This is a question I sometimes ask myself ever since I read about the great Indian statistician and central planner P.C. Mahalanobis: could he and his statistical peers in China have saved the tens of millions who perished during the famine induced by the Great Leap Forward in late 1950s China? It is not an entirely idle question in counterfactual history about what might have been. The man really was in the right time and place and could have made a big difference; and the fact that his Chinese counterparts were undermined and ignored by the Maoists helped explain why and how the famine happened. By the mid-1950s, Chinese state statisticians realised they had collected copious quantities of data on all aspects of life in Maoist China that were also practically useless. Among themselves, some admitted in internal communication: “Useless at the moment of creation!” It is not that communist China lacked talented mathematicians and statisticians. Rather Maoist ideology had caused havoc to the discipline of statistics, sanctioning some methods but not others. Crucially, it was no longer classified as a branch of mathematics, but a social science, subject to the dictates of Marxist-Leninism. This means that chance, probability and randomness, key concepts in statistics, were to be tamed, if not eliminated, under the teleological science of Marxism. Writing in the online magazine Aeon, Harvard historian Arunabh Ghosh has offered an intriguing summary of his new book, Making It Count: Statistics and Statecraft in the Early People’s Republic of China , about Mao’s assault on statistics that had enormous tragic consequences during the Great Leap Forward. During the mid-1950s, to reform statistical and data collection methods and make them useful for state planning, Chinese statisticians turned to the one man in the world who could really help them. Communist Party expels outspoken retired professor over speeches Mahalanobis was not only a brilliant mathematician who made fundamental contributions to statistics that still bear his name, he was also directly involved in the socialist central planning of the newly independent India. That gave him a unique role as both theoretician and practical planner. According to the Encyclopaedia Britannica , within statistics, he was famous for devising a method for the comparison of two data sets that has become known as the Mahalanobis distance. He applied statistics to planning for flood control and invented a method called “fractile graphical analysis” that is used to compare the conditions of different socioeconomic groups. Most importantly for his Chinese counterparts, Mahalanobis had developed pioneering techniques for conducting large-scale sample surveys, including random sampling, to determine acreages and crop yields. Ghosh’s essay gives a sense of the enormous challenges Chinese planners faced at the time: “Nowhere were these problems as pronounced as in the agricultural sector. Already neglected under the prevailing industry-centric orientation, the scale of the agricultural sector and the large variation in terrain, crops and seasons compounded the tendencies to over-report, delay and generate incommensurable data. “The waves of collectivisation that began in 1952 and forced villagers into larger and larger collectives also exacerbated the problem of reliable and standardised data.” Mahalanobis understood the problems the Chinese were facing. According to the Britannica , throughout the 1950s, he helped establish the National Sample Survey and the Central Statistical Organisation to coordinate statistical activities for the Indian government. He was a member of the Planning Commission of India from 1955 to 1967. Like Maoist China, the commission’s second five-year plan focused on the development of heavy industry in India, but it had the advantage of relying on Mahalanobis’ mathematical modelling of the Indian economy, which later became known as the Mahalanobis model. Mahalanobis spent almost a month in Beijing in 1957 to teach his Chinese colleagues his statistical and planning methods, including the use of large-scale random sampling to collect grass-roots data. The Chinese, who hoped to apply his methods especially to the agricultural sector, also visited his institutions in India the following year. Chinese art exhibition hails ‘heroic’ battle against coronavirus The problems they hoped to address were both methodical and institutional. Under the Maoists, as higher levels of the statistical system made increasingly tough demands, lower levels responded with more and more guesstimates. “As these numbers travelled up the chain, from county to province to Beijing,” Ghosh wrote, “they were combined with other estimates, leaving provincial and, eventually, national data with ever larger margins of error.” Mahalanobis’ methods, they were hoping, would cut through the statistical noise and nonsense to get to the really useful data. In 1958, Mao launched the Great Leap Forward. To the dismay of Chinese statistical reformers, random sampling and other statistical methods were to be considered ideologically out of bounds. Ghosh wrote: “True knowledge [according to the Maoists] could be gained only by a detailed, in-person investigation, not through vast exhaustive surveys nor through randomised sampling. The shift left the statistical apparatus with no reliable means to check its own data. “Most tellingly, it contributed to the state’s reduced capacity to ascertain accurately the devastating famine that overtook the countryside starting in 1959. Estimates vary, but most scholars agree that at least 30 million people, and possibly many more, lost their lives by the time the Great Leap Forward ended in 1962.” Ghosh’s research seems to have opened a genuinely new angle on what has been considered the worst man-made famine in history. It is also relevant to our own age, when “big data” is being applied to business, commerce, and state security and surveillance, in China and elsewhere.