The virtues of delegating work
- The virtues of delegating work
Posted by: "Sheetal - Karmayog" info@...
Tue Feb 28, 2012 7:57 am (PST)
The virtues of delegating work....P.V.INDIRESAN
Projects should be implemented on the principle of negative feedback control - the higher echelons of an organisation correcting errors after they have occurred rather than issuing precise instructions beforehand.
In Agatha Christie's short story, The Apples of Hesperides, Hercule Poirot observes: "There is a golden rule in life, George, never do anything yourself that others can do for you."
That is a neat explanation of the principle of delegation. In India, such delegation is like the weather: everybody talks about it but nobody does anything about it. It is understandable for the idea to be popular because it will please the many. It is equally understandable for those in authority to reject it because that way they think they will lose the authority to control. That is their view but they are wrong.
Let us consider a simple case of maintaining a road. At the present time, each repair involves enormous pre-control and paper work. For a start, the operating official prepares a note asking for permission to repair a pothole as well as for the funds to do so. That needs the approval of his bosses as well as of officials of the Finance Ministry. When both are obtained, the work is advertised and a committee is appointed to select the "lowest" quotation.
All these activities take a few months. Meanwhile potholes which, like the proverbial rabbits, have enormous capacity for reproduction, would have multiplied, say, by three. So the lowly official asks permission to use the same contract for three times as much work. All the processes have to be repeated once more with perhaps the local politician expecting more than three times his or her share.
By the time, all required approvals have been obtained, the number of potholes has probably increased not three times but seven times. Ultimately, some sort of a compromise is achieved and the price for seven - at the minimum rate - is contracted to repair the now fourteen potholes.
As a result, the potholes are covered with mud and black paint. However, the potholes do not remain idle; they reproduce again. So, the process has to start all over again. Everybody is kept busy; the clerks in the office more than the technicians on the road. A lot of money is spent but nothing is achieved.
Negative feedback control
Then, why not try Poirot's principle and delegate the entire responsibility to the operating official? Will you lose all control? You need not! To appreciate how, I suggest you try an experiment yourself; get two persons to play the game of tailing-the-donkey. To one, you make precise measurements and say how many steps forward and sideways to take, and how far the hand is to be moved. To the other, you merely mention what corrections he should take - a little forward, a little to the right and a little lowers and so on. Who will get it right?
We are talking here of the principle of negative feedback control. It is the system that nature uses to maintain a person's temperature, whether he is travelling in the Sahara or fishing in Greenland. The same principle is used to guide a satellite to explore far, far off Pluto. Feedback control is a universal principle. It operates by correcting errors after they have occurred rather than issuing precise instructions beforehand.
Delegation is the key
Delegation is the issue here; a relatively lowly official takes the final decision on his own or on her own. Then, how does the administration check that work has been done promptly and correctly? He does by feedback control, by inspection after the work - not before - has been completed. Now, the officials can make any number of promises which they may or they may not fulfil. Once the work is done, it has to be good; no official can hide potholes; there is no scope for excuses.
Unfortunately, this principle is not understood by our administrators. Some months ago, I was in a committee designing new institutions. The officials were insistent that a Joint Secretary should be on each board. They would not understand that they would have had better control if, instead of sitting on the board, they inspected the institutions same number of times. Sitting on the board, they shared the responsibility for anything and everything that went wrong. Inspecting afterwards, they had no such responsibility but full authority to check what went wrong; they would have had far better control.
Delhi Metro strategy
Then what is the ideal system? The Central and State governments restrict themselves only to post-inspection and regulation. They also retain the authority to ensure security but at the inter-unit level only. Likewise, they also control transport and communications, once again at the inter-unit level only.
All operations within a unit - be it education, healthcare, municipal services and even internal security - are decentralised. The state and the central authority will, however, perform post-operation inspection rigidly and do so dutifully.
Then, we will know who is working well and who is not. We can promote the good ones and relegate the unsuccessful ones. The Delhi Metro does that: All contractors' bills are paid immediately every two weeks and checked for accuracy in the succeeding two weeks. If the bills are accurate, the bills are paid once again in anticipation; if they are not, the contractor loses the privilege. Result: no contractor makes mistakes!
Will not that kind of decentralisation with autonomy for the operating official and regulation through post-operation inspection function better than the minute financial pre-control that we have at present?
L2, not L1
There is yet another alternative: award contracts not to L1 the lowest price demanded by a contractor but to L2 - the second lowest. I am told that it takes lot more of effort and thought to become L2 than to be L1. L2 gets a little more profit and can be expected to work more diligently. However, that is another story.
(The author is a former Director, IIT, Madras. Responses to indiresan@... and blfeedback@...)