This PLOS ONE research article investigates how digital financial literacy (DFL) relates to household consumption in China, distinguishing two complementary outcomes: total consumption expenditure (CE) and consumption structure (the proportional shares of subsistence, developmental, and enjoyment spending). Using China Household Finance Survey (CHFS) data, the authors aim to clarify whether DFL is associated with higher overall spending and whether it systematically shifts the internal composition of household consumption.
The study responds to changing income dynamics—specifically slower wage growth and rising importance of property income—and situates DFL as a potential pathway that enables households to engage in digital credit, investment, and insurance, thereby affecting consumption behavior beyond labor income.
The analysis relies on CHFS microdata. The authors measure household DFL and link it to observed consumption outcomes, then test potential mechanisms (borrowing, investment, insurance) and heterogeneity across demographic groups. Data access details are reported: CHFS data are publicly available upon application from Southwestern University of Finance and Economics. The paper reports funding support from a National Social Science Fund of China project and declares no competing interests.
The conceptual framework identifies three primary channels through which digital financial literacy could be associated with household consumption:
Borrowing (liquidity constraints): Digital credit reduces transaction costs and information frictions. Higher DFL helps households evaluate credit terms, potentially relaxing liquidity constraints and increasing current consumption, particularly for basic needs.
Investment (property income): Higher DFL is associated with greater participation in risky asset markets and wealth management. Investment returns can generate property income that may be allocated to discretionary or enjoyment spending.
Insurance (precautionary savings): Improved risk perception and insurance literacy may lead households to substitute costly precautionary savings with appropriate insurance products, freeing resources for current consumption.
These channels motivate empirical tests of both total expenditure and the shares of different consumption categories.
The study finds a robust positive association between DFL and total consumption expenditure. This relationship persists across main specifications and is particularly strong for low-income households. The authors interpret this pattern as consistent with a liquidity-constraint mechanism: DFL facilitates access to digital financial products that raise current spending capacity.
When examining consumption structure—the proportions of spending classified as subsistence, developmental, and enjoyment—the evidence for a systematic shift is limited. Although results indicate an association between higher DFL and a lower subsistence proportion alongside a higher enjoyment proportion, the authors emphasize that the overall evidence for a comprehensive upgrade in consumption structure is not robust. Thus, while DFL relates to spending levels, it does not unambiguously reorganize the internal composition of household expenditure across all specifications.
Empirical mediation analysis identifies borrowing as the main pathway linking DFL to higher household consumption. Digital credit use and the easing of short-term liquidity constraints account for a substantial portion of the observed association with total expenditure. Investment and insurance channels are present but play smaller roles in explaining the DFL–consumption relationship. This pattern supports the interpretation that immediate liquidity effects dominate over property-income or precautionary-savings channels in the observed timeframe.
The DFL–CE association exhibits heterogeneity. The link is strongest for low-income households, suggesting that targeted DFL interventions could yield greater consumption responses among disadvantaged groups. Urban–rural differences are reported as only marginally significant, while regional and educational differences do not show significant variation in the estimated association. These findings provide initial guidance for prioritizing groups in policy design.
Based on the empirical results, the authors suggest that digital financial literacy programmes can act as effective demand-side tools to boost household consumption, especially for low-income populations. However, they caution that achieving structural upgrading of consumption likely requires complementary policy efforts in education, health, and cultural services. Two practical priorities are highlighted: facilitating responsible access to digital credit (to preserve consumer protection while easing liquidity constraints) and building trust in digital insurance products (to encourage substitution away from precautionary savings). The authors stress that these measures should accompany DFL training to realize sustainable and welfare-enhancing consumption increases.
The article includes a limitations and future directions section; specific details of limitations are reported in the original manuscript. Data used are from CHFS and are publicly available upon application. The authors call for additional research to probe long-term effects, causal identification strategies, and complementary policies required to induce substantive changes in consumption structure.
This study contributes to the literature by explicitly separating total consumption expenditure from consumption structure in evaluating the role of DFL. It documents a consistent positive association between higher DFL and greater household spending—driven mainly by borrowing—and finds limited evidence that DFL alone produces a robust reallocation of spending across subsistence, developmental, and enjoyment categories. The findings point to the value of targeted DFL programmes for demand stimulation among low-income households, while underscoring the need for broader policies to achieve structural consumption transformation.