Decomposing the Gender Gap in Financial Inclusion: An Oaxaca–Blinder Analysis for Peru, 2024

Abstract

The objective of this study was to quantify and decompose the gender gap in access to and use of financial services in Peru for the year 2024, distinguishing between the portion explained by observable characteristics and the unexplained component, which is associated with discrimination or differential returns. The methodology employed a quantitative analysis based on the National Household Survey (ENAHO), using weighted and unweighted Oaxaca–Blinder decomposition models on a representative sample of 14,240 household members. The explanatory variables included age, monthly household income, years of education, area of residence, marital status, employment status, and participation in social programs. The findings revealed a significant gender gap in financial inclusion of −2.16 percentage points, with the majority attributable to the unexplained component (−0.0724), indicating structural inequalities in the returns men and women receive from their characteristics. Variables such as years of education, monthly household income, and age had significant effects but yielded lower benefits for women. It is concluded that closing the gender gap in financial inclusion in Peru requires more than improving women’s individual characteristics; it also entails addressing differential returns and promoting gender-focused public policies that ensure equitable conditions in accessing the formal financial system.

1. Introduction

In these times of globalization, financial inclusion is crucial for promoting equal opportunities, reducing poverty, and fostering sustainable economic growth. This is why, globally, the expansion of financial infrastructure and the development of digital technologies have enabled significant advances in access to and use of financial services (Arner et al. 2020; Dhahri et al. 2024; Gabor and Brooks 2017; Raffaelli et al. 2025; Sumlinski et al. 2023). However, the gender gap remains a persistent challenge at the level of developed countries and is strongest in developing countries; since before the pandemic, the gender gap in financial inclusion in the world was evident, where, in 2017, 72% of adult men in developing countries had a financial account. Conversely, only 65% of women had a financial account, evidencing a difference of 7%. In addition, women showed a tendency to have less active financial accounts, the value of which reached 31% of inactive accounts, while men reached only 25%. In the pandemic period (2020–2021), the health crisis and social confinement disproportionately affected women, mainly in the labor aspect, negatively affecting access to and use of financial services; although the digitalization of financial services accelerated given the circumstances, the digital divide limited many women from taking advantage of these new options, thus contributing to the widening of the gender gap in access to financial services. In the post-pandemic period, there has been a gradual improvement in female financial inclusion, given that, by 2021, the percentage of adults with a bank account increased to 76%, and the gender gap decreased by 6% in developing countries; however, challenges remain in the proper adoption and active use of financial services, mainly in countries where there is greater socioeconomic inequality (Gabor and Brooks 2017; Confraria et al. 2024; Ma’ruf and Aryani 2019; Orazi et al. 2021).

The aforementioned problem is not uncommon in Latin America and the Caribbean, where the gender gap in financial inclusion before the pandemic was significant since, in 2019, only 51.4% of women had access to a bank account, while men reached 57.4%, showing a difference of 6%; therefore, women had informal savings practices, with a tendency to have less access to formal credit due to the existence of structural and social barriers, complemented by the high informality of women’s labor force. In the pandemic stage, female labor participation decreased drastically, reaching 46% in 2020, aggravating the gender gap in financial inclusion. In this sense, women faced greater difficulties in accessing digital services due to the digital divide and the decrease in their economic income, affecting their ability to access formal financial products. In the post-pandemic period, improvements have been shown in female financial inclusion, translating into a reduction in the gender gap from 10% to 5% compared to the pandemic stage; however, significant differences persist in access to and use of digital payments, with 32% of men and only 28% of women currently using these financial services (Raffaelli et al. 2025; Federico and García 2019; García et al. 2024; Rivera 2021; García 2021).

At the level of Peru, this situation reflects this inequality more clearly, given that, before the pandemic, financial inclusion showed low levels and a marked gender gap. As of 2017, only 38% of adults had access to a financial account, with an average difference of 5% between men (41%) and women (36%); this inequality was related to structural factors such as high female labor informality, lower levels of financial education, limitations in access to digital technologies, and a social environment that restricted women’s economic autonomy. In the pandemic period, the state implemented a set of emergency cash transfer policies, such as the Universal Bonus and the Yanapay Bonus, in order to mitigate the economic impact on vulnerable households, where many of these transfers were channeled through digital means, such as mobile wallets and bank accounts, promoting an accelerated process of banking penetration. Women faced significant barriers to accessing these mechanisms due to limitations in the use of smartphones, limited connectivity, outdated documentation, and low familiarity with the use of digital applications; therefore, despite achieving an increase in the opening of accounts in the name of women, mainly in urban areas, the effective use of financial services remained low, with many accounts becoming inactive after receiving the bonuses, which shows that in that period, female financial inclusion expanded in terms of basic access, but it failed to close the gap in a sustained manner. In the post-pandemic stage, some advances in access to financial services were consolidated through the massive use of digital wallets such as Yape and Plin, reducing barriers to entry into the financial system. By 2024, between 55% and 60% of adult women had some type of financial account, while men reached 65% to 70%, maintaining a gender gap of between 8% and 10% on average, where the difference is concentrated in rural, indigenous, low-educated, or poor women (Raffaelli et al. 2025; García 2021; Boitano and Abanto 2020; Tenorio 2025).

