Project – Algorithmic Credit Scoring and Financial Inclusion Among Informal-Sector Entrepreneurs: An Assessment of Selected Digital Lending Platforms in Lagos State.

CHAPTER ONE

INTRODUCTION

Topic: Algorithmic Credit Scoring and Financial Inclusion Among Informal-Sector Entrepreneurs: An Assessment of Selected Digital Lending Platforms in Lagos State

1.1 Background to the Study

Financial inclusion has become an important development objective because access to appropriate financial services enables individuals and businesses to save, make payments, obtain credit, manage financial risks, and participate more effectively in economic activities. For entrepreneurs, access to credit is particularly important because it provides working capital for purchasing goods, replenishing inventory, acquiring equipment, paying employees, and expanding business operations. However, access to formal credit remains difficult for many small businesses operating outside established financial and accounting systems. Informal-sector entrepreneurs may have irregular income, limited formal financial records, inadequate collateral, and little documented borrowing history, making it difficult for conventional lenders to assess their creditworthiness. The World Bank’s Global Findex research identifies access to and use of formal financial services as important dimensions of financial inclusion, while Demirgüç-Kunt, Klapper, Singer, and Ansar (2022) emphasise the role of digital payments and financial services in expanding access and addressing inequalities in financial participation.

Traditional credit-scoring systems generally rely on information such as credit history, verified income, outstanding debt, collateral, and documented repayment behaviour. Although these indicators can help lenders assess the likelihood of repayment, they may be less effective when applicants operate businesses that generate income primarily through cash transactions and maintain limited formal records. Consequently, an entrepreneur may have an active and profitable business but still struggle to qualify for a conventional loan because the business’s financial activities are not sufficiently documented. Stiglitz and Weiss (1981) explain that information asymmetry can lead lenders to ration credit even when borrowers are willing to pay the prevailing interest rate, because lenders cannot perfectly distinguish between borrowers with different levels of risk. Berger and Udell (2006) further explain that small-business lending depends on the availability and quality of information used to assess borrowers, making the absence of reliable records an important barrier to credit access for many smaller enterprises.

The development of financial technology has created opportunities for lenders to assess credit applications using algorithmic credit-scoring systems. Algorithmic credit scoring involves the use of computational rules, statistical models, and, in some cases, machine-learning techniques to evaluate applicant information and estimate creditworthiness or the likelihood of repayment. Depending on the lender and the applicant’s consent, these systems may use conventional financial records alongside alternative data such as digital transaction histories, mobile-phone usage patterns, repayment behaviour, sales information, and other indicators of economic activity. Berg, Burg, Gombović, and Puri (2020) demonstrate that digital footprints can provide useful information for predicting default among online loan applicants, while Jagtiani and Lemieux (2019) discuss how alternative data and machine learning can influence fintech lending and credit assessment. These developments suggest that algorithmic scoring may help lenders evaluate applicants who have limited traditional credit histories, although the effectiveness of a particular model depends on the quality, relevance, and fairness of the data used.

In Nigeria, the growth of digital financial services has created additional channels through which individuals and small-business operators can apply for credit using mobile applications and other online platforms. Digital lenders may process applications, analyse available information, make preliminary lending decisions, and communicate loan offers without requiring applicants to visit a physical branch. For informal-sector entrepreneurs, these arrangements may reduce some of the time and documentation barriers associated with conventional lending. However, the availability of digital credit does not necessarily mean that all entrepreneurs receive affordable, suitable, or sustained access to finance. The World Bank’s Global Findex database provides a framework for assessing financial access, digital payments, and the differences in financial participation across population groups (Demirgüç-Kunt et al., 2022). In Nigeria specifically, Björkegren, Blumenstock, Folajimi-Senjobi, Mauro, and Nair (2025) evaluated a digital credit product and found that access to a loan improved borrowers’ reported subjective well-being after three months, although the interventions did not produce statistically significant improvements in several other welfare outcomes measured in the study. Their findings demonstrate the importance of assessing not merely whether digital credit is available, but also what access means for borrowers in practice.

Algorithmic credit scoring may contribute to financial inclusion by helping lenders process applications at scale and assess applicants whose financial circumstances are not fully represented by conventional credit records. If alternative data accurately capture an entrepreneur’s financial behaviour, an algorithm may identify creditworthy applicants who would otherwise be overlooked by traditional assessment procedures. Faster processing may also reduce application delays and lower some of the administrative costs associated with manual credit evaluation. However, algorithmic systems can also create challenges when applicants do not understand how their data are collected, how their scores are calculated, or why their applications are rejected. Data that are incomplete, unrepresentative, or strongly associated with existing social inequalities may produce unfair assessments and reinforce financial exclusion. Jagtiani and Lemieux (2019) explain the opportunities and challenges associated with alternative data in fintech lending, while Berg et al. (2020) show the predictive value of digital footprints in online lending. These studies support examining both the inclusionary potential and the limitations of algorithmic credit scoring rather than assuming that automated decisions are inherently more accurate or fair than conventional assessments.

