Sage Journals: Discover world-class research

Abstract

Income inequality is one of the most critical challenges facing global economic development today, as it undermines social fairness and impedes the healthy development of the economy. While extensive research has explored the determinants of income inequality, the role of financial literacy in addressing this issue remains underexplored. We use data from the China Family Panel Studies and apply a double machine learning model to examine the impact of financial literacy on income inequality. Our empirical analysis demonstrates that financial literacy significantly alleviates income inequality, and this conclusion holds after a series of robustness tests. Mechanism analysis identifies household entrepreneurship, financial participation, and insurance participation as critical channels through which financial literacy alleviates income inequality. Furthermore, moderation effects reveal that the development of the digital economy strengthens financial literacy’s mitigating effect on income inequality. The study also finds that financial literacy’s effect on reducing income inequality is more pronounced in eastern regions, urban areas, households with higher educational attainment, and elderly households. In summary, this research contributes to understanding the practical effects of financial literacy and provides new insights for policymakers seeking to address income inequality through financial literacy.

Keywords

financial literacy income inequality double machine learning mechanism analysis heterogeneity analysis

Introduction

As a critical issue in global economic development, income inequality has attracted widespread attention. With the rise of globalization, technological advancements, and the growth of market economies, the income gap across countries and regions has widened (Huijsmans et al., 2022). Income inequality affects economic stability and has far-reaching impacts on social justice, political stability, and public health (Lee et al., 2020; McFarland et al., 2023). For instance, societies with high inequality often experience lower social mobility, making it difficult for individuals and families to improve their economic opportunities through effort (Kim et al., 2023). Furthermore, income inequality may exacerbate social conflicts, undermine governmental legitimacy, and reduce the effectiveness of policy implementation (Kunawotor et al., 2020). The complexity of income inequality lies in its multidimensional and diverse nature, extending beyond individual or household income differences, including disparities in healthcare access, educational opportunities, and career advancement (L. K. Chu & Hoang, 2020; C. Ma et al., 2021). Theoretically, studying income inequality helps to enrich the framework of economic growth and distribution. Practically, understanding the mechanisms behind inequality can inform the development of more effective public policies to promote social equity and sustainable economic development.

Existing literature analyzes income inequality from various perspectives, including economic growth, tax policy, labor markets, and education and skill development (L. K. Chu & Hoang, 2020; Mdingi & Ho, 2021; Taghizadeh-Hesary et al., 2020). Traditional economic theory suggests that economic growth can raise income levels, reduce poverty, and narrow income gaps to some extent (Mdingi & Ho, 2021). However, the benefits of economic growth often concentrate among high-income groups, exacerbating inequality (Kavya & Shijin, 2020). Thus, relying solely on economic growth is insufficient to address income inequality effectively.

Tax policy is a vital tool for addressing income inequality. Many scholars have examined the role of progressive taxation and transfer payments in income redistribution (Taghizadeh-Hesary et al., 2020). By imposing higher tax rates on higher-income groups, progressive tax systems help narrow income gaps. Meanwhile, transfer payments through social security and welfare policies can effectively raise the living standards of low-income groups. Empirical studies show that countries with progressive tax systems and robust social security programs tend to have lower levels of income inequality (Seelkopf & Lierse, 2020). However, excessively high tax burdens may reduce economic dynamism and hinder long-term growth.

Research on labor market policies focuses on raising wages for low-income groups and promoting employment equity. Fundamental mechanisms include minimum wage policies, protection of workers’ rights, and initiatives to increase labor market participation among women and minority groups (De Pleijt & Van Zanden, 2021; Engbom & Moser, 2022). Technological advancements and globalization have also intensified skill-biased inequality in the labor market, further widening the income gap between high-skilled and low-skilled workers (Hutter & Weber, 2023). Education and skill development are widely seen as critical paths to mitigating long-term income inequality. Numerous studies indicate that improving access to education and enhancing vocational training can help low-income individuals secure better employment opportunities and raise their income levels (Nikensari et al., 2021). However, the effects of education often take time to materialize and may not significantly reduce income disparities in the short term (Liu et al., 2020). Overall, while significant progress has been made in studying income inequality, the dynamic nature of economic and social environments necessitates further exploration and ongoing evaluation of these issues.

In recent years, research on the relationship between financial literacy and household income has gained significant attention in economics. As a critical factor influencing personal and household financial decisions, financial literacy directly affects behaviors such as saving, investing, and debt management, profoundly impacting household income levels and wealth accumulation (Bayar et al., 2020). Scholars have noted that households with higher financial literacy are more capable of making effective financial decisions, such as managing expenses, selecting efficient investment tools, and mitigating financial risks. This, in turn, enables them to achieve faster wealth growth and higher income (Setiawan et al., 2022). From various perspectives, existing studies explore the relationship between financial literacy and household income. Research has shown a positive correlation between financial literacy and household income, indicating that households with higher financial literacy tend to have higher incomes (Prasad et al., 2018). This phenomenon may be linked to higher-income families having more resources to access financial education and information. Additionally, financial literacy has also been found to have a certain moderating effect on income inequality (G. Xu et al., 2024). Some studies suggest that households with lower financial literacy are more likely to fall into high debt and low savings traps, further widening the income gap (Berry et al., 2018). Furthermore, the development of the digital economy has facilitated deep integration between traditional financial services and digital platforms, significantly enhancing both accessibility and inclusivity. This trend enables financial services to reach a broader range of low-income households and families with limited financial literacy, thereby creating conditions for their entry into financial markets and improving the efficiency of resource allocation.

Although a substantial body of literature has examined the relationship between financial literacy and income, several gaps remain to be addressed. First, most existing studies focus on developed countries, and their conclusions may not apply to the institutional and economic contexts of developing economies. In China, the world’s largest developing country, income inequality is particularly pronounced, not only in terms of the rural-urban divide but also in regional disparities and unequal access to opportunities across groups. We calculate China’s Gini coefficient for the period 2013 to 2022 using data from the China Statistical Yearbook and the methodology developed by Tian (2012; Figure 1). The trend in China’s Gini coefficient from 2013 to 2022 further illustrates the widening gap in income distribution. However, limited attention has been paid to income inequality in the Chinese context, especially from the perspective of financial literacy.

Figure 1.

Gini coefficient of China from 2013 to 2022.

Second, existing research primarily investigates the effect of financial literacy on overall income levels. At the same time, relatively little attention is given to its mechanisms and transmission channels in shaping income inequality within groups. With the rapid development of China’s financial markets and the increasing complexity of financial products, financial literacy has become a crucial factor influencing household wealth management, investment decisions, and income growth. Differences in financial literacy across groups may further exacerbate income inequality, yet systematic studies on this issue remain scarce.

Finally, in terms of methodology, most prior studies rely on traditional OLS regression models for empirical analysis, which struggle to address potential endogeneity concerns adequately. By contrast, the double machine learning (DML) approach offers greater robustness and applicability for causal inference, but its use in this field remains limited. Against this backdrop, this paper not only focuses on the relationship between financial literacy and income inequality in China but also applies the DML method, thereby extending the existing literature in both theoretical and methodological dimensions.

In the context of rapid digital financial development, exploring how to improve financial literacy to reduce income inequality remains a topic of significant research value. First, does financial literacy help mitigate income inequality? Second, through what mechanisms does financial literacy reduce income inequality? Finally, how can we enhance financial literacy to alleviate income disparities in the digital age? Our study contributes in several vital ways. First, by placing financial literacy and income inequality within the same analytical framework, we employ a double machine learning approach to investigate the role of financial literacy in alleviating income inequality, thereby contributing to narrowing income gaps and alleviating inequality. Second, our findings suggest that financial literacy reduces income inequality by promoting household entrepreneurship, financial participation, and insurance participation. These insights provide a valuable understanding of enhancing household income and reducing inequality. Additionally, our moderation effects analysis reveals that the development of the digital economy strengthens the mitigating effect of financial literacy on income inequality. Lastly, heterogeneity analysis shows that financial literacy has a more pronounced effect on reducing income inequality in the central and western regions, rural areas, and among households with low education levels and elderly households. This suggests that financial literacy can help disadvantaged households overcome resource and information constraints, promote income growth, and narrow income gaps.

