# Salary Increase Matrix: Impact of Uncertain Performance Appraisal Results

Introduction

This is Part 3 of my articles on the Salary Increase Matrix

As we cannot accurately predict the final results of a performance appraisal exercise, we cannot predetermine the salary increase percentage going into each cell of the salary increase matrix so that we can cap our salary increase costs within our budget.

If the results of the performance appraisal exercise assume a normal distribution, the graph will look something like this.

Source: Lesson 4 Measures of Tendency, Miracosta.edu website

Like the example below, the number of people with the highest performance ratings appear on the right and the number of people with the poorest performance ratings appear on the left.

Source: Evaluating Performance of Employees: Top 13 Traditional Methods written by Shreyas Kammar

The results of the performance ratings will look something like the extreme right vertical column of table below.

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Exceptional 15 persons Above Expectations 35 persons Meets Expectations 48 persons Below Expectations 2 persons Unsatisfactory Nobody 20 persons 40 persons 25 persons 15 persons 100 persons

The total percentage payout in terms of salary increases is 3.7464% as shown by the following table.

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Total 0.2 0.4 0.25 0.15 1.0 Exceptional 0.15 0.03×6.2% =0.1860% 0.06×5.8% =0.3480% 0.0375×5.4% =0.2025% 0.0225×5% =0.1125% 0.8490% Above Expectations 0.35 0.07×4.8% =0.3360% 0.14×4.4% =0.6160% 0.0875×4% =0.3500% 0.0525×3.6% =0.1890% 1.4910% Meets Expectations 0.48 0.096×3.4% =0.3264% 0.192×3% =0.5760% 0.12×2.6% =0.3120% 0.072×2.4% =0.1728% 1.3872% Below Expectations 0.2 0.04×2% =0.0800% 0.08×1.4% =0.1120% 0.05×0% =0% 0.03×0% =0% 0.0192% Unsatisfactory 0 0 0 0 0 0 Total 1.0 3.7464%

Skewed Bell Curve

When we have a lot more people with better performance ratings, the curve will skew to the left (positively skewed). In this case, firms worry more about the increased costs in terms of more people getting higher salary increases.

When we have a lot more people with poor performance ratings, the curve will skew to the right (negatively skewed). In this case, firms worry more about organizational performance and ability to compete in the marketplace.

Source: Lesson 4 Measures of Tendency, Miracosta.edu website

When we have a lot more people with poor performance ratings, the curve will skew to the right.

Source: Lesson 4 Measures of Tendency, Miracosta.edu website

Assumption

Let us assume the payout salary increase percentages are fixed as follows.

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Exceptional 6.2% 5.8% 5.4% 5.0% Above Expectations 4.8% 4.4% 4.0% 3.6% Meets Expectations 3.4% 3.0% 2.6% 2.4% Below Expectations 2.0% 1.4% 0.0% 0.0% Unsatisfactory 0 0 0 0

We are now going to examine what the impact is like when the bell is negatively skewed. We are interested in this rather than a positively skewed bell shape because of the impact on manpower costs.

Example 1

Let us say that the results of this year’s performance appraisal exercise is as follows:

 Number of Staff Percentage of Staff Portion of Staff Exceptional 15 persons 15% 0.15 Above Expectations 48 persons 48% 0.48 Meets Expectations 35 persons 35% 0.35 Below Expectations 2 persons 2% 0.02 Unsatisfactory Nobody Nobody 0 Total 100 persons 100% 1.0

Next, we calculate the payout percentage in terms of salary increases.

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Total 0.2 0.4 0.25 0.15 1.0 Exceptional 0.15 0.03×6.2% =0.1860% 0.06×5.8% =0.3480% 0.0375×5.4% =0.2025% 0.0225×5% =0.1125% 0.8490% Above Expectations 0.48 0.096×4.8% =0.4608% 0.192×4.4% =0.8448% 0.12×4% =0.48% 0.072×3.6% =0.2592% 2.0448% Meets Expectations 0.35 0.07×3.4% =0.238% 0.14×3% =0.42% 0.0875×2.6% =0.2275% 0.0525×2.4% =0.126% 1.0115% Below Expectations 0.2 0.04×2% =0.08% 0.08×1.4% =0.1120% 0.05×0% =0% 0.03×0% =0% 0.0192% Unsatisfactory 0 0 0 0 0 0 Total 1.0 3.9245%

