Saturday, 7 December 2013

Dumb dumbell performance based assessment

Telekom Malaysia introduced the dumbbell performance based criteria in 2004 when Wahid took over as CEO. I thought it was uniquely Malaysian management practise until yesterday when an NUS professor revealed that they used the same performance criteria since year 2000 and in an even more extreme case where 50% of the assessed will not get any salary increment and bonus.

Telekom Malaysia only make around 10% of the staff without any increment or bonus. I suffered this fate so for the next five years, I was put into the floating pool without any real job placed under the management of a junior human resource manager. It was actually heaven but I don't think it is sustainable. It is better that I be placed in this non-performing pool compared to others who may not be able to survive and I don't mind because we have other means of earning a living. I teach part time at various institutes as well as learn other skills.

At that time I thought it was all due to dishonesty and corruption that led Wahid to implement these dumbbell curved performance based criteria because it is so obviously wrong and stupid, but when I heard about it from an Electrical Engineering educated Professor, now I realise it is just another case of mass idiocy, similar to Malaysians celebrating the formation.independence of Malaysian on the 31st of August 1957 or the people thinking that the earth is flat.

This dumbbell curve is a mathematical formula called normal probability distribution that shows a graph that is shaped like a dumbbell so it is just a simpler name for it. The word normal does not mean that it is the standard way or practise, it is just a name. There are other formulas, that can be used to describe probability distribution which is just a chance of an event in happening. Given a number of people with various performance levels, usually nature tends to follow this normal distribution but the mean, that is the highest part of the curve tends to vary for various circumstances.

For example, in IQ tests,, if the whole world is tabulated into a graph, the median(central part or mean) is 100. If your IQ is above 134, you shall be the top 1% of the total population.

What happens if you are in a group of people with an IQ of above 134? The distribution will still follow the dumbbell curve, except that the median level will be higher depending on the people who are in the sample. The majority will be at the median level, because that is how median is defined, and very few people will be above the median level. This is just nature and we cannot argue against this.

The problem comes when you apply a forced ranking such as anyone below the median will not get any salary increment or bonus whatsoever. This is despite you being the top 1% of the population in that group. In a team where everyone is excellent, with the nobel prize calibre, 50% of them will still be assessed as having no growth contribution. NUS claimed to have an exception, so does Telekom Malaysia. Everyone claims to have an exception, but it will never be implemented in practise.

When you find it necessary to have an exception, you should know that the system itself is fundamentally wrong. It has to be corrected. But this dumbbell performance criteria is not only wrong, it encourages the opposite of the aim. It encourages the reduction of performance instead of encouraging more performance in whatever worthwhile criteria.

In NUS, every lecturer is a PhD holder and must graduate with First Class Honours. If 50% of them cannot contribute anything to NUS, maybe they should certainly contribute better in other Universities. NUS and Telekom Malaysia solves this problem by promising rotations. That professor justifies it by saying that research results come in cycles. So the performance measurement is based on luck instead of performance potential.

If you study the results of the performance assessments, you should notice a pattern of rotations which tend to be evenly distributed so it no longer follows the dumbbell curve distribution which is the nature of things. So you are not following the dumbbell criteria anymore.

Worst, you discriminate against the highest performing worker who should be getting the highest reward for the talent and results that they had contributed. It is just the nature of things that once you are good, you are always good. That is the reality in life. Even in a group of excellent people, some will be consistently better than others. If the results shows otherwise, it shows that the performance measurement is faulty because it goes against the nature of things.

The greatest beneficiary of the rotation system will the be the weakest in the group who will still get the chance for a reward instead of NONE at all if the dumbbell curve is to be followed strictly. The point is, this system encourages less performance instead of more performance.

It is just a mathematial truth that the greatest benefit for the workers is when they all perform to the least of their abilities. They can get the maximum result for the least effort by not performing at all. Companies can introduce overall measurements such as an amount of bonus linked to the overall performance, but when the chance of getting the bonus is so low, maximum effort, especially from the best workers, will not be worthwhile. Human nature is such that, they will still work normally so they will expand their energy somewhere else, instead of the company which does not value their talent and effort. It is certainly advantageous to the worst workers.

The worst thing is that, it destroys team spirits. Do you want to introduce workers who will take your salary increment and bonus? You must be an idiot to do this. If you really have any intelligence and really want to improve, you'd better prepare yourself to leave this institution all the time because all humans want to get to know excellent workers so that we all can improve together, instead of being penalised.

People do improve among excellent people. I have personal experiences with this in my primary and secondary schools with clear results, as well as in Telekom Malaysia but with unclear outcomes. The dumbbell curve is the truth but it is not static. The median level can improve but not its distribution.

The true performance based criteria is still the seniority based criteria because it is certain that humans improve with more experience. Any performance criteria that denies this performance improvement is therefore certainly wrong. It is as clear as daylight and proven for generations and yet people prefer to ignore them.

No comments: