1.  Introduction

China prioritizes building a "Digital China", aligning with policies to accelerate its cyber power status and digital transformation. This initiative drives modernization, enhances national competitiveness, and fosters coordinated regional implementation. Under policy guidance, China has strengthened digital infrastructure, balanced development with security, and advanced high-quality growth through digital openness. The digital economy remains robust, with core industries contributing about 10% of GDP. China hosts 62 "lighthouse factories" (40% of the global total) and has led the world in online retail sales for 11 consecutive years, underscoring Digital China’s transformative impact.

Under China’s 14th Five-Year Plan for the Digital Economy, systematic new infrastructure—spanning 5G networks, data centers, and industrial internet—has driven rapid digital transformation, with the digital economy exceeding 50 trillion yuan (41% of GDP) by 2023. This shift reshapes online consumption ecosystems, accelerated by platforms like Douyin and Kuaishou, reflecting a structural upgrade from "basic survival" to "quality-driven" consumption. Studies highlight factors influencing online consumption, including trust in livestream hosts, product information quality, and marketing strategies[1][2]. Regional disparities persist, linked to urbanization levels, proximity to commercial hubs[3], and county-level economic development[4]. Additionally, the digital divide—shaped by income, age, and education gaps in technology adoption[5][6]—impacts consumption patterns, with higher-income, educated groups showing stronger online preferences[7].

Digital finance, driven by "technology + data," has gained academic attention for enhancing financial service efficiency and economic benefits through innovative financing tools. It leverages advantages like rapid response, broad accessibility, and cost efficiency, reducing financing costs for manufacturing[8]. Empirical studies highlight its role in advancing manufacturing upgrades, with green innovation as a key mechanism[9]. Digital finance also supports rural income growth, agricultural development, and shared prosperity[10][11]. It further aids in cutting industrial carbon emissions and fostering green economic growth[12].

China’s push to "accelerate Digital China" aligns digital finance with national priorities, leveraging the digital economy to drive consumption upgrades and industrial transformation. By lowering credit barriers and expanding consumption access, digital finance supports inclusive growth for low- to middle-income youth, while technologies like financial cloud and AI enhance financial efficiency and energize consumer markets. As key drivers of both digital engagement and social consumption, youth’s integration of online consumption with digital finance not only reshapes consumption patterns but also bolsters domestic demand and economic structural optimization.

This paper examines how digital finance reshapes youth online consumption behavior through technological empowerment and institutional innovation, linking micro-level youth actions to macro policy goals like "integrating the digital and real economy." Existing studies predominantly focus on digital finance’s macroeconomic or industry-specific impacts, leaving youth micro-behavior underexplored. The marginal contributions lie in three areas: first, shifting from macro to micro perspectives by analyzing youth as key social actors influenced by digital finance, family dynamics, and policy environments; second, bridging fragmented research on digital finance’s technical attributes and consumer behavior through empirical mechanisms connecting inclusive finance indices with online consumption levels; third, framing youth consumption as both an economic driver and social stability indicator, evaluating digital finance’s dual effects (empowerment vs. risks) to inform policies from the 20th CPC National Congress on deepening digital-real economy integration and strengthening consumption’s foundational economic role.

2.  Theoretical analysis and research hypothesis

2.1.  The impact of digital finance on young people’s online consumption

Existing research highlights digital finance's dual role in boosting consumption and addressing inequality. Studies confirm its positive impact on consumption levels and income gap reduction[13], with household digital finance use mitigating consumption inequality through income redistribution, liquidity constraint alleviation, and enhanced payment convenience[14]. While concerns about financial stability risks persist[15], evidence shows digital inclusive finance strengthens household savings[16] and elevates both middle-class consumption upgrades and low-income welfare[17]. At the enterprise level, digital finance enhances corporate financing efficiency[18] and boosts tech firms' productivity through cost reduction and innovation conversion[19]. The integration of digital payment systems (e.g., WeChat, Alipay) has revolutionized transaction efficiency, while platforms like Douyin and Taobao expand consumption scenarios.

