Rural Residents’ Digital Payment: The Use and Its Impact on Credit Availability – Evidence Using Extended UTAUT2

Mechanisms for UTAUT2 Factors to Influence Rural Residents’ Digital Payment Use

The well-known UTAUT2 model proposes seven constructs about technology behavioral intention in the consumer context: performance expectancy, effort expectancy, social influence, facilitating conditions, hedonic motivation, price value, and habit (Venkatesh et al., 2012). The original model is shown in Figure 1. In this section, we estimate the mechanism through which the factors influence rural residents’ use intention and the use behavior under the UTAUT2 model and take rural residents’ individual characteristics into account simultaneously.

OPEN IN VIEWER

Figure 1.

Original UTAUT2 model (Venkatesh et al., 2012).

Performance expectancy is defined as the degree to which using technology will provide benefits to consumers in performing certain activities (Venkatesh et al., 2012). The benefit that rural residents can obtain from using digital payment is the convenience of payment (de Blanes Sebastián et al., 2023), and it may be more prominent than that for urban residents as the financial infrastructure in rural areas is usually poorer than that in urban areas (Kong & Loubere, 2021). Many studies under the UTAUT2 model verified that performance expectancy is one important construct to influence the intention to use the technology(Al-Adwan & Al-Debei, 2024; Shareef et al., 2024), including the field of digital payment (de Blanes Sebastián et al., 2023; Manrai et al., 2024; Z. Wu & Liu, 2023). Based on empirical findings from previous literature, hypothesize 1 is proposed.

Effort expectancy is the degree of ease associated with consumers’ use of technology (Venkatesh et al., 2012). It is empirically supported by Azman Ong et al. (2023), Shah and Bhatt (2023), and Das and Datta (2024) that the simpler the digital payment process, the more likely customers are to use digital payments. Most rural residents are not familiar with digital payment due to the lack of digital literacy in the context of the digital divide (de Clercq et al., 2023; Goncalves et al., 2018; Peng & Dan, 2023; Purbo, 2017; Shenglin et al., 2017). They need to learn to use digital payment and may pay more effort to some extent. When residents feel it is easy or they can acquire this knowledge with less effort, they may be more willing to use digital payment. Thus we propose Hypothesis 2.

Social influence is the extent to which consumers perceive those important others, such as friends, relatives, or administrative staff of the corresponding technology, believe that they should use a particular technology (Venkatesh et al., 2012). Previous literature confirms that the perception or opinion of important surrounding people does affect customer’s adoption of digital payments (Al-Okaily et al., 2024; de Blanes Sebastián et al., 2023). China’s rural areas are typically relation-based societies (Qiu et al., 2020; X. Wu & Yuan, 2023), and therefore the social influence may be more significant. When the social powers around rural residents believe they should use digital payment, they are more likely to be willing to use it. Thus we propose Hypothesis 3.

Facilitating conditions refer to consumers’ perceptions of the resources and support available to perform a behavior (Venkatesh et al., 2012). The use of digital payments requires various hardware resources, software and support facilities available to help users understand how to use or to seek assistance (de Blanes Sebastián et al., 2023). Gupta et al. (2022) and Al-Okaily et al. (2024) confirm that facilitation conditions have a positive effect on technology adoption. The better the facilities and related services for digital payments in rural areas, the more willing rural residents are to use them. At the same time, the perceived facilitating conditions will also influence the use behavior of digital payment directly (Venkatesh et al., 2012). Thus we propose Hypothesis 4 and Hypothesis 5.

The price value is the consumers’ cognitive tradeoff between the perceived benefits of the applications and the monetary cost of using them (Dodds et al., 1991; Venkatesh et al., 2012). It is empirically supported by Gupta et al. (2022) and Al-Okaily et al. (2024) that under a certain level of benefits, the lower the cost of using digital payments, the more willing customers are to adopt them. Compared with urban residents, rural residents’ income is lower (Y. Gao et al., 2014; Tao et al., 2024), and they are more concerned about the price or monetary costs of using digital payment. The greater the perceived benefits from using digital payment than the monetary costs paid, the higher the price value. That is, the higher the price value, the more willing rural residents are to use digital payment. Thus we propose Hypothesis 6.

The original constructs of UTAUT2 also include habit and hedonic motivation. However, a large amount of UTAUT2-based studies do not include habit, as habit is difficult to form in the early stages of technology adoption due to the limited duration of use (Das & Datta, 2024; Tamilmani et al., 2019). Since digital payment is still a new concept in China’s rural areas and is not established enough to form habits, this study also does not include the construct of habit. Hedonic motivation under UTAUT2 is suitable for many kinds of information technology, such as personal computers and mobile phones. However, the hedonic motivation of digital payment is not obvious. Therefore, this study does not include these two constructs.

According to Fishbein and Ajzen-(1977), behavior intention can predict voluntary behavior. Digital payment by rural residents is voluntary, not coercive behavior, and intention to use digital payment will therefore be a predictor of their use of digital payment, which is verified by Azman Ong et al. (2023) and Manrai et al. (2024). Thus we propose Hypothesis 7.

Mechanisms for Other Factors to Influence Rural Residents’ Digital Payment Use

UTAUT2 does not consider consumers’ innovativeness and perceived riskiness. According to Hirschman (1980), innovativeness is people’s desire to find something new and different. People’s daring to try new technologies is considered to be a manifestation of innovativeness (Slade et al., 2015). Education or knowledge level is treated as one factor influencing innovation (Eryigit, 2020; Sáenz et al., 2023). The higher people’s level of knowledge, the better they can understand the benefits new products bring, master the use of innovative products, and manage the risks that new products may bring. Another important influencing factor is income (Eryigit, 2020; Thakur & Jasrai, 2018). High-income groups are better equipped with the financial resources and opportunities to experiment with new products or behaviors. Besides, they tend to experience higher levels of life satisfaction, which fosters a sense of youthfulness and, consequently, increases their likelihood of engaging in novelty-seeking behaviors (Eryigit, 2020). Z. Wu and Liu (2023) have proved that innovativeness is a factor that positively influences digital payment adoption. Compared with urban residents, rural residents’ innovativeness may not be salient due to their lower education and lower income (Tao et al., 2024; Xiang & Stillwell, 2023). Therefore, rural residents’ innovativeness may be of special importance. If rural residents’ innovativeness is higher, then the residents are likely to be willing to try digital payment. Thus we propose Hypothesis 8.

Perceived riskiness refers to the uncertainty or anxiety about the possible adverse consequences of an action perceived by the consumer (Mandrik & Bao, 2005). New products are usually considered inherently risky. While digital payment is essentially a new concept for rural residents and risk should be considered (Manrai et al., 2024). When digital payment is used, the payer and the payee are separated in time and space, which will make the user feel insecure. In addition, the wireless network used to transmit digital payment signals may have some defects in the process of development and is vulnerable to some insecurity factors (such as hacker attacks). As the digital payment environment is complex and different service providers may use different technologies, users may feel confused and are likely to worry about the risks caused by operational errors. Privacy breaches are often regarded as an important risk for consumers when they use online transaction payment methods (Chauhan, 2024; Park et al., 2019; Sahi et al., 2022). The negative effects of perceived riskiness on digital payment have been validated in many studies (Azman Ong et al., 2023; Liébana-Cabanillas et al., 2024; Oliveira et al., 2016; Shin, 2010; Yang et al., 2012; Slade et al., 2015). Compared with urban residents, rural residents are more sensitive to risk for (1) their lack of ways to identify and mitigate risk due to lower education and (2) their insufficient risk resilience due to lower income (Tao et al., 2024; Tian, 2005; Xiang & Stillwell, 2023). If rural residents feel that the risk of digital payment is large, their willingness to use digital payment will be reduced. Thus we propose Hypothesis 9.

Mechanisms for Digital Payment Use to Influence Credit Availability

From the perspective of financial service suppliers, digital payment enables customers to get more opportunities to use financial services, which in turn promotes suppliers to obtain more deposits and thus increase their lending capacity (Abdulai et al., 2024; Ndung’u, 2018). Digital payment technology allows suppliers to acquire the digital traces of users, which helps them to understand customer information better, thereby reducing the transaction costs of supplying credit products and reducing interest rates (Patra & Sethi, 2024; Schmidt & Cohen, 2013), and meanwhile reducing financial institutions’ requirements for mortgage guarantees (Ndung’u, 2018; Sheng, 2021). From the perspective of the financial service users, digital payment promotes users to establish contact with financial institutions, which greatly helps users learn more about financial products and can also reduce the transaction costs for customers to obtain financial services, thereby increasing their willingness to use financial products, such as applying for a loan (Barroso & Laborda, 2022; Mas, 2015). While the greatest obstacles for rural residents to obtain credit are the high interest, unawareness of what credit is available, and the lack of collateral (Zheng, 2019), digital payment can play an active role in overcoming these three obstacles. A few recent empirical studies have also confirmed that digital payments can enhance access to credit (Abdulai et al., 2024; Patra & Sethi, 2024). Therefore digital payment can help alleviate financing difficulties or improve credit availability for rural residents. Thus we propose Hypothesis 10.

Mechanism for Innovativeness to Influence Credit Availability

One manifestation of innovativeness is that people dare to try new things (Slade et al., 2015). For a long time, rural residents have mainly resorted to informal finance when they lack capital (M. Gao & Hu, 2022; Turvey et al., 2010), and as a result, formal credit, especially from online platforms, can be regarded as a new concept for most rural residents. Salvi et al. (2024) have verified that an individual’s innovativeness has a positive influence on the intention to finance using equity crowdfunding. Therefore, rural residents with higher innovativeness may be more likely to try to use formal credit. Thus, we propose Hypothesis 11.

In summary, using the existing constructs from UTAUT2 (performance expectancy, effort expectancy, social influence, facilitating conditions, price value), two other constructs (innovativeness, perceived riskiness) and the credit availability as the outcome variable of digital payment use following digital payment intention, we form the unified model of digital payment use and credit availability, shown in Figure 2. The dotted lines show the parts from the UTAUT2 model, where performance expectancy, effort expectancy, social influence, facilitating conditions and price value influence digital payment use behavior through use intention, and facilitating conditions also influence digital payment use behavior directly. The solid lines are the extension based on UTAUT2, which shows that two constructs (innovativeness and perceived riskiness) and one outcome variable (credit availability) are added. Specifically, performance expectancy, effort expectancy, social influence, facilitating conditions, price value, innovativeness, and perceived riskiness influence credit availability indirectly through digital payment intention and use, and innovativeness influences credit availability directly. Using this extended UTAUT2, we aim to find ways to improve credit availability for rural residents from the perspective of digital payment usage.

OPEN IN VIEWER

Figure 2.

Basic unified model of digital payment use and credit availability.

Questionnaire

A survey was designed using constructs and scales from the literature to test the theoretical constructs and conducted in China’s rural areas. Our target population was rural residents over the age of 18. Specifically, the scales of performance expectancy, effort expectancy, social influence, facilitating conditions, and price value are adapted from Chen and Tang (2006), Venkatesh et al. (2012), and Oliveira et al. (2016). Facilitating conditions are deleted due to collinearity with performance expectancy. Therefore, we do not report the scale for facilitating conditions here. Perceived riskiness is adapted from Lu et al. (2011), Yang et al. (2012), and Slade et al. (2015). Innovativeness and digital payment intention are adapted from Oliveira et al. (2016). Digital payment use is adapted from Venkatesh et al. (2003), Chen and Tang (2006), and Venkatesh et al. (2012). Following Black and Strahan (2002) and Mi et al. (2018), credit availability is the number of times a rural resident obtained credit in the second half of 2018. All items except those for digital payment use are measured using a five-point Likert scale, with the anchors being 1 = “strongly disagree” and 5 = “strongly agree”. Digital payment use is measured using the quantity of the categories of digital payment and the frequency of digital payment per month.

The questionnaire was originally created in Chinese and used for the field investigation. Finally, it was translated into English by language professionals. The contents of the questionnaire are shown in Appendix 1. The questionnaire was pilot-tested in January 2019 with a sample of 30 subjects from two villages in Shandong Province: Zhangjia Village, Gaoguanzhai Township, Zhangqiu District in Jinan, and Chenzhuang Village, Xiaoli Township, Jiaxiang County in Jining. The results show that the scales are reliable and valid. The data from the pilot test was not included in the formal investigation.

Mobile payment is the main form of digital payment in China’s rural areas, according to the People’s Bank of China (PBC, 2019, 2022). Therefore, digital payment in this questionnaire mainly refers to mobile payment. Mobile payment is used in various domains (e.g., the purchase of commodities and services, investments, the payment of salary and dividends, and so on) through mobile communication equipment, such as mobile phones and laptops.

To avoid the possible reverse causality that credit availability may affect digital payment use, we measure these two variables using the different time periods in the questionnaire. For credit availability, we use the second half of 2018 as the time period, while for digital payment use behavior, we use the period from the first start of digital payment use behavior to the end of 2018. As the first start may be much earlier than mid-2018 and payment habits tend to exhibit a high level of persistence (Liu et al., 2019), the period of digital payment use behavior can be roughly treated as earlier than that of credit availability and the potential reverse causality can be solved to a large extend.

Samples and Data Collection

The sample area in this study is located in Shandong Province of China. According to the China National Bureau of Statistics, Shandong Province has 38.54 million rural residents in 2019, accounting for 38.13% of the total province population, which is similar to the proportion of the whole China’s rural population in 2019 of 37.29%. Although the rural population is decreasing with the development of urbanization, there are still 36.67 million rural population in 2021 in Shandong Province, accounting for 36.05% of the total province population, which is also similar to that proportion of the whole China’s rural population in 2021 of 35.28%. Shandong Province is a large agricultural province. Its gross agricultural output value has ranked first in China for many years, and its agricultural industry is also highly diversified. Thus, the study of the behavior of rural residents in Shandong Province has a certain degree of representativeness.

This study adopts multistage sampling, which is a probability sampling method. The main advantage of this method is its ability to effectively handle large-scale and widely distributed populations, ensuring the representativeness and operability of the sample while saving costs and time (Feng, 2021). In February 2019, we carried out the formal investigation after that trial one in January 2019, mentioned in the above “questionnaire” part. Among the 17 prefecture-level cities in Shandong Province, two counties (cities or districts) were randomly selected in each prefecture-level city, 1 to 2 natural villages were randomly selected in each county (city or district), and 15 rural residents (in different households) were selected from each sample village. During the conducting of the survey in each village, the enumerators were instructed to consult the village head to understand the age distribution of the adults in the village. Based on this age distribution, the approximate number of individuals to be surveyed in each age group (four age groups shown in Table 4, a total of 15 persons) was determined, and random sampling was then conducted within each age group. This approach ensures coverage of all demographic groups and helps mitigate the issue of overrepresentation to some extent. A total of 525 questionnaires were distributed, and 518 questionnaires were returned. Due to the missing sensitive information such as income and the misunderstanding of reverse questions in some questionnaires, a total of 470 valid questionnaires were finally obtained, with an effective response rate of 90.73%.

OPEN IN VIEWER

Table 4.

Sample Characteristics.

