudinal survey and school administrative data”, PLOS ONE, Vol. 19/8, p. e0306716, https://doi.org/10.1371/journal.pone.0306716.
[175] National Agency for Education and Quality (2024), Langvarigt fravær, sygeundervisning, hjemmeundervisning og børn uden for undervisningstilbud [Long-term absence, sick leave, homeschooling and children outside of educational opportunities], https://www.uvm.dk/aktuelt/nyheder/uvm/2024/dec/241219-resultater-af-ny-undersoegelse-saadan-haandterer-kommuner-langvarigt-fravaer--tilsyn-med (accessed on 14 August 2025).
[147] National Center for Education Policy and Evaluation (2023), Neparticiparea elevilor de liceu la examenul de bacalaureat – sesiunea iunie 2023 [Non-participation of high school students in the baccalaureate exam – June 2023 session], National Center for Education Policy and Evaluation, https://www.ise.ro/neparticipare-bacalaureat-sesiunea-iunie-2023 (accessed on 14 August 2025).
[192] National Lottery Community Fund (n.d.), Improving the wellbeing and mental health of young people, https://www.tnlcommunityfund.org.uk/funding/strategic-investments/headstart (accessed on 14 August 2025).
[193] Navarro, J. et al. (2022), “Bored of the rings: Methodological and analytic approaches to operationalizing Bronfenbrenner’s PPCT model in research practice”, Journal of Family Theory & Review, Vol. 14/2, pp. 233-253, https://doi.org/10.1111/jftr.12459.
[195] Navarro, J. and J. Tudge (2022), “Technologizing Bronfenbrenner: Neo-ecological Theory”, Current Psychology, Vol. 42/22, pp. 19338-19354, https://doi.org/10.1007/s12144-022-02738-3.
[190] Niu, L. (2018), “A review of the application of logistic regression in educational research: common issues, implications, and suggestions”, Educational Review, Vol. 72/1, pp. 41-67, https://doi.org/10.1080/00131911.2018.1483892.
[184] Norwegian Directorate for Education and Training (2024), More students experience bullying, https://www.udir.no/tall-og-forskning/statistikk/analyser/2024/flere-elever-opplever-mobbing/.
[134] OECD (2026), “Bullying in education: Prevalence, impact and responses across countries”, OECD Education Working Papers, No. 341, OECD Publishing, Paris, https://doi.org/10.1787/d9f8bd9f-en.
[55] OECD (2026), OECD Data Explorer: Expenditure on educational institutions per full-time equivalent student, https://data-explorer.oecd.org/s/2np (accessed on 9 March 2026).
[181] OECD (2025), “Finite time to learn and play: Whole student development and students’ digital leisure outside of school”, OECD Education Policy Perspectives, No. 130, OECD Publishing, Paris, https://doi.org/10.1787/edbaa4bb-en.
[36] OECD (2025), OECD Policy Survey on School Attendance Problems.
[191] OECD (2024), PISA 2022 Technical Report, PISA, OECD Publishing, Paris, https://doi.org/10.1787/01820d6d-en.
[182] OECD (2024), “Students, digital devices and success”, OECD Education Policy Perspectives, No. 102, OECD Publishing, Paris, https://doi.org/10.1787/9e4c0624-en.
[61] OECD (2023), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, PISA, OECD Publishing, Paris, https://doi.org/10.1787/53f23881-en.
[3] OECD (2023), PISA 2022 Results (Volume II): Learning During – and From – Disruption, PISA, OECD Publishing, Paris, https://doi.org/10.1787/a97db61c-en.
[21] OECD (2022), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
[2] OECD (2021), Computer-based Student Questionnaire for PISA 2022, https://www.oecd.org/content/dam/oecd/en/data/datasets/pisa/pisa-2022-datasets/questionnaires/COMPUTER-BASED%20STUDENT%20questionnaire%20PISA%202022.pdf (accessed on 19 August 2025).
[45] OECD (2018), Responsive School Systems: Connecting Facilities, Sectors and Programmes for Student Success, OECD Reviews of School Resources, OECD Publishing, Paris, https://doi.org/10.1787/9789264306707-en.
[106] Ofsted and Care Quality Commission (2025), Beyond the classroom: the experiences of children with SEND who are not in school, https://www.gov.uk/government/publications/the-experiences-of-children-with-send-who-are-not-in-school-a-thematic-review/beyond-the-classroom-the-experiences-of-children-with-send-who-are-not-in-school (accessed on 19 December 2025).
[52] Öhman, A. (2016), Kartläggning av skolfrånvaro och arbetet med den i kommuner och grundskolan [Mapping school absenteeism and the work with it in municipalities and primary schools: data-report].
