03_10191

Access to financial markets, digitization, and formality have been identified as key factors associated with firm performance. This paper presents a detailed descriptive analysis of the income gap among a group of micro-businesses (MBs) in Colombia’s Pacific region, examining how various factors relate to this gap. Using data from MBs participating in a program by Foundation WWB Colombia and applying the Oaxaca-Blinder decomposition, the analysis shows that formal MBs owned by men—with stronger financial inclusion and digital skills—tend to generate higher income than their counterparts. Complementary qualitative fieldwork reveals that micro-entrepreneurs often perceive informal credit as more attractive than bank credit due to its easier access and greater payment flexibility. These findings suggest that policies promoting women’s empowerment, formalization, digital literacy, and access to financial services may contribute to narrowing these income gaps among MBs in Colombian’s Pacific Region.

Resumen

El acceso a mercados financieros, la digitalización y la formalidad han sido identificados como elementos clave, asociados al desempeño empresarial. Con esa premisa, el presente artículo da a conocer un análisis descriptivo de la brecha de ingresos, para un grupo de micronegocios (MN) de la región del Pacífico colombiano, examinando la manera como diferentes factores se asocian con esta brecha. Con información de MN participantes en un programa de la Fundación WWB Colombia, aplicando la descomposición Oaxaca-Blinder, se determinó que los MN formales, liderados por hombres con habilidades digitales e inclusión financiera, tienden a obtener mayores ingresos. Sustentado en un análisis cualitativo, se revela también que los microempresarios perciben el crédito informal más atractivo que el bancario, dada su facilidad y flexibilidad. Estos hallazgos sugieren que políticas dirigidas a fomentar el empoderamiento femenino, la formalización, la alfabetización digital y el acceso a servicios financieros podrían contribuir a cerrar las brechas entre MN del Pacífico colombiano.

Palabras clave: Financiación, competencia digital, sector informal, género, Colombia.

Introduction

The economic crisis triggered by the COVID-19 pandemic in 2020 has had far-reaching global repercussions. This shock affected both demand and supply: household incomes declined, production of goods and services was disrupted, and supply chains were interrupted. However, the impact of the crisis has been uneven. On the supply side, many entrepreneurs were forced to close their businesses permanently due to significant drops in revenue and profit. MBs, in particular, were among the hardest hit, primarily because of limited access to appropriate financing and low levels of digital adoption (Bai et al., 2021; Caballero-Morales, 2021; Khan, 2022).

According to data from the Organisation for Economic Co-operation and Development [OECD] (2021), MBs are overrepresented in sectors with a low probability of implementing social distancing, and many depend directly on commerce. As a result, they were more severely affected when mobility restrictions were introduced. Furthermore, compared to other firms, MBs are less able to capitalize on opportunities generated by the crisis (Belitski et al., 2022), placing them in a disadvantaged position during the pandemic.

Colombia is characterized by a high rate of informality, with a greater incidence of self-employment and MB activity. It is the OECD country with the highest percentage of self-employment (51 %), more than 20 percentage points above the average. According to data from the Micro-business Survey conducted by DANE (2023), Colombia has more than five million MBs that generate around eight million jobs—over a third of total employment. Given the importance of MBs in the Colombian economy and the pandemic’s documented impacts, it is crucial to understand the factors associated with their performance, particularly those that can support economic recovery.

This document examines the relevance of financial inclusion (FI) as a factor that may influence the performance of a group of MBs in the Colombian Pacific region during the COVID-19 pandemic, and its relationship with other characteristics associated with productivity, such as the adoption of digital technologies and the condition of formality. To this end, data from the WWB Colombia Foundation is used as a basis. The Foundation runs programs to monitor and strengthen business capacities in the departments of Valle del Cauca and Cauca. It implemented Plan Reactívate (Reactivation Plan) which supported 3 500 micro-entrepreneurs—primarily women—with the goal of providing training to accelerate post-pandemic recovery. Using decomposition techniques, this study quantifies the extent to which characteristics of both MBs and their owners are associated with the income gap between financially included and non-included MBs, incorporating an intersectional gender perspective into the analysis.

This study contributes to the literature on the influence of financial inclusion (FI) on micro-business performance, highlighting the importance of factors such as digital technologies and formalization in crisis contexts and specific regional settings. The relevance of FI for MBs has been widely addressed in the literature, in part due to the financial constraints commonly faced by these types of businesses (Barajas et al., 2020; Beck et al., 2005). The results suggest that access to financial services enables firms to leverage investments and capitalize on growth opportunities, which positively affects productivity and facilitates the efficient allocation of resources within the business (Clavijo et al., 2020; S. Luo et al., 2022; Y. Luo et al., 2021). However, access to such services depends not only on the characteristics of the MB, but also on those of its owners—gender, for example, plays a key role (Castro et al., 2020).

There are a variety of factors that limit access to financial products for MBs. These include cultural aspects, transaction costs and information asymmetries, and restrictions related to using assets as collateral (Barajas et al., 2020). Compared to larger firms, MBs are often more opaque, as they typically lack audited financial statements that offer a clear view of their performance. Combined with limited assets, this makes it difficult to assess their creditworthiness, leading banks either to avoid lending to them or to charge high interest rates. Additionally, the availability of information and transparency, social capital, and financial education have been shown to influence the financial inclusion of MBs (Behr & Jacob, 2018; Elliehausen & Hannon, 2018; Yoong, 2011). In particular, when owners observe others in their social group engaging with the formal financial system, they may be encouraged to adopt similar practices, thereby increasing their own participation. In turn, these behaviors can be reinforced and strengthened through greater financial knowledge.

The relationship between access to financial services and the performance of MBs became more evident during the pandemic, as conditions led to both temporary and permanent closures and contractions in employment (Bartik et al., 2020; Jin et al., 2022; Sun et al., 2022). However, MBs connected to the formal financial system were better able to mitigate the effects of the pandemic (Banna et al., 2022; Y. Luo et al., 2021; Shen et al., 2020). This was demonstrated in China, where access to credit accounted for the lower impact experienced by firms and their stronger post-pandemic recovery (Kurmann et al., 2021). In this context, public programs played a significant role by facilitating access to financing, loan forgiveness, and cost reductions (Dai et al., 2021).

Moreover, access to financial services is linked to business formalization status and technology adoption, both of which influence the performance and growth of MBs (Akpan et al., 2022; Caballero-Morales, 2021; Castro et al., 2020; Chatterjee et al., 2020; Orser & Riding , 2018). On the one hand, informal businesses have a lower probability of accessing government programs (Guerrero-Amezaga et al., 2022), which may limit their access to financial services. On the other hand, technology adoption facilitates access to financial services through tools such as electronic wallets, digital payments, and other related products.

In particular, the adoption of technologies enables MBs to access a range of financial services, including bank accounts (Hayashi et al., 2024). However, it is not only access to new technologies that matters, but also their effective use. In this regard, the ability to understand and operate digital tools is crucial for leveraging technology to improve MB performance—for instance, through the implementation of digital payments (Nguyen et al., 2024). Consequently, the use of digital channels allows financial institutions to collect information on potential clients, including transaction histories, thereby improving the availability of data for risk assessment. These improvements support more accurate credit scoring, the determination of appropriate interest rates, and the identification and rejection of high-risk applicants (Razavi & Elbahnasawy, 2025).

In the context of the pandemic, more digitized MBs experienced lower levels of impact due to their greater ease in adopting new forms of employment (Belitski et al., 2022; Y. Sun et al., 2022) and a higher propensity for networking (Caballero-Morales, 2021). Akpan et al. (2022) and Krammer (2022) argue that the adoption of technology supports businesses in adapting to new market conditions and accelerates the innovation of business models. Finally, the propensity to adopt digital technologies depends on both the characteristics of the business and those of its owner (Bai et al., 2021; Trinugroho et al., 2022).

