The Impact of Fintech on Economic Growth: Evidence from China

: Financial technology (ﬁntech) has seen fast development recently in China; however, studies exploring the contributions of ﬁntech to China’s economic growth remain limited. Thus, this study motivated by the knowledge gaps and fast expansion of ﬁntech examined: (i) the impact of ﬁntech and the submeasures of third-party payment, credit, and insurance on China’s economic growth; (ii) the regional and provincial impact of ﬁntech on China’s economic growth; (iii) the causality relationships between ﬁntech and economic growth. By using a sample of 31 provinces in China and the instrumental variable generalized method of moments (IV–GMM) technique, the study established the following: (i) ﬁntech and the submeasures of third-party payment, credit, and insurance have a statistically signiﬁcant positive effect on China’s economic growth. Speciﬁcally, a 10% rise in ﬁntech, third-party payment, credit, and insurance raises China’s economic growth by 8%, 4%, 5%, and 16%, respectively; (ii) the eastern region has the highest growth effect of ﬁntech. Moreover, Zhejiang province has the highest growth effect of ﬁntech at the provincial level; (iii) a unidirectional causality exists from third-party payment and credit to economic growth, and economic growth to insurance; a bidirectional causality exists between ﬁntech and economic growth. This article explicitly suggests substantial institutional reforms to promote the healthy development of ﬁntech in China.


Introduction
Fintech, otherwise called internet finance or digital financial inclusion, simply refers to an amalgamation of finance and information technology. It constitutes payment and settlement, risk management, networking channels, and resource allocation functions [1][2][3]. Fintech has expanded dramatically in the financial industry thanks to the rapid expansion of the Internet, information technology, mobile phones, and digital technologies. According to [4], global fintech investment stood at USD 55.3 billion in 2019. In the case of China, the Wind database reports an upsurge in the number of peer-to-peer (P2P) loan companies from 214 in 2011 to 1544 in the year 2014. The database also shows over RMB 252.8 billion outstanding loans from P2P. Again, third-party payment market size stood at RMB 190.5 trillion in 2018 and nonequity crowdfunding platforms raised RMB 2.9 billion in 2016. Finally, Internet insurance premium income increased from RMB 320,000 in 2011 to RMB 19.05 million in 2018, which represented an almost 5853% increase. Given these developments, these questions remain: (1) Does fintech contribute to China's economic growth? (2) Does the measure of fintech matter? (3) Do regional and provincial heterogeneity matter? (4) Does fintech cause China's economic growth?
Fintech eliminates the need for banks to act as a middleman between borrowers and lenders [5]. P2P lending involves leveraging the internet to link lenders and borrowers, with lenders assessing credit risk based on the information provided by borrowers and then making lending decisions [6]. Fintech increases the chances of small and mediumsized enterprises (SMEs) obtaining loans at a reasonable interest rate [7]. Furthermore, in the initial phases of fintech development; nevertheless, banks' risks and management costs increase in the growth phase of fintech. The authors of [22] estimated the effect of fintech on banks' liquidity, profitability, growth, and security in China from 2006 to 2018 and found that fintech spurs banks' growth, profitability, and security; however, it impedes banks' liquidity. Ref. [23] examined the impact of third-party payment on commercial banks' in China from 2007 to 2014 and found that third-party payment encouraged the expansion of the financial sector in China and increased the process of the industrial revolution.
Ref. [24] investigated the impact of fintech on household income in China from 2010 to 2014 and discovered that fintech raises household income, with the impact being greater for rural families. They also discovered that the wealthiest benefit less from fintech growth than the disadvantaged. Ref. [25] investigated the impact of fintech on rural farmers' vulnerability to poverty in China and discovered that fintech improves rural farmers' vulnerability to poverty. Ref. [3] also established that fintech reduces poverty in China. The authors of [26] found that fintech improves urbanization in China while using data from 2010 to 2016. Fintech is also documented to improve household consumption [27,28], entrepreneurship [29], and household welfare [30].
Ref. [31] studied the effect of fintech on China's economic growth and found no relationship between fintech and economic growth using the vector autoregressive (VAR) technique. Contrarily, ref. [32] appraised the linear and nonlinear impact of P2P lending on sustainable economic growth in China and found a U-shaped relationship between P2P lending and sustainable economic growth. Recently, ref. [33][34][35][36] also established a positive association between fintech and economic growth in China.
From the literature review, it was found that most of the studies on fintech have focused on financial development, while few studies have explored the impact of fintech on economic growth. Moreover, the few studies on fintech and economic growth have all focused on aggregate economic growth with no study focusing on labor productivity. Furthermore, the studies on fintech and economic growth have focused on aggregate fintech, with no study exploring the contributions of the submeasures of third-party payment, credit, and Internet insurance to economic growth. Finally, most of the studies on fintech and economic growth ignored the causal linkages between fintech and economic growth. Thus, this study is conducted to bridge these knowledge gaps.

