Source: Ideas and Institutions Issue #15
Analysis
A Tale of Two Payment Systems
Arthur C. Clarke once observed that any sufficiently advanced technology is indistinguishable from magic. While the world of modern technology is full of things that appear conjured by a spell, the experience of making a payment anywhere in the world in an instant is particularly magical. But the systems that enable this magical experience are quite prosaic and complex. In recent years, two systems that have stood out for their success are from India and Brazil.
Unified payment interface (UPI) is an instant payment platform in India. It was made operational in April 2016. In July 2022, it settled about 6.3 billion transactions, with the aggregate transaction value of $132 billion. UPI is reported to have 260 million users, which is a little less than 20 percent of the population. The average user transacts about 24 times a month.
Pix is an instant payment platform in Brazil. Brazil’s central bank Banco Central do Brazil created Pix. In July 2022, more than 2 billion transactions with the aggregate transaction value of about $181 billion were settled through it. It had 132 million users, including more than 122 million natural persons. So, the number of users is more than half of the population of Brazil, and the average user did about 15 transactions a month. The platform was launched in November 2020, and this scale has been achieved in twenty-one months.
Comparative successes
Both systems are remarkable successes. They seem to have improved the efficiency and ease of usage for digital payments. Since digital payment is one of the foundations of the modern digital economy, the benefits of such innovations go beyond the immediate efficiency improvements. Many innovative business models can be built because such instant, low-cost payment systems are available.
While both are successes, Pix has done better in terms of usage penetration and the volume and value of transactions. Brazil’s population is less than a sixth of India’s, but the number of users on Pix is about half of that on UPI. While the transactions per user are more for UPI, the total number of transactions on the platform are more for Pix if you scale for population. Also, the transaction value as percentage of GDP is higher for Pix. In 2021, India’s GDP was about double than that of Brazil in nominal terms and almost triple in purchasing power parity terms, but the aggregate transaction value for UPI is much smaller than that for Pix. More importantly, Pix has scaled faster than UPI. While UPI has achieved its scale in more than six years, Pix has done so in less than two years.
Similarities between Pix and UPI
Both systems reduce the time for initiating the transaction by allowing users to simply enter a key to address the payee, instead of entering the details of bank account number, bank routing number, etc. This is possible because the key is linked to a directory where information is stored. This protects the privacy of the recipients while also drastically reducing the transaction time. By enabling direct transfer from one account to another, both systems have reduced the number of intermediaries, thereby lowering the acceptance cost for payments. In both these systems, information and funds flow together. So, clearing and settlement happen at the same time, in an instant. Both systems enable their users to send or receive payment transfers at any time.
Both the systems allow open access, albeit not in the same way. UPI allows separation of the customer interface from account holding. So, an account with a bank can be transacted through third-party applications like Google Pay. This unbundling is a major innovation that has allowed UPI to grow rapidly. Pix allows non-bank accounts to directly participate in the system based on access rules, which are risk-based and publicly disclosed. Although Pix initially did not allow third-party applications to initiate transactions, under its Phase 3, it has allowed third parties to initiate payment transactions based on instruction from a user whose account is held with another financial institution.
Both Pix and UPI are initiatives under broader central bank-led strategies to open up the financial system. In India, UPI and account aggregators are among the initiatives being implemented under this strategy. Brazil is also actively pursuing various initiatives to break barriers in the financial system. While Pix is its flagship open payments initiative, it is also pursuing open banking. Brazil’s central bank governor Roberto Campos Neto stated that the Pix system could usher in the end of credit card use.
Differences between Pix and UPI
The histories and design choices of Pix and UPI also diverge in certain ways. The main aspect of divergence is in the use of government or central bank intervention to promote digital payments.
Creative destruction sometimes requires inducements from the policymakers as the incumbents in a market may have little incentive to pursue disruptive innovation. The emergence of instant payment platforms like UPI and Pix as competitors to traditional payment platforms like credit cards and prepaid wallets can be good for the system. However, since state power is coercive in nature, the inducements need to be precisely directed toward solving problems that require state intervention.
In Brazil, the central bank has focused on designing Pix as an efficient and secure platform and making it easily available as an alternative to traditional payment platforms. Neither the government nor the central bank has used much coercion to promote Pix in particular or digital payments in general. Although the central bank did prohibit charging customers directly, merchants are charged fees. There is no cap on the fees, and it is left to the market to determine the fees. Although the central bank has mandated all large financial institutions (more than 500,000 customers) to offer Pix as an option, there is no mandate on businesses to accept it. Still, the power of the innovation is such that there is widespread adoption.
