ICEQC-R-2025-10 — Costing the Teaching Workforce: Salary, Workload, Attrition and Professional Development

Teacher planning is commonly framed as the question of how many teachers a system has or needs. That formulation is incomplete. The object is the sustained provision of instruction by people with the knowledge, preparation and assignment required for the pupils and curriculum concerned. The relevant unit changes as the question changes: persons for recruitment and professional development; full-time equivalents for labour input; posts for establishment control; classes and periods for timetabling; and pupil-hours for service received.

A national total can answer whether the recorded workforce grew. It cannot establish whether the growth occurred at the level experiencing enrolment pressure, whether new teachers reached schools, whether they remained, or whether their qualifications matched the assignment. A complete account therefore follows resources through a sequence: post authorised; funds provided; person appointed; person assigned; person reports; person is present; class is scheduled; instruction is delivered. Loss at any transition has a different cause and remedy.

The framework treats these transitions as an accounting chain. The chain does not imply that teaching is reducible to an administrative transaction. It ensures that administrative claims about service can be tested. Professional judgement and classroom relationships operate within the time and conditions made available; they cannot compensate indefinitely for a post that does not exist, a vacancy that remains unfilled or a timetable that provides no teacher for a required subject.

An **authorised post** is a position permitted within the staffing establishment. A **funded post** is an authorised position for which expenditure is provided for the period. A **filled post** has an appointment recorded. A **person employed** is a unique individual with an active employment relationship. A **full-time equivalent** expresses contracted workload relative to a defined full-time workload. A **school-assigned teacher** has a current duty location in a school. A **teacher present** is observed or validly recorded at the assigned place and time. A **teacher teaching** is delivering a scheduled instructional activity.

The quantities may coincide in a small, stable system with complete records, but coincidence must be demonstrated rather than assumed. One person can fill more than one part-time post; one post can be temporarily covered by several people; a teacher can be seconded to administration while retaining a school code; a payroll record can remain active after departure; and a teacher can be present while engaged in official non-teaching work.

The service-conversion account compares quantities at adjacent stages. The funded-post gap is authorised posts minus funded posts. The vacancy gap is funded posts minus filled full-time equivalents. The assignment gap is employed teacher full-time equivalents minus school-assigned teaching full-time equivalents, after legitimate central, leave and support functions are separately classified. The availability gap is scheduled teacher time minus teacher time available after recorded absence. The delivery gap is available time minus instruction delivered.

These are not performance scores. A gap can be legitimate: authorised positions may be phased in; leave is a lawful employment condition; professional development may temporarily remove a teacher from class; emergencies may close a school. The account requires reasons and replacement arrangements. Its purpose is to distinguish explained service variation from unresolved loss and to show which authority can act.

Because each stage uses a different unit, careless subtraction can mislead. Posts cannot be subtracted directly from headcount where part-time work is material. Presence observed on one day cannot be subtracted from an annual full-time-equivalent stock. The comparison must first align unit and reference period. Where alignment is impossible, the quantities are reported side by side with the incompatibility stated.

The right to education and the commitments made at Dakar direct attention to the opportunity experienced by the child, including children in difficult circumstances and groups subject to discrimination. Workforce administration is therefore not only an internal management function. Allocation decisions determine whether curriculum, language provision, reasonable school access and continuity of instruction exist in practice.[REF-01] [REF-12]

Equity does not require identical staffing ratios in every school. Equal ratios can be inequitable where need and cost differ. A small remote school may require a minimum complement even when its ratio is low; a school serving multiple languages or children requiring specialised support may require additional capability. The public-interest obligation is to apply explicit, relevant criteria and to make the resulting distribution visible.

The 1966 Recommendation recognises the need for adequate preparation, professional conditions, consultation and particular provision for teachers in rural or remote areas. These matters are linked. Deployment cannot be sustained by transfer orders alone when housing, safety, professional support, family circumstances and career progression make a post materially more difficult to hold. A measurement system that reports the vacancy but not the repeated turnover or the conditions of service identifies only the final symptom.[REF-02]

The person is the appropriate unit for age, sex, qualification, experience and entry or exit. The post is the appropriate unit for establishment authority and vacancy. The assignment is the appropriate unit for school, level, subject, grade and function. A controlled register preserves all three and links them with dated identifiers.

Counting assignments as persons inflates the workforce when teachers serve multiple schools or subjects. Counting persons without assignments can include teachers working in administrative offices, teacher-training institutions or non-school functions. Counting filled posts can double count an individual holding fractions of more than one post. The statistical publication is to state which unit is used and how concurrent assignments are consolidated.

Each employment and assignment record requires an effective start and end date. An annual snapshot without transaction dates cannot distinguish a genuine movement from late data entry. Retrospective corrections must remain auditable: the corrected value may replace an erroneous public series, but the reason, scale and release date of the revision are to be retained.

Headcount measures people. It is essential for recruitment, retirement, professional development and occupational conditions. Full-time equivalent measures purchased or assigned labour relative to the national full-time rule. It is more suitable for resource ratios where part-time employment is material. Neither is inherently superior; they answer different questions.

If a full-time workload is 30 defined hours and three teachers work 30, 15 and 12 hours, the headcount is three and the contracted full-time equivalent is 1.9. If the hours include duties beyond teaching, this is not 1.9 instructional full-time equivalents. A further conversion using scheduled contact time is required. Definitions of full-time work, contact time and paid non-contact duty must be recorded.

International comparison of full-time equivalents is sensitive to national conversion rules. Some systems may classify a person by status; others may divide actual or contractual hours. A ratio calculated from full-time-equivalent pupils and teachers is not comparable with one using headcounts unless part-time participation is negligible or an adjustment is made. *Education at a Glance 2004* expressly bases the student–teacher ratio on full-time-equivalent students and teachers and distinguishes it from class size.[REF-08]

A teacher may undertake classroom instruction, school leadership, counselling, curriculum work, supervision or administration. Systems differ in whether school leaders retain teaching loads and whether specialist support staff are classified as teachers. The function performed in the reference period is therefore required in addition to occupational title.

The primary teacher stock for a pupil–teacher ratio ordinarily includes teachers who provide direct instruction at that level under the applicable statistical definition. Personnel wholly assigned to administration or another level are not to be included merely because their payroll grade is “teacher”. Where a principal teaches part of the timetable, only the corresponding instructional workload enters an instructional-capacity measure, while the person remains in the headcount for workforce analysis.

Non-teaching personnel are not residual or unnecessary by definition. The OECD evidence shows wide differences in the proportion and composition of non-teaching staff, reflecting organisation and management. Administrative and support functions can protect teachers' instructional time. The analytical requirement is accurate classification, not automatic transfer of every paid hour into classroom teaching.[REF-08]

Permanent, fixed-term, probationary, contract, community-supported, substitute and volunteer teachers may coexist. Employment status affects cost, security, turnover, eligibility for training and the reliability of future supply. The categories are to be reported separately where they carry materially different rights, qualification rules or workloads.

Temporary personnel can meet a short-duration gap, but a repeated sequence of temporary appointments to a continuing post indicates a structural requirement. Volunteers may extend access in an emergency or remote setting, yet their service cannot be assumed to persist in a multi-year projection without an explicit retention and remuneration basis. Contract status is not a proxy for quality; the relevant evidence concerns preparation, selection, support, workload, conditions and observed teaching.

The 2003 universal-primary-education study considered recruitment of different teacher cadres within financial scenarios and noted unresolved questions about motivation, performance, learning and political sustainability. A planning model may test such arrangements but cannot treat lower salary as an efficiency gain without accounting for selection, training, turnover, supervision and the continuity of service.[REF-07]

Three concepts require separation. **Academic qualification** concerns general or subject education. **Professional qualification** concerns completion of the recognised preparation for teaching. **Trained status** in international reporting commonly means that the teacher has received the minimum organised teacher training required in the country. The third category is nationally defined and cannot establish an internationally uniform preparation threshold.[REF-04] [REF-05]

The register is to record the credential, awarding body, field or subject, completion date, level for which preparation was intended and the national rule under which it qualifies the holder. A simple yes/no field discards information needed when rules change. Teachers in training, provisionally licensed or admitted through an alternative route are to have distinct statuses rather than being forced into “trained” or “untrained”.

Unknown status is a data-quality category. It must not be added to the trained numerator, nor automatically treated as untrained. A published proportion is to show the denominator and the share unknown. If trained status is 80 per cent among the 75 per cent of teachers with known records, the result does not establish that 80 per cent of the full workforce is trained.

For a defined population and period, the closing teacher stock equals the opening stock plus first entrants, re-entrants and transfers in, less retirements, resignations, deaths, dismissals, transfers out and other exits, plus or minus verified record corrections. The identity is to be applied separately to headcount and full-time equivalents. It also requires a boundary: movement from primary to secondary teaching is an exit from the primary stock even if the person remains employed by the same ministry.

The identity makes inconsistencies visible. If the recorded closing stock cannot be derived from opening stock and transactions, either events are missing, records are duplicated, or definitions changed. The difference is a reconciliation residual, not net recruitment. It is to be reported and investigated before the series is used for projection.

Gross entries and gross exits matter even when net change is small. A stock of 100,000 can remain constant after 10,000 entries and 10,000 exits, or after 1,000 entries and 1,000 exits. These systems have different training requirements, induction loads, experience profiles and continuity risks. A net-growth figure conceals them.

The potential entry supply route begins with eligible applicants, continues through admission to preparation, enrolment, progression, completion, recognition, application for employment, selection, appointment and reporting for duty. Attrition occurs at every transition. Training capacity is therefore not the same as employable supply.

For each cohort, planners require the number admitted, completing, meeting the applicable standard, applying, appointed and still employed after the initial period. Field of study, language, level and intended location matter. A surplus of general primary candidates does not resolve a shortage of mathematics teachers, bilingual teachers or staff willing and able to serve isolated communities.

Lead time is central. Expansion planned for the following school year cannot be met through a preparation programme lasting several years unless candidates are already in the supply route or a lawful interim route exists. A projection is to show the earliest date at which additional prepared entrants can reach schools and the capacity required in training institutions, practicum placements and induction support.

Attrition is the loss of teachers from the defined workforce over a period. The denominator is ordinarily the relevant workforce at the beginning of the period, but average-stock denominators may be used if stated. Retirement, death, resignation, dismissal and expiry without renewal are different causes. Transfers between schools are turnover for the school but not attrition from the national workforce. Transfers between levels or occupations depend on the defined boundary.

Early-career attrition requires separate observation because it tests recruitment and induction assumptions. Retirement projections require age and, where relevant, service eligibility. Mortality and long illness may be material in countries affected by HIV/AIDS; the 2003 costing study treated teacher sickness, absence, death, replacement and additional training as connected planning costs rather than as a single attrition rate.[REF-07]

Reason codes require verification. “Resignation” may conceal non-renewal, migration to another education employer or movement to a central post. “Unknown” must remain visible. Exit interviews can add explanation but are subject to non-response and may not represent those leaving under the most difficult conditions.

Growth demand is the additional workforce required because pupils, programmes, schools or curriculum obligations expand. Replacement demand is the recruitment required to offset exits and reductions in workload. Gross recruitment need is their sum, adjusted for expected re-entry and transfers in. A system with stable enrolment may still face a large requirement where the workforce is ageing or attrition is high.

The projection is iterative. New entrants themselves can leave before the horizon; recruitment in one year changes the age and experience distribution in later years. Training institutions also require teachers and may draw from the same qualified population. The model must state whether this institutional demand is included.

The age distribution provides a bounded view of retirement pressure; it is not a substitute for the governing retirement rules or actual behaviour. Experience distribution provides information about induction and career structure but must distinguish total teaching experience from tenure in the present school.

A highly experienced workforce can be a strength and a near-term replacement risk. A very young workforce may indicate successful expansion and a heavy requirement for induction, mentoring and leadership development. The planning interpretation therefore depends on the shape of the distribution, not an average age alone.

Age data are personal and require protection. Public tables use grouped ages and suppress small cells where identification is possible. Missing dates of birth are not to be assigned to the modal age; they are reported as unknown and investigated through personnel controls.

Teacher projections are conditional statements. Enrolment, repetition, progression, retirement, attrition, training yield, workload and salary assumptions all affect the result. The reportable output is not only a central estimate but a range under plausible variants and an account of which assumptions drive the range.

An upper demand case may combine faster enrolment growth, lower pupil–teacher ratios where overcrowding is acute, higher attrition and stronger leave cover. A lower case may reflect slower demographic growth, improved deployment and reduced repetition. These are not optimistic and pessimistic labels unless the educational consequences are established. A lower teacher count produced by exclusion or curtailed curriculum is not a favourable outcome.

Population by single year of age, entry policy, participation, repetition, promotion, dropout and transfer determine projected enrolment. Each component carries uncertainty. A national-age population does not reveal where pupils will attend; district and school demand also depends on migration, settlement and school choice.

The pupil projection is to preserve grade and programme because staffing rules differ. Collapsing all primary enrolment into one total assumes that teachers and classroom places are perfectly divisible and transferable. A school with 83 pupils across six grades does not create the same staffing requirement as 83 pupils in one grade at a large urban school.

Enrolment expansion can be discontinuous after fee removal, compulsory-education reform, displacement or opening of new provision. The model must permit policy shocks and show their implementation date. A straight-line trend is insufficient where the underlying rules change.

In a class-based primary model, required classes are calculated from projected pupils by grade, the maximum or planning class size, school-level minimums and rules for combined grades. Rounding matters: 41 pupils under a planning maximum of 40 require two classes, not 1.025 teachers. Applying a ratio at national level loses this indivisibility.

Multi-grade teaching may be an appropriate organisation for small schools, not evidence of deficiency by itself. Its staffing requirement depends on the grades combined, curriculum design, teacher preparation, materials and timetable. A model that assumes one teacher per grade can overstate demand in sparse areas; a model that assumes unrestricted combination can understate the professional and instructional requirement.

Double-shift arrangements increase the number of pupil places a facility can accommodate but do not automatically double teacher capacity. If teachers work both shifts, workload and contact-time rules govern feasibility and cost. If separate teachers are employed, classroom capacity and teacher demand move differently. Shift status must therefore be explicit in school records.

At levels where subjects are taught by specialists, teacher demand for subject (s) is driven by the number of scheduled pupil-group periods in that subject divided by the number of periods a full-time teacher can teach, with allowance for assignment constraints and lawful non-contact duties. The aggregate number of periods can balance while individual schools have fractional shortages that cannot be pooled across distance.

Curriculum change alters demand even with constant enrolment. Additional science, language, arts or vocational periods may require staff with different preparation. The implementation plan must therefore connect curriculum effective dates to recruitment, retraining and deployment lead times.

Subject labels require consistent coding. “Science” may denote integrated science at one level and separate disciplines at another. A teacher academically prepared in one field cannot be counted as fully deployable to every subject in a broad family without a stated equivalence rule.

Contracted working time includes teaching and may include preparation, assessment, meetings, professional development and other duties. Net teaching or contact time is the scheduled time of direct instruction. The number of teachers required for a given curriculum depends materially on contact-time obligations.

The 2004 OECD indicators show that countries differ in the organisation of teachers' working time and in annual net contact hours. These differences help explain why similar enrolment and class size can require different numbers of teachers. They also limit direct comparison of student–teacher ratios without the timetable context.[REF-08]

Longer statutory contact time does not prove more instruction was delivered. School closure, absence, examinations, training and timetable interruption can reduce actual time. Nor is maximising contact hours necessarily sound: preparation, assessment and professional collaboration are part of teaching quality. Planning requires a lawful and educationally credible allocation of total working time.