This paper also contributes to the literature by adopting an intersectional gender approach. Studies have shown a gender gap in the financial performance of MBs (Fairlie & Robb, 2009; Kiefer et al., 2020; Rosa et al., 1996). In Colombia, the gender gap in FI remains significant: 90.5 % of men have access to some type of financial service, compared to only 84.5 % of women (Banca de las Oportunidades, 2021). This gap has implications for both the labor market and educational attainment. Furthermore, during the pandemic, female-owned MBs were more adversely affected by COVID-19, particularly in developing countries (Li et al., 2016). This outcome can be explained by gender segregation across activities and economic sectors (Tusińska, 2021).

The findings in this document reveal an income gap between MBs in Colombia’s Pacific region during the COVID-19 pandemic, which is associated with differences in the characteristics of the owners and their businesses. First, we found that formal businesses or those that use financial products and services (higher level of FI), as well as those whose owners are more technologically adept, tend to generate, on average, more income. Nonetheless, as businesses grow and mature, digital skills are more likely to become less prominent and thus show a reduced association with the income gap. Finally, the findings indicate the presence of a gender gap in income: women’s MBs tend to generate lower income, which is linked to different barriers they face.

This document is organized into six sections, including this introduction. Section I outlines, in broad terms, the structure and content of the Reactivation Plan. Section II presents the characteristics of the MBs and their owners, along with the indicators constructed for the econometric analysis. Section III describes the empirical strategy used to descriptively analyze the income gap among MBs and the financial preferences of their owners. The estimates and main results are presented in Section IV. Finally, Section V offers the conclusions and final remarks drawn from the study.

I. Description of the Plan Reactívate Program

The economic reactivation program, Plan Reactívate (Reactivation Plan, RP), is a transitional assistance initiative created by Foundation WWB Colombia during the COVID-19 pandemic. It was designed to support micro-entrepreneurs—especially women—in Cali and surrounding municipalities, helping them to resume their economic activities and recover the income they had prior to the crisis. To this end, the Foundation invested more than USD 1.2 million in the development of all the actions included in the plan.

To access the training, entrepreneurs had to meet specific criteria: owning an MB that had been operating for at least six months, being self-employed, relying on the business as their main source of income, holding no more than a technical degree, and possessing basic technological knowledge. The program consisted of six months of blended (in-person and virtual) training, which included meals, as well as technical, financial, and commercial assistance, all aimed at strengthening participants’ decision-making processes.

II. Summary statistics and sample description

Using telephone surveys conducted by Foundation staff, information was collected on the sociodemographic and MB characteristics of 1 575 entrepreneurs located in Cali (34 %) and nearby municipalities (66 %). In particular, the survey gathered information on financial characteristics. FI is a crucial factor for entrepreneurship and MB sustainability, as it promotes investment in productive activities and helps mitigate unexpected shocks. However, measuring FI is complex due to its multidimensional nature. Following Demirguç-Kunt et al. (2017), FI can be classified into four dimensions: payments, savings, credit, and insurance.

The available data allow us to calculate each of these dimensions, except for insurance. For the payments dimension, MBs are classified according to whether they rely exclusively on cash or use digital payment methods. For the credit dimension, two proxies are used: one based on credit holding and another on credit access, depending on the type of financing (formal or informal). For the savings dimension, MBs are classified based on whether or not they have a bank account. Finally, we construct an FI indicator as the simple average of the four variables described above, ranging from 0 to 1.

Differences in FI levels among MBs may be explained by information asymmetries between MBs and financial markets. In this regard, the adoption of technologies by MBs helps reduce risk by providing the financial system with more and better information. As a result, financial institutions can allocate products—particularly loans—more efficiently, as they have access to a broader set of data (Mushtaq et al., 2022). Therefore, FI can be improved through the adoption of technologies.

Given this context, the available information is used to construct a weighted digital skills indicator (DSI). Respondents were asked to self-assess their ability to perform seven digital tasks, with each skill rated on a scale from 1 to 4. Each task was assigned a different weight based on its relevance to MB operations. The DSI is then calculated as the weighted average of the seven tasks and normalized using the min-max method to facilitate interpretation.

After calculating these measures, descriptive statistics (mean and standard deviation) are computed. Table 1 shows that, among the 1 575 MBs surveyed, 83 % are women, with an average age of 39. Nearly half of the sample is either married or identifies with an ethnic group. Although only 29 % have completed technical studies, they exhibit a high level of digital skills (0.86). On average, MB owners maintain some connection to the financial system (0.43), whether through savings or credit products or by using non-cash payment methods.

Regarding the characteristics of the MBs, 85 % are located in low socio-economic areas (strata 1 and 2), particularly outside Cali. They are primarily concentrated in four economic activities: 26 % in food-related businesses (such as street vending or small display cases), 16 % in clothing, 13 % in hairdressing, and 8 % in restaurants. Many of these MBs are informal: only 43 % have a Tax Identification Number (TIN) and just 13 % are registered with a commercial registry. Additionally, these MBs are relatively young, with 54 % having been in operation for three years or less.

While Table 1 presents patterns for the full sample, differences emerge across the groups analyzed in this paper. In this regard, Tables 2-5 report descriptive statistics by subgroup, according to the classifications used in the subsequent sections. Table 2 focuses on the financial characteristics of MBs, beginning with each of the dimensions (Means of payment, Access to credit, Debt holding, and Bank account) and concluding, in the final columns, with subgroup classifications based on the FI indicator. The last columns, labeled Financially included, present statistics for subgroups defined by this indicator: MB owners with an FI score above the median are considered financially included, while those below the median are classified as financially excluded.

Table 1. Summary statistics for owners and MBs (full sample)

Variable	Mean	SD
Women	0.83	0.38
Age (years)	39.08	11.56
Married	0.54	0.50
Ethnic group	0.52	0.35
Technical education	0.29	0.35
DSI	0.86	0.17
FI indicator	0.43	0.28
Cali	0.34	0.50
Socio-economic stratum
Strata 1 and 2	0.85	0.45
Stratum 3	0.15	0.47
Economic activity
Food	0.26	0.44
Clothing/Apparel	0.16	0.36
Hairdressing	0.13	0.34
Restaurants	0.08	0.28
Less than 3 years	0.54	0.50
TIN	0.49	0.50
Commercial registry	0.13	0.33

Notes: (i) Married is equal to 1 if the person is married or in a free union, and 0 otherwise. (ii) Ethnic group takes a value of 1 if the person is not considered mestizo or white, and 0 otherwise. (iii) Technical education is 1 if the owner has attained at least a technical degree, and 0 otherwise. (iv) The DSI and FI indicators range from 0 to 1, with 1 representing MB owners who are fully digitally skilled and financially included, respectively. (v) Cali is 1 if the MB operates in Cali, and 0 if it is in another municipality. (vi) Only owners belonging to socioeconomic strata 1, 2, and 3 participated in the RP program. (vii) Economic activity variables take a value of 1 if the MB belongs to the respective sector, and 0 otherwise. (viii) “Less than 3 years'' is 1 if the MB is three years old or less, and 0 otherwise. (ix) The variables Commercial registry and TIN take a value of 1 if the MB has these legal requirements, and 0 otherwise.

Source: Authors’ calculations based on RP survey.

