In this webinar from the webinar series “Financial Transformation, Credit Markets and Household Debt in Southeast Asia”, Dr Kei Kajisa, Dr Ed Lucio, and Discussant Dr Nina Anchugina examined the new credit arrangement in the Philippines’ informal finance sector: an emerging credit arrangement called “ATM sangla”.
REGIONAL ECONOMIC STUDIES PROGRAMME WEBINAR
Webinar Series 2020-2021: Financial Transformation, Credit Markets and Household Debt in Southeast Asia
Thursday, 17 December 2020 – ISEAS – Yusof Ishak Institute hosted a webinar on “Does a New Informal Credit Arrangement Improve Poor’s Welfare? The Case of Debit Card Pawning in the Philippines”, delivered by Dr Kei Kajisa, a professor of Development Economics at School of International Politics, Economics and Communication, and Dr Ed Lucio, a Data Scientist at ASB Bank Ltd. in Auckland, New Zealand. The speakers were also joined by Dr Nina Anchugina, a Deputy Director and Senior Research Fellow at the Centre for Social Data Analytics (CSDA), Auckland University of Technology, New Zealand.
Dr Kajisa led off the webinar with a summary of the results from the 2019 Financial Inclusion Survey where only 29% of the participants own formal bank accounts, while 71% remain unbanked. Insufficient funds are frequently cited as one of the reasons for not having an account. Informal sources of loan such as from informal lenders, or family and friends, remain the most common source of loans, with a mere 3% of the participants taking out loans from formal banks. As the increase in formal borrowing from 2017 to 2019 remains minimal, this indicates that penetration of the formal financial sector is growing but remains marginal. Delving deeper into the reasons behind, 47% of participants have little savings, with many mentioning that they earn just enough for their expenses. 48% of participants also alluded to the perception of it being difficult to access formal lending channels, with the lack of documentary requirements and collateral being cited as one of the top reasons. As such, this suggests that saving habits remain poor, and that there is a perceived difficulty in accessing formal lending channels.
Examining these results, Dr Kajisa remarked on the significance of informal credit arrangements in the Philippines and outlined the goal to characterise the users of a relatively new, informal credit transaction known as “ATM Sangla”. In essence, ATM Sangla – otherwise known as debit card pawning – is a form of informal lending using debit cards with salary deposits serving as collateral. The borrower surrenders his ATM card together with his passcode to the lender, and the latter withdraws the repayment out of the salary payment on each payday. Such an arrangement is unique, as it differs from existing micro loans with physical collateral and group liability. Moreover, ATM Sangla is a technology-induced credit arrangement where future income streams are collateralized by ATM technology. This mode of informal credit arrangement may affect the borrowers’ welfare – while it may increase credit opportunities, it may also lead to over-borrowing as ATM Sangla provides easy money. Dr Kajisa noted that this very worry was raised by a senator at the congress where there was a concern that the scheme not only has a high collateral, but is also legally vague and prone to abuse.
To enhance the discussion, Dr Kajisa outlined the results of a survey conducted in 3 small scale factories that produced automobile parts in the industrial estate of Laguna province in 2013. 61% of the respondents were male and approximately half of them were married, with 21% of respondents having graduated from high school or lower. 99% of the respondents also reported having their salaries deposited in their bank accounts which were distributed twice a month with an average salary level of PHP7543 fortnightly. In addition, only 26% of respondents stated that they had their own savings account in addition to the salary account. The survey further revealed that 93% of respondents were aware of ATM Sangla and 42% of them have used it at least once. Dr Kajisa observed that unlike microfinancing, ATM Sangla is not a continuous arrangement as evidenced by 46% of the transactions taking place more than a year ago. On the whole, 34% of salaries are deducted with interest rates averaging around 3% in a compounding rate. These loans were mainly used for medical expenditure, living expenses and educational purposes. Once again, Dr Kajisa noted the difference between ATM Sangla and microfinancing in which the latter is used primarily for investment purposes.