For youth—the "digital natives"—their 95.3% internet penetration rate (exceeding national averages) and financial literacy enable effective utilization of digital credit tools (e.g., Alipay's "Huabei") to overcome budget constraints. Their trust in digital platforms[20][21] and preference for tech-driven, personalized services align with digital finance's strengths, facilitating optimized consumption decisions through credit leverage and data-driven recommendations[22][23]. This synergy amplifies youth consumption efficiency while reflecting broader economic shifts toward digital-real economy integration.

2.2.  The impact of digital finance on young people’s online consumption is heterogeneous

This study highlights regional disparities in digital finance’s impact on youth online consumption, with stronger effects in eastern China due to three factors. Firstly, eastern regions benefit from advanced digital infrastructure—wider adoption of financial cloud technology and hybrid cloud architectures enhances payment and credit service efficiency[24], while higher internet and smart device penetration supports seamless transactions. Secondly, eastern youth exhibit superior digital literacy and consumption preferences: their education levels enable effective use of tools like digital wealth management and installment plans[17], while a culture favoring quality and personalization aligns with big data-driven product matching (e.g., smart home financing, new energy vehicle leasing). Thirdly, eastern economies feature deeper digital-industrial integration: synergistic "digital credit + scenario consumption" models lower barriers to high-value purchases, while robust financial ecosystems expand service inclusivity and competitiveness[25]. In contrast, central/western regions face digital divides and financial exclusion, limiting youth consumption potential despite access to tools[26][27].

3.  Research design

3.1.  Variable design

This paper selects the China Family Panel Studies (CFPS) from 2014 to 2018, the Peking University Digital Inclusive Finance Index, and relevant data from the National Bureau of Statistics for empirical research, and preprocesses the data: (1) Match the CFPS individual-level questionnaire data with variables such as the inclusive finance index according to provinces, and then combine them with time to form panel data; (2) Screen samples based on age, and only retain samples of young people aged between 18 and 35; (3) Process some abnormal samples. For samples that do not use mobile phones and computers, their online shopping amount is recorded as 0, and samples with missing key variables are deleted.

3.1.1.  Explained variable

This paper selects the annual online shopping consumption of respondents in the CFPS questionnaire as the explained variable. In general, the increase in the amount of online shopping consumption of residents is consistent with the improvement of online consumption levels, and online shopping consumption is a major part of online consumption. Therefore, this paper uses the amount of online shopping consumption to measure the online consumption level of the residents, and performs logarithmic processing on the numerical value.

3.1.2.  Explanatory variables

The core explanatory variables of this paper are the same as those of most previous literature. The digital inclusive finance index of the Digital Finance Research Center of Peking University is used as the core explanatory variable to measure the level of digital finance development in the corresponding province in a certain year. The values ​​are also logarithmized.

3.1.3.  Control variables

a. Gender

Gender differences may affect online consumption through consumption preferences and behavior patterns. Studies have shown that women are more inclined to buy non-essential items such as clothing and beauty products, while men may spend more on high-value goods (such as electronic products). In addition, women have a higher demand for shopping convenience and may rely more on online platforms.

b. Age

Although the sample is limited to young people aged 18-35, aging may still affect consumption behavior through income accumulation or changes in family responsibilities. Young groups (such as 18-25 years old) may be more familiar with digital technology and more active in online consumption; while individuals approaching 35 years old may adjust their consumption structure due to increased family expenses (such as childcare and mortgages), and the focus of consumption will shift to offline rigid demand.

c. Highest education level

Education level is closely related to digital skills and consumption concepts. Individuals with higher education are usually more receptive to new technologies and have higher income potential, which may promote online consumption.

d. Political status

Whether or not one is a member of the Communist Party of China may indirectly reflect differences in an individual's socioeconomic status or values. Party members usually have more stable careers and incomes, but their consumption behavior may be more conservative.

