Factor Comparison Method

What the factor comparison method is

The factor comparison method (طريقة مقارنة العوامل) is an approach to job evaluation in which benchmark jobs are ranked on each evaluation factor separately, and the pay of each benchmark job is then divided among those factors in money. Each factor thereby acquires a scale measured in riyals, and the remaining jobs are priced by adding up the amounts that correspond to them on these scales.

What sets it apart is that its unit of measurement is money, not points. An approach that gives a job points and then converts the total into pay keeps measurement and pricing one step apart; the factor comparison method merges them. Its output is a figure in riyals from the very first step, and its input is existing pay in riyals as well.

Where the factor comparison method sits among approaches to job evaluation

Accounts of job evaluation describe three approaches: ranking, classification and points. Factor comparison is sometimes added to them as a fourth. Of the three, it is closest to the point method, because it works on factors rather than on the job as a single whole, but it parts company with the point method over its unit, as set out above. Job classification, by contrast, assesses each job as a whole against written grade descriptions.

The difference in unit has a practical consequence. The point method produces a ranking that can survive movements in pay, because points do not move when pay does. The factor comparison method produces monetary scales that need review whenever the pay of a benchmark job moves, because that pay is the material the scale is made of.

We found no published measurement, for the Saudi market, comparing the approaches to job evaluation on which to base a claim that factor comparison performs better or worse than the others.

How the factor comparison method is carried out, in five steps

1. Choose and define the factors. A small number of factors, each described with boundaries that separate one job from the next rather than in phrases that would fit any job.
2. Choose the benchmark jobs. Jobs whose content is known and stable, held by enough people, and whose existing pay is accepted as a basis for measurement.
3. Rank the benchmark jobs on each factor separately. Four independent rankings for four factors, written down before a single figure is mentioned.
4. Divide the pay among the factors. The pay of each benchmark job is split across the factors so that the parts add up to the pay exactly. This is the point at which the ranking becomes a monetary scale.
5. Price the remaining jobs. Each job outside the benchmark set is compared with the scales factor by factor, and the amounts it matches are added together. The total is its price.

Where the method is applied strictly, steps three and four are run twice, in opposite directions: the jobs are ranked, the pay is divided, and the ranking is then derived again from the division itself. If the two rankings differ, one of the benchmark jobs does not hold, a case examined after the worked example below.

What the factor comparison method depends on

• Job analysis. The method compares the content of one job with the content of another. If that content has not been described and documented, what ends up being compared is the job titles.
• Evaluation factors. The factors are chosen and defined before a single riyal is divided among them, because the division is the yardstick for everything that follows.
• Benchmark jobs. The method works only on a set of jobs whose existing pay is accepted as a basis for measurement. That condition is its weakest link, and it is examined after the worked example.

A worked example of the factor comparison method on three benchmark jobs

Take four factors, namely required knowledge, responsibility, mental effort and working conditions, and three benchmark jobs at their existing pay. The pay of each job is divided among the four factors so that the amounts divided add up to that pay:

• An accountant, paid SAR 9,000: knowledge 4,000, responsibility 3,000, mental effort 1,500 and working conditions 500.
• A procurement specialist, paid SAR 7,500: knowledge 3,000, responsibility 2,800, mental effort 1,200 and working conditions 500.
• A warehouse clerk, paid SAR 5,000: knowledge 1,500, responsibility 1,200, mental effort 800 and working conditions 1,500.

Look at the last row. The warehouse clerk’s amount for working conditions is three times the accountant’s, although the clerk’s total pay is SAR 4,000 lower. That is not a fault in the calculation. It is what the method produces by design: the order of jobs on a single factor is not their order on the total.

Now price a new job, a custodian of assets (أمين عهدة), by comparing it on each factor with the scales already in hand:

• Required knowledge: between the warehouse clerk and the procurement specialist, so it is put at SAR 2,200.
• Responsibility: close to the procurement specialist and a little higher, because of the custody of assets, so it is put at SAR 2,900.
• Mental effort: close to the warehouse clerk, so it is put at SAR 900.
• Working conditions: the same as those of the warehouse, so they are put at SAR 1,500.

The total of 2,200, 2,900, 900 and 1,500 is SAR 7,500, exactly the procurement specialist’s pay. The custodian reaches that figure through an entirely different mix, however: SAR 800 less for knowledge, SAR 100 more for responsibility, SAR 300 less for mental effort and SAR 1,000 more for working conditions.

Here lies the most useful thing the method yields: two jobs equal in price for two different reasons. If the holder of one objects to the two being equal, the answer can be set out factor by factor rather than stated as a bare result, which is what is asked for whenever an evaluation is challenged.

The figures in this example are assumed. They are given to show how the amounts are added and how a job is priced; they are not market pay, and they are not a reference on which to build a pay table.

