A PERFORMANCE evaluation is necessary to motivate employees, provide work direction, justify promotions, and improve and reward performance. Ineffective evaluation criteria for public-service departments can drastically hinder service.

Our civil services machinery isn`t performing at its optimal because of poor evaluation criteria. According to a 2007 research paper, 93 per cent of our civil servants confirmed that bureaucracy`s performance had deteriorated over the years. The perception of our civil services by the public is no dif ferent. Do you think most civil servants should be getting stellar evaluations annually when their performance is actually poor? I hear an emphatic `no`. Based on my conversations with various ex and current civil servants, most of them still get better than decent evaluations.

Some aspects of performance evaluations in the civil services are not so relevant to certain modern management practices. For instance, performance should be evaluated against a set of expectations. Many civil servants don`t even possess their own job descriptions, and may not know what`s expected of them. Also, a supervisor`s expectations of an employee should be based on the yearly and quarterly goals of a department which will appear rudderless if expectations and department objectives are absent.

Moreover, a non-transparent culture promotes sycophancy; an annual performance report (ACR) is usually filled arbitrarily out of courtesy and without any objective feedback. Someone who is not capable in his work but adept at pleasing his superiors may manoeuvre matters to get a favourable ACR.

ACRs are tracked on paper. There are transaction costs involved when paper-based ACRs move hither and thither. Also, you cannot be promoted if you lose even one ACR. A simple, lightweight software solution can remedy this problem, since it can archive ACRs. This solution can provide an avenue for analysis and forecasting (by employing data-mining tools). We are actually depriving ourselves of all this great data that we can use to hire, retain, and motivate employees in the future.

Performance evaluations are also not given in a timely way. Submission of an ACR is delayed until it becomes absolutely necessary. A non-performing employee can keep wreaking havoc in the absence of timely feedback. I think performance evaluations should be done twice a year instead of just once.

Moreover, ACRs are filled out by the immediate boss; no feedback of the customer (who could be citizens or members of other departments)is included. The person who is at the receiving end of a civil servant`s services is important for two reasons. First, thewhole point of hiring or retaining a civil servant (who has lifelong perks)is to improve service delivery. Second, in some cases, it`s possible that the employees spend more time with their customers than with their supervisors.

Additionally, my analysis of ACRs suggests that they don`t take into account the time that has lapsed between the activity of the subordinate and the output emerging from that activity. It`s possible the results in a particular year are the fruits of efforts put in years ago, but ACRs don`t reflect that.

Further, in the current evaluation mechanism, supervisors give feedback to their subordinates, but the feedback of the latter isn`t provided to the supervisor. This is problematic, since the supervisor will never know how his subordinates want him to improve. If a supervisor`s attitude is demoralising, it will remain just that. It`s almost impossible that a supervisor is so perfect that there`s no room for improvement. It takes time to get used to bilateral feedback mechanisms, but they dobuild a culture of trust that creates a productive workplace.

Moreover, current evaluation criteria don`t contain output measures, and are the same across different departments. Output measures (eg meeting tax revenue quota) must be taken, instead of evaluating a subordinate on his or her potential. The different natures of departmentswarrant different performance criteria; one template for the ACR shouldn`t be used by all. The tax revenue of ficer cannot be gauged by the same indicators used for the crime investigation of ficer.

Current evaluation criteria also evaluate the subordinate without taking the department`s performance into account. The performance of a subordinate cannot surpass that of the department. It is ironic people might be able to get f avourable performance reviews when the public perception of the department`s performance might be otherwise.

Lastly, with existing performance criteria, we cannot identify poor performers. Without performance indicators and data, we cannot identify good performers, either. This will turn poor employees into worse ones and demoralise good ones. Our evaluation criteria need a serious rethink, and that can`t happen with so much political meddling. • The writer works in the technology sector.

Twitter: @wyounas