Ratings, Differentiation, and the “Merit” Increase

Pay seems like it should be straightforward. The employee gets a performance rating – a 5-point scale is typical. Then the employee gets a “merit” increase based on the rating.

Unfortunately, it’s not straight forward at all.

Let’s start with ratings.

Because ratings are a number they have the appearance of objectivity. But they’re no more objective than the grades you got in school.

Here are two reasons why the ratings you get at work aren’t objective:

1. Managers are biased like everyone else. There are easy graders and hard graders.
2. It’s usually easy to identify excellent performance and poor performance – the extremes. But it can be difficult to identify best overall performance from among the remaining employees. As an article from McKinsey Quarterly puts it:
   

   “… attempts to determine who is a shade better or worse yield meaningless information for managers and do little to improve performance … it’s a fool’s errand to identify and quantify shades of differential performance among the majority of employees, who do a good job but are not among the few stars.”

Nevertheless, managers are asked to make these subtle and largely meaningless distinctions.

It’s called differentiation.

Essentially, differentiation means that the manager must compare the performance of all his/her direct reports and determine who should receive more than the salary pool as a “merit increase” and who should receive less or none. That means the method for determining pay can end up driving the rating.

Theoretically, there is money to pay more than the salary pool to the better performers because the poorer performers receive less than the pool or no “merit” raise at all. But the problem is those “shades of differential performance.” Some people are better at some things sometimes; other people are better at other things sometimes.

The tool the manager has to solve the differentiation problem is the performance review. It doesn’t help.

• The performance review focuses on the individual not on how he or she compares with peers.
• Usually there is no common standard for the manager to compare the performance of employees.
• Many performance reviews include assessment of things like personal characteristics, knowledge, behavior, living the company values, contribution to the organization, etc. Presumably, each contributes to performance; how much probably varies by individual. But giving each a numerical value, a rating, doesn’t mean there is a common standard to determine the rating or that the ratings of something like living the company values is comparable to ratings of actual performance.

But once there is a rating, it’s treated the same as any other rating. What appears to be an “objective comparison” is even less “objective.”

Put this together and we have a system that invites “judgment calls” and manipulation.

For example:

• If there’s not enough money, managers can and do rotate who gets the “merit” increase. This means at least one person may be rated artificially low.
• Forced ranking and differentiation often force managers to give the rating and pay they have to give. So does senior management mandating lower ratings to provide “documentation” for a planned reduction in force.
• And of course, if managers are motivated by a desire to avoid a difficult discussion, they figure out how to give the rating they want to give.

“Merit” Increase

And so we come to the “merit” increase itself. Given the usual focus on pay, what’s interesting is that the actual stakes are so low. One reason for differentiation is to incentivize performance. But with 3% salary pools and the possibility of  receiving only an additional 2%, it’s difficult to find the incentive.

Let’s say you get 4%, 1% more than the pool. With an annual salary of $75,000, 1% = $750. Before taxes and other withholding, that’s $62.50 a month. That’s not nothing, but it’s not much: maybe a tank of gas once or twice a month.

Twice $62.50 for the exceptional performer probably won’t be enough to inspire long-term loyalty or increase that employee’s work ethic. If the employee arrived with a strong work ethic, he or she will bring it to the next job at the next company where the incentives are greater.

In the end, the unspoken goal of this entire process – rating, differentiation, and “merit” pay – seems to be to make sure that some will feel rewarded, though it is minimal and mostly symbolic, and few will feel upset. Not much of an outcome for so much work.

Recommendations

• Research shows that simply eliminating ratings does not work out well. Instead of eliminating ratings, install a structured process in which managers conduct performance discussions with their employees every few weeks. These discussions should include discussions of what the rating would be. The focus is solely on performance not pay. These discussions should be required of all managers. Simply suggesting or encouraging managers to have these discussions rarely is effective. See this post for more detail.
• Eliminate the “merit” pay process. It has little to do with “merit.” Salary pools are small, “merit” increases are negligible, and differentiation is a nice theory that doesn’t work well.
• Give most employees the same increase – for example, the 3% in most salary pools. When individual performance warrants, add a bonus for those in the middle. Add more – certainly more than 2% – for the exceptional performer. Do what is appropriate for the poor performer – less of an increase or no increase. But always explain why and discuss what to do about it.