Example 2

 Number of Staff Percentage of Staff Portion of Staff Exceptional 35 persons 35% 0.35 Above Expectations 48 persons 48% 0.48 Meets Expectations 15 persons 15% 0.15 Below Expectations 2 persons 2% 0.02 Unsatisfactory Nobody Nobody 0 Total 100 persons 100% 1.0

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Total 0.2 0.4 0.25 0.15 1.0 Exceptional 0.35 0.07×6.2% =0.434% 0.14×5.8% =0.812% 0.0875×5.4% =0.4725% 0.0525×5% =0.2625% 1.981% Above Expectations 0.48 0.096×4.8% =0.4608% 0.192×4.4% =0.8448% 0.12×4% =0.48% 0.072×3.6% =0.2592% 2.0448% Meets Expectations 0.15 0.03×3.4% =0.102% 0.06×3% =0.18% 0.0375×2.6% =0.0975% 0.0225×2.4% =0.054% 0.4335% Below Expectations 0.2 0.04×2% =0.0800% 0.08×1.4% =0.1120% 0.05×0% =0% 0.03×0% =0% 0.0192% Unsatisfactory 0 0 0 0 0 0 Total 1.0 4.4785%

Example 3

 Number of Staff Percentage of Staff Portion of Staff Exceptional 48 persons 48% 0.48 Above Expectations 35 persons 35% 0.35 Meets Expectations 15 persons 15% 0.15 Below Expectations 2 persons 2% 0.02 Unsatisfactory Nobody Nobody 0 Total 100 persons 100% 1.0

 Quartiles (Q) in Range Performance (P) 1 2 3 4 0.8 to 0.89 0.90 to 0.99 1.00 to 1.09 1.10 to 1.20 Total 0.2 0.4 0.25 0.15 1.0 Exceptional 0.48 0.096×6.2% =0.5952% 0.192×5.8% =1.1136% 0.12×5.4% =0.648% 0.072×5% =0.36% 2.7168% Above Expectations 0.35 0.07×4.8% =0.336% 0.14×4.4% =0.616% 0.0875×4% =0.35% 0.0525×3.6% =0.189% 1.491% Meets Expectations 0.15 0.03×3.4% =0.102% 0.06×3% =0.18% 0.0375×2.6% =0.0975% 0.0225×2.4% =0.054% 0.4335% Below Expectations 0.2 0.04×2% =0.0800% 0.08×1.4% =0.1120% 0.05×0% =0% 0.03×0% =0% 0.0192% Unsatisfactory 0 0 0 0 0 0 Total 1.0 4.6605%

Summary of Results and Observations

 Performance (P) Normal Distribution Example 1 Example 2 Example 3 Exceptional 15 0.8490% 15 0.8490% 35 1.981% 48 2.7168% Above Expectations 35 1.4910% 48 2.0448% 48 2.0448% 35 1.491% Meets Expectations 48 1.3872% 35 1.0115% 15 0.4335% 15 0.4335% Below Expectations 20 0.0192% 20 0.0192% 20 0.0192% 20 0.0192% Unsatisfactory 0 0 0 0 0 0 0 0 Total 3.7464% 3.9245% 4.4785% 4.6605%

It is only logical that when the salary increase percentage throughout the matrix is held constant, but when the number of staff given better performance ratings (Exceptional, Above Expectations) increases, the salary increases will consequentially increase. This means that there is an increasing threat of breaking the budget made for the salary increase.

As the human resource professional, both allowing the performance ratings to be skewed (through no fault of yours) or breaking the budget can bring serious consequences to your job or your performance ratings.

This situation is made worst is you are working in an organization where you report to a council or committee (common organizational structure of not for profit organizations such as clubs, alumini; and  non profit organization such as charities and churches.

Possible Causes

Some of the common causes are:

• Managers uses the salary increase as goodwill to win support from their staff.
• Managers afraid of offending aggressive staff by confronting them with poor but truthful appraisals.
• Managers playing favourites. As the human resource professional, sometimes this person could be your boss, the general manager.
• Discrimination, for example as a male manager bowed over by an attractive female subbordinate. Another example is racial discrimination, which sometimes can be part of disguised group bullying.
• The moderator is not doing his or her job either because it is too much work or being afraid of confronting

You can see these are not errors, as often described in literature but intentional actions.

The PAS is Failing

If you read the previous section; step back and analyze, would you think that the performance appraisal system is fair? You may work so hard but either you have a strict boss or your do not like you, and you are passed off as an average worker. Does it help the business results?

Conclusion

When the distribution of performance ratings becomes positively skewed, the costs from higher salary increases go up.