Variable	Variable assignment	Frequency	Proportion	Variable	Variable assignment	Frequency	Proportion
Gender	Male = 1	208	44.26%	Number of types of digital payment use	Never = 1	57	12.13%
	Female = 2	262	55.74%		1 type = 2	386	82.13%
Age	18–25 = 1	84	17.87%		2 types = 3	18	3.83%
	26–35 = 2	122	25.96%		3 types = 4	9	1.91%
	36–45 = 3	154	32.77%		4 types = 5	0	0
	46 and above = 4	110	23.40%	Average monthly use of digital payments	Never = 1	57	12.13%
Education	Primary school and below = 1	81	17.23%		5 and below = 2	93	19.79%
	Middle school = 2	200	42.55%		6–10 = 3	77	16.38%
	High school = 3	111	23.62%		10–15 = 4	68	14.47%
	University (college) and above = 4	78	16.60%		16 and above = 5	175	37.23%
Income (RMB)	Less than 10,000 = 1	80	17.02%	Number of loans (second half of 2018)	0 loan = 1	396	84.26%
	10,000–50,000 = 2	208	44.26%		1 loan = 2	45	9.57%
	50,000–100,000 = 3	145	30.85%		2 loans = 3	9	1.91%
	100,000+ = 4	37	7.87%		3 loans and more = 4	20	4.26%

Note. Types of digital payment use include (1) purchase of commodities or services, (2) investment, (3) salary or dividend payment to others, and (4) other kinds of payment, such as alimony to children or parents. Income is the average annual income over the past 3 years.

Data

The sample characteristics are shown in Table 4. From the view of individual characteristics, 55.74% of the respondents were female. 25.96% were aged 26 to 35 years old and 32.77% were 36 to 45 years. 42.55% had an education level of middle school and 40.21% had an education level above high school. About 44.26% of the sample had an annual income of RMB 10,000 to 50,000 in the past 3 years, and 30.85% had an income of RMB 50,000 to 100,000 in the past 3 years. Only 12.13% of the respondents reported never using digital payment; therefore, 87.87% (100% minus 12.13%) had used digital payment. About half (51.70% = 14.47% + 37.23%) of the respondents used digital payment more than 10 times per month, and only 37.23% used it more than 16 times monthly. From the perspective of the aim of digital payment use behavior, 82.13% of the respondents report they use one type of digital payment, and only 5.74% (3.83% + 1.91%) use two or more types. From the perspective of credit availability characteristics, in the second half of 2018, 15.74% of the sample had loans, of which 4.26% had loans three times or more. The sample characteristics show that most rural residents have gradually accepted and used digital payment with the improvement of education and income in rural areas. The popularization of digital payment in rural areas may greatly reduce the cost of rural residents’ access to financial resources, help financial institutions obtain rural residents’ information, and improve credit availability to rural residents. It should be noted that, however, the proportion of rural residents who use digital payment for multiple purposes and with high frequency is still low.

Data Analysis Technique

Structural equation models (SEM) combines the methods of confirmation factor analysis and path analysis to overcome the limitations of traditional multivariate data analysis techniques (Hair et al., 2021). Confirmation factor analysis (measurement model) can be used to estimate the latent variables. And path analysis (structure model) can be used to analyze the multiple and complex causal relationships (Fan et al., 2016). The first advantage of SEM is that it can accommodate latent variables corresponding to hypothetical constructs to reflect a continuum not directly observable and cope with multiple and complex causal relationships. Another advantage is that it can estimate various relationships simultaneously and provide an overall fit and make, providing a summary evaluation of multiple equations, while models similar to SEM, such as multiple linear regression, can only test each equation individually (Kline, 2015; Tomarken & Waller, 2005). This paper discusses the attitude and behavior of digital payments adoption, credit availability, and various influencing factors, which inevitably involve some unobservable variables and complex causal relationships. Given the advantages of SEM, this paper chooses SEM to illustrate the model.

Measurement Model

In this study, confirmatory factor analysis (CFA) was conducted using 8 latent variables and 28 measurement items. The number of loans in the second half of 2018 was used as the proxy of the variable of credit availability (CA), so CFA did not include credit availability (CA), and the results are shown in Table 5. According to Fornell and Larcker (1981), items that have a loading less than 0.7 should be removed from the model, and eventually a total of four items were deleted: PE4—Using mobile payment would enable me to conduct tasks more conveniently, EE4—It would be easy for me to become skillful at using mobile payment, INN3—In general, I am hesitant to try out new information technologies, and PR4—I feel secure sending sensitive information across using mobile payment systems. The modified model fits well (χ = 730.777, p < .001, χ²/df = 2.320; CFI = 0.963; TLI = 0.956; RMSEA = 0.053; SRMR = 0.045).