In addition, the effective use of financial services in the post-pandemic stage presents worrying gaps in view of the fact that, as of 2023, personal loan portfolios show that only 33% of the loans granted correspond to women, and the average delinquency rate is higher in this group due to the precariousness of economic income and the lack of solid credit history that customers have; this is accompanied by the low financial literacy of Peruvian women, whose value amounts, on average, to 15% and is lower than that of men, which becomes a limitation for making informed decisions about financial products (Raffaelli et al. 2025; Orazi et al. 2021; Armest et al. 2021; Ramirez-Asis et al. 2024; Smith 2021).

In this sense, despite the existence of consensus regarding the importance of closing these gaps, as well as the evident institutional efforts to promote financial education with a gender approach and public policies aimed at inclusion, there are still important challenges in terms of the active and productive use of financial services by women. Therefore, the post-pandemic stage represents a key opportunity to design and implement comprehensive strategies that address not only access but also the quality and sustainability of women’s financial inclusion; therefore, it is necessary to differentiate inequality that is explained by differences in aspects such as education, economic income, age, place of residence or employment status from those that correspond to structural or social factors, or even gender discrimination (Raffaelli et al. 2025; Boitano and Abanto 2020; Tenorio 2025; Vargas-Salazar et al. 2025); The Oaxaca–Blinder decomposition method, as a robust tool, can quantify these components, providing precise empirical evidence for the design of effective and focused public policies (Aslan et al. 2017; Herrera-Cano et al. 2024; Swamy 2014; Temoso et al. 2024).

Therefore, this research aims to quantify and decompose the gender gap in access to and use of financial services in Peru in 2024, using data from the National Household Survey (ENAHO). This analysis seeks to determine what proportion of the gap is explained by socioeconomic characteristics and what part remains unexplained, evidencing possible structural and gender barriers that must be addressed through specific interventions.

In addition, by applying the Oaxaca–Blinder decomposition model to analyze the gender gap in financial inclusion in Peru, using recent data from ENAHO 2024, the present research not only quantifies the gap but also breaks it down into explained and unexplained factors, offering a deeper understanding of the determinants of financial inequality between men and women. In addition, the added value lies in focusing on a post-pandemic context, which is marked by the advance of financial digitalization, integrating variables such as years of education and work experience measured by age from the perspective of human capital; this provides updated and useful empirical evidence for the design of public policies aimed at closing the gender gap in the Peruvian formal financial system.

2. Theoretical Framework

2.1. Conceptual Aspect of Financial Inclusion

Financial inclusion is defined as access to and the effective use of financial products and services in a formal manner, whether in savings, credit, insurance, or payment systems, which are at reasonable costs for all segments of the population. To guarantee financial inclusion, it is not enough to hold accounts; rather, it is key to guarantee the sustained use and quality of the services offered over time (Federico and García 2019; Singer et al. 2013).

In addition, considering financial inclusion to be a catalyst for economic development, their main potential is to mobilize savings, facilitate investment, and reduce economic vulnerabilities; however, an important criticism of this vision is that financial inclusion is often promoted under a technocratic or banking approach and does not adequately consider the social, gender, or territorial conditions that explain exclusion. In this sense, it is necessary to address financial inclusion as an economic and social right, not only as a strategy for the efficiency of the financial system (Delechat et al. 2018; Mostafa et al. 2023).

2.2. Gender Perspective and Economic Inequality

There is growing evidence highlighting how financial exclusion disproportionately affects women due to structural factors such as income inequality, informal employment, and restrictive social norms. The theory of statistical discrimination, for instance, explains how financial institutions may use observable attributes like gender as a proxy for perceived risk, thereby limiting women’s access to financial products (Anyangwe et al. 2022).

According to the capability approach proposed by other scholars, access to economic resources does not, by itself, guarantee empowerment; instead, it depends on women’s actual capacity to use those resources autonomously. Hence, the gender gap in financial inclusion becomes an indicator of restrictions on the freedom of economic choice.

The capability approach proposed by Abdu and Adem (2021) broadens our understanding, highlighting that economic empowerment does not depend solely on access to financial resources but instead on the actual capacity to use them autonomously. Therefore, this perspective includes the concept of economic freedom of choice, which allows us to analyze the financial gap as an expression of deeper limitations on women’s agency.

Furthermore, feminist approaches to development (Beirne et al. 2023) reinforce the need to recognize the intersections between gender, poverty, race, and territoriality, arguing that financial inclusion policies cannot be neutral, complemented by the adoption of differentiated strategies to transform structural inequalities. Therefore, this critical approach is vital to avoid reducing financial inclusion to a technical solution and to question its inequitable implementation.

2.3. Complementary Theories and Approaches on Financial Inclusion

2.3.1. Human Capital Theory

According to Human Capital Theory (Aronson 2007; Pérez-Vásquez and Prieto-Baldovino 2019), the level of education or years of education received and training in financial literacy determine productivity and the ability to generate economic income, which in turn has an impact on the ability to access financial services; in this sense, people with greater human capital, that is, with more years of schooling and formal work history, will show a tendency to have greater access to formal financial services (access to bank accounts, credits, insurance, and investment products) because these people have better cognitive and digital skills, greater confidence in interacting with financial institutions, and higher stable economic income, which facilitates both access to and effective use of these services.