The rest of this paper is organized as follows: Section 2 provides a literature review. Section 3 provides a theoretical analysis. Section 4 introduces the data sources, variable design, and methods. Section 5 analyzes and discusses the empirical results. Section 6 provides conclusions and discussion.

Literature Review

As a crucial economic capability, financial literacy helps individuals effectively manage wealth and make rational financial decisions, thereby influencing income levels and wealth accumulation (Bayar et al., 2020). There has been extensive discussion in the literature regarding the definition and measurement of financial literacy. Lusardi and Mitchell (2014) define financial literacy as the basic concepts and skills individuals possess related to financial behaviors, such as saving, investing, and borrowing. This knowledge includes understanding key financial principles like compound interest, risk diversification, and inflation (Twumasi et al., 2022). To measure financial literacy, scholars typically design surveys that assess respondents’ grasp of financial concepts. Common survey questions cover topics such as compound interest calculations and the trade-off between risk and return (Chang et al., 2022; J. Tan et al., 2022). These measurement tools provide a solid foundation for subsequent research on the impact of financial literacy on income.

Theoretically, financial literacy can influence individual income levels through several pathways. First, individuals with higher financial literacy tend to make more rational economic decisions, optimizing their savings and investment choices and boosting long-term income (Jappelli & Padula, 2013; Twumasi et al., 2022). Second, financial literacy enhances individuals’ ability to manage risks, enabling them to better cope with economic fluctuations or unexpected financial crises, thus reducing the likelihood of income loss (Banks et al., 2020; Lusardi & Mitchell, 2017). Additionally, financial literacy helps individuals avoid irrational financial behaviors, such as excessive borrowing, and prevents them from falling into financial distress, supporting a higher income level (Berry et al., 2018).

A substantial body of research supports that financial literacy enhances income. Lusardi et al. (2017) demonstrate that individuals with higher financial literacy tend to have greater savings rates and investment returns, leading to higher income levels. Specifically, their analysis of survey data from multiple countries reveals that individuals with basic financial literacy earn significantly more than those lacking such knowledge. Van Rooij et al. (2011) further indicate that financial literacy affects an individual’s current and future income growth by influencing investment behavior and optimizing asset portfolios. Some scholars have empirically analyzed the poverty alleviation effects of financial literacy, finding that it improves the economic conditions of poor rural populations, thereby increasing their income (J. Tan et al., 2022; H. Xu et al., 2023). The causal relationship between financial literacy and income has sparked considerable scholarly debate. Some researchers argue that higher income levels also promote financial literacy accumulation, as higher-income individuals often have more educational opportunities and access to financial information (Behrman et al., 2012; Prasad et al., 2018).

Additionally, the literature indicates that the impact of financial literacy on income may vary significantly across different demographic groups. For instance, Atkinson and Messy (2012) find that women and low-income groups typically possess lower levels of financial literacy, which limits their potential for income growth. Therefore, enhancing these groups’ financial literacy is a crucial strategy for narrowing income disparities. Moreover, factors such as the level of economic development and the maturity of financial markets in different countries may also influence the relationship between financial literacy and income (Klapper et al., 2013).

A Theoretical Analysis of the Impact of Financial Literacy on Income Inequality

Within the existing research framework, we analyze the role of financial literacy in mitigating income inequality through three main channels: entrepreneurship, financial participation, and insurance participation. The rationale is as follows. First, entrepreneurship is a crucial avenue for enhancing household income, yet it often entails high risks and uncertainties. Households with higher levels of financial literacy are better equipped to evaluate investment projects, allocate resources efficiently, and manage risks effectively. These advantages increase the likelihood of entrepreneurial success and income growth, which is particularly critical for low-income households. Second, financial participation serves as a key mechanism for wealth accumulation and income growth. Financially literate households are more likely to access formal financial markets, make informed decisions about savings, investments, and credit, and improve the efficiency of capital utilization. Over time, these behaviors help narrow the income gap with high-income households. Ultimately, insurance participation is a crucial tool for mitigating income volatility and promoting household stability. Low-income households are especially vulnerable to shocks such as illness, unemployment, or unexpected events. Improved financial literacy enhances their understanding and acceptance of insurance products, facilitates risk sharing, and reduces the likelihood of widening inequality. The theoretical framework of this study on the impact of financial literacy on income inequality can be summarized in Figure 2.

Figure 2.

Theoretical framework.

Mediation Effect Analysis

Entrepreneurship Channel

Financial literacy enhances a household’s understanding of financial products and services, enabling more effective use of tools such as loans, investments, and insurance (Twumasi et al., 2022). This knowledge facilitates access to necessary funding for entrepreneurship by reducing financial barriers, allowing families to enter the market, and creating economic value. Households with financial literacy are better equipped to engage in sound financial planning and risk management (Berry et al., 2018). This ability helps families navigate market fluctuations and financial crises during the entrepreneurial process, reducing the risk of failure. Successful entrepreneurship increases household income and fosters wealth accumulation, which helps reduce income disparities between households (Mohamad et al., 2021). Furthermore, financial literacy can be disseminated through education and training, particularly in low-income communities (Prasad et al., 2018). This process enhances overall financial capability, improving the success rate of household entrepreneurship and, on a macro level, mitigating income inequality. Therefore, financial literacy supports the reduction of income inequality by strengthening households’ entrepreneurial capacity.

Financial Participation Channel

Financial literacy improves a household’s understanding of financial products and services and enhances its decision-making abilities. With financial literacy, families are better equipped to understand and use various financial tools, such as savings accounts, investment products, and credit services, contributing to wealth accumulation (Murendo & Mutsonziwa, 2017; Sayinzoga et al., 2015). Households with financial literacy are more likely to engage in higher-return investments, such as the stock market, and achieve relatively higher investment returns (He & Li, 2020). However, as digital finance rapidly advances, low-income households often struggle to access the convenience of digital financial services. Low-income households generally possess limited financial literacy compared to high-income households, which have accumulated significant financial literacy under traditional financial systems (C. Guo et al., 2022). Financial literacy is critical in how families gather and process financial information, allowing them to make sound financial decisions. Low-income families with adequate financial literacy can skillfully use digital finance to diversify financial assets, reduce risk, and increase returns (Z. Chu et al., 2017). This contributes to reducing income inequality between low- and high-income groups. Therefore, household financial participation is a crucial channel through which financial literacy mitigates income inequality.

Insurance Participation Channel

Financial literacy enables households to understand insurance products’ function and importance, enhancing their risk management awareness. First, financially literate households are more likely to identify the risks they face, such as health, property, and unemployment risks, and proactively participate in insurance to mitigate the potential impact of these risks on household income (M. Guo et al., 2024). Second, insurance participation improves a household’s financial security. With insurance, households can receive compensation in unexpected crises, preventing sudden income loss (S. Xu et al., 2022). This financial protection reduces economic stress and boosts household consumption, contributing to economic growth. Specifically, insurance helps manage short-term financial crises, allowing families to maintain relatively stable income levels and avoid falling into financial hardship due to unforeseen events (Duan et al., 2022). Finally, broader participation in insurance can lead to wider societal benefits. As more households are insured, the overall effect of risk pooling strengthens, helping to reduce socioeconomic inequality (Barasa et al., 2021). Thus, household insurance participation is critical in the relationship between financial literacy and income inequality.

Based on the theoretical analysis above, we propose the following hypotheses:

Hypothesis 1: Financial literacy contributes to reducing income inequality.

Hypothesis 2: Financial literacy alleviates income inequality by promoting household entrepreneurship, financial participation, and insurance participation.