Hypothesis Development
The primary goal of the financial system is to move cash from surplus to deficit units [37]. This fosters economic growth by making saving and investing easier [37]. The financial system also connects borrowers and lenders, reducing the risks connected with money transfers [37]. As a result, the financial system minimizes transaction costs and information asymmetry [38]. Transaction costs include the time and money spent to connect borrowers and lenders, as well as legal fees. The uneven distribution of information between lenders and borrowers is referred to as information asymmetry. Financial institutions also offer oversight and corporate governance, which contribute to economic growth [37]. Nevertheless, the traditional financial system is frequently plagued by higher costs and risk taking, which results in financial crises [16]. The growth of fintech may effectively cut transaction costs and alleviate the problems of information asymmetry because additional complicated jobs such as customer testing, risk assessment, and information sharing are made much easier by fintech. This promotes entrepreneurship [29,39,40], trade [1,41,42], household consumption [27], financial development [23,43,44], financial inclusion [24,45], and household income [24], with these proxies widely documented to contribute to economic growth [38,46,47], thus, we hypothesize: Hypothesis 1 (H1). The impact of fintech on China's economic growth is positive.
Fintech could also have a negative effect on economic growth. As fintech expands, funds move from the real economy to fintech platforms [31]. This might reduce banks' Sustainability 2022, 14, 6211 4 of 17 capital via financial hoarding, translating into a reduction in domestic credit supply, an increase in financial crisis, and hinder economic growth through the reduction in investment and production [31]. For instance, [20] found that fintech decreases the positive association between bank deposit growth and capitalization. Ref. [22] also found that fintech impedes banks' liquidity. Thus, we hypothesize: Hypothesis 2 (H2). The impact of fintech on China's economic growth is negative.
The impact of fintech on economic growth could be heterogeneous among different regions and provinces in China. For instance, the eastern and central regions have advanced fintech development as compared with the western region [48]. This might boost the eastern and central regions' economic growth via increased employment and investment, while the western region might obtain a lower impact of fintech on economic growth. Moreover, the growth effect of fintech depends on several factors, such as financial development, investment, and employment. These factors are more developed in the eastern and central regions as compared with the western region. These factors help to extend the spillover effect of fintech on economic growth. Thus, the eastern and central regions with high levels of financial development, investment, and employment might obtain higher economic growth as compared with the western region with relatively lower financial development, investment, and employment. Thus, the following hypothesis is obtained: Hypothesis 3 (H3). The impact of fintech on economic growth is heterogeneous across different regions and provinces. Third-party payment denotes payment services offered on an unbiased payment platform other than banks that are connected to the bank payment and settlement systems of e-commerce firms and commercial banks [23] Index Credit credit

Data
In this study, credit is defined as crowdfunding. Crowdfunding is "a collective effort by consumers who network and pool their money together, usually via the Internet, to invest in and support efforts initiated by other people or organizations" [50] Index

