At the end of a panel at “Startup India Standup India”, adequately named “Disruptive Power of Technology in Financial Inclusion”, the panelists, which included Paytm CEO Vijay Shekhar Sharma, Eko Founder and CEO Abhishek Sinha, and Ispirt’s Sharad Sharma, pledged to make India a cashless economy. That was January 16th, 2016, and nearly 10 months later, Prime Minister Narendra Modi, put disruption into the financial payments space, with the move to remove (and gradually recycle) 86% of the cash in the Indian economy. The government’s narrative surrounding demonetization has changed frequently since then: first it was an attack on black money, then about addressing funding of terrorism, but the latest pitch, for a move that reportedly has seen people die, is that it moves people towards a cashless economy; what Venkaiah Naidu, the Union Minister for Urban Development and Information and Broadcasting, also referred to today as a “Cultural Revolution”, entailing, to quote him, a “behavioural modification”.
But how ready up are we to go cashless? How affordable is it for people to go cashless?
Our contention here is that there is no parity between Cash and Digital Money: a rupee paid by cash is far more convenient for a user, and affords less costs, as compared to a cashless system:
1. Number of citizens on mobile: Not all Indians are mobile, leave alone connected. The latest figures from the Indian telecom regulator TRAI show that, as of 31st July 2016 (pdf): India had a teledensity of 83%, with Bihar, Assam, Madhya Pradesh and Uttar Pradesh with teledensity of less than 70%. While state-wise data for wireless teledensity is not available, it won’t be very different, since most connections in India are wireless. Note that these are number of connections, not users, so you will have to discount this significantly, because many users have multiple SIM cards. Proof: Delhi has a teledensity of 234.77%. Urban wireless teledensity is 148%, and rural is 50.72%. Even of the 1,034.23 million connections, 88.88% are active.
So do we have enough people with mobile connections in India? Not close to a billion, especially in rural India.
2. Number of mobile users who are connected to the Internet: There were 342.65 million Internet connections by the end of March 2016, of which 20.44 million were wired connections. In total, 149.75 million were on broadband (3G + 4G + wireline broadband) and 192.9 million on “Narrowband”. Narrowband Internet subscriber base was 192.90 million (2G and wireline broadband). Click here for statewise broadband and narrowband data.
For the top four telecom operators, the number of mobile connections that are data enabled in a manner that is more than 1mb or more than 10mb per month is around 30%.
So do we have enough people with mobile connections in India who have an Internet connection? Not close to a billion.
3. How many people are online daily? These are connections. How many are online monthly? According to Facebook India MD Umang Bedi, 165 million log on to Facebook on a monthly basis. How many go online daily? Only the telecom operators know.
So do we have enough people with mobile connections in India who have an Internet connection and are active online daily? Not close to a billion.
4. Availability of reliable connectivity: “When we were doing Aadhaar,” TRAI Chairman RS Sharma said at an event WiFi in India, around two months ago, “and we said it will be an online infrastructure and identity. People said you are creating an online identity in a situation where connectivity doesn’t exist. So there was a huge amount of pressure on us to make it work offline as well. Our view was that we are creating a future-proof identity infrastructure. We don’t want an infrastructure which becomes useless tomorrow. The future is online. The future is a connected world.”But the future isn’t now, and RS Sharma knows that: “Today with Aadhaar,” he added, “I keep getting complaints that there isn’t a tower in a place and therefore we weren’t able to authenticate. Therefore, connectivity is a very very serious problem.”
If you drive from place to place, especially beyond the national highways: state highways, through pot-holed and dusty roads connecting villages and towns, to small places in the hills, you’ll find connectivity sparse and fleeting. A month ago, I had to drive 15 km to find a spot where there was sufficient connectivity to respond to a message. Pangot in Uttarakhand, where I was staying (earlier this month), has a post-office, but barely functional data from Idea, and none from Airtel, BSNL or Jio. I checked all. The fact is that in cities we’re spoilt by the connectivity we get, and we really shouldn’t judge the quality of connectivity by what high ARPU, urban, concentrated, areas get. Let’s also not forget the connectivity issues in cities, typified by call drops (though these have reduced now).
So do we have enough people with mobile connections in India who have an Internet connection, a reliable Internet connection, and are active online daily? Not close to a billion.
5. Availability of user devices: According to Idea Cellular CEO Himanshu Kapania, there are currently over a billion mobile phones in India: around 850 million feature/smartphones, and 150 million LTE enabled phones.
Vodafone’s smartphone estimate:
Airtel India MD and CEO Gopal Vittal recently said that: “What we have found is that people with smart phones, not all of them use data. That number of people with a smart phone using data is probably around 60% to 70%. And the new smart phones that are coming are not necessarily the smart phones that people are naturally using data”…”one of the big jobs for us is to actually get all of those smart phone devices to also consume data.”
This is important because the most payments business don’t support feature phones (do any?), and while we will see smartphone penetration grow, and we need feature phone support for now. Mobikwik yesterday has promised it, apart from rolling out a low bandwidth application.