Moderation Effect Analysis

Moderating Effect of the Digital Economy

The rise of the digital economy presents new opportunities for financial literacy to mitigate household income inequality. On the one hand, the widespread adoption of digital technologies has made access to financial information more convenient, allowing households to gain comprehensive knowledge about financial products and services through various online platforms and applications (Y. Tan & Li, 2022). This transparency reduces households’ knowledge barriers in financial decision-making and increases their willingness to participate in financial markets (Liang & Guo, 2015). On the other hand, the digital economy offers households more financial service options. For example, mobile payments, internet banking, and online investment platforms enable households to save and invest easily (Sudiantini et al., 2023). These services increase financial participation and enhance flexibility in asset allocation, helping households accumulate and grow wealth. By effectively utilizing digital financial tools, households can better manage risk, optimize income structures, and ultimately achieve income growth (Wu & Wu, 2023). Furthermore, the rapid growth of the digital economy has accelerated the spread of financial education, raising the overall financial literacy of society (Firmansyah & Susetyo, 2022). As financial literacy improves, households become more capable of responding to economic inequality, seizing opportunities, and participating in economic activities. In summary, the digital economy enhances financial literacy accessibility and promotes household financial participation, strengthening its role in reducing income inequality.

Based on the above theoretical analysis, we propose the following hypothesis:

Hypothesis 3: The digital economy plays a negative moderating role in the relationship between financial literacy and income inequality, meaning that the development of the digital economy enhances the mitigating effect of financial literacy on income inequality.

Data, Variable Design, and Methods

Data Source

Using data from the China Family Panel Studies (CFPS) for this study offers several advantages. The CFPS data covers urban and rural households, making it highly representative and providing comprehensive information on families from different regions and income levels. This enables a thorough analysis of income inequality (Luan et al., 2023; J. Zhang & Li, 2024). The first survey round was conducted in 2010 and has been repeated every 2 years. The latest data is from 2020. Notably, only the 2014 questionnaire from the CFPS 2010 to 2020 data set includes questions related to household financial literacy. Therefore, we selected the 2014 data as our research sample. In addition, the indicators used to measure the digital economy are drawn from the China Statistical Yearbook, provincial statistical yearbooks, statistical bulletins, and the Digital Inclusive Finance Index released by the Institute of Digital Finance at Peking University. After excluding observations with missing values for relevant variables, the final sample size used in our analysis is 3,032 households.

Variable Design

Dependent Variable

This study uses income inequality as the dependent variable, measuring it with the Kakwani index to assess household-level income inequality. The Kakwani index evaluates relative deprivation based on total household income data collected from the CFPS survey, effectively reflecting the degree of income inequality among families. The calculation method is as follows: within a group of n households, the incomes are arranged in ascending order to create the overall income distribution function $Y = (y_{1}, y_{2}, y_{3}, \cdot \cdot \cdot \cdot \cdot, y_{n})$ , where $y_{1} \leq y_{2} \leq y_{3} \leq \cdot \cdot \cdot \cdot \cdot \leq y_{n}$ . According to the definition of the income relative deprivation index, we can compare each household’s income with that of others, representing each household’s relative deprivation index as follows:

$D (y_{j}, y_{i}) = {\begin{matrix} y_{j} - y_{i}, y_{j} > y_{i} \\ 0, y_{j} \leq y_{i} \end{matrix}$ (1)

In Equation 1, $D (y_{j}, y_{i})$ represents the relative deprivation index of household j concerning household i. By summing this index for household j and dividing the result by the mean income within the group, we can determine the level of income inequality for household i. This is expressed as follows:

$\begin{matrix} Index (y_{i}) = \frac{1}{n μ_{Y}} \sum_{j} D (y_{j}, y_{i}) \\ = \frac{1}{n μ_{Y}} (n_{y_{i}}^{+} \times μ_{y_{i}}^{+} - n_{y_{i}}^{+} \times y_{i}) \\ = \frac{1}{μ_{Y}} \times \frac{n_{y_{i}}^{+}}{n} (μ_{y_{i}}^{+} - y_{i}) \end{matrix}$ (2)

In Equation 2, $μ_{Y}$ denotes the average income of all households within the group, $μ_{y_{i}}^{+}$ represents the average income of all households in group Y whose income exceeds $y_{i}$ , and $n_{y_{i}}^{+}$ is the number of households in group Y with an income greater than $y_{i}$ . Finally, $Index (y_{i})$ represents the income inequality level for household i, with values ranging from 0 to 1.

Independent Variable

The independent variable in this study is financial literacy. The 2014 CFPS Household Economic Questionnaire includes 13 questions covering topics such as interest rates, bank deposits, inflation, funds, wealth management, and investment risks. Each question has a standard correct answer, effectively measuring respondents’ financial literacy. Table 1 presents these questions and their correct answers. A correct response is coded as 1, while an incorrect response or “do not know” is coded as 0. To calculate a composite measure of financial literacy, we apply factor analysis. The suitability of the dataset for factor analysis is supported by the results of the Kaiser-Meyer-Olkin (KMO) test and Bartlett’s test of sphericity. The KMO statistic is 0.896, and the p-value for Bartlett’s test of sphericity is 0, significant at the 1% level, rejecting the null hypothesis. This indicates strong correlations among the variables, confirming that the dataset is appropriate for factor analysis. Table 2 reports the factor analysis of financial literacy. Following the criterion of retaining factors with eigenvalues greater than one, we extract two factors that together explain more than 75% of the total variance. Finally, we construct the financial literacy (FL) measure based on the factor scores and the weights of the rotated factors.

Table 1.

Financial Literacy Issues.

Serial number	Title	Financial literacy issues	Correct answer
1	One-year time deposit interest rate	What do you estimate the current interest rate for a 1-year fixed deposit at a bank to be?	1%–5%
2	Deposit maturity amount	Suppose you have a 1-year fixed deposit of 10,000 yuan with an annual interest rate of 3%. If you do not withdraw it early, how much will you have at maturity?	Equivalent to 10,300 yuan
3	Deposit renewal amount	After the deposit in the account from Question 2 matures, it is reinvested in a 1-year fixed deposit at the same interest rate. How much will be in the account after 1 year?	More than 10,600 yuan
4	Purchasing power of money	If the annual interest rate on your bank deposit is 3% and the inflation rate is 5% per year, how much can you buy with the money in this account after 1 year?	Less than now
5	Value comparison	Suppose Zhang San inherits 100,000 yuan today, while Li Si will inherit 100,000 yuan 3 years from now. Whose inheritance is worth more?	Zhang San’s inheritance has a higher value
6	Investment risks	In general, high-return investments come with high risk.	Correct
7	Stock investment risks	Generally, investing in a single stock carries less risk than investing in an equity fund.	Error
8	Decision bank	Which of the following banks has the function of formulating and implementing monetary policy?	People’s Bank of China
9	Investment product risk	In general, which of the following assets carries the highest risk?	Stock
10	The meaning of buying stocks	What does this mean if you purchase a company’s stock?	You are a shareholder of the company
11	Fund description	Which of the following descriptions of a fund is correct?	Generally, the same fund can invest in several assets, such as stocks and bonds.
12	Description of bank wealth management	Which of the following descriptions of bank wealth management products is correct?	Like other risky investments, bank wealth management products also carry the possibility of losing money.
13	Stock market function description	Which of the following statements do you think correctly describes the core function of the stock market?	The stock market facilitates the matching of buyers and sellers of stocks.

Table 2.

Financial Literacy Factor Analysis.

Factor	Eigenvalue	Cumulative variance contribution rate (%)
Factor 1	3.72425	69.583
Factor 2	1.15472	75.103
Factor 3	0.89937	78.264
Factor 4	0.85532	81.299
Factor 5	0.82602	84.200
Factor 6	0.80395	87.021
Factor 7	0.77726	89.753
Factor 8	0.71878	91.924
Factor 9	0.70114	93.974
Factor 10	0.67202	95.784
Factor 11	0.65577	97.475
Factor 12	0.63801	98.974
Factor 13	0.57338	100.000

Control Variable

Following Chen and Li (2023), we select relevant control variables that may influence household income inequality from two perspectives: household head characteristics and household characteristics. The household head characteristics include age, gender, marital status, and health status. To account for the potential non-linear impact of age on household income, we also include the square of age. The household characteristics consist of the average years of education within the household, the number of family members employed in government positions, the value of the family house, and the family size.

Mediating Variable

Entrepreneurship: Following Y. Tan and Li (2022), we use the CFPS survey question, “Has your household engaged in individual private activities in the past year?” We code a response indicating participation as 1, while we code non-participation as 0.

