It has been said that every organization runs the risk of pay compression. Pay compression is when either:
- a subordinate is paid more than their supervisor, based on regular, not over-time pay, or
- when a less tenured employee is paid more than the more senior tenured colleague in the same job.
I am sure there are other examples of pay compression, but these are probably the easiest to describe and the most prevalent.
The latter form of pay compression is the most common – or at least it has been in more recent years. This is because organizations have held pay increases to a minimum, but new employees are paid a salary to attract them. Organizations have had to do this just to fill vacant positions. This problem becomes more severe in economic downturns, such as the 2008 recession, but it occurs even in better economic times due to the way pay increases are managed.
The former situation usually is most evident in pay systems where lower level jobs, either through union contracts or other market forces, create a situation where first-line supervisors are paid less, on an hourly basis, than their subordinates.
Defining Pay Compression
Many times, we hear about the first form of pay compression when supervisory employees compare their take-home pay with the take-home pay of their subordinates who are most likely earning over-time. Technically, however, this is not pay compression. Pay compression is only when the standard rate of the supervisor’s pay is less than their subordinates when calculated for the same hours of work. Unfortunately, though, supervisory employees do not see it that way.
Either form of pay compression is a symptom of a pay system that is not being administered correctly or not keeping up with the salary movement. At its worst, pay compression generates legal action and equal pay claims (sometimes unwarranted). At its least, it can create a serious employee morale problem. Our experience indicates that it is costly to fix, and the damage to employee trust and morale takes years to resolve. In one case with which we are familiar, a 2,000-employee organization estimated that it would require $11 million to substantially reduce or remove pay compression over a three-year period. They never totally eradicated the problem because other priorities emerged in the third year. Seven years later, the problem is still unresolved and is a continual employee irritant.
Recognize that regardless of how hard you try, as long as you have employees entering, leaving or moving in the pay system, you will have some degree of pay compression. You should try to minimize it as much as possible, because once it gets out of control or its causes become institutionalized, the cost to get out of the problem will be large, the pain will be deep and the solution will be elusive.
Identifying the Situation and Finding a Solution
But how do pay administration practices create these situations, and what is a long-term solution for them?
It has been our experience that supervisors are paid less than their subordinates when there are either separate pay schedules governing pay for the subordinates and the supervisors, or there is some disconnect between the pay schedules or the adjustment to the pay schedules. Compression usually occurs at the first-line supervisory level and does not extend to the supervisor’s supervisor.
The existence of separate pay schedules may be the result of:
- developing a pay schedule that is specific to a job family, and/or
- a union contract governing the subordinate’s, but not the supervisor’s pay.
In this situation, because there is no direct connection between the two schedules, the process of adjusting them may be different in timing or amount. Thus, when a union contract governs the subordinate’s pay schedule, the pay adjustment is governed by negotiations. These negotiations may be influenced by tradeoffs in work rules, benefit costs, special pay categories, skill-based pay adjustments and previous arbitrations. In addition, most jobs at this level are paid based on a step system, whereby the employee receives a step increase and a schedule increase, as long as they have not exceeded the maximum of their pay range.
The supervisor’s pay, on the other hand, is more likely to be part of a supervisory/managerial pay schedule, which is most likely not governed by union negotiations or other restrictions. In addition, the pay schedule is more likely to be an open range, based on performance or some criteria other than years of service/step increases. As a consequence, elected officials are more likely to agree to the union contract, but try to hold pay increases down for the supervisory/management group. If there was a difference in pay between the supervisor and the subordinate before this situation occurred, then pay compression will become apparent after only a few years of this pattern of disparate pay schedule adjustments.
We have also seen this situation in recent years as organizations broaden the pay ranges, creating a greater degree of overlap between adjacent pay ranges. About 30 years ago, pay ranges for most government jobs had about a 25-35 percent spread from minimum to maximum. This was generally sufficient because the work was tightly controlled and highly specific, and it made it easy to establish about seven to 10 steps from minimum to maximum for employee pay growth. In addition, pay was usually compared to the market at the maximum of the pay range, as opposed to the 50th percent of the market, which is now the typical situation.
Over the last 30 years, we have seen the development of broad-banding, skill-based pay, competency-based pay, and a shift to look at the 50th percent of the market as the comparison point to the midpoint of the pay ranges. This has resulted in pay ranges that are broader and fewer. But this has not occurred for all jobs. Non-exempt jobs still have salary ranges that are narrower than managerial pay ranges. It is not uncommon in some organizations to have 60-percent range spreads for managers, and we have seen some as high as 85 percent.
To anyone doing the math, it is obvious that if the range spread is broadened, the minimum will be lower and the maximum will be higher than in the past. If this is done for both subordinate and supervisory pay ranges, or if it is only done for the supervisory pay ranges, what happens is that the pay ranges of subordinate and supervisory begin to overlap. Thus, a new supervisor, if paid at the minimum of their pay range, could easily be paid less than the longer-term subordinate.