Every school faces indivisible functions. A minimum complement may be required for leadership, safeguarding, subject coverage or continuity during absence. The requirement is context-specific and may create low pupil–teacher ratios in small schools. Such ratios must not be automatically labelled inefficient.

Minimum staffing policy is to specify the service protected, the school conditions to which it applies and any alternatives, such as clustering, mobile teachers or supported multi-grade provision. Transport time and reliability enter the calculation where staff serve more than one site.

The cost of minimum provision is part of the public obligation to reach dispersed populations. It is to be distinguished from avoidable fragmentation caused by poorly located schools or uncoordinated provision. The distinction requires geographic and community evidence, not the ratio alone.

A national workforce can equal the modelled national requirement while districts retain vacancies and other districts hold staff above establishment. The arithmetic balance is therefore a necessary reconciliation result, not a finding of equitable supply. Deployment measures compare actual assignment with a transparent school or district entitlement.

The entitlement itself requires review. A rule based only on pupils may fail small schools or subject-specialist programmes. A historical post schedule may preserve allocations after enrolment shifts. The measure is strongest when both the allocation rule and actual circumstances are published.

The mean pupil–teacher ratio is weighted by pupils when calculated as total pupils divided by total teachers. The unweighted mean of school ratios answers a different question and can give small schools disproportionate influence. The median school ratio describes the middle school, while percentiles show dispersion. All may be useful if labelled.

The range alone is unstable and sensitive to data error or a very small school. Interquartile ranges, tenth and ninetieth percentiles, and shares of pupils in schools beyond defined planning bands provide a more informative distribution. Planning bands are administrative thresholds, not universal quality standards, and the publication is to explain their basis.

Remote vacancies often result from a combination of recruitment, living conditions, professional isolation, safety, transport, family needs and delayed payment. A single hardship label is too coarse. The register is to identify the actual characteristics of the post and the duration for which it remains vacant or repeatedly turns over.

The 1966 Recommendation calls attention to housing, travel and hardship arrangements in remote areas. Measurement of incentives therefore includes eligibility, receipt, timeliness and retention after receipt. An allowance announced but not paid cannot be classified as an implemented deployment measure.[REF-02]

Recruitment from underrepresented regions may improve retention where candidates meet preparation requirements and training access is available. The evidence is to distinguish admission, completion, appointment and sustained service. A quota at the first stage does not establish a change at the school.

The proportion of female teachers is a legitimate workforce descriptor and may be relevant to participation, safety, role models and deployment constraints. It is not a universal measure of quality. Interpretation requires level, location, leadership distribution, subject assignment and employment status.

A national increase in the female share can coexist with continuing absence of women teachers in remote schools. Publication therefore needs cross-tabulation by location and level, subject to confidentiality. Policy responses are to address the documented constraint—such as housing, safety, recruitment opportunity or discriminatory practice—rather than infer one from the aggregate.

Deployment quality depends on match between a teacher's preparation and the assigned curriculum. The mismatch rate counts assigned instructional workload outside a defined acceptable field. A person-based rate is less precise when teachers divide time among matched and unmatched subjects.

Language of instruction may create a capability requirement not recorded by general qualifications. The data system is to record demonstrable language capacity using a lawful and relevant classification. Ethnicity must not be used as a proxy for language competence. Personal data require protection and voluntary, appropriate collection where relevant.

Early-grade assignments also warrant separate treatment. A system may place less experienced or less prepared teachers in the earliest grades although foundational instruction carries distinct demands. Experience alone does not determine fitness; the required evidence concerns preparation, support and assignment.

Vacancy is a point-in-time measure. Turnover reveals repeated disruption. School-level turnover is the number of arrivals and departures during a period relative to the average workforce, with expansion appointments identified separately. High turnover can coexist with a low vacancy rate if posts are repeatedly refilled.

Continuity matters for curriculum progression, school planning and professional collaboration. It does not justify preventing lawful mobility or preserving unsuitable assignments. The policy question is whether avoidable conditions cause disproportionate instability and whether replacements arrive before service is interrupted.

Transfer rules are to be published, consistently applied and capable of addressing documented need. A rule based on seniority alone may not resolve subject gaps; a discretionary rule without reasons creates risk of unequal treatment. Teacher participation and appeal mechanisms are consistent with professional status and can improve the evidence available for durable assignment.[REF-02]

Deployment decisions also require an effective date and confirmation of arrival. An order is not a completed transfer. The personnel register, payroll cost centre and school record are to be reconciled after movement so that the same teacher is not retained in two locations.

The 2004 *World Development Indicators* education-input table provides two measures directly relevant to the present framework: the percentage of primary teachers who have received the minimum organised teacher training required nationally, and the primary pupil–teacher ratio. The table generally refers to the 2001/02 school year and identifies values as preliminary. It draws on data compiled by the UNESCO Institute for Statistics from official national responses.[REF-05]

The source warns that education statistics commonly lag by two or three years and that coverage and collection methods vary. It defines trained teachers by national minimum requirements and cautions that preparation varies greatly. It defines the primary pupil–teacher ratio as enrolled primary pupils divided by primary teachers, regardless of teaching assignment. It also notes that definitions, grade-specific class size and hours taught affect comparability. Those qualifications are part of the result, not peripheral notes.

The table shows a broad resource gradient. The reported ratio is 40 pupils per teacher for low-income economies and 22 for middle-income economies; regional values are 45 in sub-Saharan Africa and 42 in South Asia, compared with 17 in Europe and Central Asia and in high-income economies. These figures signal materially different aggregate staffing conditions. They do not indicate the number of pupils in a class or the amount of contact time each receives.

The trained-teacher aggregate does not move in a single relationship with the ratio. Sub-Saharan Africa is reported at 80.4 per cent trained and 45 pupils per teacher, South Asia at 66.9 per cent and 42, and Latin America and the Caribbean at 76.7 per cent and 26. Because trained status follows national requirements, the percentage cannot be read as a common proficiency scale. It nevertheless reveals the scale of the nationally recognised preparation gap within each reporting boundary.

The global values also illustrate why a composite score would be unsound. Combining a trained share and a ratio would require assumptions about equivalence, substitution and weighting. Ten additional pupils per teacher cannot be offset arithmetically by a specified number of percentage points in trained status. They describe different conditions and may call for different interventions.

The country table contains very wide variation. Examples reported for 2001/02 include primary pupil–teacher ratios of 10 in Denmark and the Netherlands, 14 in Cuba and Estonia, 20 in China and Malaysia, 40 in India and Nepal, 56 in Cambodia, Mali and the Republic of Congo, 61 in Cameroon, 66 in Mozambique and 71 in Chad. These selected values are not a ranking; they demonstrate the range that a national average can occupy and the need for contextual diagnosis.[REF-05]

Training data are absent for many countries that report a ratio, and the reverse can also occur. Missingness limits any analysis restricted to complete pairs and may produce a non-representative sample. The proper publication reports the number of countries and population coverage contributing to an aggregate when the source makes that information available; it does not interpret an absent value as zero or full compliance.

Values for individual countries can also refer to 2002/03 where marked, and source notes may modify coverage. A comparative extract is therefore to carry the country-specific note. The convenience of a single table does not remove this obligation.

*Education at a Glance 2004* defines the ratio of students to teaching staff using full-time-equivalent students and teachers. It defines class size through the number of students in a class or teaching group. The report demonstrates that two countries can have similar student–teacher ratios and different class sizes because teacher working time, pupil instruction time and the organisation of groups differ.[REF-08]

The relation can be expressed conceptually. Average class size rises with the number of pupils and pupil instructional periods and falls with the number of teachers and teacher contact periods, subject to scheduling and school constraints. Consequently, a system can reduce a resource ratio by employing more teachers without reducing class size if those teachers have lower contact loads or undertake other functions. It can operate relatively small classes with a less favourable ratio if pupil instruction time or teacher organisation differs.

Cross-country reporting is therefore to place ratio and class-size results in separate columns and avoid causal language. A class-size policy affects rooms, class formation and timetables; a staffing-ratio policy affects aggregate labour resources. They interact but are not the same instrument.

The OECD publication records annual net contact time in public institutions and explains that teaching time is only part of teachers' working time. Preparation, assessment, professional activity and administration may be regulated differently. The comparison is useful because the same teacher FTE can support different quantities of scheduled instruction under different contact-time rules.[REF-08]

Annual measures also depend on the length of the school year and period definitions. Weekly loads cannot be compared without the number of teaching weeks. Statutory time is not observed time. A complete comparative account identifies whether a value is required, scheduled or delivered.

The 2004 *World Development Report* reports preliminary absence estimates from unannounced visits. For the primary-school observations reproduced in its table 1.2, the reported rates were 16 per cent in Ecuador, 25 per cent for an average of 14 Indian states, 18 per cent in Indonesia, 15 per cent in Papua New Guinea, 13 per cent in Peru, 17 per cent in Zambia and 26 per cent in Uganda. The related state table for India ranged from 15 per cent in Gujarat to 39 per cent in Bihar among the listed states.[REF-06]

The 2003 World Bank study examined 55 large low-income countries and used a detailed subset for service-delivery analysis. In its adjusted sample, primary pupil–teacher ratios ranged from 13:1 to 79:1. The group characterised by relatively higher primary completion had an average ratio of 39:1; the full sample average was 44:1. The study used 40:1 as a 2015 parameter in simulations that also adjusted salary, repetition, non-teacher inputs, public finance and private enrolment.[REF-07]

The source itself provides reasons not to detach the number from its model. It acknowledges that pupil–teacher ratio is not perfectly correlated with class size, that country geography can constrain deployment, and that the parameter forms one part of a balanced package. Its regression table reported essentially no independent association for the ratio within the specified model after other variables were included, while repetition and spending variables carried different relationships. That result does not prove ratio is irrelevant; it warns against presenting the benchmark as an isolated causal threshold.

The study's group data also show why cost per graduate and cost per pupil differ. One group combined high enrolment with low completion and very high repetition. Another combined low coverage, high teacher salaries relative to national income and high ratios. In each case, changing teacher numbers alone would leave a material constraint unresolved.[REF-07]

The available evidence supports a clear but limited assessment. Large differences in aggregate teacher resources exist; nationally recognised preparation remains incomplete in many reporting systems; teaching-time organisation affects the meaning of resource ratios; and presence losses can materially reduce service. The evidence does not support a universal staffing number, a common training-quality scale or a global estimate of delivered instructional FTE.

Accordingly, comparative reporting is to use a set of non-substitutable measures. It may identify systems or regions for further investigation, but policy selection follows national and subnational diagnosis. The stronger the claim—especially a claim about quality, causation or individual conduct—the more direct the evidence required.

Presence records whether a teacher who was expected at a specified location and time was there. It is narrower than attendance over a term and broader than unauthorised absence. A valid presence measure begins with a roster of persons expected, excludes those not scheduled, and records the reason and authorisation status separately from the initial observation.

Unannounced observation can reduce reporting bias but one visit has high sampling uncertainty for an individual school. Repeated visits across days and seasons are required for system estimates. The 2004 *World Development Report* tables define absence as a person supposed to be present but not present during an unannounced visit, including authorised and unauthorised reasons. The preliminary country estimates ranged from 13 to 26 per cent for the primary-school observations shown, while the reported state values within India ranged more widely. These figures demonstrate measurement relevance; they do not establish a current rate for every school or a cause applicable across countries.[REF-06]

Administrative attendance systems cover more days but can be affected by incomplete signing, retrospective alteration or unclear schedules. Observation and administrative data therefore provide complementary checks. Payroll deductions are not a presence measure, since authorised leave, delayed procedure and sanctions follow different rules.

Absence is to be classified by reason: authorised leave, illness, professional development, official non-teaching duty, delayed assignment, industrial action where applicable, unauthorised absence and unknown. The classification is completed after verification, not inferred by an observer. Duration matters because one extended absence and repeated single-day absence have different replacement and continuity effects.

A service measure also records whether a qualified substitute or another planned arrangement maintained instruction. The educational loss is scheduled periods not delivered or materially curtailed, not the number of absent persons alone. Replacement by combining classes may preserve supervision while reducing instructional conditions; it is reported as a distinct arrangement.

The official school calendar states the days or hours intended. School-level closure can result from weather, insecurity, elections, examinations, public-health events, building failure or local administrative decisions. Teacher presence cannot restore time when the school is closed, so closure is measured separately.

Within an open day, late start, early finish, extended breaks and timetable gaps reduce available instruction. Reliable measurement may use sampled observation, timetable records and school logs. The burden is to be proportionate: the objective is a defensible system estimate and timely local response, not continuous surveillance of every professional action.

The multiplication assumes independent and correctly aligned proportions; actual records are preferable. The example is not a performance target. It shows why a payroll stock can overstate delivered capacity and why double-counting losses must be avoided. If absence is already included in delivered-period data, it is not applied again.

Instructional capacity is an enabling condition. It cannot establish clarity of explanation, appropriateness of pedagogy, classroom climate, formative assessment or learning. The quality framework in the 2005 EFA Global Monitoring Report connects learner characteristics, context, enabling inputs, teaching and learning procedure, and outcomes; the workforce account occupies only part of this wider structure.[REF-03]

For this reason, no composite “teacher quality score” is proposed. Credentials, experience, presence and contact time are not interchangeable and cannot be added without arbitrary weights. Each measure supports a specific decision and retains its limitation.

The personnel register is the authoritative record of unique persons and dated employment events. Its minimum fields include identifier, employment status, workload, preparation, subject or grade field, employing authority, assignment, start and end dates, leave status and source of verification. Sensitive information is protected and public outputs are aggregated.

Identity controls must accommodate legitimate name variation without creating duplicate persons. A correction procedure is required for errors, and access to personal records is logged. The statistical function does not require publication of names, salary details or individual attendance.

The establishment records posts and organisational authority; payroll records payment. They are related but neither proves school service. Reconciliation identifies active payments without a current personnel record, filled posts without payment, persons charged to the wrong location, duplicated payments and delayed removal after exit.

An exception is a case for investigation, not proof of wrongdoing. Timing differences, lawful arrears and approved secondment can explain a mismatch. The control result records resolution and prevents unresolved exceptions from silently entering published teacher totals.

The school census links teachers to pupils, grades, classes and facilities at a common reference date. It is indispensable for deployment ratios. It may nevertheless count staff reported by the head teacher rather than verified persons, omit teachers temporarily away or include visitors. Stable school and person identifiers permit reconciliation.

Timetables translate assignment into scheduled service. They require period length, subject, grade or class, teacher and day. A timetable that assigns one person simultaneously to different classes is internally invalid. Rule-based validation may identify the conflict, but its correction remains an administrative task supported by the school record.

Training providers report admissions, enrolment, completion and field of preparation. These data require cohort identifiers so that completion yield is not calculated from unrelated annual totals. Recognition status and programme changes are dated.

Linkage to employment statistics is performed under lawful data-protection arrangements. Public analysis reports aggregate transitions and does not expose applicant-level results. Where linkage is unavailable, the uncertainty between completions and appointments is stated rather than filled by assumption.

Common codes are required for education level, programme, subject, school, district, employment status, qualification and exit reason. ISCED 1997 supports international level mapping, while national detail is preserved beneath that map.[REF-10]

Classification change creates a break unless records can be consistently recoded. Version numbers and effective dates are therefore part of the statistical record. A new definition of “trained” is not to be presented as a change in teacher preparation unless historical data have been recoded on the same basis.