[98] Öhman, A. (2016), Skolans tomma stolar: Om frånvaro i grundskolan och hur kommuner och skolor arbetar med frågan [Schools’ empty chairs: On absenteeism in compulsory school and how municipalities and schools are addressing the issue], https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fprestationsprinsen.se%2Fwp-content%2Fuploads%2F2019%2F08%2FPrestationsprinsen_skolans_tomma_stolar_om_fr%25C3%25A5nvaro_i_grundskolan.pdf&data=05%7C02%7Csamo.varsik%40oecd.org%7C66553ad0e07 (accessed on 14 November 2025).
[74] Pak, A. (ed.) (2025), “Nearby arrests and violent crime as predictors of student absenteeism”, PLOS One, Vol. 20/7, p. e0323565, https://doi.org/10.1371/journal.pone.0323565.
[91] Parentkind (2024), The National Parent Survey for Wales 2024, https://www.parentkind.org.uk/assets/resources/NPS-24-Wales-Hyperlinked_compressed.pdf (accessed on 12 August 2025).
[42] Patton, G. and R. Viner (2007), “Pubertal transitions in health”, The Lancet, Vol. 369/9567, pp. 1130-1139, https://doi.org/10.1016/s0140-6736(07)60366-3.
[158] Pijl, E. et al. (2021), “The prevalence of sickness absence among primary school pupils – reason to be worried?”, BMC Public Health, Vol. 21/1, https://doi.org/10.1186/s12889-021-10193-1.
[83] Plăeșu, A. et al. (2024), Liceenii din România: Calitatea vieții liceenilor din România [High school students in Romania: Quality of life of high school students in Romania], National Center for Educational Policies and Evaluation, https://www.ise.ro/liceenii-din-romania-calitatea-vietii-liceenilor-din-romania (accessed on 12 August 2025).
[46] Potia, A. et al. (2025), “Transition From Primary to Secondary School: Igniting Attendance and Engagement Among Aboriginal and Torres Strait Islander Students Through National Policy Reform”, Australian Journal of Social Issues, https://doi.org/10.1002/ajs4.70041.
[156] Rambøll (2018), Undersøgelse af hjemmeundervisning, fravær og børn uden for undervisningstilbud [Study of homeschooling, absenteeism and children outside educational opportunities], https://www.uvm.dk/publikationer/folkeskolen/2018-undersoegelse-af-hjemmeundervisning--fravaer-og-boern-uden-for-undervisningstilbud (accessed on 14 August 2025).
[90] RED C (2025), Unpublished market research poll provided to the OECD by Tusla Education Support Service (TESS).
[130] Richards, A. and K. Clark-Howard (2023), “What are the reasons for absenteeism in secondary education in New Zealand/Aotearoa? A systematic review of the qualitative literature using student voice research”, Kairaranga, Vol. 24/2, pp. 73-94, https://doi.org/10.54322/3w9hk758.
[25] Romero, M. and Y. Lee (2007), A National Portrait of Chronic Absenteeism in the Early Grades, https://academiccommons.columbia.edu/doi/10.7916/D89C7650 (accessed on 19 December 2025).
[103] Rowlands, M. (2022), Attendance review – implications of the COVID-19 pandemic for school attendance, Welsh Government, https://www.gov.wales/sites/default/files/publications/2023-12/attendance-review-implications-of-the-covid-19-pandemic-for-school-attendance.pdf (accessed on 12 August 2025).
[129] Saelzer, C. and A. Lenski (2016), “Mind the Gap: How Students Differentially Perceive Their School’s Attendance Policies in Germany”, Journal of Education for Students Placed at Risk (JESPAR), Vol. 21/4, pp. 246-260, https://doi.org/10.1080/10824669.2016.1206824.
[68] Seelman, K. and M. Walker (2018), “Do Anti-Bullying Laws Reduce In-School Victimization, Fear-based Absenteeism, and Suicidality for Lesbian, Gay, Bisexual, and Questioning Youth?”, Journal of Youth and Adolescence, Vol. 47/11, pp. 2301-2319, https://doi.org/10.1007/s10964-018-0904-8.
[47] Sergejeff, J. (ed.) (2023), Yhteisellä koulutiellä [On the common school road], Ministry of Education and Culture, http://urn.fi/URN:ISBN:978-952-263-924-0 (accessed on 12 August 2025).
[44] Short, M. and S. Rosenthal (2008), “Psychosocial Development and Puberty”, Annals of the New York Academy of Sciences, Vol. 1135/1, pp. 36-42, https://doi.org/10.1196/annals.1429.011.