Internet insurance insu
Internet insurance is the marriage of insurance services (brokerage and underwriting) and technology, and it is often termed as InsurTech [51] Index Labor force l Total employment Yuan Gross fixed capital formation k Investment Yuan Table 2 presents the descriptive statistics of the variables. We observe that China's economic growth has an average rate of 10.74% with a maximum of 13.41%. Investment has an average of 9.13% with a maximum of 10.56%. Similarly, labor has an average of 6.01% with a maximum of 7.59%. Fintech has an average of 4.973% with a maximum of 5.819%. With regards to the submeasures of fintech, insurance has the highest average (5.664%), followed by third-party payment (4.805%), while credit has the lowest average (4.547%).  Table 2 also shows that the Livin-Lin-Chu (LLC) (LLC tests the null hypothesis of all panels holding unit roots against the alternative hypothesis that some panels are stationary [13]) [52] unit-roots test, at the log levels and first difference, demonstrated the variables are stationary. Figure 1 shows the correlation relationship between third-party payment, credit, insurance, fintech, and economic growth. The results show that there exists a positive correlation between third-party payment, credit, insurance, fintech, and economic growth.

Methodology
We build on the neoclassical aggregate production function (APF) following [53]. This theory examines the relationship between inputs and outputs in production [13,54].
The model is stated as Equation (1) ) , ( where it Y denotes output, it k denotes capital, and it l denotes labor. it A denotes total factor productivity (TFP). i denotes province, and t denotes time. To examine fintech's influence on economic growth, we substituted fintech into it Taking the log of either end gives Equation (5) From Equation (5), we estimated the following growth equation:

Methodology
We build on the neoclassical aggregate production function (APF) following [53]. This theory examines the relationship between inputs and outputs in production [13,54].
The model is stated as Equation (1) where Y it denotes output, k it denotes capital, and l it denotes labor. A it denotes total factor productivity (TFP). i denotes province, and t denotes time. To examine fintech's influence on economic growth, we substituted fintech into A it (see Equation (2)) Substituting Equation (2) into Equation (1) gives Equation (3) Expressing Equation (3) in econometrics gives Equation (4) Taking the log of either end gives Equation (5) From Equation (5), we estimated the following growth equation: Equation (6) ln where ln gdp it denotes the natural logarithm of economic growth, ln k it signifies natural logarithm of investment, ln l it signifies natural logarithm of labor while ln i f it denotes the natural logarithm of fintech, which is further disaggregated into the natural logarithm of third-party payment (ln tpp it ), the natural logarithm of credit (ln credit it ), and the natural logarithm of insurance (ln insu it ). The parameter b 3 represents the long-run elasticity estimates of fintech (third-party payment ln tpp it , credit ln credit it , and insurance ln insu it ). ε it signifies the error term of each observation. The goal of this study is to compute b 3 . We anticipate b 3 to have either a positive or negative significant impact on China's economic growth. We used the IV-GMM technique for our empirical analysis following [13,18]. This model produces efficient outcomes since it addresses endogeneity and omitted variables problems [18]. Following [13,18], we employed the robust option. This option shows the Hansen test [55], which tests the null hypothesis of instrument overidentification; restriction should not be rejected in both models [55]. To solve endogeneity issues, we took the first and second lags of the fintech proxies as instruments.
For our robustness analysis, we used the ordinary least-squares estimator (OLS) and the random effects (RE) technique.