Table 2. Summary statistics for owners and MBs by financial characteristics

	Means of payment				Access to credit				Debt holding				Bank account				Financially included
	No		Yes		No		Yes		No		Yes		No		Yes		No		Yes
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Women	0.85	0.36	0.76	0.43	0.83	0.37	0.81	0.39	0.84	0.36	0.79	0.40	0.87	0.34	0.82	0.39	0.86	0.34	0.79	0.41
Age (years)	39.82	11.7	36.77	10.84	38.56	11.55	40.2	11.53	38.25	11.71	40.66	11.12	37.61	12.03	39.4	11.44	38.43	11.92	39.67	11.2
Married	0.54	0.50	0.54	0.50	0.53	0.50	0.57	0.50	0.53	0.50	0.57	0.50	0.48	0.05	0.56	0.50	0.52	0.50	0.57	0.50
Ethnic group	0.55	0.50	0.45	0.50	0.54	0.50	0.48	0.50	0.55	0.50	0.48	0.50	0.58	0.49	0.51	0.50	0.57	0.50	0.48	0.50
Technical education	0.24	0.42	0.44	0.50	0.27	0.44	0.33	0.47	0.28	0.45	0.30	0.46	0.21	0.41	0.30	0.46	0.21	0.41	0.35	0.48
DSI	0.84	0.18	0.93	0.11	0.86	0.17	0.86	0.16	0.86	0.17	0.86	0.17	0.84	0.19	0.87	0.16	0.85	0.18	0.87	0.15
Cali	0.30	0.46	0.46	0.50	0.33	0.47	0.36	0.48	0.32	0.47	0.38	0.49	0.36	0.48	0.33	0.47	0.29	0.45	0.39	0.49
Strata
Strata1 and 2	0.89	0.31	0.73	0.44	0.87	0.34	0.82	0.38	0.88	0.32	0.80	0.40	0.88	0.33	0.85	0.36	0.91	0.29	0.80	0.40
Stratum 3	0.11	0.31	0.27	0.44	0.13	0.34	0.18	0.38	0.12	0.32	0.20	0.40	0.12	0.33	0.15	0.36	0.09	0.29	0.20	0.40
Economic activity
Food	0.30	0.46	0.11	0.31	0.26	0.44	0.24	0.43	0.27	0.45	0.22	0.42	0.30	0.46	0.25	0.43	0.31	0.46	0.21	0.40
Clothing/Apparel	0.14	0.35	0.21	0.41	0.15	0.36	0.17	0.38	0.15	0.36	0.17	0.38	0.13	0.33	0.16	0.37	0.13	0.34	0.18	0.38
Hairdressing	0.15	0.35	0.10	0.30	0.13	0.34	0.13	0.34	0.13	0.33	0.15	0.35	0.16	0.37	0.13	0.33	0.15	0.35	0.12	0.33
Restaurants	0.09	0.28	0.07	0.25	0.09	0.28	0.07	0.26	0.08	0.28	0.08	0.27	0.11	0.32	0.08	0.26	0.09	0.29	0.07	0.26
Less than 3 years	0.54	0.50	0.53	0.50	0.58	0.49	0.45	0.50	0.59	0.49	0.44	0.50	0.65	0.48	0.52	0.50	0.62	0.49	0.47	0.50
TIN	0.45	0.50	0.62	0.49	0.47	0.50	0.53	0.50	0.46	0.50	0.54	0.50	0.26	0.44	0.54	0.50	0.40	0.49	0.57	0.50
Commercial registry	0.10	0.30	0.22	0.42	0.10	0.30	0.18	0.39	0.10	0.30	0.19	0.39	0.06	0.24	0.14	0.35	0.07	0.26	0.18	0.38

Notes: (i) Married is equal to 1 if the person is married or in a free union, and 0 otherwise. (ii) Ethnic group takes a value of 1 if the person is not considered mestizo or white, and 0 otherwise. (iii) Technical education is 1 if the owner has attained at least a technical degree, and 0 otherwise. (iv) The DSI indicator ranges from 0 to 1, with 1 representing MB owners who are fully digitally skilled. (v) Cali is 1 if the MB operates in Cali, and 0 if it is in another municipality. (vi) Only owners belonging to socioeconomic strata 1, 2, and 3 participated in the RP program. (vii) “Less than 3 years'' is 1 if the MB is three years old or less, and 0 otherwise. (viii) The variables Commercial registry and TIN take a value of 1 if the MB has these legal requirements, and 0 otherwise. (ix) Economic activity variables take a value of 1 if the MB belongs to the respective sector, and 0 otherwise. (x) Division of sample in the columns labeled “Financially included” is based on whether the MB has a FI indicator below or above the median.

Source: Authors’ calculations based on RP survey.

Among MBs that rely exclusively on cash as a Means of payment, 85 % of the owners are women, compared to 76 % among those who use other payment methods—indicating that women are less likely to adopt digital forms of payment. This pattern holds across the other FI dimensions, suggesting that women are, on average, less integrated into the financial system. Similar disparities appear across other key variables related to FI dimensions and the FI indicator. As with women, owners belonging to ethnic groups tend to be less connected to the financial system: a lower proportion uses digital payment methods (45 %, compared to 55 % who rely on cash), accesses or holds debt (48 % in both cases), and has a bank account (51 %).

This pattern is reflected in the final columns: 57 % of financially excluded owners belong to an ethnic group, while this share falls to 48 % among the financially included. Educational attainment also shows a positive association with FI, as 35 % of financially included owners have a technical diploma, compared to only 21 % among the financially excluded. Finally, formality is strongly correlated with financial characteristics. Owners who use digital payment systems or hold financial products (such as credit or savings accounts) are more likely to meet the formal documentation requirements to operate as MBs. In this regard, 57 % and 18 % of financially included owners have a TIN or a commercial registry, respectively, compared to 40 % and 7 % among the financially excluded.

Based on holding a TIN and a commercial registry, Table 3 summarizes the statistics for formal and informal MBs. As in Table 2, disadvantaged groups in society, such as women and ethnic minorities, are less likely to own a formal MB. Among informal MBs, the share owned by women is higher than among formal ones, particularly when the commercial registry is used as the formalization criterion: 85 % of informal MBs are owned by women, compared to 66 % of formal MBs. Additionally, informal MBs are more often owned by individuals belonging to an ethnic group (55 % and 54 % in informal MBs, compared to 50 % and 41 % in formal MBs, according to the TIN and commercial registry, respectively).

In terms of education and FI, formal MBs have a higher proportion of owners with a technical diploma and stronger ties to the financial system. Likewise, informal businesses tend to be younger than three years and are more frequently located in areas classified as strata 1 and 2.

Table 3. Summary statistics for owners and MBs by formality

	TIN				Commercial registry
	No		Yes		No		Yes
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Women	0.86	0.34	0.79	0.41	0.85	0.36	0.66	0.47
Age (years)	38.62	11.69	39.56	11.42	38.89	11.49	40.39	12.04
Married	0.54	0.50	0.55	0.50	0.54	0.50	0.56	0.50
Ethnic group	0.55	0.50	0.50	0.50	0.54	0.50	0.41	0.49
Technical education	0.26	0.44	0.31	0.46	0.27	0.44	0.41	0.49
DSI	0.85	0.17	0.87	0.16	0.86	0.17	0.88	0.16
FI indicator	0.38	0.28	0.48	0.26	0.41	0.27	0.57	0.27
Cali	0.34	0.47	0.34	0.47	0.34	0.47	0.34	0.48
Strata
Strata1 and 2	0.88	0.32	0.83	0.38	0.88	0.33	0.70	0.46
Stratum 3	0.12	0.32	0.17	0.38	0.12	0.33	0.29	0.46
Economic activity
Food	0.29	0.45	0.22	0.41	0.26	0.44	0.20	0.40
Clothing/Apparel	0.16	0.37	0.15	0.36	0.16	0.37	0.10	0.31
Hairdressing	0.15	0.35	0.12	0.32	0.14	0.34	0.10	0.31
Restaurants	0.09	0.29	0.07	0.26	0.08	0.28	0.07	0.25
Less than 3 years	0.58	0.49	0.50	0.50	0.56	0.50	0.41	0.49

Notes: (i) Married is equal to 1 if the person is married or in a free union, and 0 otherwise. (ii) Ethnic group takes a value of 1 if the person is not considered mestizo or white, and 0 otherwise. (iii) Technical education is 1 if the owner has attained at least a technical degree, and 0 otherwise. (iv) The DSI and FI indicators range from 0 to 1, with 1 representing MB owners who are fully digitally skilled and financially included, respectively. (v) Cali is 1 if the MB operates in Cali, and 0 if it is in another municipality. (vi) Only owners belonging to socioeconomic strata 1, 2, and 3 participated in the RP program. (vii) “Less than 3 years'' is 1 if the MB is three years old or less, and 0 otherwise. (viii) The variables Commercial registry and TIN take a value of 1 if the MB has these legal requirements, and 0 otherwise. (ix) Economic activity variables take a value of 1 if the MB belongs to the respective sector, and 0 otherwise.