Financial Participation: We determine financial participation based on the CFPS survey question, “Does your household own stocks, bonds, funds, or other financial products in the past year?” If the household holds at least one financial product, we code it as 1; otherwise, we code it as 0.

Insurance Participation: We analyze insurance participation using the CFPS survey question, “What was your household’s expenditure on commercial insurance (such as commercial medical insurance, auto insurance, property insurance, or life insurance) in the past year?” For analysis, we apply a natural logarithmic transformation to this expenditure.

Moderating Variable

This study uses the digital economy as the moderating variable. We adopt the approach of Zhao et al. (2020) and measure the digital economy from five aspects at the provincial level: internet penetration rate, the number of internet-related workers, internet-related output, the number of mobile internet users, and digital inclusive finance. We detail additional specific indicators in Table 3. We calculate the digital economy index using the entropy method, denoting it as Digital.

Table 3.

Construction of the Digital Economy Index.

Level 1 indicators	Level 2 indicators	Level 3 indicators	Indicator properties
Digital economy	Internet penetration	Internet users per 100 population	+
	Internet-related practitioners	Percentage of employees in computer services and software	+
	Internet-related outputs	Total telecommunication services per capita	+
	Number of mobile Internet users	Number of mobile phone users per 100 population	+
	Development of digital Inclusive finance	China digital inclusive finance index	+

Definition of Variables and Descriptive Statistical Analysis

We present the definitions and descriptive statistics of the variables in Table 4. The results indicate that the mean of household income inequality (Index) is 0.275, with a standard deviation of 0.273, suggesting a relatively high level of income inequality among households in China and significant variability. In contrast, the mean financial literacy (FL) is 0.061, with a standard deviation of 0.698, reflecting a low overall level of financial literacy among households and notable differences in financial literacy across families. These findings suggest that although the average level of income inequality is high, there is considerable disparity in financial literacy and capabilities among households, which may influence their economic decisions and wealth management abilities. Consequently, these statistical results provide crucial context for exploring the impact of household financial literacy on income inequality and lay a foundation for subsequent analyses.

Table 4.

Definition of Variables and Descriptive Statistical Analysis (N = 3,032).

Variable symbol	Definition	Mean	SD
Dependent variable
Index	The Kakwani index of income inequality calculated in this paper	0.275	0.273
Independent variable
FL	The financial literacy index calculated in this paper	0.061	0.698
Control variable
age	Age of the household head	48.755	14.54
age2	Square of the age of household head/100	25.884	14.802
Gender	Gender of household head: male = 1; female = 0	0.443	0.497
Marry	Marital status of household head: married = 1; other = 0	0.883	0.373
Healthy	Health status of household head: healthy = 1; unhealthy = 0	0.877	0.328
Edu	Mean years of education in the household	10.227	3.682
Government	Number of people in the family working in the government sector	0.492	0.741
Lnhouse	Natural logarithm of the value of the family house	10.721	4.909
Famsize	Family size	3.223	1.484
Mediating variable
Entre	Family entrepreneurship = 1; family not entrepreneurship = 0	0.130	0.336
Finance	Family financial participation = 1; family no financial participation = 0	0.155	0.362
Insurance	Natural logarithm of expenditure on family business insurance (Yuan)	2.138	1.125
Moderating variable
Digital	The natural logarithm of the digital economy index calculated in this paper	0.309	0.071

We categorize households into two groups based on the mean financial literacy level: low financial literacy and high financial literacy. Figure 3 presents the mean differences between the two groups for income inequality. Households with low financial literacy have an average income inequality of 0.326, while those with high financial literacy have a mean of 0.226, revealing a significant difference. This finding suggests that higher financial literacy contributes to reducing household income inequality.

Figure 3.

Mean difference between low and high financial literacy.

Methods

Baseline Regression Model

The double machine learning (DML) model is an advanced method for causal inference. Its core idea lies in disentangling the complex relationships among the treatment variable, outcome variable, and control variables to obtain unbiased estimates. The DML approach involves two stages. In the first stage, machine learning algorithms are used to model the treatment and outcome variables separately based on the control variables, producing corresponding prediction residuals. In the second stage, these residuals are used to construct a new model to estimate the treatment effect, thereby eliminating the influence of confounding factors. Compared to traditional econometric models, the DML model offers significant advantages in handling high-dimensional data, relaxing model assumptions, and improving estimation accuracy and flexibility. Following the approach proposed by Chernozhukov et al. (2017), we construct a partially linear regression model as follows:

$Inde x_{i} = θ_{0} F L_{i} + g_{0} (X_{i}) + U_{i}, E [U_{i} | X_{i}, F L_{i}] = 0$ (3)

$F L_{i} = f_{0} (X_{i}) + V_{i}, E [V_{i} | X_{i}] = 0$ (4)

Where $Inde x_{i}$ denotes the income inequality index, $F L_{i}$ represents financial literacy, and $X_{i}$ is a vector of control variables. $U_{i}$ and $V_{i}$ are error terms, assumed to have zero mean. $θ_{0}$ captures the estimated causal effect of financial literacy on income inequality. The functions $g_{0} (X_{i})$ and $f_{0} (X_{i})$ are unknown components estimated using machine learning algorithms. In the baseline regression, we employ the random forest algorithm to estimate both functions. First, regarding the parameters of the base learner, we set the number of trees (n_estimators) to 500 to ensure model stability and convergence. For the maximum depth of the trees (max_depth), we did not impose strict restrictions; instead, we adopted a data-driven approach that allows the model to capture nonlinear relationships among variables more effectively. Second, other key hyperparameters, such as the minimum number of samples required to split a node (min_samples_split) and the minimum number of samples required at a leaf node (min_samples_leaf), were set according to commonly used empirical values. Finally, for the cross-validation strategy, this study employed five-fold cross-validation.

Mediation Effect Model

Building on the DML framework, we construct the following model to examine the underlying mechanisms:

$Mediato r_{i} = θ_{0} F L_{i} + g_{0} (X_{i}) + U_{i}, E [U_{i} | X_{i}, F L_{i}] = 0$ (5)

$F L_{i} = f_{0} (X_{i}) + V_{i}, E [V_{i} | X_{i}] = 0$ (6)

$Inde x_{i} = θ_{0} Mediato r_{i} + g_{0} (X_{i}) + U_{i}, E [U_{i} | X_{i}, Mediato r_{i}] = 0$ (7)

$Mediato r_{i} = f_{0} (X_{i}) + V_{i}, E [V_{i} | X_{i}] = 0$ (8)

Where $Mediato r_{i}$ denotes the mechanism variables: entrepreneurship, financial participation, and insurance participation. The meanings of the other variables remain consistent with those in Equations 3 and 4.

Moderation Effect Model

Since the double machine learning approach cannot directly assess moderation effects, we employ a traditional ordinary least squares (OLS) model to conduct the analysis. We establish the following moderation effect model based on the testing and analytical procedures for moderation effects found in existing research (Wan et al., 2024).

$\begin{matrix} Inde x_{i} = ϕ_{0} + ϕ_{1} F L_{i} + ϕ_{2} F L_{i} \times Digita l_{p} + ϕ_{3} Digita l_{p} \\ + ϕ_{4} X_{i} + λ_{p} + ε_{i} \end{matrix}$ (9)

where $F L_{i} \times Digita l_{p}$ represents the interaction term between household’s financial literacy and the digital economy. $Digita l_{p}$ represents the level of the digital economy at the provincial level, while the other variables retain the same definitions as in Equation 3.

Results

Baseline Regression Results

Before conducting the empirical regression analysis, a correlation test is carried out among the variables to identify and address potential multicollinearity issues. We found a significant negative relationship between financial literacy and income inequality (Figure 4). The correlation coefficients between any two variables are low, indicating no presence of multicollinearity.

Figure 4.

Correlation between the main variables used in this study.

