a day ago
National Statistics Day: 5 things about PC Mahalanobis, who could decipher the stories data tells
June 29 is celebrated as National Statistics Day. It is the birth anniversary of Prasanta Chandra Mahalanobis, the father of statistics in India. Known as 'The Professor', Mahalanobis played a key role in setting up the Indian Statistical Institute (ISI) Kolkata and the Planning Commissions of independent India. Among his many gifts, the one that benefitted India the most was the ability to accurately and efficiently read the stories that datasets contain.
Born in Kolkata in 1893 to an influential Brahmo Samaj family, Mahalanobis studied at Presidency College before leaving for London for further studies. However, he came back to India and spent his life in both setting up statistics — the science of collecting and reading data — as an academic discipline, and applying the academic knowledge to practical scenarios.
Here are five snapshots from his illustrious life that will tell you about his contributions, and why you should care about them.
Brain behind India's sample surveys
The National Sample Surveys that India conducts have long been admired the world over for the granular data they capture about households. This information, about income, employment status, health status, consumption habits, etc. then informs policy decisions. It was Mahalanobis who in 1950 helped launch the first such survey.
Nobel laureate economist Angus Deaton with co-author Valerie Kozel wrote in 2005: 'Where Mahalanobis and India led, the rest of the world has followed, so that today, most countries have a recent household income or expenditure survey… Most countries can only envy India in its statistical capacity.'
In a country as huge and diverse as India, it was very difficult to get an idea of the actual living conditions of people across the length and breadth of the country. Mahalanobis did pioneering work in efficient sample surveys, which basically involve collecting information from small representative samples to get an accurate idea of the big picture. For example, interviewing a sample of say 50 families from rural and urban areas, of different religious, caste and economic backgrounds, living in different geographical conditions, can give a reasonably good idea of how the country at large is earning, spending, and living.
India has Mahalanobis to thank for coming up with a method to do this and then implementing that method.
In the world of statistics, Mahalanobis is best known for coming up with the 'Mahalanobis Distance' in 1936, which is still used for understanding and analysing data.
Put very simply, the Mahalanobis Distance tells you how far from the normal or ideal a certain piece of data is. For example, suppose a survey looks at households with comparable incomes in a certain region and the health of the children in these households. All the children are likely to be at different levels of height, weight, haemoglobin, etc. The Mahalanobis distance can look at all these variables — income, health indicators — and flag if a child is an outlier, like being remarkably stunted for an affluent household.
Two chance happenings
Two of the most important decisions in Mahalanobis's life were made due to a missed train and a delayed boat.
In 1913, after graduating from Calcutta's Presidency College, he went to England to study BSc. at the University of London. He made a trip to Cambridge, where he was awe-struck by the chapel of King's College. He then happened to miss the train back to London, and stayed the night at a friend's house. 'In the friend's house he met a student who was studying at King's College and, hearing that Mahalanobis found the chapel so attractive, suggested he apply to study there. Remarkably, he was interviewed the next day and offered a place,' MacTutor, a mathematicians' biographies collection maintained by the University of St Andrews, Scotland, says.
Mahalanobis's tryst with statistics also happened by chance. His boat to India had been delayed due to World War I, so he had some extra time to spend in the library of King's College. Here, he came across the journal Biometrika. '…fascinated by what he read, he bought a complete set of volumes and took them back to India…He saw that statistics was a new science connected with measurements and their analysis…This was the turning point in his scientific career,' the MacTutor article says about Mahalanobis.
Similar to Biometrika, Mahalanobis started the journal Sankhya in India.
Work in flood prevention
When Mahalanobis passed away in 1972, the great statistician CR Rao wrote his obituary for the Indian National Science Academy, and talked about Mahalanobis's early work related to floods in Bengal and Odisha.
In 1922, a flood devastated North Bengal. The government was considering building expensive retarding basins to hold up the flood waters when the problem was referred to Mahalanobis.
'A statistical study of rainfall and floods extending over a period of 50 years showed that the proposed retarding basins would be of no value in controlling floods in North Bengal. The real need was improvement of rapid drainage… Specific remedies were recommended, many of which were implemented and proved effective,' Rao wrote.
Mahalanobis made a similar study for Odisha in 1926, where an expert committee had concluded that the bed of the river Brahmini had risen, leading to floods. The statistical study by Mahalanobis 'covering a period of about sixty years showed that no change had occurred in the river bed, and the construction of dams for holding up of excessive flood water in the upper reaches of the river would provide an effective control,' Rao wrote.
Mahalanobis even provided the first calculations for a dam project on the Mahanadi, 'which formed the basis of the Hirakud Hydroelectric Project inaugurated about thirty years later in 1957,' Rao wrote.
In the 1950s, Mahalanobis was trying hard to get the new American computer, the UNIVAC, to India. In the book 'Planning Democracy', Nikhil Menon, professor at the University of Notre Dame, writes that the Americans were refusing his request because of Mahalanobis's political beliefs.
'India was unable to get a digital computer from the United States through the 1950s partly because of Mahalanobis' reputation as a Soviet sympathiser during the Cold War…The Indian government, like the Professor, remained unaware that the person spearheading the mission to bring computers to India was himself an obstacle,' Menon wrote.
Menon quotes President of Brooklyn College, Harry Gideonse, to describe the 'threat' the Americans believed Mahalanobis was. Gideonse in a report described Mahalanobis as a person of 'exceptional personal charm and broad cultural background'. 'To me, Mahalanobis is far more significant than straight communist propaganda. He has personal and moral authority, apparent integrity, and an impressive command of relevant information. His ideas are in my judgment a direct preparation for an authoritative solution to India's economic problems,' Gideonse wrote.