Source: Authors’ calculations based on RP survey.

Table 4. Summary statistics for owners and MBs by digital skills

	Digitally skilled (p30)				Digitally skilled (p50)				Digitally skilled (p70)
	No		Yes		No		Yes		No		Yes
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Women	0.84	0.36	0.82	0.39	0.85	0.36	0.81	0.4	0.85	0.36	0.78	0.42
Age (years)	45.76	11.23	35.76	10.22	43.7	11.48	34.18	9.45	41.74	11.42	32.81	9.27
Married	0.57	0.50	0.53	0.50	0.58	0.49	0.51	0.50	0.58	0.49	0.45	0.50
Ethnic group	0.56	0.50	0.51	0.50	0.55	0.50	0.50	0.50	0.54	0.50	0.50	0.50
Technical education	0.14	0.35	0.36	0.48	0.16	0.37	0.41	0.49	0.21	0.40	0.48	0.50
FI indicator	0.40	0.26	0.44	0.28	0.41	0.27	0.45	0.28	0.41	0.27	0.47	0.29
Cali	0.28	0.45	0.37	0.48	0.31	0.46	0.37	0.48	0.32	0.47	0.38	0.49
Strata
Strata1 and 2	0.90	0.30	0.83	0.37	0.88	0.33	0.83	0.38	0.87	0.34	0.82	0.39
Stratum 3	0.10	0.30	0.17	0.37	0.12	0.33	0.17	0.38	0.13	0.34	0.18	0.39
Economic activity
Food	0.33	0.47	0.22	0.41	0.30	0.46	0.21	0.41	0.29	0.45	0.19	0.39
Clothing/Apparel	0.15	0.35	0.16	0.37	0.16	0.36	0.16	0.36	0.16	0.36	0.15	0.36
Hairdressing	0.10	0.29	0.15	0.36	0.12	0.32	0.15	0.36	0.12	0.33	0.16	0.37
Restaurants	0.09	0.29	0.08	0.27	0.09	0.29	0.08	0.26	0.09	0.28	0.08	0.26
Less than 3 years	0.45	0.50	0.58	0.49	0.47	0.50	0.62	0.49	0.50	0.50	0.63	0.48
TIN	0.47	0.50	0.50	0.50	0.47	0.50	0.51	0.50	0.46	0.50	0.55	0.50
Commercial registry	0.10	0.30	0.14	0.35	0.12	0.32	0.14	0.35	0.12	0.32	0.15	0.36

Notes: (i) Married is equal to 1 if the person is married or in a free union, and 0 otherwise. (ii) Ethnic group takes a value of 1 if the person is not considered mestizo or white, and 0 otherwise. (iii) Technical education is 1 if the owner has attained at least a technical degree, and 0 otherwise. (iv) The DSI and FI indicators range from 0 to 1, with 1 representing MB owners who are fully digitally skilled and financially included, respectively. (v) Cali is 1 if the MB operates in Cali, and 0 if it is in another municipality. (vi) Only owners belonging to socioeconomic strata 1, 2, and 3 participated in the RP program. (vii) “Less than 3 years’’ is 1 if the MB is three years old or less, and 0 otherwise. (viii) The variables Commercial registry and TIN take a value of 1 if the MB has these legal requirements, and 0 otherwise. (ix) Economic activity variables take a value of 1 if the MB belongs to the respective sector, and 0 otherwise.

Source: Authors’ calculations based on RP survey.

As for digital skills, MBs are divided into two groups based on the DSI, using different thresholds. As a reference, the 30th, 50th, and 70th percentiles of the DSI are applied, resulting in three alternative classifications. Owners are labeled as “Digitally skilled – Yes” if their DSI is above the corresponding threshold, and “Digitally skilled – No” otherwise. As shown in Table 4, women tend to be less digitally skilled across all threshold levels. Similarly, owners with low digital skills are more likely to belong to an ethnic group (56 %, 55 %, and 54 %) compared to their digitally skilled counterparts (51 %, 50 %, and 50 %). Regarding education and FI, digitally skilled owners are more likely to have attained a technical education degree and to use a broader range of financial products. Furthermore, MBs owned by digitally skilled individuals tend to be younger (58 %, 62 %, and 63 %) than those owned by individuals with lower digital skills (45 %, 47 %, and 50 %). Finally, digital proficiency also correlates positively with holding the documentation required to operate formally.

Gender gaps are a common phenomenon in Colombian society. Table 5 shows the differences between male and female MB owners in terms of both owner and business characteristics. First, a higher share of women belong to ethnic groups (54 %) compared to men (47 %). However, men and women tend to have similar education levels and digital skills. Regarding financial products, men are more likely to be involved in the financial system. In terms of MB characteristics, women tend to own younger MBs (49 %) compared to men (55 %). Nevertheless, MBs owned by women are predominantly informal, particularly when measured by commercial registry.

Regarding the dimensions of FI, Figure 1 displays the income distribution of MBs across the financial categories defined. As shown in Figure 1(a), MBs that use digital payment methods tend to have higher incomes than those relying solely on cash. Although the difference is less marked, a similar pattern appears in panels (b), (c), and (d), where MBs with access to formal financing, formal credit, or a bank account, respectively, show higher incomes compared to those without access to these financial products.

The descriptive analysis above reveals income differences across the MB subgroups analyzed, supporting the application of a decomposition method based on grouping variables such as financial characteristics, formality, digital skills, and gender. Moreover, the observed variation in subgroup composition suggests at least correlations between these characteristics and income levels.

Table 5. Summary statistics for owners and MBs by gender

	Gender
	Men		Women
	Mean	SD	Mean	SD
Age (years)	37.96	11.52	39.31	11.57
Married	0.63	0.48	0.52	0.50
Ethnic group	0.47	0.50	0.54	0.50
Technical education	0.28	0.45	0.29	0.45
DSI	0.88	0.16	0.86	0.17
Financial inclusion	0.49	0.27	0.42	0.28
Cali	0.37	0.48	0.33	0.47
Strata
Strata1 and 2	0.81	0.40	0.86	0.34
Stratum 3	0.19	0.40	0.14	0.34
Economic activity
Food	0.20	0.40	0.27	0.44
Clothing/Apparel	0.06	0.24	0.18	0.38
Hairdressing	0.10	0.29	0.14	0.35
Restaurants	0.08	0.27	0.08	0.28
Less than 3 years	0.49	0.50	0.55	0.50
TIN	0.60	0.49	0.46	0.50
Commercial registry	0.25	0.43	0.10	0.30

Notes: (i) Married is equal to 1 if the person is married or in a free union, and 0 otherwise. (ii) Ethnic group takes a value of 1 if the person is not considered mestizo or white, and 0 otherwise. (iii) Technical education is 1 if the owner has attained at least a technical degree, and 0 otherwise. (iv) The DSI and FI indicators range from 0 to 1, with 1 representing MB owners who are fully digitally skilled and financially included, respectively. (v) Cali is 1 if the MB operates in Cali, and 0 if it is in another municipality. (vi) Only owners belonging to socioeconomic strata 1, 2, and 3 participated in the RP program. (vii) “Less than 3 years’’ is 1 if the MB is three years old or less, and 0 otherwise. (viii) The variables Commercial registry and TIN take a value of 1 if the MB has these legal requirements, and 0 otherwise. (ix) Economic activity variables take a value of 1 if the MB belongs to the respective sector, and 0 otherwise.