Official workforce statistics require impartial release, disclosed methods and correction of material error. Political or administrative preference cannot determine whether an unfavourable result is published. The Fundamental Principles of Official Statistics provide the relevant basis for professional methods, transparency and confidential treatment of individual data.[REF-11]

Revisions are dated and explained. A census update, payroll cleaning exercise or reclassification may alter prior years without indicating an actual workforce change. Both the revised series and the break explanation are necessary for public interpretation.

A workforce scenario begins with the service objective: projected pupils to be taught under a defined curriculum, school network and set of access conditions. It then states class formation or subject-period demand, teacher workload, minimum school staffing, expected availability and replacement assumptions. Only after those choices is the teacher requirement calculated.

At least three cases are advisable where uncertainty is material: a central case, a demand-pressure case and an implementation-capacity case. Each case uses coherent assumptions. Combining high enrolment with low training capacity but leaving vacancies at zero is internally inconsistent.

Ratios can provide a rapid aggregate check and a scenario parameter. The 2003 study of universal primary completion observed an average pupil–teacher ratio of 39:1 among a group of higher-completion low-income countries and used 40:1 in its 2015 costing scenarios. The authors also documented substantial country variation and placed the ratio alongside salaries, repetition, non-salary inputs and finance.[REF-07]

This evidence supports two restrained conclusions. Extremely high or low ratios warrant diagnosis, and balanced resource parameters matter. It does not establish a universal optimum. Country geography, school size, curriculum, contact time, teacher preparation and fiscal capacity alter both cost and feasibility.

The salary bill is calculated from filled FTE by grade, salary point, allowances, employer contributions and time. Average salary multiplied by teacher count is insufficient where the grade distribution or hardship payments change. Vacancies, arrears and newly recruited cadres are separately shown.

Teacher expenditure cannot be planned in isolation. The 2003 costing study explicitly protected non-teacher inputs, including materials, development, supervision, assessment and management, because unchecked salary pressure can displace conditions required for effective teaching.[REF-07]

Expansion requires the cost of preparation places, practicum, candidate support, recognition, recruitment, induction and mentoring. Rapid change can also require temporary classrooms, teacher housing, transport or additional supervision. These costs have different timing from salaries and must appear in the implementation schedule.

Where unprepared teachers are already serving, in-service routes require release time and replacement. Counting the same teacher as fully available for instruction and full-time preparation over the same period overstates capacity.

The recurring salary obligation continues after construction or initial external support ends. A scenario therefore reports recurrent cost at the horizon, its share of the relevant public resource envelope and the sensitivity to salary and attrition. External finance may support transition, but an unfunded future wage bill is not a sustainable staffing plan.

The United Nations Millennium Project report available in January 2005 calls for country-based strategies that connect investment needs to public management and financing. Workforce projections serve that procedure when assumptions, sequencing and domestic responsibilities are explicit.[REF-09]

The national profile reports the reference-year workforce in headcount and FTE, the pupil–teacher ratio, the proportion meeting the current national preparation requirement, funded vacancies, verified attrition and the distribution of the principal measures. It states coverage of non-government provision, temporary staff and missing records.

The headline is not to describe a system as having “enough teachers” solely because one ratio meets a planning value. A defensible statement identifies the scope: for example, aggregate filled FTE equalled modelled national demand under stated assumptions, while specified districts and subjects retained shortages.

District profiles support deployment and resource decisions. School-level data are used administratively but public release is governed by privacy, data quality and risk of misinterpretation. Small cells concerning qualification, absence or personal characteristics are suppressed or combined.

Publication can identify schools lacking a service where this is necessary for public action, but it must not infer individual failure from system data. The responsible authority and correction route accompany material deficiencies.

Every comparative table includes level, sector, reference year, teacher unit, pupil unit, full-time-equivalent rule, training definition and coverage. If one country reports headcount and another FTE, direct ranking is withheld or clearly qualified.

Comparative results also disclose whether private institutions and teachers on leave are included. A wider coverage can raise or lower a ratio without a change in service. The user is to be able to reconstruct the statistical boundary from the note.

The framework supports workforce budgeting, training-intake planning, identification of distributional gaps, subject scheduling, monitoring of vacancies and assessment of data integrity. It supports questions, not automatic sanctions.

It is inappropriate to use a pupil–teacher ratio as an individual teacher-performance measure, to equate trained status with teaching effectiveness, to interpret a one-day absence observation as misconduct, or to infer learning from contact time. It is also inappropriate to compare salary cost without workload and price context.

Pre-primary staffing cannot be derived by transferring a primary pupil–teacher ratio. The age of children, duration of sessions, group organisation, care and protection responsibilities, adult support and the boundary between education and other early-childhood services affect the required workforce. Titles and preparation routes also vary, making international teacher counts especially sensitive to classification.

The planning register is to distinguish teachers with direct educational responsibility, classroom assistants and personnel whose principal function lies outside education. Their contributions are complementary, but merging them into a single denominator obscures both cost and professional preparation. Where sessions are part day, child and staff full-time equivalents require a stated conversion.

Small enrolments and community provision may produce fragmented employment. Headcount can substantially exceed FTE, and annual turnover may be masked by short contracts. A pre-primary profile therefore reports hours, setting type, preparation status and continuity in addition to persons.

In a primary class-teacher system, the core demand unit is ordinarily the class or multi-grade group. National enrolment divided by a ratio is insufficient because every school has grade-specific, integer requirements. Entry expansion first raises demand in early grades and then moves through the cycle; the location and timing of the cohort must be retained.

Preparation for foundational literacy and numeracy, language of instruction and multi-grade practice is part of deployability. A teacher qualified for primary education under the national rule may still require specific support for an unfamiliar language or combined-grade assignment. Such support is not represented by the trained-teacher percentage.

Primary planning also needs continuity. Repeated movement of class teachers can disrupt progression even if each vacancy is filled quickly. The school turnover rate, days without a regular teacher and replacement arrangement provide a stronger account than an annual vacancy snapshot.

The transition from a class-teacher to a subject-specialist model can create a structural break in demand. The same pupils generate several subject-period requirements, and teachers' timetables contain non-contact preparation and assessment time. A national surplus in one subject does not offset a shortage in another.

Small lower-secondary schools face fractional subject loads. Permissible solutions include teachers prepared across related fields, coordinated timetables across nearby schools and carefully scheduled itinerant provision. Every solution requires a valid subject-match rule and travel or coordination time. Treating fractional needs as nationally poolable can produce a calculated balance that no timetable can operate.

Expansion of secondary access also increases demand for laboratories, materials and specialised rooms. Hiring subject teachers without the enabling conditions may increase payroll but not the intended curriculum. Workforce and non-salary planning remain connected.

Upper-secondary academic and vocational programmes require staffing by programme, subject and workshop rather than level alone. Occupational expertise may be acquired outside the teacher-training route, while pedagogical preparation remains relevant. The register is to preserve both dimensions instead of classifying all specialists as trained or untrained through one field.

Part-time practitioners can provide current occupational expertise. Their headcount is a poor measure of capacity if each contributes few periods. FTE, timetable availability and continuity across the programme are necessary. Equipment safety and supervision can also set minimum staffing independent of enrolment.

Programme closures and labour-market change can create subject-specific surplus that is not readily transferable. Redeployment scenarios state the additional preparation required and do not count a person as immediately deployable merely because an employment post remains funded.

A policy that removes fees, extends compulsory education or opens a new grade can change demand faster than the preparation supply route. The first-year plan identifies teachers already qualified but not employed, possible re-entrants, lawful temporary routes, deployable staff and the training capacity required for later years. Each source is reported separately because its cost and sustainability differ.

Rapid expansion carries a distributional risk. New posts may be filled first in accessible areas, leaving the intended beneficiaries with larger classes or less-prepared staff. Monitoring therefore compares the pace of enrolment growth, post creation, appointment and arrival by district and school.

The 2003 universal-primary-education study documents how large access expansion can be accompanied by rising pupil–teacher ratios and increased use of unqualified staff if the workforce response lags. The planning implication is not to delay access until every input is ideal; it is to make the transition condition, duration and corrective path explicit.[REF-07]

Conflict and displacement can alter pupil location, destroy records, interrupt payment and make ordinary deployment unsafe. Historic establishment figures rapidly lose relevance. A temporary register may be necessary, but it records the source and verification status of each entry so that emergency data do not become an unexamined permanent payroll.

Planning uses short review intervals and location ranges rather than precise long-term forecasts where movement is continuing. Minimum service, safe access, language and psychosocial conditions may require staffing arrangements not captured by ordinary ratios. Teachers themselves may be displaced and require recognition of prior service or credentials where records can be verified.

Security information is handled carefully. Publishing exact teacher or school locations can create risk. Public accountability can report aggregate coverage and service gaps without releasing operational details that endanger staff or pupils.

At the 2008 cut-off, HIV/AIDS constitutes a material workforce and pupil-planning concern in heavily affected countries. The 2003 costing study identifies sickness-related absence, teacher mortality, replacement recruitment, training demand and increased numbers of orphans requiring support as connected consequences. It estimates costs through explicit epidemiological and demographic assumptions rather than treating the epidemic as a general narrative risk.[REF-07]

National planning requires confidential, aggregate analysis. Individual health status is not needed for routine teacher statistics and must not be inferred. Mortality and long-term absence data may inform projections when collected lawfully; workplace support, leave and replacement are handled through applicable employment and health arrangements.

Sensitivity cases are preferable to false precision. The projection states mortality, long-absence and replacement assumptions and varies them within evidence-based ranges. It also tests the training system's ability to replace losses without lowering preparation standards or stripping experienced staff from training roles.

Geographic dispersion changes the relation between efficiency and access. A school or island may require at least one teacher even when enrolment is very low. National ratios can then appear favourable while the system faces high unit costs, single-teacher vulnerability and limited curriculum breadth.

Planning maps travel time, seasonal accessibility, communications and the feasibility of shared staff. The 2003 costing study expressly notes that low population density and geography can make movement towards a common ratio difficult. A model that ignores this constraint turns legitimate access cost into apparent inefficiency.[REF-07]

Resilience requires reserve arrangements. If a one-teacher school loses its teacher, instructional capacity falls to zero. A modest system-level substitute pool, housing provision, scheduled relief and cluster support may have greater service value than reducing the nominal ratio elsewhere.

Where non-government schools educate a material share of pupils, a public system profile that excludes their teachers does not describe the full educational workforce. Coverage is to follow the legal and statistical mandate and is clearly labelled. Public funding, employment responsibility and data reporting are separate questions.

Community-supported teachers may not appear on central payroll. Their workload, preparation, payment regularity and continuity still affect pupils. Inclusion in statistics must not be presented as a change in governmental employment status or an endorsement of employment conditions.

Double counting is a risk where a teacher works in both government and non-government institutions. Person headcount is consolidated within the defined population, while assignment FTE is allocated to each sector under a stated rule.

Decentralisation can place establishment, appointment, payment and supervision at different levels. The national information system must therefore distinguish who owns the record and who is accountable for each transition. A central payroll cannot correct a local timetable; a school cannot fund an unauthorised post.

Comparable national statistics require common definitions and exchange schedules, not necessarily centralised administration. Local systems can retain operational detail while submitting a controlled set of person, post, assignment and school records. The central publication reports coverage and late or non-reporting authorities.

Local discretion is valuable where circumstances differ, but exceptions to the staffing rule are coded and reviewable. Otherwise, variation in ratios cannot be separated into legitimate adaptation, historical allocation or unequal treatment.

The first phase is reconciliation: establish unique persons, dated assignments, valid posts and a common school list. The second phase is diagnostic: calculate stock, flow, deployment and time measures with uncertainty. The third phase is projection: connect pupil demand, curriculum, workload and exits. The fourth phase is implementation: align training, recruitment, deployment, conditions and finance. The fifth phase is public review: publish results, revisions and unresolved gaps.

Sequence matters. A sophisticated projection cannot repair duplicated persons or invalid school codes. Equally, data cleaning without a decision purpose can consume capacity while vacancies persist. Each data improvement is tied to a planning decision it enables.

No single office holds all evidence. Finance authorities control funding and payroll; employer authorities control appointments; districts and schools confirm assignment and service; training authorities govern preparation; statistical offices maintain classifications and publication standards. A written data-responsibility schedule is therefore required.

Professional participation is also material. Teachers and their organisations hold evidence about workload, feasibility, rural conditions and the consequences of timetable rules. Consultation does not replace governmental responsibility; it improves the factual basis and legitimacy of decisions.[REF-02]

Priority research concerns the relationship between deployment, effective instructional time and pupil outcomes; the retention effects of different rural-service arrangements; the transition from preparation to appointment; subject mismatch; and the reliability of administrative presence measures. Evaluation designs need comparison groups or credible counterfactuals where causal claims are intended.

International work is also required on definitions of trained status, FTE conversion and attrition. Harmonisation is valuable where it preserves national detail and does not imply that differently constituted credentials are equal.

Comparative teacher data remain incomplete and depend on national definitions. Several sources use different reference years, and school-level distributions are not consistently available. This report therefore provides a measurement and planning architecture rather than a new global estimate of teacher shortage.

The framework cannot determine the proper curriculum, salary or employment model for a jurisdiction. It cannot remove uncertainty from demographic projection or predict individual behaviour. It can require that these choices and uncertainties be visible.

The comparative evidence is predominantly descriptive. Differences in pupil–teacher ratios, nationally defined preparation status, reported presence or completion do not isolate the effect of one workforce condition from finance, pupil composition, curriculum, school organisation or wider social circumstances. The report therefore does not estimate a causal return to recruitment, class-size change, qualification reform or attendance control. Those questions require designs capable of distinguishing the intervention from concurrent change and of examining implementation over time.

Geographic evidence is also uneven. National and regional aggregates can indicate where further inquiry is required, but they cannot reveal every school, subject or population gap. Conversely, a small number of documented local cases cannot establish a national pattern. The planning method addresses this limitation by moving from aggregate signal to reconciled school and timetable evidence before an allocation decision is made. Where such evidence is absent, the limitation is a reason to narrow the conclusion, not to substitute an assumed distribution.

Teacher supply is sufficient only when it becomes sustained, appropriately prepared and equitably distributed instructional capacity. Headcount alone cannot establish that condition. Neither can a favourable national ratio, a training credential without its definition, or a funded post without a person present to teach.

The necessary discipline is straightforward in principle: count the correct unit; preserve the transitions from post to service; project entries and exits; translate curriculum and school organisation into demand; examine distributions; measure time without confusing it with quality; and publish limitations beside results. Where the chain breaks, the break is the planning finding.

The public value of teacher statistics lies in their capacity to direct resources to children whose instruction is least secure. A system that makes missing assignments, unresolved vacancies, unequal deployment and lost instructional time visible is better placed to act than one that protects an aggregate appearance of balance.

All indicators require a stated education level, sector, geography, reference date or period, and source. “Teacher” denotes a person assigned to direct instruction under the applicable national and international statistical definition. Mixed-function staff are apportioned in FTE measures where valid workload data exist; otherwise they are separately classified.

Rates use non-zero denominators and report missing components. Percentages are not calculated where coverage is materially incomplete unless the covered population is explicit. Results are rounded only after calculation. Suppression for confidentiality is not treated as zero.

The pupil denominator is aligned to the same level, sector and reference period as the teacher numerator. Where pupil FTE is unavailable, headcount-based ratios are labelled accordingly. Boarding, distance and part-year participation rules are disclosed where material.

