Mobile Money and Financial Inclusion in India

This is an analytical paper that I submitted to the '2014 FII Open Data Analysis Challenge'. Data for this paper was made available by Intermedia, which is an independent consultancy specializing in strategic research and evaluation. IMRB International, a leading market research organization, conducted the survey to collect data.

Introduction and Key Question

According to a recent World Bank report [i], thirty percent of the Indian population is still below poverty line. Worse, according to the same report, the extreme poor in India has gone up from 22% in 1981 to 33% in 2010 meaning there is much to be accomplished in bringing formal financial services to the unbanked and under-banked poor in India. Given that Mobile money is considered a vehicle for financial inclusion [ii], [iii], and India boasts of over ninety percent mobile phone subscribers (in terms of population), with thirty seven percent of them in rural areas [iv], when the country entered the mobile money market with much fanfare in 2006, it seemed destined for success. But surprisingly, as of today, mobile money service in India is severely lagging behind even smaller countries like Bangladesh and Pakistan in terms of the amount that flows through digital network [v].

Around the same time in 2007, Safaricom (registered trademark), Kenya’s leading Mobile Network Operator (MNO), launched M-Pesa mobile money transfer platform. This preeminent digital money service has been a phenomenal success and has put Kenya at the forefront of mobile money (27 percent of Kenya’s GDP flows through M-Pesa). It has since inspired many developing economies to experiment with the service on varying scales. Several reasons have been attributed to this success: one of them is that with the then political unrest, state-owned banks did not enjoy the trust of the common people who desperately needed to send money from the cities to villages (Safaricom’s slogan was “Send Money Home”); another reason is the absence of regulations which allowed the industry to evolve freely before guidelines were imposed; and a controversial reason was Safaricom’s monopoly in the market which helped it gain an initial stronghold.

The reasons for success of mobile money in Kenya could be perceived as unique to the country. The government and policy makers (including those in India) have realized that the time and economic situation under which M-Pesa took off cannot be replicated elsewhere. Still, some of the deeper questions remain the same, relating to regulation of mobile money industry, partnerships between MNOs and Banks, and if the focus should be on scale or profit margin.

While addressing these policy concerns could make for a successful implementation, this paper tries to answer some fundamental microeconomic questions that may help us identify potential customers and successfully steer mobile money towards financial inclusion. The Intermedia survey data [vi] provided as part of this project samples 45k respondents all over India. In this respect, the key questions about the surveyed population that we seek to address are outlined below:

1. What percentage of the respondents has a Know Your Customer (KYC) ID?
2. How aware are the respondents about mobile money?
3. What are the reasons why a respondent does or does not have a traditional bank account?
4. Who do people trust their money with?

Summary of Findings

The findings of this study are summarized on the basis of mobile money awareness, access to banking facilities, and perception of financial institutions.

More people have KYC id and mobile phone access, but mobile money awareness is low: Over 90% respondents across zones and Progress Out of Poverty Index (PPI) cutoff have a KYC ID. Additionally, over 95% respondents above the cutoff and 82% below have access to mobile phones. But only an average 11% have heard about mobile money services.

Fewer poor people have bank accounts, opened mainly to receive money; those that have one are farther away from banking facilities: 66% above and 43% below cutoff have a bank account. Furthermore, for poor people, ability to receive money from someone is the main reason to have an account. Across all zones, over 95% of bank account holders, both above and below PPI use Over The Counter (OTC) services. Poor people are farther away from OTC and ATM facilities.

State owned institutions enjoy people’s trust more than privately owned institutions: Over 90% respondents across the spectrum would trust state owned banks than private or foreign banks. Around the same percentage said they would trust the Post Office (PO).

Description of Analytical Approach

Data for this paper was made available by Intermedia, which is an independent consultancy specializing in strategic research and evaluation. IMRB International [vii], a leading market research organization, conducted the survey to collect data.

To analyze diversity in geography and culture across India, the data was divided into the following categories:

• Zone – North, East, West, and South
• Because mobile money has almost always been discussed in tandem with financial inclusion, statistics have been split by below and above PPI cutoff

The raw data available in MS Excel readable format was loaded onto R (http://www.r-project.org) for statistical analysis. The raw dataset had approximately 1050 variables and a little more than 45,000 data points. The available weights in the raw dataset were activated using the svydesign command under package survey.

The resulting weight-activated survey data was then used to compute averages across multiple variables including ‘owning a mobile phone’, ‘having a KYC id’, and ‘having a bank account’. These variables were then compared across the levels to find subsets of population to whom mobile money services could be sold.

Finally, data visualization was done using Tableau Desktop version 8.2 where summaries from R were loaded to generate charts and graphs.

Discussion of Results

More people have KYC id and a mobile phone access, but mobile money awareness is low

As seen in Figure 1, though majority of respondents have access to mobile phones, only a small fraction of this total is aware of mobile money services. When prompted, awareness is highest in North at 9%. Numbers are lower for other regions. Combining both spontaneous and prompted recall, only 11% respondents are aware of mobile money service. Regardless of the service provider, better part of respondents has heard about the service either via television or through other relatives, friends, and neighbors.

