Cafe Classic: Increases as a Percentage of Job Value

Editor's Note:  A better way to do merit pay?  Jim Brennan lays out the Classic case for delivering merit increases as a percent of job value (or range midpoint or market reference point) rather than as a percent of actual base pay.  How many of you are already doing this?  How many are considering?

The Percentage of Job Value approach seems far superior to the antique habit of awarding increases according to base pay.  Instead of giving a percentage of the base rate as the reward, make it a percentage of the midpoint job value or Market Reference Point (MRP).  That simple method completely avoids the many problematic issues of merit increase grids, which are complex systems, both difficult to create and hard to explain. 

Granting the same percentage of a common midpoint or MRP for the same performance is eminently fair.  Also, it is far clearer and easier to communicate than awarding weird personal awards that compound if not modified by an awkward merit increase grid. 

If two peers holding equally valued jobs perform at the exact same performance level for a specific evaluation period, it seems right for them to get the same merit-related cash reward.  Allocating increase amounts or bonus pools according to personal pay, on the other hand, can create terribly dysfunctional distributions.  Makes no sense to say that superior performance is worth more money to a well-paid incumbent than to a lower-paid peer, when logic would argue the opposite.  Great output results from a person working at a bargain rate would be remarkable, while such top levels of performance should be normal for a highly paid incumbent who has earned premium pay for a long time. 

The size of a personal salary accumulated over many years of past compounded base pay increases should have no relevance to the value of output results in a later period.  If retention is a concern, that should be a much bigger issue for the low-paid worker than for their compatriot who does the same work at an inflated rate that far exceeds the market norm.  People who have been historically generously compensated for many years by hefty guaranteed base rates have no right to an infinite entitlement to the largest increases. 

Merit increase grids are therefore used by many who apply increase percentages to base pay.  They can prevent uncontrolled compounding but have many drawbacks.  Merit grids are complicated to design, construct, administer and explain; and they earn criticism for “punishing” high earners by imposing lower increase percentages on those with higher compa-ratios.  There are many potential negative consequences to applying reward percentages to base salaries, and the traditional corrective matrix grid has accumulated a wide variety of defects over the years.

There is a better way.  Instead of referencing a complicated subjectively created table with projected estimates of frequency distributions to slow the range penetration trajectory for high-earners, you can just apply your increase percentages to the Job Value (or Range Midpoint, Control Point or MRP).  That much simpler approach provides perfect symmetry in absolute pay distributions that are controlled by the value of the position and the personal performance of the individual worker without regard to the size of their current wage or salary.  No table.  No cut-off points.  No arbitrary parameters.  No complexity.

A number of strings in various discussion forums cover this topic, generally findable under “using the Job Value” or “Brennan Payout” approach.   The practice has been around for a long time in both the U.S. and Canada, with instructions published various places (i.e., in The Personnel Journal and the Performance Management Workbook) in the past.  It has been extremely successful when used; yet many compensation professionals are unaware of the method.  

“Relatively absolute” payments computed against normative job values are far more equitable than percentages applied to personal base pay amounts.  Same job worth, same output value, same absolute reward dollars.  How could such perfectly fair treatment be wrong?

E. James (Jim) Brennan is an independent compensation advisor with extensive total rewards experience in most industries.  After corporate HR posts and consulting CEO roles, he was Senior Associate of pay surveyor ERI before returning to consulting in 2015. A prolific writer (author of the Performance Management Workbook), speaker and frequent expert witness in reasonable executive compensation court cases, Jim also serves on the Advisory Board of the Compensation and Benefits Review.

Image "Percent And Dollar Symbols" courtesy of renjish krishnan/FreeDigitalPhotos.net

Comments

[Jack Loring]

AMEN!

Jim,
I appreciate the step-by-step instructions on this approach that you provided in your book.

Some organizations utilize a forced placement of ratings in order to replicate a normal distribution (bell curve). In my view, that approach is more about budget management not performance management.

The "Brennan Payout" approach will match the salary increase budget (to the penny) without forcing managers to alter their ratings in order to fit the budget allocation.

Thanks for sharing your wisdom.

Jack

[E. James (Jim) Brennan]

Pleased to learn that you took it to the next step, Jack! One of these days, someone will "suddenly discover" it and force-feed it down their clients' throats until all have abandoned the ugly forced distribution edicts and the awkward merit grid techniques.

Glad to hear that you also found value in it... hopefully during the years that my book royalties were still in effect! It sells for less than the postage cost, today. LOL!

[Amber B]

Thank you for sharing this information! It's a different approach and I agree with the logic behind it, I'm considering how this change could work at my organization.

[E. James (Jim) Brennan]

For those who agree that this approach seems promising to simplify communications about equity (same amount for same performance in same-valued jobs) without the prejudicial effects of past years, I suggest some Beta Testing.

Carefully select some unit with a smart articulate supervisor who wants to volunteer his section for a trial run.

Do some preliminary rehearsals, pitching all the possible objections and protests you can imagine. If you can't think of them, I'll give you a bunch of samples, like "us highly paid senior people deserve a bigger piece of the pie because we're fatter"... I mean, "..because we ALWAYS get a bigger raise for the same level of performance as a mere beginner in the same job we have."

Run the concept past your upper management to feel them out and keep them in the loop so you don't go behind their backs. Ideally, you would have a slight premium added to the budget for the volunteered section or group, to supply an incentive to the manager to test it. Their pleasure at the subsequent clarity and improvement of the prior relatively unjustifiable internal equity relationships should win the day for expansion.

This simple approach is ideal to deftly dodge the awful compounding effects that drain off the majority of merit budget distributions to senior incumbents who are outperformed by starving junior peers. Without the eminently "fair" grant of identical actual raises to same-grade identical performers, higher-performing juniors will never catch up with the already grossly overpaid seniors who can coast while the beginners who outdo them without equal reward consequences will soon despair and quit.

[E. James (Jim) Brennan]

P.S. At the end, I should have said "will face continuing wider gaps" rather than "catch up".

Identical absolute merit amounts for performance of identical quality in identically graded/valued jobs will never close the existing lead of longer-service ("senior") peers, but it will reduce the wildly disproportionate merit budget allocations to historically more prosperous peers. The gap earned by many years of consistently high performance remains, rightly... but it is NOT right to continue to "over-reward" them just because they have accrued a base income at a high compa-ratio level.

Divert more cash to the young struggling new high achievers rather than starve them to stuff more in the pockets of the already-rich.

That's the basic selling point. It can be clearly and simply demonstrated by showing Excel projections showing the difference in relative/absolute outcomes between the old method and the new "Brennan" solution over a number of years. I created one such demo years ago.

As others have noted, it also completely eliminates any need for forced performance distributions in any simultaneous merit-based pay distribution. That's the bigger background topic.