The operation of digital lending platforms also raises questions concerning consumer protection, responsible lending, privacy, transparency, and the treatment of borrowers. A platform may provide rapid access to credit while charging costs that are difficult for borrowers to understand, imposing repayment periods that do not match their income patterns, or relying on data practices that borrowers cannot easily evaluate. These concerns are especially relevant to informal-sector entrepreneurs whose earnings may fluctuate from day to day and who may depend on short-term loans to maintain their businesses. In Nigeria, the Federal Competition and Consumer Protection Commission introduced the Limited Interim Regulatory/Registration Framework and Guidelines for Digital Lending in 2022 to address issues associated with digital money-lending services. The Commission subsequently issued the Digital, Electronic, Online, or Non-Traditional Consumer Lending Regulations in 2025, addressing matters including transparency, data privacy, abusive recovery practices, and consumer redress (Federal Competition and Consumer Protection Commission [FCCPC], 2022, 2025). These regulatory developments demonstrate that digital lending must be assessed not only in terms of convenience and approval speed but also in terms of responsible practices and borrowers’ ability to use financial services safely.

Lagos State provides a relevant setting for studying algorithmic credit scoring and financial inclusion because it contains extensive commercial activity, dense markets, trading networks, small enterprises, and a growing digital-financial-services environment. Informal-sector entrepreneurs in the state include market traders, artisans, small-scale distributors, food vendors, repair operators, personal-service providers, and other business owners whose financial activities may not be fully captured by conventional accounting systems. Selected digital lending platforms may use automated assessment methods to determine eligibility, assign credit limits, set loan terms, or decide whether an application should be approved. Nevertheless, the extent to which these scoring processes enable informal-sector entrepreneurs to obtain suitable credit—and whether particular groups encounter persistent barriers—requires empirical investigation. Evidence from Nigerian digital-credit research demonstrates that digital loans have become a meaningful financial channel, while scholarship on fintech lending highlights the importance of alternative data, borrower experience, and consumer protection (Björkegren et al., 2025; Jagtiani & Lemieux, 2019). Accordingly, this study examines how algorithmic credit scoring relates to financial inclusion among informal-sector entrepreneurs using selected digital lending platforms in Lagos State, with attention to access, affordability, timeliness, transparency, and the practical experiences of borrowers.

1.2 Statement of the Problem

Access to finance remains a significant concern for informal-sector entrepreneurs who require working capital to sustain and expand their businesses but may lack the formal documentation expected by conventional lenders. Many of these entrepreneurs conduct transactions in cash, operate without audited financial statements, and have limited formal credit histories. As a result, lenders may find it difficult to assess their repayment capacity using traditional credit-scoring methods, potentially restricting access to loans even for entrepreneurs with viable businesses. Algorithmic credit scoring has emerged as a possible solution because it can analyse different types of information to estimate creditworthiness and support faster lending decisions. However, the availability of these systems does not establish that they have significantly improved financial inclusion among informal-sector entrepreneurs in Lagos State. Stiglitz and Weiss (1981) explain why information asymmetry can lead to credit rationing, while Berg et al. (2020) demonstrate the predictive value of digital footprints in online lending. The central problem is whether algorithmic credit scoring is enabling informal-sector entrepreneurs to obtain credit who might otherwise have been excluded from formal lending.

A second problem concerns the transparency and fairness of algorithmic credit-scoring decisions. Although digital lending platforms may assess applications rapidly, borrowers may not understand the information used to determine their eligibility, credit limits, interest rates, or repayment conditions. Entrepreneurs whose transactions are mainly cash-based or whose digital records are limited may receive lower scores or be rejected despite having the capacity to repay. Similarly, data reflecting existing social or economic inequalities may affect the way applicants are assessed. If these problems are not properly managed, algorithmic systems could reproduce or deepen existing patterns of financial exclusion instead of reducing them. Jagtiani and Lemieux (2019) identify opportunities and challenges associated with alternative data and machine learning in fintech lending, while Berg et al. (2020) demonstrate the predictive usefulness of digital footprints without establishing that every algorithmic model will be fair across all borrower groups. The problem therefore includes determining how entrepreneurs experience credit assessment and whether algorithmic scoring practices are perceived as accessible, transparent, and fair.