The teacher denominator for qualification shares includes only records with known status, while the percentage unknown is shown against the full stock. For absence observation, the denominator contains persons expected at the observed time, not all employees.

School ratios are calculated before percentile analysis. Pupil-weighted exposure attaches each pupil to the ratio of the school attended. District aggregates are calculated from district totals rather than averaging school ratios unless the statistic is expressly the mean school ratio.

For level (l), location (g), subject or function (s), and period (t):

T_{lgs,t+1}=T_{lgs,t}+E_{lgs,t}+R_{lgs,t}+M^{in}_{lgs,t}-A_{lgs,t}-M^{out}_{lgs,t}+C_{lgs,t}

where (T) is stock, (E) first entry, (R) re-entry, (M) transfer or migration, (A) attrition and (C) verified correction. Every term uses the same headcount or FTE basis.

Gross appointments required for a target stock equal positive growth demand plus expected exits and workload reductions, less credible re-entry and transfers in. Candidate completions required equal appointments divided by expected appointment yield. If preparation completion yield is below one, admissions required are calculated separately using a cohort model.

For each school and grade, classes required are the upward-rounded quotient of projected pupils and the applicable class-formation value, modified by approved multi-grade rules and minimum staffing. Aggregation occurs after school calculation. Rounding only after national aggregation understates demand where many small residual classes exist.

For subject (s), required FTE equals total required subject periods divided by full-time teacher contact periods, adjusted for valid cross-school pooling and minimum viable assignments. Fractional needs cannot automatically be pooled across schools where travel makes schedules incompatible.

Availability-adjusted demand divides scheduled service by the expected delivered share only where the loss is not already built into reserve staffing. The adjustment is scenario-specific. It is not a licence to normalise avoidable absence; improvement cases show the workforce and service consequences of reducing preventable loss.

Tests identify duplicate protected identifiers, identical biographical records under different identifiers, concurrent full-time assignments, assignment after recorded exit and payroll activity before appointment. Matches are reviewed; rule-based flags do not determine identity.

Start dates precede end dates. Assignment, qualification and employment statuses are valid on the reference date. A school code is active for the same period. Transactions entered late retain both event and entry dates.

No person exceeds the permitted combined workload without a documented exception. Timetable periods do not overlap. Teacher level and subject codes are compatible with the assignment rule or carry a mismatch status. Funded vacancies cannot exceed funded establishment.

Personnel records are compared with payroll, school census and establishment totals. Differences are classified by timing, coverage, definition or unresolved error. A reconciliation statement reports both counts and the number resolved.

A published scenario record contains the base year; projection horizon; population source; enrolment, repetition and progression assumptions; school network; class-formation rules; curriculum periods; teacher workload; minimum staffing; qualification boundary; attrition by reason; retirement behaviour; training yield; appointment yield; salary and allowance path; and price basis.

At minimum, the record varies enrolment, attrition, contact-time demand and appointment yield. Where salary expenditure is material, salary growth and grade distribution are varied. Where remoteness is material, the feasibility of redistribution and the cost of minimum schools are tested.

The adopted scenario identifies which assumptions are policy choices and which are forecasts. It names the authority responsible for each implementation condition and the date on which actual results will be compared with the projection.

The sampling frame is the current school list with region, level, management and accessibility strata. Probability selection supports system inference. Purposeful visits can investigate known risks but are not combined with probability estimates without appropriate weighting.

Observers record school opening, persons expected, persons present, scheduled classes and classes operating at the visit time. They do not classify the reason for absence without subsequent verification. Visit date and time are protected during fieldwork to reduce anticipation.

Preliminary discrepancies are returned through a controlled procedure to the competent authority and school for factual verification. Individual due procedure is preserved. The statistical report uses aggregate, reviewed classifications and states the share unresolved.

No pupil or teacher is publicly identified. Observation does not intrude on instruction beyond what is necessary. Data are retained only for the authorised purpose, with access limited and correction available.

Before release, the responsible statistical body confirms that units and reference dates align; source coverage is stated; unknown values remain visible; training definitions are current for the year; revisions are logged; tables reconcile with source files; and narrative claims do not exceed the data.

The publication confirms that pupil–teacher ratio is not called class size; trained status is not called teaching quality; presence is not equated with unauthorised absence; contact time is not equated with learning; and a national balance is not called equitable deployment without distributional evidence.

No weakness in the data justifies silence about teacher distribution. It requires a result matched to the strength of evidence. A verified national headcount may be published as headcount; it is not converted rhetorically into classroom availability. A sampled presence estimate may inform system planning; it is not used to declare individual fault.

The framework is complete when every headline can be traced to a defined unit, period and source, and when every limitation that could change the decision is visible to the reader. That standard protects teachers from unsupported inference, protects public funds from false reconciliation and protects pupils from the assumption that a recorded post is the same as instruction received.

An establishment review determines whether authorised and funded posts correspond to the services that schools are required to provide. It is not a general reduction exercise. Its reference population comprises every active school and programme within the stated sector, together with central or district teaching-support functions that draw from the same occupational establishment.

The review begins from service requirements rather than inherited post totals. For each school it records projected pupils, grades, classes, curriculum periods, approved school hours, leadership arrangements, specialised obligations and justified minimum staffing. It then compares calculated entitlement, authorised posts, funded posts, filled FTE and assigned instructional FTE.

A common baseline date is selected before calculations. Appointments, transfers, exits and school changes occurring after that date remain in a transaction window and are not mixed into the snapshot. Without this separation, a person may appear at an old school in the personnel extract and at a new school in the school census.

The baseline is not frozen administratively: urgent appointments and transfers continue. It is frozen statistically so that every comparison refers to a reproducible point. The final review applies validated transactions and produces a second dated position.

The entitlement function is documented in executable rules or an auditable worksheet. It specifies upward rounding, treatment of combined grades, leadership release, subjects, shifts, small schools and travelling staff. Exceptions carry an authority, reason and expiry or review date.

Entitlement is calculated at school level before aggregation. Applying a national ratio first can suppress the indivisible staffing requirements of small schools. Conversely, school-level calculation without a rule for feasible multi-grade organisation can overstate need. Both the entitlement and its educational assumptions are reviewed.

A school may be below entitlement, at entitlement or above entitlement in aggregate FTE. A second classification tests subject, grade and qualification fit. A school at aggregate entitlement but lacking a teacher for a required subject is not fully staffed. A school above aggregate entitlement may nevertheless have a justified minimum or a mismatch that prevents redistribution.

An apparent surplus is not automatically available for transfer. The review tests minimum school staffing, actual timetable, part-time fractions, impending retirement, approved leave, language or specialist functions, travel feasibility and expected enrolment. Only the remaining deployable excess enters a redistribution scenario.

Transfer feasibility is constrained by employment rules, consultation, notice, family circumstances and professional requirements. A scenario that counts every numerical surplus as immediately movable overstates implementation capacity and can disrupt schools without filling the target vacancy.

A shortage is classified by duration and remedy horizon. A timetable adjustment may resolve a fractional subject gap within a term; recruitment may require a year; preparation of new teachers may require several years. Temporary measures are identified as such and carry an end condition.

Shortage caused by inaccurate enrolment, duplicated posts or an unrecorded transfer is corrected in the data before resources are allocated. Shortage caused by a genuine service obligation becomes part of the recruitment, deployment or school-organisation plan.

Schools receive the data and calculation underlying their classification and can submit factual correction. The review distinguishes correction of fact from appeal against policy. Aggregate findings are published with methods, while personal records remain protected.

The qualification rule register records each recognised route to teaching, its effective dates, education level, subject or phase, required academic education, professional preparation and transitional provisions. Historical records are evaluated under the rule applicable at the reference date unless the policy question expressly requires current status.

A change in requirements can move teachers between categories without any new training. Time-series publication therefore provides both the legal-definition series and, where feasible, a constant-definition analytical series. It does not describe the discontinuity as improvement or decline in capability.

Verification confirms awarding body, programme, completion and identity. It does not evaluate the quality of an individual teacher. Where original records are unavailable, the competent authority applies a documented alternative-evidence procedure. Pending cases remain pending or unknown.

Foreign or cross-jurisdiction credentials are classified only after the applicable recognition decision. The statistical office does not infer equivalence from title similarity. Recognition for employment and evidence of subject preparation are maintained as separate fields where national rules distinguish them.

The match matrix maps fields of preparation to curriculum assignments. It may contain full match, accepted related field, provisional match requiring support, and no established match. The matrix is level-specific and dated. Broad occupation codes are not sufficient.

In-service participation is recorded by content, duration, provider and completion. Attendance at a short activity does not alter trained status unless the national rule expressly provides that result. Professional development is analysed as continuing preparation rather than a substitute label for initial qualification.

Release time is included in workforce planning. If training occurs during scheduled instruction, the replacement or rescheduling arrangement is recorded. A programme can strengthen future capacity while temporarily reducing current contact time; both effects belong in the plan.

The match rate indicates formal alignment of preparation and assignment. It does not prove command of content, pedagogy or pupil learning. Conversely, a mismatch classification does not prove ineffective teaching: experienced teachers may have acquired substantial competence outside the recognised route. The measure identifies a planning and support question under the governing rule.

The primary estimand is the proportion of teacher-observation opportunities in which a person expected to be present is observed present. Alternative estimands include the proportion of teachers absent at least once across visits, the proportion of scheduled periods delivered and the proportion of pupil-hours lost. They are not interchangeable.

The denominator is constructed from the valid timetable and leave information applying at the visit. Persons not scheduled at that time, no longer assigned, or lawfully working at another location are excluded from “expected” only after verification.

Schools are commonly the first-stage sampling unit, with teachers or periods observed within selected schools. Selection probabilities are retained. Estimates use weights and account for clustering. A simple mean across visited schools can overrepresent small schools if the target is teacher opportunities or pupil exposure.

Stratification by region, level, management and accessibility can improve precision and ensure coverage of policy-relevant groups. Oversampled strata are returned to their population weight for national estimates. Separate estimates are published only where sample size supports them.

One visit estimates a cross-sectional system rate under the design but provides limited evidence about persistent school or person patterns. Repeated visits permit analysis of recurrence and seasonality. The interval and visit window are chosen without creating a predictable schedule that changes behaviour.

Observers record presence, schedule and immediate factual information. A separate verification procedure checks authorised leave, training, official duty and assignment. “Unknown reason” remains a result where evidence is not available. The analysis can report total absence and the verified distribution of reasons without converting unknown cases into unauthorised absence.

A closed or inaccessible selected school is not removed without classification. If closure is itself part of service availability, it enters the appropriate estimate. Security or access non-response is reported and may require adjustment; substitution with an easier school can bias results.

Results state field dates, school-year position, sampling frame, achieved sample, weighting, visit definition, reason-verification procedure and uncertainty. Preliminary results are labelled. The report does not publish names or use a system estimate to allege individual misconduct.

Suppose removal of a direct charge is expected to increase grade-one entry sharply in the next school year. The immediate requirement is estimated by school and grade, not by applying a national ratio to total enrolment. Existing classrooms, double-shift feasibility, teacher candidates already prepared and the distribution of growth determine what can open on time.

The plan distinguishes temporary and permanent demand. A one-year entry bulge progresses through the cycle; its teacher and classroom requirement moves by grade. If participation remains permanently higher, later cohorts sustain the expansion. Recruitment, preparation and budget profiles differ between these cases.

Quality protection requires complementary materials, induction and supervision. The experience documented in the 2003 universal-primary-education study, including instances where rapid expansion raised ratios and increased reliance on unqualified teachers, shows why access and workforce preparation must be costed together.[REF-07]

Four small schools have low ratios but each lacks reliable coverage during leave and cannot offer every curriculum area. A national efficiency screen might identify surplus. The appropriate analysis maps travel time, road seasonality, grade combinations, teacher skills, community access and the minimum adult presence required.

Clustering may permit a mobile specialist or shared professional development, but daily pooling is infeasible if travel consumes the teaching period. Closure or consolidation is not treated as a staffing calculation alone because distance can exclude children. The scenario compares service received, not only FTE saved.

A district reports an overall surplus of secondary teachers and simultaneous vacancies in mathematics and science. Person counts cannot resolve the conflict. Timetables are converted into subject periods, credentials are mapped under the match matrix, and feasible cross-school assignments are tested.

Responses may include targeted recruitment, retraining, timetable coordination or temporary supported assignments. Each has lead time and quality conditions. Redeploying an unmatched surplus into a vacancy changes the post count but not subject-matched capacity.

A system with stable pupil numbers faces a concentration of retirement eligibility over five years. The projection uses actual eligibility and observed retirement timing, not age alone. Candidate admissions are increased early enough to account for preparation duration and completion yield.

Mentoring capacity is protected because simultaneous departure of experienced staff can weaken induction. Phased retirement or return arrangements are modelled only if legally available and supported by evidence of uptake.

A region shows a high visit-based absence estimate. Before sanction or additional recruitment, the authority verifies leave, official duties, training schedules, payment delays, housing, travel and replacement. It also examines school closure and delivered periods.

If much absence is authorised training without replacement, the response concerns scheduling and reserve capacity. If assignment records are wrong, the response is reconciliation. If unexplained recurrent absence remains after due procedure, employment controls may apply. The same rate can therefore lead to different action according to cause.

Payroll FTE rises while school-assigned instructional FTE remains stable. Reconciliation tests central assignments, new leadership release, leave, delayed deployment, duplicates and expansion in non-school functions. The finding is not automatically payroll fraud or administrative waste.

Public reporting states how much of the difference is explained and what remains unresolved. Only validated instructional assignments enter the pupil–teacher ratio used for service planning.

The worked cases share no universal remedy. Each begins by selecting the unit appropriate to the decision and tracing the point at which recorded resources cease to be deployable service. This discipline prevents ratio targets from substituting for diagnosis and keeps administrative action proportionate to evidence.

The projection base is reconstructed before it is advanced. Opening pupil and teacher stocks refer to the same school year and education boundary. Late census returns, duplicate schools, teachers assigned outside the level and unresolved payroll records are identified. A base-year adjustment is documented rather than silently absorbed into the first projected change.

Pupil stocks are held by school, grade, sex and programme where the input supports that detail. Teacher stocks are held by person, FTE, assignment, subject or grade capability and employment status. Aggregation is delayed until the organisational constraints have been applied.

The pupil model advances each grade through promotion, repetition, dropout and transfer. New entrants are derived from the relevant age population and entry assumptions or from a separately estimated intake series. The sum of outcomes for a grade cohort is reconciled to the opening cohort, with migration and record corrections explicit.

The model can operate with fewer parameters where data are limited, but the simplification is disclosed. A gross-enrolment trend does not permit precise grade-flow projection. In that case, broader demand ranges are preferable to invented transition rates.

Projected pupils are converted to classes at the school-grade level. The upward-rounding rule is applied, and approved combined-grade arrangements are incorporated. For specialist teaching, classes are multiplied by curriculum periods by subject. The resulting timetable demand is checked against the total prescribed instructional period.

School opening and closure assumptions are separate inputs. A new school changes minimum staffing and distribution even if total enrolment remains unchanged. Consolidation changes travel and access and is not introduced solely as an arithmetic means to reduce FTE.

The teacher stock is advanced by age or service group, field, location and status where reliable. First entry, re-entry, retirement, resignation, mortality, dismissal, transfer and workload change have separate rates or counts. The model prevents a transfer between districts from becoming both a national exit and entry unless the national boundary genuinely changes.