e. Marital status

Marital status may affect consumption decisions through family responsibilities and spending structure. Married individuals may pay more attention to household necessities (such as home appliances and maternal and child products), while unmarried people tend to focus on personal consumption (such as entertainment and clothing).

f. Health status

Individuals with poor health may rely more on online shopping due to mobility problems, but their spending power may be limited by medical expenses.

g. Income

Income level is the core determinant of spending power. Total work income (including bonuses and subsidies) reflects career stability, while total personal income covers a wider range of income sources (such as investment and part-time jobs). In theory, the higher the income, the stronger the online spending power.

h. Provincial GDP

The level of regional economic development affects online consumption through infrastructure and residents' purchasing power. Provinces with high GDP usually have a more complete e-commerce ecosystem, which may promote online consumption.

3.2.  Descriptive statistics

Based on the data of the China Family Tracking Survey from 2014 to 2018, the Digital Inclusive Finance Index and the data of the National Bureau of Statistics, this study conducted descriptive statistics on the selected samples of the youth group (18-35 years old). As shown in Table 2, the core variables contain 13,433 to 15,417 valid observations. The sample size differences of different variables reflect the situation of missing data. Among them, the mean of the explained variable "online consumption" (the value after taking the pair) is 4.45, and the standard deviation is 3.90, indicating that there are significant differences in the online consumption level between individuals. The maximum consumption amount of the sample is 15.79, but the consumption record of some samples is 0. The mean of the explanatory variable "digital finance" index is 5.43, and the standard deviation is less than 0, indicating that the overall level of digital financial development in various provinces is relatively close, but there is still a certain gap between different regions.

Further observing the control variables, the mean of gender is 0.50, indicating that the gender ratio of men and women in the sample is balanced. The mean age is 26.91 years old, and the standard deviation is 4.19, which is in line with the age distribution characteristics of the youth group. The mean of the highest education level is 3.88, corresponding to the level of "high school/technical secondary school/technical school/vocational high school", but the standard deviation is large, reflecting the significant differences in the educational background of the samples. The mean of marital status is 0.81, indicating that about 80% of the samples are married. Among the income variables, the logarithmic means of total work income and personal total income are 9.71 and 7.02 respectively, and the standard deviation is large, highlighting the income stratification phenomenon of the youth group. In addition, the logarithmic mean of provincial GDP is 10.12, indicating that the sample covers regions with large differences in economic development levels. Overall, the data characteristics are highly consistent with the target group and variable selection of the research design.

This study uses VIF to test the multicollinearity problem of relevant variables in the model. As shown in Table 3, the VIF values ​​of each variable are between 1.01 and 2.39, among which the VIF of the core explanatory variable digital finance is 2.39. The collinearity level of income level in the control variable is slightly higher than that of other variables, but still significantly lower than the empirical threshold of 10. The average VIF of all variables is 1.65, indicating that there is no serious multicollinearity problem in the model. In addition, the 1/VIF values ​​are all greater than 0.4, and the regression analysis results are less likely to be disturbed by collinearity, and the research conclusions are robust.

3.3.  Model design

In order to test the impact of digital finance level on online consumption, this paper constructs the following multivariate regression model based on the least squares method and time fixed effects, which is expressed as follows:

$$ \mathit{ln}\left(OnlineConsu{m}_{it}\right)={\alpha }_0+{\beta }_{1}DFinanc{e}_{jt}+\gamma X_{ijt}+{\lambda }_t+{ϵ}_{it}$$(1)

Among them,  $$ \mathit{ln}\left(OnlineConsu{m}_{it}\right)$$  is the logarithm of the amount of online shopping consumption of individual  $$ i$$  in period  $$ t$$ , and  $$ DFinanc{e}_{jt}$$  is the digital financial inclusion index of province  $$ j $$ where individual  $$ i$$  is located in year  $$ t$$ .  $$ X_{ijt}$$ represents the control variables, including gender, age, highest education, political status, marital status, health status, income level, etc. at the individual level; and GDP per capita, etc. at the provincial level.  $$ {\lambda }_t$$  is the time fixed effect, which controls time trends and macro policy shocks, etc., and  $$ {ϵ}_{it}$$  is the random disturbance term.