Causality Model
We used the Dumitrescu and Hurlin [56] causality test to examine the causal nexus between fintech, economic growth, and the other macroeconomic variables in China. The test is specified as follows: Equation (7) We tested the following assumption: H 0 : γ i = 0, ∀i = 1, . . . , n (null hypothesis) against the alternative: are parameters. m is the length of lags. n is the number of provinces. Y it denotes the dependent variables while X it denotes the independent variables. We tested the null hypothesis of no causal relationship against the alternative of a causal relationship among the variables. Table 3 shows the full sample results. The results show that third-party payment, credit, insurance, and fintech have a positive and statistically significant effect on China's economic growth in both models at the 1% level. The IV-GMM results show that a 1% rise in third-party payment increases China's economic growth by 0.39%. This implies that a 10% rise in third-party payment increases China's economic growth by approximately 4%. The findings also show that a 1% rise in credit increases China's economic growth by 0.50%. This implies that a 10% rise in credit increases China's economic growth by 5%. Third-party payment platforms and credit platforms propel general economic efficiency by reducing transaction costs and information asymmetry [8]. This generates liquidity, triggers savings, and advances capital accumulation, leading to increased production, trade, consumption, and reproduction purposes [31], thus a positive contribution to China's economic growth. Our results also show that a 1% rise in insurance consumption increases China's economic growth by 1.59%. This implies that a 10% rise in insurance consumption raises China's economic growth by 16%. Internet insurance, just like traditional insurance sectors, provides risk management and financial intermediary services to businesses and individuals. This offers protection to firms and relieves pressure to cover damages, facilitates commercial transactions, and provides credit by mitigating losses; it promotes entrepreneurial attitude, investment, boosts technological innovation, ensures efficient capital allocation, and enables risk-averse individuals and entrepreneurs to undertake higher return activities, thus a positive contribution to China's economic growth. A rise in fintech increases China's economic growth by 0.79%. This implies that a 10% rise in fintech increases China's economic growth by 8%. The significant positive effect of third-party payment, credit, insurance, and fintech on China's economic growth supports our H1. This conflicts with the findings of [31], which found no relationship between fintech and economic growth. These differences in results could be attributed to differences in datasets. These results are robust using the OLS and RE models; however, the coefficients of the IV-GMM model are slightly higher than that of the OLS and RE models, indicating that the IV-GMM model addresses endogeneity and omitted-variable issues. Notes: ( ) = Robust standard errors, J = Hansen J-statistics, JP = Hansen J-statistics p-value, F = F-statistics, RMSE = roots mean square error, VIF = variance inflation factor. ** p < 0.05, and *** p < 0.01.

Full Sample Results
Investment significantly increases economic growth according to the RE models and the IV-GMM model (11). An increase in investment increases China's economic growth by 0.3%. This supports the findings of [57] for 5 OPEC countries, [58] for 40 SSA countries, and [49] for 54 African countries, all of which found that investment increases economic growth. Labor has an insignificant effect on China's economic growth in both models.
Turning to the stability of the model, the null hypothesis of instrument overidentification restriction was not rejected in most of the cases, except in Model 12; however, the Cragg-Donald Wald F test statistic showed that the instruments used are not weak. The variance inflation factors (VIF) of the OLS estimates are also below 10, demonstrating that our study does not suffer from multicollinearity issues. Table 4 shows the regional effect of fintech and the submeasures on economic growth using the IV-GMM technique. The results show that third-party payment, credit, insurance, and fintech have a statistically significant positive effect on economic growth in both the eastern and central regions; however, only credit and fintech are significant drivers of economic growth in the western region. Precisely, a 1% increase in third-party payment, credit, insurance, and fintech increases economic growth by: 0.49%, 0.84%, 1.59%, and 0.94%, respectively, in the eastern region; 0.12%, 0.13%, 0.48%, and 0.19%, respectively, in the central region. Moreover, a 1% increase in credit and fintech increases economic growth by 0.17% and 0.31%, respectively, in the western region. The findings also show that the eastern region has the highest growth effect of fintech as compared with the western and central regions. This is in tandem with our H3. Notes: ( ) = robust standard errors, J = Hansen J-statistics, JP = Hansen J-statistics p-value, F = F-statistics, RMSE = roots mean square error, * p < 0.1, ** p < 0.05, and *** p < 0.01.

Regional Analysis
The results further show that investment drives economic growth in the central and western regions, while it impedes economic growth in the eastern region. Labor drives economic growth in the eastern region, while it impedes economic growth in the western and central regions.