[17] Simon, O. et al. (2020), “Elementary absenteeism over time: A latent class growth analysis predicting fifth and eighth grade outcomes”, Learning and Individual Differences, Vol. 78, p. 101822, https://doi.org/10.1016/j.lindif.2020.101822.
[188] Singer, J. et al. (2021), “Advancing an Ecological Approach to Chronic Absenteeism: Evidence from Detroit”, Teachers College Record: The Voice of Scholarship in Education, Vol. 123/4, pp. 1-36, https://doi.org/10.1177/016146812112300406.
[139] Sinmaz, E. (2025), Majority of girls and young women in UK alter behaviour to feel safe, study finds, https://www.theguardian.com/education/2025/sep/23/majority-of-girls-and-young-women-in-uk-alter-behaviour-to-feel-safe-study-finds (accessed on 5 May 2026).
[75] Smart, M. et al. (2020), “Is a School’s Neighborhood Physical Disorder Related to Its Academic Outcomes?”, Child & Youth Care Forum, Vol. 50/2, pp. 247-259, https://doi.org/10.1007/s10566-020-09572-3.
[117] Sorrenti, L. et al. (2025), “The Mediation Role of School Alienation Between Perceptions of the School Atmosphere and School Refusal in Italian Students”, Education Sciences, Vol. 15/7, p. 786, https://doi.org/10.3390/educsci15070786.
[141] Sosu, E. et al. (2021), “Socioeconomic status and school absenteeism: A systematic review and narrative synthesis”, Review of Education, Vol. 9/3, https://doi.org/10.1002/rev3.3291.
[34] State Service of Education Quality (2020), Report on general education and vocational education institutions for long-term unjustified absences of students delays and actions to prevent them in the 2019/2020 academic year.
[84] Stein, M. and J. Grigg (2019), “Missing Bus, Missing School: Establishing the Relationship Between Public Transit Use and Student Absenteeism”, American Educational Research Journal, Vol. 56/5, pp. 1834-1860, https://doi.org/10.3102/0002831219833917.
[56] Stronati, C. (2023), “The design of upper secondary education across OECD countries: Managing choice, coherence and specialisation”, OECD Education Working Papers, No. 288, OECD Publishing, Paris, https://doi.org/10.1787/158101f0-en.
[30] Swedish National Agency for Education (2024), Attityder till skolan 2024 [Attitudes towards school in 2024], Swedish National Agency for Education, https://www.skolverket.se/statistik-och-utvarderingar/uppfoljning-och-utvardering/attityder-till-skolan (accessed on 11 August 2025).
[101] Swedish National Agency for Education (2023), Skolverkets bedömning av läget i skolväsendet 2023 [The National Agency for Education’s assessment of the situation in the school system in 2023], https://www.skolverket.se/sok-publikationer/publikationsserier/rapporter/2023/skolverkets-bedomning-av-laget-i-skolvasendet-2023 (accessed on 12 August 2025).
[51] Tanya Lereya, S. et al. (2022), “How does the association between special education need and absence vary overtime and across special education need types?”, European Journal of Special Needs Education, Vol. 38/2, pp. 245-259, https://doi.org/10.1080/08856257.2022.2059631.
[80] TESS (2025), School Attendance Data Primary and Post-Primary Schools And Student Absence Reports Primary and Post-Primary Schools: 2023-2024, https://www.tusla.ie/uploads/content/Analysis_of_School_Attendance_Data_2023-24.pdf (accessed on 19 December 2025).
[79] TESS (2024), School Attendance Data Primary and Post-Primary Schools And Student Absence Reports Primary and Post-Primary Schools: 2022-2023, https://www.tusla.ie/uploads/content/AAR_SAR_2022_2023_School_Year.pdf (accessed on 21 August 2025).
[78] TESS (2023), School Attendance Data Primary and Post-Primary Schools And Student Absence Reports Primary and Post-Primary Schools: 2019-2022, https://www.tusla.ie/uploads/content/AAR_SAR_2019_22.pdf (accessed on 21 August 2025).
[77] TESS (2022), School Attendance Data Primary and Post-Primary Schools And Student Absence Reports Primary and Post-Primary Schools: 2018/19, https://www.tusla.ie/uploads/content/TESS_School_Attendance_Absence_Returns_2018.2019.pdf (accessed on 21 August 2025).
[161] Thomas, C. and G. Melendez‐Torres (2024), “The experiences of menstruation in schools in high income countries: A systematic review and line‐of‐argument synthesis”, Psychology in the Schools, Vol. 61/7, pp. 2820-2844, https://doi.org/10.1002/pits.23192.