In the specific case of women, the theory of human capital explains part of the gender gap in financial inclusion since, historically, they have had lower levels of schooling, more interrupted work trajectories, and greater participation in informal employment; these differences directly affect their ability to access financial products, to understand their conditions and risks, and to be considered credit subjects by financial institutions. In this sense, according to this theory, disparities in education and work experience between men and women contribute significantly to the persistence of the gender gap in financial inclusion.

In addition, a critical limitation of this approach is that it tends to hold the individual responsible for their level of inclusion, leaving aside the structural factors that condition unequal access to human capital. That is why, if women have less formal experience, the reason is not only individual decisions, but also barriers such as a lack of care networks, labor discrimination, or social norms; in this sense, although useful for the component “explained” in an Oaxaca–Blinder model, the theory of human capital alone is not enough to understand the deep roots of the gender gap in financial inclusion.

2.3.2. Social Exclusion Approach

This approach emphasizes that financial exclusion is a multidimensional phenomenon, driven by economic, social, and cultural barriers. It highlights the importance of intersectoral policies that address not only physical access to financial services but also barriers related to information, trust, and discriminatory regulations (Olarte 2017; Ramos-Zaga et al. 2024; Roa and Villegas 2023).

An advantage of this approach is that it helps to identify the multidimensionality of financial exclusion and justify intersectoral public policies since its empirical application is more complex, as many of the factors identified are not easily measurable with quantitative variables, making their integration into econometric models such as Oaxaca–Blinder a limitation. Therefore, the greatest strength originates in the explanatory and diagnostic capacity rather than in its direct statistical operationalization.

2.3.3. Structuralist Approach

This approach highlights labor market segmentation, informality, and the precariousness of female employment as key determinants of inequality in financial inclusion. It underscores the importance of analyzing these factors in contexts of high female labor informality, as they limit credit history development and access to collateral (Arner et al. 2020).

In addition, this approach relates inequalities in financial access to the macro-functioning of the economic system and labor policies, and at the level of the Peruvian case, its analysis is complex because informality affects more than 70% of employed women. However, the most important strength is that it offers a systemic explanation of exclusion, relating the financial market to the labor market, whose challenge is that it does not always allow specific policy solutions to be derived at the micro- or individual level, and its integration with empirical models requires simplifications that may lose part of their theoretical richness.

2.4. Oaxaca–Blinder Decomposition Models

The Oaxaca–Blinder decomposition method (Temoso et al. 2024; Blinder 1973) is considered an econometric technique widely used to break down wage gaps or other inequalities between groups. This model separates the average difference observed in a dependent variable into two components. The first is the explained component, which is associated with differences in observable characteristics such as education, economic income, age, geographical area, and employment status, among others. The other component is the unexplained component, which is attributable to differences in returns to those characteristics or to unobservable factors, such as discrimination, social norms, or institutional barriers.

There are some recent studies that have applied this technique to explain gender gaps in financial inclusion in similar contexts, such as Arun and Kamath (2015), who applied this technique to break down differences in labor income by gender in Latin America, while Demirgüç-Kunt et al. (2018) highlighted the methodology’s potential to analyze inequalities in access to financial products. Specifically, in the case of Peru, taking into account the information existing in the ENAHO database, it was not possible to identify any research that worked with it, but it can be used to estimate the probability of access and use of financial services considering key socioeconomic variables.

Therefore, this model is a useful methodological tool that allows us to break down differences between groups, such as the gender gap in financial inclusion, into two components: one explained by observable characteristics (education, income, age, etc.) and another unexplained, attributed to differences in the returns of these characteristics or to unobserved factors such as discrimination or social norms. Therefore, it is compatible with the theory of human capital and with structural and gender approaches since it empirically quantifies the relative weight of different factors.

However, this model has important limitations since it is based on assumptions of linearity and correct specification. In addition, the “unexplained” component is ambiguously interpreted, which does not necessarily reflect direct discrimination, and at the same time may include omitted variables, measurement errors, or specification errors.

Despite the above, its application in scenarios such as the Peruvian case, where there is little quantitative evidence on financial inclusion with a gender approach, represents a relevant methodological advance and establishes an empirical basis for the design of more equitable policies.

3. Materials and Methods

3.1. Research Approach and Design

The present research considers a quantitative approach, with a non-experimental, cross-sectional, and explanatory design, where secondary source information from the National Household Survey (ENAHO) of the National Institute of Statistics and Informatics (INEI) for the year 2024 is used due to access to existing information, at the same time comparing the results with cross-sectional data. The cross-sectional design allows us to assess the financial inclusion situation during a specific period, while the explanatory focus aims to identify and quantify the factors that determine the gender gap in access to and use of formal financial services (Mendoza 2014; Hernández Sampieri 2018).

3.2. Population and Sample

The study population consists of individuals aged 18 and above residing in private households in Peru. The sample is probabilistic, stratified, and nationally representative, in accordance with ENAHO’s sampling design. For analysis, individuals with complete data in Modules 2, 3, 5, 7, and 34 were selected to obtain the relevant study variables.