4.  Empirical analysis

4.1.  Regression test

The regression test results are shown in columns (1), (2), and (3) of Table 4, which show the baseline regression results of the impact of digital financial development on online consumption among young people. The model is divided into three columns: (1) without control variables and fixed effects; (2) with control variables but no fixed effects; (3) with both control variables and time fixed effects. The coefficient of the core explanatory variable "digital finance" is significantly positive in all models, indicating that the improvement of digital finance has a significant promoting effect on online consumption. Among the control variables, the coefficients of age (negative), highest education level (positive), and provincial GDP (negative) are all significant, which is in line with theoretical expectations. The results show that the degree of development of digital finance has a significant impact on the online consumption behavior of young people.

***p<0.01", "**p<0.05", "*p<0.10

4.2.  Heterogeneity analysis

This table is based on geographical differences. It divides regions into the East and the West according to their geographical location, economic development level, etc. The "East" refers to the coastal provinces in eastern China (such as Beijing, Shanghai, Guangdong, etc.), which have developed economies, complete digital financial infrastructure, and active resident consumption; the "West" covers the inland provinces in the central and western regions (such as Sichuan, Gansu, Xinjiang, etc.), whose economic development is relatively lagging and digital financial penetration is low. Test the heterogeneity of the impact of digital finance on online consumption at the regional level. The results show that the coefficient of digital finance in the eastern region (1) is significantly positive (7.848, p<0.01), while the coefficient in the western region (2) is not significant (-2.847), and the inter-group difference Prob>chi2=0.0252, the value is less than 0.1, and there is an inter-group difference. This shows that the promotion effect of digital finance has regional heterogeneity, which may be affected by relevant factors such as economic development level or infrastructure differences.

4.3.  Robustness test

This table verifies the robustness of the model by replacing the core explanatory variables, and replaces digital finance with Internet penetration rate for regression analysis. The results show that the coefficient of Internet penetration rate is significantly positive (2.132, p<0.01), which is consistent with the direction of digital finance in the baseline model, supporting the robustness of the core conclusion. The sign and significance of the control variables are basically consistent with the baseline model, further indicating that the influencing mechanism of online consumption is stable and strengthening the reliability of the empirical results.

The mobile payment penetration rate is further used as an alternative explanatory variable, and the year fixed effect is added for robustness test. The coefficient of mobile payment penetration rate is significantly positive (2.534, p<0.01), which once again verifies the role of digital finance-related variables in promoting online consumption. Among the control variables, the effects of key variables such as gender, age, and education are consistent with the baseline model, and the negative effect of provincial GDP is enhanced (-0.350, p<0.05), suggesting the moderating role of regional economic differences and confirming the universality of the relationship between digital finance and online consumption.

5.  Conclusion and policy recommendations

This study examines how digital finance shapes youth online consumption using data from the China Family Panel Studies (2014–2018), Peking University’s Digital Financial Inclusion Index, and national statistics. Findings reveal that digital finance significantly boosts youth consumption through enhanced payment convenience, credit accessibility, and expanded consumption scenarios, though regional disparities persist. Eastern China exhibits stronger effects due to advanced digital infrastructure and higher digital literacy, while central and western regions lag due to infrastructure gaps and economic structural imbalances. Risks such as overconsumption and debt accumulation among credit-constrained youth highlight the need for balanced policy interventions. Policy recommendations emphasize improving financial literacy—particularly in central and western regions—to foster rational consumption, strengthening oversight of digital finance platforms to curb fraud and excessive credit, bridging regional digital divides through infrastructure investment and tailored financial products, and integrating digital finance with real-economy sectors like e-commerce and logistics to expand consumption opportunities while stimulating regional economic growth.