[135] Thornton, M., M. Darmody and S. McCoy (2013), “Persistent absenteeism among Irish primary school pupils”, Educational Review, Vol. 65/4, pp. 488-501, https://doi.org/10.1080/00131911.2013.768599.
[9] Tong, P. and I. An (2024), “Review of studies applying Bronfenbrenner’s bioecological theory in international and intercultural education research”, Frontiers in Psychology, Vol. 14, https://doi.org/10.3389/fpsyg.2023.1233925.
[62] Tran, L. and S. Gershenson (2021), “Experimental Estimates of the Student Attendance Production Function”, Educational Evaluation and Policy Analysis, Vol. 43/2, pp. 183-199, https://doi.org/10.3102/0162373720984463.
[194] Tudge, J. et al. (2016), “Still Misused After All These Years? A Reevaluation of the Uses of Bronfenbrenner’s Bioecological Theory of Human Development”, Journal of Family Theory & Review, Vol. 8/4, pp. 427-445, https://doi.org/10.1111/jftr.12165.
[69] UNICEF (2025), Learning interrupted: Global snapshot of climate-related school disruptions in 2024, https://www.unicef.org/reports/learning-interrupted-global-snapshot-2024 (accessed on 30 April 2026).
[149] UNICEF and Argentine Society of Pediatrics (n.d.), 2° Informe Especial del Observatorio de la Infancia y Adolescencia SAP-UNICEF [2nd Special Report of the SAP-UNICEF Observatory on Childhood and Adolescence], https://sub.sap.org.ar/storage/app/media/pdf/files_informe-especial-del-observatorio-de-la-infancia-y-adolescencia-final-11-22sap-unicef_1668805381.pdf (accessed on 5 May 2026).
[167] Vargas cruz, S. et al. (2016), “Relación entre el estado nutricional y el ausentismo escolar en estudiantes de escuelas rurales”, Revista de Salud Pública, Vol. 17/6, pp. 861-873, https://doi.org/10.15446/rsap.v17n6.48709.
[38] Varsik, S. (2025), “Transitions through education and into the labour market: Exploring educational outcomes and equity gaps using international surveys”, OECD Education Working Papers, No. 324, OECD Publishing, Paris, https://doi.org/10.1787/18be3c52-en.
[100] Vennerød-Diesen, F. et al. (2024), «Et bevis på at man trenger skolen, egentlig», Delrapport 2 [“Proof that you really need school”, Interim report 2], https://www.udir.no/tall-og-forskning/finn-forskning/rapporter/2024/koronapandemiens-langsiktige-konsekvenser-for-elever-pa-ungdomstrinnet-delrapport-2/ (accessed on 14 November 2024).
[63] von Hippel, P. (2024), “The effect of smaller classes on infection-related school absence: evidence from the Project STAR randomized controlled trial”, BMC Public Health, Vol. 24/1, https://doi.org/10.1186/s12889-023-17503-9.
[138] Williford, A. et al. (2020), “Associations between different forms of peer victimization and school absences: The moderating role of teacher attachment and perceived school safety”, Psychology in the Schools, Vol. 58/1, pp. 185-202, https://doi.org/10.1002/pits.22438.
[163] Yon, D. (ed.) (2023), “School absenteeism in children with special health care needs. Results from the prospective cohort study ikidS”, PLOS ONE, Vol. 18/6, p. e0287408, https://doi.org/10.1371/journal.pone.0287408.
[152] Zheng, Z. et al. (2022), “Associations of Parental Cancer With School Absenteeism, Medical Care Unaffordability, Health Care Use, and Mental Health Among Children”, JAMA Pediatrics, Vol. 176/6, p. 593, https://doi.org/10.1001/jamapediatrics.2022.0494.
As mentioned in Box 2.1, the long-term absence rates used in this chapter may be biased due to sampled students who were absent during the assessment, and due to students misinterpreting the questions as referring to school closures.
Panel A in Annex Figure 2.A.1 displays the association between long-term absence (percentage of students who reported that they had missed school for more than three consecutive months at the secondary level) and the share of students who had been sampled but did not participate in PISA. Based on the R2, the association is small (Cohen, 1988[189]). Pearson correlation coefficient equals 0.04, also considered small (ibid.). However, if long-term absent students were to miss PISA across countries systematically, we would expect to see a negative correlation coefficient: the absence rate would increase due to the non-participation of long-term absent students, but the long-term absence rate would decrease. The evidence presented here does not align with this expectation.