OPEN IN VIEWER

Table 5.

Constructs, Items, and Factor Loadings.

Construct	Item	STD. estimate	SE	p
Performance expectancy (PE)	PE1	0.866	0.014	***
	PE2	0.905	0.011	***
	PE3	0.841	0.016	***
	PE5	0.833	0.016	***
Effort expectancy (EE)	EE1	0.926	0.008	***
	EE2	0.946	0.007	***
	EE3	0.940	0.007	***
Price value (PV)	PV1	0.870	0.013	***
	PV2	0.908	0.011	***
	PV3	0.928	0.010	***
Social influence (SI)	SI1	0.885	0.014	***
	SI2	0.923	0.012	***
	SI3	0.837	0.017	***
Perceived riskiness (PR)	PR1	0.806	0.021	***
	PR2	0.871	0.018	***
	PR3	0.841	0.019	***
Innovativeness (INN)	INN1	0.886	0.014	***
	INN2	0.831	0.017	***
	INN4	0.888	0.014	***
Digital payment intention (DPI)	DPI1	0.944	0.011	***
	DPI2	0.878	0.014	***
	DPI3	0.752	0.022	***
Digital payment use (DPU)	PAYQ	0.683	0.032	***
	PAYF	0.868	0.029	***

Note. Perceived Riskiness (PR) is a reversed item. Facilitating conditions are deleted due to collinearity with performance expectancy.

***

p < .01.

This study evaluates the construct reliability using composite reliability (CR). As shown in Table 6, the CR values for all the constructs range from 0.755 to 0.956 and are higher than the threshold value of 0.7 (Fornell & Larcker, 1981), which suggests that the constructs are reliable. The average variance extracted (AVE) and the factor loadings of the items are used to test the convergent validity. The AVE values of the latent variables in Table 6 range from 0.610 to 0.879, all higher than the threshold value of 0.5 (Fornell & Larcker, 1981). Moreover, except for PAYQ (the quantity of the categories of digital payment), the factor loadings of the items are higher than 0.7, and the p value shows that these results are significant at the level of .001. Thus we can conclude that the measurement model has a good convergent validity.

OPEN IN VIEWER

Table 6.

Composite Reliability, Convergent Validity, and Discriminant Validity of the Constructs.

Construct	CR	AVE	PE	EE	PV	SI	PR	INN	DPI	DPU
PE	0.920	0.743	0.862
EE	0.956	0.879	0.774	0.938
PV	0.929	0.814	0.497	0.495	0.902
SI	0.913	0.779	0.471	0.425	0.385	0.883
PR	0.878	0.705	−0.043	0.071	0.016	−0.034	0.840
INN	0.902	0.755	0.488	0.591	0.406	0.396	0.079	0.869
DPI	0.896	0.743	0.544	0.552	0.548	0.414	0.021	0.480	0.862
DPU	0.755	0.610	0.579	0.653	0.498	0.315	0.135	0.581	0.578	0.781

Note. The bold numbers on the diagonal are the square root of AVE, and the lower triangle is the Pearson correlation coefficient matrix of the constructs.

Discriminant validity of the constructs is evaluated using Fornell-Larcker criteria, which states that the square root of AVE should be greater than all correlations between each pair of constructs. As shown in Table 6, there is no case where the correlation coefficient of the two constructs is greater than the square root of AVE, so the scales have good discriminant validity.