So do we have enough people with mobile connections in India who have an Internet connection, a reliable Internet connection, with handsets that support Indian payment apps, and are active online daily? Not close to a billion.
6. Merchant acceptance: India had 712.5 million debit cards, and 130.53 million transactions, as of August 2016. That’s around 18 transactions for every 100 cards. Credit Cards? Only 26.38 million in India as of August 2016, accounting for 83.95 million transactions.
Demonetization might lend itself to greater utilization of cards, but there were only 1,461,672 point of sales machines in India, as of August 2016, according to the RBI. In the entire country. In all likelihood, concentrated in major cities, with some merchants with have more than one machine, as backup. A 2015 Ernst and Young report said that India has the dubious honour of having one of the lowest POS terminal penetration, with only 693 machines per million. Brazil had 32,995 terminals per million people and China and Russia, had around 4000 terminals per million people.
So we don’t have enough people with debit or credit cards, and not enough PoS machines to use them. This is apart from not enough people with mobile connections in India who have an Internet connection, a reliable Internet connection, with handsets that support Indian payment apps, and are active online daily? Not close to a billion.
7. Payment and mobile network capacity: What we’ve seen with Demonetization and the increase in usage of cards and online payments is that somewhere in the value chain, banks and/or payment gateways were not in a position handle the load. Transactions failed. What we were told was that Visa wasn’t able to handle the load. At present, there isn’t sufficient capacity for the escalation in usage if everyone starts transacting digitally. More importantly, do we have the network capacity to deal with this? What happens in an emergency situation, when networks are down because everyone is trying to call everyone, as we’ve seen previously in India? If you don’t have cash, and there is insufficient connectivity, how will you be able to buy anything, use public transportation etc?
8. Time taken for a transaction: If you’ve driven through a toll booth, or paid for parking, you know that operators keep exact change because they expect notes to come in with a specific denomination. The time taken isn’t usually to tender change for notes, but for printing a receipt. Watch a small shop selling high frequency purchases like mobile recharge cards, candies or cigarettes, and you’ll see that the pace at which they close a transaction with a customer is critical for them: they don’t typically give a bill for each transaction, and that’s a problem when it comes to taxation. But from a user’s perspective, think of the additional time it takes:
8.1 For a card, you need to place it in a PoS machine, get a user to input a PIN, and if there is connectivity, wait for the merchant to get a confirmation before you can leave.
8.2 For digital transactions, you need to get a user to scan a merchant QR code, authenticate with a PIN (ideally). Or, you need the merchant to send a payment link to a customer, for the customer to receive it, open a page, type in details and complete a transaction. Then wait for the merchant to receive a confirmation of the transaction before you can leave. Can you imagine doing this while exiting a parking or at a toll booth? If you’re buying a Rs 2 Pulse candy, imagine the friction involved, as indicated here. The quickest means of payment is an NFC machine, but most phones aren’t NFC enabled in India, nor do merchants accept NFC payments.
9. Security issues: The weakest security link in any transaction is not the technology system, but the user, and their lack of understanding of security issues. To get a sense of this, to withdraw money from ATM’s, some people were giving others their card and PIN numbers. For example: this and this.But there are other risks too: In 2011, it was believed that payment gateway CC Avenue was hacked. HDFC Bank too. Last month, HDFC and Axis Bank were hacked too.
Last month, Surekha Pillai found that her card had been used for international purchases, and had to jump through several hoops, from customer care to customer care, to get transactions reversed. The questions she raised indicate that even a global behemoth like Amazon isn’t geared up to deal with such situations, for a customer who is more informed than your average person:
- Why isn’t Amazon India helping customers resolve these issues instead of making them run around in circles?
- Why isn’t there a two step authentication process for Indian cards used outside India?
- Why does Amazon save credit card details by default?
- Why doesn’t Amazon report suspicious activity like Google does? (for logins from a different country, device, IP address)
- Why aren’t customer care executives trained to deal with issues related to security breaches?
The difference between cash and digital is that cash limits the damage to the the loss of a note or of a number of notes. In digital, the risks are higher: the advantage of wallets was that you could transfer money to them bit by bit, and lower your risk of exposure. That of course doesn’t mean that digital shouldn’t be an option – I’m not saying that – but it shouldn’t necessarily be the only option. It also doesn’t mean that storing all your cash in your house is risk-free, but the move to make India a cashless country increases security risks for all citizens, with each account/wallet company becoming a single point of failure.
10. No privacy with cashless: a switch to cashless means that each and every transaction is tracked and documented. This is great for governance, with taxation, but there is no protection for citizens, as to who owns that data, whom they can share it with, and how it will be utilised. If I’m using a wallet, where is the law that prevents usage of that data for advertising to me? By switching to cashless, you’re not giving users a choice. India doesn’t have a privacy and data protection law, and shamefully enough, the Indian government has gone to court arguing that there isn’t a fundamental right to privacy in the country. To quote the Attorney General of India, representing the Union of India, in August last year:“Violation of privacy doesn’t mean anything because privacy is not a guaranteed right”
Cash offers that relative privacy and anonymity, that the Government of India is trying to deny its citizens. The only cashless currency that affords anonymity is bitcoin.