In this study, we employ a random forest algorithm and apply five-fold cross-validation to conduct the empirical analysis. We follow standard model tuning procedures for selecting and adjusting key parameters to ensure the reliability and robustness of the estimation results. Table 5 reports the baseline regression results on the impact of financial literacy on income inequality. Column (1) includes only the household head’s control variables. The coefficient on financial literacy (FL) is −0.066 and statistically significant at the 1% level, indicating that financial literacy significantly mitigates income inequality. Columns (2) and (3) progressively add province fixed effects and household-level control variables. The coefficient on FL remains significantly negative throughout, confirming the robustness of the result. Column (4) further controls for provincial fixed effects, yielding a coefficient of −0.037 for FL, significant at the 1% level. This finding suggests that a one-unit increase in a household’s financial literacy leads to a 3.7 percentage point reduction in income inequality. The baseline regression results indicate that enhancing financial literacy reduces household income inequality, supporting Hypothesis 1.

Table 5.

Baseline Regression Results.

Variable	(1)	(2)	(3)	(4)
Variable	Index	Index	Index	Index
FL	−0.066*** (0.007)	−0.069*** (0.008)	−0.033*** (0.008)	−0.037*** (0.008)
Constant	0.001 (0.005)	−0.003 (0.005)	−0.012** (0.005)	−0.006 (0.005)
Household head level	Yes	Yes	Yes	Yes
Household level	No	No	Yes	Yes
Province fixed effects	No	Yes	No	Yes
N	3,032	3,032	3,032	3,032

Note. Robust standard errors in parentheses.

p < .05. ***p < .01.

Mediating Effect Analysis

Table 6 presents the results of the mediation mechanisms through which financial literacy reduces income inequality. In columns (1) and (2), the impact of financial literacy on household entrepreneurship is positive, with the coefficient of FL being 0.027, statistically significant at the 1% level. Meanwhile, the effect of household entrepreneurship on income inequality is negative, with the coefficient of Entre at −0.027 (p < .1). This indicates that enhanced financial literacy promotes household entrepreneurship, which somewhat reduces income inequality, suggesting that entrepreneurship may be a potential channel for alleviating income inequality. In columns (3) and (4), financial literacy has a significant positive effect on household financial participation (FL = 0.126, p < .01), while financial participation significantly reduces income inequality (Finance = −0.032, p < .01). This implies that greater financial literacy increases household financial participation, which in turn helps narrow income inequality. In columns (5) and (6) show that financial literacy has a significant positive impact on household insurance participation (FL = 0.037, p < .1), and insurance participation significantly decreases income inequality (Insurance = −0.014, p < .05). This suggests that enhanced financial literacy significantly promotes insurance participation, which in turn reduces income inequality. Thus, Hypothesis 2 of this study is supported.

Table 6.

Regression Results for Mediation and Moderation Effects.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Variable	Entre	Index	Finance	Index	Insurance	Index	Index
FL	0.027*** (0.010)		0.126*** (0.011)		0.037* (0.023)		−0.055* (0.030)
Entre		−0.027* (0.014)
Finance				−0.032*** (0.012)
Insurance						−0.014** (0.007)
FL * Digital							−0.295** (0.118)
Digital							0.034 (0.091)
Constant	−0.008 (0.007)	−0.006 (0.005)	−0.012* (0.006)	−0.006 (0.005)	0.007 (0.013)	−0.006 (0.005)	0.229*** (0.071)
Household head level	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Household level	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	3,032	3,032	3,032	3,032	3,032	3,032	3,032

Note. Robust standard errors in parentheses.

p < .1. **p < .05. ***p < .01.

Moderating Effect Analysis

According to the moderation effect results in Column (7) of Table 6, financial literacy significantly negatively impacts income inequality, with an estimated coefficient of −0.055, significant at the 10% level. This indicates that improvements in financial literacy help reduce income inequality. The estimated coefficient of the interaction term FL * Digital is −0.295, significant at the 5% level, suggesting that the digital economy significantly amplifies the negative effect of financial literacy on income inequality. In other words, the digital economy plays a crucial moderating role in this relationship. When the digital economy is more developed, the impact of financial literacy in alleviating income inequality becomes even more pronounced. These findings suggest that the development of the digital economy not only affects income distribution through its channels but also further reduces income inequality by enhancing the effect of financial literacy. This highlights the importance of the synergistic interaction between financial literacy and the digital economy. Therefore, supporting Hypothesis 3 of this study.

Heterogeneity Analysis

Regional Differences

According to the regional grouping regression results in columns (1) and (2) of Table 7, In the eastern region, the estimated coefficient for FL is −0.030, which is significant at the 5% level, indicating that improving financial literacy helps alleviate income inequality. By contrast, in the central and western regions, the estimated coefficient of FL is −0.034 but not statistically significant. This suggests that the role of financial literacy in reducing income inequality is more pronounced in the eastern region. This primarily stems from the more developed financial market system and information infrastructure in eastern areas. The high level of financial service provision and information accessibility amplifies the marginal effect of financial literacy, enabling individuals to allocate assets and manage risks more effectively. Additionally, the employment structure in the eastern region is more concentrated in high-value-added industries, where improved financial literacy enables individuals to seize opportunities, thereby contributing to a more equitable income distribution.

Table 7.

Heterogeneity Analysis: Region and Urban-Rural.

	(1)	(2)	(3)	(4)
	East	Central-West	Urban	Rural
Variable	Index	Index	Index	Index
FL	−0.030** (0.012)	−0.034 (0.011)	−0.041*** (0.008)	−0.010 (0.027)
Constant	−0.010 (0.007)	−0.007 (0.007)	−0.003 (0.005)	−0.011 (0.017)
Household head level	Yes	Yes	Yes	Yes
Household level	Yes	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes	Yes
N	1,368	1,664	2,798	234

Note. Robust standard errors in parentheses.

p < .1. **p < .05. ***p < .01.

Urban-Rural Differences

According to the urban-rural grouping regression results in columns (3) and (4) of Table 7. In urban areas, the estimated coefficient for FL is −0.041, which is significant at the 1% level. In rural areas, although the estimated coefficient of financial literacy is −0.010, it is not statistically significant. This suggests that financial literacy plays a more substantial role in alleviating income inequality in urban regions. One possible explanation is that urban areas have more developed financial market systems and a greater variety of financial products, which provide broader opportunities for financial literacy to have an impact. Individuals with higher financial literacy are better positioned to access wealth management, credit, and investment opportunities in urban areas, enabling them to diversify income sources and grow their assets, thereby narrowing income disparities. Moreover, urban residents generally have higher levels of education, which strengthens their ability to translate financial knowledge into practical behavior, enhancing both their income-generating capacity and risk-avoidance skills.

Differences Between Education

We categorize households based on education level following X. Ma (2024), dividing them into two groups: high education level (more than 9 years of education, High-edu) and low education level (9 years of education or less, Low-edu). According to the educational level grouping regression results in columns (1) and (2) of Table 8, the impact of financial literacy on household income inequality significantly differs between high and low-education level households. In high-education level households, the estimated coefficient for FL is −0.040, which is significant at the 1% level. In contrast, for households with low education levels, the estimated coefficient for FL is −0.018, it is not statistically significant. This indicating that financial literacy has a stronger effect on reducing income inequality in high-education level households. This may be because highly educated households possess a stronger ability to understand and apply financial knowledge. Higher education enhances individuals’ capacity to identify and process financial information efficiently, enabling them to participate more effectively in asset allocation, risk management, and wealth accumulation. As a result, they benefit more from income distribution mechanisms and experience a reduction in intra-group income inequality.

Table 8.

Heterogeneity Analysis: Education and Age.

	(1)	(2)	(3)	(4)
	High-edu	Low-edu	Middle-Young	Old
Variable	Index	Index	Index	Index
FL	−0.040*** (0.010)	−0.018 (0.014)	−0.036*** (0.009)	−0.048*** (0.016)
Constant	−0.003 (0.006)	−0.004 (0.008)	−0.005 (0.005)	−0.004 (0.010)
Household head level	Yes	Yes	Yes	Yes
Household level	Yes	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes	Yes
N	1,832	1,200	2,290	742

Note. Robust standard errors in parentheses.

p < .1. **p < .05. ***p < .01.