New entrants are drawn from the preparation and recruitment supply route. Candidate cohorts complete after the applicable programme duration, and only the expected eligible and appointed share reaches the teacher stock. An unconstrained recruitment case may be shown to describe need, but the operational case cannot recruit more candidates than the supply route can supply.

Available national stock is allocated to school needs through a stated rule. Perfect allocation is not assumed. Historical vacancy duration, subject scarcity, geographic constraints and transfer limits inform the deployable share. The difference between total supply and feasible deployment is reported.

A redistribution case identifies donor and recipient locations only after minimum staffing, subject match and lawful movement are tested. The model records the year in which movement can occur. Immediate transfer at the base date is not assumed when consultation and notice are required.

Scheduled contact periods are adjusted for school opening, teacher availability and replacement. The sequence is defined to avoid applying the same loss twice. Where only one availability estimate exists, scenarios use an interval and do not distribute identical absence mechanically to every school.

Subject-matched capacity is calculated after availability because an unmatched substitute may preserve general supervision without delivering the required subject. Pupil-received hours additionally require pupil attendance; this lies outside the teacher-stock equation and is reported separately.

Where earlier data are available, the model is initialised in a past year and projected to an observed year. Differences are decomposed into demographic, enrolment, attrition, recruitment, deployment and data-revision components. Back-testing does not validate an unchanged future, but it reveals unstable assumptions and coding errors.

Every projection release records input vintage, model version, policy assumptions and correction history. A new release does not overwrite the prior scenario. Decision-makers can therefore distinguish a changed outlook from a changed model or revised baseline.

A regional profile translates the national workforce account into comparable subnational evidence while preserving differences in settlement, programme and responsibility. It is not a league table. Its purpose is to locate service constraints, determine whether they are widespread or concentrated and identify the authority able to respond.

The common core contains pupil and teacher FTE, trained status, vacancies, attrition, turnover, subject coverage and availability under nationally controlled definitions. The contextual module records remoteness, school size, language, conflict or displacement, shift operation and other conditions relevant to interpretation.

Comparability is strongest for the common core. Contextual variables explain why equal ratios do not necessarily imply equal provision. They are not used to excuse avoidable inequality without evidence.

Administrative regions may vary greatly in population and area. Publication therefore includes both administrative aggregates and school-distribution statistics. Boundary changes are versioned; historical values are recast only where the source permits a defensible mapping.

Teachers serving multiple regions are apportioned by workload or assigned to a distinct mobile category. Counting the full person in each region inflates the national total. Pupils are assigned to the school attended for service ratios and to residence for access analysis where both data exist.

The profile can compare rural and urban schools, poverty groups, language communities, girls' and boys' exposure, and other nationally relevant groups. The stratifier must be measured at the appropriate unit. A district poverty category is not treated as a household characteristic for every pupil.

Disaggregation is constrained by data quality and confidentiality. Small groups are combined or suppressed, and suppression does not erase the need for administrative response. Narrative findings can identify a protected service gap without publishing identifiable cells.

A typology may group regions facing similar planning conditions, such as rapid enrolment growth, remote small-school provision, subject mismatch or retirement pressure. Grouping is based on disclosed variables and does not replace the published underlying measures.

Typologies are reviewed because regions can change category. They are not institutional labels or permanent judgements. Their value lies in matching technical assistance and planning tools to a documented problem.

Extreme values are validated before publication. A very low ratio may result from a small denominator, closure, duplicated staff or justified minimum provision. A very high ratio may reflect missing teachers, double shifts or an enrolment surge. The validated value remains in the distribution; it is not removed merely because it is inconvenient.

The regional narrative begins with recorded facts, then states interpretation and limitation. It avoids adjectives of praise or blame. Where a region differs from the national result, the report identifies the magnitude, period and relevant context. Recommendations are assigned to the authority with the competence and resources to act.

The workforce finance model includes salary, allowances, employer contributions, preparation, recruitment, induction, professional development, substitute provision and material deployment supports where these are part of the staffing strategy. Capital costs such as housing or new training facilities are separated from recurrent expenditure.

The price year and currency basis are fixed. Nominal and real series are not mixed. Foreign-financed and domestically financed components are separately visible where this affects sustainability, without implying that one source changes the educational value of the input.

Salary cost is calculated by FTE, grade and step, not by multiplying all teachers by one average when the composition is expected to change. Entry cohorts begin at their applicable point; promotions, increments and retirement alter the distribution over time. Part-year appointment and vacancy timing are included.

Allowances are modelled by eligibility and expected payment, including rural or hardship provisions. An allowance that is legally due but historically delayed remains a liability; its absence from cash execution is not an efficiency saving.

Preparation cost includes places required to yield the necessary eligible completions. If the completion yield is 80 per cent, funding only the number of eventual appointments as entrants understates capacity. Practicum, instructor supply and facilities can constrain expansion even when funds are nominally available.

Candidate grants or living support may affect access to preparation and the regional composition of entrants. The model records these as policy components rather than burying them in an average training cost.

Advertising, selection, verification, initial travel, posting and induction have timing and administrative costs. Large recruitment exercises also require procedure capacity. Delayed appointment can leave funded posts vacant for part of the school year, producing apparent underspending and real service loss.

Mentoring creates workload for experienced teachers. Its cost can appear as release time or additional staffing rather than a cash payment. Excluding it assumes that induction has no effect on existing instructional capacity.

Housing, transport, hardship allowances and scheduled travel can be required to sustain remote service. The 1966 Recommendation provides an international basis for attention to accommodation, travel and hardship in remote areas. The finance model tests receipt and retention, not announcement alone.[REF-02]

Some supports serve several public services and require interdepartmental planning. The education budget records only its share where an agreed allocation exists; the staffing scenario nevertheless states the dependency.

Teacher supply cannot consume the entire recurrent envelope without regard to materials, maintenance, supervision and assessment. The 2003 universal-primary-education model used a distinct share for inputs other than teacher salaries and explained their role in quality and system management.[REF-07]

The present framework does not adopt a universal expenditure share. It requires the scenario to show what remains after workforce costs and whether the curriculum and professional support can operate within that amount.

Relevant results include total workforce recurrent cost, cost per teacher FTE, cost per pupil, salary share of education recurrent expenditure, cost of vacancies, and medium-term liability at the horizon. These measures require context; a high salary share can reflect underfunded non-salary inputs, large staff need or narrow accounting coverage.

Cost per delivered instructional hour can be estimated where time data are reliable. It is an efficiency diagnostic, not a wage judgement, and is not compared across systems without price, workload and curriculum context.

The financing gap equals the cost of the specified feasible service plan less credible domestic and external resources over the same period. Aspirational funds are not counted as available. A gap can be reduced by additional finance, changed sequencing or genuine efficiency improvement; it cannot be removed by omitting required costs.

The horizon-year recurrent obligation is compared with projected resources under stated economic and revenue assumptions. Sensitivity tests cover salary, exchange-rate exposure where relevant, attrition and enrolment. A plan that can be financed only during the initial project period is identified as transitional and requires a continuation decision.

The personnel system is event-based. Appointment, confirmation, workload change, qualification recognition, assignment, leave, return, promotion, transfer and exit are separate dated events. A current-status table may be derived from them, but it does not replace the history required for stock-and-flow analysis.

Every event contains a protected person identifier, event type, effective date, recording date, competent authority, source document and prior status. Corrections identify the field corrected and retain the superseded value under controlled access. Deletion is restricted to invalid duplicate records and remains auditable.

The post record contains post identifier, organisation, function, level or subject, authorised FTE, funding status, effective dates and authority. A post transferred between schools is closed at the origin and opened or reauthorised at the destination; its history is not overwritten.

Temporary project posts, substitutes and permanent establishment are separately coded. A person can be linked to more than one fractional post, but the combined workload control applies. A post without a person is a vacancy only if it is funded and available for filling under the stated definition.

The school census records each teacher expected to serve the school on the census date, protected identifier, function, workload at the school, grades and subjects, employment status and temporary absence. The school head certifies the return, and the district validates unresolved identity or assignment differences.

The module does not request data already held reliably in the personnel register merely for convenience. It collects school-level facts needed for reconciliation. Pre-populated fields can reduce burden, but schools must be able to report a factual discrepancy without altering the authoritative personnel record directly.

Each record describes a timetable cycle, day, period, duration, class or group, subject, assigned teacher and location. The cycle start and end dates permit term changes. Cancelled or replaced periods are held in a delivery record, not by editing the approved timetable after the event.

The system validates simultaneous assignments, missing teacher periods, teachers exceeding contact load and classes lacking required curriculum periods. A flagged conflict is returned to the school for correction or a documented exception.

Providers report programme identifier, recognition status, field, mode, expected duration, entrants by cohort, active students, completions, withdrawals and completion date. Candidate linkage uses protected identifiers under lawful arrangements. Aggregate publication distinguishes provider-reported completion from recognition and employment eligibility.

Programme capacity includes instructor FTE, practicum places and facilities, since enrolled places alone can overstate sustainable expansion. Changes to programme duration or requirements receive an effective date.

The field form records the roster expected at the visit time, observed status, scheduled class, whether the class is operating, replacement and immediate evidence. It does not ask the observer to judge whether absence is lawful. Verification fields are completed later by authorised personnel.

School closure, inaccessible visit and missing timetable are distinct outcomes. They remain in the field dataset and are handled under the sampling protocol. Free-text notes support resolution but are not substituted for controlled categories in aggregate statistics.

Transfers use a documented structure, encoding and reference-code version. Each submission has reporting period, producing authority, creation date, record count and checksum. Rejected records return a specific error; valid records are not discarded because other records fail.

Validation occurs at entry, consolidation and pre-publication. Entry checks format and code validity. Consolidation checks cross-source identity and timing. Pre-publication checks totals, revisions, tables and disclosure risk.

Personal event history is retained according to lawful administrative and statistical requirements. Public analytical files remove direct identifiers and limit combinations that permit re-identification. Access roles distinguish payroll operation, personnel administration, statistical analysis and public release.

The right to correct an inaccurate personnel record is separate from alteration of a published aggregate. When correction materially changes an aggregate, the statistical release follows the revision policy.

Measurement risk arises when a statistic does not represent the intended condition. Examples include headcount used as labour input, payroll used as presence, trained status used as effectiveness and ratio used as class size. The control is conceptual before it is technical: the indicator name, unit and permissible claim are fixed before collection.

A binding ratio or vacancy target can change recording behaviour. Posts may be reclassified, pupils shifted between dates, teachers temporarily reassigned for the census or unknown qualifications coded as compliant. Independent reconciliation and multiple measures reduce this risk.

Targets are therefore accompanied by counter-indicators. A drive to lower vacancies is checked against arrival and retention; a drive to increase trained share is checked against unknown status and subject match; a drive to increase contact time is checked against preparation and lawful workload.

Mechanical equalisation can remove teachers from small or disadvantaged schools because their ratios appear low. The safeguard is a service-based entitlement with minimums and contextual evidence. Redistribution is tested for access, curriculum and continuity before implementation.

Administrative data can affect pay, career and reputation. Statistical anomalies are not disciplinary findings. Individuals receive applicable notice, evidence and review before adverse employment action. Aggregate reporting does not name teachers in absence or qualification tables.

Professional consultation is required where staffing rules alter workload or conditions. The 1966 Recommendation provides a basis for participation and objective assessment within the education service.[REF-02]

Teacher records contain personal, financial and potentially sensitive health or location information. Collection is limited to necessary fields. Health status is not gathered for routine workforce planning when aggregate mortality or absence evidence is sufficient.

Access logs, separation of duties and secure transfer protect records. Public tables use suppression, aggregation or controlled rounding where small cells could identify a person. Security-sensitive school locations can be aggregated during conflict.

Projection and sample uncertainty are reported, but administrative incompleteness is not disguised as a confidence interval. Sampling error, model range, classification uncertainty and coverage gap are separate limitations. A narrow model interval does not compensate for missing private-school teachers or unknown assignments.

An association between ratios and completion in a selected sample does not establish that changing the ratio alone causes completion. A credential share does not establish classroom practice. The safeguard is a claim-evidence matrix reviewed before release.

Every intervention is tested against five questions: Is the service problem verified? Is the proposed action connected to its cause? Is a less disruptive effective action available? Are impacts on underserved pupils and teachers assessed? Is the action reviewable against a dated result?

The test prevents broad transfer, surveillance or employment measures from being justified by weak aggregates. It also prevents data limitations from indefinitely delaying a necessary response where the service failure is directly observed.

Residual risks are recorded in the release. A publication can pass statistical preparation while noting that a material share of credentials remains unverified or that presence estimates exclude inaccessible areas. The limitation appears with the affected finding and in the technical register.

Implementation begins with a twelve-domain assessment: legal authority, establishment, payroll, personnel identity, school list, assignment, qualification, timetable, training supply route, projection, finance and publication. Each domain is classified by evidence available, decision prevented and feasible next improvement.

The assessment avoids a maturity score. A single score would allow strength in payroll to offset absence of assignment data. The domains remain visible because each can block a different decision.

The first cycle establishes the common reference date, consolidates school and person codes, reconciles active payroll with personnel and school records, and publishes a bounded baseline. It need not await a complete historical series. Unknown status and unresolved exceptions are included as results.

Priority is given to errors that change service decisions: duplicate active persons, teachers with no location, schools with no valid staff record, funded vacancies and subject gaps. Less consequential descriptive fields can follow.

Once the baseline closes, pupil and teacher flows are projected under central and sensitivity cases. The first projection may use broad attrition or training-yield ranges if detailed cohorts are unavailable. The range narrows only as evidence improves.

The workforce and finance teams produce one reconciled scenario. Separate projections that use different teacher counts, salary bases or school years are not combined in a final total.

School entitlements are calculated and returned for factual review. The authority then classifies shortage, justified minimum, feasible surplus and mismatch. Movement, recruitment and support actions are scheduled with effective dates and responsibilities.

Implementation monitoring records appointment, arrival, timetable assignment and retention. A transfer order or signed contract is an intermediate event, not the service result.

The release contains the baseline date, revision status, core indicators, distribution, scenario assumptions and limitations. Tables are generated from the same controlled data used for the narrative. A correction protocol remains open after publication.

Capability requirements include personnel administration, education statistics, demography, finance, timetabling, data protection and field verification. Training is directed to actual functions. A central statistical course does not equip district officers to resolve transfer and assignment transactions unless those tasks are included.

Documentation is an operational asset. Definitions, code lists, validation rules and recurring procedures are maintained so that the system does not depend on one individual.

Implementation is monitored through products rather than training attendance alone. Relevant indicators include the share of active persons with valid assignment, the proportion of post and payroll exceptions resolved within the period, reporting timeliness, credential verification coverage, projection residuals and publication correction time.

These indicators assess the administrative system, not individual teacher performance. They are owned by the responsible office and accompanied by the resources needed to improve them.

After the initial reconciliation, events are procedure continuously and snapshots produced on a fixed schedule. Annual clean-up cannot substitute for timely transaction control. Code changes, school openings and policy revisions enter through governed procedures.

Independent periodic review tests whether published measures still answer the planning questions and whether target pressure has changed recording behaviour. Measures that no longer serve a decision are revised or retired with a documented series break.

The purpose of capability development is not to create the largest personnel database. It is to produce timely, limited and trustworthy evidence that connects public expenditure to instruction. The implementation programme is successful when vacancies, deployment gaps and lost time can be identified early enough for a responsible authority to act and when the public account remains accurate about what the data do not establish.