A third problem is that access to digital loans does not necessarily translate into meaningful financial inclusion. Financial inclusion involves more than the approval of a loan; it also concerns whether suitable financial products are available, affordable, accessible, and useful to the people who need them. Informal-sector entrepreneurs may obtain short-term digital credit but face high borrowing costs, repayment schedules that do not match their business cash flows, limited opportunities to challenge decisions, or repeated borrowing to meet existing obligations. Such circumstances may limit the developmental benefits of digital lending, even when applications are processed quickly. The Nigerian digital-credit evaluation by Björkegren et al. (2025) found that access to a digital loan improved reported subjective well-being in the short term but did not significantly improve several other measured welfare outcomes. The FCCPC (2025) also introduced regulatory provisions addressing transparency, data privacy, and consumer protection in digital lending. These developments demonstrate the need to assess whether algorithmic credit scoring provides informal-sector entrepreneurs with meaningful and responsible access to credit rather than merely increasing the speed of lending decisions.

The fourth problem concerns the limited contextual evidence required to understand how algorithmic credit scoring relates to financial inclusion among informal-sector entrepreneurs using selected digital lending platforms in Lagos State. The effects of automated credit assessment may vary according to the type of business operated, the availability of digital transaction records, the entrepreneur’s borrowing history, the platform’s lending criteria, and the affordability of its loan products. A platform may increase approval opportunities for some applicants while excluding others whose financial activities are not adequately represented in its data. Existing research on digital footprints and fintech lending provides important insights into credit prediction, while research on Nigerian digital credit highlights the need to distinguish access to loans from broader improvements in borrowers’ welfare (Berg et al., 2020; Björkegren et al., 2025). However, further evidence is needed to establish how users of selected digital lending platforms in Lagos State experience algorithmic credit scoring and how these experiences relate to their financial inclusion. This study therefore seeks to examine the relationship between algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs in selected areas of Lagos State.

1.3 Aim and Objectives of the Study

The main aim of this study is to assess the relationship between algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs using selected digital lending platforms in Lagos State.

The specific objectives are to:

1. Examine the extent to which algorithmic credit scoring facilitates access to digital credit among informal-sector entrepreneurs in selected areas of Lagos State.

2. Assess the influence of algorithmic credit-scoring processes on the accessibility, timeliness, and affordability of credit for informal-sector entrepreneurs.

3. Examine entrepreneurs’ perceptions of the transparency and fairness of algorithmic credit-scoring decisions on selected digital lending platforms.

4. Determine the relationship between algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs in Lagos State.

1.4 Research Questions

The study will be guided by the following research questions:

1. To what extent does algorithmic credit scoring facilitate access to digital credit among informal-sector entrepreneurs in selected areas of Lagos State?

2. How do algorithmic credit-scoring processes influence the accessibility, timeliness, and affordability of credit for informal-sector entrepreneurs?

3. How do informal-sector entrepreneurs perceive the transparency and fairness of algorithmic credit-scoring decisions on selected digital lending platforms?

4. What relationship exists between algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs in Lagos State?

1.5 Research Hypothesis

The following null hypothesis will be tested at the 0.05 level of significance:

H₀: There is no statistically significant relationship between algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs using selected digital lending platforms in Lagos State.

1.6 Significance of the Study

This study is significant because it examines how automated credit-assessment systems may influence access to finance among entrepreneurs who are often underserved by conventional lending arrangements. Its findings may be useful to the following groups.

Informal-sector entrepreneurs: The study may help entrepreneurs understand how digital lending platforms assess credit applications and how their financial records and digital transaction histories may affect lending decisions. The findings may also improve awareness of loan terms, borrowing costs, data privacy, and the importance of responsible credit use.

Digital lending platforms and fintech companies: The study may provide evidence about how users experience algorithmic credit scoring. The findings could help platforms identify difficulties associated with application procedures, credit eligibility, decision transparency, and the treatment of applicants with limited formal financial histories.

Banks and other financial institutions: The study may contribute to discussions on alternative data, automated credit assessment, and the provision of financial services to customers who lack conventional credit records. The findings may help institutions consider how digital assessment methods can complement existing lending procedures.

Government and regulatory agencies: The study may be useful to institutions responsible for financial inclusion, data protection, consumer protection, and digital lending oversight. Its findings could support policies that encourage responsible innovation while safeguarding borrowers against unfair treatment, privacy violations, and misleading loan terms.

Business associations and entrepreneurship-support organisations: Organisations that support traders, artisans, market operators, and other informal-sector entrepreneurs may use the findings to develop financial-literacy programmes and improve awareness of digital credit products, responsible borrowing, and borrowers’ rights.

Researchers and students: The study may contribute to literature on algorithmic credit scoring, financial technology, and financial inclusion in Nigeria. It may also provide a basis for further studies examining algorithmic fairness, alternative credit data, and the financial experiences of underserved business groups.

1.7 Scope of the Study

The study focuses on algorithmic credit scoring and financial inclusion among informal-sector entrepreneurs using selected digital lending platforms in Lagos State.

Content scope: The study covers automated credit assessment, access to digital loans, credit approval and rejection, loan-processing time, affordability, transparency, perceived fairness, and financial inclusion.