The consistency test in the factor comparison method that removes a benchmark job

Factor comparison carries within it a check that not every approach to job evaluation carries: the ranking of the benchmark jobs on the four factors must be consistent with their existing pay. If the total divided for a job comes out different from its pay, or if its ranking on one factor requires its ranking on another to be reversed for no reason that can be stated, the fault lies in its existing pay, not in the division.

The correct response is then to remove that job from the benchmark set, not to adjust the figures until they fit. A job kept in the set after its division has been adjusted to match a doubtful wage passes that doubt on to every job priced after it.

What happens in the factor comparison method when a benchmark job’s pay moves

Raise the accountant’s pay in the example above from SAR 9,000 to SAR 9,900, an increase of 10%, because the market has moved in that specialism alone. The additional SAR 900 has to be divided among the accountant’s four factors. If all of it goes to knowledge, that factor’s share becomes SAR 4,900 instead of 4,000.

The effect does not stop at the accountant. The custodian’s knowledge amount of SAR 2,200 was set from the custodian’s position between the warehouse clerk and the procurement specialist. Neither of those has moved, so the estimate stays where it was and the custodian’s total remains SAR 7,500. Had the increase gone instead to a factor shared by more than one benchmark, responsibility for example, every job whose responsibility was estimated by reference to the accountant would move with it.

The practical rule follows from this. Before an increase on a benchmark job is divided, the question is why it was given, because the answer decides which factor carries it, and the factor that carries it decides how many other jobs will move. An increase spread evenly across the four factors because that is easier to calculate moves the whole table with nothing in the market to justify the movement. How an increase budget is shared out across employees in a pay cycle is described under salary review, and pay that rises outside the approved increase is measured as wage drift.

The inherent flaw of the factor comparison method: its input is its output

Factor comparison builds its scales from the pay the organisation is paying today. If that pay carries an accumulated imbalance, the method does not reveal it. Instead it reproduces the imbalance in the form of a scale that looks objective, because it has become figures attached to named factors.

For that reason the factor comparison method cannot on its own answer the question of internal pay equity, because that question is precisely what the method assumes has already been settled. The sound sequence is to test the pay of the benchmark jobs against a reference outside the organisation, such as a salary survey, and only then to build the scales on the jobs that remain after the test.

How the factor comparison method differs from paired comparison, the point method and the wage for comparable work

• The paired comparison method. It compares people with one another on performance, whereas factor comparison compares jobs on factors. The word comparison is shared, but what is compared is not: there the comparison is of a jobholder, here of a job regardless of who holds it.
• The point method. It gives ratings on the factors and adds them up as points, and the total becomes money only through a later conversion step in which the value of a point is decided. Factor comparison has no such step, because it starts with money.
• The wage for comparable work (أجر المثل). That term describes what is paid for similar work, while factor comparison is a procedure within the organisation that produces a price. The two may agree or may diverge, and neither stands as evidence for the other.

Where the factor comparison method does not hold

• Few benchmark jobs. A scale built on three jobs and used to price forty stretches estimation across distances with no evidence in between. The worked example above uses three jobs for illustration, not to suggest that three are enough.
• Closely similar jobs. Where roles are very alike, the differences on each factor become smaller than the evaluators’ ability to tell them apart, and personal judgement enters without being seen.
• Volatile pay. Every increase on a benchmark job reopens the division of its pay across all four factors, so an organisation that reviews pay twice a year has to review its scales twice a year.
• Nobody left to explain it. The method produces figures that are easy to present and hard to defend without the person who carried it out. If the reason for each division is not documented when it is made, a year later only the figures remain.

What the factor comparison method leaves to the organisation

The definition of the factor comparison method sets neither the number of factors nor their names. The four factors in the worked example are an illustration, and choosing and defining the factors is a decision taken within the organisation.

In what we have read of the Saudi Labor Law (نظام العمل), its Implementing Regulation (اللائحة التنفيذية) and the model work regulation (النموذج الموحّد), we found no provision requiring a particular method of job evaluation. So we do not write that the factor comparison method is required, and we do not write that it is prohibited.

Before a pay table is built on the factor comparison method

The factor comparison method suits an organisation that has a reasonable number of jobs whose pay it is confident in, and that wants to price the jobs around them in a way that can be explained factor by factor. It does not suit a situation in which the question being asked is whether that pay is sound in the first place.

The test before starting is a single question. Can you name five jobs in your organisation, say of each one that its pay is sound, and state what you based that judgement on? If you cannot, the step that comes before evaluation is testing that pay, not choosing an approach with which to evaluate it.

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

Related terms

• Job Classification
• Range Spread
• Pay Compression
• Salary Review
• Wage Drift