There is a common question, which lingers usually in the memory of management cadre: Can all management information be computerized? Nay, certainly not. It can be only computerized for certain functions as it has its own limitations. What I have found while wading through some of the reports produced in Nepal about computer-based information system that somehow these reports are couched and inferred with the very vague statement somewhat like, This engenders a false impression among the top level that in the imminent future they will have a total rein on their official empire through the use of computer-based information system. Attempts are steered to provide midwifery services to this idea couched and inferred in the so-called computer alchemist report to envelop top management cadre which eventually lead to the impaired means of management information. What has been found that more time is dawdle on drawing lines between top and middle management which is futile as it is very difficult to segregate these two levels with any degree of precision. None has been able to define it meaningfully Attempts to make logical generalizations is an utter waste of time. Instead of dwelling on this approach, it would be worthwhile to spotlight on the following pragmatic approach:

Basically, management functions can be broadly assorted into three segments:
        a) Strategic Planning
        b) Management Control
        c) Operational Control

Strategic planning entails predicting uncertain future events. In the strategic planning, complete precision of results is neither desired nor feasible. The job of strategic planning is of non-repetitive type and requirements of large amount of data is also not compulsive. The handling of large amount of data is only needed in case of statistical analysis when done as a part of the strategic planning for deducing at any particular conclusion. These are the blatant facts of strategic planning in front of our naked eyes then how is it possible to automate all decisions of uncertainty. The only possibility of automating is to analyze those information through computers which consist of a lot of interacting variables. The instance of this would be a business simulation aimed at testing the large-range effects of different courses of action before making a decision of the best course of action. Therefore, it has a certain extent of automation.

    The approach and technique applicable in the operational control can not be exactly replicated in the management control even if the job seems of the same nature. The prominent difference between the operational control and the management control is subtly programming the decisionmaking rules. Decision-making rules can be programmed at operational level as the job of operational level is more of repetitive and routine types. But the same can not be done at the management control level. The most spectacular part of computer is that it can be programmed to the great extent for making routine-type decision-making. Management control decision-making is usually not of routine type. Therefore, it is a gigantic task to think of programming computer. In addition, management control decisions are made only after a long deliberation. Therefore, the speed is the undermining factor.

Almost all gamut of operational control activities can be computerized with the present state-of-the-art technology. Most of the activities of operational control involve a lot of interacting variables and the speed is also the critical factor.Instantaneous information is desirable as any unanticipated change in one segment of the system can affect the other part and the consequent alteration is instantly required. Operational control decisions can be programmed as these decisions are more structured, repetitive and routine type.

It is crystal while dissecting each management functions that all information required for all levels of management can not be programmed with the present state-of-the-art technology. It is unwarranted assumptions that all levels of information requirement can be automated to chum out desired level of information. The question may arise here what would be the consequences of making such assumptions. Firstly, it would decrease the quality of management information even if developed by spending ponderous amount of time and money. Secondly, it will direct more attention towards impaired means of improving management information rather than the better means. Even if we develop a computer-based system for the top level then what would be the question which management will pose to the computer. None of us have a crystal bail to foretell categorically that this would be the question.