Analysis of the effects of currency demonetization in India using Bayesian prediction

In the ongoing political battle emerging out of the demonetization of currency notes in India, the ruling party and oppositions are ostensibly defending the interests of the poor. Some politicians are relying on anecdotal evidence to advocate the difficulties being faced by the common man. However, using the large and real sample data from Reserve Bank of India (RBI) and careful analysis for different income categories of people, we infer that the politicians advocating the difficulties faced by the common man are being disingenuous in pushing their claims for political gains. Further, the observation indicates that the long queues seen at banks and post offices is from three sections of the population: (i) the richer folks, who want to exchange their honestly earned savings for new currency, (ii) people who are acting as agents for the dishonest, and (iii) people who are in the category of below poverty line (BPL). This work proposes prediction scheme using Bayesian network to predict the outcome of demonetization, considering several dependencies. The proposed scheme is compared with other prediction approaches. The work reveals that the effect of demonetization of ₹ 500 and ₹ 1000 currency has largely affected BPL population, than the other categories. We further argue that this problem could have reduced the pains of all the sections of the society (e.g, printing ₹ 50 and ₹ 100 notes in plenty) with better planning and accurate predictions by the Government.