Academic studies have endorsed the benefits of demonetization

Several academic studies have now examined the effects of demonetization carefully. To make the correct inference, policymakers need to distinguish between correlation and causality. This subtle, but crucial, distinction is made in our folk tales, where a story is related of a fruit falling from the branch of a tree when a crow comes and sits on it. While the crow thinks it “caused” the fruit to fall, the two events may only be correlated with no causation between them. Other influences such as the fruit being ripe or a gush of wind blowing while the crow sat on the branch may have caused the fruit to fall. To infer that the crow caused this specific event, the effect of confounding events has to be eliminated, which is precisely what careful academic research does.

A study by International Monetary Fund chief economist Gita Gopinath published in The Quarterly Journal of Economics (Chodorow-Reich et al., 2018) acknowledges such confounding factors with the demonetization done by India in late 2016: “…other economic shocks and policies besides demonetization affected the global economy and India specifically during this period. Salient examples include the election of Donald Trump which occurred on the same day as the demonetization announcement, a rise in the global price of crude oil of 60% from January to October 2016.” In fact, the study does not find any aggregate effects on the macro-economy, and thereby relies on differences across districts in the two quarters following demonetization to estimate some costs in the short run.

As a trade-off between costs and benefits is always inherent to any economic policy, the benefits of demonetization determine its net outcome. Several academic studies by scholars at top universities throw careful light on these benefits.

Lahiri (2020) shows that India’s direct-tax-to-gross domestic product (GDP) ratio had been falling steadily from 6.4% in 2008 to 5.4% in 2016—a fall of 1% over the 8-year period. However, this declining trend reversed from 2017 and has steadily increased since then to 6% in 2019. Compared to the trend that prevailed, demonetization has increased the direct tax-to-GDP ratio by 0.2%, 0.8% and 1% amounting to ₹40,000 crore, ₹1.25 trillion and ₹1.89 trillion in direct taxes in 2017, 2018 and 2019, respectively.

Lahiri also describes the effects on the digital economy: “Demonetization of November 2016 caused the volume of digital transactions to shoot up on impact, while simultaneously causing a drop in the volume of traditional transactions… digital transactions have consistently exceeded traditional transactions both in levels and growth rates since 2017.”

In a pair of National University of Singapore working papers, Agarwal et al. (2020) show that demonetization led to a permanent increase in the use of digital transactions, especially among the young. Even two years after the event, those who switched to digital transactions have not returned to cash payments. In a Harvard University working paper, Bandi et al. (2020) find that post-demonetization, customers who switched to digital payments on e-commerce platforms—instead of cash-on-delivery—spend more per transaction and are less likely to return their purchases. In a Northwestern University working paper, Crouzet et al. (2020) find a 60% permanent increase in fintech payments after demonetization.

The permanent increase in digital payments represents a significant move towards formalization of India’s economy. As this phenomenon is more pronounced among the young, who will grow over their career into higher wage earners, this change will persist. As digital economies exhibit significant network effects, these gains from demonetization will enhance the contribution of the formal economy and thereby increase the country’s tax base.

As demonetization was a move against those who had aggregated wealth by stealth, examining the redistributive effect of demonetization is crucial. Chanda and Cook (2020) find that the districts that experienced large increases in deposits were poorer and worse-off on several widely-used socio-economic indicators. Yet, these districts were the ones that recorded higher levels of economic activity in the year-and-a-half that followed. Using a longitudinal survey of household expenditures and incomes, they also find that poorer households had larger increases in expenditures and incomes in the following 18 months. These findings, which carefully control for various confounding factors, thus demonstrate its benefits accruing to the poor.

All these academic studies rely on evidence from large samples and thereby avoid the biases that can result from inferences based on anecdotes. Furthermore, by using careful econometric techniques to control for various confounding factors, they enable robust inference. In contrast, lay inferences tend to rely on anecdotes, which has severe limitations because it may possibly work in the specific setting where it is described but cannot be generalized. The limitations of a lay inference drawn from anecdotes are particularly crucial to appreciate in a country like India, with so much heterogeneity in social, economic and cultural factors. Aggregates and averages drawn from large samples across India are therefore necessary—but not sufficient—for careful inference. The sufficiency condition is satisfied only if various confounding factors are carefully controlled for in studies using large samples.

In sum, while demonetization imposed short-term costs, there have been important benefits via the above-mentioned increase in direct tax collections as a proportion of GDP, and the sharp, persistent rise in digital payments. As these represent long-term benefits from greater formalization, they will continue to accrue in a globalized economy that is undergoing a digitally-enabled Fourth Industrial Revolution.

K.V. Subramanian is chief economic adviser to the Government of India