3.3. Variables and Operationalization

Financial inclusion was considered a dependent variable, which was defined as access to and use of formal financial services, measured by the possession of a bank account and use of financial products. In addition, the independent variables considered were the years of education, whose indicator is years of schooling received by the person; household income, measured by a natural logarithm of total monthly household income; occupation, measured by the person’s main occupational category (formal/informal/unemployed/independent); the geographical location, which measures the area of residence (urban/rural); household size, as measured by the total number of household members; the person’s marital status (single, married, cohabiting, separated, widowed); and gender, which measures the sex of the person (male/female) and the age of the person in years completed, the definition of which is detailed in

Table 1. Operationalization of research variables.

Type	Variable	Indicator	Scale	Source/Instrument
Dependent	Financial inclusion	Access and use of financial services (bank account ownership, usage)	Binary (Yes/No)	ENAHO, Module 500
Independent	Years of education	Total years of schooling	Discrete quantitative	ENAHO, Module 300
	Household income	Natural log of total monthly household income	Continuous (S/.)	ENAHO, Module 500
	Occupation	Main occupational category (formal, informal, unemployed, self-employed)	Nominal categorical	ENAHO, Module 500
	Geographic location	Area of residence (urban/rural)	Nominal categorical	ENAHO, Module 100
	Household size	Total number of household members	Discrete quantitative	ENAHO, Module 200
	Marital status	Marital status (single, married, cohabiting, separated, widowed)	Nominal categorical	ENAHO, Module 200
	Gender	Sex of the respondent (male/female)	Nominal categorical	ENAHO, Module 200
	Age	Completed years of age	Continuous	ENAHO, Module 200

.

3.4. Econometric Model Specification

In the regression model, the dependent variable (financial inclusion) is binary (1 = has access and use; 0 = otherwise). A binary response model is specified, and the Logit model is selected due to its suitability for qualitative dependent variables.

The probability of financial inclusion is modeled as follows:

$$P\left(Y_i=1∣X_i\right)=F\left(X_i^{\prime }\beta \right), \mathrm{con} F\left(z\right)={\displaystyle \frac{e^z}{1+e^z}}$$

where

• $Y_i=1$ if individual i has access and uses financial services; 0 otherwise.
• $X_i$ is the vector of independent variables (education, income, occupation, geographic location, household size, marital status, gender, age).
• $\beta$ is the vector of estimated parameters.
• F(·) is the cumulative logistic distribution function.

The specific model proposed in the research is

$$\begin{gathered} \mathrm{P}\left({\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{i}\mathrm{a}\mathrm{l} \mathrm{I}\mathrm{n}\mathrm{c}\mathrm{l}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{i}}\right) \\ =F({\beta }_0+{\beta }_1{\mathrm{E}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s}\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s} \mathrm{o}\mathrm{f} \mathrm{e}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_i \\ +{\beta }_2{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{i}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}}_i+{\beta }_3{\mathrm{O}\mathrm{c}\mathrm{c}\mathrm{u}\mathrm{p}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_i \\ +{\beta }_4{\mathrm{G}\mathrm{e}\mathrm{o}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{p}\mathrm{h}\mathrm{i}\mathrm{c} \mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_i+{\beta }_5{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{s}\mathrm{i}\mathrm{z}\mathrm{e}}_i \\ +{\beta }_6{\mathrm{M}\mathrm{a}\mathrm{r}\mathrm{i}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{s}}_i+{\beta }_7{\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{r}}_i+{\beta }_8{\mathrm{A}\mathrm{g}\mathrm{e}}_i+e_i) \end{gathered}$$

In order to quantify the gender gap in financial inclusion, the Oaxaca–Blinder decomposition technique was applied (Temoso et al. 2024; Blinder 1973; Jann 2008). This technique separates the mean difference in the dependent variable between men and women into two components:

• Explained component (E): Due to differences in observable characteristics (education, income, occupation, location, etc.).
• Unexplained component (U): Due to differences in returns to those characteristics or unobserved factors (e.g., discrimination, social norms).

The mathematical formulation is

$$∆\overline{Y}={\left({\overline{X}}_M-{\overline{Y}}_F\right)}^{\prime }{\widehat{\beta }}^{\ast }+{\overline{X}}_F^{\prime }\left({\widehat{\beta }}_M-{\widehat{\beta }}_F\right)$$

where

• $∆\overline{Y}$ is the average difference in financial inclusion between men (M) and women (F).
• ${\overline{X}}_M,{\overline{Y}}_F$ is the average of the characteristics for each group.
• ${\widehat{\beta }}^{\ast }$ is the vector of reference coefficients (can be ${\widehat{\beta }}_M,{\widehat{\beta }}_F$ or combined weight).

The first term is the explained component (differences in characteristics).

The second term is the unexplained component (differences in returns or unobserved factors).

3.5. Ethical Aspects

This research used the publicly accessible and anonymous ENAHO database for the year 2024, complying with the principles of confidentiality and research ethics established by the INEI (National Institute of Statistics and Census).

3.6. Limitations of the Research

Using ENAHO’s cross-sectional data prevents the analysis of individual or causal dynamics in financial inclusion since the same individuals cannot be followed over time; so, for future investigations, the Oaxaca–Blinder extensions can be applied.

In addition, when using the Oaxaca–Blinder decomposition method, the fact that it is descriptive in nature, not causal, does not allow for the identification of the direct effects of the variables. In addition, the “unexplained” component may include discrimination, specification errors, or omitted variables, and its interpretation should be performed with caution.