Causality Analysis
The Dumitrescu and Hurlin [56] causality analysis was also used in this study to examine the causality relationships among the variables. Chord diagrams are used to show the findings (see Figures 2-5). The F-statistics produced by the Dumitrescu and Hurlin causality test are used to calculate a variable arc length in the chord diagram. The connection between any two variables defines causation between the variables. The results show that there exists a unidirectional causality running from third-party payment and credit to economic growth (see Figures 2 and 3). There exists a unidirectional causality running from economic growth to insurance (see Figure 4). There also exists bidirectional causality between fintech and economic growth (see Figure 5). Thus, an increase in fintech will stimulate China's economic growth and vice versa. The results further show that there exists a unidirectional causality running from investment to credit (see Figure 3), investment to insurance (see Figure 4), investment to labor (see Figure 4), and investment to economic growth (see . A unidirectional causality also exists from fintech to investment (see Figure 5). There exists bidirectional causality between investment and labor (see Figures 2,3 and 5). Finally, there exists a unidirectional causality from labor to third-party payment, credit, and insurance, while there exists bidirectional causality between labor and fintech. These imply that further improvement in human capital and infrastructure investment in China will drive fintech development and vice versa.

Further Analysis
To provide further robustness to our previous findings, we used labor productivity (Labor productivity refers to output per worker) as a different measure of economic growth. The results are presented in Table 5. In consonance with our earlier findings, third-party payment, credit, insurance, and fintech retained their statistically significant positive effect. The IV-GMM model results show that a 1% rise in third-party payment raises labor productivity by 0.08%. This implies that a 10% rise in third-party payment increases labor productivity by 0.8%. A 1% rise in credit raises labor productivity by 0.09%. This implies that a 10% rise in credit raises labor productivity by 0.9%. A 1% rise in insurance raises labor productivity by 0.32%. This implies that a 10% rise in insurance consumption increases labor productivity by 3.2%. Finally, a 1% rise in fintech raises labor productivity by 0.14%. This implies that a 10% rise in fintech raises labor productivity by 1.4%. These are original findings. The statistically significant and positive effect of third-party payment, credit, insurance, and fintech on labor productivity supports our H1. These results are robust using the OLS and RE models. third-party payment, credit, and insurance, while there exists bidirectional causality be tween labor and fintech. These imply that further improvement in human capital and in frastructure investment in China will drive fintech development and vice versa.   third-party payment, credit, and insurance, while there exists bidirectional causality between labor and fintech. These imply that further improvement in human capital and infrastructure investment in China will drive fintech development and vice versa.

Further Analysis
To provide further robustness to our previous findings, we used labor productivity (Labor productivity refers to output per worker) as a different measure of economic growth. The results are presented in Table 5. In consonance with our earlier findings third-party payment, credit, insurance, and fintech retained their statistically significant positive effect. The IV-GMM model results show that a 1% rise in third-party payment raises labor productivity by 0.08%. This implies that a 10% rise in third-party payment increases labor productivity by 0.8%. A 1% rise in credit raises labor productivity by 0.09%. This implies that a 10% rise in credit raises labor productivity by 0.9%. A 1% rise in insurance raises labor productivity by 0.32%. This implies that a 10% rise in insurance consumption increases labor productivity by 3.2%. Finally, a 1% rise in fintech raises labor productivity by 0.14%. This implies that a 10% rise in fintech raises labor productivity by

Further Analysis
To provide further robustness to our previous findings, we used labor productivity (Labor productivity refers to output per worker) as a different measure of economic growth. The results are presented in Table 5. In consonance with our earlier findings third-party payment, credit, insurance, and fintech retained their statistically significan positive effect. The IV-GMM model results show that a 1% rise in third-party paymen raises labor productivity by 0.08%. This implies that a 10% rise in third-party paymen increases labor productivity by 0.8%. A 1% rise in credit raises labor productivity by 0.09%. This implies that a 10% rise in credit raises labor productivity by 0.9%. A 1% rise in insurance raises labor productivity by 0.32%. This implies that a 10% rise in insurance consumption increases labor productivity by 3.2%. Finally, a 1% rise in fintech raises labor productivity by 0.14%. This implies that a 10% rise in fintech raises labor productivity by The variance inflation factors (VIF) of the OLS estimates are below 10, showing that our study does not suffer from multicollinearity issues. Notes: ( ) = robust standard errors, J = Hansen J-statistics, JP = Hansen J-statistics p-value, F = F-statistics, RMSE = roots mean square error, VIF = variance Inflation Factor. * p < 0.1, and *** p < 0.01.

