But do appraisals improve job performance?
In 1999, the Gallup Organization published a study of 80,000 employees from 400 organizations to determine what creates high performance. The conclusion was that the performance orientation of a workplace is determined by 12 issues. Interestingly, not one of the 12 was related to appraisal or pay.
In 2002, the Corporate Leadership Council surveyed 19,000 employees in 34 companies and 29 countries to mathematically rank the effectiveness of 106 performance drivers (e.g., selection, training, career paths). The most effective driver was "fairness and accuracy of informal feedback." It increased performance by more than 35%.
But the study found that pay ("recognizing and rewarding achievement") had only a small impact -- it improved performance by 4.4%. Appraisals were even worse. For example, semi-annual appraisals reduced performance by 1%.
Yet, in attempting to build a high-performing organization, senior leaders push the appraisal and pay buttons over and over again: "This year, let's change our five-point scale to four points."
At my former employer, the Chief HR Officer (CHRO) decided to fix a persistently low score on the climate question, "People are rewarded according to their job performance."
The first year, the CHRO announced the company would "strengthen" the pay-performance link through mandated distributions (e.g., only 10% can get a 5 rating, etc.). Ironically, after years of stable ratings, the pay-performance score dropped precipitously while the cross-industry average remained stable.
The second year, the CHRO decided he needed to push harder. This time, he took away managers' discretion to allocate bonus dollars and mandated fixed payouts by performance rating. The result was another statistically significant drop in pay-performance satisfaction.
The third year, the CHRO decided that high performers would get a fixed 20% increase and mid and low performers would get nothing.
By year four, 34% of employees felt their pay and performance were linked -- down from a starting point of 47%. With two percentage points considered statistically significant, perception of pay-performance dropped 13 percentage points in four years.
The problem never was pay. The problem was appraisal fairness. When appraisals are not seen as fair, increasing appraisal-pay leverage exacerbates dissatisfaction.
How to Fix Appraisal Fairness
There are two ways to improve appraisal fairness. The first is to use objective measures. The second is to use subjective measures assessed by multiple raters.
Using Objective Measures: In jobs where performance measures are indisputable, hard-wiring appraisal and pay can be very effective.
I once spoke to a successful McDonald's restaurant manager -- a B+ or A- performer. I asked how he rewards his people. He said, "Often, when we have a $1,200 lunch, I buy a pizza and we celebrate." Then, I asked one of the top 10 U.S. managers how he rewards his people. He said, "When we have a $1,200 lunch, I buy a pizza and we celebrate."
Those were two completely different answers. The first was telling his people, "When your performance is worthy, in my judgment, I may or may not buy you a pizza." The second said, "$1,200 is the bar. Jump it and get a pizza -- every time. My job is to get you over the bar."
Unfortunately, performance in most jobs is not so clear.
Several years ago, the quarterback of my favorite college team bested numerous school records. But I thought he was a terrible quarterback. He played very well when the stakes were low, but consistently choked when they were high. If this were a business, should pay for quantitative achievements be measured or should we use quantitative measures as evidence to make a fair judgment?
Using Subjective Measures: In jobs where quantitative measures are not clear, create a quantitative score from multiple subjective observations. This is the standard method for customer satisfaction research.
Using 360-degree feedback in an appraisal is one way to turn multiple subjective observations into objective measures. In the CLC study, 360-degree feedback increased performance by 8.1% -- not bad.
Goldman Sachs uses 360-degree appraisals for all employees. While 360s for all seems administratively impossible, it seems to work at Goldman Sachs. It's the only company I know where employees consistently tell me the appraisal system improves performance.
Manage Throughout the Year: The most important method for ensuring fairness and improving performance is to keep managers and employees talking throughout the year.
Of CLC's 106 performance drivers, six of the top eight were related to communication between managers and employees.
1. Fairness and accuracy of informal feedback
3. Emphasis in formal review on performance strengths
4. Employee understanding of performance standards
5. Internal communication
6. Manager knowledgeable about performance
8. Feedback that helps employees do their jobs better
How can you improve manager performance on the six? It's very simple -- hold monthly business reviews.
Business reviews are one-hour one-on-ones between managers and employees to report on progress to last month's commitments, to set next month's commitments and to make adjustments in annual performance objectives.
A good business review will directly address each of the six most effective performance drivers. Sure, once a month certainly is not enough, but for most companies this simple practice will be a big leap forward.
Managers who say they are too busy to spend an hour each month with their direct reports are not managers; they are individual contributors and need to be replaced by managers.
This article is commentary by an independent contributor, separate from TheStreet's regular news coverage.