Staff Writer 
HR Perspective with MEMORY NGUWI
A PAY structure is, in essence, a pricing system for labour. It establishes the relative worth of every job inside the organisation and then tests that internal hierarchy against what the external market pays for comparable work. Fairness, retention and cost control all follow from those two decisions.
Most boards do not see it that way. They treat the structure as something HR must administer grades and midpoints that HR refreshes when it gets a new salary survey. Ask what it is for, and you get a policy answer instead of a full understanding of labour marketing pricing.
Every grade band is a price. Every midpoint is a posted price for a bundle of skill, responsibility, effort and risk. When you slot a job into a grade, you are quoting a price to two audiences at once: the market outside and the person already doing the work inside.
Those two prices answer different questions. The internal price says what a job is worth next to the one above it and the one below it, which is what job evaluation does, whether you use Paterson, Castellion or any other method you can defend. It is a ranking exercise before it is a money exercise. The money only becomes credible once the ranking is credible and defensible.
The external price is the test. You take the internal hierarchy to the market and ask whether the world agrees with your ordering and your levels. When the two disagree, you have a judgement call. A structure that is never tested against external market prices is not very useful internally for retention.
Start with the survey, because that is where most of the confusion begins. There is no single going rate waiting to be found. If one existed, two people with identical skills doing identical work would earn roughly the same wherever they went, and they do not. Research matching millions of workers to their employers finds that firm specific pay premiums explain about 20% of all variation in wages.
A meta-analysis pooling 1,320 estimates from 53 separate studies found that employers hold real wage setting power, because workers cannot simply cross the street. Searching takes time and moving carries risk.
So when the survey shows your grade C1 at the 40th percentile, that is not a discovery about the market. It is a description of a price you chose, and most of the time you chose it by accident. Benchmarking tells you what other buyers are paying, but it cannot tell you what a job is worth inside your operation, because that depends on what the job produces for you.
Most pay structures I come across were never designed. They accumulated, one exception at a time, until nobody could explain them. Ask three managers why a grade pays what it pays, and you will get three different answers.
That matters because people price their own work based on the person next to them. Pay was never a private matter between an employer and one employee. The instinct to compare is not petty; it is how buyers in every market decide whether a price is reasonable.
When researchers randomly told University of California staff about a public salary database, those earning below the median in their unit reported lower satisfaction and became far more likely to start job hunting. Those above the median barely reacted. Nobody’s bank balance changed, only what people believed their work was worth.
The effect extended not only to output but also to job satisfaction. In a month long field experiment with Indian manufacturing workers, those paid less than teammates doing the same job cut output by 0.33 standard deviations and cut attendance by 12 percentage points, losing pay for every day missed. Then the researchers found something every grading committee should know.
When the pay differences were tied to visible differences in productivity, the concerns disappeared. Workers were not objecting to unequal pay. They were objecting to unexplained pay, which is a different problem with a different fix.
That is not a one-off result. A review of the pay dispersion literature concluded that gaps between people help or harm depending almost entirely on whether the variation can be explained by something the workforce can see. A gap you cannot defend in a corridor conversation will cost you.
Direction matters too. At one large bank, employees who found out their managers earned more than expected worked longer hours and sold more. Employees who found out their peers earned more did the opposite. People stay motivated when they see a clear path upward, but they grow bitter when coworkers are treated differently for no obvious reason.
Ask employees to rank what motivates them, and pay usually lands around fifth. Whole wellbeing strategies have been built on that ranking. But when researchers compared what people say with what people do, they found employees systematically understate how much pay drives them, because admitting money matters feels thoughtless.
The reverse belief is also wrong. Paying more does not buy contentment at any reliable rate. Across 92 independent samples, the link between how much people earn and how satisfied they are with their jobs was only marginal at 0.15, and only slightly stronger for satisfaction with pay itself at 0.23. A larger synthesis covering 240 samples across 35 years reached the same conclusion about what drives pay satisfaction, and it is comparison, not the absolute number.
Bonuses are usually sold as a way to make people try harder. Price them properly and they are a transfer of uncertainty from the employer to the employee. A review of 106 studies on schemes that reward collective results found the link to performance was real but modest, which is what happens when people discount money they are not sure they will see.
Examining 138 state level minimum wage increases in the United States across 37 years, researchers found jobs below the new minimum disappeared and jobs just above it appeared in almost identical numbers, leaving total low wage employment essentially unchanged while affected workers earned about 6.8% more. That only makes sense if employers had been pricing labour below its value. Underpricing is not free, it is invisible for a while, and the consequences are seen during your next recruitment drive.
Disclosure is important. When California required cities to post senior salaries online, top managers’ pay fell by roughly 7% and quit rates jumped by about 75%.
So build a structure you can defend. Grade jobs with a method you can explain in one sentence to the person being graded. Write down why each grade sits where it does, because a reason you cannot write down is a reason you cannot defend. Make sure the distance between grades reflects a real difference in contribution, keep the promotion ladder visible, and choose where you sit against the market instead of letting that happen by default. Underprice a job today, and the market will reprice it for you tomorrow, through a resignation letter, at a rate worse than the one you refused to pay.
l Nguwi is the managing consultant of Industrial Psychology Consultants and a registered occupational psychologist.