Figure 1: % population with mobile access for above and below PPI; % overall population that have heard about mobile money service (Prompted Recall)

Table 1: 4-Sample test for equality of proportions

A zone level 4-sample test for equality of proportions without continuity correction at 95% Confidence Level shows if these fractions are significantly different across zones. Null hypothesis is proportions are same across zones. Each metric (shown in plot) has a p-value less than 0.05 denoting that the proportions across zones are significantly different.

Opening a bank account or registering for mobile money in India requires a potential customer to provide KYC documents, the list of which is provided on this ICICI bank website [viii]. ICICI partners with Vodafone to provide M-Pesa mobile money service in India. Data shows that for populations both above and below PPI cutoff, only a very small fraction of respondents did not have any form of KYC ID. As seen in Figure 2, across zones and cutoff, over 90% of respondents have an ID suggesting that identification is not a hurdle.

Figure 2: Comparison of % population with KYC ID for above and below PPI cutoff

Fewer poor people have bank accounts, opened mainly to receive money; those that have one are farther away from banking facilities

Although a huge majority has some form of KYC ID, sizable respondents do not have a personal bank account. Given their financial woes, those below PPI fared worse than those above. In South, 50% respondents below cutoff have a bank account compared to 32% in the East. Over the cutoff, more than 60% have a bank account, with South touching almost 70%.

Figure 3: Population above PPI - % with KYC ID and % with a bank account

Table 2: 2-Sample test for equality of proportions

Results of cutoff level 2-sample test for each zone to check proportion of KYC ID and bank account is shown in Table 2. A p-value less than 0.05 denotes the proportions are  different for each zone, except for South where both groups above and below cutoff have same proportion with KYC ID (p-value of 0.035).

As seen in Table 3, why someone did not own a bank account depends on their poverty level. Respondents below PPI cited ‘No Income’ as the foremost reason. Above cutoff, ‘Do not need an account’ came out as the primary reason. In North, half of those without a bank account felt there was no need for one exposing a financially sound subset of people that feel all forms of financial services are irrelevant. This number is significantly lower for the other regions, with West being the lowest at about 28%.

Table 3: Reasons why respondents do not have a bank account

As seen from figures 4 and 5, for both below and above cutoff, majority of those without income got by with help from family and friends. One surprising detail stands out – percent respondents without income paying job is higher for above cutoff across the board. It is likely these poor people are doing petty jobs that pay them sporadically, while those above cutoff may not entertain such prospects because of dignity. Darker shade indicates higher percent receiving support for maintenance. In South, 82% above cutoff receive support, while the same is 93% for North.

Figure 4: % population above PPI without a job and obtaining support from family - darker shade indicates higher % getting support

Figure 5: % population below PPI without a job and obtaining support from family - darker shade indicates higher % getting support

Figure 6: Respondents that received money from family over last 90 days

Although sizable population with no regular income mentioned that they got support from family and friends, data shows that only a percentage of them are getting that support regularly as shown in Figure 7. These may be family members working in cities and sending home money for support. Percent population that received financial support from family/ friends anywhere from one to six times over the most recent ninety- day period is lowest for South. About 40% people without income in North get family support. East follows at 24% (above) and 16% (below). West is around 13%, and less than 10% poor people in South received regular family support. There is a small group that received support more than six times and is not included in the analysis here.

For those obtaining regular family support (one to six times over 90 day period), it would make sense to incent the sender to switch to mobile money. Safety and convenience need to be the focus for this along with emphasis on reduced charges. Those receiving financial support largely obtain it as cash and usually collect the money themselves or through family/ friends. Based on data, identification will not be a bottleneck when registering new customers. The recipients will be willing to sign up because of their financial needs and perhaps also because a recommendation from someone they know will invariably carry more weight. Some incentives for senders could be:

• Reduced transaction charges if transfers are done at regular intervals
• Incentives for a new customer referral
• Increased airtime 

Above PPI, average amount transferred in South was almost double that of the other regions (INR 4,000 compared to INR 2,000), meaning fewer people transferred more money. For below PPI, East, North and West sample means are fairly representative of the population means as seen by the low Standard Error (SE). For South however, the significantly high SE shows that this number is not representative of the population mean. 

Figure 7: Comparison of average transfer amount in INR for populations above and below PPI

Those respondents not included in the above group are getting by with support from family intermittently. They did not receive any financial support over the three-month period that was surveyed. In the absence of regular financial support, and being excluded from mainstream banking, only a fraction of this total population (~6000) has ever availed a loan hence the statistics are not split by PPI. About 30% of this population in South availed a loan at some point, followed by North at 23% and then East at about 8%. Numbers are negligible for West. In the South, half of these loans came from friends, relatives, and neighbors and one-third from private moneylenders. Over 80% of these loans in the North came from relatives, neighbors and friends. People in South are more inclined to seek out moneylenders when in difficulty. Policy makers should encourage microfinance organizations to provide incentives to these people to use mobile money for loan repayment. Albeit the profit margins of microfinance institutions are extremely small, a temporary reduction in interest rate or an extended repayment period may motivate them to try mobile services. 

