Grade Drift

What grade drift is

Grade drift (تضخم الدرجات, in Arabic literally “grade inflation”) is the gradual movement of employees or jobs to higher grades on the ladder of job grades at a pace that exceeds what a change in the work justifies. Over the years more employees sit in higher grades, while the tasks and responsibilities they perform do not grow. The Arabic name borrows from inflation: the grade is handed out generously and loses its power to distinguish one level from another.

Grade drift is a phenomenon in the structure of grades and not a judgement on any particular promotion, since a single promotion may be deserved. What follows is an explanation of the management concept and of ways of reading it. It does not establish a right or a duty to promotion.

How grade drift differs from wage drift, grade placement, job classification and job evaluation

These concepts belong to the field of pay structure, and they differ in what they measure:

• Wage drift is the amount by which the wage base actually exceeded what the approved raise decision required. It is a gap in money, whereas grade drift is a gap in where people stand on the ladder. Grade drift may be one cause of wage drift.
• Employee grade placement puts each employee in a grade and at a position within that grade’s pay range when the ladder is built or when the employee joins. Grade drift is what happens to placement afterwards as time passes.
• Job classification builds the ladder, with the descriptions of the grades and the steps for placing a job in it. Grade drift is what affects that construction when jobs move away from it without any change in their descriptions.
• Job evaluation ranks jobs by their relative value. Comparing the current grade of a job with what its evaluation justifies is the first of the two ways of measuring grade drift.

Where grade drift comes from

These are possible sources of ours, not a statistic of what happens in establishments, and they need not all be present in one establishment:

• Promotion to justify a raise. An establishment uses the grade as a means of raising pay when the range of the current grade has no room. Wide pay bands remove the need for such a nominal promotion, as broadbanding describes.
• Reclassification under pressure. The grade of a job is raised because its holder asked for it or hinted at leaving, and a technical reason is written for the decision after it has been taken.
• A change of title without a change of content. The title is raised and the grade follows, while the tasks stay as they were.
• Hiring from a higher grade. A candidate is appointed at a grade above that of the advertised job because the market requires it, and the grade of the job is then treated as the grade of its holder.
• Accumulated exceptions. Each case is handled separately and the original evaluation is not reviewed, until the exceptions become the actual ladder.

How grade drift is measured

Grade drift is measured in two ways, and they answer two different questions:

• The distance from the evaluation. This is the distance between the current grade of a job and the grade that its evaluation justifies. It needs an updated evaluation of jobs.
• The share in the upper grades. This is the share of employees in the upper grades out of all employees at two dates. It is not the average grade number, because grade numbers are an ordering and not a scale on which the gaps between numbers are equal. The worked example below uses this measure.

Two readings can be added to them. The first is the share of approved requests to raise a grade out of all requests in the period, together with the number approved without any change in the job description. The second is the list of jobs whose grade was raised in the period, with what changed in their written tasks before and after the decision.

These measures are better applied at the level of a job family and not of the whole establishment, since grades may drift in one family and stay stable in another.

A hypothetical worked example

Take a hypothetical establishment with 120 employees spread over six grades. In the first year the distribution was 12 in grade 1, 30 in grade 2, 36 in grade 3, 24 in grade 4, 12 in grade 5 and 6 in grade 6. The total is 12 + 30 + 36 + 24 + 12 + 6 = 120. Two years later the distribution was 6, 24, 36, 30, 18 and 6, and its total is 6 + 24 + 36 + 30 + 18 + 6 = 120, so the number of employees did not change.

In the first year, 42 employees were in grades 4 to 6 (24 + 12 + 6 = 42), which is a share of 42 ÷ 120 = 35%. Two years later, 54 employees were in those grades (30 + 18 + 6 = 54), which is a share of 54 ÷ 120 = 45%. The share rose by 10 percentage points. The number of additional employees in the upper grades is 54 minus 42 = 12.

The establishment then reviewed job descriptions and found that five of these 12 movements were supported by a documented change in the tasks of the job. That leaves seven movements without support (12 minus 5), which is 7 ÷ 12 = 58.3% of the movement and 7 ÷ 120 = 5.8% of the employees.

