Soulless Debt Engines

My phone lights up with an urgent message: “You are getting into the danger zone. Pay your overdue loan of KSh.11,404.47¹ to KEEP your limit! IDEA: you can get another loan immediately!”

This message was the fifth in a week about an outstanding loan of KSh 10,750.00. An amount of KSh 95.53 had been automatically debited from my savings account, reducing the balance to KSh 10,654.47. As the full sum remained unpaid past the due date, the loan lapsed into overdue status and was rolled over for an additional thirty days. A penalty fee of KSh 750.00 was imposed, raising the total amount owed to KSh 11,404.47.

More messages warning me about losing my borrowing limit followed. Reflecting on these messages five years later evokes distressing memories and the trauma of being trapped in a cycle of debt during my early post-college years. Around when I started my first job, an internship barely covering my expenses, these notifications became regular. I was a newcomer in the city, just settling into a new apartment. Access to digital loans was a quick solution to my money problems without the burden of social debt. I preferred the anonymity offered by apps, bypassing the complexities of interpersonal relationships that come with borrowing money. Little did I realize that this anonymity only bought me time at a premium and a fleeting financial respite. What followed was a relentless cycle of borrowing, repaying, and borrowing anew, all while managing a steady stream of nudges and messages. After a six-month probationary period, I received a performance review at work and a salary increase, letting me escape the cycle. This was my lucky break!

My experience isn’t unique; only about three million² Kenyans are engaged in salaried employment, with most living paycheck to paycheck. Low on money and wages, any unanticipated life demands can lead to a spiraling cycle of borrowing. Reflecting on my past debt cycle, I quickly transition from feelings of shame and trauma towards gratitude for the fortunate combination of chance and choice that has prevented me, at least for some time, from resorting to the high-interest borrowing options. It’s crucial to understand that those who resort to borrowing from these high-interest digital credit apps aren’t doing so out of laziness; rather, the challenges they face are complex and multi-layered, including precarious work conditions, rising costs of living, limited financial and credit literacy, and lack of access to formal bank loans. The accessibility of digital credit is particularly appealing to those working in the informal sector, often involved in casual jobs or running small businesses. Most digital credits are built around their borrowing behavior, and the adverse effects of high-interest rates have hit this group the hardest. I spoke with Victor, a gig worker who recently became one of the four million Kenyans lucky enough to have their debts waived. (Victor asked for his real name to be withheld.)

“I was frequent on Mshwari and Branch until I defaulted and lost my limit. Then, I jumped on Fuliza when it came, but over time, I realized I was paying way more than I borrowed, especially when late payments happened and penalties added up. You think you have a clear plan of paying then the job you were hoping for doesn’t happen. These apps don’t know that.”

Private investors, aid organizations, and the state are significant players in the Kenyan digital credit sector, advocating for the inclusion of marginalized groups, specifically women, youth, and informal workers who lack representation in formal financial circles. The Kenyan government recently launched the Hustler Fund—smaller short-term personal loans with lower interest rates—as one of its main pillars of economic development. Still, the Hustler Fund’s growing loan defaults intensify apprehensions over digital credit amid the emergence of new players in the market.

The business of digital loans relies on credit scoring algorithms to evaluate borrowers’ risk profiles. The algorithms leverage datasets like demographic information, GPS, payment history, social connections, device details, contact lists, and other personal data. Essentially, any information accessible to the service provider becomes a credit risk data point. The process of scoring is ambiguous and has been called a ‘black box algorithm,’ an opaque system where inputs and processes are hidden from the general public. Borrowers have limited ability to influence or contest the credit limit assigned to them. Analysis of how these algorithms classify and make decisions echoes historical patterns of exclusion, with no clear way of pointing out the biases.

Most users have no idea how to improve their chances for better loans, lower interest, and higher limits. As one borrower I spoke to explained, “I don’t know what happened to my limit since COVID, I have not been able to grow it, I don’t know how to, maybe they want me to save and use their apps more, but where is the money to do that?” Another shared how she established a mechanism with her husband to grow their limit together stating, “We use one line (sim card) for borrowing, this makes sure we are growing our limit. If each of us has to grow it alone, we will not get far.” With their shared SIM card they are continuously figuring out how to grow their credit under that reputation ID collectively, an endeavor full of trial and error.