We examined the potential common method bias (CMB) by using three statistical analysis methods. First, Harman’s single factor test was used to examine the extent of CMB. The unrotated principal components factor analysis indicated that 39.34% of the total variance was explained by one single factor, which is less than the 50%, suggesting that CMB was not a critical issue (Podsakoff et al., 2003). Second, CFA single factor method test was conducted, and the results showed that the model fitness was not up to the standard (CFI = 0.507, TLI = 0.460, RMSEA = 0.196, SRMR = 0.124). Third, the Unmeasured Latent Method Construct (ULMC) method by including a common method factor was conducted, and the results showed that the model fitness became worse (ΔCFI =is −0.063, ΔTLI =is −0.071, ΔRMSEA =is 0.037, ΔSRMR =is 0.228), which means that the model fitness became worse. Thus common method bias can be regarded as nonexistent. In addition, the presence of multicollinearity is estimated through the variance inflation factor (VIF). All the VIF values are below 3.3 (Cenfetelli & Bassellier, 2009), indicating that the multicollinearity problem is nonexistent.

Hypothesis Testing

The results of the structural model are shown in Table 7 and Figure 3. The model fitness is robust (χ² = 640.855, p < .001, χ²/df = 2.543; CFI = 0.958; TLI = 0.950; RMSEA = 0.057; SRMR = 0.059). The results state that PE (β = .162, SE = 0.069, p < .05), EE (β = .160, SE = 0.071, p < .05), PV (β = .293, SE = 0.045, p < .01), SI (β = 0.096, SE = 0.046, p < .05) and INN (β = .173, SE = 0.051, p < .01) have positive impacts on digital payment use intention (DPI). Hence, hypotheses 1, hypotheses2, hypotheses 3, hypotheses 6, and hypotheses 8 are supported. Among these constructs, PV has the most prominent positive influence, followed by INN. The impact of perceived riskiness (PR) on the digital payment intention (DPI) is not significant, so hypothesize 9 is rejected. Our model also tests the influence of digital payment intention (DPI) on digital payment use (DPU), which is significantly positive (β = .627, SE = 0.04, p < .01), and hypothesis 10 is supported. The influence of digital payment use (DPU) on credit availability (CA) is shown to be statistically positive (β = .120, SE = 0.062, p < .05), thereby supporting hypothesize 10. Lastly, innovativeness (INN) is found to have a positive impact on credit availability (CA; β = .121, SE = 0.058, p < .05), and then Hypothesis 11 is supported.

OPEN IN VIEWER

Table 7.

Estimation Results for the Structural Model.

DV	IV	STD. estimate	SE
DPI	PE	0.162**	0.069
	EE	0.160**	0.071
	PV	0.293***	0.045
	SI	0.096**	0.046
	PR	0.012	0.04
	INN	0.173***	0.051
DPU	DPI	0.627***	0.04
CA	DPU	0.120**	0.062
	INN	0.121**	0.058

Note. ns = insignificant., ***representsp < .01, **representsp < .05, *representsp < .1.

OPEN IN VIEWER

Figure 3.

Estimation results for the structural model in graph.

Theoretical Contributions

Previous research on the use of digital payment seldom considers the outcome. However, to consider the outcome will make the use of digital payment more meaningful. Besides, previous related literature pays little attention to rural areas. In rural areas with relatively weak financial infrastructure, digital payment that rely less on traditional financial branches will be of greater significance to rural residents. Considering these two shortcomings and combining them with the expectation that digital payment may improve rural residents’ credit availability, this paper introduces credit availability as an outcome variable in the digital payment adoption model UTAUT2, with rural residents as the objects of observation. Correspondingly, the first contribution of the study brings a relatively new direction in extending the digital payment adoption model by adding outcome variables. The second contribution is that it expands the applications of the UTAUT2 model in the context of rural residents. The third contribution is that the study’s results broaden the ways to improve rural financial services, including payment and credit. The last contribution is that this study may give other countries some inspiration on the aspects of rural residents’ digital payment adoption and credit availability.