Province-Specific Analysis
Given the provincial differences in fintech development [48], the sample mean might be dissimilar from the individual provincial mean. Thus, we used the IV-GMM results in Table 3 Figure 8 displays the province-specific effect of insurance on economic growth. The findings indicate that the impact of insurance on economic growth is positive for all the provinces. The findings further suggest that Zhejiang (10.7227), Jiangsu (10.5578), Shanghai (10.473), Tianjin (10.3458), and Beijing (10.2735) have a high growth effect of insurance as compared with Ningxia (9.7007), Shanxi (9.6307), Qinghai (9.5112), Gansu (9.0922), and Henan (8.7208), which have a relatively low growth effect.

Conclusions and Policy Recommendations
The literature on fintech has received great attention among academicians and policymakers in recent times; however, studies assessing the contributions of fintech to economic growth are rare. Accordingly, this study contributes to the limited studies on fintech and economic growth by examining: (i) the effect of fintech on China's economic

Conclusions and Policy Recommendations
The literature on fintech has received great attention among academicians and policymakers in recent times; however, studies assessing the contributions of fintech to economic growth are rare. Accordingly, this study contributes to the limited studies on fintech and economic growth by examining: (i) the effect of fintech on China's economic growth, (ii) the regional and provincial effect of fintech on China's economic growth, and (iii) the causality relationships between fintech and economic growth in China.
Using the IV-GMM technique, our findings indicated that: (i) Fintech, third-party payment, credit, and insurance are significant drivers of China's economic growth. Specifically, a 10% rise in fintech, third-party payment, credit, and insurance raises China's economic growth by 8%, 4%, 5%, and 16%, respectively. (ii) The eastern region has the highest growth effect of fintech. Moreover, Zhejiang province has the highest growth effect of fintech at the provincial level. (iii) A unidirectional causality exists from third-party payment and credit to economic growth, economic growth to insurance, and a bidirectional causality exists between fintech and economic growth.
The study's conclusions have substantial practical consequences for Chinese policymakers. The findings indicate that fintech, as well as its submeasures of third-party payment, credit, and insurance, are key determinants of China's economic growth. As a result, this study suggests that policymakers should encourage the growth of fintech in China. This article explicitly suggests substantial institutional reforms to promote the healthy development of fintech in China. Some of the reforms are: there should be restrictions on who to fund for a new project and the amount they fund because some projects fail; there should be a threshold on the amount of loans one can borrow from P2P platforms. Moreover, it was established that economic growth and labor explain fintech development; thus, this study suggests that policies that drive China's economic growth and labor will drive fintech development. Furthermore, Guizhou, Ningxia, Gansu, Qinghai, and Xizang have a relatively low growth effect of fintech; thus, further development of fintech in these provinces is essential to enhance China's economic growth.
This study has the following shortcomings that future studies can build upon: First, we focused on China, given data availability. However, fintech is a global phenomenon. Thus, future studies should employ more countries to examine the links between fintech and economic growth. Additionally, we examined the linear effect of fintech on economic growth. Thus, future studies should examine the nonlinear effect of fintech on economic growth. Moreover, we focused on aggregate economic growth; thus, future studies should explore the sectoral effect of fintech. Moreover, future studies should use change in economic growth as an alternative measure of economic growth to reassess these associations. Furthermore, given data management problems, we focused on fintech and the sumeasures of third-party payment, credit, and insurance. Thus, future studies should use alternative measures of fintech proxies such as the index of coverage breadth, credit investigation, monetary fund, investment, and the level of digitalization. Finally, due to the problem of multicollinearity, we considered fintech as the only influence factor of total factor productivity for the aggregate production function; thus, future studies should add more control variables by increasing the sample size.