---

Kenya emerged as a pioneer in mobile money after the success of MPESA, a mobile phone-based money transfer service, in 2007. Currently, the three major digital credit services are all operated by Safaricom, the leading telco in Kenya, through their MPESA network. Two are in partnership with local banks and one runs as an overdraft service. These three products have a combined market share of 74%. M-Shwari, launched in 2012, leads at 34%. Since the early days, M-Shwari has been a preferred channel for savings and loans. Fuliza, a continuous overdraft service on MPESA, follows at 25%. Fuliza has a one-off 1% access fee and a daily maintenance fee dependent on loan sizes. KCB-MPESA is the third largest with a market share of 15%. Launched in 2015 to build strategic partnerships, it allows a higher borrowing limit. Accessing these digital credits requires one to have an MPESA account and a Kenyan ID, but creating an account does not always guarantee financial access. For Fuliza, a clause states that “not all are eligible for a loan limit upon activation.” They go further, noting that “upon opting in, one will be assigned a limit that defines the maximum amount of overdraft they can access.” This limit is based on usage and timely repayment of the overdraft. Yet even for those unable to access loan limits, these services have become an omnipresent part of Kenya’s financial system.

Digital credit apps have influenced societal and economic structures by altering credit assessment methods, borrowing habits, and employment dynamics. Nowadays, it’s common for hiring managers to request a Credit Reference Bureau (CRB) clearance certificate to assess the potential risk of employees. A credit rating falls within a range of 200 to 900, with a score below 400 indicating a borrower is a defaulter and a score above 800 demonstrating a high confidence in repayment.

Today, Kenya is on a journey of credit repair, with plans by the Central Bank of Kenya to recategorize the credit scoring model to one that will be price and risk-based rather than imposing blanket denials on defaulters. Hopes of more inclusive financial services abound. The current credit assessment model promises to broaden access to financial services but can exacerbate inequality, but those with limited access to traditional banking systems and minimal digital data still face challenges accessing favorable credit terms. When subjected to credit scoring systems, a new economic class stratification emerges. Borrowers are grouped into high-risk and low-risk profiles, limiting how much they can access. Further, the reliance on credit scores as a determinant for loans and even job opportunities widens the gap between socioeconomic classes, playing an outsized role in more than just money. Individuals with lower credit scores may find themselves locked out of opportunities and forced into high-interest financial options, further perpetuating inequality.

Digital credits further influence the money culture. Fuliza, for instance, subtly encouraged open discussions about loan limits, turning them into a sort of social currency. Some feel lucky to have high limits, while others feel the limits assigned to them are unfair. Reactions to questions I asked about Fuliza limits were mixed. Some felt they deserved a higher limit, considering their frequent use of telco services. Others questioned how their limit had increased significantly even though they no longer required the overdraft, and ended up opting out. This unequal allocation of limits based on service usage,seems to favor those who didn’t need much while shortchanging those reliant on the service. The growing transparency among borrowers on Fuliza limits contrasts with our traditional money culture of secrecy. Digital creditors have capitalized on this by marketing privacy, offering quick downloads and sign-ups, and enabling individuals to navigate financial strains without the need for familial or social support. However, the social relations of money reemerge in different ways, whether it’s through shared bragging messages on Fuliza limits or the unfortunate instances where debt collectors have turned to debt shaming to get money back.

“I felt they should have been more understanding, I had been borrowing and paying on time, until when I lost a job at the beginning of COVID. Now the limit is gone and I don’t think they can give me again,” one borrower said when asked about late repayments after she lost her job. Reviewing some of the messages sent to her, it’s evident how the relationship between the creditor and borrower evolves. The pattern begins with overly friendly messages of enticement, followed by congratulatory messages upon receiving a loan. Then there are friendly reminders or nudges to encourage early payment for a discount on facility fees. Then the tone shifts to a warning if one is at risk of being late with a payment, and in cases of lateness, the messages adopt a stern tone indicating that access has been restricted, the borrower is locked out and the limit seized.