Source: Authors’ calculations based on RP survey.

III. Methodology

A. Oaxaca-Blinder decomposition

Counterfactual decomposition methods are commonly used to investigate the sources of variation in gaps between two groups. This method separates the average difference (or any other statistic of interest) into two effects or components: the composition effect and the structure effect. The composition effect, also known as the Explained component, refers to the part of the gap caused by differences in the population’s composition, i.e., observable characteristics. The structure effect, or Unexplained component, refers to differences in the returns to those characteristics (coefficients) or to unobserved factors.

The main objective is to account for the mean difference in income between two comparison groups (A and B), using a linear model of the form:

where i represents the MB and g the group used in the OB decomposition. Y_i represents the income (logs.) of MB i.¹ The vector of covariates, X_i, includes owner characteristics: gender, marital status, ethnicity, age, educational level (specifically, whether the owner has obtained a technical degree), as well as financial characteristics, and digital skills.² Likewise, MB characteristics are incorporated into the estimates through X_i, such as holding a commercial registry and TIN (proxies for formality), the socioeconomic stratum, city, and economic activity.

We follow the decomposition suggested by Oaxaca (1973) and Blinder (1973). The resulting equation is:

On the right-hand side, the first term of the equation corresponds to the Unexplained component, while the second term represents the Explained component. As noted in the introduction, analyzing the performance of MBs (measured through their income) is key to identifying differences between groups across dimensions such as FI, formality, digital skills, and gender. Therefore, comparison groups are defined according to these characteristics. For example, with FI, group A corresponds to MBs that are not financially included, while group B includes those with a certain degree of FI. More generally, the groups used to split the sample for the OB decompositions correspond to those presented in Tables 2-5.

Nonetheless, the conclusions drawn from the OB decomposition are subject to some limitations due to the characteristics of the sample. Participants in the RP program were not randomly selected; instead, they were drawn from the Pacific region of Colombia based on specific eligibility criteria. This limits the extent to which the results can be generalized to other MBs across the country. Furthermore, due to financial and time constraints, the survey contains limited information, preventing us from controlling for crucial unobservable attributes (i.e., entrepreneur ability) that are likely to be correlated with both observed characteristics and the variables used to split the sample. Consequently, the OB decomposition may affect the magnitude of the estimated coefficients. Therefore, the interpretation of the OB decomposition should be considered descriptive.

Despite these limitations, the analysis provides valuable information on the potential gaps faced by MBs. Financial factors, formality, digital skills, and gender are characteristics that can create significant barriers for entrepreneurs. The use of OB decomposition still offers valuable descriptive insights into a particularly vulnerable segment of the entrepreneurial population—those whom the Foundation WWB Colombia and many other organizations aim to support. Thus, the analysis sheds light on important patterns and associations that can inform future research and guide the design of programs aimed at fostering MBs, as well as the development of targeted policies and the implementation of initiatives focused on reducing gaps and barriers for MBs, which represent a major source of employment in Colombia.

B. Focus group

To further investigate the factors associated with access to financing and the use of digital resources during the pandemic among entrepreneurs participating in the RP, two group interviews (5 participants) were conducted. Participant selection considered gender and financing access. Accordingly, one group included owners without access to financing (two women), while the other comprised owners with access to financing (two women and one man).

The interviews focused on financing access and usage across three distinct periods: prior to the mandatory lockdown, during the lockdown, and following the relaxation of lockdown measures. For each period, the type of financing (formal or informal) and its determinants were identified. Additionally, the interviews explored other strategies employed by entrepreneurs to cope with the challenges arising during the pandemic. This qualitative information provides valuable context for interpreting and validating the quantitative results obtained from the decomposition analyses.

IV. Results

A. OB decomposition

Following the methodology outlined in Section III, this section presents the estimation of the components of the Oaxaca-Blinder (OB) decomposition for each variable of interest. The income variable was log-transformed to facilitate the interpretation of the results. The findings are organized into four subsections. The first subsection focuses on FI, given its well-documented relevance for MB survival, as established in the literature. The subsequent subsections examine the results related to formality, digital skills, and gender, providing additional insights into the income generation gaps observed among MBs.

1. Financial inclusion

Table 6 presents the estimates for each of the financial dimensions: Digital means of payment (column 1), Access to formal credit (column 2), Holding a formal debt (column 3), and Possession of a savings or checking account (column 4); column 5 shows the results based on the FI indicator. In all columns, the difference in means between the two groups is negative and statistically significant, indicating that MBs without access to financial products perform worse on average. These MBs tend to generate 23.3 to 42.2 pp less income than those with financial products. The differences are decomposed into Explained and Unexplained components, most of which are statistically significant. The income gap between those who use and do not use digital payment methods is mainly explained by differences in coefficients rather than observables, as indicated by the large share of the Unexplained component (87.4 %).³