An indicator cannot be separated from its metadata. Name, purpose, population, unit, numerator, denominator, reference period, source, collection method, coverage, calculation, revision and limitation are held in one controlled record. A change in any defining element receives a new version.

The public note may be shorter than the internal record, but it contains every fact needed to prevent a materially false interpretation. Detail that affects only procedure can remain internal; detail that affects meaning cannot.

The teacher-headcount record states whether the population includes public and non-government schools, temporary staff, teachers on leave, central staff and teachers with multiple assignments. It defines the reference date and the rule for active employment.

The calculation deduplicates persons within the defined system. If cross-sector identity cannot be reconciled, sector totals are published without claiming that their sum is a unique-person total. A teacher headcount is not described as available or classroom teaching unless assignment and function are part of the measure.

The FTE record defines the full-time workload, data used for fractions, treatment of overtime, multiple posts, unpaid leave and mixed functions. Contracted FTE, assigned teaching FTE and instructional FTE have separate indicator codes.

Where national rules differ by level or occupation, the conversion is performed within each rule and then aggregated. One universal divisor is not applied to incompatible workloads. Comparisons identify whether pupil FTE is also used.

The numerator is enrolled pupils or pupil FTE at a stated level and date; the denominator is teachers or teacher FTE under a stated instructional boundary. The record specifies sector coverage, assignment treatment and whether teachers are counted regardless of actual teaching assignment, as in the definition reported in *World Development Indicators 2004*.[REF-05]

The limitations field states that the ratio is not class size, does not show subject match or presence and may be affected by enrolment error and hours taught. If a time series crosses the ISCED 1976 to ISCED 1997 transition, the break is disclosed.[REF-05] [REF-10]

The record quotes or precisely summarises the applicable national minimum organised preparation rule and identifies its version. The numerator includes only teachers verified as meeting it. The denominator for the percentage and the full-stock unknown share are both reported.

The limitations state that preparation content and duration vary across countries and that the measure does not include all competence acquired through experience or self-instruction. This preserves the caution set out in the 2004 international source.[REF-05]

A vacancy record states whether the denominator is authorised posts, funded posts or posts open for recruitment. Filled-but-unoccupied, temporarily covered and frozen posts are distinct statuses. The reference date and duration bands are included.

The measure does not by itself establish teacher shortage, because the establishment may be outdated. Comparison with calculated service entitlement provides the planning interpretation.

The defined workforce boundary determines exit. The metadata lists included reasons, treatment of transfers, denominator, time at risk and late-recorded events. Retirement eligibility and actual retirement are separate measures.

Where an average stock denominator is used, its construction is stated. Rates by subject or region identify movement out of that group even when the person remains in the national system; these are labelled group exits, not national attrition.

The presence record states the target population, expected roster rule, observation method, visit window, number of visits, achieved sample, weights, treatment of closure and verification of reasons. It reports uncertainty and non-response.

The indicator name is “observed presence” or “observed absence” when both authorised and unauthorised reasons are included. “Unauthorised absence” is reserved for verified cases under the governing rule.

The record defines scheduled contact time, school-open time, teacher availability, replacement and delivered period. It explains whether capacity is estimated from separate factors or directly from period records. The same loss is not applied more than once.

The limitation states that delivered time is not evidence of pedagogical quality or learning. Subject-matched capacity has an additional dated match rule.

Indicator owners review definitions before each cycle. A change request states why the current measure no longer serves its purpose, expected series effect and recoding feasibility. Statistical and policy users are informed before release where a break is material.

The decision record begins with gross recruitment need by year, field and location. It subtracts eligible candidates already in the supply route, credible re-entrants and feasible redistribution. It then applies completion and appointment yields to determine admissions.

Preparation capacity is checked for instructors, practicum and facilities. The decision states whether additional intake is permanent or temporary. A temporary bulge has an exit strategy so that institutions and candidates are not left with unsupported capacity after demand falls.

New posts are justified by service entitlement, not vacancy alone. Evidence includes projected pupils, curriculum, class or period formation, existing assignments, feasible redistribution and duration of need. Funding covers salary and associated costs over the planning horizon.

A post created for a new school is linked to the school's opening and facility readiness. If the school does not open, the appointment and deployment plan is revised rather than leaving a teacher notionally assigned to a non-operating location.

The record identifies the apparent excess, service minimum, subject capability, legal procedure, receiving vacancy, travel or relocation, consultation and implementation date. It estimates both donor-school and recipient-school effects.

Redeployment is not recorded as complete until the teacher reports and appears on the receiving timetable. Non-arrival, appeal or delayed movement remains visible. The staffing account avoids counting the same FTE at both schools during the transition.

Evidence includes vacancy duration, turnover, applicant rates, housing, safety, payment, transport and professional support. The decision identifies which constraint the measure is intended to change. A general allowance may be poorly targeted if housing or delayed pay is the binding problem.

Evaluation compares eligible and relevant non-eligible posts over a stated period, while recognising selection differences. Receipt and timeliness are measured. Retention without presence or service continuity is insufficient evidence of success.

The record defines the urgent service gap, duration, eligibility, selection, preparation, supervision, employment conditions and transition to a regular arrangement. It considers whether the temporary route draws candidates away from full preparation.

Temporary appointments are monitored by location and population served so that lower standards or insecure conditions are not concentrated among disadvantaged pupils without explicit justification and corrective plan.

The decision begins from curriculum periods, current workload, preparation and assessment duties, class size, non-teaching requirements and applicable employment arrangements. It estimates staffing and service effects, not only salary savings.

Increasing contact time may reduce FTE demand arithmetically but can weaken preparation or professional work if total workload is unchanged. Reducing contact time may support quality but requires additional staff. The policy record states the evidence and trade-off.

The analysis maps pupils, distance, travel time, safety, seasonal access, language, disability access, school quality, teacher continuity and cost. Staffing efficiency is one component. A consolidation that makes attendance impracticable does not improve educational access.

Alternatives such as supported multi-grade teaching, cluster services and transport are compared. Community evidence is included, but the decision remains subject to the child's educational opportunity and public responsibility.

Publication is justified by a public decision need and tested for accuracy, confidentiality and foreseeable harm. A school-level ratio may be suitable; individual absence or qualification details ordinarily are not. Small cells and exceptional circumstances require protection.

The release provides correction and contextual notes. School publication does not substitute for the authority's duty to correct a known shortage.

The record distinguishes an analytical recommendation from an authorised decision. It records departure from the evidence with the authority and reason. This protects the statistical function from being recast after the event and permits public assessment of whether implementation matched the stated basis.

The 2005 EFA Global Monitoring Report was published in 2004 and is therefore available at the evidence cut-off. It frames quality through learner characteristics, context, enabling inputs, teaching and learning, and outcomes. Its teacher evidence supports the proposition that numbers, preparation, conditions and instructional procedure are connected, but it does not create a single teacher-quality index.[REF-03]

This report uses that framework to bound claims. Workforce statistics describe enabling capacity and part of the procedure; they do not replace learning evidence. Where the EFA report presents country or regional observations, later revisions are not imported.

The *Global Education Digest 2004* provides internationally compiled education data and the statistical context for programme comparison. Its role here is definition and coverage rather than creation of a new country ranking. National education structures are mapped through ISCED 1997, with national detail retained.[REF-04] [REF-10]

Publication year does not mean every observation refers to 2004. Reference years are taken from the table and note. A report using these data in March 2005 therefore describes the latest reported observation, not the current school condition without qualification.

The absence figures are from unannounced visits and were presented as preliminary. The source defines the rate as the percentage of staff supposed to be present but not present, including excused and unexcused absence.[REF-06]

This report accordingly uses “observed absence” and requires reason verification. It does not extend the figures to countries not observed, to later years or to individual misconduct. The India figure is identified as an average for 14 states, not a national census.

The wider service-delivery framework in the source concerns accountability relationships among citizens, policymakers and providers. This report uses the narrower measurement implication: financed employment must be connected to service received. It does not adopt every institutional prescription for every country.

The 2003 study analyses primary completion and financing in low-income countries. It reports wide ranges in service-delivery parameters and constructs simulations around country reform paths. Its 40:1 pupil–teacher ratio is a model parameter associated with the observed average of a relatively higher-completion group, not a universal standard.[REF-07]

The report also models teacher salaries, repetition, non-teacher recurrent inputs, public finance, private enrolment and HIV/AIDS costs. Selective citation of 40:1 without those conditions would misrepresent the source. The present study uses the value only to demonstrate scenario practice and benchmark limits.

The underlying study often uses ratios as proxies and acknowledges geographic constraints. It does not supply school-level entitlement rules. National planners must therefore move from aggregate scenario to school and timetable calculation.

The OECD publication defines student–teacher ratios using full-time equivalents and distinguishes them from average class size. It reports instruction time, teaching time, working-time organisation, salaries and decision structures for participating systems.[REF-08]

Its country coverage and institutional context differ from the low-income-country studies. Values are not combined into one global distribution without harmonisation. The conceptual distinction between ratio, class size and time is transferable; the numerical results are not assumed to apply outside the reported population.

The 1966 Recommendation concerns preparation, professional status, conditions, participation and remote service. It guides the normative context but does not prescribe a global staffing ratio.[REF-02]

The Convention on the Rights of the Child informs attention to non-discrimination, access, attendance and educational purpose. It does not convert every workforce indicator into a direct legal compliance test without analysis of the applicable obligation and facts.[REF-12]

The January 2005 report provides the immediate development-planning context for costed national strategies. Its use in this report is to support integrated and sequenced planning. It is not used as evidence for workforce outcomes that had not yet occurred by the cut-off.[REF-09]

Definitions and official values are taken from the issuing official source. Analytical propositions are attributed to their study and retain scope. ICEQC calculations are labelled as illustrations or calculations based on cited data. No later source is used to revise the contemporaneous position.

A position established by the competent authority within the education workforce structure. Authorisation does not establish that funding is available, a person is appointed or service is delivered.

An authorised position with budget provision for the stated period. A funded post may be vacant, frozen, temporarily covered or filled.

A post to which an appointment is active. The term does not establish that the appointee has reported, is assigned to a school, is present or is teaching.

The number of unique persons meeting the stated teacher and active-status definition at the reference date. Headcount gives equal weight to full-time and part-time persons.

The sum of teacher workloads expressed relative to the nationally defined full-time workload. Contracted, assigned and instructional FTE are distinct.

Personnel directly involved in instruction under the applicable statistical definition. The treatment of principals, specialists and mixed functions must be stated.

For the international indicator used here, a teacher who has received the minimum organised teacher training required nationally, whether before or during service. The term is nationally defined and does not establish common preparation across countries.[REF-05]

A teacher meeting the applicable recognised professional qualification rule. This may coincide with trained status but is retained separately where national law or data distinguish them.

The relationship between recognised preparation and assigned curriculum under a dated national match rule. It is calculated by period where assignments are mixed.

Pupils divided by teachers within a stated level, sector, period and unit. It is a resource ratio and is not average class size.

Pupils divided by teachers meeting the stated national preparation requirement. It combines resource quantity and nationally defined status but does not measure teaching quality.

The number of pupils in a class or teaching group during a defined instructional arrangement. A pupil can belong to different groups by subject; a teacher can serve more than one class.

Scheduled time during which a teacher is responsible for direct instruction. It is part of total working time and differs from pupil instruction time where schedules or groups vary.

Delivered contact hours divided by the annual full-time contact-hour rule, subject to the stated calendar and assignment. It measures delivered time capacity, not pedagogy or learning.

A post meeting the specified authorisation and funding conditions that lacks a filled or serving FTE at the reference date. The vacancy definition states whether posts temporarily covered or not open for recruitment are included.

The gap between the teacher capacity required to deliver the defined service and feasible available capacity. It depends on entitlement, capability, deployment and time, not vacancies alone.

Movement of a person into the defined teacher workforce during the period. First entry, re-entry and transfer in are separately classified.

Exit from the defined workforce during the period. Retirement, resignation, death, dismissal and other reasons are separated; internal transfers are treated according to the boundary.

Movement into and out of a school, district or group. It can be high when national attrition is low because internal transfers contribute.

The assignment and distribution of teacher capacity among schools, locations, levels, subjects and functions. A deployment order is complete for service measurement only after effective arrival and timetable assignment.

The observed or validly recorded condition of an expected teacher being at the required location and time. It does not by itself show that instruction occurred.

The condition of an expected teacher not observed present under a stated visit design. It can include authorised and unauthorised reasons until verification.

Absence determined under the applicable rule to lack authorisation after verification and due procedure. It is not inferred solely from an unannounced visit.

The portion of intended calendar time during which the school is operating and available to pupils. It is distinct from teacher presence and pupil attendance.

A scheduled instructional period that occurred under the stated delivery rule. Replacement, combined-class and shortened-period arrangements are separately coded.

The volume and composition of instruction that available, appropriately assigned staff can deliver under the curriculum and timetable. It is an enabling condition and not a learning outcome.

Recruitment or return required to offset expected exits and workload reduction. It exists even where enrolment and target workforce stock are stable.

Additional capacity required because pupil numbers, programmes, curriculum, access conditions or staffing policy expand.

Growth demand plus replacement demand, less credible re-entry and transfers in, with the same boundary and period.

Recognised completions from a defined entrant cohort divided by entrants to that cohort. Annual completions divided by annual admissions is not a cohort yield when programme duration exceeds one year.

Eligible selected candidates who accept appointment and report for duty divided by the relevant eligible candidate population. The precise stage used as denominator is stated.

The teacher FTE calculated under the approved staffing rule for a school or programme. It is distinct from inherited authorised posts and is subject to review of the rule itself.

The smallest staff configuration capable of maintaining the defined service and protection requirements in a particular setting. It can produce a low ratio in small schools without implying avoidable surplus.

The difference between the observed closing stock and the stock derived from verified opening stock and events. It indicates unresolved data or boundary change and is not labelled recruitment or attrition.

A discontinuity caused by definition, classification, coverage, source or method change. It is distinguished from an actual change in workforce or service.

A reference value used to test implications under stated evidence and context. It is not automatically a standard, target or causal threshold.

A conditional calculation under an internally coherent set of demographic, educational, workforce and financial assumptions. It is not a prediction independent of those assumptions.

The population, sector, geography and records represented in an indicator. Coverage gaps remain explicit and can differ across numerator and denominator.

A case for which the required evidence is unavailable or unresolved. Unknown is neither compliant nor non-compliant and remains visible in denominators and limitations.

A change to a published statistic resulting from corrected data, updated source or method. Revision date, reason and series effect are disclosed.

Consider a school reporting 18 persons as teachers. Twelve work full time in classroom teaching. Two work half time. One principal teaches 40 per cent of the full-time classroom load. One full-time teacher is on long authorised leave with a separately appointed replacement. One person has transferred but remains on the school list, and one visiting teacher is counted both by this school and the home school.

The reported headcount of 18 is not a valid instructional denominator. After removing the transferred duplicate, retaining unique persons, allocating the visitor by workload and avoiding double count between leave holder and replacement, the employment, assignment and instructional measures differ. The exact result depends on the applicable treatment of leave and the visiting fraction; the calculation record must show those choices.

The lesson is not that FTE always produces a smaller figure. A school may have several part-time assignments from staff based elsewhere, increasing assignment headcount while producing a modest FTE. The purpose of reconciliation is to select the unit that matches the question.