Practical Implications

The results of this study show that rural residents’ digital payment use significantly promotes credit availability. Performance expectancy, effort expectancy, social influence, price value, and innovativeness all significantly and positively affect digital payment intention and behavior, thereby indirectly improving credit availability. In addition to indirectly affecting credit availability through digital payment intention and behavior, innovativeness has a positive and significant impact on credit availability directly. Since rural residents’ digital payment use can promote credit availability, it is necessary to encourage rural residents to use digital payment more, that is more often and in more situations. At the same time, financial institutions need to better use the information obtained from the digital traces left by digital payment.

Several approaches can be used to encourage rural residents to use digital payment more. First, continue to strengthen the construction of rural digital infrastructure. It is necessary to improve the breadth and depth of network coverage to ensure that rural areas have access to high-speed and smooth network services. Besides, it needs to accelerate the expansion of 5G networks into rural areas and enhance the level of mobile internet applications. Second, strengthen cross-industry integration and encourage deep collaboration between digital payment, finance, technology, business, and other sectors to expand more new application scenarios to improve the performance expectancy of digital payment. Third, continue to promote the simplification of digital payment, such as increasing the use of near-field transactions, to reduce the effort expectancy of rural residents, which is very important for rural residents, especially the older. At the same time, continue to enhance rural residents’ digital skills and literacy through education or training, especially refined and detailed guidance, to reduce the difficulty of using digital payments, thereby lowering their effort expectancy. Fourth, continue to reduce the cost of digital payment, both the price of hardware like mobile phones and the network use fee, to improve the price value. Meanwhile, continue to raise rural residents’ income to decrease their price sensitivity. Fifth, increase the publicity about digital payment in rural areas through various channels to enhance the overall understanding of digital payment among rural residents, thus expanding digital payment’s social influence. Sixth, promoting the concept of mass innovation to enter rural areas will be conducive to fostering rural residents’ individual innovativeness, which thus helps them try digital payment and various types of credit. Lastly, continue improving protective measures, such as insurance, rights protection, remedial actions, etc., in case of risks or losses.

As digital payment can increase credit availability, three ways are recommended to ensure financial service entities better use the traces left by digital payment. First, financial service entities are encouraged to enrich the digital payment scenarios, which is consistent with the suggestions about encouraging rural residents to use digital payment. Digital payment is the entry to various financial services, and in the context of the digital economy, digital payment has become a battleground for financial institutions. Having more use scenarios means financial service entities may know more about rural residents’ credit status, income status, and capital use status. Second, financial institutions must conduct efficient data mining after obtaining the traces left by various digital payments. Financial institutions need to increase investments in data mining, such as fintech and big data algorithms. Third, competition should be encouraged in the field of digital payment. Monopolization of this market would be detrimental to the diversification of application scenarios and, more importantly, would lead to higher interest rates on credit from these platforms, which would counter the “affordability” of financial inclusion.

These suggestions are also applicable to other countries to some extent. Countries with underdeveloped digital infrastructure or lacking mature digital service providers may seek international support and collaboration in cases where they don’t have the domestic capacity to foster these services.

Limitation and Scope of the Future Research

This study has some limitations and thus provides impetus for future improvements. Firstly, this study has not yet considered the moderating effects of demographic characteristics of individuals, such as gender, age, education level, and other factors, which might provide more detailed results that need to be further enriched and improved in future research. Secondly, for the factors that affect both digital payment adoption and credit availability, only one factor, innovativeness, is considered in this study. Still, there are many other factors, such as income and wealth. More such factors can be added to future research to increase the model’s explanatory power. Thirdly, for the outcome variable, this study only considers credit availability. However, digital payment can bring different outcomes, such as consumption and transaction convenience, which can be further estimated in future studies. Lastly, this study mainly focuses on rural residents in Shandong Province of China. Although the behavior of rural residents in Shandong Province can be somewhat representative of the whole country due to the similarity of the proportion of the rural population in Shandong Province to the whole country, the fact that its gross agricultural output value is the highest in the entire country, and the high diversity of its agricultural industry, the study cannot be fully representative of the whole country because of the fact that there are always differences among regions. Therefore, future studies can be based on rural residents across the whole of China to draw more general conclusions.