Respondents that currently have a bank account cited the following reasons as primary motivation: 

• Safety 
• Savings facilities
• To receive money from Government (to a lesser extent)
• To receive money from another person (to a lesser extent)

Figure 8: Reasons why respondents have a bank account

For account holders above cutoff, savings is a huge motivation to open an account, with West leading at 41%, followed by 33% for both East and North and 26% for South. In the East and North, safety factor is as important at 34%. Safety is a lower priority for those in the West and South at 24% and 11% respectively. 

For those below PPI cutoff, safety and savings features are a concern, albeit to a lesser extent. Given their uncertain financial status, they need a bank account to receive money from a person (P2P) or from government (G2P).

For those with bank account, the mobile money service needs to be as good or better than banking facilities. Across all zones, over 95% of account holders, both above and below PPI use OTC services. Use of ATM services is highest in South at 55% compared to North at 37%. There is a significant difference in ATM usage between above and below PPI cutoff across all zones. 

Figure 9: Comparison of ATM usage for people above and below PPI cutoff

As seen in Table 4, poor people are farther away from branches and ATM locations. But they are still using OTC services as much as the people above cutoff. This shows that rural poor are not banks’ most sought after customers.  

Table 4: Distance to OTC and ATM facilities - % population by PPI and distance band

State owned institutions enjoy people’s trust more than privately owned institutions

Data shows that across the country, more than 90% of respondents would fully trust or rather trust state owned banks than private or foreign banks. Over 90% would fully trust or rather trust the post office indicating strong support for state run institutions. 

Conclusion

Using Intermedia’s survey data, this paper analyzes subsets of population based on their demographics and financial behavior. Specifically, using identification, mobile money awareness, and reasons why people have a bank account, gaps are identified within mobile money awareness, access, and perception. Based on the summary of findings the following distinct marketing strategies and policy initiatives are recommended. 

More people have KYC id and mobile phone access, but mobile money awareness is low: A targeted marketing campaign to educate potential customers about mobile money is necessary with emphasis on safety, ease of use, low costs, and convenience. Because television and existing subscribers haven’t been able to scale up adoption, service providers need to engage agents proactively along with current users in rural areas by providing live demonstrations. Agents should to be rewarded for booking new business, which will establish a competitive network. If mobile money is to become a tool for financial inclusion, KYC ID requirement should be completely waived for the extremely poor. A referral from a family or neighbor should be enough and an upper limit on transfer amount may be set to avoid money laundering. 

Fewer poor people have bank accounts, opened mainly to receive money; those that have one are farther away from banking facilities: While selling mobile money to customers with bank account, safety and full range of financial options available needs to be stressed. For people who provide financial support to their family, it may be an effective marketing strategy to incent the sender to switch to mobile money by providing incentives such as reduced transaction charges, incentives for new referral, and increased airtime. Airtime top-ups and money transfers can be handled through MNO agents. For savings and loan products, banks can plan a “mobile satellite office” with one employee, who will serve a certain rural area by liaising with existing agents. Multiple satellites will be able to provide wide coverage. Banks will thus be able to reach the unbanked rural poor, which is not cost effective for them at present. 

Another strategy could involve microfinance institutions (MFIs) partnering with MNOs to popularize mobile money among the poor. Specifically, incentives such as marginally lower rates and extended repayment period may motivate people to try mobile money services. At the same time, despite small profit margins, MFIs will be able to increase revenue in the long term through wider adoption of mobile money services.

State owned institutions enjoy people’s trust more than privately owned institutions: Department of Posts, which is part of India Post, offers mobile money transfer service in partnership with Bharat Sanchar Nigam Limited (BSNL). The sender needs to physically remit the money at a post office to obtain a PIN and send it to the receiver to be able to transfer funds. Commission for transfer varies depending on the amount. As MNOs and post office are not offering the same kind of service, MNO agents could be housed within a small area of post office to attract potential mobile money customers.

References

[i] The State of the poor: Where are the poor and where are the poorest (http://www.worldbank.org/content/dam/Worldbank/document/State_of_the_poor_paper_April17.pdf)

[ii] Financial Inclusion by Extension of Banking Services - Use of Business Facilitators and Correspondents, Retrieved August 2014, (http://www.rbi.org.in/Scripts/BS_CircularIndexDisplay.aspx?Id=2718)

[iii] Baijlal, H. “Promoting Financial Inclusion: Is Mobile Money the Magic Bullet?” (http://blogs.worldbank.org/psd/promoting-financial-inclusion-is-mobile-money-the-magic-bullet)

[iv] Highlights on Telecom Subscription Data as on March 31st, 2014. Press Release No. 25/2014, Telecom Regulatory Authority of India

[v] Mirani, L. “Why mobile money has failed to take off in India” (http://qz.com/222964)

[vi] InterMedia, Research for Global Development (http://www.intermedia.org/)

[vii] IMRB International (http://www.imrbint.com/)

[viii] ICICI Bank, FAQs, Get to Know m-pesa (http://www.icicibank.com/Personal-Banking/faq/detail.page?identifier=prod-mpesa-20133011015030012)