Suppose three schools enrol 40, 160 and 800 pupils and have 2, 5 and 20 teacher FTE. Their school ratios are 20:1, 32:1 and 40:1. The unweighted mean school ratio is approximately 30.7:1. The national aggregate ratio is 1,000 pupils divided by 27 FTE, or approximately 37.0:1. The results differ because the aggregate weights schools by pupils and teachers through their totals.

Neither statistic is incorrect. The aggregate answers how total pupils relate to total resources. The mean describes the average of three school ratios, giving the 40-pupil school the same influence as the 800-pupil school. A public table must name the statistic and preferably show the distribution.

Assume four schools each have 41 pupils in a particular grade and the planning rule permits no more than 40 pupils in a class. Each school requires two classes, for a total of eight. Aggregating first gives 164 pupils divided by 40, or 4.1, which rounds upward to five. National-first calculation would understate the requirement by three classes because the classes cannot be pooled across schools.

If approved two-grade multi-grade groups are feasible in two of the schools, demand may change, but that is an organisational assumption requiring curriculum and preparation evidence. It is not a mathematical correction to the class-size rule.

A lower-secondary school has ten pupil groups, each requiring five mathematics periods per week: 50 group-periods. A full-time mathematics teacher has an 18-period contact load. The arithmetic requirement is 2.78 FTE. Because timetables require actual people and periods, three appropriately matched full-time teachers may be necessary unless a lawful fractional appointment or cross-school assignment is feasible.

If the school employs three mathematics-qualified persons but one has only half of the timetable available to mathematics, the matched capacity is 2.5 FTE, not a headcount of three. The uncovered periods remain visible even if overall school teacher headcount exceeds entitlement.

A district reports 1,000 teacher FTE. Seven hundred and twenty are verified as meeting the national preparation rule, 80 are verified not to meet it, and 200 have unresolved records. The trained share among known cases is 720 divided by 800, or 90 per cent. Verified trained FTE represent 72 per cent of the full stock, and the unknown share is 20 per cent.

Publishing only “90 per cent trained” invites the reader to apply the result to the full workforce. The proper presentation gives the known-case percentage, verified full-stock proportion and unknown share together. Unknown cases are then prioritised for verification rather than converted to either category.

A national system opens with 10,000 teachers. During the year, 300 retire, 100 resign, 50 die, 30 are dismissed and 20 leave for other recorded reasons. Five hundred teachers transfer between districts. National attrition is 500 exits divided by 10,000, or 5 per cent. The 500 internal transfers do not change national stock.

For the affected districts and schools, those transfers contribute to turnover. If all movements were added to national attrition, replacement demand would be overstated because each transfer in is also an available teacher elsewhere in the national system. If transfers were ignored locally, staffing instability would be understated.

If 2,000 candidates enter a two-year preparation programme and 1,600 complete and meet the recognised requirement, cohort training yield is 80 per cent. If 1,360 of those completers apply and are eligible, 1,200 accept appointment and 1,140 report for duty, the report-to-duty yield from recognised completers is 71.25 per cent.

To secure 2,000 reporting entrants under the same combined yield, more than 3,500 admissions would be required, subject to rounding and the stability of each transition. The correct planning response may also include improving completion, appointment timing or acceptance rather than expanding admissions alone.

A weighted survey observes 2,500 expected teacher opportunities and estimates 2,050 present and 450 absent. The observed absence rate is 18 per cent. Verification classifies 180 absent opportunities as authorised leave, 90 as official duty or training, 120 as unauthorised and 60 as unresolved.

The rate of all-reason observed absence remains 18 per cent. The unauthorised share among all expected opportunities is 4.8 per cent under the verified classification; the unresolved share is 2.4 per cent. It would be incorrect to describe all 18 per cent as unauthorised, and it would also be incorrect to exclude authorised absence when estimating immediate replacement need.

The sampling interval accompanies each estimate. The weighted counts presented here are illustrative and do not carry a calculated interval because no sample design is specified.

A system schedules 10 million teacher-periods. School closures cancel 400,000. Among the 9.6 million periods in open schools, teacher unavailability leaves 700,000 uncovered; replacement staff deliver 300,000 of those. Of the remaining scheduled periods with an available teacher, 100,000 are not delivered for other recorded reasons.

Delivered periods are therefore 9.1 million if the replacement periods are included and each loss is counted once. The delivered-period rate is 91 per cent of intended periods, while the rate conditional on schools being open is approximately 94.8 per cent. The two denominators answer different questions and are both labelled.

A district has 1,100 authorised posts, 1,050 funded posts and 1,000 filled FTE. Fifty funded posts are vacant, giving a funded-vacancy rate of about 4.8 per cent. A current school-based demand calculation, however, indicates an entitlement of 1,080 FTE. Relative to current service need, the filled gap is 80 FTE, not 50.

The additional 30 may reflect an establishment that has not kept pace with demand. Conversely, if entitlement were 970 FTE, a vacancy rate would still exist within the funded establishment while the district held an apparent aggregate surplus relative to the current rule. Both comparisons are required before action.

Two regions can share a 35:1 aggregate ratio. In one, most schools lie between 32:1 and 38:1. In the other, remote schools operate at 15:1 while large schools reach 60:1. The first has a relatively concentrated distribution; the second has marked dispersion and likely organisational constraints.

An equalisation policy would be premature without testing minimum schools and feasibility. The relevant finding is not merely that the second region is unequal. It is that pupils encounter very different staffing conditions and the reasons must be separated into justified minimum provision, avoidable concentration and data error.

Suppose 10,000 teacher FTE have an average base salary of 3,000 currency units, and employer contributions and allowances add 20 per cent. Annual employment cost is 36 million units before preparation, substitution and administration. If the workforce grows by 5 per cent and the salary scale rises by 4 per cent, the next-year employment cost is not 36 million plus 9 per cent when grade composition, part-year recruitment and allowances change; a grade-level calculation is required.

The simple approximation—39.312 million—can serve as a check. Material deviation from the detailed calculation is explained by timing or composition rather than hidden in a total.

A central scenario requires 15,000 deployable FTE. If projected enrolment is 3 per cent higher, attrition rises from 5 to 7 per cent, and appointment yield falls from 85 to 75 per cent, candidate requirement can rise by much more than 3 per cent because growth, replacement and supply route conversion compound.

Sensitivity is calculated sequentially and jointly. Reporting only one-factor changes can understate combined implementation pressure. The joint case must remain plausible; it is not constructed by combining every extreme regardless of coherence.

Before a result enters the narrative, the drafter applies four tests. The **unit test** asks whether the noun in the sentence matches the statistical unit. The **boundary test** asks whether level, sector and geography match. The **time test** asks whether the date supports a current, trend or projection statement. The **claim test** asks whether the measure can support the verb and conclusion.

Thus “the district has 500 teachers” fails if the value is post FTE; “class size fell” fails if only a pupil–teacher ratio changed; “absence is misconduct” fails if reasons are unverified; and “teacher quality improved” fails if only trained status rose. Correction is made in the sentence, not deferred to a distant footnote.

All quantities in this appendix are hypothetical. They demonstrate calculation boundaries and failure modes; they do not describe a country, region or school. Their value lies in making assumptions inspectable. A national application replaces every illustrative value with verified evidence and retains the same controls.

The decision review is used before a workforce plan is approved or materially revised. It asks whether the proposed stock, recruitment, deployment and financing decisions can produce the stated instructional service. It does not assess individual teachers and does not convert statistical exceptions into employment findings.

The review confirms the education level, sector, geography, pupil population, curriculum and horizon. Teacher headcount, FTE, posts and instructional periods are aligned to that boundary. A national balance is not accepted as evidence of district or subject sufficiency.

Opening stock is reconciled across personnel, establishment, payroll and school assignment. Unresolved duplicates, unassigned persons and unknown qualifications remain visible. A material residual prevents precise projection and is carried as a sensitivity or corrected before decision.

Pupil projections are translated through school, class and curriculum organisation. Upward rounding, minimum schools, subject periods, contact time and travel constraints are included. The review rejects national-first arithmetic that cannot produce a feasible school timetable.

Recruitment need separates growth and replacement. Training admissions, recognised completion, appointment acceptance and report-to-duty yields are not merged. The plan cannot appoint more suitable candidates than the supply route supplies without identifying an interim route and its safeguards.

Apparent surplus is tested against minimum staffing, subject match, workload, location and lawful movement. Transfers have realistic dates and conditions. A person is not counted at the receiving school until assignment becomes effective and service is confirmed.

The plan distinguishes intended calendar, school opening, teacher availability and delivered periods. Replacement provision is not counted twice. Presence evidence retains its observation design and reason limitations.

Salary, allowances, contributions, preparation, recruitment, induction, substitute capacity and deployment supports are costed over the horizon. Complementary materials and management are protected from unexamined salary crowding. Recurrent obligation remains visible after transitional finance ends.

The review examines rural and remote service, subject and language gaps, sex distribution, accessibility, turnover and groups exposed to weak provision. It also examines teacher conditions, consultation, privacy and due procedure. An efficiency measure that removes minimum service fails the review.

The competent authority approves, revises, phases or rejects the plan with reasons. Approval identifies critical assumptions and the date on which actual recruitment, deployment, cost and instructional coverage will be compared with the plan. A material assumption failure reopens the relevant decision rather than being concealed in an aggregate variance.

An early-warning system is not a prediction that a vacancy will certainly occur. It is an ordered body of evidence showing that the probability, expected scale or likely duration of an interruption has changed enough to justify verification or preventive action. The distinction matters because education authorities operate under uncertainty. Retirement eligibility is known more precisely than actual retirement; an application for transfer is more informative than a stated preference; a delayed salary payment may precede departure without causing every affected teacher to leave. Each signal therefore requires a defined decision use, a review interval and a named authority able to respond.

The warning chain begins with exposure, moves through pressure and intention, and ends with an observed workforce event. Exposure indicators include age structure, remoteness, housing scarcity, insecurity, high living costs, disease burden and dependence on a narrow preparation supply route. Pressure indicators include salary arrears, repeated excess workload, unfilled leadership posts, deteriorating housing, exceptional absence and unsuccessful recruitment rounds. Intention indicators include applications for transfer, non-renewal notices, retirement inquiries and withdrawal after appointment. Event indicators include resignation, retirement, death, dismissal, transfer, failure to report and a funded post remaining unfilled after its recruitment window. Authorities should not collapse these stages into one attrition rate.

A useful signal must change a possible decision. A rising share of teachers eligible to retire may justify preparation-intake review where pupil demand is stable or growing. It should not automatically trigger recruitment if enrolment is falling, subject requirements are changing or eligible staff commonly remain. Repeated vacancy advertisements with no qualified applicants may justify attention to pay, housing, preparation supply or appointment delay. They do not establish which response will work. The warning record should state the observed measure, its threshold or comparison, the mechanism requiring investigation, the response available and the cost of a false alarm or missed warning. This is the discipline that converts workforce statistics into service protection.[REF-13] [REF-14]

The annual attrition rate should use a clearly bounded opening workforce. Its numerator comprises members of that workforce who, by the end of the observation interval, are no longer active in the defined teaching service. Its denominator is normally the opening active workforce, not the closing workforce and not all persons ever appearing on the payroll during the year. Entries during the year belong in the stock-and-flow account but do not belong in the opening cohort. Where records cannot identify a unique person across years, apparent attrition may be produced by changed identifiers, delayed returns or reassignment between administrative units.

At least four rates may be needed. Total service attrition records departures from the defined education workforce. Occupational attrition records departure from teaching even when the person remains employed by the ministry. School attrition records departure from a particular school and therefore includes transfers within the system. Contract attrition records non-renewal or non-return among a specified temporary cohort. These measures answer different questions. A national service may retain nearly all teachers while schools serving remote or disadvantaged communities experience repeated loss. Conversely, high school mobility may partly reflect planned promotion or correction of an earlier maldistribution.

Reason coding is essential. Retirement, death, long-term incapacity, resignation for another occupation, international migration, domestic transfer, promotion to non-teaching work, dismissal, contract expiry and unknown departure should remain separate. Planned and unplanned losses impose different lead times. Preventable and unavoidable are evaluative categories requiring evidence, not codes to be assigned from assumption. Unknown status must be published because a falling unknown share can raise recorded attrition without any change in actual behaviour.

Cohort survival complements the annual rate. Authorities should follow new appointees from acceptance through reporting, completion of the first term, completion of induction and continued service after one, three and five years where records permit. Loss before reporting points to a different institutional failure from loss after several years. Comparable reporting requires the service boundary, employment categories, treatment of leave, observation dates, matching procedure and revision policy. OECD comparative work available by the cutoff reinforces the importance of distinguishing policies for the whole profession from targeted measures for particular teachers and schools.[REF-14]

Vacancy is meaningful only in relation to a post system. The authorised-post vacancy rate compares unfilled authorised full-time-equivalent posts with all authorised posts. The funded-post vacancy rate narrows the denominator to positions for which resources have been released. A recruitment vacancy may be defined still more narrowly as a funded post approved for recruitment and unfilled after the normal appointment interval. These indicators should be published together where differences are material. Otherwise a fiscal restriction can make the vacancy rate appear to improve merely because needed posts were never funded.

A service-gap measure is also required. It compares filled, appropriately matched instructional capacity with capacity implied by current enrolment, curriculum, class-formation and minimum-school rules. The service gap may exceed recorded vacancies when establishments have not been updated for population growth, new grades, language provision or small-school requirements. It may be smaller than authorised vacancies where enrolment has declined or duplicate posts remain. Neither result licenses automatic abolition or appointment: it triggers reconciliation of legal entitlement, educational demand and finance.

Duration changes the seriousness of a vacancy. Authorities should distinguish posts vacant at the reference date, vacancy-days accumulated during the term, and repeated vacancy episodes affecting the same school or subject. A post filled on census day after being empty for most of the year should not be treated as uninterrupted provision. A short vacancy covered by an appropriately prepared substitute differs from one covered by merged classes, cancelled subjects or unqualified assignment. The early-warning profile therefore records opening date, recruitment steps, applicant count, qualified-applicant count, offers, acceptances, reporting, interim coverage and final disposition.

The most consequential hidden vacancy is an obligation without an authorised post. Evidence may appear as classes without a teacher, systematic multi-grade combinations exceeding policy, subjects omitted from the timetable, teaching by personnel outside their preparation, or learners refused entry because staff are unavailable. These conditions should be counted as service deficits and investigated without pretending that they are internationally comparable vacancies. International comparison can align concepts and disclose differences; it cannot make different establishment laws identical.[REF-13] [REF-15]

The opening dashboard should contain a small set of linked measures rather than an undifferentiated collection. Stock measures comprise active teacher headcount, active full-time equivalent, filled instructional post equivalent, trained status under the national definition, and subject or grade match. Flow measures comprise entrants, returns, retirements, other exits, transfers in and out, and changes of teaching function. Vacancy measures comprise authorised, funded and recruitment vacancies, their duration and qualified-applicant yield. Exposure measures comprise age bands, contract expiry, temporary assignment, location, subject, school size and dependence on one-person functions.

Each measure should be reported by education level, sex where relevant, geographic unit, school disadvantage or remoteness category, subject or curricular function, employment status and qualification status when cell size and lawful confidentiality allow. The purpose is not to rank teachers. It is to identify concentrations of service risk. A national total can remain stable while one district loses its mathematics teachers, while beginning teachers leave particular schools, or while small schools operate without a headteacher. Distributional publication makes those differences visible.