Table 6. OB decomposition grouping by financial characteristics

	Means of Payment (1)	Access to Credit (2)	Debt holding (3)	Bank Account (4)	Financial Inclusion (5)
General
MBs lacking a financial product (cols 1-4) or financially excluded (col 5)	13.044***	13.057***	13.039***	12.957***	12.948***
	(0.0312)	(0.0309)	(0.0324)	(0.0652)	(0.0365)
MBs owning a financial product (cols 1-4) or financially included (col 5)	13.466***	13.341***	13.351***	13.190***	13.329***
	(0.0522)	(0.0531)	(0.0482)	(0.0298)	(0.0389)
Difference	-0.422***	-0.283***	-0.312***	-0.233***	-0.381***
	(0.0608)	(0.0615)	(0.0581)	(0.0717)	(0.0534)
Explained	-0.054	-0.088***	-0.130***	-0.087***	-0.157***
	(0.0616)	(0.0329)	(0.0322)	(0.0308)	(0.0325)
Unexplained	-0.369***	-0.196***	-0.182***	-0.146**	-0.224***
	(0.0806)	(0.0592)	(0.0577)	(0.0716)	(0.0561)
Explained
Women	-0.033**	-0.011	-0.027**	-0.031**	-0.041***
	(0.0140)	(0.0163)	(0.0136)	(0.0134)	(0.0129)
Stratum	-0.006	-0.009	-0.024**	-0.012*	-0.035**
	(0.0282)	(0.0070)	(0.0116)	(0.0073)	(0.0140)
Married	-0.000	0.007	0.003	-0.003	0.001
	(0.0028)	(0.0065)	(0.0044)	(0.0047)	(0.0043)
Ethnicity	-0.011	-0.004	-0.006	-0.005	-0.009
	(0.0112)	(0.0052)	(0.0065)	(0.0044)	(0.0068)
Age owner	-0.003	0.000	0.007	0.000	0.004
	(0.0156)	(0.0072)	(0.0109)	(0.0049)	(0.0047)
Technical education	0.008	0.000	-0.000	-0.008	-0.002
	(0.0209)	(0.0071)	(0.0038)	(0.0066)	(0.0113)
Cali	-0.013	-0.001	-0.007	0.005	-0.013*
	(0.0161)	(0.0033)	(0.0064)	(0.0051)	(0.0078)
Age business	-0.002	-0.024*	-0.037**	-0.015*	-0.026**
	(0.0063)	(0.0132)	(0.0153)	(0.0086)	(0.0115)
DSI	0.062	-0.000	-0.002	-0.002	0.008
	(0.0446)	(0.0017)	(0.0032)	(0.0041)	(0.0073)
Commercial Registry	-0.083***	-0.060***	-0.057***	-0.050***	-0.071***
	(0.0224)	(0.0193)	(0.0168)	(0.0135)	(0.0156)
TIN	0.020	0.005	0.004	0.011	-0.000
	(0.0198)	(0.0072)	(0.0080)	(0.0163)	(0.0134)
Food	0.013	0.003	0.011	0.010	0.020*
	(0.0321)	(0.0044)	(0.0079)	(0.0076)	(0.0110)
Clothing/Apparel	-0.013	0.001	-0.001	0.000	-0.002
	(0.0105)	(0.0027)	(0.0029)	(0.0026)	(0.0049)
Hairdressing	0.000	-0.001	0.004	-0.003	-0.001
	(0.0083)	(0.0026)	(0.0046)	(0.0039)	(0.0031)
Restaurants	0.008	0.006	0.002	0.016	0.010
	(0.0078)	(0.0059)	(0.0057)	(0.0100)	(0.0075)
Unexplained
Women	-0.187	0.243*	0.059	0.000	0.071
	(0.1220)	(0.1264)	(0.1198)	(0.1769)	(0.1188)
Stratum	0.153	-0.023	-0.106	-0.125	-0.166
	(0.1292)	(0.1366)	(0.1271)	(0.1664)	(0.1185)
Married	-0.039	0.128**	0.071	-0.008	0.045
	(0.0622)	(0.0605)	(0.0579)	(0.0660)	(0.0520)
Ethnicity	0.024	0.011	0.022	-0.006	0.041
	(0.0650)	(0.0634)	(0.0614)	(0.0841)	(0.0601)
Age owner	0.076	-0.006	0.141	-0.038	0.183
	(0.2325)	(0.2215)	(0.2115)	(0.2653)	(0.1947)
Technical Education	0.036	0.026	0.031	0.006	0.027
	(0.0298)	(0.0345)	(0.0338)	(0.0356)	(0.0251)
Cali	-0.008	0.024	-0.015	-0.079	-0.041
	(0.0361)	(0.0394)	(0.0362)	(0.0526)	(0.0315)
Age business	0.030	0.057	0.101	-0.156	0.080
	(0.0655)	(0.0695)	(0.0674)	(0.0992)	(0.0658)
Digital Skills Index	0.714	0.463	0.548*	0.132	0.661**
	(0.4416)	(0.3414)	(0.3197)	(0.3723)	(0.2957)
Commercial Registry	-0.002	-0.032*	-0.004	0.007	-0.011
	(0.0163)	(0.0185)	(0.0165)	(0.0184)	(0.0129)
TIN	0.047	0.046	0.039	0.025	-0.020
	(0.0551)	(0.0567)	(0.0532)	(0.0429)	(0.0427)
Food	0.080	0.051	0.026	0.064	0.048
	(0.0562)	(0.0406)	(0.0403)	(0.0531)	(0.0419)
Clothing/Apparel	-0.029	0.022	-0.003	0.027	-0.001
	(0.0218)	(0.0257)	(0.0240)	(0.0293)	(0.0203)
Hairdressing	-0.001	0.020	0.032	0.053	0.009
	(0.0292)	(0.0257)	(0.0226)	(0.0373)	(0.0247)
Restaurants	-0.005	0.009	0.000	-0.027	-0.014
	(0.0195)	(0.0198)	(0.0175)	(0.0274)	(0.0181)
Constant	-1.256**	-1.234**	-1.123**	-0.020	-1.139**
	(0.6256)	(0.5160)	(0.4930)	(0.6103)	(0.4666)
Observations	1 494	1 494	1 494	1 494	1 494

Notes: (i) Standard errors in parentheses (∗ p < 0,1, ∗∗ p < 0,05, ∗∗∗ p < 0,01). (ii) The weight of each component corresponds to its proportion relative to its total difference.

Source: Authors’ calculations using Stata.

Regarding the Explained component, the results indicate that gender and commercial registry are positively associated with the widening of the income gap. Similarly, MBs utilizing digital means of payment tend to exhibit higher income levels. While these associations do not establish causality, they underscore the significance of formalization and gender as determinants of variation in MB performance. These findings further support the relevance of promoting FI through a gender-sensitive framework, as a means to mitigate economic vulnerabilities that disproportionately affect women (Roa, 2021).

The analysis of the credit and savings dimensions reveals an income gap ranging from 23.3 to 31.2 pp between those who have access to (or hold) formal credit or a bank account and those who do not. Both the Explained and Unexplained components of the decomposition are statistically significant, with the latter contributing more substantially to the overall difference. At a more disaggregated level, MB age, socioeconomic stratum, commercial registry, and gender are positively associated with the widening of the income gap.

Finally, the FI indicator exhibits patterns consistent with those observed for the individual dimensions. Financially excluded MBs earn 38.1 pp less income, a gap comparable to that found between the groups defined by means of payment (column 1). However, in this case, the gap is more evenly distributed between the Explained and Unexplained components, with weights of 41.2 % and 58.8 %, respectively. Within the Explained component, gender, stratum, MB age, holding a commercial registry, and being located in Cali are positively associated with a broader income gap between financially included and excluded MBs. In the Unexplained component, only the DSI contributes to the income gap. As with the estimates based on formal credit holding, digital skills appear to be associated with a reduction in the income gap.

The OB decomposition further enables the grouping of covariables according to selected criteria. In this case, variables are classified into categories related to owner characteristics, MB characteristics, and formality status. This grouping allows for a more detailed characterization of the income gap across both components of the decomposition. Table 7 presents the aggregated results for these categories. However, the first panel is omitted, as the estimates do not differ from those already reported in Table 6.

Table 7. OB decomposition grouping by financial characteristics (aggregate)

	Means Payment	Access Credit	Possession Credit	Bank Account	Financial Inclusion
Explained
Owner	-0.046	-0.016	-0.047**	-0.058***	-0.082***
	(0.0420)	(0.0223)	(0.0219)	(0.0189)	(0.0236)
Business	0.055	-0.017	-0.031	0.011	-0.004
	(0.0521)	(0.0156)	(0.0194)	(0.0162)	(0.0199)
Formality	-0.063**	-0.055***	-0.052***	-0.039**	-0.071***
	(0.0259)	(0.0189)	(0.0169)	(0.0195)	(0.0184)
Unexplained
Owner	0.063	0.378	0.218	-0.17	0.202
	(0.2977)	(0.2940)	(0.2803)	(0.3687)	(0.2672)
Business	0.781*	0.646*	0.688**	0.013	0.743**
	(0.4614)	(0.3557)	(0.3347)	(0.4047)	(0.3138)
Formality	0.044	0.015	0.035	0.032	-0.030
	(0.0528)	(0.0542)	(0.0511)	(0.0417)	(0.0412)
Constant	-1.256**	-1.234**	-1.123**	-0.02	-1.139**
	(0.6256)	(0.5160)	(0.4930)	(0.6103)	(0.4666)
Observations	1 494	1 494	1 494	1 494	1 494

Notes: (i) Standard errors in parentheses (* p < 0.1, ** p < 0.05, *** p < 0.01). (ii) The owner aggregate variable includes sex, stratum, marital status, ethnicity, age, and educational level. (iii) The aggregate business variable includes geographic location, economic activity, seniority, and DSI. (iv) Formality aggregate variable includes both the TIN holding and commercial registry status. (v) The weight of each component is calculated as its ratio to the total difference.

Source: Authors’ calculations using Stata.

Analyzing the Explained component, formality is associated with a widening of the income gap across all columns. Similarly, owner-related variables appear to contribute to greater disparities in MB income. With respect to the Unexplained component, MB characteristics are negatively correlated with the income gap. Thus, the characteristics associated with the owner, the MB, and formality complement one another in explaining income disparities through both the Explained and Unexplained components, particularly when credit holding and the FI indicator are used in the OB decomposition (columns 3 and 5).

2. Formality

The second dimension of interest is formality, measured using two proxies based on data availability: TIN and Commercial Registry. As reported in Table 8, formal MBs exhibit higher income levels—19.5 and 82.8 pp more, respectively—compared to their informal counterparts. These differences indicate a strong association between formalization status and income disparities among MBs.

