P.C. Mahalanobis: The Statistician Who Built India's First National Sample Survey

Prasanta Chandra Mahalanobis remains one of the most consequential figures in the development of modern statistics in the Global South. While household surveys are now taken for granted as instruments of public policy, in 1950 India launched the world's first continuous national sample survey under his direction, a system that quietly reshaped how developing nations measured themselves. His career offers a rare example of how a single researcher can influence both the technical and bureaucratic architecture of a country.

For readers familiar with the Australian Bureau of Statistics and the national Census, Mahalanobis's ambition will feel surprisingly familiar. The parallel between Canberra's statistical culture and Kolkata's ISI is often noted by demographers at the Australian National University, where Indian-trained statisticians have long collaborated on population research projects across South Asia. Indian-Australian academic networks, particularly in Melbourne and Sydney, have kept this intellectual conversation active for decades.

A Formative Education in Physics and Mathematics

Mahalanobis was born in 1893 in Calcutta into a Bengali family that valued both literature and science. He studied physics at Presidency College and later crossed the Atlantic to study at King's College Cambridge, where exposure to biometricians like Karl Pearson shifted his interest toward statistics. A brief stop at the Cavendish Laboratory left him convinced that measurement, when applied rigorously to social phenomena, could be just as illuminating as in the physical sciences.

Returning to India in the early 1920s, he took up a teaching post at Presidency College and began publishing papers on anthropometric measurements and meteorological data. These early studies, modest in scale, planted the seeds for what would become a sweeping institutional project.

The Indian Statistical Institute Takes Shape

In 1931 Mahalanobis founded the Indian Statistical Institute in a single room in Calcutta, with a handful of students and almost no funding. Within a decade the institute had attracted international attention for its work on sample design, and it had begun training a generation of Indian statisticians. Visiting scholars from universities in Melbourne and Sydney in the late 1940s often remarked on the institute's distinctive blend of theoretical rigour and field-level pragmatism.

The institute's growth coincided with India's push toward planned development. Mahalanobis served as an advisor to Prime Minister Nehru and helped draft the statistical framework for the country's early industrial planning. His insistence that policy must rest on measured evidence, rather than ideology, set a tone that shaped several decades of public administration.

Designing the National Sample Survey

The launch of the National Sample Survey in 1950 was Mahalanobis's most ambitious undertaking. Drawing on techniques he had refined for studying floods in Bengal and crop yields across rural districts, he designed a rotating panel of households whose data would feed directly into planning ministries in New Delhi. The method allowed India to estimate employment, consumption, and land use at a fraction of the cost of a full census.

Feature Mahalanobis's NSS (1950) Australian Bureau of Statistics Household Surveys UK Office for National Statistics Continuous Survey
Frequency Continuous, with annual rounds Annual and quarterly cycles Quarterly rolling panels
Primary use Planning and poverty estimation Policy and labour market analysis Economic monitoring
Sampling innovation Multistage stratified design Probability-proportional-to-size Rotating panel design
International influence Model for many post-colonial statistical offices Model for Pacific Island partners Model across the Commonwealth

The comparison shows how a method developed in post-partition Calcutta travelled outward, leaving traces in statistical offices that operate on very different political and geographical scales.

The Mahalanobis Model and Five-Year Planning

By the mid-1950s Mahalanobis had articulated what became known as the Mahalanobis model, a planning framework that prioritised heavy industry and capital goods in order to maximise long-run economic growth. The model was deeply contested within India, and subsequent planners diluted many of its sharper edges. Yet its core idea, that statistical evidence should arbitrate between competing economic visions, persisted well beyond the second Five-Year Plan. Economists at the University of Queensland have used the model as a case study in how technocratic advice interacts with democratic politics.

Critics at the time, including several colleagues at ISI, argued that the model underweighted agriculture and consumer goods. The debate itself became a teaching example in evidence-based policy courses taught across Australian and South Asian universities.

Recognition, Honours and the Wider World

Mahalanobis was elected a Fellow of the Royal Society in 1945, only the second Indian to receive the honour after Srinivasa Ramanujan. He also served as president of the International Statistical Institute and won the Padma Vibhushan. These honours confirmed a stature that extended well beyond South Asia, and his correspondence with R.A. Fisher, who spent his later years in Adelaide, remains a fascinating record of mid-century statistical thinking.

The international acclaim did not lure him abroad. Mahalanobis chose to remain in India, declining several offers from European and American universities. He continued to direct ISI until his death in 1972, leaving behind an institution that still trains much of the country's statistical workforce and a method of inquiry that quietly underpins much of what the modern data economy now takes for granted.

Studying Mahalanobis Today

For readers drawn to the intersection of data, policy, and national development, Mahalanobis's life offers rich material. A few starting points may help.

The most direct way to engage further is to read Mahalanobis's 1944 paper "On the Generalised Distance in Statistics," which introduced the distance measure still named after him. From there, the natural next step is to open a current issue of Sankhya, the journal he founded, and see how his questions are still being answered.