Mobile money adoption in Kenya captured the interest of Silicon Valley investors who were keen to capitalize on this opportunity, funding startups in the space, notably Tala³ and Branch⁴ in the early days. The rapid growth fueled overreliance on debt, encouraged predatory lending, and has resulted in more exclusion than inclusion. Many borrowers overextended themselves and are still suffering the effects of their debt-soiled mobile wallets. How did the borrowers get here? Well, it started with the unfair hand dealt to them. With informal and insecure work, they turned to credit to get by and with time, the debt increased as did interest rates, fees, and penalties. The dependency then has little to do with their borrowing culture but with the operation of digital credits. For most, what was meant to be a temporary relief has become a permanent survival mechanism poweredfueled by soulless debt enginesmachines.

---

In the early days of Mshwari, customer data showed that the main customer was the market woman who would borrow early in the morning and pay in the evening after a day’s work, then borrow again the following day. This worked effectively when borrowing was for the business, and especially a business that had clear repayment plans. Today many borrow for utility needs. While these began as exceptional cases, the narrative was borrowed by many other digital lending startups that came after. This market has expanded further to target other informal entrepreneurs and the youth group. Some people find ways to use digital technologies to their advantage while others feel betrayed, especially when they need to work long hours to repay credit. Today, over 50⁵ fintech apps are in the market even as many get pulled off Google Play. A recent, long-awaited policy change by the Kenyan government focused on protecting consumers from predatory lending required fintech firms to reapply for licensing. Out of 288 lenders, only 10 have passed the licensing, with others awaiting approval.

A subtle rebellion is brewing among clients who are trying to contest the algorithm. A discreet message might be shared before a payment is made: “nitume kwa hii namba” (should I send to this number instead?). It’s a hack for managing money that comes into the MPESA wallet without having the automatic deduction to pay in full immediately, a small hack challenging the system’s rules. Guides like “Here’s how to increase your Fuliza limit” abound. These contentions reveal people finding ways to navigate and defy the restrictions imposed upon them, and trying to challenge the hand dealt to them. Even as creditors devise new ways of debt collection that involve social shaming, a sense of indifference might emerge: “there’s no shame anymore; we do not care.” The assertion is a pushback against the stigma associated with borrowing, a step towards independence from societal judgments, and a sign of growing apathy for debt repayment. Amid this resistance lies the murky territory of loan stacking and serial borrowing that gets debtors deeper into dependency. What sometimes starts as a strategic move to outsmart the system’s limitations leads to loan accumulation, borrowing from one to pay the other. Most times this leads to overextension and becoming high consumers of digital loans. In this resistance, there’s also the simple act of ignoring calls and messages. Debt collectors’ persistent reminders are met with silence, a way for clients to reclaim control in a sector governed by commodified data, debt enginesmachines, and impersonal client relations.

As I reflect on my time spent relying on digital credit, the scoring algorithm appears soulless. While it helped during tough times, the system’s constant nudges and penalties seemed oblivious to life's complexities. Sadly, this isn't just my story; it's happening to many others locked in a dependent relationship with the scoring machine, borrowing for daily needs while in precarious employment states. Lacking contextual understanding, the algorithm turns affordable credit into a fleeting financial respite with high costs, shaping both our financial and credit behaviors while leaving little literacy. If we want to use tech to create a more inclusive financial system, it's time to rethink the digital crediting framework and place humane relations at its core.

Notes

Footnotes

1. 11,404.47 Kenyan Shillings, or approximately 71 USD ↩

2. https://www.knbs.or.ke/wp-content/uploads/2022/05/2022-Economic-Survey1.pdf ↩

3. https://tala.co.ke/ ↩

4. https://branch.co.ke/ ↩

5. https://www.similarweb.com/apps/top/google/store-rank/ke/finance/top-free/ ↩

---