Panel B in Annex Figure 2.A.1displays the association between long-term absence at the secondary level and the percentage of students who reported that the school building had been closed for three or more months because of COVID-19 in the three years before PISA 2022. Based on the R2 = 0.1 the association is small to medium (Cohen, 1988[189]). Pearson correlation coefficient equals 0.32, which is considered medium (ibid.). These associations are estimated at the system level, but the correlation between long‑term absence and school closures can also be modelled at the student level (Annex Table 2.B.1). Broadly speaking, the model coefficients maintain their values and the conclusions from the models do not change regardless of whether the school closure variable is included. Moreover, the variable itself is not significant.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025); and OECD (2023[61]), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, Table I.A2.6, https://doi.org/10.1787/53f23881-en.
Logistic models are commonly applied to estimate relationships between a binary dependent variable and independent variable(s). In this chapter, the binary dependent variable is whether students reported missing school for more than three consecutive months at the secondary educational level. Independent variables include a range of student-level characteristics and school-level factors (Annex Table 2.B.1).
Parameters that are estimated using a logistic regression are not easily interpretable (log-odds). Therefore, it is suggested that they are transformed into coefficients with a more straightforward interpretation. Various approaches can be applied (most commonly odds ratios). In line with recommendations in the literature, the parameters are transformed into marginal effects, specifically, average marginal effects (Niu, 2018[190]). In the models presented in this chapter, average marginal effects report the predicted change in the probability of long-term absence at the secondary educational level for a one-unit change in an independent variable, while holding other variables constant.
PISA applies a two-stage sampling (first selects schools, then students within schools). Thus, students attending the same school cannot be considered as independent observations (violating an assumption of logistic regression). This is because students within a school usually share more common characteristics than students from different schools (e.g. access to the same teachers and resources). To consider this, standard errors are computed with a balanced repeated replication methodology, in line with recommendations by the OECD (2024[191]). Observations are also weighted by the final student weights.
Coefficients display average marginal effects in percentage points on the probability of long-term absence in secondary education
Note: SE = standard error. Coefficients in bold are statistically significant at 5% level. See the main text for a description of the variables used. The models display average marginal effects, which report the predicted change in the probability of long-term absence at the secondary educational level for a one-unit change in an independent variable, while holding other variables constant. The models account for country fixed effects. Model 1 includes the following countries: Argentina, Australia, Belgium, Brazil, Bulgaria, Chile, Colombia, Croatia, Czechia, Denmark, Estonia, Finland, France, Greece, Hungary, Iceland, Korea, Latvia, Lithuania, Malta, Mexico, the Netherlands, New Zealand, Peru, Poland, Portugal, Romania, the Slovak Republic, Slovenia, Switzerland, Thailand, Türkiye, the United Kingdom and the United States. Model 2 excludes Denmark and accounts for school closures because of the COVID-19 pandemic. Model 3 excludes non-EU countries from Model 2.
Source: OECD (2022[21]), PISA 2022 (dataset), https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 19 May 2025).
← 1. The term “drivers” is used throughout the chapter as an umbrella label for factors empirically associated with SAP and, where supported by longitudinal evidence, plausibly contributing to their onset, persistence or escalation. Related terms such as “risk factors” or “influencing factors” are treated as largely overlapping but are only retained when reflecting the terminology used in the original sources.
← 2. Caution is required when interpreting estimates using PISA 2022 because one or more PISA sampling standards were not met in the following countries: Australia, Canada, Denmark, Ireland, Latvia, the Netherlands, New Zealand, the United Kingdom and the United States. See Reader’s Guide, Annexes A2 and A4 in OECD (2023[61]).
← 3. Additional conceptual grounding for the explanation of the model comes from broader developmental work by Tudge et al. (2016[194]), Navarro et al. (2022[193]), and Navarro and Tudge (2022[195]), alongside sources that apply or extend Bronfenbrenner’s ideas in the school attendance field, including Gottfried and Gee (2017[196]), Melvin et al. (2019[197]), Childs and Scanlon (2022[200]), Enderle et al. (2025[104]), Bond et al. (2024[198]), Enderle (2025[199]) and Heyne (2025[11]).
← 4. Quotations from students, teachers, parents and other education stakeholders are used throughout this chapter as illustrative examples of lived experiences related to the topics discussed. They should not be interpreted as representative of the country or jurisdiction in which they were collected, nor as an assessment of that education system as a whole.
← 5. Whether the coefficient of determination is small, medium or large follows Cohen’s proposed magnitudes for R2, i.e. 0.02, 0.13 and 0.26 for very weak, weak and moderate fits, respectively (Cohen, 1988[189]).
← 6. Pain during menstruation.
← 7. HeadStart aims to explore and test new ways to improve the mental health and well-being of young people aged 10 to 16 and prevent serious mental health issues from developing (National Lottery Community Fund, n.d.[192]).