Finally, another limitation is the complexity that exists in the management process of the variables analyzed, given that the ENAHO database does not explicitly or accurately provide the variables studied; on the contrary, it must be generated from other variables that exist.

4. Results

4.1. Descriptive Analysis of Variables Involved in the Gender Financial Inclusion Gap

According to the descriptive results shown in

Table 2. Financial inclusion gap by gender.

Variables	Category	Gender		Total	Statistic
		Men	Women
Financial Inclusion	No	3957	2961	6918	Pearson chi2(1) = 22.7286
	%	57.20	42.80	100.00
	Yes	3897	3425	7322
	%	53.22	46.78	100
Total		7854	6386	14,240	Pr = 0.000
%		55.15	44.85	100.00

, in general terms, 51.42% of households are financially inclusive, meaning they have access to and use financial services, while 48.58% are not financially inclusive. More specifically, when analyzing the distribution of financial inclusion in Peruvian households by gender, based on a sample of 14,240 individuals, it is evident that 55.15% are men and 44.85% are women, indicating a slightly higher proportion of men in the analyzed sample. When disaggregated by financial inclusion status, 48.6% are not financially included, whereas 51.4% are.

Among those without access to formal financial services, 57.20% are men and 42.80% are women. Conversely, among those who do have access, 53.22% are men and 46.78% are women. This comparison indicates a slight improvement in the relative participation of women within the financially included group. However, men continue to outnumber women in both absolute and proportional terms across both groups (Table 2).

Based on Pearson’s Chi-Square (χ²) test of independence for this distribution, which yielded a value of 22.73 with a p-value < 0.001, there is a statistically significant association between gender and financial inclusion. In other words, access to financial services is not independent of gender, and a measurable gap exists that cannot be attributed to random variation (Table 2).

These results underscore the persistence of a gender gap in financial inclusion, with women still showing lower participation than men. Although the percentage difference may not appear very large, its statistical significance indicates that this gap is structural and should be considered in the design of public policies that promote equal access to the formal financial system.

The analysis of socioeconomic differences by gender, detailed in

Table 3. Mean comparison of key socioeconomic variables by gender.

Variable	Men’s Mean	Women’s Mean	Difference	T-Value	p-Value
Years of education	10.343	9.880	0.463	7.179	0.0000
Log household monthly income	6.561	5.911	0.650	28.088	0.0000
Age	45.275	44.302	0.973	4.339	0.0000
Household size	3.904	3.863	0.041	1.338	0.1808

, highlights gender disparities in key socioeconomic variables that help explain the determinants of financial inclusion in Peruvian households. A difference-of-means test (Student’s t-test) was used to evaluate whether men and women differ significantly in structural characteristics that typically influence access to and use of financial services.

There is a significant gap in years of education attained: men have an average of 10.34 years of schooling, while women have 9.88 years—a difference of 0.46 years, nearly half a year. Despite appearing moderate, this difference is statistically significant (p < 0.001), pointing to a persistent educational lag among women. This educational disadvantage, though smaller than in past decades, may still hinder women’s financial literacy and their ability to understand and use complex financial products.

Regarding monthly household income, a substantial gap exists: men report an average log household income of 6.56 compared to 5.91 for women. This 0.650 difference in natural logarithmic units represents a significant disparity in absolute income terms. The difference is highly significant (p < 0.001) and reflects structural labor market inequalities such as occupational segregation, informality, and the glass ceiling, which disproportionately affect women. These income gaps have direct implications for financial inclusion, given that income is a critical factor in bank usage and access to formal financial services (Table 3).

The age of the members of the household is a variable that, although it does not measure work experience directly, is a common and useful proxy, especially if there is no detailed data on work history. In this sense, it explains financial inclusion, since people of working age are more likely to be banked than people who do not meet this condition; therefore, analyzing this variable, we can see that men are on average 45 years old and women 44 years old, showing a difference of almost one year (p < 0.001). Therefore, although the magnitude of this difference is low, its statistical significance demonstrates a slightly greater presence of men in the most advanced age groups, which has marginal implications for financial inclusion, insofar as age is associated with greater work experience, economic stability, and cumulative access to financial services (Table 3).

Regarding household size, there is no significant difference between men and women (3.90 vs. 3.86 people, p = 0.1808), indicating that both genders live in households of similar sizes. Thus, this variable is not a relevant differentiating factor in gender-based financial inclusion in the analyzed sample (Table 3).

When analyzing financial inclusion by gender and area of residence, 34.33% of included individuals are urban women, and only 10.52% are rural women, while 36.03% are urban men and 19.12% are rural men. This reveals a high concentration in urban areas for both men and women.

In terms of access to the financial system, the data show that urban men represent the group with the highest inclusion (39.58% of included individuals), followed by urban women (37.15%). This indicates that although a slight gender gap exists, access to formal financial services in urban areas tends to be relatively high and equitable (

Table 4. Financial inclusion gaps by urban–rural residence, disaggregated by gender.

Variable	Category	Group				Total
		Urban Women	Rural Women	Urban Men	Rural Men
Financial Inclusion	No	2168	793	2233	1724	6918
	%	31.34	11.46	32.28	24.92	100.00
	Yes	2720	705	2898	999	7322
	%	37.15	9.63	39.58	13.64	100
Total		4888	1498	5131	2723	14,240
%		34.33	10.52	36.03	19.12	100.00

).