Based on a series of questions, Dr Kajisa classified the respondents into 3 groups. The first group of respondents valued the satisfaction of consumption in the near future and may make myopic decisions such as borrowing more to fulfil this consumption. Amongst the first group, the more myopic respondents can be further characterised into sophisticated and naïve respondents. The former group was aware of their bias and attempt to avoid myopic decisions by using commitment devices to control themselves, while the latter was not aware of their bias and remain myopic. The second group of respondents were more patient and appreciated consumption in the future more. The third group of respondents did not value consumption in the present or the future any more than the other. In the survey, it was found that 34%, 22% and 43% of the respondents could be classified into the first, second and third group of respondents, respectively. Comparing these numbers with the national average of the Philippines, United States and India, Dr Kajisa discerned that the proportions of more myopic respondents were quite similar. However, Dr Kajisa also commented that it was quite unique that the number of respondents in the second group who were more patient was higher in the survey and in the Philippines as compared to the United States.
Apart from classifying the respondents, the survey also attempted to understand the respondents’ borrowing behaviour. It was revealed that a larger share of more myopic respondents had outstanding balances from relatives and friends. Furthermore, while the more patient respondents had more outstanding loans at ATM Sangla than others, the relation was not statistically significant.
The apparent ease of ATM Sangla in expanding credit access to more myopic respondents further raises the question of whether this arrangement is merely a channel for more easy money which could lead to overborrowing, or if it can serve as a commitment device for the sophisticated but myopic users. To answer this, Dr Kajisa proceeded to outline the analysis conducted should the more myopic but sophisticated respondents used ATM Sangla as a commitment device. The results indicate that these respondents were not aware of their bias and that ATM Sangla – unlike microfinance – cannot substitute for commitment savings. The latter result can be attributed to 3 reasons: firstly, there are no group meetings in ATM Sangla, secondly, ATM Sangla is not a form of continuous borrowing and thirdly, the loans obtained are not confined to investment purposes. These results lie in sharp contrast with another study on Indian microfinance borrowers where more myopic but sophisticated female participants use microcredit as a commitment device. However, one anomaly exists: a relatively large proportion of the more patient respondents in the second group use ATM Sangla as compared to those in the third group. Dr Kajisa explained that one possible reason could be that the former was aware that they value future consumption more and hence, attempt to use ATM Sangla to intentionally consume now. This is known as future bias which implications is still not well-understood.
Dr Kajisa also introduced an index to measure the temptation of fun consumption which includes smartphone usage, eating at Jollibee restaurants and access to Facebook. When the same analysis is conducted using this index, the results support Dr Kajisa’s hypothesis that the more myopic respondents will be more tempted to spend on fun goods as compared to the more patient ones.
In closing, Dr Kajisa reflected that a significant proportion of respondents are relatively more myopic. Moreover, both the more myopic respondents and those who suffer from future bias are more likely to borrow with ATM Sangla. Among the former, many are likely to be naïve or sophisticated without commitment devices. This is consistent with the analysis used with respect to the consumption index.
The webinar was handed over to Dr Lucio who pointed out that ATM Sangla addresses financing needs by those marginalised by the formal sector which is pertinent to majority of the Philippines. In addition, Dr Lucio emphasised that the family network in the Philippines is extremely strong and is a source of loans. Also, the Philippines has a strong degree of consumerism. As such, the income earned to the price of these goods is very disproportionate.
Dr Anchugina then proceeded to discuss the challenges of experimental design and emphasised the importance of behavioural economics in understanding the tension between current actions and future plans. This has implications outside saving behaviour, which includes understanding addiction and policymaking. Moreover, Dr Ancuhgina discussed some experimental design challenges where people may introduce additional realisations to the experiment when it is incentivised rather than hypothetical. People may also tend to underreport their debt level. This problem is exacerbated by the fact that informal debt may be impossible to declare. Finally, a lack of understanding and accounting of risk attitudes may be another challenge. Dr Anchugina also addressed some points in Dr Kajisa’s results which suggests that informal lending could lead to overborrowing. Medical expenses are cited as the second most common reason for ATM Sangla loans which indicates that respondents might not be protected from negative shocks. Lastly, Dr Anchugina also suggested if the findings can be used for other countries with developing financial systems or even serve as a lesson.
The webinar concluded with Dr Kajisa, Dr Lucio and Dr Anchugina engaging in a question-and-answer session with the participants. Questions answered included the potential for ATM Sangla to be formalised and expanded to a larger borrowing group, the relevance of using big data to extract consumption patterns for loans, as well as comparisons to the Chinese and Brazilian credit system.