By Memory Nguwi
MOST people who work in human resources have heard of the Paterson Job Evaluation System.
Far fewer have heard of Castellion. That is surprising, because the Castellion system was built to do something Paterson was never designed to do: capture the full picture of what makes one job more demanding than another by measuring six separate factors rather than relying on a single dimension.
Job evaluation, at its core, is a way of working out which jobs in an organisation are worth more than others so that pay can be compensated equitably. It does not look at the person doing the job. It looks at the job itself. The Castellion system does this by breaking every job into six measurable factors, scoring each one, and combining the scores to arrive at a total that determines the grade.
The system was developed by Arthur Cortis of the South African Breweries Group between 1956 and 1962. The name fuses two famous SAB beer brands, Castle Lager and Lion Lager, permanently linking the methodology to its brewing industry origins. Its theoretical foundation draws on the “period of discretion” concept developed by Dr Elliott Jacques, which holds that the most reliable way to distinguish between jobs at different levels is to look at how long a person can work before someone reviews their decisions.
Castellion was introduced to Zimbabwe through Rhodesian Breweries in the early 1970s and has since become one of the most widely used grading systems across industrial, commercial, and parastatal organisations in Southern Africa. It is also completely open source. Unlike systems that require licensing fees or royalties, any organisation can adopt Castellion without financial barriers.
The system uses six factors grouped into three pairs: Effort (the complexity and intensity of thinking), Responsibility (independence and consequences of error), and Competence (education and practical experience). By scoring all six factors and combining them through a precise formula, the system produces a richer picture of each job’s demands.
Factor 1: Decision making (1-200 points): This is the most heavily weighted factor. It recognises four categories. rote memory decisions (1-16 points) involve simple, repetitive choices such as checking whether a floor is clean or sorting documents into folders. Pragmatic experience decisions (17-52 points) require drawing selectively on past experience in varied but clearly defined situations, such as a security guard responding to a suspicious incident or a supervisor deploying workers. Adaptive thinking decisions (53-122 points) involve analytical and evaluative thinking, such as an engineer diagnosing a recurring machine failure or a finance manager reasoning through conflicting data. Strategic decisions (123-200 points) require conceptual, imaginative thinking where neither approach nor content is fully defined, such as a chief executive formulating a long-term strategic plan.
Factor 2: Pressure of work (Coefficient of 1.0 to 2.0): This factor measures how frequently decisions must be made, the variety of issues handled simultaneously, and the pressure of switching between tasks. It is expressed as a coefficient multiplied by the decision making score. A coefficient of 1.0 means decisions are taken at leisure. A coefficient of 2.0 means urgent decisions across a full range of organisational problems under considerable pressure. This ensures that two jobs with identical decision complexity score differently if one demands rapid, simultaneous decision making. This is the pressure of making decisions not necessarily ordinary workload.
Factor 3: Controls and checks (1-112 points): This factor measures independence. At the lowest level, work is checked continuously as it progresses. Moving up, controls shift to completion reviews, then to indirect monitoring through monthly or quarterly reports. At the top, the only controls are organisational results measured over multiple years, and the person reports directly to a board.
Factor 4: Consequences of error of judgement (1-128 points): This factor assesses the potential damage when judgement proves wrong. It does not cover simple mistakes. At the lowest levels, standard procedures provide safeguards, and errors cost mainly wasted time. At higher levels, errors affect departments, injure people, or disrupt programmes. At the top, faulty long-term decisions can threaten the organisation’s survival.
Factor 5: Education (0-68 points): This factor scores the formal knowledge the job requires, not what the incumbent holds. No education scores zero. Ordinary Level passes score 17 points. A bachelor’s degree scores 35. A master’s scores 62. A doctoral qualification scores 68. The scoring table specifically recognises Southern African qualification structures.
Factor 6: Experience (Coefficient of 1.1 to 5.0): Experience is multiplied by the education score. Less than one year of required experience uses a coefficient of 1.1. Fifteen or more years uses 5.0. This means experience amplifies education. A job requiring a master’s degree and 15 years of experience scores far higher on competence than the same degree with one year of experience.
How the final score is calculated
The six factors combine into three subtotals. The effort subtotal equals decision making multiplied by the pressure of work. The responsibility subtotal equals controls and checks plus consequences of error. The competence subtotal equals education multiplied by experience. Adding all three produces the final score, which ranges from 4 to 1,248 points.
The 16-grade structure
The final score places each job into one of 16 grades grouped into five levels. Labour grades 16 to 13 (4-104 points) cover unskilled and semi skilled work. Standard grades 12 to 9 (105-270 points) cover positions requiring definite skills and increasing complexity. Senior grades 8 to 5 (271-623 points) cover supervisory, specialist, and professional positions. Executive grades 4 to 2 (624-950 points) cover senior managerial and professional roles. Top executive grade 1 (951-1,248 points) covers the most senior organisational leadership. Jobs scoring near grade boundaries are flagged as borderline and reviewed carefully before final assignment.
The Castellion system operates on non-negotiable rules. Evaluators examine the job, never the person in it. The job is evaluated as it currently exists, not based on future projections. Competent performance is assumed. Unlikely events are disregarded. Evaluation should result from consensus, not a single individual’s judgement. And there must always be someone on standby to present (job holder) oral evidence when required or to answer questions.
How a Castellion project works in practice
A typical Castellion project moves through sensitisation (ensuring all stakeholders understand what job evaluation is and that it does not automatically lead to pay increases), job description development (the foundation of all grading accuracy), job grading itself (through committee, expert, or hybrid approaches), results presentation to executive management, communication to staff, an appeals process for employees who believe their job was incorrectly graded, and final implementation with knowledge transfer so the organisation can maintain the structure independently.
Job evaluation establishes relative worth. It tells you Job A is more demanding than Job B. But it does not tell you how much anyone should be paid. Pay is determined through a separate pay structuring exercise that combines grading results with market data and organisational affordability. An organisation that completes an evaluation without pay structuring has done only half the work.
Nguwi is an occupational psychologist, data scientist, speaker and managing consultant at Industrial Psychology Consultants (Pvt) Ltd, a management and human resources consulting firm.