However, this situation changes drastically in rural areas, where only 13.64% of the included are rural men, and just 9.63% are rural women. These figures reflect both a geographic and a gender-based gap, with rural women being the group with the lowest level of financial inclusion in both absolute and relative terms (Table 4).

Furthermore, among the financially excluded, the highest percentages are also found in urban areas: 32.28% are urban men, and 31.34% are urban women. Although these values are high due to larger urban populations, proportionally, exclusion is more severe in rural areas: 24.92% of the excluded are rural men, and 11.46% are rural women. These results highlight a double vulnerability for rural women, who face barriers due to both gender and geography. The gender–financial inclusion area gap is thus a structural issue that requires differentiated attention in public policy (Table 4).

Therefore, by conducting the density analysis of the key variables in the study (Figure 1), a deeper understanding of behavioral patterns can be achieved. The analysis of the age distribution reveals that both men and women with financial inclusion exhibit similar density curves, although with subtle differences in concentration. The density corresponding to men (bandwidth = 2.2447) shows a slight shift toward older ages compared to women (bandwidth = 2.2665), suggesting that, within the group with access to financial services, men tend to be slightly older. This difference may be related to longer career trajectories or greater financial stability in later stages of life, which facilitates access to financial products.

Regarding the logarithm of total monthly household income, a marked gender difference is observed. The distribution for men (bandwidth = 0.1660) is skewed toward higher income values compared to that of women (bandwidth = 0.2434), indicating a higher concentration of men in the upper income levels among users of the financial system. This asymmetry highlights a structural economic gap, even among those who already have access to financial services. Furthermore, the greater dispersion observed in the female density curve reflects a higher degree of income heterogeneity among financially included women, which may be explained by their participation in less stable or lower-paid occupations.

In the case of area of residence, the results show that most financially included individuals, both men and women, live in urban areas. However, the density for women (bandwidth = 0.0715) indicates a slightly higher representation of rural women compared to men (bandwidth = 0.0752) within the financially included group. This may be related to targeted interventions or policies aimed at rural women, such as microfinance programs or conditional cash transfers, which have contributed to their increased incorporation into the formal financial system.

When analyzing the distribution of years of education received, the data show that financially included men tend to have higher and more homogeneous educational levels, as evidenced by the greater concentration of the curve (bandwidth = 0.3830). In contrast, the density for women (bandwidth = 0.6550) is more dispersed, reflecting greater variability in the educational levels of financially included women. This pattern highlights a persistent reality: although more women are gaining access to the financial system, many do so from less favorable educational conditions, which may limit their ability to use or fully benefit from the available financial services.

In this regard, Figure 1 collectively demonstrates that gender gaps in financial inclusion are not limited to access alone but are also mediated by significant differences in age, income, residence, and education. These findings underscore the importance of promoting differentiated policies that address structural inequalities in order to achieve effective, equitable, and sustainable financial inclusion.

4.2. Application of the Oaxaca–Blinder Model to Explain the Gender Gap in Financial Inclusion

4.2.1. Gender Gap in Financial Inclusion: Magnitude and General Decomposition

Based on the results presented in Table 5, the existence of a negative gender gap of −2.16 percentage points in financial inclusion is confirmed. This indicates that women have a lower probability of being financially included compared to men, a difference that is statistically significant at the 10% level. Furthermore, this gap is decomposed into three subcomponents:

• Explained component (endowments), which accounts for 3.21 percentage points (p < 0.01). This indicates that if women had the same observable characteristics as men (e.g., education, income, and age), their financial inclusion would actually be higher.
• The unexplained component (coefficients), valued at −7.24 percentage points (p < 0.01), shows that the largest portion of the gap is due to differences in how the financial system values these characteristics between men and women.
• Interaction component (combined effect of characteristics and returns), valued at 1.88 percentage points (p < 0.05), indicates a partial moderation of the gap.

Table 5. Gender gap model approach: Oaxaca–Blinder decomposition.

	(1)	(2)	(3)	(4)
Variables	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition
Natural logarithm of total monthly household income		0.0192 ***	0.108 **	0.0136 **
		(0.00473)	(0.0515)	(0.00653)
Age		−0.00413 ***	0.106 **	0.00301 **
		(0.00141)	(0.0421)	(0.00143)
Years of education received		0.0132 ***	0.118 ***	0.00702 ***
		(0.00228)	(0.0304)	(0.00207)
Marital status		−0.00655 ***	−0.0294	0.00412
		(0.00215)	(0.0196)	(0.00278)
Household size		2.31 × 10−5	−0.0601 **	−0.000708
		(0.000237)	(0.0265)	(0.000717)
Main occupational category		0.0103 ***	0.121	−0.00372
		(0.00228)	(0.0772)	(0.00240)
Area of residence		−0.00127	0.0425 **	−0.00391 **
		(0.00149)	(0.0212)	(0.00197)
Geographical region of household		0.00103	−0.0106	−0.000143
		(0.00106)	(0.0267)	(0.000388)
Participation in a social program		0.000253	−0.0248	−0.000528
		(0.000341)	(0.0164)	(0.000592)
Group 1	0.509 ***
	(0.00791)
Group 2	0.531 ***
	(0.00876)
Difference	−0.0216 *
	(0.0118)
Endowments	0.0321 ***
	(0.00662)
Coefficients	−0.0724 ***
	(0.0130)
Interaction	0.0188 **
	(0.00795)
Constant			−0.443 ***
			(0.131)
Observations	14,240	14,240	14,240	14,240

Standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

Table 5. Gender gap model approach: Oaxaca–Blinder decomposition.