Differences Between Ages

We categorize household heads’ ages into two groups based on Luan et al. (2023): middle-young (under 60 years) and old (60 years and older). According to the age grouping regression results in columns (3) and (4) of Table 8, the estimated coefficient for FL in middle-young households is −0.036. In contrast, the estimated coefficient in old households is −0.048, indicating a more significant effect of financial literacy in alleviating income inequality for older households. This distinction may arise because older households typically rely on fixed income sources, such as pensions or savings, which limit opportunities for income growth. Improved financial literacy helps them manage finances and invest more effectively, optimizing resource allocation and reducing economic inequality. In contrast, middle-young households often participate more actively in the labor market and have diversified income sources, leading to a weaker marginal effect of financial literacy on income inequality.

Robustness Test

In examining the impact of financial literacy on household income inequality, potential endogeneity issues may arise, including reverse causality and omitted variable bias. While financial literacy may influence income inequality, income inequality itself could also affect a household’s ability to acquire financial literacy. Additionally, unobserved factors such as individual intelligence and risk tolerance could simultaneously influence financial literacy and income inequality, leading to biased model estimates. To mitigate the influence of these factors and ensure the reliability of our conclusions, it is necessary to conduct a series of robustness checks to strengthen the validity of our findings.

Instrumental Variable Methods

We attempt to select appropriate instrumental variables to address potential endogeneity issues and identify the net effect of financial literacy on income inequality. We follow the approach of Rozelle et al. (1999) and use the average financial literacy of other households in the same community, excluding the target household, as the instrumental variable (FL_village). This approach satisfies the relevance requirement, as the financial literacy environment in the community likely influences individual households’ financial literacy. Although the average financial literacy level of other residents in the same village can influence a household’s own financial literacy, its impact on income inequality operates primarily through this indirect channel. There is no direct causal relationship between neighbors’ financial literacy and a household’s income level or position in the income distribution. Therefore, the instrument satisfies the condition of exogeneity.

Column (1) of Table 9 reports the regression results using the DML approach. The estimated coefficient of financial literacy is −0.007 and is statistically significant at the 10% level. This suggests that the mitigating effect of financial literacy on income inequality remains robust even after addressing endogeneity concerns. We employ a two-stage least squares (2SLS) regression to mitigate potential endogeneity, following Yi et al. (2023). Column (2) of Table 9 shows that the estimated coefficient for FL_village is positive at the 1% significance level, indicating a significant positive correlation between the instrumental variable and financial literacy, thus meeting the relevance condition for an instrumental variable. Column (3) presents the second-stage results. After incorporating the instrumental variable, the estimated coefficient for FL remains significantly negative, further supporting the conclusion that financial literacy mitigates household income inequality, reaffirming Hypothesis 1. Additionally, the minimum eigenvalue of the instrumental variable is 967.555, which is well above the critical value of 16.38, eliminating concerns about the weak instrumental variable.

Table 9.

Robustness Test: Instrumental Variable Method.

	(1)	(2)	(3)	(4)	(5)
	DML	2SLS		ERM
Variable	Index	FL	Index	FL	Index
FL	−0.007* (0.018)		−0.030* (0.017)		−0.035*** (0.013)
FL_village		0.770*** (0.021)		1.000*** (0.024)
Constant	−0.006 (0.005)	−0.368*** (0.113)	0.079 (0.057)	0.000 (0.010)	0.072 (0.061)
Household head level	Yes	Yes	Yes	Yes	Yes
Household level	Yes	Yes	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes	Yes	Yes
N	3,032	3,032	3,032	3,032	3,032

Note. Robust standard errors in parentheses.

p < .1. ***p < .01.

In columns (4) and (5) of Table 9, we re-examine the effect of financial literacy on household income inequality using the Extended Regression Model (ERM). Jiang and Wang (2020) highlight that ERM is suitable for addressing various treatment issues, including omitted variables, sample selection, and bidirectional causality. The estimated coefficient for FL_village in Column (4) remains significantly positive, confirming the relevance assumption. Meanwhile, the estimated coefficient for FL in Column (5) is still significantly negative, further validating the positive role of financial literacy in mitigating household income inequality.

Replacing the Dependent Variable and the Empirical Model

We recalculated the Kakwani index of household income inequality (Property_Index) based on household property income. In Column (1) of Table 10, the estimated coefficient for FL is negative at the 1% significance level, indicating that financial literacy significantly reduces household income inequality. Since the Kakwani index measures truncated inequality data bounded between [0, 1], we employed a Tobit model to empirically analyze income inequality based on total household and property income. The results in columns (2) and (3) of Table 10 show that the estimated FL coefficients are negative at the 1% significance level, confirming that financial literacy effectively reduces household income inequality.

Table 10.

Robustness Test: Replacing the Dependent Variable and the Empirical Model.

	(1)	(2)	(3)
	DML	Tobit	Tobit
Variable	Property_Index	Index	Property_Index
FL	−0.031*** (0.008)	−0.049*** (0.010)	−0.040*** (0.010)
Constant	−0.003 (0.005)	−0.212*** (0.080)	−0.376*** (0.080)
Household head level	Yes	Yes	Yes
Household level	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes
R ²		0.175	0.176
N	3,032	3,032	3,032

Note. Robust standard errors in parentheses.

***

p < .01.

Deletion of the Special City Sample

Given the significant economic and social differences between China’s four directly governed municipalities (Beijing, Shanghai, Tianjin, and Chongqing) and other cities, we excluded samples from these four cities and re-tested the model. The regression results in Table 11 show that all estimated coefficients for FL remain significantly negative, indicating that financial literacy mitigates household income inequality. These findings confirm the robustness of our study’s results.

Table 11.

Robustness Test: Deletion of the Special City Sample.

Variable	(1)	(2)
Variable	Index	Property_Index
FL	−0.033*** (0.009)	−0.034*** (0.009)
Constant	−0.006 (0.005)	−0.007 (0.005)
Household head level	Yes	Yes
Household level	Yes	Yes
Province fixed effects	Yes	Yes
N	2,464	2,464

Note. Robust standard errors in parentheses.

***

p < .01.

Replace the Double Machine Learning Algorithm

This study employs the random forest algorithm for the DML method. We conduct a robustness check to address potential errors arising from algorithm selection by replacing the random forest algorithm with alternative DML algorithms, including Lasso regression, gradient boosting, and neural networks. As shown in Table 12, the mitigating effect of financial literacy on income inequality remains statistically significant across all alternative models, further confirming the robustness of the baseline regression results.

Table 12.

Robustness Test: Replace the Double Machine Learning Algorithm.

Variable	(1)	(2)	(3)
	Lasso regression	Gradient boosting	Neural networks
	Index	Index	Index
FL	−0.045*** (0.008)	−0.043*** (0.008)	−0.035*** (0.008)
Constant	−0.000 (0.005)	−0.001 (0.005)	−0.003 (0.005)
Household head level	Yes	Yes	Yes
Household level	Yes	Yes	Yes
Province fixed effects	Yes	Yes	Yes
N	3,032	3,032	3,032

Note. Robust standard errors in parentheses.

***

p < .01.

Discussion and Conclusion

Discussion

Addressing income inequality is critical due to its profound implications for social stability, economic growth, and sustainable development. Income inequality exacerbates social divisions, worsens the living conditions of low-income groups, and ultimately hinders overall economic dynamism (Lee et al., 2020; McFarland et al., 2023). Tang et al. (2022) highlight that low-income groups often rely on a single source of income, primarily wages, and face unstable employment conditions. These factors directly widen the income gap between low- and middle- to high-income groups. In recent years, academic research on income inequality has expanded significantly, focusing on policy interventions, tax reforms, and equitable access to education (L. K. Chu & Hoang, 2020; Taghizadeh-Hesary et al., 2020). However, beyond policy measures, promoting financial literacy is equally essential. Financial literacy equips individuals with the skills to manage personal finances effectively and engage in financial markets through investment, enhancing their income potential (Twumasi et al., 2022). Studies reveal that financially literate individuals excel in asset accumulation and risk management, enabling them to maintain financial stability during economic fluctuations (Banks et al., 2020; Lusardi & Mitchell, 2017). By improving financial literacy across society, individuals can achieve more significant wealth accumulation, reducing income inequality (G. Xu et al., 2024).