In India, the government has banned any fees to be charged from customers or merchants for UPI. The government has also made it mandatory for businesses with more than Rs. 500 million (about $6 million) annual sales to accept UPI payment, with a penalty to be enforced by the joint commissioner of income tax for each day of non-compliance. This was done in 2019 through an amendment to the Payment and Settlement Systems Act that was included in the finance bill, which was a money bill. The amendment did not seem to pertain to the matters listed in Article 110 of the Constitution, which defines the scope of money bills. This suggests that the government was willing to take a legal risk to impose these restrictions. Further, the government and its surrogates have often defended the demonetization decision of November 2016, in which about 86 percent of the currency in circulation was cancelled, by showing the increase in digital payments since the decision was taken.
Occam’s razor of public policy
The story of Pix comes close to serving as a counterfactual for what might have happened with UPI without the use of shock (demonetization) and coercion (banning fees and mandating businesses). Pix has achieved as much, and perhaps more, success than UPI in less than a third of the time. If an innovation constitutes a considerable improvement, it can grow without such interventions. Considering the creditable innovation and system-building that have gone into creating UPI, the innovation could have scaled on its own accord.
The UPI-related restrictions and mandates also have other consequences. They impede further innovation in the payments market because the commercial incentives to invest have been considerably weakened. They also concentrate market power in the hands of the National Payments Corporation of India (NPCI), which runs the UPI platform. The policymakers seem to be assuming that NPCI will always do a good job and that there is no need to promote further competition in the sector. Experience shows that such market power creates problems sooner or later. In the long run, a dominant player has little incentive to provide effective grievance redressal, pursue further fundamental innovations, invest in high performance systems, give fair terms to stakeholders, and so on.
The economist Ajay Shah once stated the Occam’s razor of public policy as: “When two alternative tools yield the same outcome, we should prefer the one which uses the least coercion.”
This is a useful principle to apply when a policy decision is being made. It can also be applied to reflect on the decision after it has been made—to understand whether a less coercive path could have been chosen. It seems clear that for all the talk about the necessity of shock and coercion for promoting digital payments, the example of Pix has shown that perhaps a path with much less coercion could also have worked. India should pay heed to this example and take steps to dilute the restrictions and mandates.
—By Suyash Rai
Review
Of Censorship, Repression, and Literature
How does censorship shape literature and expression, other than in its most direct form of disallowing some of it?
There are two contemporary trends that should sensitize us to the importance of understanding the different forms censorship can take and the differing ways in which censors can perform their job even within seemingly similar institutional structures. The first is the debate on whether to make social media platforms responsible for regulating speech, and the second is the increasing trend to advocate social media platforms for the removal of politically objectionable content and the associated “cancel culture” targeting individuals. While the regulation of speech on social media is unlikely to be exactly the same as the censorship practiced against books and newsprint, the motivations and behavior of stakeholders are likely to remain the same.
In his book Censors at Work: How States Shaped Literature (2014), Robert Darnton provides three richly detailed case studies of state censorship under regimes with different ideological objectives and different levels of bureaucratization: pre-revolution Bourbon France, British India after the 1857 Mutiny, and East Germany in the 1980s. Through each study, Darnton first highlights how censorship always involves a struggle between free speech on the one hand and state and non-state elites on the other. The second aspect he highlights in each case is the numerous ways in which communication can be constrained.
The most interesting parts of Darnton’s analysis are of his exploration of the positive aspects of censorship, most clearly exemplified in pre-revolution France. Bourbon France treated publication as a privilege, and every legally published book had to have royal sanction. In dispensing this sanction, the French censors endorsed the book’s quality. Censorship was not merely negative, designed to remove objectionable phrases and material. Censors provided “approbations,” as Darnton points out:
“One censor . . . remarked in his approbation, “I had pleasure in reading it; it is full of fascinating things.” Another, who was a professor of botany and medicine, stressed the book’s usefulness for travelers, merchants, and students of natural history . . .”
Censors could either provide approbations of this nature, providing tacit permissions to publish work that did not deserve a positive endorsement, or refuse to sanction its publication. Censors worked hard—many corrected errors and spellings and “often objected to harshness of tone, defending an ideal of moderation and propriety.”
Darnton, however, notes that most controversial books on issues like religion and politics were actually not submitted for censorship, but in fact printed outside French borders and then sold illegally within France. Censors were hesitant to publish even the books that were “defenses of orthodoxy.” Such defenses had to be convincing enough:
“One censor rejected a pious attempt to refute deism on the grounds that it was too feeble: ‘To produce a weak defense of religion is inadvertently to expose it . . .’”