	(1)	(2)	(3)	(4)
Variables	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition	Gender Gap: Oaxaca–Blinder Decomposition
Natural logarithm of total monthly household income		0.0192 ***	0.108 **	0.0136 **
		(0.00473)	(0.0515)	(0.00653)
Age		−0.00413 ***	0.106 **	0.00301 **
		(0.00141)	(0.0421)	(0.00143)
Years of education received		0.0132 ***	0.118 ***	0.00702 ***
		(0.00228)	(0.0304)	(0.00207)
Marital status		−0.00655 ***	−0.0294	0.00412
		(0.00215)	(0.0196)	(0.00278)
Household size		2.31 × 10−5	−0.0601 **	−0.000708
		(0.000237)	(0.0265)	(0.000717)
Main occupational category		0.0103 ***	0.121	−0.00372
		(0.00228)	(0.0772)	(0.00240)
Area of residence		−0.00127	0.0425 **	−0.00391 **
		(0.00149)	(0.0212)	(0.00197)
Geographical region of household		0.00103	−0.0106	−0.000143
		(0.00106)	(0.0267)	(0.000388)
Participation in a social program		0.000253	−0.0248	−0.000528
		(0.000341)	(0.0164)	(0.000592)
Group 1	0.509 ***
	(0.00791)
Group 2	0.531 ***
	(0.00876)
Difference	−0.0216 *
	(0.0118)
Endowments	0.0321 ***
	(0.00662)
Coefficients	−0.0724 ***
	(0.0130)
Interaction	0.0188 **
	(0.00795)
Constant			−0.443 ***
			(0.131)
Observations	14,240	14,240	14,240	14,240

Standard errors in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

Thus, these patterns demonstrate that the inequality does not stem from women’s observable capacities; rather, it results from how these capacities are differentially valued by the financial system.

4.2.2. Comparison Between the Unweighted Model and the Weighted Model (Simple Adjustment)

A comparison of the unweighted model (Table 5, column 1 and overall differences) reveals the baseline gap, where the difference in financial inclusion between groups (Group 1: men; Group 2: women) is −0.0216, confirming lower financial inclusion among women. The explained portion has a value of 0.0321, indicating that if women had the same levels of income, education, and other characteristics as men, their likelihood of financial inclusion would be higher. However, the effect of the returns on these characteristics is negative and of greater magnitude (−0.0724), thereby confirming the existence of a structural bias in favor of men.

In the weighted model (Table 5, columns 2, 3, and 4), an adjustment is made to correct for potential sampling imbalances. The results reveal the following:

• The natural logarithm of household income has a significant and positive effect on both the endowment component (0.0192, p < 0.01), especially on the return’s component (0.108, p < 0.05), indicating that monthly household income contributes more to improving financial inclusion for men than for women.
• In the case of age, the value is negative in the endowment component (−0.00413, p < 0.01) but positive and significant in the return’s component (0.106, p < 0.05), demonstrating that age is valued differently and more favorably for men.
• Years of education have a positive effect on both components, though with a stronger impact on returns (0.118, p < 0.01), highlighting education as a crucial factor, but one that is more highly rewarded for men.
• Marital status has a negative and significant effect only on the endowment component (−0.00655, p < 0.01), indicating that women in certain family conditions face additional barriers to financial access.
• Household size, although showing no effect in the endowment component, has a significantly negative effect in the return’s component (−0.0601, p < 0.05), suggesting that this variable disproportionately reduces financial inclusion for women.
• Regarding area of residence (urban vs. rural), the endowment component shows no significant effect, while the returns component is significantly positive (0.0425, p < 0.05), indicating that living in urban areas offers greater financial benefits for men than for women.
• Occupational category (working in the formal sector) shows favorable endowments for men (0.0103, p < 0.01), although the returns are not statistically significant.
• Other variables, such as the household’s geographical region (Coast, Highlands, or Jungle) and participation in social programs, do not show statistically significant contributions, although they may be relevant from a qualitative or public policy perspective.

Therefore, when interpreting the constant term of the returns model (−0.443, p < 0.01), it can be understood as representing the portion of the gap that cannot be explained by either endowments or differential returns. Given its large and negative value, it reflects the presence of a deep and persistent structural bias within the Peruvian financial system—one that penalizes women for factors not captured by the model, such as social norms, institutional biases, and lack of financial history.

In this regard, the results from both models indicate that the gender gap in financial inclusion in Peru in 2024 is primarily explained by the unequal treatment of similar characteristics between men and women, rather than by deficiencies in women’s profiles. The weighted model further confirms that monthly household income, age, and years of education are key variables for financial inclusion, yet their effects are more highly valued in men. Moreover, the findings make it clear that the inequality stems not from lower capabilities but from unfair and unequal valuation, highlighting the urgent need for inclusive financial policies with a gender perspective, particularly those aimed at addressing biases in credit evaluation mechanisms, access to financial products, and the development of women’s economic capacities.