Increasing residents’ income channels is the primary way to narrow the income gap between different groups. Building on this premise, our study specifically examines the role of financial literacy in addressing income inequality, exploring the mechanisms through which the two are connected. We find that financial literacy, due to its ability to optimize resource allocation (Twumasi et al., 2022) and improve risk management (Banks et al., 2020), primarily impacts income inequality by encouraging households to engage in entrepreneurship, participate in financial markets, and utilize insurance. These findings validate and extend previous research by Halvarsson et al. (2018), X. Zhang (2024), and Pu et al. (2022). Halvarsson et al. (2018) emphasize the pivotal role of entrepreneurship in reducing income inequality. X. Zhang (2024) demonstrates that financial market participation effectively broadens income sources, mitigating income inequality. Similarly, Pu et al. (2022) highlight the significant value of insurance participation in narrowing income gaps. Furthermore, our study investigates the moderating effect of the digital economy on the relationship between financial literacy and income inequality. The rise of the digital economy amplifies the impact of financial literacy by significantly lowering the costs of accessing financial knowledge and services (Y. Tan & Li, 2022). This increased accessibility enables more low-income groups to benefit from financial innovations, further enhancing the role of financial literacy in reducing income inequality. These findings offer valuable insights into the processes and mechanisms through which financial literacy alleviates income inequality. They also give policymakers a fresh perspective for designing more targeted financial education programs and inclusive financial policies.

Limitations and Future Research Directions

This study provides valuable insights into the relationship between financial literacy and income inequality. However, several limitations should be noted. First, the analysis focuses exclusively on the Chinese context. Given the substantial differences across countries, future research could adopt a comparative approach and expand the scope to a global level, thereby enhancing the generalizability of the findings. Second, the mediation effect is analyzed using a two-step method. However, the DML model and traditional regression analysis may differ in the identification of factors. Future studies may therefore consider applying the DML approach to mediation analysis to strengthen the robustness of causal inference. Third, this study relies solely on cross-sectional data from 2014, since this is the only year within the 2010 to 2022 CFPS micro-survey that includes financial literacy questions. This data constraint may introduce measurement bias. Future research could utilize more recent panel data to enhance both the accuracy and robustness of the results.

Conclusion

In today’s society, income inequality has emerged as one of the most significant challenges to global economic development. It affects social equity and justice and constrains sustainable economic growth. Therefore, identifying effective ways to alleviate income inequality and promote equitable wealth distribution is crucial. This study utilizes data from the 2014 CFPS and applies a DML model to systematically examine the effect of financial literacy on mitigating income inequality. The results demonstrate that financial literacy significantly reduces income inequality, a conclusion that remains robust after a series of robustness tests confirming the reliability of the findings. Mechanism analysis shows that financial literacy alleviates income inequality primarily through three channels: promoting entrepreneurship, increasing financial participation, and enhancing insurance participation. Moreover, the rise of the digital economy plays a critical moderating role, digital economy enhances the effect of financial literacy on reducing income inequality. The heterogeneity analysis further reveals that the mitigating effect of financial literacy is particularly pronounced in eastern regions, urban areas, highly educated households, and elderly households. These findings offer new perspectives on the causes of income inequality and highlight how financial literacy can serve as a tool to address this issue.

Based on the research conclusions in this paper, we propose the following policy recommendations:

First, it is essential to enhance residents’ financial literacy. Since financial literacy is significantly negatively correlated with income inequality, expanding financial knowledge and skills can effectively mitigate income disparities. The government can integrate financial literacy courses into the education system, encourage schools and communities at all levels to provide universal financial education, and design targeted, tiered training programs for different groups to improve their financial decision-making capacity.

Second, financial literacy should be encouraged to play a role through channels such as entrepreneurship, financial participation, and insurance participation. Specifically, the government can improve the support system for small and micro enterprises, expand financing channels, and lower barriers to entrepreneurship. It can also promote inclusive finance, enhancing access to financial services for rural and low-income groups. At the same time, strengthening regulation and public awareness of the insurance market can help households use insurance products more effectively to diversify risks, thereby reducing income inequality.

Third, the amplifying effect of the digital economy should be leveraged. Research shows that the digital economy can enhance the role of financial literacy in reducing income inequality. Therefore, efforts should focus on accelerating digital infrastructure development and promoting “Internet + financial education” platforms to improve residents’ online learning and financial participation. At the same time, support for digital inclusive finance in rural and underdeveloped regions should be strengthened to bridge the digital divide.

Ultimately, policy design should take into account the diverse needs of various regions and populations. Research indicates that financial literacy has a stronger effect in reducing income inequality among residents in eastern regions, urban areas, highly educated individuals, and the elderly. Therefore, greater investment in financial education and digital inclusive finance is needed in central and western regions as well as rural areas. Additionally, accessible and easy-to-understand financial literacy programs should target low-education groups. Creating more digital entrepreneurship and employment opportunities for younger populations promotes a more balanced income distribution.

Footnotes

ORCID iD

Wang Li

Funding

The author received no financial support for the research,authorship,and/or publication of this article.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Data Availability Statement

Data will be made available on request.

References

Atkinson

Messy

F. A.

(2012). Measuring financial literacy: Results of the OECD/International Network on Financial Education (INFE) pilot study (OECD Working Papers on Finance, Insurance and Private Pensions, No. 15). OECD Publishing.

Banks

Bassoli

Mammi

(2020). Changing attitudes to risk at older ages: The role of health and other life events. Journal of Economic Psychology, 79, Article 102208.

Barasa

Kazungu

Nguhiu

Ravishankar

(2021). Examining the level and inequality in health insurance coverage in 36 sub-Saharan African countries. BMJ Global Health, 6(4), e004712.

Bayar

Sezgin

H. F.

Ozturk

O. F.

Sasmaz

M. U.

(2020). Financial literacy and financial risk tolerance of individual investors: Multinomial logistic regression approach. Sage Open, 10(3), 1–11.

Behrman

J. R.

Mitchell

O. S.

Soo

C. K.

Bravo

(2012). How financial literacy affects household wealth accumulation. American Economic Review, 102(3), 300–304.

Berry

Karlan

Pradhan

(2018). The impact of financial education for youth in Ghana. World Development, 102, 71–89.

Chang

Chen

Tai

(2022). The impact of financial literacy on rural household self-employment: The mediating role of financial ability. Emerging Markets Finance and Trade, 58(11), 3297–3308.

Chen

(2023). Middle Class vulnerability in China: Measurement and determinants. Sustainability, 15(8), 6717.

Chernozhukov

Chetverikov

Demirer

Duflo

Hansen

Newey

(2017). Double/debiased/neyman machine learning of treatment effects. American Economic Review, 107(5), 261–265.

10.

Chu

L. K.

Hoang

D. P.

(2020). How does economic complexity influence income inequality? New evidence from international data. Economic Analysis and Policy, 68, 44–57.

11.

Chu

Wang

Xiao

J. J.

Zhang

(2017). Financial literacy, portfolio choice and financial well-being. Social Indicators Research, 132, 799–820.

12.

De Pleijt

Van Zanden

J. L.

(2021). Two worlds of female labour: Gender wage inequality in western Europe, 1300-1800. The Economic History Review, 74(3), 611–638.

13.

Duan

Wang

Zheng

(2022). Financial literacy and commercial insurance participation behavior. Financial Engineering and Risk Management, 5(4), 113–127.

14.

Engbom

Moser

(2022). Earnings inequality and the minimum wage: Evidence from Brazil. American Economic Review, 112(12), 3803–3847.

15.

Firmansyah

Susetyo

D. P.

(2022). Financial behavior in the digital economy era: Financial literacy and digital literacy. Jurnal Ekonomi Dan Bisnis Digital, 1(4), 367–390.

16.

Guo

Wang

Yuan

(2022). Digital finance and the efficiency of household investment portfolios. Emerging Markets Finance and Trade, 58(10), 2895–2909.

17.