So, what is the solution to this issue? It appears obvious to us, but it is not an easy solution. That is, the jobs need to be placed on the same internal hierarchy and priced together. This sounds simple, but it involves the adaptation of a job evaluation system or an internal hierarchy scheme that is common to all jobs. While internal equity has always been an important consideration for valuing jobs within a schedule, it has not always been as important to be used across schedules. If applied correctly, there will be one hierarchy across all comparisons with an agreed upon comparison group. Then, when market data is introduced, and internal equity is merged with the market data, the pay structure of grades and ranges will dictate pay ranges that will, for the most part, create a significant pay difference between the subordinate and the supervisor’s pay ranges, provided that the overlap between ranges is kept to a minimum, and the grade of the supervisor is usually more than one grade higher than the subordinate job grade.
Establishing Internal Equity
Since most organizations already have salary ranges in place, this solution means that some effort will need to be made to establish the internal equity link between the jobs of the supervisor and subordinate. This can be accomplished in the next round of pay adjustments or over time, depending on the severity of the problem, or it can be accomplished by adjusting the differences between the ranges and the range spread. Ideally, there should be at least a five percent difference between adjacent pay ranges, calculated at the midpoint of the range (or the minimum or maximum). Normally there will be a two-grade difference between the grade of the subordinate and the supervisor. If this is the case, a broader salary range will not create significant pay compression issues, if any. If you can manage a larger difference between grade midpoints, like 7 to 10 percent, it would be even better, but you need to make sure that the market and internal equity differences are consistent with that degree of spread.
The second form of pay compression, where a more senior employee is paid less than, equal to or nearly equal to a newer employee in the same job title, is generally a costlier solution. In addition, in order to maintain the solution without slipping back into pay compression in the future requires a different way of managing pay and pay schedule adjustments.
Recall that this form of pay compression usually emerges when a current employee does not receive an increase in pay that is matched by the market movement. Then, when a vacancy is filled, the new employee who is less constrained by current pay administration practices receives a salary offer that is sufficient to attract them to the job, but also sufficient to set off pay compression issues.
Of course, the applicant is well worth the money (or so we are told) until about a year later when we find that their value was really not at the level of the salary that was paid. So now you are faced with the proposition of either trying to smooth the “compressed” employee’s ego, or throwing money at the problem in the hopes that this resolves the situation.
The irony of the issue is that the former does not work in the long run, and the latter could have been avoided had prudent salary management been used.
Prudent Salary Management
What is prudent salary management? It consists of the following two issues: Salary range adjustment and employee salary movement.
I do not know where this idea came from, but it has been around for a long time. When most salary ranges were managed with steps, it probably made sense, but if you don’t have steps, it doesn’t. That is, when salary ranges are moved by a standard factor, such as a “cost of living” adjustment, then everyone on the salary range moved up by the same amount. Now, if the salary range has steps, and an employee also moves up a step, then most everything is kept in sync. Especially, if new employees are hired at the first step without exception.
However, when you move from steps to open ranges, and adjust ranges and employee pay at the same amount, the problems begin. This is because an employee who is hired at the minimum of the range will remain at the minimum of the salary range regardless of what adjustments are made to the salary ranges in the market. Then, when a new employee is hired, their pay is usually at least as high as the more senior employee.
There is a permanent solution and a one-time solution. The permanent solution involves adjusting the ranges by an amount that is less than the “cost of living,” or adjusting the ranges every other year. In conjunction, adjust employee pay at an amount that is greater than the cost of living.
Regardless of how you define the cost of living, this will keep the pay compression issue at bay. This may seem like a foreign concept, but the private sector has been doing this for years. This solution recognizes that the purpose of a salary range is to guide overall hiring rates to be reasonably competitive with the market and to control top-end salaries. It should not be used to adjust employee pay. If it is considered in this perspective, an adjustment of the salary ranges should not necessarily trigger a salary increase for any employee, unless they are at the minimum and performing at below-average levels of performance.
Many elected officials will complain that in lean financial years (when have we ever had boatloads of money?) they could not afford this solution. However, what we are finding now is that these same organizations have hired new employees and have a pay compression problem that is seriously affecting morale and increasing turnover. The solution now becomes a temporary one-time solution and is very costly. It involves reviewing the seniority of all employees in their current positions and adjusting pay accordingly. When we have performed this analysis for clients, the cost is usually significantly more than if the organization had followed the prescription outlined above that would have avoided the issue in the first place.
The solution usually involves adopting some form of algorithm that specifies how employee pay should be adjusted. For example, one model follows these guidelines:
- • An employee receives an X percent pay increase for every year of service up to a certain number of years.
- • For every two years of service after that, they receive an X percent pay increase up to a certain number of years.
- • For every three years of service after that, the employee receives an X percent increse up to the maximum of the salary range.
This will normally spread pay out over about a 20- to 30-year career, but the timing can be adjusted at any point. For example, some organizations have capped increases at the midpoint instead of the maximum. The correct solution is dependent on the money available, as well as how the organization wants to manage pay adjustments in the future. It should be noted, however, that unless the organization adopts the first solution in the future, pay compression will quickly re-emerge.
Pay compression is a tough issue with which many organizations struggle. In our experience, no one resolves the issue once and for all. It requires constant vigilance, board education and improved salary management practices. It may not be totally resolved, but its negative effects can be substantially reduced. And for that there is usually a monetary cost of some form or another.
The Comp Doctor™ is Jim Fox and Bruce Lawson, Managing Directors in Gallagher's Human Resources & Compensation Consulting practice. He specializes in assisting governments in fixing their compensation and classification systems.