Four derived signals provide a practicable warning sequence. The replacement-pressure ratio compares expected losses over the planning horizon with the current active stock. The supply route-coverage ratio compares plausible new entrants, after observed completion and appointment yields, with replacement and growth demand. The persistent-vacancy share identifies funded vacancies exceeding the ordinary recruitment period. The continuity-risk share identifies schools or required subject services experiencing repeated departure, vacancy or unmatched coverage. None has a universal favourable threshold. Authorities should set review thresholds from their appointment calendar, service standards, historical variation and capacity to intervene.

Dashboard governance is as important as selection. Every release needs a reference date, coverage statement, numerator and denominator definitions, revision marker and responsible data steward. Red, amber and green categories may assist operational reading, but the numerical value and reason for classification must remain visible. A colour must not substitute for uncertainty. Suppression rules should prevent identification of individual staff in small schools, while district officers with lawful duties may receive controlled case information necessary to arrange coverage. Public reporting should concentrate on system obligations, not individual blame.[REF-11] [REF-12]

A trigger is a rule for attention, not an automatic personnel decision. Three types should be distinguished. A level trigger is crossed when an indicator exceeds a stated value, such as a funded vacancy remaining open beyond the normal appointment window. A change trigger is crossed when deterioration from the previous comparable period exceeds ordinary variation. A conjunction trigger requires several observations, such as rising transfers, unsuccessful recruitment and loss of subject coverage in the same district. Conjunctions often reduce false alarms but may delay recognition; single severe events, including loss of the only teacher at a small school, require immediate action regardless of trend.

Verification should follow a fixed sequence. The data steward confirms identity matching, reference dates, denominator and late transactions. The establishment authority confirms that the post exists and is funded. The district verifies the school, timetable and interim coverage. The personnel authority establishes recruitment or departure status. The finance authority confirms salary and allowance execution where relevant. The result is classified as confirmed risk, data correction, resolved interruption or unresolved case. The record should preserve both the original signal and its disposition so that the warning rule can later be evaluated.

Escalation should correspond to educational consequence. A local staffing officer may resolve an isolated short absence. A persistent subject vacancy spanning several schools may require regional deployment, preparation or distance-support measures. A national cluster of expected retirements may require intake, finance and regulatory decisions several years in advance. Emergency interruption may require temporary measures, but temporary appointments, enlarged classes or compressed timetables must carry expiry dates, minimum preparation, supervision and restoration plans. The ease of a temporary response must not turn a warning mechanism into a route around professional standards.

Authorities should monitor the performance of the warning system itself. Relevant measures include the share of signals verified within the required time, positive predictive value where outcomes can be observed, false-alarm burden, missed interruptions discovered outside the system, time from warning to coverage, recurrence after closure and distribution of unresolved cases. These measures are not a release judgement on the report. They are ordinary management evidence about whether the early-warning arrangement protects instruction.[REF-11] [REF-14]

Cross-country comparison should begin from functions that all systems must perform: estimate demand, identify active teachers, recognise exits, authorise and finance positions, recruit, deploy, ensure presence and sustain required teaching. Administrative forms differ. A central civil service, decentralised employer, school-based contract system or mixed public and non-government arrangement may locate these functions in different bodies. Comparison is sound when it aligns the function, population and observation period; it is unsound when it assumes that one administrative label has the same legal meaning everywhere.

The UNESCO Institute for Statistics projection work demonstrates the scale of teacher requirements implied by universal primary education and the need to account for both expansion and attrition. Its comparative value lies in a consistent scenario framework, while national decisions still require school, subject, training and financing detail. The 2008 Education for All assessment similarly places teacher recruitment and quality within the remaining distance to 2015. These sources support warning about aggregate pressure; they do not by themselves identify which school will be vacant next term.[REF-13] [REF-15]

In systems undergoing rapid enrolment expansion, growth demand may dominate replacement demand, and shortened preparation routes may conceal a future quality and retention problem. In systems with stable or declining cohorts, an ageing workforce can still create abrupt subject or location shortages because pupil decline and retirement are not distributed alike. Small island and remote systems face indivisible minimum staffing: a national surplus cannot replace the only qualified teacher required at a distant school. Conflict-affected systems may lose records, facilities and staff simultaneously, making conventional rates least reliable when warning is most necessary.

European comparative evidence available by 2008 shows the limits of the available common indicators. Teacher age, pupil–teacher ratios and prospective youth cohorts describe parts of supply and demand but do not yield a universally accepted shortage measure. This limitation is analytically useful. It directs authorities to publish national vacancy, applicant, attrition and service-gap measures with full definitions, then compare patterns and institutional responses without claiming false equivalence. The proper comparative conclusion may be that two systems face similar educational risk through different administrative mechanisms.[REF-16] [REF-17]

Attrition and vacancy are distributional questions because the learners with the least alternative provision often bear the longest interruptions. A national warning system should test whether rural, remote, poor, minority-language, disability-inclusive, conflict-affected and small-school settings experience higher departure, longer recruitment duration, lower qualified-applicant yield or more frequent unmatched assignment. The categories must reflect lawful national data practice and should not be used to attribute risk to learners or communities. They describe obligations placed on the system.

Professional conditions belong inside the indicator framework. Salary level relative to relevant alternatives, payment regularity, housing, travel, safety, workload, class composition, leadership, access to preparation, career progression and participation in decisions may influence recruitment and retention. These are not interchangeable and their association with departures does not prove causation. They are nevertheless legitimate diagnostic evidence. The 1966 Recommendation concerning the Status of Teachers connects employment conditions, professional preparation, consultation and educational quality; it does not support treating teachers as movable units without rights or context.[REF-02]

Corrective action should be proportionate. Hardship allowances may be suitable where a verified cost or amenity difference impedes recruitment, but they require timely payment and evaluation. Local recruitment may improve continuity where candidates meet standards and receive support, but it must not confine disadvantaged areas to a lower qualification rule. Compulsory transfer may fill a post quickly while worsening morale or producing rapid return movement. Housing provision, induction, mentoring, safe transport, workload adjustment and career recognition may operate together. The warning record should identify the hypothesised constraint and the response selected so that continuation depends on evidence.

The Convention on the Rights of the Child requires education to be available on equal terms and directed to the development of the child. The Convention on the Rights of Persons with Disabilities entered into force on 3 May 2008 and requires an inclusive education system. At the cutoff, this strengthens the case for tracking specialist and support capacity required for inclusion without defining children as a staffing burden. A vacant or unmatched post matters because it interrupts an educational entitlement.[REF-12] [REF-18]

The first action is to freeze and reconcile a base date before the next appointment cycle. National authorities should match personnel, payroll, establishment and school-census records; assign unique case identifiers; express part-time service consistently; and publish the residual that cannot be reconciled. Districts should confirm schools with no assigned teacher, no head, no required subject coverage, prolonged vacancies or repeated temporary coverage. Corrections should remain auditable through dated transactions rather than overwriting the starting position.

The second action is to construct cohort flows. Every active teacher at the base date should have a subsequent status: continuing in the same school, transferred, promoted outside teaching, on defined leave, retired, resigned, deceased, dismissed, contract ended or unknown. New appointees should be followed from offer to acceptance and reporting. Preparation cohorts should be followed from entry to completion, eligibility, application, appointment and reporting. These linked transitions expose where apparent supply is lost and permit plausible lead-time estimates.

The third action is to establish warning rounds aligned with decisions. A term-opening round addresses immediate coverage. A pre-budget round identifies posts, allowances, preparation places and transition costs. A pre-recruitment round identifies subjects and locations requiring targeted action. A mid-year round verifies persistent vacancies, unplanned departures and the effectiveness of temporary coverage. The schedule should state who receives each signal, the time allowed for verification and the escalation path when responsible institutions disagree.

The fourth action is to publish a bounded national profile. It should show workforce stock and flow, attrition by reason, vacancy by entitlement basis and duration, recruitment yield, age and contract exposure, geographic and subject concentration, instructional coverage and uncertainty. It should explain breaks in series and refrain from a single international league table. Authorities should accompany the profile with corrective commitments, responsible bodies and next reporting dates. This closes the loop between statistics and educational service.

By 18 October 2025, the global evidence supports urgency but not a universal numerical alarm threshold. Education for All expansion, teacher retirement, uneven deployment, health and economic pressures, and the difficulty of attracting candidates to particular schools create different risk combinations. The defensible response is a common measurement architecture with locally valid thresholds, rapid verification and professional safeguards. Early warning succeeds when it preserves continuous, appropriately prepared teaching where learners would otherwise lose it; it fails when a favourable national average conceals that loss.[REF-01] [REF-02] [REF-13] [REF-14] [REF-15]

Warning is valuable only when it arrives before the last feasible decision date. Teacher preparation, budget authorisation, establishment creation, recruitment, appointment, relocation and reporting each have a different lead time. A system that observes a retirement on the day a teacher leaves has produced an accurate event record but no useful advance notice. The planning account should therefore work backwards from the date on which instruction must be available. It should identify the latest dates for approving finance, opening recruitment, completing selection, issuing offers, arranging travel or housing, confirming acceptance and preparing interim coverage.

Retirement exposure ordinarily permits the longest notice because age and service records are known, although eligibility is not the same as intention. Contract expiry provides a definite review date but may leave uncertainty about renewal. Pregnancy, illness and family leave require lawful and confidential administration; they should inform substitute capacity without becoming grounds for adverse treatment. Resignation may offer only the contractual notice period. Death, displacement, insecurity and sudden disaster may offer none. The warning system therefore needs both a planned replacement channel and a contingency reserve.

Preparation lead time is particularly important. Increasing admission to a multi-year programme cannot fill next term's vacancy. Authorities should distinguish immediate candidates already eligible for appointment, near-term candidates completing preparation, and prospective candidates whose entry can affect only later years. Completion, eligibility, application, appointment and reporting yields should be applied sequentially. A plan based on admitted students alone will overstate supply where dropout, examination failure, occupational choice, appointment delay or refusal of location is material.[REF-14] [REF-15]

Decision calendars should also recognise the school year. A vacancy discovered after national recruitment closes may persist for a full cycle even if many candidates exist. Late fiscal release can produce the same result. The dashboard should measure administrative time as well as labour-market scarcity: days from confirmed need to recruitment authority, advertisement, shortlist, offer, acceptance and reporting. This separation locates delay within the responsible institution. It also prevents a general claim of shortage when qualified applicants are waiting for an appointment decision.

An intervention should be chosen at the earliest stage where evidence is sufficient. supply route weakness may justify outreach or additional preparation capacity. Low completion may require student support or programme review. Low application yield may point to employment attractiveness. Low offer acceptance may be concentrated by location or subject. Failure to report may reflect late notice, transport or housing. Early-career exit may call for induction, manageable assignment and leadership support. Each response has its own time to effect and should be evaluated against the transition it was intended to improve.

Thresholds should be calibrated from the authority's own comparable history and educational obligations. A simple approach begins with several years of school-level or district-level observations, identifies the signals known before each staffing cycle, and records whether a defined interruption subsequently occurred. The authority can then examine how alternative thresholds would have changed true warnings, false alarms and missed events. The purpose is not to maximise one statistical score. It is to choose a tolerable balance given the educational harm of missed coverage and the administrative cost of verification.

Back-testing must respect information dates. A model should use only data that would actually have been available when the decision was due. Later corrections, final payroll transactions and explanations discovered after departure cannot be treated as advance knowledge. Where historical records are incomplete, the authority should state that calibration is provisional and begin accumulating a prospective record. Apparent precision from reconstructed data is less useful than a transparent rule subject to scheduled revision.

Different risks warrant different loss functions. Missing the probable closure of a one-teacher school may leave an entire community without nearby provision; a false alarm may require only a telephone verification. A national age-structure alert can justify long-range review but need not generate individual case action. A warning concerning a rare specialist may need a lower threshold because replacement time is long. A large urban district may absorb one departure more readily but still face serious subject clusters. Thresholds should therefore be set by decision class, not by a desire for one visually simple national score.

Confidence intervals are appropriate for survey estimates, including sampled presence measures, but administrative indicators also have uncertainty. Coverage error, duplicate persons, unmatched identifiers, late transactions, unknown exit reasons and outdated establishments can materially alter results. Authorities should publish a range or quality class where a numerical interval cannot be calculated. They should test whether the warning remains under plausible corrections. A signal that disappears when one late transaction is entered should require verification; a large persistent gap across sources supports stronger action.

Calibration can itself create perverse incentives if managers are judged on whether an indicator remains below a threshold. Posts may be suppressed, exits miscoded as transfers, recruitment opened briefly to reset duration, or difficult schools administratively combined. Counter-measures include publishing related quantities, retaining transaction histories, using independent statistical review and examining educational service directly. A falling vacancy rate alongside more cancelled lessons is not improvement. The test is whether required, appropriately prepared instruction became more continuous.[REF-06] [REF-11]

The 2008 planning environment requires attention to shocks that can affect teachers and households simultaneously. Rising food and transport costs can reduce the real value and reliability of remuneration, especially for personnel posted far from markets or dependent on long travel. Fiscal pressure can delay hiring, salary release, maintenance and housing. Health burdens can increase absence, mortality, caring responsibilities and training loss. Conflict or displacement can change both pupil demand and teacher availability quickly. These pressures do not justify a speculative attrition estimate; they justify more frequent observation of the transitions through which service may fail.

A shock-sensitive profile should monitor payroll timeliness, exceptionally high absence, unplanned leave, transfer and resignation applications, applicant and acceptance yield, travel or housing constraints reported through lawful channels, and rapid enrolment changes. It should also monitor whether schools respond by merging classes, shortening the day, cancelling subjects or relying repeatedly on unprepared personnel. Changes should be compared with the same period of prior years where seasonality matters. National economic indicators can identify exposure, but school and personnel records establish the educational consequence.

Continuity planning should protect regular institutions. A contingency roster may include qualified retired teachers willing to return temporarily, teachers available for short assignment, supervised candidates near completion and mobile specialist teams. The rules should define eligibility, safeguarding, remuneration, duration, supervision and return to normal staffing. Community or volunteer assistance can support schools but should not be counted as equivalent teacher capacity unless it meets the applicable professional definition. Emergency measures must not convert temporary necessity into a permanent lower standard for disadvantaged learners.

Coordination is required because responsibility is distributed. Finance ministries control releases; public-service bodies may control establishments; education ministries set curriculum and teacher policy; training institutions control programme capacity; local authorities and school leaders observe service interruption. A warning that stops at the statistical office has not protected instruction. Each high-risk category should have a pre-agreed communication path and a body authorised to decide. Disputes over records or responsibility should have a time limit and an interim coverage rule.

Fiscal restraint should be evaluated over the full cost horizon. Leaving a post vacant may appear to save salary while increasing class size, dropout risk, replacement costs, supervision burden and inequality. Conversely, indiscriminate recruitment can create an unsustainable salary liability and crowd out materials or maintenance. Scenario comparison should show the cost of prevention, temporary coverage, delayed appointment and prolonged interruption. The educational consequence and distribution should accompany the financial result.

The appropriate stance at the evidence cutoff is vigilant but bounded. Authorities should shorten reporting intervals where shocks are intensifying, preserve definitions so that change remains interpretable, and avoid attributing an individual departure to a macroeconomic pressure without case evidence. The early-warning framework is designed to reveal accumulating exposure and service loss soon enough for proportionate action. It does not transform association into certainty, nor does it permit a crisis narrative to displace professional rights, due procedure or minimum educational conditions.[REF-02] [REF-03] [REF-13]