Update: We’ve done an analysis of what kind of sensitive data Wallet apps are collecting.
11. Language compatibility: Paytm has recently updated their application with some features enabled in Indian languages. Mobikwik has done English and Hindi. PhonePe works in English, Hindi and Tamil. However, most mobile handsets don’t have an Indian language interface, as don’t most applications and services. Ola is available in Indian languages only for drivers, not passengers. Apart from Snapdeal, no ecommerce company tried going the Indian language way. There’s a part of the population in India which still isn’t able to read and write, leave alone being able to read and write English, while we don’t have phones that are are in Indian languages and apps that aren’t in Indian languages. The digital divide here is massive. Physical notes are a visual medium of exchange.
12. Interoperability issues (between payment systems): cash is interchangeable: you don’t need a connection, an application or an account to exchange cash. Here, you have a situation where State Bank of India doesn’t allow payment into a Paytm wallet via netbanking, or wallet to wallet transfer isn’t allowed. There’s the Unified Payments Interface, set up by the bank owned group NPCI, where the Reserve Bank of India has not allowed wallet to wallet transfers. Customers are locked-in whether it is to their bank account (because you need banking systems functional to transfer money) or to their wallets.
13. Cost of transaction:
13.1 Merchant costs: Merchants need a working Internet connection to accept digital payments. They need to pay a monthly rental for a machine, or a smartphone with an application to accept payments. On Credit cards, merchants are charged a merchant discount rate (MDR), an inter-bank exchange fee, of 2.5-1.7% per transaction. On debit cards, they need to pay 0.75% per transaction below Rs 2000 and 1%for transactions above Rs 2,000. For UPI, merchants are charged 0.75% per transaction plus other costs (on par as debit cards.)
13.2 Customer costs: You need a smartphone, an Internet connection and/or have to pay USSD charges (Rs 0.5 per session) and data charges when applicable.
13.3 Cost are applicable when cashless is converted into cash: From an RBI paper on processing costs on cheques and ATMs: “The feedback received from different banks revealed the following – a total cost of Rs.1.95 per Rs.1000/- which excluded the cost of insurance and dispensing cash at ATMs; the cost of dispensing cash through ATMs alone is approximately Rs.17 per transaction; the opportunity loss for holding idle cash would be approximately 9%; the cost per transaction at ATMs ranges from Rs.6.60 to Rs.15.88 in case of fully outsourced operations depending upon the service provider and area of operation.”
So where are we today?
Cash isn’t the same as cashless (digital payments) because:
- Not enough people have mobile connections, an Internet connection (which can survive massive usage in times of emergency), or use it regularly, on a smartphone, which supports all Indian languages, with an application that supports all Indian languages. Internet connectivity isn’t reliable or available or as cheap for users as cash.
- The process of making digital payments in India is not easy, and is time consuming.
- Making digital payments is costlier either for the merchant or the customer, or both.
- Digital payments can lead to major security risks, with adequate processes not in place for easy redressal, for either merchant or customer. Above all, not enough is being done to educate the consumer, the weakest chain in the link.
- Digital payments aren’t a single standard like cash: money in one type of account is not the same as in another type of account, and it is not interoperable, unlike cash.
Here’s the thing: Cash might be more expensive for the government, because of tax evasion, corruption and the need to keep recirculating old, spoilt, currency, and enabling transfers, but digital is very expensive for citizens. What is happening here is a transfer of cost of money from government to citizens, and a massive collection of data.
But does that mean that there shouldn’t be any cashless transactions?
Certainly not. The point is that we’re not ready yet. Many of these issues mentioned above will be addressed one by one: connectivity will (hopefully) improve; indic languages interfaces and operating systems developed, security improved, customer care improved, smartphone prices will come down, but the idea to force people into adopting cashless payments is foolish and unnecessary, when you don’t have the wherewithal to meet the demand at that scale, this quickly. People are hurting, and there are no means of meeting that demand in the near term.
The important thing is to give people choice, and switch people to cashless gradually.
Parity between cash and digital money is probably impossible to achieve, but there are means of getting closer to it: by creating an incentive structure for that switch, and that involves making cash more expensive than cashless, and better enforcement.
Here are two ways in which cashless can be encouraged
- Giving an indirect tax rebate for using cashless methods of payment, which brings parity between cash and cashless. Even online, merchants can be incentivised to charge less for digital payments, and more for cash on delivery.
- Digital Payments businesses have tried their hand with cashbacks, and lower rates for digital purchases have already encouraged digital payments. Incentives could be given to businesses, which they can transfer to customers.
Tell us more alternatives either in the comments, or at firstname.lastname@example.org. Share more issues with us as well, and we’ll update this post with credit.
(With inputs from Shashidhar KJ and Sneha Johari)