Guo

Chen

Wang

Cao

(2024). Financial literacy and the household participation in commercial insurance: Evidence from China. Research in International Business and Finance, 71, Article 102478.

18.

Halvarsson

Korpi

Wennberg

(2018). Entrepreneurship and income inequality. Journal of Economic Behavior & Organization, 145, 275–293.

19.

(2020). Can online social interaction improve the digital finance participation of rural households? China Agricultural Economic Review, 12(2), 295–313.

20.

Huijsmans

Rijken

A. J.

Gaidyte

(2022). The income gap in voting: Moderating effects of income inequality and clientelism. Political Behavior, 44(3), 1203–1223.

21.

Hutter

Weber

(2023). Labour market effects of wage inequality and skill-biased technical change. Applied Economics, 55(27), 3063–3084.

22.

Jappelli

Padula

(2013). Investment in financial literacy and saving decisions. Journal of Banking & Finance, 37(8), 2779–2792.

23.

Jiang

Wang

(2020). Is linking social capital more beneficial to the health promotion of the poor? Evidence from China. Social Indicators Research, 147(1), 45–71.

24.

Kavya

T. B.

Shijin

(2020). Economic development, financial development, and income inequality nexus. Borsa Istanbul Review, 20(1), 80–93.

25.

Kim

M. S.

Lee

S. J.

(2023). Income inequality, social mobility, and electoral participation in the US counties: Revisiting the inequality-participation nexus. Political Studies, 71(4), 1129–1150.

26.

Klapper

Lusardi

Panos

G. A.

(2013). Financial literacy and its consequences: Evidence from Russia during the financial crisis. Journal of Banking & Finance, 37(10), 3904–3923.

27.

Kunawotor

M. E.

Bokpin

G. A.

Barnor

(2020). Drivers of income inequality in Africa: Does institutional quality matter? African Development Review, 32(4), 718–729.

28.

Lee

C. C.

Lee

C. C.

Lien

(2020). Income inequality, globalization, and country risk: A cross-country analysis. Technological and Economic Development of Economy, 26(2), 379–404.

29.

Liang

Guo

(2015). Social interaction, Internet access and stock market participation—An empirical study in China. Journal of Comparative Economics, 43(4), 883–901.

30.

Liu

Luo

Zhang

Rozelle

Hagist

Hou

(2020). The returns to education in rural China: Some new estimates. Australian Journal of Agricultural and Resource Economics, 64(1), 189–208.

31.

Luan

Zou

Huang

(2023). Digital divide and household energy poverty in China. Energy Economics, 119, Article 106543.

32.

Lusardi

Michaud

P. C.

Mitchell

O. S.

(2017). Optimal financial knowledge and wealth inequality. Journal of Political Economy, 125(2), 431–477.

33.

Lusardi

Mitchell

O. S.

(2014). The economic importance of financial literacy: Theory and evidence. American Economic Journal: Journal of Economic Literature, 52(1), 5–44.

34.

Lusardi

Mitchell

O. S.

(2017). How ordinary consumers make complex economic decisions: Financial literacy and retirement readiness. Quarterly Journal of Finance, 7(3), Article 1750008.

35.

Song

Zong

(2021). Urban-rural inequality of opportunity in health care: Evidence from China. International Journal of Environmental Research and Public Health, 18(15), Article 7792.

36.

(2024). Internet use and income gaps between rural and urban residents in China. Journal of the Asia Pacific Economy, 29(2), 789–809.

37.

McFarland

M. J.

Hill

T. D.

Montez

J. K.

(2023). Income inequality and population health: Examining the role of social policy. Journal of Health and Social Behavior, 64(1), 2–20.

38.

Mdingi

S. Y.

(2021). Literature review on income inequality and economic growth. MethodsX, 8, Article 101402.

39.

Mohamad

N. M.

Masron

T. A.

Ibrahim

(2021). The role of entrepreneurship on income inequality in developing countries. Journal of Poverty, 25(6), 520–542.

40.

Murendo

Mutsonziwa

(2017). Financial literacy and savings decisions by adult financial consumers in Zimbabwe. International Journal of Consumer Studies, 41(1), 95–103.

41.

Nikensari

S. I.

Santosa

P. B.

Sugiyanto

F. X.

(2021). Still gap issues of education, skill and capital to income inequality. Iranian Economic Review, 25(2), 205–218.

42.

Prasad

Meghwal

Dayama

(2018). Digital financial literacy: A study of households of Udaipur. Journal of Business and Management, 5, 23–32.

43.

Wang

Zhang

Zeng

(2022). Can basic medical insurance reduce elderly family income inequality in China? Frontiers in Public Health, 10, Article 838733.

44.

Rozelle

Taylor

J. E.

DeBrauw

(1999). Migration, remittances, and agricultural productivity in China. American Economic Review, 89(2), 287–291.

45.

Sayinzoga

Bulte

E. H.

Lensink

(2015). Financial literacy and financial behaviour: Experimental evidence from rural Rwanda. The Economic Journal, 126(594), 1571–1599.

46.

Seelkopf

Lierse

(2020). Democracy and the global spread of progressive taxes. Global Social Policy, 20(2), 165–191.

47.

Setiawan

Effendi

Santoso

Dewi

V. I.

Sapulette

M. S.

(2022). Digital financial literacy, current behavior of saving and spending and its future foresight. Economics of Innovation and New Technology, 31(4), 320–338.

48.

Sudiantini

Rizky

P. P.

Hazarika

(2023). Digital economy and financial inclusion in reviving the national economy: A management strategy. Revenue Journal: Management and Entrepreneurship, 1(1), 64–75.

49.

Taghizadeh-Hesary

Yoshino

Shimizu

(2020). The impact of monetary and tax policy on income inequality in Japan. The World Economy, 43(10), 2600–2621.

50.

Tan

Cai

Han

Zhou

(2022). Understanding peasant household’s land transfer decision-making: A perspective of financial literacy. Land Use Policy, 119, Article 106189.

51.

Tan

(2022). The impact of internet on entrepreneurship. International Review of Economics & Finance, 77, 135–142.

52.

Tang

Gong

(2022). Narrowing urban-rural income gap in China: The role of the targeted poverty alleviation program. Economic Analysis and Policy, 75, 74–90.

53.

Tian

(2012). Calculation of the Gini coefficient in China and analysis of its changing trends, Journal of Humanity, 2, 56–61.

54.

Twumasi

M. A.

Jiang

Ding

Wang

Abgenyo

(2022). The mediating role of access to financial services in the effect of financial literacy on household income: The case of rural Ghana. Sage Open, 12(1), 1–11.

55.

Van Rooij

Lusardi

Alessie

(2011). Financial literacy and stock market participation. Journal of Financial Economics, 101(2), 449–472.

56.

Wan

Fan

Wang

(2024). Local government financing vehicle restructuring, implicit government guarantees and debt financing costs: Evidence from the Chinese municipal corporate bond market. Economic Systems, 48(2), Article 101179.

57.

(2023). Impacts of digital inclusive finance on household entrepreneurship. Finance Research Letters, 56, Article 104114.

58.

Feng

Wang

Liang

(2024). Digital financial literacy and rural income inequality. Sage Open, 14(3), 1–17.

59.

Song

Ankrah Twumasi

(2023). The relationship between financial literacy and income structure of rural farm households: Evidence from Jiangsu, China. Agriculture, 13(3), 711.

60.

Yang

Ali

S. T.

Cui

(2022). Does financial literacy affect household financial behavior? The role of limited attention. Frontiers in Psychology, 13, Article 906153.

61.

Han

Long

(2023). Does Internet use increase public perception of environmental pollution? Evidence from China. Social Indicators Research, 166(3), 665–685.

62.

Zhang

(2024). Digital technology access, labor market behavior, and income inequality in rural China. Heliyon, 10(14), e33528.

63.

Zhang

(2024). A study on the impact of financial literacy on the Urban-Rural Income Gap. International Journal of Global Economics and Management, 3(1), 242–248.

64.

Zhao

Zhang

Liang

(2020). Digital economy, entrepreneurship, and high-quality economic development: Empirical evidence from Urban China. Journal of Management World, 36(10), 65–76.