Why govts fail in making the right personnel requirement estimates

The recent fuss over the government’s order regarding summoning India Administrative Service (IAS) officers to serve at the Centre actually points to a larger crisis wherein the governments at the Centre and the states will not only have poorly qualified employees, but they will also not have enough people working for them.
 
What they will have in fact is a shortage of competence and an excess of incompetence as well as an overall shortage. This has already happened. The short point, as Brijeshwar Singh, a former IAS officer of the Tamil Nadu cadre, says, is that the Department of Personnel and Training (DoPT) has had no scientific way of calculating future requirements. It’s all guesswork.

 
One example of this inability to forecast happened in 1962 or 1963. Bharat Electronics had invented a copying machine. It was a forerunner to the Xerox copier. Prime Minister Nehru, however, decided not to use it. He asked what would happen to the typists in the government. So the prototype could not be commercialised. But the government had no way of predicting how many typists it would require as its role expanded. So it hired more than it needed. The result was over-staffing at the least productive levels of government. We are still paying for that.
 

Then, in 1986, when Rajiv Gandhi introduced massive technology inputs, the problem became even worse. The government then overcorrected, but at the senior levels. The result is a huge shortage of competent officers and oversupply of incompetent employees.

 
Nor is this only a civilian problem. The inability to predict how many people would be needed over time afflicts the defence services also. As technology improves, how much infantry do you need? Or seamen or pilots? Or anything else? Does the government know, except by making linear projections that a new technology can render irrelevant? It’s only when you look at the pension bill that you get an idea. This is now almost 30 per cent of the defence budget. Clearly, the government didn’t factor in the new technologies for killing the enemy.
 
The private sector, meanwhile, operates on the reverse principle: never employ two people if just one can do the job with longer hours. This often leads to understaffing in key areas, and thus to missed opportunities. Thus, you may have a mechanised production line but because your HR function is given low priority, you also have a persistent mis-allocation of employees. Indeed, even if the HR department is given the resources it needs, it would be very surprising if it used the appropriate forecasting techniques.

 
Brijeshwar Singh’s solution is that the DoPT should use what is called Markov analysis to make the predictions about how many people will be needed. When I told him that someone would have to explain the Markov process to the DoPT, he said “you are a journalist, you do it”. So, here it goes.
 
It is named after the Russian mathematician Andrei Markov who postulated it in 1906. It’s been applied to all sorts of random things since then.
 
To quote Wikipedia: “Markov analysis is… used to forecast the value of a variable whose predicted value is influenced only by its current state, and not by any prior activity. In essence, it predicts a random variable based solely upon the current circumstances surrounding the variable.” Without getting into the mathematical complexities, suffice it to say here that you can use it for forecasting the future behaviour of a variable (or system) whose current state or behaviour is random, meaning it can go any which way. We simply don’t know.

 
The short point is this: whether it is Markov analysis, or some other method for forecasting personnel requirements, the government is way behind in the use of modern techniques. It should catch up rapidly because otherwise it will be in even deeper trouble than it is now.