5. Discussion

Following the results confirming that the gender gap in financial inclusion in Peru during 2024 is not exclusively due to observable differences between men and women (such as income, education, or age) but primarily to differences in the returns these characteristics generate for each group, it becomes evident that structural issues are at play. The unexplained component of the Oaxaca–Blinder decomposition accounts for more than 70% of the total gap, indicating the existence of structural and potentially discriminatory factors that hinder women’s financial inclusion, even when they possess similar characteristics to men.

In this regard, the findings of the present study align with those of Ñopo (2012), who, using a matching technique to analyze the gender wage gap in Latin America, found that women, despite having similar characteristics to men, do not receive equivalent returns in the labor market. This suggests that similar mechanisms may be present in the financial system, where women face institutional or social biases that reduce their chances of inclusion, even when they meet comparable conditions.

These results are also consistent with the findings of Chalup et al. (2023), who examined the evolution of the gender wage gap in Peru between 1997 and 2021 using the same Oaxaca–Blinder technique. They found that, although differences in endowments have diminished over time, the unexplained component attributable to discrimination or unobservable factors remained constant, demonstrating persistent structural inequality in the labor market, even amid improvements in women’s education. In this sense, the scenario mirrors that observed in the financial system, where women, despite considerable educational and economic progress in recent years, continue to face disadvantages in accessing and utilizing financial services.

Similarly, the findings are in line with those of Demirgüç-Kunt et al. (2022), who showed that the gender gap in financial inclusion persists globally, despite advances in access to digital banking services. Their analysis, based on the 2021 Global Findex database, demonstrates that in middle-income countries such as Peru, women are less likely to have a bank account or access to formal credit even if they are employed or have a regular income. This inequality is more pronounced in rural areas and among those with lower levels of education, which is supported by the present research: variables such as “area of residence” and “years of education received” show an unequal impact across genders.

Along the same lines, the results also align with Andrade and Buvinic (2019), who emphasize that a key strategy to close the gender gap in financial inclusion is to improve the availability and use of gender-disaggregated data. Their experience in Mexico showed that systematic data collection enabled the design of more inclusive, women-focused policies. This highlights that, for Peru to move forward, it is essential to have information systems capable of accurately assessing how financial inclusion varies between men and women and how each group responds to different public policies.

Consistent with the above, similar findings were reported in the Peruvian context by Boitano and Abanto (2020), who noted that although access to the financial system has increased in recent years due to digitalization and social programs, the gender gap remains constant or is even widening, especially among women in rural areas or with lower educational attainment. Like the present study, they argue that improving women’s objective conditions (such as income or education) is not enough; what is also needed is a transformation in how the financial system values these characteristics.

In addition, the findings support the arguments of Sahay et al. (2025), who contend that greater financial inclusion of women not only reduces inequality but also drives economic growth. Through multivariate analysis, they demonstrated that countries with smaller gender gaps in financial access tend to have better indicators of well-being and development. Applying this to the Peruvian case, they showed that reducing the gender gap in financial inclusion is not merely a matter of equity but also one of economic efficiency and sustainable development.

Lastly, the recent study by Beirne et al. (2023), which offers a cross-country analysis, emphasizes that financial institutions must redesign products and service channels to better meet women’s specific needs, such as digital security, geographic accessibility, and financial education support. This conclusion is consistent with the findings of the present study, which show that despite controlling for observable variables, inequality persists—suggesting that the solution lies not only in closing observable gaps, but also in reforming the incentives and structures within the financial system.

6. Conclusions

The results of this research confirm the existence of a statistically significant gender gap in financial inclusion in Peru, even after controlling for socioeconomic and demographic factors; on average, women are 2.16 percentage points less likely to be financially included compared to men, showing a persistent and structural disadvantage in accessing and using formal financial services.

Through the Oaxaca–Blinder decomposition model, it was evidenced that most of this gap is not due to differences in observable characteristics (such as age, education, household income, or geographical location); on the contrary, it is due to the differentiated returns that these characteristics generate according to gender. That is why the unexplained component represents a negative contribution of −0.0724, significantly higher than the explained component (0.0321); therefore, the presence of institutional, social, and cultural mechanisms disproportionately limits women’s financial inclusion, even when they have similar conditions to those of men.

Variables such as years of schooling, monthly household income, and age are important determinants of financial inclusion, although their effect varies between genders; factors such as marital status, main occupation, and area of residence show effects differentiated by sex, evidencing that the financial system in Peru does not operate in a gender-neutral manner, which reaffirms that improving only the individual characteristics of women is not enough to close the gap, so it is necessary to transform the structures of the Peruvian financial system that perpetuate inequalities of access.

Finally, the limitations of this research are the use of cross-sectional data from ENAHO, which does not allow for the analysis of changes over time or the establishment of causal relationships; the Oaxaca–Blinder model presents methodological restrictions, such as difficulty in interpreting the unexplained component, and the measurement of financial inclusion is limited to basic access, without considering quality, autonomous use, or other relevant dimensions. Therefore, it is recommended that future research use longitudinal data or panels to capture temporal dynamics and incorporate qualitative approaches that can explore invisible barriers such as social norms or gender discrimination, including in their analysis of other variables, such as financial education, digital use, and economic autonomy.