Now suppose that the 120 employees belong to two families, the first with 40 employees and the second with 80. Split by family, the figures show what the overall figure hides. The first family had 10 employees in the upper grades in the first year and 20 two years later, so its share is 10 ÷ 40 = 25% and then 20 ÷ 40 = 50%, a rise of 25 points. The second family had 32 and then 34, so its share is 32 ÷ 80 = 40% and then 34 ÷ 80 = 42.5%, a rise of 2.5 points.

The two families together had 10 + 32 = 42 in the first year and 20 + 34 = 54 two years later, which matches the overall figure. The movement, however, was concentrated in the first family, and that is where the question begins.

This calculation is approximate, because the net difference between two distributions does not distinguish those who moved within the establishment from those who left or joined. It does not prove that the remaining seven movements were undeserved, since their support may exist unwritten. The figures are hypothetical and set to illustrate. They do not represent an establishment or a study, and the example does not mean that a share like this is customary.

How grade drift shows in daily work

The figures may not show drift until its signs have already appeared in daily work. These signs are signals that merit a question and not evidence:

• After each raise cycle there are many requests to raise a grade, and many of those requests come with a competing or outside offer attached.
• An employee holds the same grade as a subordinate, or one close to it. This may happen when grades are raised without regard to the reporting order of the structure.
• The upper grades are full and have few vacancies, so that those who have actually qualified cannot rise.
• Job descriptions have not been updated for years and are still used as the reference for decisions to raise a grade.
• Pay ranges overlap in adjacent grades, which may be caused by pay in the lower grades catching up with pay in the grade above.

Why grade drift is a concern

The effect of grade drift shows in three places, all of them possibilities that depend on the establishment:

• Grades may lose the ability to describe the difference between one job and another, which weakens the pay fairness that rests on them.
• The cost of pay may rise over time, because the pay range available to an employee changes with the grade.
• The door may close on deserved promotions, because the upper grades have filled with movements that had no support.

None of this can be measured except by comparing what happened with what was expected, so a single figure in a single year is not enough for a judgement.

What is not grade drift

• Growth in the actual work. An establishment whose activities expanded may have jobs that are larger than before, and raising the grade of such a job after evaluating it again is a correct decision. If the job description was not updated, the movement may look like drift while having a basis, which is why a change is documented in the description.
• A change in the composition of the workforce. The arrival of a new branch with senior jobs raises the share of the upper grades without any single job moving.
• Deserved promotion. Promotion rate measures the share of those promoted in the period out of the employees covered, and a rise in it may reflect employees who have qualified for larger roles.
• Succession between generations. An establishment whose senior jobs fell vacant through retirement or departure fills them from inside.

The question that separates drift from these cases is whether each movement has a change in the work behind it that can be described. A neighbouring concept is the Peter principle, which concerns the person who is promoted until reaching a role that the person does not master, and not the structure of the ladder.

What can limit grade drift

The measures below are possibilities that we propose for an establishment to adopt according to its situation, and they are not a prescription:

• Requiring support for a raise. Any request to raise a grade carries evidence of a change in tasks, taken from job analysis or from an updated description.
• Separating the retention route from the grade. The establishment has another way to recognise those it wants to keep, so that the grade is not the only means.
• Approving from outside the direct reporting line. The person to whom the employee reports does not alone decide the raising of the grade.
• Publishing the grade distribution to management each year. Management can then see a trend before it builds up.
• Evaluating jobs again at intervals. The original descriptions do not remain the basis of grades that have actually changed.

Another question arises once drift is found: what to do with an employee whose grade was raised without support. The legal side of lowering a grade is covered in demotion, and it lies outside the concept described here. We suggest that an establishment check each movement before it is made and not after, by requiring support for a raise.

How much movement towards the upper grades is acceptable

We set no acceptable share for the upper grades and no acceptable speed for its rise. In the sources we reviewed, we found no published source that fixes a figure for this, and we propose no figure of our own. The share varies with the activity, the age and the structure of the establishment, and it is judged in each establishment against the changes in the work that justify the movement.

Grade drift remains a reading of what happens to the structure of grades over time and not a prescription applied to people. It shows the establishment where a question is worth asking, and what is decided afterwards rests on job descriptions and on the facts of each case.

This is an explanation of the concept, not legal advice.