The main concern for the censors was that manuscripts submitted to them had text that would either explicitly or implicitly upset powerful families and lineages in the French nobility. An endorsement of a potentially controversial book could spell disaster for the censor, in addition to the author and publisher.
Censorship in pre-revolution France was therefore both repressive and elevating. It shaped literature not just by doing what our understanding of censorship is—preventing publication of objectionable literature. But because publication was a royal privilege, censorship also involved improving the text, correcting its deficiencies, and endorsing its quality.
In British India, on the other hand, censorship was exercised through a mixture of surveillance and repression. After 1857, the British Indian state sought to understand Indian culture and prepared extensive, confidential catalogues of all literature published:
“. . . the new Indian Civil Service (ICS) began to keep track of it. The paper trail in the archives of the ICS begins with occasional ‘returns’ about the output of books. Then it leads through quarterly ‘catalogues’ . . . and it culminates in annual ‘reports,’ which quantified and analyzed book production in each presidency or province.”
Between 1868 and 1905, the catalogue covered 200,000 titles—more than the total output in France during the Age of Enlightenment. Even more pernicious and dangerous from the British point of view was the spread of literature through plays and other dramatic performances. Their catalogues enabled them to surveil vernacular literature to watch for danger. Through networks of spies, the state kept tabs on dramatic performances, literary sessions, and schools. This came especially handy in the aftermath of the unpopular partition of Bengal province in 1905. Extremist and revolutionary movements spiked as a reaction to the partition, and the leading individuals wrote frequently, drawing inspiration from literature.
The British state adopted two tactics—“repression by the police and prosecution in the courts.” Laws were passed in the service of such tactics—the Newspapers Act of 1908 and the Indian Press Act of 1910 both sought to prosecute sedition.
In contrast to Bourbon France, both the ideological objectives of censorship and the manner in which censorship took place were different. The chief aim of censorship in British India was to maintain Britain’s imperial rule and prevent nationalist literature from causing Indians to rebel. Censorship was practiced ex post, by raids on publishers and prosecution for libel and seditious writing, as opposed to ex ante censorship in Bourbon France. Darnton puts this difference down to the ideological tensions embedded in liberal imperialism. The British wanted to be seen as the champions of free speech and expression on the one hand. On the other, their rule in India depended in part on their ability to restrict the growth of nationalist writing.
Censorship in East Germany (GDR) under Communist rule was perhaps the clearest expression of ideological control. Darnton interviewed two GDR censors about how they saw their work of censorship:
“What, then, was censorship as he had practiced it? He answered with a single word: ‘Planning.’ In a socialist system, he explained, literature was planned like everything else, and to demonstrate the point, he reached into a drawer and handed me a remarkable document entitled ‘Subject Plan 1990: Literature of the GDR.’”
All literature in the GDR had to conform to the annual plan. In 1989, for example, the 40th year of the GDR, historical novels would be focused on anti-fascism, and novels set in the present would conform to the “principle of socialist realism.”
From Darnton’s book, one can derive self-censorship to be a distinct feature of literature in the GDR. According to Darnton, the combination of totalitarian control and the threat of loss of liberty and livelihood led to most censorship taking place in the heads of writers and then on the part of the editors in publishing houses. The censors interviewed by Darnton stated that by the time texts reached them, there was very little left to eliminate:
“On average, they said, they rejected only a half dozen of the two hundred or so manuscripts in East German fiction that they examined each year.”
This does not mean that writers were hapless accomplices to the planned literary objectives of the GDR state. They still had a great deal of leeway to negotiate, fight, and plead with editors and publishers over the censorship of texts and passages. Darnton goes to lengths to highlight how much, even the comprehensively planned literature of the GDR was open to negotiation, bargaining, and exploiting loopholes.
Through these detailed ethnographic accounts of censorship in practice under three distinct regimes, Darnton provides fodder for some common conclusions about the nature of censorship and the behavior of writers and censors. One important theme from Darnton’s account of censorship is that for the censors, literature that directly attacked or challenged existing elites and institutions was more dangerous than abstract philosophical treatises. In Bourbon France, censors fretted most about veiled insults at noble families or the nobility. In British India, nationalist literature was seen as seditious. Censors in each of these countries were also immersed in the literature they engaged with and often behaved permissively until they were unable to do so.
This leads to the second important implication of Darnton’s account—the constant negotiation, compromise, and the identification of loopholes that all stakeholders went through to successfully publish literature. Most individuals who published successfully chose not to die on the sword of abstract philosophical ideas and instead published their works through constant engagement with their censors, publishers, and state authorities. When faced with great adversity, it was not merely the power of abstract ideals but the appeal to common sensibilities that allowed literature to flourish.
—By Anirudh Burman