When formality is measured by TIN holding, only the Explained component is statistically significant, indicating that the income difference is primarily attributable to observable MB characteristics. In contrast, when formality is measured through the commercial registry, the Explained component remains significant and carries substantial weight (69.2 %) in explaining the income difference.

Table 8. OB decomposition grouping by formality

	(TIN)	(Commercial registry)
	Formality	Formality
General
Informal MBs	13.051***	13.040***
	(0.0359)	(0.0275)
Formal MBs	13.246***	13.868***
	(0.0408)	(0.0837)
Difference	-0.195***	-0.828***
	(0.0543)	(0.0881)
Explained	-0.128***	-0.255***
	(0.0280)	(0.0782)
Unexplained	-0.067	-0.573***
	(0.0545)	(0.1079)
Explained
Owner	-0.084***	-0.158**
	(0.0202)	(0.0671)
Business	0.016	0.011
	(0.0159)	(0.0449)
FI	-0.061***	-0.108**
	(0.0168)	(0.0470)
Unexplained
Owner	0.449	-0.006
	(0.2750)	(0.4138)
Business	-0.147	1.171**
	(0.3211)	(0.5429)
FI	-0.061	-0.117
	(0.0476)	(0.0968)
Constant	-0.308	-1.621**
	(0.4726)	(0.7561)
Observations	1 494	1 494

Notes: (i) Standard errors in parentheses (∗ p < 0.1, ∗∗ p < 0.05, ∗∗∗ p < 0.01). (ii) The owner aggregate variable includes sex, stratum, marital status, ethnicity, age, and educational level. (iii) The aggregate business variable includes geographic location, economic activity, seniority, and DSI. (iv) FI aggregate variable includes the FI indicator. (v) The weight of each component is calculated as its ratio to the total difference.

Source: Authors’ calculations using Stata.

Furthermore, Table 8 presents the covariables grouped according to owner characteristics, MB characteristics, and FI. The results suggest that higher levels of FI are consistently associated with a wider income gap, regardless of the formality measure employed. Additionally, owner characteristics are significant across both formality measures, contributing to the explanation of income differences. With respect to the Unexplained component, no significant factors emerge when formality is measured by TIN holding. However, when using the commercial registry as the formality measure, the variables grouped under MB characteristics appear to contribute to the income gap, favoring MBs operating informally.

3. Digital skills

Additional estimates are conducted using digital skills as the grouping variable. Table 9 presents the results, applying the thresholds specified in Section III (30th, 50th, and 70th percentiles). First, income differences are statistically significant across all threshold levels, indicating that owners with lower digital skills earn, on average, between 10.6 and 16.8 pp less income compared to their more digitally skilled counterparts. Nevertheless, the magnitude of the gap diminishes as higher percentiles are employed as thresholds.

When examining the aggregated estimates by groups (owner, MB, formality, and FI), the results indicate that owner characteristics do not appear to explain income disparities among MBs. By contrast, MB characteristics are associated with a narrowing of the gap, whereas formality and FI are linked to a modest widening. Regarding the Unexplained component, Table 10 reveals virtually no statistically significant differences in the returns to owner characteristics. However, MB characteristics, formality, and FI continue to be associated with the size of the income gap.

Table 9. OB decomposition grouping by digital skills

	DSI
	(Percentile 30)	(Percentile 50)	(Percentile 70)
General
MBs with a DSI below the corresponding percentile	13.034***	13.078***	13.115***
	(0.0502)	(0.0393)	(0.0326)
MBs with a DSI above the corresponding percentile	13.202***	13.218***	13.221***
	(0.0322)	(0.0374)	(0.0492)
Difference	-0.168***	-0.140***	-0.106*
	(0.0596)	(0.0543)	(0.0590)
Explained	0.004	0.026	0.013
	(0.0413)	(0.0454)	(0.0598)
Unexplained	-0.172***	-0.166**	-0.119
	(0.0659)	(0.0647)	(0.0770)
Explained
Owner	-0.036	-0.008	-0.026
	(0.0379)	(0.0432)	(0.0556)
Business	0.071***	0.061***	0.077***
	(0.0184)	(0.0184)	(0.0235)
Formality	-0.020**	-0.014	-0.025*
	(0.0098)	(0.0093)	(0.0148)
FI	-0.011	-0.013*	-0.013
	(0.0068)	(0.0075)	(0.0088)
Unexplained
Owner	-0.197	-0.500*	-0.499
	(0.3157)	(0.2902)	(0.3115)
Business	-0.076	-0.055	-0.091
	(0.1012)	(0.0929)	(0.1000)
Formality	-0.057	0.028	-0.007
	(0.0532)	(0.0490)	(0.0522)
FI	0.059	0.005	0.040
	(0.0555)	(0.0530)	(0.0591)
Constant	0.098	0.357	0.439
	(0.3376)	(0.3063)	(0.3223)
Observations	1 494	1 494	1 494

Notes: (i) Standard errors in parentheses (* p < 0.1, ** p < 0.05, *** p < 0.01). (ii) The owner aggregate variable includes gender, stratum, marital status, ethnicity, age, and educational level. (iii) The business aggregate variable contains geographic location, economic activity, and seniority. (iv) Formality aggregate variable includes both the TIN holding and commercial registry status. (v) FI aggregate variable includes the FI indicator. (vi) The weight of each component is calculated as its ratio to the total difference.

Source: Authors’ calculations using Stata.

4. Gender gap

Finally, the owner’s gender is introduced as a grouping variable in the OB decomposition. Given that the RP aims to support entrepreneurs, particularly women, the sample includes a high proportion of female participants. Thus, performing a gender-based decomposition enables fair comparisons that account for differences in the observable characteristics of male and female business owners.

Table 10 reports a positive and statistically significant difference, indicating that male-owned MBs generate 66.9 pp more income than female-owned MBs. Both the Explained and Unexplained components are significant, with the latter accounting for the largest share of the income gap (85.7 pp). In interpreting the decomposition results, it is important to note that, given the positive sign of the gap, a positive coefficient denotes a contribution to widening the gap, while a negative coefficient indicates an association with its narrowing.

Regarding the Explained component, characteristics related to the owner, formality, and FI are associated with a widening of the income gap between MBs (Table 10). These attributes are more prevalent among male-owned MBs, which, on average, exhibit higher income levels compared to female-owned MBs. This indicates that differences in these characteristics partially account for the observed income disparity between men and women. Regarding the Unexplained component, only formality remains statistically significant, suggesting that differences in the returns to formality contribute to income inequalities.

Table 10. OB Decomposition grouping by gender

General	Gender
MB owned by men	13.697***
	(0.0669)
MB owned by women	13.028***
	(0.0287)
Difference	0.669***
	(0.0728)
Explained	0.116***
	(0.0338)
Unexplained	0.553***
	(0.0721)
Explained
Owner	0.026*
	(0.0138)
Business	-0.020
	(0.0190)
Formality	0.083***
	(0.0222)
FI	0.028***
	(0.0104)
Unexplained
Owner	-0.037
	(0.3461)
Business	-0.137
	(0.4106)
Formality	0.149*
	(0.0891)
FI	0.047
	(0.0900)
Constant	0.532
	(0.5906)
Observations	1 494

Notes: (i) Standard errors in parentheses (* p < 0.1, ** p < 0.05, *** p < 0.01). (ii) The owner aggregate variable includes stratum, marital status, ethnicity, age, and educational level. (iii) The business aggregate variable includes geographic location, economic activity, and seniority. (iv) Formality aggregate variable includes both the TIN holding and commercial registry status. (v) FI aggregate variable includes the FI indicator. (vi) The weight of each component is calculated as its ratio to the total difference.

Source: Authors’ calculations using Stata.

