The most-documented, least-addressed problem in UK education.
For thirty years, research has shown that teacher absence depresses pupil attainment, behaviour and workforce morale. For thirty years, schools have answered it with a clipboard, an agency phone call, and a holding lesson.
Two teaching days lost, every school day.
In 2023/24, 65.7% of teachers in England's state-funded schools took at least one period of sickness absence, averaging 8.3 working days each — around ten points above pre-pandemic levels. Across 468,300 FTE teachers, that is on the order of 2.6 million teaching days lost to sickness in a single year, before maternity, bereavement, jury service, training or unfilled vacancies are added. [1]
For the typical 70-teacher secondary school, this works out at roughly 380 teaching days lost a year — just over two teaching days for every school day. That is the number of cover events a school must orchestrate, document and quality-control across a year. [1]
Six studies, one finding.
Teacher absence is not a neutral event. Its effect on pupils is measurable, replicable, and disproportionately borne by those least able to absorb it.
| Study | Setting | Key finding |
|---|---|---|
| Borgen, Markussen & Raaum (2023) [46] | Norway, lower-secondary, full population | Exam grades fall by 2.3% — effect concentrated among disadvantaged pupils |
| Miller, Murnane & Willett (2008) [47] | US urban district | Ten additional absence days reduce maths attainment by 1.7–3.3% |
| Clotfelter, Ladd & Vigdor (2009) [48] | North Carolina | Significant negative effect; mediated by substitute quality |
| Herrmann & Rockoff (2012) [49] | New York City | Effect larger when the substitute is less effective |
| Boyd-Swan & Herbst (2017) [50] | US national CALDER data | Confirms Miller et al.; effect persists across subjects |
| Brookings Institution (2020) [138] | US synthesis | Effect larger in schools serving low-income or minority pupils |
Most leaders report occasions when they have had to accept non-specialists or supply workers without QTS to cover absences, particularly at short notice.
Department for Education, Use of Supply Teachers in Schools (2024) [5]The harm is not evenly shared.
The most important finding in this literature is distributional. Borgen et al. are explicit that the effect they identify is driven by pupils from low socioeconomic backgrounds; US syntheses reach the same conclusion. [46, 138] Families with the cultural capital to compensate — through tutoring, supervised homework, private provision — partly insulate their children from the cover-quality gap. Families without it cannot.
The documented Pupil Premium gap is this pattern in operation. A cover lesson that does not reach its learning objective is not a neutral event for a disadvantaged pupil; it is a deficit that compounds across the school year. [52]
The second-order costs.
Behaviour clusters in cover lessons.
Ofsted estimates pupils lose up to 38 days of teaching a year to low-level disruption; a quarter of teachers lose more than 10 minutes per 30-minute lesson. [9, 35] The EEF is unambiguous: consistency of behaviour-policy application by all adults is the single biggest cover lever — and consistency is impossible if the cover adult has never been told what the policy says. [33]
Cook et al. (2018) demonstrated that positive greetings at the door — possible only when the adult knows pupils' names — reduced disruptive behaviour by roughly 9 percentage points and improved academic engagement by roughly 20 percentage points. [151]
SEND pupils carry the disproportionate burden.
Around 18% of secondary pupils have an identified SEND status. [2] EHCP provision is a legal entitlement that does not pause when a teacher is absent. A covering adult who does not know which pupils have an EHCP, what it specifies, and how the regular teacher implements it is — through no fault of their own — failing to deliver legally required provision. [21, 69]
The statutory duty is clear. The operational mechanism to discharge it — giving the cover adult the relevant information before the lesson — has, until now, had no dedicated tool.
The £1.4 billion is only the visible layer.
Direct agency spend is the top of a stack of partly-overlapping costs. The layers beneath are systematically larger and systematically less visible.
| Cost category | Mechanism | Annual UK sector estimate |
|---|---|---|
| Direct agency spend | Mark-ups on the supply daily rate | £1.4 billion [111, 121] |
| Internal cover labour | Teacher / leader / HLTA reassignment | ~£500m–£800m (modelled) |
| Handover preparation | Absent teacher producing cover work | ~£200m–£400m (modelled) |
| Behaviour-incident time | Lost teaching from disruption | Up to 38 days/pupil/year [35] |
| Compliance overhead | Repeat KCSIE induction, SCR upkeep | Variable; severe if absent |
| Attrition replacement | ~£15k per leaving teacher [107] | ~£600m–£1bn (modelled) |
| Pupil-attainment loss | Disadvantaged amplification | Externalised to pupils & families |
| Ofsted-grade risk | Downgrade hits enrolment & funding | School-specific; severe in the tail |
This is the problem. The next page is the response.
Cover My Class is not a timetabling tool, not a recruitment agency. It is the missing infrastructure for the cover moment itself — designed to give the covering adult, before the first lesson, exactly what the absent teacher would have brought to the room.