Geographical scope: The research will be conducted among informal-sector entrepreneurs in selected areas of Lagos State. The specific locations should be stated in Chapter Three after the sampling framework has been established.

Population scope: The target population comprises informal-sector entrepreneurs who have applied for, received, or attempted to access credit through the selected digital lending platforms. Depending on the research design, the study may include both approved and rejected applicants to avoid examining only the experiences of successful borrowers.

Conceptual scope: Algorithmic credit scoring is the principal explanatory variable, while financial inclusion is the outcome of interest. The study will assess the relationship between credit-scoring processes and access to suitable financial services without assuming that every digital lending platform uses the same scoring methods.

Platform scope: The selected digital lending platforms should be explicitly identified in the methodology. Their inclusion should be based on relevance to the research questions, accessibility of eligible respondents, and appropriate regulatory considerations.

1.8 Limitations of the Study

One potential limitation is limited access to proprietary algorithms. Digital lending platforms may not disclose the exact variables, weights, or computational procedures used to assess creditworthiness. Consequently, the study may rely on borrowers’ experiences and observable lending outcomes rather than directly auditing the underlying algorithms.

A second limitation is respondents’ limited knowledge of credit-scoring procedures. Some entrepreneurs may know whether their applications were approved or rejected without knowing how the decisions were reached. The questionnaire should therefore distinguish between respondents’ direct experiences, their understanding of the process, and their perceptions of fairness.

A third limitation is recall bias and self-reporting. Respondents may not accurately remember application dates, loan terms, interest charges, or rejection reasons. Where feasible and with appropriate consent, the researcher may use loan notifications, application records, or other available documentation to support respondents’ accounts.

A fourth limitation is variation among lending platforms. Platforms may use different scoring models, eligibility requirements, pricing structures, and repayment periods. The study should identify the platforms included and interpret the results with regard to these differences.

Finally, geographical and sampling limitations may restrict the generalisability of the findings. Results from selected entrepreneurs in Lagos State may not represent all informal-sector entrepreneurs in Nigeria. The research should clearly report its sampling procedures and the characteristics of the participants.

1.9 Operational Definition of Terms

Algorithmic Credit Scoring: The use of computational rules, statistical models, or machine-learning techniques to assess an applicant’s creditworthiness and support decisions about loan eligibility, credit limits, pricing, or repayment risk.

Financial Inclusion: The extent to which individuals and businesses can access and use appropriate, affordable, and useful financial services that meet their needs.

Informal-Sector Entrepreneurs: Individuals who own or operate businesses that function partly or substantially outside formal registration, accounting, taxation, or employment arrangements. The specific eligibility criteria will be defined in the methodology.

Digital Lending Platforms: Online or mobile-based services through which applicants can request credit and lenders can assess applications, communicate decisions, and administer loans.

Alternative Data: Information beyond conventional credit reports and formal financial statements that may be used to assess creditworthiness, such as digital payment records, transaction histories, or other relevant behavioural and business information.

Creditworthiness: The estimated ability and willingness of a borrower to repay a loan according to the agreed terms.

Credit Accessibility: The extent to which eligible entrepreneurs can identify, apply for, and obtain suitable credit without unreasonable procedural or eligibility barriers.

Credit Affordability: The extent to which the interest, fees, repayment schedule, and other borrowing conditions are manageable in relation to a borrower’s financial circumstances.

Algorithmic Transparency: The extent to which applicants can obtain understandable information about the factors influencing automated credit decisions and the procedures available for clarification or review.

Algorithmic Fairness: The extent to which credit-scoring systems assess applicants consistently and avoid unjustified discrimination against individuals or groups.

Digital Credit: A loan applied for, assessed, approved, or administered through a digital channel, including a mobile application or online platform.

1.10 Organisation of the Study

The study is organised into five chapters.

Chapter One presents the general introduction, including the background to the study, statement of the problem, aim and objectives, research questions, research hypothesis, significance, scope, limitations, and operational definitions.

Chapter Two reviews relevant literature on algorithmic credit scoring, financial inclusion, digital lending, alternative data, and informal-sector entrepreneurship. It presents the theoretical framework, reviews empirical studies, and identifies the research gap.

Chapter Three explains the research methodology, including the research design, population, sample size, sampling techniques, research instrument, validity and reliability procedures, methods of data collection, and techniques for analysing the data.

Chapter Four presents and analyses the data collected from respondents, answers the research questions, tests the hypothesis, and discusses the findings in relation to the objectives and existing literature.

Chapter Five presents the summary of findings, conclusion, recommendations, limitations of the study, and suggestions for further research.

Project – Algorithmic Credit Scoring and Financial Inclusion Among Informal-Sector Entrepreneurs: An Assessment of Selected Digital Lending Platforms in Lagos State.