B. Comparative analysis

The interviews reveal consistent patterns in financing behavior during the pandemic. Most respondents reported a preference for obtaining credit from family or friends rather than from formal banking institutions. This tendency is attributable to the income volatility experienced by MB owners, which generates significant uncertainty regarding their repayment capacity. As a result, informal financing through relatives or friends is favored, given their greater flexibility in renegotiating debt terms. This preference is reflected in the word clouds presented in Figure 2. In particular, Figure 2(a) highlights frequently mentioned words among those with access to the formal financial system, including ease, loans, and references to relatives (sister and uncle), illustrating their inclination toward informal credit options and their reluctance to engage with formal banking services.

Conversely, Figure 2(b) indicates that the term business is the most frequently used word among those without access to the formal financial system, suggesting a search for additional business alternatives beyond their primary activity as a strategy to mitigate the effects of the pandemic-induced crisis. Figure 2(b) also highlights the prominence of the word pandemic among respondents lacking formal financing. Thus, while both groups express similar views regarding access to financing through relatives, for those without formal financial access, income generation appears to hold greater relevance than obtaining loans.

The focus groups further reveal that the preference for informal financing arises from a fear of the formal financial system, likely attributable to limited knowledge and compounded by negative experiences reported by close acquaintances. Nevertheless, irrespective of the financing source, micro-entrepreneurs generally turn to the financial system only as a last resort and often lack specific investment plans. Consequently, even when formal credit becomes accessible, it may not necessarily translate into productive outcomes.

C. Discussion

As highlighted in the interviews, necessity-driven MBs exhibit limited access to formal sector loans and credits, primarily due to negative attitudes (fears) and obstacles such as unfavorable credit bureau records. Consequently, their primary sources of financing tend to be personal networks, including family members and trusted acquaintances.

With respect to attitudes, the interviews consistently revealed a significant lack of knowledge regarding the functioning of the financial system, which fostered aversion toward formal financing. As a result, participants frequently perceived the formal system as risky and intimidating, a perception often reinforced by negative experiences within their social networks. Furthermore, many respondents systematically compared formal financing with credit obtained from relatives and close acquaintances. In these comparisons, the formal financial system was characterized as “rigid” or “strict,” marked by fixed payments and deadlines and limited flexibility for renegotiation, in contrast to the familiarity and trust inherent in family loans, where repayment terms can be more easily adjusted.

This perception of rigidity stems from the absence of a close relationship with potential customers, limiting banks’ ability to offer low-risk credit. In this context, MBs resort to informal credit at high interest rates, which constrains their growth and keeps them trapped in the informal system. Therefore, access to and proficiency in new technologies that facilitate the use of financial services, such as digital payments, may contribute to the creation of payment histories and credit records. In turn, this can reduce information asymmetries and foster closer relationships between formal banks and potential customers. Increased trust could then lower interest rates and promote access to credit.

Furthermore, adverse attitudes and obstacles underscore the need to strengthen financial education among these population segments, who tend to maintain a distant and skeptical stance toward formal financial institutions. However, these beliefs and perceptions are also shaped by their living conditions. Engagement in activities characterized by variable incomes and lower returns heightens the perceived risk of default when accessing credit. From this position, living under conditions of uncertainty and vulnerability, incurring debt does not represent a “viable” alternative.

As a result, when needing a loan, these individuals primarily rely on their close networks, especially family members. The main advantages of this choice lie in the flexibility and trust that enable them to manage the debt and its repayment under conditions perceived as more compatible with their income and economic circumstances. This reliance on close relationships was also recurrent during the pandemic, where support from family members, alongside knowledge of and access to institutional programs, played a central role. In short, these family and social ties constitute a key resource for the operation and growth of necessity-driven businesses.

However, when individuals accessed loans, whether formal or informal, the common denominator was the absence of an investment plan, with loans generally understood as responses to “need” or “urgency.” From their perspective, borrowing occurs as a “last resort”: when their own capacity is insufficient, they lack savings, and no family member is able to lend them money. Only one interviewee mentioned using credit with specific goals for business development. Overall, for the individuals interviewed, debt is not viewed as something that can be planned or used strategically, but rather as a necessity and an extraordinary situation to be avoided.

V. Conclusions

Since the onset of the COVID-19 pandemic, governments have implemented a range of public policies aimed at mitigating the effects of the crisis and supporting households and businesses. However, it is not possible to reach all individuals, particularly those operating informally. In this context, Foundation WWB Colombia launched the RP program in 2021 to assist hundreds of MBs. Drawing on data collected through the program, this document analyzes the income gap and its relationship with FI among MBs located in Cali and its surrounding areas. Additionally, other factors such as the use of digital technologies, formality, and gender are examined. To this end, the analysis combines OB decompositions and focus group discussions.

The results show that formal MBs with higher levels of FI are more likely to generate higher income compared to those that are informal and less connected to the formal financial system. This may reflect the fact that formal MBs tend to be larger and more established in the market. Likewise, a high level of FI is associated with a greater likelihood of accessing credit during times of crisis, which may help mitigate the negative impact on economic performance. Furthermore, the findings suggest that owners with stronger digital skills tend to generate higher income in their MBs on average, highlighting the potential relevance of these skills (particularly for those with limited technological familiarity). Finally, women-owned MBs appear to earn lower income on average, a pattern that is also accompanied by lower reported levels of digital skills and education.

Regarding the focus groups, the results primarily indicate that accessing the formal financial system can be a deliberate choice for the MBs studied. The interviews reveal that limited access to credit products is often associated with adverse attitudes (such as fear of default) and obstacles like negative credit bureau reports. These insights underscore the importance of strengthening financial education, enabling individuals to make more informed financial decisions. As individuals gain the capacity to make choices under improved conditions, access to and utilization of formal financial products may increase, potentially enhancing MB performance.

Nonetheless, reducing supply-side barriers remains essential. In this regard, public policy should foster the adoption of new technologies, such as electronic wallets, fintech solutions, and digital payments, which are critical for generating information within the financial system. Moreover, alternative data sources should be explored, including open banking and payment histories from utilities or telecommunications services. These innovations can enable banks and other financial institutions to allocate credit more efficiently — both in terms of loan amounts and interest rates — while helping to prevent discrimination based on gender, race, or other socioeconomic characteristics. Consequently, MBs also benefit from reduced information asymmetries, which lead to lower interest rates and an eased financial burden, thereby improving their performance.

Declarations

Funding

This study was funded by the program Productive and Social Inclusion: Programs and Policies for the Promotion of a Formal Economy (Code 60185), part of Colombia Científica-Alianza EFI, under Contingent Recovery Contract No. FP44842-220-2018.

Origin of the study

This article results from a collaboration between Universidad del Rosario and Fundación WWB Colombia.

Acknowledgments

The authors thank the Fundación WWB Colombia team for their collaboration and guidance throughout the study. We also express our gratitude to the anonymous reviewers for their constructive comments, which substantially improved the scope of the research and strengthened its theoretical foundations. We would also like to thank Tiziana Laudato for her diligent copyediting and proofreading of the English version of this article.

Conflict of interest

Data availability

The dataset used in this study is not publicly available, as it contains proprietary information obtained under license. Requests for access to the data and replication of the analysis may be directed to the corresponding author (david.rodriguezg96@outlook.com).

Impacts of financial inclusion on micro-businesses: Factors explaining business growth

Impactos de la inclusión financiera en las microempresas: factores que explican el crecimiento empresarial

Abstract

Resumen

Introduction

I. Description of the Plan Reactívate Program

II. Summary statistics and sample description

III. Methodology

A. Oaxaca-Blinder decomposition

B. Focus group

IV. Results

A. OB decomposition

1. Financial inclusion

2. Formality

3. Digital skills

4. Gender gap

B. Comparative analysis

C. Discussion

V. Conclusions

Declarations

Funding

Origin of the study

Acknowledgments

Conflict of interest

Data availability

Notas al pie

References